article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
europium hexaboride is part of the large and heterogeneous class of materials that exhibit colossal magnetoresistance ( cmr ) .
the ferromagnetic transition in eub@xmath0 is accompanied by a dramatic change in resistivity .
there is a large body of experimental data available on the magnetic and electric properties , but a thorough understanding is lacking .
eub@xmath0 has a cubic unit cell with eu - ions at its vertices and a boron octahedron at its center .
the material is ferromagnetic and shows two magnetic transitions : at @xmath1k and at @xmath2k @xcite .
these have been associated with a spin reorientation and a ferromagnetic transition , respectively .
neutron diffraction experiments @xcite have given a magnetic moment @xmath3 .
this is exclusively due to the localized half - filled @xmath4-shell in the eu@xmath5 ions.@xcite electronic structure calculations @xcite , shubnikov - de haas and de haas - van alphen measurements @xcite show that eub@xmath0 is a semimetal .
the fermi surface consist of two ellipsoidal pockets , one electron - like and one hole - like , centered on the @xmath6 point in the brillouin zone .
the pockets contain very few carriers : hall effect measurements yield @xmath7 carriers per formula unit @xcite at low temperatures .
small dilations of the boron octahedra cause overlap of the conduction and valence bands at the @xmath6 points rendering eub@xmath0 semimetallic .
the carrier concentration decreases smoothly as temperature increases .
the electrical resistivity is metallic in the ferromagnetic regime .
it shows a sharp peak near @xmath8 . above this temperature ,
the resistivity decreases with an almost activated temperature dependence until it reaches a minimum at about 30k . at higher temperatures it increases and eventually starts to saturate at about room temperature .
the application of a magnetic field produces sharp changes in the resistivity .
close to the magnetic transition , _ negative _ magnetoresistance ( mr ) values of up to 100% have been observed @xcite .
this decrease in resistivity is accompanied by a large decrease in the ( negative ) hall coefficient@xcite and an increase in the plasma frequency @xcite .
the change in the plasma frequency is more gradual than the changes in resistivity and hall coefficient . in the ferromagnetic regime , on the other hand , the mr is large and _ positive _ : at 1.7 k resistivity changes of up to 700% have been observed in a transversal applied field of 7 t @xcite .
the mr depends quadratically on the applied field strength at low temperatures@xcite . just above t@xmath9 and up to @xmath10 ,
the existence of magnetic polarons has been proposed as the cause of the stokes shift measured with spin flip raman scattering ( sfrs)@xcite . the resistivity is activated at these temperatures .
however , this data contains a previously unremarked puzzle , in that the energy scale turns out to be considerably lower - by a factor of thirty - than expected based on reliable estimates of the exchange interaction .
wigger et al.@xcite showed how the crossover between large positive and large negative mr from well below to well above the ferromagnetic transition can be explained by the dominance of orbital scattering at @xmath11 to spin scattering at @xmath12 .
the model we shall use for the carrier transport in these regimes is similar to that of ref.@xcite and we shall thus suppress most of the details .
the key feature of the model is its multiband nature - there are two types of carrier .
in this paper we concentrate principally on the regime close to @xmath13 and analyze the evidence for the existence of magnetic polarons in europium hexaboride .
we show how the sfrs results can be explained using a multiband model , resolving the conundrum of the anomalously small energy associated with the carrier spin flip .
we model eub@xmath0 as a ferromagnetic semimetal with a low carrier density .
both electrons and holes are itinerant and are coupled to the local moments @xmath14 .
this can be described by the following general hamiltonian : @xmath15 here , the hopping parameter is roughly @xmath16 ev@xcite .
@xmath17 is the itinerant carrier spin operator and the subindices @xmath18 and @xmath19 stand for electrons and holes respectively .
@xmath20 ( @xmath21 ) is the on - site coupling between the spins of the electrons ( holes ) and the local moments .
@xmath22 is the magnetic exchange between local moments .
first of all we need to discuss what is the origin of the ferromagnetism and the order of magnitude of the magnetic couplings . ferro- and
antiferro - magnetism of the insulating eu - chalcogenides ( eux , x@xmath23 o , s , se , te ) has been explained as due to superexchange interaction between neighbor eu ions @xcite through the anion between them .
the density of carriers in the undoped chalcogenides is too low to expect any indirect rkky ( ruderman - kittel - kasuya - yosida ) interaction .
the ferromagnetic interaction arises instead from the overlap between the 4@xmath4- and 5@xmath24-orbitals at different cations .
this overlap leads to an effective exchange interaction in third order in perturbation theory@xcite .
this does not apply directly to eub@xmath0 due to the different crystalline structure , but nevertheless one expects that the superexchange coupling @xmath22 is small .
moreover , the increase of magnetic critical temperature and concomitant decrease of resistivity under high pressures @xcite has revealed that the magnetic exchange in eub@xmath0 is mainly due to the rkky interaction .
therefore , @xmath22 in eq .
[ eq : hamiltonian ] is negligible .
the rkky magnetic exchange is mediated by the itinerant carriers via their coupling with the lattice magnetic moments .
an effective heisenberg - like magnetic exchange can be written in terms of the local hund s like exchange coupling @xmath25 ( @xmath20 or @xmath21 ) @xcite @xmath26 where @xmath27 , @xmath28 is the density of carriers , and @xmath29 is the fermi energy .
@xmath30 is an oscillating function of @xmath31 but is ferromagnetic for small @xmath31 .
this is the relevant limit for europium hexaboride , as its low carrier density implies @xmath32 . to estimate the value of @xmath25 we use the mean field relation between t@xmath9 and j@xmath33 , @xmath34 , where @xmath35 is the coordination number for eu , and @xmath36 is the @xmath35-component of the local moments .
using a critical temperature @xmath37 and a parabolic approximation to the bands , @xmath38 is found , consistent with reported data for isolated eu - ions@xcite . in this estimation
we are considering that only one kind of carrier is responsible for the magnetism .
when the local exchange coupling @xmath25 is large enough , carriers can be localized by ferromagnetic clusters and form composite objects called magnetic polarons .
ferromagnetic polarons can exist in the low temperature phases of antiferromagnets but here we are interested in those formed in the paramagnetic phase .
a necessary condition for the existence of magnetic polarons is that the density of carriers is very low compared to the inverse of the correlation volume , namely @xmath39 .
when this condition is fulfilled , polarons are well - defined non - overlapping entities .
there are two kinds of magnetic polarons : free and bound .
a free magnetic polaron is a carrier localized in a ferromagnetic cluster embedded in a paramagnetic background .
a carrier that is coupled strongly to local moments via a hund s like coupling tends to align the moments that are within a bohr radius .
this causes a trapping potential that localizes the carrier .
the potential can be enhanced by random fluctuations of the magnetization that produce an alignment of local moments in the carrier s vicinity@xcite .
the carrier thus traps itself by the magnetization it causes .
it could increase the alignment of the local moments and hence decrease its energy by localizing itself in a smaller volume .
however , this would lead to an increase in kinetic energy .
the quantity that determines the stability and size of these objects is therefore @xmath40 where @xmath25 is the coupling of the carrier spin to the local moments and @xmath41 is the hopping parameter .
the ratio @xmath40 needs to be typically larger than one @xcite to guarantee stability of the free magnetic polaron . on the other hand , in bound magnetic polarons the main driving force trapping the carrier is not the local magnetic interaction but the electrostatic potential created by impurities .
the formation of the ferromagnetic cluster described above does occur .
however , it is a second order process , as the magnitude of hund s coupling is much smaller than the coulomb interaction .
mean field @xcite and monte - carlo @xcite calculations have shown that magnetic polarons can exist within a temperature window above t@xmath9 whose width depends on the ratio @xmath40 . at higher temperatures ,
magnetic fluctuations are strong enough to destroy the magnetic polarons .
below @xmath42 , the condition @xmath43 is not fulfilled and the polarons overlap . if a magnetic field is applied within the existence temperature window , the size of a polaron increases until eventually the polarons overlap and produce a ferromagnetic transition .
free and bound magnetic polarons can be differentiated by their dynamics and the resistivity they cause .
bound magnetic polarons are bound to an impurity in the system so the only way of transport is via an activated process : when the trapped carrier is `` ionized '' it is free to move until it is trapped by another impurity .
therefore they produce a resistivity @xmath44 such that @xmath45 .
in contrast , free magnetic polarons are able to move to adjacent areas when random fluctuations of the nearby spins produce an aligned region .
there is not a barrier to overcome in this process .
this transport mechanism has been called `` fluctuation - induced hopping '' @xcite and produces a metallic resistivity @xmath46 .
magnetic polarons have been largely studied in connection with eu - chalcogenides ( euo , eus , euse , eute ) @xcite and diluted magnetic semiconductors such as cd@xmath47mn@xmath48te and pb@xmath47mn@xmath48te with @xmath31 , the concentration of magnetic ions , small .
experimental evidence included photoluminescence spectra @xcite and magneto - optical experiments as spin flip raman scattering @xcite .
the raman scattering spectrum shows an inelastic peak at low frequencies ( stokes shift ) which , for the diluted magnetic semiconductors , depends as follows on polaronic properties:@xcite @xmath49 where @xmath50 is the density of magnetic ions participating in the formation of polarons @xcite , and @xmath25 is the local exchange interaction between the @xmath51 itinerant electrons and the @xmath24 electrons localized in the mn ions .
the low density of local moments makes for a small stokes shift , which - for cd@xmath47mn@xmath48se - is consistent with experiment@xcite .
sfrs measurements done in eub@xmath0 have similarly revealed a zero - field peak in scattered intensity of the order of @xmath52 @xcite at @xmath53 , just above the magnetic critical temperature .
the behavior of this peak with temperature and external magnetic field is consistent with the stability conditions theoretically established for magnetic polarons .
free magnetic polarons are not expected to be stable in eub@xmath0 as @xmath25 and @xmath41 are comparable .
moreover , we have argued above that the activated behavior of the resistivity is better explained by means of bound magnetic polarons .
eu - site vacancies would produce the binding coulomb potential for electrons .
we expect the energy of the stokes shift to be given by eq .
[ eqn : stokesshift ] but now @xmath54 as eub@xmath0 has a local moment on every eu site in the cubic lattice .
very few site vacancies are expected in this fairly clean material .
mean - field theory@xcite predicts that bound magnetic polarons are fully spin polarised so @xmath55 . using this value and the energy of the stokes shift we obtain @xmath56mev
this is far too low compared to the values reported in the literature for @xmath57 in isolated eu ions @xmath58mev @xcite .
we are therefore left with a conundrum : the peak in the light scattering intensity follows all the trends calculated for an object with magnetic origin but the energy of that peak is almost two orders of magnitude smaller than expected .
the solution to this problem lies in the fact that both polarised electrons and holes are found at the fermi energy .
electrons and holes come from different b and eu orbitals and therefore their magnetic couplings to the localized spin in the eu @xmath59 orbitals , @xmath20 and @xmath21 respectively , can be very different .
electronic structure calculations @xcite reveal that the hole pocket comes from the highest intraoctahedron b @xmath60 band . on the other hand , the electron pocket comes from bonding combinations of the cation @xmath24 orbitals pointing along the cartesian directions with some hybridization with the b atoms and some free - electron - like character on the ( 110 ) axes between the cations . in other words , the electron charge density distribution is mainly found around the eu ions while the holes are found around the b. therefore , the coupling of the electrons is expected to be much larger than that of the holes .
consistently , fig .
10 in ref.@xcite shows a much larger majority - minority spin band splitting for electrons than for holes . in conclusion , we propose that the ferromagnetic ordering is produced by the itinerant electrons coupled to the localized spin in eu with @xmath61 , while the itinerant holes , much more weakly coupled ( @xmath62 ) , account for the sfrs stokes shift .
a corollary of this identification is that there is likely a much higher energy feature in the sfrs , so far unobserved , that correspond to spin - flip of the electron state . in the following section
we will see how the existence of both electrons and holes is necessary to explain other electronic properties of eub@xmath0 .
at low temperatures the magnetoresistance is due to the presence of two types of carriers and we will call it `` orbital '' mr .
there are two effects involved .
the same physics that causes the hall effect is the most important cause of the mr .
in addition to this , there is also a small shift of the bands with applied magnetic field that causes a small change in the carrier density . in a simple metal with one type of carrier and a simple fermi surface ,
there is no mr . a current that flows perpendicular to a magnetic field
is initially deflected due to the transverse lorentz force .
this causes an electric field , the hall field @xmath64 , to build up : @xmath65 the carriers that are deflected hit the edge of the sample and accumulate .
the field that is thus built up counteracts the lorentz force . when it cancels the lorentz force , the current is undeflected .
this is well known ; it means that there is no mr in a normal metal .
the component of the current density that is parallel to the applied electric field is not affected by the magnetic field .
therefore the resistivity remains unaffected as well . in a semimetal ,
on the other hand , there are by definition two kinds of carriers .
these two kinds will almost invariably have different scattering times and different masses .
[ eqn : hall_voltage ] shows that the hall voltage depends on the scattering time and the mass of the carrier .
therefore , the hall voltages of the different kinds of carriers are also different .
there is thus no voltage at which the two carriers will travel through the sample without deflection .
the difference in the hall voltages is proportional to the applied magnetic field , so that the component of the current parallel to the applied electric field decreases with increasing field .
the resistivity increases therefore when a magnetic field is applied , and the mr is therefore positive .
we call this orbital mr , since it is due to the difference of the masses and scattering times of the two types of carriers .
these properties derive from particularities of the atomic orbitals in the material .
the mr can be calculated from a linearized boltzmann equation that includes a magnetic field under the assumption of a spherical fermi surface .
this is a standard calculation , which can be found in @xcite and @xcite .
we quote the result for the magnetoresistance : @xmath66 where @xmath67 and @xmath68 are the electron and hole conductivity respectively , and @xmath69 and @xmath70 are their mobilities , both in the absence of any magnetic field . the right hand side of the above formula is easily seen to vanish if both types of carriers have the same mass and scattering time : the difference of the mobilities in the numerator vanishes in that case . at low temperatures the orbital mr is largest : the scattering time is largest , so that the deflection is largest as well .
as the scattering time decreases the effect becomes less important .
for sufficiently small scattering times orbital mr becomes negligible .
the second effect that causes mr in eub@xmath0 is the shifting of the bands when a magnetic field is applied .
the shift is caused by the coupling @xmath25 of the carriers to the eu local moments .
let us write the local moments as the sum of their average and the deviations therefrom : @xmath71 .
we now use this expression in the hamiltonian in eq .
[ eq : hamiltonian ] , and obtain a term that couples the carrier s energy to the magnetization . as the magnetization grows , the electron - like band is shifted to lower energies , and the hole - like band to higher energies .
this causes an increase in the number of carriers as carriers spill over from one band into the other
. the change in carrier density can be obtained from the following two requirements .
firstly , the band - shift introduced by the change in magnetization @xmath72 is @xmath73 for electrons and @xmath74 for holes .
secondly , the increase in the number of carriers of both kinds is equal : as the hole like band is shifted up , new holes are created as negatively charged particles spill into the electron - like band and vice versa .
overall charge balance is maintained .
we also assume a spherical fermi surface , an accurate assumption in the case of eub@xmath0 @xcite .
we can use the requirement that charge neutrality be conserved to calculate the shift in the fermi level .
the shift is then used to obtain the number of carriers by integrating the density of states up to the fermi level .
the increase resulting from the shift of the bands is small , even at full saturation of the magnetization .
its effect on the mr at low temperatures is then negligible .
the magnetization as a function of applied field is obtained from a curie - weiss model .
we included this change in carrier density due to the bands shifting in our model for the orbital mr .
the contribution to the mr due to band shifting is opposite to that of the orbital effects .
an external magnetic field increases the carrier density , so that it decreases the resistivity .
since the magnetization is almost saturated at low temperatures , the carrier density does not change much with applied field , and the mr is affected only slightly . the orbital contribution to
the mr dominates .
we plot the mr obtained for the combined effect of the band shifts and the orbital effects in fig .
[ fig : mrlowt ] .
the temperature dependent scattering times at zero field are obtained from experimental data in @xcite .
the electron and hole mobilities are obtained from the conductivity at zero field , the masses of the carriers@xcite and the carrier densities .
the latter were obtained from the plasma frequency in @xcite .
we introduced a small imbalance between the carrier densities of @xmath75 per unit cell , in accordance with @xcite .
this imbalance is thought to arise from impurities .
these numbers provide input to the model .
our simple model , which depends only on parameters measured at zero field , can reproduce the large positive mr at low temperatures accurately .
[ fig : mrlowt ] shows that it reproduces the magnitude of the mr well .
it also predicts the ( nearly ) quadratic dependence in applied field . from fig .
[ fig : mrlowt ] it is clear that the orbital contribution to the mr dwindles at higher temperatures
. the model proposed will lose its validity near the magnetic transition and in the paramagnetic phase , when other effects dominate .
the mechanisms that govern the low - temperature mr become insignificant near the ferromagnetic transition . close to @xmath13 the scattering time is so short that the positive orbital contribution to the mr is negligible . on the other hand ,
the shift of the bands caused by an applied magnetic field becomes substantial .
we estimate that the carrier density changes by about @xmath77 as the applied field saturates the magnetization .
this could explain only part of the negative mr in the critical regime .
aditionally , near the critical point , spin fluctuations may provide a large contribution to the electrical resistance .
the dominant modes near a ferromagnetic transition are those with small @xmath78 .
they only produce substantial backscattering if @xmath79 is itself small @xcite .
this is the case of eub@xmath0 as its carrier density is very small .
the suppression of the spin fluctuations when a magnetic field is applied is largely responsible for the mr found in the critical regime .
moreover , as shown in ref .
@xcite ( see fig .
3 ) , the localization of the carriers in magnetic polarons further increases the magnetoresistance . in the temperature regime - just above @xmath13 - where the sfrs data gives evidence for bound magnetic polarons , the magnetoresistivity is large , and strongly negative , as expected since an applied magnetic field suppressed the magnetic polarons . at temperatures greater than about @xmath80 where the polarons are destabilized
, we have a smaller ( but still negative ) mr dominated by local spin fluctuation scattering .
as mentioned in the introduction , the negative mr is accompanied by a shift in the plasma frequency @xcite .
the carrier density change produced by the band shifting alone is too small to account for that shift .
this is in contrast to the results by wigger _
et al _ @xcite who claim consistency between the experiment and their calculation , though the functional dependence @xmath81 @xcite is not reproduced .
there are two main differences between their model and ours : they consider eub@xmath0 to be a strongly compensated n - type magnetic semiconductor and use the same local coupling @xmath25 for electrons and holes .
eub@xmath0 is a low carrier density ferromagnet with unusual properties : the resistance changes from metallic to activated and then to metallic as the temperature increases and the magnetoresistance changes sign close to @xmath42 .
the activated region has been ascribed to the existence of bound magnetic polarons .
we discuss their existence in the light of spin flip raman scattering measurements reported in ref.@xcite .
we conclude that the signature seen by those experiments is due to the hund s like coupling of holes with the local spins while electrons are responsible for the magnetic ordering through the rkky interaction .
this resolves the puzzle that the rkky transition temperature implies an exchange coupling of the carriers to the local moment of about 0.1 ev , about 30 times larger than the measured spin - flip energy of a carrier trapped in a bound polaron .
the existence of bound magnetic polarons is also consistent with a large _ negative _ magnetoresistance above @xmath13 .
the _ positive _ magnetoresistance in the ferromagnetic phase is also produced by the interplay of two kinds of carriers with different masses and scattering times . in the diluted magnetic semiconductors of the type cd@xmath47mn@xmath48se , the mn are antiferromagnetically coupled via superexchange .
this is a shortrange interaction . at very low @xmath31 ,
the probability of having mn clusters is very low so @xmath82 . for larger @xmath31 , antiferromagnetically coupled clusters of mn form , so not all the mn ions will participate in the magnetization of the system .
see , for instance @xcite . | eub@xmath0 is a low carrier density ferromagnet which exhibits large magnetoresistance , positive or negative depending on temperature .
the formation of magnetic polarons just above the magnetic critical temperature has been suggested by spin - flip raman scattering experiments .
we find that the fact that eub@xmath0 is a semimetal has to be taken into account to explain its electronic properties , including magnetic polarons and magnetoresistance . |
there are three distinct types of solar wind identified by @xcite .
first , there is relatively high coronal electron temperature wind originating from loops in the streamer stalk region @xcite .
second , there is solar wind from the outside of this region .
this wind includes coronal hole wind that has relatively low coronal electron temperatures and high wind speeds , as well as slower solar wind with lower coronal electron temperatures than the stream stalk region .
the third type of solar wind is the transient interplanetary coronal mass ejections ( icmes ) which are caused by the coronal mass ejections ( cmes ) @xcite .
the streamer stalk region is the narrow region in the middle of the streamer belt , which has the highest density fluctuations and the lowest solar wind speeds @xcite .
@xcite provide observational evidence that the streamer stalks can be the coronal sources of the slow solar wind .
@xcite suggests that slow solar wind originates from regions of rapidly expanding flux - tubes located above small coronal holes and at the boundaries of the large polar holes .
based on a comparison of solar remote and in - situ observations , @xcite suggests that low - speed wind with higher @xmath2 ratios may originate from open fields in or near active regions .
fisk and collaborators ( e.g. * ? ? ?
* ; * ? ? ?
* ) suggest that reconnection between open and closed field lines releases material to form the solar wind .
this model provides a reasonable explanation for the differences between fast and slow solar wind .
the distribution of the three types of wind varies with the solar cycle . at solar minimum ,
the coronal holes concentrate at both poles and high latitude coronal hole wind is observed @xcite .
the heliospheric current sheet is flat and lies near the equatorial plane , and the streamer belt stalk wind occurs in a band around the current sheet @xcite .
the icme rate is roughly proportional to the solar activity levels and therefore is very low at solar minimum @xcite . at solar maximum ,
the current sheet tilts to high latitudes , and the streamer - stalk wind , which still occurs in a band around the current sheet , now can reach high latitudes .
the polar coronal holes shrink , resulting in less coronal hole wind in the heliosphere .
the increasing rate of icmes can temporarily enhance the open magnetic flux of the sun .
subsequently , interchange reconnection between the large icmes loops and the open field of the sun eliminates the increased magnetic flux @xcite .
further , there is no compelling observational evidence to suggest that disconnection of open magnetic flux occurs at the heliospheric current sheet @xcite .
hence , the expectation was , prior to this solar minimum , the open magnetic flux would return to a constant background level , as it had in previous minima @xcite .
in the current solar minimum , both the open magnetic flux and the mass flux of the solar wind are reduced compared to any previous solar minimum for which there are good space observations . in this paper , we offer a possible explanation for the decrease in open magnetic flux in the current minimum .
we point out that the streamer - belt - stalk - associated wind originates from a narrower region in the current solar minimum compared to the previous one , and thus the region outside the streamer belt stalk region is larger .
when we calculate the increase in area outside the stalk region , we find it is equal and opposite to the decrease in open magnetic flux , suggesting that the total magnetic flux in the region outside the stalk region remains constant in each solar minimum .
the implication for the transport models of open flux developed by fisk and colleagues is then discussed .
the _ ulysses _ 18-year mission started in 1991 @xcite and terminated in 2009 , providing sufficient data to compare the different conditions in the two minima . here
we use the data from the year of 1995.07 - 1998.2 ( carrington rotation 1892 - 1933 ) at last minimum and 2005.83 - 2008.96 ( carrington rotation 2036 - 2077 ) at the current minimum and compare the difference .
as shown in figure [ figure1 ] , the radial component of the heliospheric magnetic field , the so - called open magnetic flux of the sun ( @xmath3 ) , decreases by 35.4% ; @xmath2 , @xmath4 , and fe / o all decrease by 61.5% , 66.2% and 10.1% , respectively . especially , when considering the region outside of the streamer stalk ( as identified later ) , we find the open magnetic flux decreased by 30% . , @xmath2 , @xmath5 , and fe / o at the previous ( black ) and the current ( dotted ) minimum from _
ulysses_.,title="fig : " ] + .in - situ signatures of three types of solar wind [ cols="^,^,^,<",options="header " , ]
fisk and colleagues developed a model for the global transport of open magnetic flux on the sun , which is illustrated in figure [ figure3]a @xcite .
differential rotation drives the open flux across the polar coronal hole and then into closed field regions where open flux does not disconnect at the current sheet , but rather the flow patterns turn as shown .
the process by which the magnetic field is transported through the closed field region is diffusion due to reconnection with loops .
this model accounts for a number of features of the observed behavior of open magnetic flux ; e.g. , the slow solar wind appears to come from large coronal loops outside of coronal holes , and be released by reconnection @xcite .
-axis is the solar rotation axis .
p marks the open line ( green ) that connects to the pole .
the curves with arrows ( red ) are the trajectories of the open field lines , and the yellow region is the streamer stalk region .
( b)the open lines reconnects and diffuses outside the streamer stalk region . , title="fig : " ] + this picture now needs to be revised , as shown in figure [ figure3]b .
the open magnetic flux in regions outside the streamer - stalk region is unable to penetrate into this region .
thus , disconnection of this component of open flux , which must occur at the heliospheric current sheet , is not possible .
rather , the turning of the flow patterns of open flux must occur outside the streamer stalk region , as shown .
the total open magnetic flux outside of the streamer - stalk region , which can not now disconnect , is conserved , as is observed to be the case .
there are several points worth emphasizing .
the signature we use to identify our streamer - belt - stalk wind is the @xmath2 ratio , or the inferred coronal electron temperature , not the solar wind speed , as was used in other studies .
moreover , it is important to note that the streamer - stalk wind is not the entire slow speed solar wind .
rather , it is only the very slow , high coronal electron temperature wind , which we identify as originating from the streamer stalk underlying the heliospheric current sheet .
there is a broader slow solar wind region @xcite , as constrained by the black dotted line in figure [ figure3]b .
the conservation of the total magnetic flux in the non - streamer - stalk region during the two solar minima suggests that the open magnetic field of the sun in the current solar minimum is behaving as it did in previous minima , the only difference being the width of the streamer belt stalk region , which controls the magnetic field strength in the region outside the streamer belt stalk region . this work was supported in part by nasa headquarters under the nasa earth and space science fellowship program - grant nnx09av13h , by the heliophysics theory program , by nasa / jpl contract 1268016 , and by nsf grant atm 0632471 | to explore the difference between the most two recent solar minima , we analyze the in - situ _ ace _ and _ ulysses _ observations and examine the distributions of the three types of solar wind ( streamer - stalk - associated wind , wind from outside the streamer stalk that can be associated , in part , with coronal holes , and interplanetary coronal mass ejections ) .
we use the taxonomy provided by @xcite to identify the three types of solar wind .
we then map the in - situ observations to the 2.5 solar radii surface . with the aid of the potential - field - source - surface model ( pfss )
, we calculate the normal distance from the solar wind `` foot point '' to the local helisopheric current sheet on that surface .
we find that the source region of the streamer stalk wind is narrower ( @xmath0 ) compared to the previous minimum ( @xmath1 ) .
the area outside the streamer stalk is accordingly larger , but the magnetic field strength is observed to be lower , with the result that the total amount of the magnetic open flux from the outside of streamer stalk region is conserved in the two successive solar minima .
the implications of the conservation of open magnetic flux for models of the behavior of the solar magnetic field are discussed . |
since the confirmation by wilkinson microwave anisotropy probe ( wmap ) observations of the cosmic microwave background that the large scale ( @xmath0 in comoving units ) auto - correlation function of temperature fluctuations is close to zero [ fig . 16 , @xcite ( orthogonally projected spatial scale ) ; fig . 1 of @xcite ( non - projected spatial scale ) ] and difficult to reconcile with a standard cosmic concordance model ( 7 of @xcite
; though see also @xcite ) there has recently been considerable observational interest in understanding multiply connected models of the universe .
the poincar dodecahedral space ( pds ) has been proposed as a better model of the 3-manifold of comoving space rather than an `` infinite '' flat space ( e.g. @xcite ; though see also @xcite ) .
curvature estimates are consistent with the model : analysis of the 3-year wmap data @xcite results in best estimates of the total density parameter @xmath1 ( when combined with hst key project on @xmath2 data ) and @xmath3 ( when combined with supernova legacy survey data ) , or together with the third , fourth and fifth acoustic peaks , as estimated from their observations near the south galactic pole using the arcminute cosmology bolometer array receiver ( acbar ) , @xmath4 is obtained @xcite , consistently with that expected from the pds analyses , which require positive curvature in this range of @xmath5 values .
it has also recently become noticed that global topology in a universe containing density perturbations can , in principle , have at least some effect on the metric , even though the effect is expected to be small . at the present epoch , in the case of a @xmath6 model of fundamental lengths which are slightly unequal by a small fraction @xmath7
, a new term appears in the acceleration equation , causing the scale factor to accelerate or decelerate in the directions of the three fundamental lengths in a way that tends to equalise them @xcite . in this context , the properties and implications of the twin paradox of special relativity in a multiply connected space need to be correctly understood .
it has already been shown @xcite that resolving the twin paradox of special relativity in a multiply - connected `` minkowski '' space - time implies new understanding of the paradox relative to the case in simply connected minkowski space . moreover , it is known that , at least in the case of a static space with zero levi - civita connection , multiple connectedness implies a favoured space - time splitting .
this should correspond to the comoving reference frame @xcite .
this could be of considerable importance to the standard cosmological model , since it would provide a novel physical ( geometrical ) motivation for the existence of a favoured space - time foliation , i.e. the comoving coordinate system .
the difference between the twin paradox of special relativity in a multiply connected space relative to that in a simply connected space is that in a multiply connected space , the two twins can move with constant relative speed and meet each other a second time , _ without requiring any acceleration_. the paradox is the apparent symmetry of the twins situations despite the time dilation effect expected due to their non - zero relative speed .
it is difficult to understand how one twin can be younger than the other why should moving to the left or to the right be somehow favoured ? does the time dilation fail to occur ?
as shown by several authors @xcite , the apparent symmetry is violated by the fact that ( at least ) one twin must identify the faces of the fundamental domain of the spatial 3-manifold _ non - simultaneously _ , and has problems in clock synchronisation .
@xcite suggested that the apparent symmetry is also violated by an asymmetry between the homotopy classes of the worldlines of the two twins . here , we reexamine this suggestion . the multiply connected space version of the twins paradox is briefly reviewed in [ s - paradox ]
, the question of clarifying the asymmetry is described in [ s - asymmetry ] , and a multiply connected space - time , with a standard minkowski covering space - time , and the corresponding space - time diagrams , are introduced in [ s - st - diagrams ] . in [ s - worldlines - etc ] ,
the worldlines of the two twins , projections from space - time to space , homotopy classes and winding numbers are expressed algebraically . in [ s - results ] , the resulting projections of the twins paths from space - time to space are given and their homotopy classes inferred .
we also calculate whether or not the generator of the multiply connected space and the lorentz transformation commute with one another , in [ s - lorentz ] .
discussion suggesting why this result differs from that of @xcite is presented in [ s - discuss ] , including a brief description in [ s - tperiodic ] of how the non - favoured twin can choose a time - limited instead of a space - limited fundamental domain of space - time .
conclusions are summarised in [ s - conclu ] . a thought experiment to develop physical intuition of the non - favoured twin s understanding of the spatial homotopy classes
is given in appendix [ s - stretchable - cord ] .
for a short , concise review of the terminology , geometry and relativistic context of cosmic topology ( multiply connected universes in the context of modern , physical cosmology ) , see @xcite ( slightly outdated , but sufficient for beginners ) .
there are several in - depth review papers available @xcite and workshop proceedings are in @xcite and the following articles , and @xcite . comparisons and classifications of different _ observational _ strategies ,
have also been made @xcite .
we set the conversion factor between space and time units to unity throughout : @xmath8 .
the first articles presenting the special relativity twins paradox in a multiply connected space were apparently @xcite and @xcite .
there are several more recent articles , including e.g. @xcite and @xcite .
consider the following description . in a one - dimensional ,
multiply connected , locally lorentz invariant space , one twin moves to the left and one to the right in rockets moving at constant relative speed to one another .
the two twins meet twice , at two distinct space - time events . at the earlier space - time event ,
the two twins are of equal ages . at the later space - time event
, each twin considers the other to be younger due to lorentz time dilation .
however , this later space - time event is a single space - time event each twin has undergone physical aging processes . if necessary
, each twin could carry an atomic clock in order to more precisely measure proper time than with biological clocks .
so there can only be one ordinal relation between the two twins ages at the second space - time event : either the leftward moving twin is younger , or the rightward moving twin is younger , or the two twins are of equal age .
which is correct ?
there is no acceleration ( change in velocity ) by either twin , so the usual explanation of the paradox ( in simply connected space ) is invalid .
however , in this case , the situation is , or at least seems to be , perfectly symmetrical .
it would be absurd for either `` leftwards '' or `` rightwards '' movement to yield a younger age . on the other hand , time dilation implies that the `` other '' twin must `` age more slowly '' .
the twins physically meet up ( for an instant ) at the second space - time event , which is a single location in space - time , so the `` first '' twin objectively measures that the `` other '' twin is younger , so the two twins can not be equally aged at the second space - time event .
but then which twin is `` the first '' and which is `` the other '' ?
this question illustrates the apparent symmetry of the situation and the apparent implication that `` time dilation fails '' . alternatively ,
if the situation is _ not _ symmetrical and time dilation occurs as is expected , then what breaks the apparent symmetry ? why should leftwards movement by favoured relative to rightward movement , or vice - versa ?
which is correct : is the situation symmetrical with a failure of time dilation , or is the situation asymmetrical with time dilation taking place ? in @xcite , @xcite , @xcite and @xcite
it was shown that the apparent symmetry in the question as stated above is not mathematically ( physically ) possible .
there is a hidden implicit assumption related to the usual intuitive error common to beginners in special relativity : the assumption of absolute simultaneity .
the necessary asymmetry can be described in different ways .
one way of explaining the asymmetry is as follows .
one twin is able to consistently synchronise her clocks by sending photons in opposite directions to each make a loop around the universe and observing their simultaneous arrival time , and the other twin measures a delay between receiving the two photons ( or coded signal streams ) and is forced to conclude that something is asymmetrical about the nature of her `` inertial '' reference frame @xcite .
carrying out this test requires a metrical measurement , that of proper time intervals . here , in order to examine @xcite s suggestion that there is a homotopy asymmetry , i.e. an asymmetry of topology rather than an asymmetry depending on the metric , it is easier to first explain the asymmetry of the apparently symmetrical paradox in a more geometric way , similar to the presentation in @xcite , but with some additional figures . and coordinate system @xmath9 , made multiply connected in each spatial section at constant time @xmath10 via the generator ( translation ) @xmath11 .
a twin moving to the right at constant relative velocity @xmath12 ( hereafter , the `` rightmoving twin '' ) has worldline @xmath13 and simultaneity axis @xmath14 where @xmath15 , defining her coordinate system @xmath16 .
space - time points a@xmath17 and a@xmath18 are identical space - time events under the generator @xmath11 , i.e. the single event may in general be called a. similarly , space - time points b@xmath17 , b@xmath18 and b@xmath19 are a single space - time event b. space - time events c and d are distinct from each other and from a and b ; c and d occur at the same spatial location for the rightmoving twin .
the worldlines of the leftmoving and rightmoving twins are @xmath20 and @xmath21 respectively .
[ f - spacetimeone ] , width=302 ] figure [ f - spacetimeone ] shows a standard minkowski space as covering space for simplicity with only one spatial dimension , for two twins moving with constant velocity relative to one another , hereafter , the `` leftmoving '' and `` rightmoving '' twins respectively . specifically , we set the rightmoving twin moving at a velocity @xmath12 , in units of the space - time conversion constant @xmath22 , towards the right .
as a covering space , this space is a standard locally and globally lorentz invariant space - time @xmath23 .
we choose a generator @xmath11 which favours , arbitrarily , but without loss of generality , the leftmoving twin .
this generator , @xmath11 , a translation of constant length @xmath24 with @xmath25 generates the quotient , multiply connected space , @xmath26 , where @xmath27 is the group generated by @xmath11 , i.e. @xmath28 .
this arbitrary choice reveals where an implicit assumption was made in the presentation of the paradox above : a generator matching space - time events in a way that preserves time unchanged in one reference frame , or in other words , a generator which `` is simultaneous '' in one reference frame , is not simultaneous in other frames .
hence , symmetry is not possible . , but shown in the rest frame of the rightmoving twin .
the identities due to the generator @xmath11 remain correct : a@xmath29a@xmath18 and b@xmath29b@xmath30b@xmath19 , _ even though they are non - simultaneous _ ; @xmath11 could be described as a non - simultaneous generator in the rightmoving twin s reference frame .
the worldline @xmath20 of the leftmoving twin is mapped by the generator to a physically equivalent copy , @xmath31 , running from a@xmath18 to b@xmath18 .
figure [ f - cylindertwo ] helps show this is possible by using our three - dimensional intuition .
[ f - spacetimetwo ] , width=302 ] the generator @xmath11 identifies points ( in three - dimensional space , these would be faces of the fundamental domain rather than points ) in a spatial section at any given time @xmath10 : a@xmath29a@xmath18 and b@xmath29b@xmath30b@xmath19 .
the rightmoving twin has @xmath14 and @xmath13 axes different from those of her leftmoving twin , in order to preserve lorentz invariance .
she disagrees with the leftmoving twin about simultaneity of events , finding , e.g. that space - time event a@xmath18 occurs before space - time event a@xmath17 , space - time event b@xmath18 occurs before space - time event b@xmath17 , etc .
this is shown in fig .
[ f - spacetimetwo ] , where @xmath32 is the doppler boost .
so far , this is identical to the situation in simply connected minkowski space - time , until we realise that both twins must agree that a@xmath29a@xmath18 and that b@xmath29b@xmath30b@xmath19 .
, embedded in 3-d euclidean space and projected onto the page , or informally , `` rectangle a@xmath17a@xmath18b@xmath18b@xmath17 rolled up into a cylinder and stuck together to make it multiply connected '' .
the leftmoving twin s worldline @xmath20 is shown .
[ f - cylinderone ] , width=132 ] , embedded in 3-d euclidean space and projected onto the page , or informally , `` rectangle a@xmath17cdb@xmath18 rolled up and stuck together to make it multiply connected '' . _ the spatial boundaries of this region , @xmath33 and @xmath34 , are offset by a time interval a@xmath18c @xmath35 before being matched : the result is not a cylinder .
_ note that in space - time , there are two geodesics joining space - time events a@xmath17 and @xmath36 : one at constant spatial position ( appearing vertical here ) , and one at constant time ( appearing nearly horizontal , but sloped at a moderate angle in this projection ) , and similarly for b@xmath18 and @xmath37 .
however , only one of these two geodesics the vertical ( timelike ) one can be a worldline of a physical ( non - tachyonic ) particle , so there is no causality violation . a similar diagram to this one
has earlier been published in figure 5b in @xcite .
the rightmoving twin s worldline @xmath21 from a@xmath17 to b@xmath18 is shown .
[ f - cylindertwo ] , width=94 ] while both twins agree that a@xmath29a@xmath18 , they _
disagree _ as to whether or not these are simultaneous events . using the terminology of @xcite , the leftmoving twin is able to synchronise clocks , while the rightmoving twin is _ unable to synchronise clocks_. writing the lorentz transformation from @xmath9 to @xmath16 as a matrix , @xmath38 the generator @xmath11 , initially expressed in the first twin s coordinates in eq .
( [ e - generator ] ) , can be rewritten using the second twin s coordinates as @xmath39 an intuitive , geometric way of describing this is that according to the rightmoving twin , _ after cutting the covering space , non - simultaneous points are pasted together_. figure [ f - cylinderone ] shows the cylinder `` cut and pasted together '' out of a space - time region with constant space and time boundaries according to the leftmoving twin .
note that a trapezium in fig .
[ f - spacetimeone ] , e.g. a@xmath17a@xmath18b@xmath19b@xmath18 would serve just as well as the rectangle a@xmath17a@xmath18b@xmath18b@xmath17 for this `` rolling up '' process . as long as there are boundaries of constant time @xmath10 ,
the result of identifying the other two sides of the trapezium is a cylinder .
this trapezium , a@xmath17a@xmath18b@xmath19b@xmath18 , is particularly interesting when we shift to the reference frame of the rightmoving twin in fig .
[ f - spacetimetwo ] , since the boundaries a@xmath17b@xmath18 and a@xmath18b@xmath19 now become spatial boundaries : @xmath40 cutting and pasting from the rightmoving twin s point of view must still identify identical space - time events to one another : _ either _ identifying a@xmath17b@xmath17 to a@xmath18b@xmath18 , or a@xmath17b@xmath18 to a@xmath18b@xmath19 , will correctly apply the isometry to the covering space and `` paste '' together our spatially finite interval in order to obtain a manifold without any boundaries .
( the time domain extends infinitely from this rectangle in both the negative and positive directions . )
so , one option for embedding this identification in 3-d euclidean space and projecting it onto the page would be to use the same trapezium .
this corresponds to the rightmoving twin s intuition of identifying `` two spatial points '' to one another while trying to ignore the nature of space - time as a two - dimensional continuum : the set of points along the line segment a@xmath17b@xmath18 constitute a single `` spatial point '' @xmath41 , while the set of points along the line segment a@xmath18b@xmath19 constitute a single `` spatial point '' @xmath42 , a `` spatial point '' is really a worldline it is not just a single point , it s a curve in space - time . however , rather than identifying the `` spatial borders '' of a@xmath17a@xmath18b@xmath19b@xmath18 to one another , it is helpful to follow the rightmoving twin s naive intuition even further
. let us try to cut out and then paste together the space - time region with both constant space boundaries and constant time boundaries , i.e. the region a@xmath17cdb@xmath18 .
the result is shown in fig .
[ f - cylindertwo ] ( cf . fig .
5b in @xcite ) .
this clearly shows the non - simultaneity of the cutting / pasting process for the rightmoving twin .
the rectangle in @xmath16 space - time has to be given a time mismatch when it s pasted together .
this visually illustrates the error in the twins paradox in a multiply connected space ( as described in [ s - paradox ] ) : _ the implicit assumption of absolute simultaneity_. if we implicitly assume absolute simultaneity , then we implicitly assume that there is no inconsistency in supposing that both twins can identify spatial boundaries without any time offsets .
however , lorentz invariance is inconsistent with absolute simultaneity ; hence , the asymmetry : at most one twin can simultaneously identify spatial boundaries .
of course , neither the leftmoving twin nor the rightmoving twin are necessarily favoured .
a complete , precise statement of the problem needs to arbitrarily favour one twin over the other : either the leftmoving or the rightmoving twin may be chosen , but one of them must be chosen to be favoured in order for the space - time to be self - consistent .
this also illustrates the potential interest for cosmology : several authors note the existence of a favoured inertial reference frame implied by the multiple - connectedness of a ( static ) space - time whose covering space - time is minkowski and that this could provide a physical motivation for the comoving coordinate system @xcite .
is the apparent ( erroneous ) symmetry of the two twins situations broken by asymmetry between the homotopy classes of the two twins projected worldlines ( `` spatial paths '' ) in some way ?
in @xcite it was suggested that the twin who simultaneously identifies spatial boundaries ( in this paper , the leftmoving twin ) has a spatial path of zero winding index , while the rightmoving twin has a spatial path of non - zero winding index .
we consider the worldlines of the two twins between the two space - time events a and b , using figs [ f - spacetimeone ] and [ f - spacetimetwo ] and eqs ( [ e - generator ] ) , ( [ e - lorentz ] ) and ( [ e - generator - moving ] ) .
a superscript `` l '' or `` r '' is used to denote which reference frame is being used for expression in a particular coordinate system . the leftmoving twin s worldline , @xmath43 can be written in her own reference frame as @xmath44 using her proper time @xmath45 ( see fig .
[ f - spacetimeone ] ) .
a copy of the leftmoving twin s worldline as mapped by @xmath11 can be written , using eq .
( [ e - generator ] ) , as @xmath46 we can write this in the rightmoving twin s frame using eq .
( [ e - lorentz ] ) : @xmath47 the rightmoving twin s worldline , @xmath48 can be written in her own reference frame as @xmath49 using her own proper time @xmath50 ( see fig .
[ f - spacetimetwo ] ) , or equivalently in the leftmoving twin s frame , using @xmath51 and eq .
( [ e - lorentz ] ) , as @xmath52 @xcite state that `` the asymmetry between [ the two twins ] spacetime trajectories lies in a topological invariant of their spatial geodesics , namely the homotopy class '' . in order to consider spatial geodesics in the two twins reference frames , we need to project from space - time to space in each of their reference frames .
let us write these projections @xmath53 and @xmath54 algebraically in the leftmoving and rightmoving twins frames in the most obvious way : @xmath55 and @xmath56 using eqs ( [ e - generator ] ) and([e - generator - moving ] ) , the generator @xmath11 can then be rewritten in _ projected _ form as @xmath57 and @xmath58 in the two twins frames respectively .
the two spaces which space - time is projected to are clearly both equivalent to @xmath59 , where the group @xmath27 is defined by the generator @xmath60 or @xmath61 for the leftmoving or rightmoving twin respectively .
the length ( circumference ) of @xmath62 is @xmath24 or @xmath63 respectively .
it is important to note here that _ in the absence of metric information , the two twins do not know anything more than this about the nature of their respective spaces , and they do not know how their spaces relate to spatial sections , i.e. subspaces of the covering space @xmath23 . _
in fact , metric information gives an important difference between these two ( in this paper , one - dimensional ) spaces : the rightmoving twin s constant time hypersurface has a time mismatch after `` going around '' the universe once .
however , here the question is to examine the _ topological _ nature of the twins spatial paths in the _ absence _ of metric information .
the space - time point of view which the rightmoving twin would have , if she had metric information available ( e.g. by sending out photons and measuring proper time intervals ) , is irrelevant . similarly to @xcite
, we write that two loops in @xmath64 at a are homotopic to one another , @xmath65 `` if they can be _ continuously _ deformed into one another , i.e. that there exists a continuous map '' @xmath66 satisfying @xmath67 as in @xcite , we will assume that no proof is needed that @xmath68 , and we define the winding number as the `` the number of times a loop rolls around '' the circle @xmath62 . since the aim is to see if the leftmoving and rightmoving twins points of view can be distinguished in order to determine who is favoured , we need indices on the winding number to indicate whether it refers to the worldline of the leftmoving or rightmoving twin , and from whose point of view . again , we use the superscript `` l '' or `` r '' to denote which reference frame is being used . a twin s point of view of her own winding number is denoted by an empty subscript , and the subscript `` other '' is used for a twin s point of view of the other twin s winding number .
that is , the leftmoving twin measures her own winding number @xmath69 and the other twin s winding number @xmath70 , and the rightmoving twin measures her own winding number @xmath71 and the other twin s winding number @xmath72 . in each case , the winding number is positive in the direction of motion of the `` other '' twin . , including the worldlines @xmath20 and @xmath21 of the leftmoving and rightmoving twins respectively , corresponding to @xmath73 and @xmath33 in the covering space - time as shown in figure [ f - spacetimeone ] , and showing their projections from space - time into space under the projection @xmath53 given in eq .
( [ e - phi - l ] ) .
arrows indicate increasing proper time along each worldline .
the worldlines are given algebraically in [ s - worldlines ] .
[ f - cylinderone_homotopy ] , width=226 ] , including the worldlines @xmath74 of the leftmoving and rightmoving twins respectively .
these correspond to @xmath33 and @xmath75 in the covering space - time as shown in fig .
[ f - spacetimetwo ] . also shown
are their projections from space - time into space under the projection @xmath54 given in eq .
( [ e - phi - r ] ) .
arrows indicate increasing proper time along each worldline .
the worldlines are given algebraically in [ s - worldlines ] .
[ f - cylindertwo_homotopy ] , width=226 ]
figure [ f - cylinderone_homotopy ] shows that @xmath76 projects to a point , and @xmath77 projects to a closed loop . as stated in @xcite , the former path has a zero winding index , while the second has a unity winding index : from the point of view of the leftmoving twin , there is a difference in the two twins winding numbers .
however , we havent yet checked if the winding numbers are relative or absolute .
we evaluate the winding numbers as follows . using the two twins worldlines from the leftmoving twin s point of view , eqs ( [ e - p - l ] ) and ( [ e - q - l ] ) , and projecting them with eq .
( [ e - phi - l ] ) , we have @xmath78 and @xmath79 since @xmath80 from eq .
( [ e - generator - proj ] ) , clearly @xmath81 and @xmath82 in the sense of eq .
( [ e - defn - homotopy ] ) , so that the winding index of the leftmoving twin is zero , i.e. @xmath83 and that of the rightmoving twin is unity , @xmath84 figure [ f - cylindertwo_homotopy ] shows that @xmath77 projects to a point , and @xmath85 projects to a closed loop .
the former path has a zero winding index , while the latter has a unity winding index : from the point of view of the rightmoving twin , there is a difference in the two twins winding numbers .
however , if each twin each assume that she is stationary and that the other is moving away , as we have assumed throughout , then they also disagree regarding the winding numbers : neither simultaneity nor winding number is absolute .
we evaluate this algebraically as follows .
using the two twins worldlines from the rightmoving twin s point of view , eqs ( [ e - pg - r ] ) and ( [ e - q - r ] ) and projecting them with eq .
( [ e - phi - r ] ) , we have @xmath86 \label{e - pg - r - projected } \ ] ] and @xmath87 eq . ( [ e - q - projected ] )
shows that in this case , the rightmoving twin has winding index zero : @xmath88 since @xmath89 from eq .
( [ e - generator - moving - proj ] ) , eq .
( [ e - pg - r - projected ] ) clearly gives @xmath90 using the sign convention which we use globally here for the @xmath91 or @xmath14 coordinates , the winding index of the leftmoving twin in this case would be @xmath92 .
more usefully , using the winding number convention described in [ s - winding - defn ] , i.e. the winding number is positive in the direction of motion of the `` other '' twin , this becomes @xmath93 in summary , from eqs .
( [ e - nl ] ) , ( [ e - nl - other ] ) , ( [ e - nr ] ) and ( [ e - nr - other ] ) , we have @xmath94 clearly , the homotopy classes of the twins worldlines do not reveal the asymmetry of the system : each twin considers herself to have winding index @xmath95 and the other twin to have @xmath96 .
@xcite also suggested in their eqs ( 6)(8 ) that the generator @xmath11 does not commute with the lorentz transformation .
here we calculate this explicitly .
equation ( [ e - lorentz ] ) gives @xmath98 applying the generator in the appropriate reference frame , eq .
( [ e - generator - moving ] ) , gives @xmath99 \nonumber \\ & = & \gamma [ x -\beta t + l , t -\beta ( x + l ) ] .
\label{e - gofl - two } \end{aligned}\ ] ] equation ( [ e - generator ] ) gives @xmath100 the lorentz transformation , eq .
( [ e - lorentz ] ) , gives @xmath101 \nonumber \\ & = & \gamma [ x -\beta t + l , t - \beta ( x+l ) ] \nonumber \\ & = & g \circ { \cal l } ( ( x , t ) ) , \label{e - gofl - is - lofg}\end{aligned}\ ] ] where the last equality uses eq .
( [ e - gofl - two ] ) .
these last two equations give @xmath102 this result is in contrast to @xcite s suggestion that @xmath103 .
it is not obvious why our results differ . here
, we consider @xmath104 to be a holonomy transformation on the covering minkowski space @xmath23 , independently of any particular coordinate representation . as in @xcite
, we consider @xmath105 to be a transformation `` from one frame of space - time coordinates to another system '' .
the results summarised in [ s - results - both ] differ from the conclusion in @xcite , where it was pointed out that since winding indices are topological invariants , `` neither change of coordinates or reference frame ( which ought to be continuous ) can change [ the winding indices ] values '' , i.e. the rightmoving twin and the leftmoving twin should agree that the leftmoving twin has a zero winding index and the rightmoving twin a unity winding index . , but showing that the two worldlines together form a single closed loop in space - time .
neither worldline constitutes a closed curve alone , prior to projection .
[ f - cylinderone_easy ] , width=170 ] this argument does not take into account the nature of the projection from space - time to space .
the argument is correct in that the topologically invariant nature of winding indices is valid in _ space - time _ , but is not necessarily valid in the worldlines _ after projection from space - time to space_. the projection from an @xmath106-dimensional manifold to an @xmath107-dimensional manifold does not ( in general ) conserve all topological properties of subspaces : while a continuous projection conserves continuity , it does not conserve non - continuity .
for example , consider a discontinuous path @xmath108 in the leftmoving twin s reference frame , where @xmath109 is a step function , i.e. a function discontinuous for some values @xmath110 .
this projects under @xmath53 to a continuous , closed loop . similarly in higher dimensions : consider @xmath111 , which can be continuously deformed into something resembling a mess of string with the two ends tied together and not touching itself anywhere
. project this from euclidean 3-space into the euclidean 2-plane .
the projection will ( in general ) be a graph with many nodes , not @xmath62 .
in fact , the worldlines ( e.g. those labelled 1 , 2 , 3 , 4 in fig
. 2 in @xcite ) are all open curves in space - time , i.e. each of them is in the same homotopy class as a single point .
figures [ f - cylinderone_homotopy ] and [ f - cylindertwo_homotopy ] show that these open curves only become either a loop or a point after projection .
the projection can convert a discontinuity into a continuity ( or a continuum of points into a single point ) . in space - time , it is the union of the two twins worldlines ( two different paths in space - time from space - time event a to space - time event b ) which forms a closed curve , not either worldline alone .
this is shown in fig .
[ f - cylinderone_easy ] .
this is why a `` change of coordinates or reference frame '' using a continuous function _ can _ change the winding index values of two worldlines , if those two worldlines are projections from @xmath112-space - time to @xmath106-space , considered in two different reference frames . nevertheless , with the use of metrical information and space - time rather than just space , it is interesting to note some different possibilities of what `` space '' could correspond to if thought of as a subspace of the covering space - time @xmath23 .
let us now use our knowledge of metrical information for considering the nature of `` space '' , i.e. spacelike sections of space - time , for the two twins .
, again in the reference frame of the rightmoving twin , showing part of a hypersurface at constant time for the rightmoving twin as thick horizontal line segments .
space - time events c@xmath17 and c@xmath18 are identical to one another ( the generator @xmath11 identifying equal space - time events is illustrated ) ; space - time events e@xmath17 , e@xmath18 and e@xmath19 are identical to one another .
a is a single space - time event , c is a single space - time event , and e is a single space - time event .
the fundamental domain ( fd ) can be written as a space - limited region with unconstrained time ( shaded with a backslash `` @xmath113 '' ) and a time offset when matching boundaries .
it can also be written as a _
time_-limited region with unconstrained space ( shaded with a forward slash `` / '' ) and a space offset when matching boundaries . within this fd ,
consider an observer at rest in this frame , capable of making metric measurements , located at space - time event a and shown in the covering space at a@xmath17 .
for this observer , the events c and e happen `` over there and simultaneously to a '' .
the observer will arrive at those spatially distant events without any motion , simply by waiting .
see fig .
[ f - cylindertwo_tperiodic ] for a cylindrical representation of the time - limited fd .
[ f - spacetimetwoconst ] , width=302 ] a spacelike section of space - time at constant time for the leftmoving twin connects with itself and is clearly @xmath62 . and [ s - tperiodic ] .
c@xmath17 and c@xmath18 are a single space - time location ; e@xmath17 and e@xmath18 are a single space - time location .
time boundaries are identified to one another with a spatial offset .
some of the multiple topological images of the worldlines of the leftmoving and rightmoving twins are labelled as previously , i.e. @xmath31 and @xmath21 respectively , as well as @xmath114 .
the positive spatial and time directions are shown as @xmath14 and @xmath13 respectively .
[ f - cylindertwo_tperiodic ] , width=302 ] in contrast , a spacelike section of space - time at constant time for the rightmoving twin does not connect with itself : there is a time offset of @xmath115 after one loop . for a spacelike section to connect with itself
, it would need to be at non - constant time .
this is not a problem for topological measurements in the absence of metrical information : the rightmoving observer who sends out slow moving probes equipped with neither clocks nor rods and having speeds which can not be precisely calibrated will assume that space is @xmath62 . in the covering space ,
this could be most simply interpreted as replacing eq .
( [ e - phi - r ] ) by a projection from @xmath23 to a subspace @xmath116 defined : @xmath117 in the rightmoving twin s reference frame , or equivalently , @xmath118 in the leftmoving twin s reference frame .
in other words , the subspace @xmath62 of @xmath23 correspond to the rightmoving twin s interpretation of space from topological measurements alone would most easily be modelled as a spatial hypersurface of constant time for the leftmoving twin .
while the rightmoving ( non - favoured ) twin making only topological measurements does not _ know _ that her `` space '' @xmath62 consists of a set of space - time points at differing times @xmath13 , the nature of what would be her constant time hypersurface , if she were able to determine it , reveals an interesting property of a multiply connected space - time for a non - favoured observer : _ the fundamental domain can be chosen by matching time boundaries instead of space boundaries .
_ figures [ f - spacetimetwoconst ] and [ f - cylindertwo_tperiodic ] illustrate this situation .
the fundamental domain of this space - time can either be described as the region @xmath119 expressed as the direct product of a spatial interval as the first parameter and a time interval as the second , or as @xmath120 again with space first and time second .
the former , @xmath121 , is the `` obvious '' fundamental domain , satisfying @xmath122 where the generator @xmath11 is easiest to use if expressed in the rightmoving twin s reference frame , i.e. as in eq .
( [ e - generator - moving ] ) , and @xmath123 , @xmath124 , etc .
the latter , @xmath125 , is the more surprising fundamental domain , satisfying @xmath126 again where @xmath11 is best expressed using eq .
( [ e - generator - moving ] ) .
what is surprising is that a verbal description of @xmath125 could be `` time is periodic for the non - favoured twin '' .
however , this description omits important information ; a more complete description would be `` time is periodic , but with a spatial offset , for the non - favoured twin '' .
in other words , the ability to choose a time - limited fundamental domain does _ not _ mean that events periodically repeat themselves from the rightmoving twin s point of view , since the generator of the periodicity does _ not _ yield an offset by a vector @xmath127 . what it yields is a space - time offset by ( an integer multiple of ) the vector @xmath128 .
we could also say that for a non - favoured twin , the space - time periodicity is `` diagonal '' to the space - time axes , i.e. @xmath129
finding an asymmetry in the twin paradox of special relativity in a multiply connected space is less obvious than in a simply connected space , since neither twin accelerates .
it was already known that the asymmetry required is the fact that ( at least ) one twin must identify space - time events non - simultaneously and has problems in clock synchronisation . here
, space - time diagrams and the corresponding algebra have been presented in order to examine whether or not the homotopy classes of the twins worldlines provide another asymmetry .
they show that homotopy classes ( numbers of windings ) do _ not _ show which of the two twins of the twin paradox has a preferred status , contrary to what was previously suggested : each twin finds her own spatial path to have zero winding index and that of the other twin to have unity winding index ( in the direction of travel of the other twin ) .
although the twins apparent symmetry is broken by the need for the non - favoured twin to non - simultaneously identify spatial domain boundaries , and by the non - favoured twin s problems in clock synchronisation ( provided that she has precise clocks ) , the non - favoured twin _ can not _ detect her disfavoured state by measuring the topological properties of the two twins worldlines in the absence of precise metric measurements with clocks or rods . on the other hand , a non - favoured twin capable of making precise metric measurements will notice many surprising properties of space - time .
for example , a non - favoured twin could define the fundamental domain of space - time to be @xmath130 where the first dimension is space and the second is time . in words , time would be `` periodic with period @xmath115 and a spatial offset when matching time boundaries '' as an alternative explanation to having space be `` periodic with period @xmath63 and a time offset when matching space boundaries '' .
we thank agnieszka szaniewska for a careful reading and useful comments on this paper . st.b .
acknowledges suport form polish ministry of science grant no nn203 394234 .
r. , lustig s. , steiner f. , 2005a , cqg , 22 , 3443 , http://arxiv.org/abs/astro-ph/0504656 [ ] r. , lustig s. , steiner f. , 2005b , cqg , 22 , 2061 , http://arxiv.org/abs/astro-ph/0412569 [ ] j. d. , levin j. , 2001 , phys .
a , 63 , 044104 , http://arxiv.org/abs/gr-qc/0101014 [ ] v. , roukema b. f. , eds , 2000 , `` cosmological topology in paris 1998 '' paris : blanlil & roukema , http://arxiv.org/abs/astro-ph/0010170 [ ] c. h. , stewart d. r. , 1973 , phys .
d , 8 , 1662 g. , 2004 , mnras , 348 , 885 , http://arxiv.org/abs/astro-ph/0310207 [ ] j. , 2005 , arxiv preprints , http://arxiv.org/abs/astro-ph/0503014 [ ] m. , luminet j. , 1995 , phys .
, 254 , 135 , http://arxiv.org/abs/gr-qc/9605010 [ ] r. j. , 1990 , european journal of physics , 11 , 25 j. , roukema b. f. , 1999 , in nato asic proc . 541 : theoretical and observational cosmology topology of the universe : theory and observation .
p. 117 , http://arxiv.org/abs/astro-ph/9901364 [ ] j. , weeks j. r. , riazuelo a. , lehoucq r. , uzan j. , 2003 , nature , 425 , 593 , http://arxiv.org/abs/astro-ph/0310253 [ ] j .- p .
, 1998 , acta cosmologica , xxiv-1 , 105 , http://arxiv.org/abs/gr-qc/9804006 [ ] a. , jaffe a. , 2007 , physical review letters , 99 , 081302 , http://arxiv.org/abs/astro-ph/0702436 [ ] p. c. , 1983 , am . j. phys . , 51 ,
791 p. c. , 1986 , am .
j. phys . , 54 , 334 m. j. , gomero g. i. , 2004 , braz
. j. phys . , 34 , 1358 , http://arxiv.org/abs/astro-ph/0402324 [ ] c. l. , ade p. a. r. , bock j. j. , bond j. r. , brevik j. a. , contaldi c. r. , daub m. d. , dempsey j. t. , goldstein j. h. , et al .
2008 , arxiv e - prints , 801 , http://arxiv.org/abs/0801.1491 [ ] b. f. , 2000 , bull .
astr . soc .
india , 28 , 483 , http://arxiv.org/abs/astro-ph/0010185 [ ] b. f. , 2002 , in marcel grossmann ix conference on general relativity , eds v.g .
gurzadyan , r.t .
jantzen and r. ruffini , world scientific , singapore observational approaches to the topology of the universe .
p. 1937 , http://arxiv.org/abs/astro-ph/0010189 [ ] b. f. , bajtlik s. , biesiada m. , szaniewska a. , jurkiewicz h. , 2007 , a&a , 463 , 861 , http://arxiv.org/abs/astro-ph/0602159 [ ] b. f. , buliski z. , szaniewska a. , gaudin n. e. , 2008 , a&a , 486 , 55 , http://arxiv.org/abs/0801.0006 [ ] b. f. , lew b. , cechowska m. , marecki a. , bajtlik s. , 2004 , a&a , 423 , 821 , http://arxiv.org/abs/astro-ph/0402608 [ ] d. n. , bean r. , dor o. , nolta m. r. , bennett c. l. , hinshaw g. , jarosik n. , komatsu e. , page l. , peiris h. v. , verde l. , barnes c. , halpern m. , hill r. s. , kogut a. , et al .
2007 , apjsupp , 170 , 377 , http://arxiv.org/abs/astro-ph/0603449 [ ] d. n. , verde l. , peiris h. v. , komatsu e. , nolta m. r. , bennett c. l. , halpern m. , hinshaw g. , jarosik n. , kogut a. , limon m. , meyer s. s. , page l. , tucker g. s. , weiland j. l. , wollack e. , wright e. l. , 2003 , apjsupp , 148 , 175 , http://arxiv.org/abs/astro-ph/0302209 [ ] g. d. , 1998 , cqg , 15 , 2529 j .- p .
, lehoucq r. , luminet j .-
, 1999 , in proc .
of the xix@xmath131 texas meeting , paris 1418 december 1998 , eds .
e. aubourg , t. montmerle , j. paul and p. peter , article n@xmath132 04/25 new developments in the search for the topology of the universe , http://arxiv.org/abs/gr-qc/0005128 [ ] j .- p . ,
luminet j .-
, lehoucq r. , peter p. , 2002
, 23 , 277 , http://arxiv.org/abs/physics/0006039v1 [ ] o. , 2004 , submitted to found .
, http://arxiv.org/abs/gr-qc/0403111 [ ]
since we are interested in topology , suppose that neither twin has _
precise _ measuring rods or clocks , though both twins may have _ very approximate _ methods of measuring metric properties .
neither twin is aware that when she completes a loop of the universe , she may detect a time offset .
however , both twins have read history books and are aware of claims that `` space is multiply connected '' , so they attempt to verify this experimentally .
we can imagine that at event a , the two twins instantaneously create a highly stretchable physical link between them , such as a light - weight string or cord of negligible mass and extremely high strength against breaking .
as they move apart , the cord stretches , preserving the topological properties of space connectedness , while ignoring time .
we could alternatively imagine that one of the twins leaves behind a `` trail '' of some sort , e.g. like the vapour trail of an aircraft visible by human eye from the ground .
however , in this case , we have to be careful to avoid thinking of the particles in the vapour trail as being at rest in any particular frame , since otherwise we favour one of the two twins arbitrarily .
now consider the state of the cord `` at '' event b , when the two twins meet up again and join the two ends of the cord together .
b is a single space - time event .
even though the two twins disagree about space - time coordinates of the event , they agree that it is a single event and agree that they have physically joined the two ends of the cord . clearly the cord now forms a closed loop , of winding index one .
firstly , it is misleading because event b is just one space - time point among a whole set of space - time points where the particles constituting the cord are located , but `` the state of the cord '' is only of interest at this point of the discussion in the local neighbourhood of event b. it is difficult to avoid intuitively thinking of `` the state of the cord '' , i.e. of the state of a spatially extended object `` at '' the _ time _ of the event b , which is wrong , because it assumes simultaneity . a better way of thinking of the cord is presented below . secondly , it is misleading because the word `` at '' suggests a mono - valued time coordinate for a single event .
the reality is that just as in a multiply connected space , a single ( physical ) spatial point exists at many spatial points in the covering space , the situation is similar in a multiply connected space - time : a twin ( observer ) finds that a single space - time event exists at many ( in general ) non - simultaneous space - time points in the covering space - time .
one twin happens to be favoured and finds that the multiple space - time copies of a single event are simultaneous , but the other twin , moving with a different velocity , has a generator which is `` diagonal '' to her space - time axes . from the leftmoving twin s point of view , in fig .
[ f - spacetimeone ] , the cord can always be considered as a simultaneous object , i.e. a series of successive `` snapshots '' of the cord consist of horizontal line segments joining @xmath73 and @xmath33 , starting at a and sliding up to a final state of @xmath133 which for the leftmoving twin , is the state of the cord `` at '' the time of space - time event b. the rightmoving twin s point of view is similar , except that as can be seen in fig .
[ f - spacetimetwo ] , `` simultaneous '' snapshots of the cord , i.e. horizontal line segments joining @xmath73 and @xmath33 , starting at a and sliding upwards , have a problem when the right - hand end of the cord arrives at b@xmath18 . at this point , the left - hand end of the cord has not yet arrived at b@xmath17 according to the righmoving twin s notion of simultaneity .
however , b@xmath17 and b@xmath18 are a single physical space - time event : the space - time event of joining the two ends of the cord together occurs _ non - simultaneously _ according to the rightmoving twin .
this is intuitively difficult to imagine .
one way that the rightmoving twin could think about this could be that `` as '' the cord slides up from event a , it tilts in some way so that when / where the two ends of the cord are joined at b , the cord can be imagined as stretched along the line segment @xmath133 along a series of space - time events which are non - simultaneous .
this requires the use of some arbitrary affine parameter to define `` as '' for the rightmoving twin , i.e. a parameter that is something like time but is not physical time .
of course , the simplest option for this parameter is the leftmoving twin s time coordinate , but this does not make it any easier for the rightmoving twin to develop her intuition about it . if we consider the cord to be `` stretched '' rather than `` unrolled '' , so that the parts of the cord closest to each twin are ( nearly ) stationary with respect to that twin , and if the cord is created containing a mix of isotopes of radioactive elements in initially known proportions , then at event b , the proper times at the two ends of the cord will be measurable by measuring the isotopal mixes . in this case
, both twins will agree that not only the rightmoving twin has aged less , but also that the end of the cord `` held '' by the rightmoving twin is younger than the end of the cord `` held '' by the leftmoving twin .
so although the joined - up cord forms a single closed loop , its non - simultaneous nature is revealed by the discordant ages ( isotope mixes ) of the two ends that are joined up at b. this is dependent on the thought experimental setup requiring the cord to be locally ( nearly ) at rest with respect to each twin , i.e. the cord is `` stretched '' . with a different experimental setup for the behaviour of the cord ,
the aging of the cord occurs differently , and can be calculated by studying the worldlines of the particles composing the cord .
now that we have some way of seeing either twin s way of thinking of this closed loop from b@xmath17 to b@xmath18 , whose path through space ( projection of worldline to a spacelike hypersurface ) does this loop represent ?
each twin considers herself to be stationary , and the other twin to be moving `` rightwards '' or `` leftwards '' , respectively .
so each twin considers her own path through space to be a single point a path of zero winding index and that of the other twin to be a closed loop a path of winding index unity represented by the cord .
each twin considers the _ other _ twin to have pulled and/or stretched the cord so that eventually the two ends could be joined , not herself .
this is just an intuitive way of thinking of the projections described above : an observer in a spatially multiply connected , locally lorentz space - time , who is unable to make high - precision spatial and temporal measurements but can measure topological properties of space is unable to use the homotopy class of her spatial path ( projected worldline ) to detect the fact that she is either a favoured or a non - favoured observer .
of course , if we understand the full nature of this space - time , then we can note that the nature of `` the cord '' for at least one of the twins is a cross - section through space - time at non - constant time , as noted above .
this is necessary in order for event b to be a single event in space - time .
it can also to help to remember a key idea in resolving the `` pole in the barn paradox '' of simply connected minkowski space : neither a pole nor the door - to - door path of a barn is a one - dimensional object both are _ two - dimensional space - time objects_. the `` length '' of any such object depends on the choice of the reference frame , or in other words , the choice of spacelike cross - section .
yet another useful way of thinking of a pole is as a `` worldplane '' a collection of worldlines .
our ordinary intuition of a pole as a one - dimensional object is due to our implicit assumption of absolute simultaneity .
we can think of the cord stretched between the two twins and joined up at event b to be the entire filled - in area of the triangle @xmath134 in figs [ f - spacetimeone ] and [ f - spacetimetwo ] a two - dimensional space - time object . depending on various possible
thought experimental setups for creating / producing / stretching the cord , various sets of worldlines for the particles composing the cord are possible , but in each case would fill in this triangle . | in a multiply connected space , the two twins of the special relativity twin paradox move with constant relative speed and meet a second time without acceleration .
the twins situations appear to be symmetrical despite the need for one to be younger due to time dilation . here
, the suggestion that the apparent symmetry is broken by homotopy classes of the twins worldlines is reexamined using space - time diagrams .
it is found that each twin finds her own spatial path to have zero winding index and that of the other twin to have unity winding index , i.e. the twins worldlines relative homotopy classes are symmetrical .
although the twins apparent symmetry is in fact broken by the need for the non - favoured twin to non - simultaneously identify spatial domain boundaries , the non - favoured twin _ can not _ detect her disfavoured state if she only measures the homotopy classes of the two twins projected worldlines , contrary to what was previously suggested .
we also note that for the non - favoured twin , the fundamental domain can be chosen by identifying time boundaries ( with a spatial offset ) instead of space boundaries ( with a temporal offset ) .
reference systems time cosmology : theory |
[ [ numerical - calculations . ] ] numerical calculations .
+ + + + + + + + + + + + + + + + + + + + + + + starting from the master equations , we can formulate equations of motion for the elements of the density matrix @xmath81 with the first index being the cavity photon number and the second being the atomic excitation ( @xmath2 ) .
the explicit equations can be written as @xmath82\varrho_{n,1;m,1}-ig[\sqrt{n}\varrho_{n-1,2;m,1}-\sqrt{m}\varrho_{n,1;m-1,2}]\nonumber\\ & -i\omega_\text r[\varrho_{n,2;m,1}-\varrho_{n,1;m,2}]+\gamma\varrho_{n,2;m,2}+\kappa\sqrt{(n+1)(m+1)}\varrho_{n+1,1;m+1,1},\\ \dot\varrho_{n,1;m,2}=&[i(\delta_\text a-(n - m)\delta_\text c)-\tfrac{\gamma+\kappa(n+m)}{2}]\varrho_{n,1;m,2}-ig[\sqrt{n}\varrho_{n-1,2;m,2 } -\sqrt{m+1}\varrho_{n,1;m+1,1}]\nonumber\\ & - i\omega_\text r(\varrho_{n,2;m,2}-\varrho_{n,1;m,1})+\kappa\sqrt{(n+1)(m+1)}\varrho_{n+1,1;m+1,2},\\ \dot\varrho_{n,2;m,1}=&-[i(\delta_\text a+(n - m)\delta_\text c)+\tfrac{\gamma+\kappa(n+m)}{2}]\varrho_{n,2;m,1}-ig[\sqrt{n+1}\varrho_{n+1,1;m,1 } -\sqrt{m}\varrho_{n,2;m-1,2}]\nonumber\\ & - i\omega_\text r(\varrho_{n,1;m,1}-\varrho_{n,2;m,2})+\kappa\sqrt{(n+1)(m+1)}\varrho_{n+1,2;m+1,1},\\ \dot\varrho_{n,2;m,2}=&-[i\delta_\text c(n - m)+\gamma+\tfrac{\kappa}{2}(n+m)]\varrho_{n,2;m,2}-ig[\sqrt{n+1}\varrho_{n+1,1;m,2}-\sqrt{m+1}\varrho_{n,2;m+1,1}]\nonumber\\ & -i\omega_\text r(\varrho_{n,1;m,2}-\varrho_{n,2;m,1})+\kappa\sqrt{(n+1)(m+1)}\varrho_{n+1,2;m+1,2}.\end{aligned}\ ] ] we truncate the set of equations at a sufficiently large photon number @xmath83 . by varying @xmath83 ,
the validity of the calculations can be checked .
using @xmath84 , we can eliminate one element of the main diagonal , in our case , we chose @xmath85 .
this introduces an inhomogeneity into the equations , allowing us to calculate the steady state density matrix simply by inverting the matrix of coefficients and multiplying with the inhomogeneity .
finally , the expectation values of interest can be directly obtained .
the normally ordered variance of a light field is usually measured via balanced homodyne detection , see e.g. @xcite . in case of single atom fluorescence the complications stem from the small collection efficiency , substantially reducing the effect to be measured .
this problem can be resolved by correlation measurements @xcite , since in this case all correlation functions include the quantum efficiencies only as proportionality factors .
one may detect the intensity correlation function @xmath86 by the scheme in fig . 2 of ref .
@xcite . here
the indices @xmath87 refer to the different output channels of a beamsplitter .
following @xcite , we consider the difference between correlations at equal times and the corresponding steady state value , @xmath88 for ( stationary ) single atom fluorescence as the signal and a 50/50 beamsplitter we obtain @xmath89 with fl and lo denoting fluorescence and local oscillator , respectively , and @xmath90 is the intensity . for @xmath91 ,
the second term is dominant .
squeezing is detected if @xmath92 .
the effect of our optimization of squeezing , which yields a factor of two , is clearly observed by this method . | squeezing of atomic resonance fluorescence is shown to be optimized by a properly designed environment , which can be realized by a quasi - resonant cavity .
optimal squeezing is achieved if the atomic coherence is maximized , corresponding to a pure atomic quantum state .
the atomic - state purification is achieved by the backaction of the cavity field on the atom , which increases the atomic coherence and decreases the atomic excitation . for realistic cavities ,
the coupling of the atom to the cavity field yields a purity of the atomic state of more than 99% .
the fragility of squeezing against dephasing is substantially reduced in this scenario , which may be important for various applications . [
[ introduction . ] ] introduction .
+ + + + + + + + + + + + + a single atom and its coupling to the electromagnetic field is a system of fundamental interest for the understanding of the quantum phenomena of light and matter . it was predicted that a driven two - level atom emits antibunched light @xcite , which can not be described by the classical maxwell theory .
the first experimental demonstration of this effect was based on the resonance fluorescence of an atomic beam @xcite . in a related experiment
it was demonstrated that a sub - poissonian photon statistics may occur @xcite .
later on , photon antibunching could be demonstrated with single trapped ions @xcite .
squeezing was predicted to occur in the single - atom resonance fluorescence @xcite .
it can also be realized in the fluorescence of many atoms , via regular arrangement of the atoms @xcite , detection in the forward direction with respect to the pump - beam @xcite , and bistability in a strong driving field @xcite .
the latter two cases , could be experimentally demonstrated @xcite . squeezing in single - atom resonance fluorescence
could not be observed yet .
based on homodyne correlation measurements with a weak local oscillator , an efficient measurement technique was proposed @xcite , which is not limited by the collection efficiency of the fluorescence light .
its feasibility was demonstrated in resonance fluorescence experiments @xcite .
very recently , squeezed light has been observed in the output channel of a weakly driven high - q cavity , containing a single atom @xcite .
since an empty driven cavity can not produce squeezing , the observed squeezing of the cavity output field is clearly based on the coupling of the atom to the cavity field .
this work demonstrates , that squeezed light originated from a single atom is a prevailing subject .
the squeezed light in the output field of a cavity containing a single atom is undoubtedly a fundamental issue for its own . under general conditions , however , one can not expect that this is equivalent to the demonstration of squeezing in resonance fluorescence . hence a direct fluorescence measurement would still be of fundamental interest . the present letter deals with the optimization of atomic resonance fluorescence with respect to squeezing .
it is shown that squeezing becomes optimal , when the atomic coherence is perfectly controlled , which is equivalent to the purification of the atomic quantum state . atomic coherence control and state - purification
can be realized by a properly adjusted cavity , even for different choices of the mean atomic excitation .
this implies the possibility to control the fluorescence intensity to some extend .
similar to ordinary fluorescence measurements , the light is observed out the side of the cavity .
the setup under study substantially reduces the fragility of squeezing against dephasing .
this achievement is expected to be of great relevance for applications in miniaturized systems , such as quantum dots in semiconductor microcavities . [ [ optimal - squeezing . ] ] optimal squeezing .
+ + + + + + + + + + + + + + + + + + let us consider a general two - level atom ( tla ) in an arbitrary environment .
the atomic source field can be written as @xmath0 where @xmath1 ( @xmath2 ) is the flip operator of the atom with ground state @xmath3 and excited state @xmath4 . here
@xmath5 describes the atom - light - field coupling and the phase @xmath6 includes the phase of the driving laser .
the field is squeezed , if the normally ordered field variance becomes negative , corresponding to a noise reduction below the vacuum level .
optimizing with respect to the phase , we get for a tla @xmath7 see @xcite .
the full field , @xmath8 , is composed of the free field @xmath9 ( assumed to be in the vacuum state ) and the atomic source field @xmath10 .
the `` @xmath11 '' prescription denotes normal ordering .
the atomic expectation values in eq .
( [ eq.atvarsol1 ] ) are readily derived from the atomic density operator , @xmath12 .
the density matrix @xmath13 reads as @xmath14 from the positive semidefiniteness of quantum states it follows that @xmath15 , that is @xmath16 here we made use of the completeness relation of the tla , @xmath17 .
this inequality defines the maximal atomic coherence , @xmath18 , for any atomic excitation @xmath19 . using this result , for arbitrary atomic excitation the minimal variance follows for the maximal atomic coherence as @xmath20 the absolute minimum follows for @xmath21 , @xmath22 this value
can not be attained in free - space resonance fluorescence @xcite .
let us consider the structure of the atomic quantum state for optimal squeezing .
the purity of the state is given by @xmath23 compared with eq .
( [ eq.csi ] ) , purity of the atomic state is equivalent to maximal atomic coherence .
that is , optimal squeezing of the resonance fluorescence is achieved for a pure state of the atomic subsystem .
hence , to optimize squeezed emission from a single atom , the task is to find an environment for which the atomic subsystem is in a ( nearly ) pure state . note that the above results may be used to estimate the squeezing in the resonance fluorescence of a tla , without the need of homodyne detection . the minimal variance , optimized with respect to the phase
, @xmath24 can be expressed by the minimal variance for maximal atomic coherence ( first term ) , and the atomic - state purity ( occurring in the second term ) . the first contribution is solely determined by the atomic excitation , cf
eq .
( [ var - min ] ) .
the excitation can be observed by comparison with the saturation intensity .
simple methods also exist for detecting the purity of low - dimensional systems @xcite . in view of the relevance of purity measurements in the field of quantum information
, this may open an alternative possibility to infer squeezing in resonance fluorescence from independent measurements .
[ [ squeezing - in - cavity - assisted - fluorescence . ] ] squeezing in cavity - assisted fluorescence .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + in the following we will deal with a method for the practical realization of optimal squeezing by atomic - state purification .
we will show that a single - mode cavity may serve as an environment of the tla to optimize squeezing in resonance fluorescence .
in fact , our system may purify the atomic state for different values of the atomic excitation @xmath19 .
the cavity - assisted optimization of squeezing from a tla is feasible with current technology , cf . the required scheme in fig . [ fig.sys ] .
consider the tla being driven by a coherent light source of frequency @xmath25 , with the rabi frequency @xmath26 ( chosen to be real ) .
its spontaneous emission is characterized by the energy relaxation rate @xmath27 .
furthermore the atom is coupled to a single mode cavity of frequency @xmath28 , with coupling strength @xmath29 .
the cavity excitation is described by bosonic creation and annihilation operators , @xmath30 and @xmath31 , respectively .
the cavity emits light with a rate @xmath32 .
the hamiltonian for this system , in the frame rotating with @xmath25 and in the rotating - wave approximation , reads as @xmath33 where @xmath34 and @xmath35 .
the density operator @xmath36 of the full system obeys the von - neumann equation with lindblad terms for the different decay channels , @xmath37+\frac{\gamma}{2}\mathcal{l}_{\hat a_{12}}[\hat\varrho]+\frac{\kappa}{2}\mathcal{l}_{\hat a}[\hat\varrho],\\ \mathcal{l}_{\hat x}[\hat\varrho]&=2\hat x\hat\varrho\hat x^\dagger-\hat x^\dagger\hat x\hat\varrho-\hat\varrho\hat x^\dagger\hat x.\end{aligned}\ ] ] in general this system can not be solved analytically .
the numerical calculations are based on solving the steady - state equations of the density matrix , by truncation at a sufficiently large number of cavity photons , for details see @xcite .
let us now consider the cavity - assisted scenario for a strong driving field , @xmath38 , but detuning of the atom by @xmath39 .
hence the effective pumping is still limited , so that no saturation occurs , @xmath40 . a similar scenario has been studied in @xcite , but not in the context of squeezing . under such conditions the mollow - triplet is clearly visible in the spectrum , and the sidebands are well separated , at ( nearly ) the same frequencies as in free space : @xmath41 following the argumentation in @xcite , for a detuning of the cavity mode of @xmath42 its excitation depends on the ratio of the atomic decay to the cavity damping . if the cavity emission rate significantly exceeds the atomic decay , the excitation of the cavity is proportional to @xmath43 , hence the cavity is almost empty . in this case
the outcoupling of the cavity photons is fast compared to the emission rate of the atomic fluorescence out the side of the cavity .
it should be noted that we do not need a very good cavity , solely one with @xmath44 . despite the tiny excitation ,
the cavity effects become visible if a fluorescence sideband is close to a cavity resonance .
single - photon transitions in the cavity occur if the detuning , @xmath45 , is resonant to a sideband according to eq .
( [ eq.rfsidebands ] ) , @xmath46 at such a resonance , besides the fluorescence out of the side of the cavity , we also obtain enhanced emission of the cavity itself .
the excitation of the cavity increases slightly , which is consistent with the argumentation in @xcite .
nevertheless , the cavity emission increases strongly , as it scales with @xmath32 which is large compared to @xmath27 .
a similar situation was recently considered in @xcite , where steady - state inversion of a tla in a cavity was predicted . in our scenario
the cavity mode diverts a significant portion of the energy from the atom , which would otherwise contribute to the fluorescent light .
this yields a reduction of the atomic excitation . from this point of view
a not too good cavity is needed , in order to avoid a too strong backaction onto the atom . on the other hand , it has to be good enough to preserve the coherence of the atom , which would be lost in free - space fluorescence .
the coherent part of the atomic excitation ( cpae ) , @xmath47 , is determined by the interaction of the atom with both the cavity mode and the vacuum modes in free space , leading to the fluorescence . as a consequence of the above discussion , for the resonance condition ( [ eq.cavres ] ) ,
the cpae is increased due to the coupling to the cavity .
hence , we expect @xmath19 to decrease , while @xmath18 increases , yielding an obvious purification of the atomic state , according to eqs .
( [ eq.csi ] ) , ( [ eq.purity ] ) .
a critical condition in this setup is the requirement of @xmath44 and @xmath48 , so that the atom - cavity coupling significantly exceeds the atomic decay , @xmath49 .
experimental works , such as @xcite , suggest that large ratios of @xmath50 can be achieved in the optical frequency range . in this experiment a value of @xmath51
was realized , which will be used in our calculations reported below . combining the above arguments
, we can determine the regime of optimized squeezing in cavity - assisted resonance fluorescence .
we need strong pumping and atomic detuning , such that the atomic excitation in free space would slightly exceed @xmath52 .
for example , this would be the case for @xmath53 and @xmath54 , for details see @xcite .
the cavity , which obeys the condition @xmath55 , is tuned to a resonance according to eq .
( [ eq.cavres ] ) .
this yields a reduction of the atomic excitation to approximately @xmath52 , as needed for maximal squeezing according to eq .
( [ abs ] ) .
the cpae increases , resulting in a purification of the atomic state together with optimized squeezing , cf .
eq .
( [ eq.sq-pur ] ) .
the cavity emission rate @xmath32 is chosen to be intermediate , @xmath56 , since for large @xmath32 values the situation becomes close to that in free space . for smaller cavity losses
the atomic excitation is not sufficiently reduced , due to the backaction of the cavity photons on the atom . [
[ numerical - results . ] ] numerical results .
+ + + + + + + + + + + + + + + + + + in fig .
[ fig.sqmax ] we show the dependence of the atomic excitation and cpae on the atomic detuning @xmath57 .
the other parameters are chosen according to the requirements discussed above .
the atomic excitation @xmath58 behaves similar to the free - space scenario . for a wide parameter range ,
the cpae is rather close to its maximum value , cf .
eq .
( [ eq.csi ] ) , indicating significant purity of the atomic state .
the purity is given by @xmath59 .
the cavity excitation is very small , even at the cavity resonance according to eq .
( [ eq.cavres ] ) , occurring for the chosen parameters at @xmath60 .
this is consistent with the results of @xcite at the cavity resonance , which would predict @xmath61 for our parameters . :
atomic excitation @xmath19 ( blue , dashed line ) , cpae @xmath18 ( green crosses ) and maximum cpae @xmath62 ( red circles ) , cavity excitation @xmath63 ( violet , dashed - dotted line ) , normally ordered field variance of the fluorescence of a tla in the cavity ( light blue , solid curve ) , and in free space ( black , dotted curve ) .
the straight line ( black , solid ) marks the free space maximal squeezing of @xmath64 .
the system parameters are : @xmath65 , @xmath66 , @xmath67 , @xmath68.,width=302 ] around the cavity resonance we see , that the normally ordered field variance attains the value of @xmath69 .
this is more than 94% of the maximum possible squeezing of @xmath70 , which can be readily measured as discussed in @xcite .
the purity tr@xmath71 of the atomic subsystem ( not depicted in fig .
[ fig.sqmax ] ) , shows a clear maximum of about 99.5% at the cavity resonance .
the corresponding atomic excitation is @xmath72 . our numerical results for the obtained squeezing effect , the excitation and purity of the atomic state are in full agreement with the analytical relations of these quantities as given in eq .
( [ eq.sq-pur ] ) together with ( [ var - min ] ) . for larger values of @xmath50 , even more
than 99% of the absolute squeezing limit , @xmath73 , can be achieved .
this exceeds the free - space result substantially .
let us now consider , for our optimized environment , the sensitivity of squeezing with respect to dephasing , which is crucial in free - space fluorescence .
for this purpose we assume that , in addition to dephasing due to radiative damping , there is also radiationless dephasing described by the rate @xmath74 . for our considerations of the cavity - assisted atomic fluorescence we supplement the equations of motion ( [ eqmo ] ) with another lindblad - term , @xmath75+\frac{\gamma}{2}\mathcal{l}_{\hat a_{12}}[\hat\rho]+\frac{\gamma_\text d}{2}\mathcal{l}_{\hat a_{22}}[\hat \rho]+\frac{\kappa}{2}\mathcal{l}_{\hat a}[\hat\rho].\ ] ] the additional dephasing only increases the decay of the off - diagonal matrix elements of the density operator , that is , the atomic coherence and hence the cpae . in free space , squeezing does not occur anymore if the dephasing rate @xmath74 exceeds the energy relaxation @xmath27 . the atomic coherence needed for squeezing decays on a time scale which is faster than that of the emission of the fluorescence radiation . as stated above , in our setup , on a cavity resonance
the cpae is increased due to the increased transition amplitudes between atom and cavity .
the backaction of the cavity onto the atom preserves the coherence on a longer time scale , thus increasing the actual coherence in the steady state .
hence , the robustness against dephasing is enhanced . in this context
it is important that the cavity emission rate @xmath32 significantly exceeds the atomic decay rate @xmath27 .
this renders it possible to observe squeezing in cavity - assisted resonance fluorescence for surprisingly strong atomic dephasing . for different dephasings @xmath74 .
from bottom to top : @xmath76 .
the inset shows the dependence of the squeezing on @xmath74 at the cavity resonance , @xmath77 .
all other parameters are as in fig .
[ fig.sqmax].,width=302 ] in fig .
[ fig.deph ] , we show the normally ordered field variance for the same cavity as in fig
. [ fig.sqmax ] , for different values of @xmath74 .
the dependence of the minimal field variance on @xmath74 is shown in the inset . for @xmath78 ,
the minimal variance is below @xmath64 .
this is the maximal squeezing in free space , which could only be achieved for @xmath79 .
the squeezing in the cavity setup under study vanishes for @xmath80 .
this result may be of great importance for applications in light - emitting systems other than single atoms .
for example , significant dephasing is usually expected to occur in condensed matter systems .
a typical candidate could be quantum dots in semiconductor microstructures .
the scenario under study opens the possibility to create squeezing for relatively strong dephasing and intermediate cavity coupling , as it is typical for semiconductor microcavities @xcite .
[ [ conclusions . ] ] conclusions .
+ + + + + + + + + + + + we have studied the possibility of the optimization of squeezing in the resonance fluorescence of a two - level atom .
it is based on a special design of the environment of the atom , which can be achieved by a cavity with properly adjusted parameters .
the maximal squeezing of the light emitted from a coherently driven atom in an optimal environment is twice as strong as the maximal squeezing in free space experiments . more importantly , squeezing can be optimized for different atomic excitations , so that it is no longer limited to weakly driven systems .
the resulting possibility to increase both the fluorescence intensity and the squeezing will substantially widen the possibility of detection and application of squeezing in resonance fluorescence .
the realization of optimal squeezing is accompanied by the purification of the atomic state for different atomic excitations , which can be relevant for any other application which requires pure atomic states for different mean atomic excitations .
it is also important that the strong limitation of squeezing by dephasing can be substantially reduced in our system .
this may open important possibilities to control the emission of squeezed light from more complex , miniaturized systems , such as quantum dots in semiconductor microcavities .
[ [ acknowledgments . ] ] acknowledgments .
+ + + + + + + + + + + + + + + + the authors gratefully acknowledge stimulating conversation with p. r. rice and j. sperling .
this work was supported by the deutsche forschungsgemeinschaft through sfb 652 .
100 h. j. carmichael and d. f. walls , j. phys .
b * 9 * , l43 ( 1976 ) ; h. j. kimble and l. mandel , phys . rev .
a * 13 * , 2123 ( 1976 ) .
h. j. kimble , m. dagenais , and l. mandel , phys .
rev .
lett .
* 39 * , 691 ( 1977 ) .
r. short and l. mandel , phys .
rev .
lett . * 51 * , 384 ( 1983 ) .
f. diedrich and h. walther , phys .
rev .
lett .
* 58 * , 203 ( 1987 ) .
m. schubert , i. siemers , r. blatt , w. neuhauser , and p. e. toschek , phys .
rev .
lett . * 68 * , 3016 ( 1992 ) .
d. f. walls and p. zoller , phys .
rev .
lett . * 47 * , 709 ( 1981 ) .
w. vogel and d .-
g .
welsch , phys .
rev .
lett . * 54 * , 1802 ( 1985 ) .
a. heidmann and s. reynaud , j. physique * 46 * , 1937 ( 1985 ) .
m. d. reid and d. f. walls , phys .
rev .
a * 32 * , 396 ( 1985 ) .
z. h. lu , s. bali , and j. e. thomas , phys .
rev .
lett . * 81 * , 3635 ( 1998 ) .
m. g. raizen , l. a. orozco , min xiao , t. l. boyd , and h. j. kimble , phys .
rev .
lett .
* 59 * , 198 ( 1987 ) .
w. vogel , phys .
rev .
lett . * 67 * , 2450 ( 1991 ) ; phys .
rev .
a * 51 * , 4160 ( 1995 ) .
s. gerber , d. rotter , l. slodicka , j. eschner , h. j. carmichael , and r. blatt , phys .
rev .
lett . * 102 * , 183601 ( 2009 ) .
a. ourjoumtsev , a. kubanek , m. koch , c. sames , p.w.h .
pinske , g. rempe , and k. murr , nature * 474 * , 623 ( 2011 ) .
w. vogel and d .-
g .
welsch , _ quantum optics _ ( wiley - vch , berlin , 2006 ) .
r. filip , phys .
rev .
a * 65 * , 062320 ( 2002 ) .
a. k. ekert , c. m. alves , d. k. l. oi , m. horodecki , p. horodecki , and l. c. kwek , phys .
rev .
lett . * 88 * , 217901 ( 2002 ) .
h. nakazato , t. tanaka , k. yuasa , g. florio , and s. pascazio , phys .
rev .
a * 85 * , 042316 ( 2012 ) .
see our supplementary information at the end of the document for details on the numerical procedure and the proposal of a measurement technique .
h. freedhoff and t. quang , j. opt .
soc .
am .
b * 10 * , 1337 ( 1993 ) .
t. quang and h. freedhoff , opt .
comm . * 107 * , 480 ( 1994 ) .
s. hughes and h. j. carmichael , phys .
rev .
lett . * 107 * , 193601 ( 2011 ) . c. j. hood , m. s. chapman , t. w. lynn , and h. j. kimble , phys .
rev .
lett . * 80 * , 4157 ( 1998 ) .
g. khitrova , h. m. gibbs , m. kira , s. w. koch , and a. scherer , nat . phys . * 2 * , 81 ( 2006 ) .
a. laucht , n. hauke , j. m. villas - bas , f. hofbauer , g. bohm , m. kaniber , and j. j. finley , phys .
rev . lett . * 103 * , 087405 ( 2009 ) .
s. reitzenstein and a. forchel , in _ cavity qed in quantum dot - micropillar cavity systems _ , edited by p. michler , single semiconductor quantum dots , ( springer , berlin , 2009 ) .
a. dousse , j. suffczyski , a. beveratos , o. krebs , a. lamatre , i. sagnes , j. bloch , p. voisin , and p. senellart , nature * 466 * , 217 ( 2010 ) .
w. vogel , phys .
rev .
a * 51 * , 4160 ( 1995 ) . |
the purpose of this paper is to study the uniqueness and non - degeneracy of solutions to a nonlinear schdinger - type equation , arising from the minimization of the following energy functional @xmath2 under the mass constraint @xmath3 . here
@xmath4 is the positive part , @xmath5 is a 2-spinor that describes the quantum state of a nucleon ( a proton or a neutron ) , @xmath6 are the pauli matrices and @xmath7 .
the equation of interest is @xmath8 with @xmath9 the lagrange multiplier associated with the mass constraint .
this equation can as well be written in the form of a system of two coupled dirac - like equations @xmath10 indeed , the above model can formally be deduced from a relativistic model involving one dirac particle coupled with two auxiliary classical fields ( the so - called _ @xmath11 model _ ) , in a specific non - relativistic limit that will be described in detail below . in this limit
, the equations for the classical fields can be solved explicitly , leading to the nonlinear system and the corresponding nonlinear energy functional , expressed in terms of @xmath12 only .
the term @xmath13 in is the usual nonlinear schrdinger attraction which describes here the confinement of the nucleons . on the other hand
, the denominator @xmath14 can be interpreted as a mass depending on the state @xmath12 of the nucleon , and it describes a phenomenon of saturation in the system .
a high density @xmath15 generates a lower mass , which itself prevents from having a too high density . mathematically speaking ,
this term enforces the additional constraint @xmath16 , which is very important for the stability of the energy . without the @xmath12-dependent mass ,
the model is of course unstable and the energy functional is unbounded from below .
the mass term @xmath14 allows us to consider the minimization of the energy in space dimensions @xmath17 without any limitation on @xmath18 and @xmath19 , even if @xmath20 is the interesting physical case .
we remark that the upper bound @xmath21 on the particle density @xmath22 arises after an appropriate choice of units .
let us emphasize that , in the model presented above , spin is taken into account since @xmath12 takes values in @xmath23 . under the additional assumption that the state of the nucleon is an eigenfunction of the spin operator
, the energy must be restricted to functions of the special form @xmath24 leading to the simpler functional @xmath25 it is an open problem to show that minimizers of the original energy are necessarily of the special form . in principle , the spin symmetry could be broken . in this paper
we will however restrict ourselves to the simplified functional , which we study in any space dimension @xmath17 .
the corresponding euler - lagrange equation simplifies to @xmath26 to our knowledge , the above model was mathematically studied for the first time in @xcite , where esteban and the second author of this paper formally derived the equation from its relativistic counterpart , and then proved the existence of radial square integrable solutions of ( [ eq : nonlinear_nospin ] ) .
this result has then been generalized in @xcite , where the existence of infinitely many square - integrable excited states ( solutions with an arbitrary but finite number of sign changes ) was shown . in @xcite ,
esteban and the second author used a variational approach to prove the existence of minimizers for the spin energy , for a large range of values for the parameter @xmath27 .
the model is translation - invariant , hence uniqueness can not hold .
usual symmetrization techniques do not obviously apply due to the presence of the pauli matrices @xmath28 s but a natural conjecture is that all minimizers are of the form @xmath29 after a suitable space translation and a choice of spin orientation .
the approach of @xcite applies as well to the simplified no - spin model , and the proof works in any dimension .
the result in this case is the following .
[ thm : minimizers ] let @xmath17 and @xmath30 there exists a universal number @xmath31 such that @xmath32 for @xmath33 , @xmath34 and there is no minimizer ; @xmath32 for @xmath35 , @xmath36 and all the minimizing sequences are precompact in @xmath37 , up to translations .
there is at least one minimizer @xmath38 for the minimization problem @xmath39 and it can be chosen such that @xmath40 , after multiplication by an appropriate phase factor .
it solves the nonlinear equation for some @xmath41 .
the method used in @xcite to prove theorem [ thm : minimizers ] is based on lions concentration - compactness technique @xcite and the main difficulty was to deal with the denominator @xmath42 , for which special localization functions had to be introduced . because the energy depends linearly on the parameter @xmath27 , the function @xmath43 is concave non - increasing , which is another important fact used in the proof of @xcite .
the critical strength @xmath44 of the nonlinear attraction is the largest for which @xmath34 and it can simply be defined by @xmath45 it can easily be verified that @xmath46 in dimension @xmath47 , that @xmath48 is related to the gagliardo - nirenberg - sobolev constant in dimension @xmath49 , and that @xmath50 in higher dimensions .
estimates on @xmath44 have been provided in dimension @xmath20 in @xcite and similar bounds can be derived in higher dimensions by following the same method .
after the two works @xcite , it remained an open problem to show that minimizers are all radial and unique , up to a possible translation and multiplication by a phase factor .
the purpose of this paper is to answer this question .
our main result is the following .
[ thm : main ] the nonlinear equation has no non - trivial solution @xmath51 in @xmath52 when @xmath53 . for @xmath54 , the nonlinear equation admits a unique solution @xmath55 that tends to 0 at infinity , modulo translations and multiplication by a phase factor .
it is radial , decreasing , and non - degenerate .
this theorem is the equivalent of a celebrated similar result for the nonlinear schrdinger equation ( see , e.g. , ( * ? ? ?
b ) and @xcite for references ) .
our main contribution is the remark that the equation can be rewritten in terms of @xmath56 as a simpler nonlinear schrdinger equation @xmath57 applying a classical argument of mcleod @xcite ( as explained in ( * ? ? ?
b ) and in @xcite ) allows to prove the non - degeneracy and uniqueness in the radial case . that any solution of is necessarily radial decreasing then follows from the moving plane method @xcite .
the proof of theorem [ thm : main ] is provided in section [ sec : proof ] below .
let us remark that , since equation is invariant under multiplications by a phase factor , we can always suppose that a solution @xmath58 is real - valued .
hence , in ( * ? ? ?
* appendix a.1 ) it has been proved that any solution @xmath59 is such that @xmath60 a.e . in @xmath61 whenever @xmath62 . as a consequence , the change of variables @xmath63 makes sense
whenever @xmath62 . as an application of theorem [ thm : main ] , we are able to construct a branch of solutions of the underlying dirac equation , that converges to the non - relativistic solution @xmath38 in the limit , thereby justifying the formal arguments of @xcite .
we explain this now . we restrict ourselves to @xmath20 for simplicity ( but the results are similar in other dimensions ) .
we consider one relativistic nucleon in interaction with two scalar fields @xmath64 ( the _ @xmath0field _ ) and @xmath65 ( the _ @xmath1field _ ) .
as described for instance in @xcite , the corresponding equation is @xmath66 where @xmath67 are the dirac matrices and @xmath68 is now a 4-spinor .
the wavefunction @xmath69 should in principle be normalized in @xmath70 but , here , we think of fixing @xmath71 instead of imposing @xmath72 .
any non - trivial solution @xmath69 to also gives a normalized solution after an appropriate change of parameters .
in most physics papers , the equation for the @xmath0-field @xmath64 contains a nonlinear term as well ( for instance including vacuum polarization effects @xcite ) , @xmath73 but we restrict ourselves to the simpler linear case for convenience . the fields @xmath64 and @xmath65 are respectively focusing and defocusing , which can be seen from the different signs in the two klein - gordon equations . on the other hand
, they have very different effects , since @xmath64 modifies the mass @xmath74 in the same way for the upper and lower spinors , whereas @xmath65 is repulsive for the upper spinor and attractive for the lower spinor .
this statement is clarified when the dirac equation is written in terms of @xmath75 as @xmath76 we see that @xmath77 and @xmath78 respectively appear in the two equations . in our units
, the non - relativistic limit corresponds to @xmath79 , with all the masses being of the same order . on the contrary to atomic physics , in nuclear physics
the coupling constants @xmath80 and @xmath81 are very large , comparable to the masses .
it is therefore customary to work in a regime where @xmath82 and @xmath83 are fixed or , even , large . in the two klein gordon equations , the laplacian
can then be neglected in such a way that @xmath84 and hence @xmath85 @xmath86 as usual , in the non - relativistic regime , the lower spinor @xmath87 is of order @xmath88 .
simple effective equations will then be obtained in the limit . in order to better illustrate the regime of interest for the @xmath0@xmath1 model ,
let us first discuss the case of the @xmath0 model , in which @xmath89 and @xmath90 .
the equation then reduces to @xmath91 the interesting regime is then @xmath83 of order 1 , say @xmath92 fixed .
it can be proved that @xmath93 and the usual nls equation is recovered in the limit , after a simple scaling .
the precise result is the following .
[ thm : sigma ] let @xmath94 be positive constants . then for @xmath74 large enough , the equation admits a branch of solutions of the special form @xmath95 with @xmath96 in the limit @xmath97 , we have @xmath98 strongly in @xmath99 , where @xmath100 is the unique positive radial solution of @xmath101 functions of the form have the lowest possible total angular momentum ( * ? ? ?
. the theorem can be shown by following step by step the method of section [ sec : relativistic_limit ] , using the non - degeneracy of the nls ground state .
its proof will not be provided in this paper for shortness .
theorem [ thm : sigma ] is not satisfactory from a physical point of view .
indeed , the limit @xmath100 is considered physically unstable since the corresponding energy functional is unbounded from below in dimension 3 .
furthermore , in practice @xmath102 is very large and the corresponding @xmath100 is then very peaked at the origin . in real nuclei
, many forces are in action and they tend to compensate in order to avoid this collapse at @xmath103 .
it is therefore important to take the @xmath1 field into account . for the @xmath0@xmath1 model ,
the interesting regime is when the parameters @xmath104 and @xmath105 behave like @xmath74 , whereas @xmath106 stays bounded , which is the cancellation between the two scalar fields mentioned before .
even if @xmath104 diverges , the model still has a nice bounded limit @xmath38 , which is precisely the non - relativistic ground state studied in the previous section .
[ thm : dirac ] let @xmath107 be positive constants such that @xmath108 .
then for @xmath74 large enough , the equation admits a branch of solutions of the special form @xmath109 with @xmath110 in the limit @xmath97 , we have @xmath111 strongly in @xmath99 , where @xmath38 is the unique positive solution of with @xmath112 and @xmath113 .
we refer to @xcite , ( * ? ? ?
* sec . 3 ) and ( * ? ?
2.3 ) for a discussion of the validity of this regime for standard nucleons .
typical physical values for the parameters of the model are provided in ( * ? ? ?
* table 3.1 ) .
the proof of theorem [ thm : dirac ] is provided in section [ sec : relativistic_limit ] and it is based on the implicit function theorem . in other words , we see as a small perturbation of and we use the non - degeneracy of @xmath38 to construct a solution .
remark that , thanks to the non - degeneracy property proved in section [ sec : nondegeneracy_l2 ] , this argument gives also the local uniqueness of the solution to around @xmath58 , modulo translations and multiplication by a phase factor .
the exact same reasoning can be used for proving theorem [ thm : sigma ] .
a similar argument has for instance been used in @xcite .
we hope that our work will stimulate further research on this model .
* acknowledgement . *
the authors acknowledge financial support from the european research council ( fp7/2007 - 2013 grant agreement mniqs 258023 ) and the anr ( nonap 10 - 0101 ) of the french ministry of research .
moreover , the research of the second author was supported by the labex cempi ( anr-11-labx-0007 - 01 ) .
this section is devoted to the proof of theorem [ thm : main ] , which is split in several steps . in the next section
, we explicit the change of variable @xmath114 and combine it with symmetric rearrangement to deduce that the minimization problem @xmath39 can be restricted to radial decreasing functions .
this step is not necessary for our analysis but we mention it for completeness , as it gives a simpler existence proof than in @xcite .
then , in section [ sec : moving_planes ] , we use the moving plane method to conclude that positive solutions to the nonlinear equation are radial decreasing . section [ sec : uniqueness_radial ] is devoted to the uniqueness of radial solutions . finally , we prove the non - degeneracy of the linearized operator in the whole of @xmath115 ( modulo the trivial symmetries of the problem ) in section [ sec : nondegeneracy_l2 ] .
we recall that the energy functional is @xmath116 which we study on the subset of @xmath38 s in @xmath115 such that @xmath117 and @xmath118 since @xmath119 , it is clear that any such @xmath38 must be in @xmath37 .
it was proved in ( * ? ? ?
2.1 ) that it must also satisfy @xmath120 a.e .
the nonlinear term @xmath121 is then well defined and , since @xmath120 , we conclude that @xmath122 . by using rearrangement inequalities and the change of variable @xmath114 , we are able to prove that minimizers are always radial - decreasing .
this can be used to simplify the proof of theorem [ thm : minimizers ] of @xcite .
[ lem : rearrangement ] for every @xmath123 , the minimization problem @xmath39 can be restricted to radial non - increasing functions .
furthermore , any minimizer of @xmath39 , when it exists , is positive and radial - decreasing , after a possible translation and multiplication by a phase factor .
first we recall that @xmath124 a.e .
, see ( * ? ? ?
hence @xmath125 and the minimization problem can be restricted to functions satisfying @xmath40 , which we assume from now on .
let then @xmath126 be the schwarz rearrangement of @xmath38 . using
that @xmath127 we see that @xmath128 next , we have @xmath129 for all @xmath130 and , since @xmath131 is increasing , @xmath132 , by ( * ? ? ?
3 & lem . 7.17 ) . we conclude that @xmath133 and the minimization can be restricted to radial non - decreasing functions .
if @xmath38 is a ( possibly non - symmetric ) minimizer with @xmath40 , then @xmath126 is also a minimizer and we have @xmath134 . in general
, this does not imply that @xmath38 is itself radial - decreasing , but this will be proved using the nonlinear equation . denoting @xmath114 and @xmath135
, we see that @xmath136 must solve the euler - lagrange equation @xmath137 in particular , @xmath138 must be the first eigenvector of the schrdinger operator @xmath139 and therefore @xmath140 .
the real - analyticity of @xmath138 ( see , e.g. , @xcite ) combined with the equality @xmath141 now implies that @xmath142 , hence @xmath143 , after an appropriate space translation , by @xcite .
finally , if @xmath38 is an arbitrary minimizer , the equality @xmath144 implies @xmath145 by ( * ? ? ?
this concludes the proof of the lemma . in the previous section ,
we have shown using rearrangement inequalities that minimizers of @xmath146 are necessarily radial - decreasing . here
we use the moving plane method to prove that non - negative solutions of the equation are also all radial decreasing , which of course also implies lemma [ lem : rearrangement ] .
we recall that the nonlinear equation can be rewritten in terms of @xmath114 as @xmath147 we also remark that @xmath130 when @xmath148 and @xmath149 . for simplicity of notation
, we denote @xmath150 let @xmath151 and @xmath152 be a non - trivial solution of with @xmath153 .
then , @xmath138 is radial decreasing about some point in @xmath154 .
elliptic regularity gives that @xmath155 at infinity .
then the result follows immediately from the famous moving plane method . indeed , noticing that @xmath156 and @xmath157
, we may use ( * ? ? ?
we have proved that any solution to the equation must be radial - decreasing .
the next step consists in studying the uniqueness of radial solutions .
in this section , we study radial solutions to the equation , which then solve @xmath158 and we concentrate on showing the uniqueness of positive solutions such that @xmath159 when @xmath160 . in dimensions
@xmath161 , the condition @xmath162 is necessary to avoid a singularity at the origin . in dimension @xmath47 , the solution is known to be even about one point and , after a suitable translation we may always assume @xmath162 as well .
more precisely , to prove the existence of solutions in dimension @xmath47 , we use the fact that in this case the local energy @xmath163 is conserved along the trajectories
. however , in dimension @xmath161 , the energy @xmath164 defined by , decreases : @xmath165 the solutions @xmath166 to are parametrized by @xmath167 . using the same arguments as in the proof of (
2.6 ) and in particular the fact that the energy @xmath164 is non - increasing , we can easily show that a solution starting at @xmath168 stays bigger than @xmath169 and hence can not tend to @xmath103 at infinity .
moreover , note that the equation has the three stationary solutions @xmath170 , @xmath171 and @xmath172 .
hence @xmath173 is necessary .
the following is a reformulation of the result of @xcite that was expressed in terms of @xmath174 .
[ thm : shooting ] for @xmath175 , there is no non - trivial solution @xmath138 to , such that @xmath155 at infinity . for @xmath54 ,
there exists one positive solution @xmath176 to , such that @xmath177 at infinity .
it is decreasing , starts at @xmath178 and has the following behavior at infinity : @xmath179 for some @xmath180 . including the ground state @xmath176 , for @xmath181 and @xmath182 , in dimension @xmath20.[fig : portrait],width=340 ] the proof used in @xcite , which is presented for @xmath20 but can be generalized for all @xmath161 ,
is based on a shooting method consisting in increasing @xmath183 continuously starting from @xmath103 ( figure [ fig : portrait ] ) .
a byproduct of the proof is that all the other solutions @xmath166 with @xmath184 do not tend to @xmath103 at infinity .
hence it will remain to prove that the solutions @xmath166 with @xmath185 necessarily vanish at some point @xmath186 .
the explicit decay rate was not stated in @xcite , but it is a classical fact whose proof can for instance be read in @xcite . as remarked above , the result of lemma [ lem : rearrangement ] can be use to simplify this proof . for completeness
we quickly explain the non - existence part in theorem [ thm : shooting ] which is needed below and is itself taken from ( * ? ? ?
* prop 2.1 ) .
the idea is to use that the local energy is non - increasing .
this implies that any solution satisfying @xmath162 and @xmath187 at infinity must be such that @xmath188 hence @xmath189 is a necessary condition for the existence of @xmath138 .
moreover , we see that @xmath190 which is strictly above the stationary solution @xmath191 .
the main result of this section is the following [ thm : radial ] for @xmath54 and @xmath17 , the solution @xmath176 of theorem [ thm : shooting ] is the only non - trivial positive solution @xmath138 of such that @xmath187 at infinity . furthermore , @xmath176 is non - degenerate : the unique solution @xmath192 to @xmath193 diverges exponentially fast when @xmath160 . more precisely , if @xmath161 , @xmath192 satisfies @xmath194 and @xmath195 exponentially fast when @xmath160 . in dimension @xmath47
, the result follows immediately from the hamiltonian feature of the problem , based on the energy and the fact that @xmath196 is a non - degenerate critical point of @xmath164 , when @xmath41 .
in particular , the divergence of the solution @xmath192 to the linearized equation can be proved by computing the wronskian @xmath197 , using that @xmath198 .
we deduce that @xmath199 which can not converge to @xmath103 at infinity . in the following we assume @xmath161 .
there are many existing results dealing with the uniqueness ( and , often , the non - degeneracy as well ) of radial solutions to semi - linear equations of the type @xmath200 .
after the pioneering works on the nls nonlinearity @xcite , many authors introduced various conditions on the function @xmath201 that ensure uniqueness , see , e.g. @xcite . our particular function @xmath201 as defined in satisfies some of the assumptions required in these works .
for instance uniqueness can be directly obtained from ( * ? ? ?
1 ) in dimensions @xmath202 , by means of lemma [ lem : prop_f ] below . on the other hand ,
the non - degeneracy is sometimes not explicitly stated in those works , although often shown in the middle of the proof . for clarity
, we will therefore quickly explain the proof of the theorem , following the approach of mcleod in @xcite and its summary in ( * ? ? ?
b ) and @xcite . the main properties of the function @xmath201 that make everything works are summarized in the following [ lem : prop_f ]
let @xmath201 be defined as in on @xmath203 , with @xmath54
. then 1 .
@xmath201 is negative on @xmath204 and positive on @xmath205 with @xmath206 ; 2 .
@xmath207 is decreasing on @xmath205 ; 3 . for every @xmath208 ,
the function @xmath209 has exactly one root @xmath210 , at which we have @xmath211 .
the above properties of @xmath201 are somehow inherited from the nls case , since @xmath212 with @xmath213 . below we will not use the property _ ( 2 ) _ , but rather _ ( 3 ) _ ( which itself follows from _ ( 2 ) _ ) .
we however state _ ( 2 ) _ since the monotonicity of @xmath214 appears in many works , including for instance @xcite and @xcite .
the proof of lemma [ lem : prop_f ] will be provided at the end of the proof of the theorem .
the ` @xmath215 ' function appears as well in @xcite , where an additional assumption on the behavior of @xmath216 was required .
now , we assume that @xmath54 and we look at the solutions @xmath166 of with @xmath217 and @xmath162 , and we let @xmath183 vary in @xmath218 .
note that the function @xmath219 is smooth ( indeed real - analytic since @xmath201 is analytic ) .
following @xcite , we introduce the sets @xmath220 @xmath221 @xmath222 which form a partition of @xmath218 .
as we have recalled above , since the energy @xmath164 is decreasing along a solution , we have @xmath223 .
this was actually shown in @xcite , where the solution @xmath224 is constructed by looking at the supremum of @xmath225 .
in particular , @xmath226 . if @xmath227 we let for convenience @xmath228 . since @xmath229 is smooth , it can easily be proved that @xmath230 is open .
the same holds for @xmath225 , but the proof is more difficult .
the idea is that the points of @xmath230 are characterized by the fact that the trajectory in phase space crosses first the horizontal axis ( that is , @xmath166 vanishes before @xmath231 ) , whereas for @xmath232 it only crosses the vertical axis ( @xmath231 vanishes and @xmath166 does not ) , see figure [ fig : portrait ] .
[ lem : prop_s_- ] let @xmath233 .
then @xmath234 on @xmath235 , that is , @xmath166 vanishes before @xmath231 .
in particular , @xmath166 is strictly decreasing on @xmath235 .
the proof is again based on the monotonicity of the energy @xmath164 and it can for instance be read in ( * ? ? ?
the idea is the following .
we denote for simplicity @xmath236 and @xmath237 .
first , since @xmath238 , then we have from @xmath239 and hence @xmath240 for small @xmath241 . on the other hand @xmath242 ( since @xmath243 is the first root of @xmath236 and the latter can not have double zeroes ) . assuming that @xmath244 changes sign before @xmath243 implies that @xmath138 must have a local strict minimum at some point @xmath245 , at which @xmath246 .
then , since @xmath247 , there must be another later point @xmath248 at which @xmath249 .
however , we have @xmath250 a contradiction .
[ lem : prop_s_+ ] let @xmath232 .
then @xmath251 vanishes at least once and , for the first positive root @xmath252 of @xmath251 , we have @xmath253 .
the set @xmath225 is open .
the proof follows the presentation of @xcite and it goes as follows .
we denote for simplicity @xmath236 and @xmath237 . if @xmath254 , then @xmath255 and @xmath256 for all @xmath257 .
hence , let @xmath258 .
first we claim that @xmath244 must vanish .
otherwise @xmath138 is decreasing whenever @xmath259 and increasing if @xmath260 . in both cases
, @xmath138 has a positive limit @xmath261 at infinity .
using the equation , we see that @xmath262 , hence @xmath263 .
next , following @xcite , we look at @xmath264 which solves the equation @xmath265 at infinity we have @xmath266 , hence @xmath267 , which easily leads to a contradiction .
we conclude that @xmath244 vanishes and we denote by @xmath252 its first root .
next we distinguish two cases .
first , if @xmath268 , then @xmath269 and , by , @xmath256 for all @xmath241 and in particular @xmath253 .
second , if @xmath270 , then @xmath244 is negative for small @xmath271 ( due to the fact that @xmath272 ) . since @xmath273 ( otherwise @xmath274 and @xmath138 is constant )
, we see that @xmath138 must attain a local minimum at @xmath252 . from the equation , this yields @xmath275 and hence @xmath276 , which implies @xmath253 .
finally we prove that @xmath225 is open .
we already know that @xmath277 .
let then @xmath278 .
for @xmath279 in a neighborhood of @xmath183 , @xmath280 possesses a local minimum at @xmath281 at which @xmath282 . since @xmath283
, we get @xmath284 for all @xmath285 and some @xmath286 , and therefore @xmath287 .
let now @xmath288 be the unique solution to @xmath289 the main remark is that @xmath290 is the variation of @xmath138 with respect to the initial condition @xmath217 , which implies the following result assume that @xmath227 and that @xmath291 when @xmath292 .
then there exists @xmath286 such that @xmath293 and @xmath294 .
this is ( * ? ? ?
3(b ) ) and the argument goes as follows .
choose first @xmath295 such that @xmath296 on @xmath297 , and then @xmath298 such that @xmath299 for all @xmath300 .
finally , choose @xmath301 such that @xmath302 and @xmath303 .
for @xmath304 , we then have @xmath305 and @xmath306 .
the function @xmath307 is negative at @xmath308 with @xmath309 . if @xmath310 or if @xmath287 , then @xmath311 must tend to @xmath103 or become positive at some point , and therefore it must have a first local ( strict ) minimum at some point @xmath312 , with @xmath313 for all @xmath314 .
from the equation we can then write @xmath315 for some @xmath316 . here
@xmath317 because of our assumption that @xmath318 and @xmath319 by choice of @xmath320 .
now @xmath321 and @xmath322 , which is a contradiction .
the argument is the same for @xmath323 .
the lemma implies that any @xmath227 for which @xmath324 diverges to @xmath325 must be an isolated point .
now , if we can prove that @xmath326 for all @xmath227 then we would clearly be done . indeed , we know that @xmath225 and @xmath230 are open and they can only be separated by points in @xmath327 .
but the lemma says that points in @xmath327 can only serve as a transition between @xmath225 below and @xmath230 above .
therefore , there can be only one such transition , and we conclude that @xmath327 is reduced to one point .
so our goal will be to prove that all the points in @xmath327 have @xmath326 .
our argument will be based on the wronskian identity @xmath328 for various functions @xmath329 s .
a simple calculation shows that @xmath330 @xmath331 and @xmath332 these three test functions correspond respectively to variations of @xmath166 using multiplication by a constant , dilations and translations .
[ lem : v_has_one_root ] for every @xmath227 , the function @xmath288 vanishes exactly once . for simplicity
we denote again @xmath236 and @xmath333 .
assume on the contrary that @xmath334 for all @xmath335 ( if @xmath192 does not vanish it must be strictly positive since it can not have double zeroes ) .
using with @xmath336 , we find @xmath337 and , therefore , @xmath338 is decreasing and vanishes at @xmath339 , hence @xmath340 . since @xmath341 , we conclude that @xmath342 for @xmath343 and thus @xmath344 .
as we have said @xmath345 vanishes at @xmath339 and it is decreasing , hence @xmath346 for @xmath343 .
however @xmath347 decays exponentially at infinity and hence @xmath348 for @xmath271 large enough . using
, this proves that @xmath349 .
therefore @xmath350 diverges to @xmath325 exponentially at infinity , which contradicts the assumption that @xmath351 .
next , the proof that @xmath192 can only vanish once is the same as in @xcite
. indeed , start with @xmath352 at which the solution @xmath353 is stationary .
the function @xmath354 vanishes exactly once since @xmath355 decreases from @xmath356 to 0 .
taking @xmath357 and using that @xmath358 can not have double zeroes gives that @xmath192 can vanish at most once .
we are now able to show that @xmath192 and @xmath350 diverge to @xmath325 .
let @xmath227 . then @xmath359 and @xmath360 diverge to @xmath325 as @xmath160 . for simplicity
we denote again @xmath236 and @xmath333 .
let @xmath361 be the unique root of @xmath192 , at which we must have @xmath362 .
let now @xmath363 , which is chosen such that @xmath364 vanishes at the zero @xmath361 of @xmath192 . recall that @xmath240 and @xmath365 for all @xmath241 , by lemma [ lem : prop_s_- ] .
then we have from @xmath366 next we remark that the function @xmath367 vanishes both at @xmath339 and at @xmath368 .
therefore , its derivative must vanish at least once on @xmath369 , that is , @xmath370 vanishes before @xmath361 . since @xmath138 is strictly decreasing , and since @xmath371 vanishes only once by lemma [ lem : prop_f ] , we conclude that @xmath372 is negative for @xmath373 , hence @xmath367 is strictly decreasing after @xmath361 .
in particular , @xmath374 since @xmath375 and @xmath376 go to 0 exponentially at infinity , we conclude that @xmath377 must diverge .
more precisely , we have for @xmath271 large enough @xmath378 since @xmath379 and by .
hence @xmath380 and after integrating we get @xmath381 . as a consequence , @xmath192
diverge to @xmath325 exponentially . finally using that @xmath382 for large @xmath271 ( since @xmath383 )
, we conclude that @xmath350 diverges to @xmath325 exponentially as well .
as we have explained , the fact that all the points @xmath227 are non - degenerate with @xmath326 implies uniqueness , and concludes the proof of theorem [ thm : radial ] .
let @xmath213 be the nls polynomial , which is such that @xmath384 .
we have @xmath385 which is positive decreasing on @xmath386 .
noticing that @xmath387 we find @xmath388 this is negative for @xmath389 .
let now @xmath208 , and consider the function @xmath215 in . note that @xmath390 , and hence @xmath391 is positive for small @xmath392 . on the other hand , @xmath393 hence @xmath215 must vanish at least once on the interval @xmath218 .
next we remark that @xmath394 .
\label{eq : calcul_vanish_once}\end{aligned}\ ] ] note that , for @xmath395 , @xmath396 and @xmath397 from this we conclude that @xmath398 when @xmath399 hence @xmath215 can only vanish on @xmath205 . on this interval
@xmath214 is strictly decreasing , as we have shown before , hence @xmath215 can only have one root .
the linearized operators at our solution @xmath400 are defined by @xmath401 and @xmath402 more precisely , the linearized operator is @xmath403 .
the operator @xmath404 describes variations with respect to @xmath38 for real functions , whereas @xmath405 is related to the invariance of our problem under multiplication by a phase factor .
it is easy to verify that both @xmath404 and @xmath405 are self - adjoint operators on @xmath115 , with domain @xmath406 and form domain @xmath37 .
the main result of this section is [ thm : non - degenerate ] in @xmath115 , we have @xmath407 and @xmath408 .
the operators @xmath404 and @xmath405 both satisfy the perron - frobenius property that their first eigenvalue , when it exists , is necessarily non - degenerate with a positive eigenfunction .
this follows for instance from the fact that @xmath409 and from harnack s inequality ( * ? ? ?
6.4 , thm 5 ) which gives the strict positivity of eigenfunctions .
since @xmath410 and @xmath38 is positive , we deduce that it must be the first eigenfunction of @xmath405 , and that it is non - degenerate .
thus @xmath408 . next , in dimension @xmath47
, we know that @xmath411 and @xmath412 has a constant sign . hence , @xmath103 is the first eigenvalue of @xmath404 and it is non - degenerate which implies @xmath413 .
the argument for @xmath404 in dimension @xmath161 is slightly more complicated .
a lengthy but straightforward computation shows that @xmath414 since @xmath415 , the multiplier @xmath416 is bounded away from @xmath103 and we deduce that @xmath417 if and only if @xmath418 . hence @xmath419 if and only if @xmath420 .
the argument is now classical .
the operator @xmath421 commutes with space rotations and it may be written as a direct sum @xmath422 corresponding to the decomposition @xmath423 with @xmath424 the @xmath425th eigenspace of the laplace - beltrami operator on the sphere @xmath426 . in dimension @xmath20 ,
@xmath427 where @xmath428 are the usual spherical harmonics .
the formula for @xmath429 is @xmath430 with an appropriate boundary condition at @xmath339 ( neumann for @xmath431 and dirichlet for @xmath432 ) .
each @xmath429 has the perron - frobenius property .
since @xmath433 and @xmath434 has a constant sign , we conclude that @xmath103 is the first eigenvalue of @xmath435 and it is non - degenerate , thus @xmath436 .
next , for @xmath437 , we simply use that @xmath438 in the sense of quadratic forms , which shows that the first eigenvalue of @xmath429 must be positive and hence @xmath439 .
finally , for @xmath431 , the operator @xmath440 was studied in theorem [ thm : radial ] , where we proved that the unique solution to @xmath441 with @xmath442 diverges exponentially at infinity , hence can not be in @xmath443 .
we have therefore shown that @xmath444 since @xmath445 , this says that @xmath446 which concludes our proof of theorem [ thm : non - degenerate ] .
we want to prove the existence of a branch of solutions to the dirac equation @xmath447 which can be rewritten for @xmath448 as @xmath449 here the parameters are chosen as @xmath450 with @xmath451 fixed such that @xmath452 . it will be convenient to introduce the new fields @xmath453 then , imposing the special form @xmath454 with real - valued functions @xmath455 and @xmath456 , and using and , we obtain the following system @xmath457 which is equivalent to for functions of the above form . next , we consider the following rescaling @xmath458 and we find @xmath459 finally , denoting @xmath460 the perturbative parameter and recalling that @xmath461 we obtain @xmath462 with @xmath463 @xmath464 when @xmath465 , we obtain the system of equations @xmath466 which is equivalent to with @xmath174 and was studied in @xcite .
we introduce the map @xmath467 defined by @xmath468 here the spaces @xmath469 are the projections of the usual sobolev spaces @xmath470 to the sector of minimal total angular momentum ( they in particular contain a boundary condition at @xmath339 ) , whereas @xmath471 is the usual projection of @xmath472 to the subspace of radial functions . in what follows ,
we let @xmath473 , @xmath474 and @xmath475 .
solving the system is equivalent to solving @xmath476 .
we construct a branch of solutions , by means of an implicit function - type argument .
the first step is to prove that @xmath477 is a smooth operator from @xmath478 into @xmath479 .
the proof is tedious but elementary .
it relies on the fact that @xmath99 is an algebra and that all the functions appearing in the definition of @xmath477 are polynomials in the unknowns @xmath484 .
also , it uses that @xmath485 is an isomorphism from @xmath486 to @xmath487 .
indeed , this map is related to the restriction of the dirac operator @xmath488 to functions of the form . since @xmath489 ,
the operator is an isomorphism from @xmath99 to @xmath490 and the same holds in the radial subspaces @xmath486 and @xmath487 .
similarly , the operator @xmath491 is the fourier multiplier with the matrix @xmath492 and we claim that @xmath493 for all @xmath494 and all @xmath495 .
this estimate shows that the corresponding map is bounded on @xmath99 , as needed . in order to prove , we recall that @xmath496 changing @xmath497 into @xmath498 , it suffices to show that x@xmath499 for all @xmath500 , with @xmath501 the matrix @xmath502 has two real positive eigenvalues @xmath503 and @xmath504 with @xmath505 .
hence @xmath506 as a consequence , @xmath507 , for all @xmath494 and for all @xmath495 .
finally , the fact that @xmath482 is also continuous can be proved with the same arguments .
first we prove that @xmath513 is a one to one operator .
let @xmath514 a nontrivial solution to @xmath515 . then , since @xmath516 is bounded , @xmath517 and @xmath518 solves @xmath519 a calculation shows that the radial function @xmath329 solves @xmath520 where @xmath404 is the linearized operator defined in . since the restriction of @xmath404 to radial functions is invertible by theorems [ thm : radial ] and [ thm : non - degenerate ] , we conclude that @xmath521 and @xmath513 is one - to - one .
next , we observe that @xmath513 can be written as the sum of two linear operators @xmath522 as we have already said before , the upper part of the operator @xmath523 is a restriction of the dirac operator @xmath524 and it is an isomorphism from @xmath486 to @xmath487 , by definition of these spaces . on the other hand , @xmath525 is compact .
therefore @xmath513 is a one - to - one operator that can be written as a sum of an isomorphism and a compact perturbation and it is then an isomorphism . as a conclusion
, we can apply the implicit function theorem to find that there exists @xmath526 and a function @xmath527 such that @xmath528 and @xmath529 for @xmath530 .
this concludes the proof of theorem [ thm : dirac ] .
, _ symmetry of positive solutions of nonlinear elliptic equations in @xmath533 _ , in mathematical analysis and applications , part a , vol . 7 of adv . in math .
, academic press , new york - london , 1981 , pp .
369402 . height 2pt depth -1.6pt width 23pt , _ the concentration - compactness principle in the calculus of variations .
the locally compact case , part ii _ , ann .
h. poincar anal .
non linaire , 1 ( 1984 ) , pp .
223283 . , _ nonlinear dispersive equations _ , vol .
106 of cbms regional conference series in mathematics , published for the conference board of the mathematical sciences , washington , dc , 2006 .
local and global analysis . | we prove the uniqueness and non - degeneracy of positive solutions to a cubic nonlinear schrdinger ( nls ) type equation that describes nucleons .
the main difficulty stems from the fact that the mass depends on the solution itself . as an application
, we construct solutions to the @xmath0@xmath1 model , which consists of one dirac equation coupled to two klein - gordon equations ( one focusing and one defocusing ) . |
one of the main goals of percolation theory @xcite in recent decades has been to understand the geometric structure of percolation clusters .
considerable insight has been gained by decomposing the incipient infinite cluster into a _
backbone _ plus _ dangling bonds _ , and then further decomposing the backbone into _ blobs _ and _ red bonds _ @xcite . to define the backbone , one typically fixes two distant sites in the incipient infinite cluster , and defines the backbone to be all those occupied bonds in the cluster which belong to trails between the specified sites @xcite .
the remaining bonds in the cluster are considered dangling .
similar definitions apply when considering spanning clusters between two opposing sides of a finite box @xcite ; this is the so - called _ busbar _ geometry .
the bridges in the backbone constitute the red bonds , while the remaining bonds define the blobs . at criticality ,
the average size of the spanning cluster scales as @xmath2 , with @xmath3 the linear system size and @xmath4 the fractal dimension .
similarly , the size of the backbone scales as @xmath5 , and the number of red bonds as @xmath6 . while exact values for @xmath4 and @xmath7 are known @xcite ( see ) , this is not the case for @xmath8 . in @xcite
however , it was shown that @xmath9 coincides with the so - called monochromatic path - crossing exponent @xmath10 with @xmath11
. an exact characterization of @xmath12 in terms of a second - order partial differential equation with specific boundary conditions was given in @xcite , for which , unfortunately , no explicit solution is currently known .
the exponent @xmath12 was estimated in @xcite using transfer matrices , and in @xcite by studying a suitable correlation function via monte carlo simulations on the torus . in this paper , we consider a natural partition of the edges of a percolation configuration , and study the fractal dimensions of the resulting clusters .
specifically , we classify all occupied bonds in a given configuration into three types : branches , junctions and nonbridges .
a bridge is a _ branch _ if and only if at least one of the two clusters produced by its deletion is a tree .
junctions are those bridges which are not branches .
deleting branches from percolation configurations produces _ leaf - free _ configurations , and further deleting junctions from leaf - free configurations generates bridge - free configurations .
these definitions are illustrated in fig .
[ fig : diagram ] .
it is often useful to map a bond configuration to its corresponding baxter - kelland - wu ( bkw ) @xcite loop configuration , as illustrated in fig .
[ fig : diagram ] .
the loop configurations are drawn on the medial graph @xcite , the vertices of which correspond to the edges of the original graph .
the medial graph of the square lattice is again a square lattice , rotated @xmath13 .
each unoccupied edge of the original lattice is crossed by precisely two loop arcs , while occupied edges are crossed by none .
the continuum limits of such loops are of central interest in studies of scharmm lwner evolution ( sle ) @xcite . at the critical point , the mean length of the largest loop scales as @xmath14 , with @xmath15 the hull fractal dimension .
a related concept is the accessible external perimeter @xcite .
this can be defined as the set of sites that have non - zero probability of being visited by a random walker which is initially far from a percolating cluster .
the size of the accessible external perimeter scales as @xmath16 with @xmath17 . in two dimensions , coulomb - gas arguments @xcite
predict the following exact expressions for @xmath4 , @xmath7 , @xmath15 and @xmath18 @xmath19 where for percolation the coulomb - gas coupling @xmath20 . ] .
we note that the magnetic exponent @xmath21 , the two - arm exponent @xcite satisfies @xmath22 , and that for percolation the thermal exponent @xmath23 @xcite .
the two - arm exponent gives the asymptotic decay @xmath24 of the probability that at least two spanning clusters join inner and outer annuli ( of radii o(1 ) and @xmath3 respectively ) in the plane .
we also note that @xmath18 and @xmath15 are related by the duality transformation @xmath25 @xcite .
the most precise numerical estimate for @xmath8 currently known is @xmath26 @xcite .
we study critical bond percolation on the torus @xmath27 , and show that as a consequence of self - duality the density of bridges and nonbridges both tend to 1/4 as @xmath28 .
using monte carlo simulations , we observe that despite the fact that around 43% of all occupied edges are branches , the fractal dimension of the leaf - free clusters is simply @xmath4 , while their hulls are governed by @xmath15 .
by contrast , the fractal dimension of the bridge - free configurations is @xmath8 , and that of their hulls is @xmath18 .
[ fig : configuration ] shows a typical realization of the largest cluster in critical square - lattice bond percolation , showing the three different types of bond present . in more detail ,
our main findings are summarized as follows . 1 .
the leading finite - size correction to the density of nonbridges scales with exponent @xmath29 , consistent with @xmath30 .
it follows that the probability that a given edge is not a bridge but has both its loop arcs in the same loop decays like @xmath24 as @xmath28 .
the leading finite - size correction to the density of junctions also scales with exponent @xmath29 , while the density of branches is almost independent of system size .
the fractal dimension of leaf - free clusters is @xmath31 , consistent with @xmath32 for percolation clusters .
the hull fractal dimension for leaf - free configurations is @xmath33 , consistent with @xmath34 .
the fractal dimension for bridge - free clusters is consistent with @xmath8 , and we provide the improved estimate @xmath35 .
the hull fractal dimension for bridge - free configurations is @xmath36 , consistent with @xmath37 .
the remainder of this paper is organized as follows .
section [ model_algorithm_quantities ] introduces the model , algorithm and sampled quantities .
numerical results are summarized and analyzed in section [ results ] .
a brief discussion is given in section [ discussion ] .
we study critical bond percolation on the @xmath38 square lattice with periodic boundary conditions , with linear system sizes @xmath39 , 16 , 24 , 32 , 48 , 64 , 96 , 128 , 256 , 512 , 1024 , 2048 , and 4096 . to generate a bond configuration
, we independently visit each edge on the lattice and randomly place a bond with probability @xmath40 . for each system size
, we produced at least @xmath41 independent samples ; for each @xmath42 we produced more than @xmath43 independent samples .
a _ leaf _ in a percolation configuration is a site which is adjacent to precisely one occupied bond . given a percolation configuration we generate the corresponding _ leaf - free _ configuration via the following iterative procedure , often referred to as _
burning_. for each leaf , we delete its adjacent bond . if this procedure generates new leaves , we repeat it until no leaves remain .
the bonds which are deleted during this iterative process are precisely the branches defined in section [ introduction ] .
the bridges in the leaf - free configurations are the junctions . deleting the junctions from the leaf - free configurations then produces bridge - free configurations .
the algorithm we used to efficiently identify junctions in leaf - free configurations is described in sec .
[ algorithm ] . given an arbitrary graph @xmath44
, the bridges can be identified in @xmath45 time @xcite . rather than applying such graph algorithms to identify the junctions in our leaf - free configurations however , we took advantage of the associated loop configurations .
these loop configurations were also used to measure the observable @xmath46 , defined in section [ measured quantities ] .
consider an edge @xmath47 which is occupied in the leaf - free configuration , and denote the leaf - free cluster to which it belongs by @xmath48 . in the planar case ,
it is clear that @xmath47 will be a bridge iff the two loop segments associated with it belong to the same loop .
more generally , the same observation holds on the torus provided @xmath48 does not simultaneously wind in both the @xmath49 and @xmath50 directions .
if @xmath48 does simultaneously wind in both the @xmath49 and @xmath50 directions , loop arguments may still be used , however the situation is more involved .
it clearly remains true that if the two loop segments associated with @xmath47 belong to different loops , then @xmath47 is a nonbridge .
suppose instead that the two loop segments associated with @xmath47 belong to the same loop , which we denote by @xmath51 .
deleting @xmath47 breaks @xmath51 into two smaller loops , @xmath52 and @xmath53 .
for each such loop , we let @xmath54 and @xmath55 denote the winding numbers in the @xmath49 and @xmath50 directions , respectively , and we define @xmath56 . as we explain below , the following two statements hold : 1 .
[ bridge test ] if @xmath57 or @xmath58 , then @xmath47 is a bridge .
[ nonbridge test ] if @xmath59 and @xmath60 , then @xmath47 is a nonbridge . as an illustration , in fig .
[ fig : diagram ] edge 1 is a junction while edge 2 is a nonbridge , despite both of them being bounded by the same loop .
edge 1 can be correctly classified using statement [ bridge test ] , while edge 2 can be correctly classified using statement [ nonbridge test ] . by making use of these observations ,
all but very few edges in the leaf - free clusters can be classified as bridges / nonbridges .
we note that in our implementation of the above algorithm , the required @xmath61 values can be immediately determined from the stored loop configuration without further computational effort . for the small number of edges to which neither of the above two statements apply ,
we simply delete the edge and perform a connectivity check using simultaneous breadth - first search .
this takes @xmath62 time per edge tested @xcite .
we now justify statement [ bridge test ] . in this case , the loop @xmath52 is contained in a simply - connected region on the surface of the torus .
the cluster contained within the loop @xmath52 is therefore disconnected from the remainder of the lattice , implying that @xmath47 is a bridge .
edge 1 in fig .
[ fig : diagram ] provides an illustration .
finally , we justify statement [ nonbridge test ] . in this case ,
@xmath52 and @xmath53 either both wind in the @xmath49 direction , or both in the @xmath50 direction ( one winds in the positive sense , the other in the negative sense ) .
suppose they wind in the @xmath50 direction .
it then follows from the definition of the bkw loops that there can be no @xmath49-windings in the cluster @xmath63 . by assumption
however , @xmath48 does contain an @xmath49-winding , so it must be the case that @xmath47 belongs to a winding cycle in @xmath48 that winds in the @xmath49 direction .
the edge @xmath47 is therefore not a bridge .
edge 2 in fig .
[ fig : diagram ] provides an illustration . from our simulations , we estimated the following quantities . 1 .
the mean density of branches @xmath64 , junctions @xmath65 , and nonbridges @xmath66 .
2 . the mean size of the largest cluster @xmath67 3 . the mean size of the largest leaf - free cluster @xmath68 4 .
the mean size of the largest bridge - free cluster @xmath69 5 .
the mean length of the largest loop , @xmath46 , for the loop configuration associated with leaf - free configurations 6 .
the mean length of the largest loop , @xmath70 , for the loop configuration associated with bridge - free configurations we note that fewer samples were generated for @xmath67 and @xmath70 than for other the quantities .
in sections [ bond density ] , [ fractal dimension of clusters ] , [ fractal dimension of loops ] , we discuss least - squares fits for @xmath64 , @xmath65 , @xmath66 and @xmath68 , @xmath69 , @xmath46 , @xmath70 .
the results are presented in tables [ tab : bond_density ] , [ tab : fd_clusters ] and [ tab : fd_loops ] . in section [ fitting methodology ] , we first make some comments on the anstze and methodology used
. let @xmath71 ( @xmath72 ) denote the mean density of occupied edges whose two associated loop segments belong to the same ( distinct ) loop(s ) . from lemma [ loop lemma ] in appendix [ loop lemma appendix ] , we know that for @xmath40 bond percolation on @xmath27 we have @xmath73 for all @xmath3 . in the plane however , an edge is a bridge iff the two associated loop segments belong to the same loop .
we therefore expect that both @xmath66 and @xmath74 should converge to 1/4 as @xmath28 .
furthermore , there is a natural interpretation of the quantity @xmath75 .
as noted in section [ algorithm ] , if the two loop segments associated with an edge belong to different loops , then that edge can not be a bridge .
this implies that @xmath75 is equal to the probability of the event that `` a given edge is not a bridge but has both its loop arcs in the same loop '' .
let us denote this event by @xmath76 .
studying the finite - size behaviour of @xmath66 will therefore allow us to study the scaling of @xmath77 .
since @xmath78 , it follows that @xmath79 is also equal to @xmath77 .
armed with the above observations , we fit our monte carlo data for the densities @xmath65 , @xmath64 and @xmath66 to the finite - size scaling ansatz @xmath80 we note that since @xmath81 for all @xmath3 , the finite - size corrections of @xmath74 should be equal in magnitude and opposite in sign to the finite - size corrections of @xmath66 . since @xmath82
, the latter should be positive and the former negative .
finally , we note that the event @xmath76 essentially characterizes edges which _ would _ be bridges in the plane , but which are prevented from being bridges on the torus by windings . by construction
, branches always have at least one end attached to a tree , suggesting that they can not be _ trapped _ in winding cycles in this way .
this would suggest that it should be @xmath65 that contributes the leading correction of @xmath74 away from its limiting value of 1/4 .
the observables @xmath68 , @xmath69 , @xmath46 , @xmath70 are expected to display non - trivial critical scaling , and we fit them to the finite - size scaling ansatz @xmath83 where @xmath84 denotes the appropriate fractal dimension . as a precaution against correction - to - scaling terms that we failed to include in the fit ansatz , we imposed a lower cutoff @xmath85 on the data points admitted in the fit , and we systematically studied the effect on the @xmath86 value of increasing @xmath87 .
generally , the preferred fit for any given ansatz corresponds to the smallest @xmath88 for which the goodness of fit is reasonable and for which subsequent increases in @xmath88 do not cause the @xmath86 value to drop by vastly more than one unit per degree of freedom . in practice , by `` reasonable '' we mean that @xmath89 , where df is the number of degrees of freedom . in all the fits
reported below we fixed @xmath90 , which corresponds to the exact value of the sub - leading thermal exponent @xcite . leaving @xmath91 free in the fits of @xmath65 and @xmath66 we estimate @xmath92 .
we therefore conjecture that @xmath93 , which we note is precisely equal to the two - arm exponent @xmath94 .
we comment on this observation further in section [ discussion ] . for @xmath64 by contrast
, we were unable to obtain stable fits with @xmath91 free .
fixing @xmath95 , the resulting fits produce estimates of @xmath96 that are consistent with zero .
in fact , we find @xmath64 is consistent with 0.21405018 for all @xmath97 . this weak finite - size dependence of @xmath64 is in good agreement with the arguments presented in section [ fitting methodology ] .
all the fits for @xmath64 , @xmath65 and @xmath66 gave estimates of @xmath98 consistent with zero .
we therefore set @xmath99 identically in the fits reported in table [ tab : bond_density ] . from the fits , we estimate @xmath100 , @xmath101 and @xmath102 .
we note that @xmath103 within error bars , as expected .
the fit details are summarized in table [ tab : bond_density ] .
we also note that the estimates of @xmath96 for @xmath65 and @xmath66 are equal in magnitude and opposite in sign , which is as expected given that @xmath96 is consistent with zero for @xmath64 . in fig .
[ fig : bond_density ] , we plot @xmath64 , @xmath65 and @xmath66 versus @xmath104 .
the plot clearly demonstrates that the leading finite - size corrections for @xmath65 and @xmath66 are governed by exponent @xmath94 , while essentially no finite - size dependence can be observed for @xmath64 .
.fit results for @xmath64 , @xmath65 , and @xmath66 . [ cols="<,<,<,<,<",options="header " , ]
we have studied the geometric structure of percolation on the torus , by considering a partition of the edges into three natural classes . on the square lattice , we have found that leaf - free configurations have the same fractal dimension and hull dimension as standard percolation configurations , while bridge - free configurations have cluster and hull fractal dimensions consistent with the backbone and external perimeter dimensions , respectively .
in addition to the results discussed above , we have extended our study of leaf - free configurations to site percolation on the triangular lattice and bond percolation on the simple - cubic lattice , the critical points of which are respectively 1/2 and 0.24881182(10 ) @xcite .
we find numerically that the fractal dimensions of leaf - free clusters for these two models are respectively 1.8957(2 ) and 2.5227(6 ) , both of which are again consistent with the known results @xmath105 and @xmath106 @xcite for @xmath4 . in both cases ,
our data show that the density of branches is again only very weakly dependent on the system size .
it would also be of interest to study the bridge - free configurations on these lattices .
in addition to investigating the fractal dimensions for cluster size , and in the triangular case also the hull length , it would be of interest to determine whether the leading finite - size correction to @xmath66 is again governed by the two - arm exponent .
the two - arm exponent is usually defined by considering the probability of having multiple spanning clusters joining inner and outer annuli in the plane . as noted in section [ fitting methodology ]
however , our results show that for percolation , the two - arm exponent also governs the probability of a rather natural geometric event on the torus : the event that a given edge is not a bridge but has both its loop arcs in the same loop .
this provides an interesting alternative interpretation of the two - arm exponent in terms of toroidal geometry .
let us refer to an edge that is not a bridge but has both its loop arcs in the same loop as a _
pseudobridge_. we note that an alternative interpretation of the observation that @xmath107 is that the number of pseudobridges @xmath108 scales as @xmath6 . a natural question to ask is to what extent the above results carry over to the general setting of the fortuin - kasteleyn random - cluster model .
consider the case of two dimensions once more . in that case , we know that if we fix the edge weight to its critical value and take @xmath109 we obtain the uniform spanning trees ( ust ) model .
for this model all edges are branches , and so the leaf - free configurations , which are therefore empty , certainly do not scale in the same way as ust configurations . despite this observation , preliminary simulations and the chayes - machta algorithm @xcite for @xmath110 . ]
performed on the toroidal square lattice at @xmath111 , @xmath112 , @xmath113 , @xmath114 , @xmath115 , @xmath116 and @xmath117 suggest that , for all @xmath118 $ ] , the leaf - free configurations have the same fractal dimension and hull dimension as the corresponding standard random cluster configurations . in the context of the random cluster model ,
the behaviour of the leaf - free configurations for the ust model therefore presumably arises via amplitudes which vanish at @xmath119 .
in addition , these preliminary simulations suggest that the number of pseudobridges in fact scales as @xmath6 for the critical random cluster model at any @xmath118 $ ]
. it would also be of interest to determine whether the fractal dimensions of cluster size and hull length for bridge - free random cluster configurations again coincide with @xmath8 and @xmath18 when @xmath120 .
the authors wish to thank bob ziff for several useful comments / suggestions , and t.g . wishes to thank nick wormald for fruitful discussions relating to the density of bridges .
this work is supported by the national nature science foundation of china under grant no .
91024026 and 11275185 , and the chinese academy of sciences .
it was also supported under the australian research council s discovery projects funding scheme ( project number dp110101141 ) , and t.g .
is the recipient of an australian research council future fellowship ( project number ft100100494 ) .
the simulations were carried out in part on nyu s its cluster , which is partly supported by nsf grant no .
in addition , this research was undertaken with the assistance of resources provided at the nci national facility through the national computational merit allocation scheme supported by the australian government . j.f.w and y.j.d acknowledge the specialized research fund for the doctoral program of higher education under grant no .
y.j.d also acknowledge the fundamental research funds for the central universities under grant no . 2340000034 .
let @xmath121 ( @xmath122 ) denote the fraction of occupied edges whose two associated loop segments belong to the same ( distinct ) loop(s ) .
let @xmath124 denote the number of edges in @xmath125 .
since @xmath126 is a cellularly - embedded graph @xcite , it has a well - defined geometric dual @xmath127 and medial graph @xmath128 . for any @xmath129
we denote its dual by @xmath130 . for @xmath129 ,
let @xmath131 be the event that the two loop segments associated with @xmath47 both belong to the same loop , and let @xmath132 be the event that they belong to distinct loops .
the key observation is that for any @xmath133 we have @xmath134 to see this , first note that the number of terms on either side of is @xmath135 , and that each term is either 0 or 1 .
then note that there is a bijection between the terms on the left- and right - hand sides such that the term on the left - hand side is 1 iff the term on the right - hand side is 1 , as we now describe .
let @xmath136 with @xmath137 , and let @xmath138 denote the _ dual _ configuration : include @xmath139 in @xmath138 iff @xmath140
. with the term on the left - hand side corresponding to @xmath141 , associate the term @xmath142 appearing on the right - hand side .
this is clearly a 1 - 1 correspondence .
let @xmath143 denote the loop configuration corresponding to @xmath144 . by construction , @xmath145 .
the loop configuration @xmath146 differs from @xmath143 only in that the loop arcs cross @xmath139 in @xmath143 but cross @xmath47 in @xmath146 .
if @xmath147 , then it follows that @xmath148 .
the converse holds by duality , and so is established .
see fig .
[ fig : loops diagram ] for an illustration . | we investigate the geometric properties of percolation clusters , by studying square - lattice bond percolation on the torus .
we show that the density of bridges and nonbridges both tend to 1/4 for large system sizes . using monte carlo simulations , we study the probability that a given edge is not a bridge but has both its loop arcs in the same loop , and find that it is governed by the two - arm exponent .
we then classify bridges into two types : branches and junctions .
a bridge is a _ branch _ iff at least one of the two clusters produced by its deletion is a tree .
starting from a percolation configuration and deleting the branches results in a _ leaf - free _ configuration , while deleting all bridges produces a bridge - free configuration . although branches account for @xmath0 of all occupied bonds , we find that the fractal dimensions of the cluster size and hull length of leaf - free configurations are consistent with those for standard percolation configurations . by contrast , we find that the fractal dimensions of the cluster size and hull length of bridge - free configurations are respectively given by the backbone and external perimeter dimensions .
we estimate the backbone fractal dimension to be @xmath1 . |
throughout this article , we will let @xmath8 , @xmath9 , and @xmath10 denote the set of positive integers , the set of nonnegative integers , and the set of prime numbers , respectively . the lowercase letter @xmath11 will always denote
a prime number , and @xmath12 will denote the exponent of @xmath11 in the prime factorization of a positive integer @xmath13 .
furthermore , for any nonzero complex number @xmath14 , we let @xmath15 denote the principal argument of @xmath14 with the convention that @xmath16 . for any complex number @xmath0 ,
the divisor function @xmath17 is the arithmetic function defined by @xmath2 for all @xmath3 .
the function @xmath5 is a multiplicative arithmetic function that satisfies @xmath18 for all primes @xmath11 and positive integers @xmath19 . of course , if @xmath20 , then we may write @xmath21 .
divisor functions are some of the most important functions in number theory ; their appearances in various identities and applications are so numerous that we will not even attempt to list them .
however , divisor functions other than @xmath22 , and @xmath23 are rarely studied .
recently , the author @xcite has studied the ranges of the functions @xmath5 for real @xmath0 and has shown that there exists a constant @xmath24 such that if @xmath25 , then the range of the function @xmath26 is dense in the interval @xmath27
if and only if @xmath28 . for any complex @xmath0 , we will let @xmath29 be the range of the function @xmath5 . in this article , we will study the basic topological properties of the sets @xmath6 for various complex numbers @xmath0 . more specifically , we will direct the bulk of our attention toward answering the following questions : 1 .
for which complex @xmath0 is @xmath6 bounded ?
for which complex @xmath0 does @xmath6 have isolated points ? 3 .
what can we tell about the closure @xmath30 of the set @xmath6 for given values of @xmath0 ?
in particular , what are the values of @xmath0 for which @xmath6 is dense in @xmath7 ?
we begin with a number of useful lemmas .
some of these lemmas not only aid in the proofs of later theorems , but also provide some basic yet interesting information that serves as a nice introduction to the sets @xmath6 .
henceforth , @xmath0 will denote a complex number with real part @xmath31 and imaginary part @xmath32 .
[ lem1.1 ] for any @xmath33 , @xmath34 .
we have @xmath35 lemma [ lem1.1 ] tells us that @xmath36 is simply the reflection of the set @xmath6 about the real axis . in many situations ,
this simple but useful lemma allows us to restrict our attention to complex numbers @xmath0 in the upper half plane and then use symmetry to deduce similar results for values of @xmath0 in the lower half - plane .
[ lem1.2 ] we have @xmath37 if and only if @xmath38 and @xmath39 for some prime @xmath11 and some rational @xmath40 that is not an even integer .
first , suppose @xmath38 and @xmath41 , where @xmath11 is a prime and @xmath40 is a rational number that is not an even integer . as @xmath40 is not an even integer , @xmath42 .
we may write @xmath43 for some nonzero integers @xmath44 and @xmath45 with @xmath46 .
then @xmath47 so @xmath37 .
conversely , suppose @xmath37 .
then there exists some @xmath3 with @xmath48 .
clearly @xmath49 , so we may let @xmath50 be the canonical prime factorization of @xmath13 . then @xmath51 , so @xmath52 for some @xmath53 .
let @xmath54 and @xmath55 .
we know that @xmath20 because , otherwise , we would have @xmath56 .
therefore , @xmath57 , so @xmath58 .
now , @xmath59 so we must have @xmath38 and @xmath60 for some integer @xmath61 . letting @xmath62 , we see that @xmath32 has the desired form . finally , @xmath40 is not an even integer because @xmath63 .
[ lem1.3 ] suppose @xmath38 and @xmath64 .
let @xmath65 , and let @xmath66 be the circle @xmath67 .
then @xmath68 is a dense subset of @xmath66 . by lemma [ lem1.1 ]
, it suffices to prove our claim in the case @xmath69 .
furthermore , because @xmath70 for all primes @xmath11 , it suffices to show that the set @xmath71 is a dence subset of the circle @xmath72 .
we know that every point in @xmath73 lies on the circle @xmath74 because @xmath75 for all primes @xmath11 . now , choose some @xmath76 and some @xmath77
. we may write @xmath78 for some @xmath79 $ ] .
we wish to show that there exists a prime @xmath11 such that @xmath80 for some integer @xmath81 .
equivalently , we need to show that there exists a prime @xmath11 and a positive integer @xmath13 such that @xmath82 . setting @xmath83 , @xmath84 , and @xmath85
, we may rewrite these inequalities as @xmath86 .
it follows from the well - known fact that @xmath87 that such a prime @xmath11 is guaranteed to exist for sufficiently large @xmath13 ( here , we let @xmath88 denote the @xmath89 prime number ) . [ lem1.4 ] if @xmath90 , then @xmath91 for all @xmath3 .
suppose @xmath90 .
for any prime @xmath11 and positive integer @xmath19 we have @xmath92 therefore , for any @xmath3 , @xmath93\left[\prod_{\substack{p\vert n \\ p\geq 2^{1/a}}}\left\lvert\sigma_c\left(p^{\nu_p(n)}\right)\right\rvert\right]\ ] ] @xmath94\left[\prod_{\substack{p\vert n \\
p\geq 2^{1/a}}}(p^a-1)\right]\geq\prod_{p<2^{1/a}}(p^a-1).\ ] ] in the third question that we posed above , we asked if we could find the values of @xmath0 for which @xmath6 is dense in @xmath7 .
lemma [ lem1.4 ] gives us an immediate partial answer to this question .
if @xmath90 , then @xmath6 can not be dense in @xmath7 because there is a neighborhood of @xmath95 of radius @xmath96 that contains no elements of @xmath6 .
we will see in theorem [ thm2.2 ] that , in some sense , @xmath6 is very far from being dense when @xmath90 .
the following lemma simply transforms an estimate due to rosser and shoenfeld into a slightly weaker inequality which is more easily applicable to our needs .
[ lem1.5 ] if @xmath97 , then @xmath98}\left(1-\frac{1}{p}\right)<\frac{\log y}{\log x}+\frac{2}{\log^2y}.\ ] ] rosser and shoenfeld s estimate ( * ? ? ?
* theorem 7 ) states that if @xmath99 , then @xmath100 where @xmath101 is the euler - mascheroni constant .
therefore , if @xmath97 , then @xmath98}\left(1-\frac{1}{p}\right)<\left(\frac{e^{-\gamma}}{\log x}\left(1+\frac{1}{2\log^2x}\right)\right)\left(\frac{e^{-\gamma}}{\log y}\left(1-\frac{1}{2\log^2y}\right)\right)^{-1}\ ] ] @xmath102 @xmath103 [ lem1.6 ] suppose @xmath104 and @xmath69 .
fix some @xmath105 . for each nonnegative integer @xmath61 ,
let @xmath106}$ ] .
let @xmath107 .
if @xmath108 , then @xmath109 and @xmath110 in addition , @xmath111 suppose @xmath112 . then @xmath113 , so @xmath114 . by the definition of @xmath115 ,
@xmath116 $ ] for some nonnegative integer @xmath61 .
therefore , @xmath117 .
the inequality @xmath118 follows from the law of cosines because @xmath119 @xmath120 we now wish to prove that @xmath121 .
this sum makes sense ( the order of the summands is immaterial ) because all summands are positive by the preceding inequality . for sufficiently large @xmath61
, we may use lemma [ lem1.5 ] to write @xmath122 @xmath123 @xmath124 also , if @xmath125 , then @xmath126 hence , for sufficiently large @xmath61 , we have @xmath127 @xmath128 the desired result then follows from the fact that @xmath129 [ lem1.7 ] suppose @xmath130 and @xmath69 . fix some @xmath105 . for each nonnegative integer @xmath61 ,
let @xmath131}$ ] .
let @xmath132 . if @xmath133 , then @xmath134 and either @xmath135 or @xmath136 .
in addition , @xmath137 the proof is quite similar to that of lemma [ lem1.6 ] .
suppose @xmath138 .
then @xmath113 , so @xmath114 . by the definition of @xmath139 ,
@xmath140 $ ] for some positive integer @xmath61 .
therefore , either @xmath135 or @xmath136 .
the law of cosines yields @xmath141 we now show that @xmath142 first , we need to check that the order of the summands in this summation does not matter . for all @xmath143 , we have @xmath144 because @xmath145 and @xmath146 . therefore , @xmath147 for all @xmath143 .
in fact , if @xmath11 is sufficiently large , then @xmath148 as all summands are negative , their order does not matter . using lemma [ lem1.5 ]
, we see that if @xmath61 is sufficiently large , then @xmath149 @xmath150 hence , for sufficiently large @xmath61 , we have @xmath151 @xmath152 we now obtain the desired result from the fact that @xmath153 we omit the proof of the following lemma because it is very similar to ( and quite a bit easier than ) the proofs of lemmas [ lem1.6 ] and [ lem1.7 ] .
we remark that , for the proof of the following lemma , it is more convenient to use theorem 5 in @xcite than it is to use theorem 7 .
[ lem1.8 ] suppose @xmath69 .
for each nonnegative integer @xmath61 , let @xmath154 $ ] , and let @xmath155 . if @xmath156 , then @xmath157 also , @xmath158 .
it turns out that questions @xmath159 and @xmath160 posed in the introduction are not too difficult to handle , so we will give complete answers to them in this section .
[ thm2.1 ] the set @xmath6 is bounded if and only if @xmath161 . if @xmath161 , then @xmath6 is bounded because @xmath162 for all @xmath33 .
now , suppose @xmath104 .
by lemma [ lem1.1 ] , we see that it suffices to prove the result for @xmath163 .
if @xmath164 , then @xmath165 if @xmath69 , then the proof follows from lemma [ lem1.6 ] because @xmath166 where @xmath115 is defined as in the lemma . note that we have used the fact that @xmath5 is multiplicative .
[ thm2.2 ] if @xmath167 and @xmath168 , then @xmath6 has no isolated points .
if @xmath90 or @xmath169 , then every point of @xmath6 is an isolated point of @xmath6 .
first , suppose @xmath145 , and let @xmath170 .
note that @xmath171 by lemma [ lem1.2 ] .
we may write @xmath172 for some @xmath3 .
choose some @xmath77 . to show that @xmath173 is not an isolated point of @xmath6 , we simply need to exhibit a positive integer @xmath174 such that @xmath175 . as @xmath145
, we may choose some prime @xmath176 such that @xmath177 .
let @xmath178 . as @xmath40 is relatively prime to @xmath13 , we have @xmath179 @xmath180 this also shows that @xmath181 because @xmath182 .
we now handle the case @xmath38 , @xmath168 . in this case , @xmath183 .
we wish to show that @xmath6 has no isolated points .
in fact , we will prove the much stronger assertion that @xmath6 is dense in @xmath7 . by lemma [ lem1.1 ]
, we see that it suffices to prove this claim when @xmath69 .
fix some @xmath184 and some @xmath185 $ ] .
choose @xmath186 .
we wish to exhibit a positive integer @xmath174 such that @xmath187 and @xmath188 for some integer @xmath81 .
this will show that @xmath189 is either in @xmath6 or is a limit point of @xmath6 . because we may choose @xmath189 to be any nonzero complex number
, this will prove the assertion that @xmath6 is dense in @xmath7 . by lemma [ lem1.3 ] ,
it is possible to find distinct primes @xmath190 such that @xmath191 for each @xmath192 .
note that @xmath193 , so @xmath194 for each @xmath192 .
for each @xmath192 , we know that @xmath195 lies on the circle @xmath67 and that @xmath196 , so it is easy to verify that @xmath197 . therefore , one may easily verify that @xmath198 let @xmath199 be a positive integer such that @xmath200 now , if @xmath11 is any prime such that @xmath201 , then it is easy to see from the law of cosines ( just as in the proof of lemma [ lem1.6 ] ) that @xmath202 because @xmath203 .
lemma [ lem1.3 ] tells us that it is possible to choose distinct primes @xmath204 such that @xmath205 and @xmath206 for all @xmath207 .
let @xmath208 , and let @xmath209 .
using the fact that @xmath210 , we have @xmath211 therefore , @xmath212 .
this implies that there exists a complex number @xmath173 such that @xmath213 , @xmath214 , and @xmath215 . by lemma [ lem1.3 ] ,
it is possible to choose distinct primes @xmath216 and @xmath217 such that @xmath218 , @xmath219 , @xmath220 , @xmath221 , @xmath222 , and @xmath223 .
essentially , we have just chosen @xmath216 and @xmath216 so that @xmath224 is sufficiently close to @xmath173 and @xmath225 is sufficiently close to @xmath226 .
if we let @xmath227 , then @xmath228 also , there exists some integer @xmath81 such that @xmath229 @xmath230 @xmath231 @xmath232 this complete the proof of the fact that @xmath6 is dense in @xmath7 when @xmath38 and @xmath64 .
now , assume @xmath90 .
choose some @xmath233 , and let @xmath45 be a positive integer such that @xmath234 .
suppose @xmath235 for some prime @xmath236 and some positive integer @xmath237 .
let us write @xmath238 , where @xmath239 and @xmath240 are integers , @xmath241 and @xmath242 .
note that @xmath243 and @xmath244 by lemma [ lem1.4 ] . for the sake of brevity , let @xmath245 .
then @xmath246 if we fix @xmath236 , we see that @xmath237 must be bounded above .
similarly , if we fix @xmath237 , we see that @xmath236 must be bounded above
. consequently , there are only finitely many prime powers @xmath247 that can divide @xmath45 .
this implies that there are only finitely many positive integers @xmath45 such that @xmath248 .
it follows that any disk in the complex plane contains finitely many points of @xmath6 , so every point of @xmath6 is an isolated point of @xmath6 .
the final case we have to consider is when @xmath169 .
this is easy because @xmath249 is simply the number of divisors of @xmath13 .
we see that @xmath250 , so every point in @xmath251 is an isolated point of @xmath251 .
the third question posed in the introduction proves to be more difficult than the first two . for one thing ,
the first part of the question is fairly open - ended .
what exactly would we like to know about the sets @xmath30 ? in order to ask more specific and interesting questions , we wish to gain a bit of basic information .
first of all , if @xmath90 or if @xmath169 , then there is not much use in inquiring about the set @xmath30 because this set is the same as @xmath6 by theorem [ thm2.2 ] . if @xmath164 ( so that @xmath0 is real ) , then @xmath6 is a subset of @xmath252 . in that case
, there are many fascinating questions we may ask .
for example , as mentioned in the introduction , the author has classified those real @xmath0 for which @xmath30 is a single closed interval @xcite . here , however , we will not pay too close attention to the sets @xmath30 for real @xmath0 . for the moment , let us streamline our attention toward the sets @xmath30 when @xmath253 and @xmath64 . by theorem [ thm2.1 ] , these sets are not bounded . at the same time , theorem [ thm2.2 ] tells us that @xmath6 has no isolated points for such values of @xmath0 , so we might hope to obtain more interesting sets than those that arise when @xmath90 .
in fact , if we recall the second part of the third question posed in the introduction and decide to embark on a quest to find those complex @xmath0 for which @xmath254 , then we need only consider the case @xmath253 , @xmath64 . to begin this quest
, we define a set @xmath255 $ ] for each complex @xmath0 by @xmath256 .
the set @xmath257 is simply the set of arguments of nonzero points in @xmath6 .
clearly , if @xmath254 , then @xmath258 $ ] .
in fact , the truth of the converse of this assertion ( in the case @xmath253 ) will allow us to deduce our main result .
first , we need the following lemma .
[ lem3.1 ] if @xmath130 and @xmath64 , then @xmath258 $ ] .
choose some @xmath259 .
we will find a positive integer @xmath13 such that @xmath260 for some integer @xmath81 , and this will prove the claim .
consider the sets @xmath261 and @xmath262 that were defined in lemma [ lem1.8 ] , and let @xmath263 , where @xmath264 .
pick some positive integer @xmath61 .
note that @xmath265 . because lemma [ lem1.8 ] tells us that @xmath266 , it is not difficult to see from a bit of geometry that @xmath267 using the fact that @xmath268 along with the inequality @xmath269 , which holds for all real @xmath270 , we have @xmath271 it then follows from lemma [ lem1.8 ] that @xmath272 . because @xmath145 , @xmath273 .
let @xmath274 be a positive integer such that @xmath275 for all @xmath276 .
as @xmath277 , there exists some integer @xmath278 such that @xmath279 . setting @xmath280
, we have @xmath281 for some integer @xmath81 , from which we obtain the desired inequalities @xmath260 .
[ thm3.1 ] the set @xmath6 is dense in @xmath7 if and only if @xmath253 and @xmath64 .
we have seen that @xmath253 and @xmath64 if @xmath254 , so we now wish to prove the converse . if @xmath38 and @xmath64 , then we saw in the proof of theorem [ thm2.2 ] that @xmath6 is dense in @xmath7 .
therefore , let us assume that @xmath130 . with the help of lemma [ lem1.1 ] , we may also assume that @xmath69 . fix @xmath0 ( with @xmath282 ) , and choose some @xmath283 and some @xmath184 . let @xmath13 be a positive integer such that @xmath284 .
we wish to show that @xmath189 is a limit point of @xmath6 .
if we can accomplish this goal , then we will know that any arbitrary nonzero complex number @xmath14 is a limit point of @xmath6 .
indeed , lemma [ lem3.1 ] allows us to choose @xmath285 arbitrarily close to @xmath15 , and we may set @xmath286 .
it will then follow that @xmath95 is also a limit point of @xmath6 so that @xmath254 . if @xmath287 , then we are done by theorem [ thm2.2 ] .
therefore , let us assume that @xmath288 .
in particular , @xmath289 because @xmath290 .
choose some @xmath291 with @xmath292 . if @xmath293 , let @xmath294 ; if @xmath295 , let @xmath296 . we will produce a positive integer @xmath174 such that @xmath297 and @xmath298 for some integer @xmath81 .
as @xmath299 and @xmath300 are fixed , we may make @xmath301 as small as we wish by initially choosing sufficiently small values of @xmath302 and @xmath303 .
therefore , the construction of such an integer @xmath174 will prove that @xmath189 is a limit point of @xmath6 ( technically , we must insist that @xmath304 , but this follows from our assumption that @xmath288 ) .
let us define @xmath305 , @xmath306 , @xmath115 , and @xmath139 as in lemmas [ lem1.6 ] and [ lem1.7 ] .
we will also write @xmath307 and @xmath308 , where @xmath309 and @xmath310 . because @xmath145 , we have @xmath311 and @xmath312 for all sufficiently large integers @xmath61 .
let us fix some positive integer @xmath274 large enough so that for all integers @xmath313 , we have @xmath311 , @xmath312 , @xmath314 , @xmath315 , @xmath316 , and @xmath317 .
suppose @xmath293 .
because @xmath311 for all @xmath313 and @xmath318 by lemma [ lem1.6 ] , there must exist some integer @xmath278 such that @xmath319 this yields the inequalities @xmath320 let @xmath321 . because @xmath322 is relatively prime to @xmath13 for all @xmath313
, we have @xmath323 therefore , @xmath324 , so @xmath325 as desired .
we now prove that @xmath298 for some integer @xmath81 .
we have @xmath326 for some integer @xmath81 .
therefore , @xmath327 let us fix some @xmath328 .
because @xmath329 , we know from lemma [ lem1.6 ] that @xmath330 . also , @xmath331
. using basic trigonometry , we have @xmath332 if we write @xmath333 , then lemma [ lem1.6 ] tells us that @xmath334 .
recall that we chose @xmath274 large enough to ensure that @xmath335 .
also , @xmath336 because we chose @xmath337 .
we have @xmath338 , so @xmath339 @xmath340 this shows that @xmath341 from these inequalities , we get @xmath342 @xmath343 @xmath344 we now use the fact that @xmath345 to conclude that @xmath346 this completes the proof of the case @xmath293 .
let us now assume @xmath295 .
the proof of this case is quite similar to the previous case . because @xmath312 for all @xmath313 and @xmath347 by lemma [ lem1.7 ] , there must exist some integer @xmath348 such that @xmath349 this yields the inequalities @xmath350 in this case
, we will let @xmath351 .
because @xmath352 is relatively prime to @xmath13 for all @xmath313 , we have @xmath353 therefore , @xmath354 , so @xmath297 once again .
we have @xmath355 for some integer @xmath81 . therefore , @xmath356 let us fix some @xmath357 .
because @xmath358 , we know from lemma [ lem1.7 ] that @xmath359 or @xmath360 . also , @xmath361 .
trigonometry tells us that @xmath362 recalling that we chose @xmath274 large enough to ensure that @xmath363 , we see that @xmath364 . in addition , because @xmath365 , we have @xmath366 . if we write @xmath367 , then lemma [ lem1.7 ] tells us that @xmath368 .
we also see that @xmath369 , so @xmath370 .
using the inequality @xmath371 , which holds for all @xmath372 , we have @xmath373 thus , @xmath374 . combining these inequalities ,
we get @xmath375 @xmath376 @xmath377 using the fact that @xmath378 , we have @xmath379 hence , we conclude that @xmath380 this completes the proof of the case @xmath295 .
recall the section of the proof of theorem [ thm2.2 ] in which we proved that @xmath254 whenever @xmath38 and @xmath64 .
we proved this fact by describing the construction of an integer @xmath174 and showing that we could make @xmath381 arbitrarily close to any predetermined complex number .
the observant reader may have noticed that the integer @xmath174 that we constructed was squarefree .
furthermore , we constructed @xmath174 using primes that could have been arbitrarily large .
in other words , there was never an upper bound on the sizes of the required primes .
the same observations are true of the proofs of lemma [ lem3.1 ] and theorem [ thm3.1 ] .
in particular , we chose @xmath13 in the proof of theorem [ thm3.1 ] to be an arbitrary element of @xmath257 , and we could have chosen @xmath13 to be a squarefree number with large prime divisors . therefore , we obtain the following corollary .
[ cor3.1 ] let @xmath382 be a positive integer , and let @xmath383 be the set of squarefree integers whose prime factors are all greater than @xmath382 .
suppose @xmath253 and @xmath64 .
let @xmath384 .
then @xmath385 is dense in @xmath7 .
we now focus on the sets @xmath6 that arise when @xmath161 .
we saw in theorem [ thm2.1 ] that these sets are bounded , so it is natural to inquire about the shapes of their closures .
we will prove two theorems in order to provide a taste of the questions that one might wish to ask about these sets .
the first theorem is very straightforward , but leads to an interesting question .
[ thm3.2 ] if @xmath161 , then @xmath386 . as @xmath387 , we have @xmath388 @xmath389 now , @xmath390 , and the result follows .
it is clearly not difficult to extract explicit bounds for @xmath391 from the proof of theorem [ thm3.2 ] , but we are more concerned with the following obvious corollary .
[ cor3.2 ] if @xmath161 , then @xmath392 .
suppose @xmath161 .
since @xmath30 is lebesgue measurable , we may define @xmath393 to be the area of @xmath30 .
provided @xmath394 , we may let @xmath395 be the centroid of @xmath30 ( where @xmath396 represents a differential area element in the complex plane ) .
corollary [ cor3.2 ] endorses @xmath397 as a potential candidate for @xmath398 , although it certainly does not provide a proof .
unfortunately , a rigorous determination of the value of @xmath398 seems to require knowledge of the shape of @xmath30 because of the apparent necessity of calculating the integrals involved in the definitions of @xmath393 and @xmath398 .
the limit in corollary [ cor3.2 ] does not appear to be too useful for these purposes because its value depends on the ordering of @xmath8 .
that is , if we let @xmath399 be some enumeration of the positive integers , then it could very well be the case that @xmath400 ( or that this limit does not exist ) .
hence , for now , we will let @xmath398 be .
it seems natural to ask for the values of @xmath0 with @xmath161 for which the sets @xmath30 are connected .
the following theorem will show that if @xmath31 is sufficiently negative ( meaning negative and sufficiently large in absolute value ) , then @xmath30 is separated . in fact , the theorem states that for any positive integer @xmath174 , if @xmath31 is sufficiently negative , then @xmath30 is a disjoint union of at least @xmath174 closed sets .
this is somewhat unintuitive since , for each @xmath3 , @xmath401 as @xmath402 . in other words , the sets @xmath30 shrink " while becoming more separated " as @xmath402 .
[ thm3.3 ] for each positive integer @xmath174 , there exists a real number @xmath403 such that if @xmath404 , then @xmath30 is a union of at least @xmath174 disjoint closed sets .
we will assume that @xmath161 throughout this proof .
note that we may set @xmath405 .
we will let @xmath406 denote the image under @xmath5 of a set @xmath407 of positive integers . for each @xmath408
, let @xmath409 denote the @xmath410 prime number .
let @xmath411 be the set of positive integers that are not divisible by any of the first @xmath412 primes , and let @xmath413 denote the set of positive integers @xmath13 such that the smallest prime divisor of @xmath13 is @xmath409 . in symbols , @xmath414 , and @xmath415 , where we convene to let @xmath416 . fix some positive integer @xmath61 .
we will show that there exists a real number @xmath417 and a function @xmath418 such that if @xmath419 , then @xmath420 for all @xmath421 and @xmath422 .
this will show that @xmath423 is disjoint from @xmath424 whenever @xmath419 .
in particular , it will follow that @xmath423 is disjoint from @xmath425 for all integers @xmath426 whenever @xmath419 .
setting @xmath427 , we will see that if @xmath428 , then @xmath429 are @xmath430 disjoint closed sets whose union is @xmath30 . choose some @xmath421 and @xmath422 .
note that we may write @xmath431 , where @xmath81 is a positive integer and @xmath432 .
if we let @xmath433 for all @xmath434 , then we have @xmath435 similarly , @xmath436 .
observe that @xmath437 , so there exists some number @xmath438 such that @xmath439 for all @xmath419 .
let us define @xmath418 by @xmath440 if @xmath419 , then @xmath441 @xmath442 @xmath443 @xmath444 so @xmath445 .
therefore , @xmath446
@xmath447 @xmath448 @xmath449 @xmath450 for all @xmath419 , and this is what we sought to prove .
the expressions in the proof of theorem [ thm3.3 ] get a bit messy , and we probably could have simplified the argument by using more careless estimates . however , we organized the proof in order to give an upper bound for the values of @xmath451 .
that is , we showed that we may let @xmath438 be any number such that holds for all @xmath419 .
in particular , we have the following corollary . [ cor3.3 ]
if @xmath452 , then @xmath30 is separated .
preserve the notation from the proof of theorem [ thm3.3 ] .
note that @xmath453 therefore , when @xmath454 , becomes @xmath455 we may rewrite this inequality as @xmath456 noting that @xmath457 and dividing each side of this last inequality by @xmath458 yields @xmath459 therefore , we simply need to show that holds for all @xmath460 . if @xmath461 , then @xmath462 , so @xmath463 @xmath464 hence , holds for @xmath461 .
we now show that holds for @xmath465 .
the reader who wishes to evade the banalities of the following fairly computational argument may wish to simply plot the values of @xmath466 and @xmath467 to see that appears to hold for these values of @xmath31 .
let @xmath468 and @xmath469 so that our goal is to prove that @xmath470 for all @xmath471 . in order to reduce the number of necessary computations
, we will partition @xmath472 into the two intervals @xmath473 $ ] and @xmath474 .
for all @xmath471 , we have @xmath475 @xmath476 now , if @xmath477 , then @xmath478 numerical calculations show that if @xmath479 then @xmath480 . because @xmath481 is continuous
, we see that @xmath482 for all @xmath477 . if @xmath483 , then @xmath484 for all @xmath485 numerical calculations show that @xmath486 . because @xmath481 is continuous ,
we see that @xmath487 for all @xmath483 .
= 8.5 cm + = 8.29 cm we have obtained a decent understanding of the ranges @xmath6 of the functions @xmath5 for values of @xmath0 with real part @xmath488 , but many problems concerning the sets @xmath6 that arise when @xmath161 remain open . for example , it would be quite interesting to determine the values of @xmath0 for which @xmath30 is connected .
corollary [ cor3.3 ] provides an initial step in this direction , but there is certainly much work that remains to be done .
for example , it seems as though the bound @xmath460 in that corollary is quite weak since the estimates used to derive it are far from optimal . recall that we derived that bound by showing that if @xmath460 , then @xmath30 is the disjoint union of the closed sets @xmath489 and @xmath490 defined in the proof of theorem [ thm3.3 ] .
we remark , however , that it might be more useful to look at the value of @xmath0 for which @xmath491 is disjoint from @xmath492 . indeed ,
if we confine @xmath0 to the real axis and decrease @xmath0 from @xmath493 to @xmath494 , we will see that @xmath30 first separates into the disjoint union of the two connected sets @xmath491 and @xmath492 .
more formally , one may use the methods described in @xcite to show that there exist constants @xmath24 and @xmath495 such that if @xmath0 is real and @xmath496 , then @xmath30 is the disjoint union of the two connected sets @xmath491 and @xmath492 ( this phenomenon essentially occurs because the largest value of @xmath497 occurs when @xmath498 ) . nonetheless , a full determination of those complex @xmath0 for which @xmath30 is connected seems to require knowledge about the specific shapes of the sets @xmath30 .
using mathematica to plot points of some of these sets allows one to see the emergence of sets that appear to have certain fractal - like properties ( see figure 1 ) .
however , as we can only plot finitely many points , it is difficult to predict the shapes of the full sets @xmath6 and their closures .
we mention once again the open problem of determining the values of @xmath393 and @xmath398 defined after corollary [ cor3.2 ] above .
observe the point @xmath397 plotted in the top image of figure 1 ( corresponding to @xmath499 ) . upon visual inspection
, this point does not appear to be the centroid @xmath398 of @xmath30 .
this is likely due to the fact that points of @xmath6 clustered in the lower part of the image are packed more densely than those in the middle part .
finally , we remark that , while investigating the topics discussed in section 3 , the author found that it would be useful to have good upper and lower bounds for @xmath257 .
thus , we state the derivation of such bounds as an additional potential topic for future research . | for any complex number @xmath0 , let @xmath1 denote the divisor function defined by @xmath2 for all @xmath3 , and define @xmath4 to be the range of @xmath5 .
we study the basic topological properties of the sets @xmath6 .
in particular , we determine the complex numbers @xmath0 for which @xmath6 is bounded and determine the isolated points of the sets @xmath6 . in the third section , we find those values of @xmath0 for which @xmath6 is dense in @xmath7 .
we also prove some results and pose several open problems about the closures of the sets @xmath6 when these sets are bounded .
.2 in 2010 _ mathematics subject classification _ : primary 11b05 ; secondary 11a25 . _
keywords : divisor function ; complex number ; range ; dense ; isolated point ; closure . _ |
recently the validity of general relativity ( gr ) has been brought to question by yilmaz , et al . although such interpretations allow for gravitation to be mathematically consistent and singularity free .
such revisions fail to describe the behavior of test particles as adequately as gr , elevating gr as the correct theory .
today certain questions about gr remain relevant , such as how does it relate to vacuum energy and quantum mechanics in general .
it has been shown in previous works that gr remains self consistent when including the quantum vacuum , or zero - point field .
however , the search for a self consistent theory of quantum gravity , " remains a major theoretical challenge today . among the theoretical arguments against the standard interpretation of gr
is the choice of mathematical coordinate systems .
special relativity ( sr ) , is based upon the structure of a flat minkowski spacetime given in a four - dimensional coordinate system .
recently attempts have been made describing coordinate systems with fractal spaces as opposed to natural ones .
such an adaptation as the case with the yilmaz approach eliminates singularities within the field equations .
recent observational and experimental data have also put into question the validity of gr . the national aeronautics and space administration ( nasa )
has reported an anomalous acceleration of @xmath3 , on spacecraft on the outer edge of the solar system .
this data was obtained from information gathered by the jet propulsion laboratories ( jpl ) , and the deep space network ( dsn ) .
thus far , no satisfactory conclusion has been given to explain the so called anomalous acceleration towards the sun .
" not only have spacecraft provided some fundamental flaws with gravitation , but laboratory results as well .
eugene podkletnov , has reported a gravitational shielding " effect with composite bulk @xmath4 ceramic plates . in light of all of these developments
it is hard to consider gr as the correct theory .
it is the opinion of the author that gr is a theory that works , " however it does nt necessarily make it the correct theory . the goal of this letter is to show that gr is not the correct theory of gravitation , but just works exceptionally well . just as previously newton s law of universal gravitation worked exceptionally well .
this letter is not intended to be a replacement for gr , nor is it intended to present theoretical flaws of the that theory .
this letter is only presented as an introductory work for an alternative theory of gravitation .
the general theme of this letter is given by the following postulates : ( virtual gravitation ) .
spacetime is not a null energy field , it consist of asymptotic vacuum fluctuations , and behaves as a virtual energy - sheet . "
( planckian invariance ) .
the planck length is a gauge invariant function for all ( interacting ) brane observers .
[ an adaptation to the postulate of new relativity . ]
this letter is presented in the following format in section [ uni ] a brief introduction into unified field theories are given .
in section [ qgh ] a few quantum gravity approaches are introduced . in section [ fg ] fractal geometry
is introduced and its relations to a complex system are given . in section [ qedf ] the meaning of fractal geometry for qed is discussed . in section [ qcdf ] the meaning of a fractal geometry is discussed for qcd . in section [ card ] a new theoretical particle is introduced utilizing fractal geometry . in section [ flt ] a relationship between n - dimensional and two - dimensional systems are given . in section [ geo ]
a philosophy of geometry is given . in section [ vac ]
the effects of the quantum vacuum are discussed . in section [ stg ] a relationship between fractal geometries and the quantum vacuum
are discussed . in section [ feystg ] the meaning of feynman diagrams are discussed . in section [ qm ] the validity of quantum mechanics is brought into question . in section [ bm ] an alternative description of gravity is given which may explain the epr paradox . in section [ cqg ]
an overview of a canonical non riemannian gravitational field is given . in section
[ can ] the planck length results as a function in canonical quantum gravity . in section [ aa ] a possible alternative for the anomalous acceleration " of spacecraft
is given . in section [ pgeo ]
pseudo geodesic equations are presented . in section [ dis ] a general discussion of this work
is presented . in section [ pl ] a discussion of the meaning of the planck length
the conclusions of this work is drawn in section [ con ] , which gives stronger definitions to equivalence principle in appendix [ ep ] . finally it is suggested that there may exist a detectable from of yang - mills gravity " in appendix [ ymg ] .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ i am convinced that he [ god ] does not play dice . _
einstein _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ einstein , quit telling god what to do . _ " n .
bohr _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the unification of gravitation with quantum mechanics began with einstein s objections to the newly developed quantum theory .
although acknowledging the successes of the new theory he believed it to be incomplete .
einstein was convinced that there was a deeper theory involved , one which would also include gr , a unified field theory was christened .
soon came the work of kaluza and klien , giving a pseudo mathematical unification of electromagnetism and gravitation .
the theory would soon die out and loose interest , until quantum mechanics came around . and asked the simple question , how does gravity behave at the quantum level ; the answer kalzua - klien gravity .
research in this area soon exploded , extra dimensions were soon added to the field equations , superstring theory was born .
particle physics began unifying fundamental forces as well , the weak force , the strong force , the electromagnetic force .
but no gravitational force , cosmologists helped out , the big bang and nucelarsynthesis would help to explain the problem .
soon physics became littered with grand unified theories ( guts ) and theories of everything ( toes ) , they all have the approach of a unified field theory .
" however , they missed the simple point einstein was trying to make , how do quantum mechanics and relativity relate ?
this is a hierarchical question , the relevant question is how do macroscopic and microscopic worlds communicate ?
historically there have been two models formulated for the construction of a consistent quantum gravity theory .
they are the canonical and hamiltonian approaches .
these two approaches have had limited success , however more recently the theory of loop quantum gravity has been introduced into the sea . out of the three approaches presented loop quantum gravity
is generally accepted as the correct approach . however , for a more accurate description of the historic developments of quantum gravity see . in this letter
i will focus on the canonical approach as it relates to gauge invariance .
_ if we re built from spirals while living in a giant spiral , then is it possible that everything we put our hands to is infused with the spiral ? _
" + max cohen in the motion picture @xmath5 the presence of matter within gr disturbs the field equations by the existence of singularities or point - particles . "
how can one avoid this eye sore in the equations , simple fractal geometry .
if matter is fractal it can not condense into points , however this allusion can still take place above the planck energy scale .
it is hard to believe that this simple approach has only been attempted in recent times , fractal sets are more common in nature than simple polygons .
first let us begin with a simple construction of a fractal set with the simple equation @xmath6 .
one must also realize that a fractal is composed of a complex number system , i.e. a+ib .
using this form one may wish to construct a complex averaging of the mean , which results from the golden mean @xmath7 .
thus we have constructed a complex mean which has two possible solutions as seen below : c_m = + = \ {
l ; a = c + a ; a c + .
this complex mean thus has non communcating solutions .
which from the stand point of imaginary numbers yields the general statements : = i = - i thus these statements would appear to be in agreement of the theory of quaternions . which is interesting enough in itself , a four - dimensional version of the complex number system . with this preliminary work set
we can now construct a minkowski spacetime which that takes advantage of fractal dimensions .
first one can construct the generalized three - dimensional manifold as a three - brane , and thus incorporating time as a fractal set .
thus the fractal construction on spacetime is presented in the form @xmath0 a similar approach was made in ref . .
it is here postulated the origin for this three - dimensional brane arises from planck scaling .
the reason for this postulate is seen when the golden mean is applied to @xmath1-dimensions @xmath8 , the higher n the closer the mean is to 1 .
thus only when @xmath9 approaches infinity will we see that @xmath0 will yield the standard minkowski space of 3 + 1 dimensions .
this is of course in agreement with any system that we apply c=@xmath10=@xmath11=@xmath12=1 more evidently this corresponds to a time dilation effect in terms of special relativity .
such that we have the following revision to the flat four - dimensional minkowski space : ( , z^2)=c^2-()z^2_1-()z^2_2-()z^2_3 [ k ] it is interesting to note that this pseudo metric appears to be an inverse of the standard minkowski spacetime , this importance is seen in section [ bm ] .
so far this method has only left intriguing consequences , however it diverges from the point made earlier in this section . in string theory
the atomism view of matter is replaced with vibrating strings , these vibrating strings correspond to a real geometry .
however , fractal geometries are allowed to break of these strings into imaginary components .
these imaginary components thus make the string a complex function , yielding a pseudo point - like string . to analyze this premise
allow us to view the nambu - goto action @xmath13 .
if strings can indeed be made to subside with fractal geometry , then they would break off into imaginary components by the empirical action : s=_0 ^2 where @xmath14 .
this pseudo string forms many more string components in @xmath1-dimensions , that is the string fragment continues in infintium .
however , in the physical sense the string exist a pseudo point - like particle , do to the scaling nature of the planck length . these fractal strings then interact within a field , known as gravitational or zero - point fields .
when the fractal strings converge with other fractals , a self organization takes place , i.e. the production of virtual particles .
this production is made possible through the non communacating mathematics associated with quantum mechanics .
as the particles are produced they destroy one another , such that their world - sheets reverberate in a complex form .
this complex reverberations in @xmath1-dimensions is responsible for the vibration of the string , which we navely associate with mass .
electromagnetic waves are the result of four - dimensional interactions , and its real wave would correspond to the results found in classical mechanics .
however , it would intern have a fractal complex field , which would cause the field to break periodically yielding a string fragment , a quanta .
this quanta in non self regulating , i.e. it is the nature of the real wave which causes the string to reverberate .
the above consideration may carry some controversy in the well known theory of quantum electrodynamics ( qed ) .
if a quanta is just a fractal string , then what is the proper approach for the exchange of energy between the two systems ?
well , the result appears as classical approximation , the quanta hits the string as a solid body , causing a change in geometry , which virtual particles oppose .
this causes the string to bounce " back to its original form , emitting a real wave , but not necessarily a quanta , remember a quanta is given by a complex field .
since quantum mechanics is sketched out onto a point particle - like environment , its consequences would agree with the qed model .
in fact the fractal model yields a much physical picture for non communacating relationships than the quantum theory .
the above result would agree with qed , however , its definitions are quite weak , in fact one may expand these definitions to quantum chromodynamics ( qcd ) .
thus a nucleon may be made to reflect electromagnetic radiation as well , however , this dose not defy the documented experiments in any magnitude .
it would be the interaction of the system , i.e. what string perimeters give way to the colored , and other gauged forces , that yields its particles " .
each string fragment consist of its own local vibrations ( gauge invariance ) , which attributes its mass , i.e. differentiates between a higgs particle and a quark .
these states then have their own local statistics , their real waves would then correspond differently than the electromagnetic field . which results in the production of the celebrated yang - mills field , and thus yields the production of colored particles such as the gluon .
cantor pioneered the study of infinities with his new theory of cardinal numbers , however he faced opposition in his time for this new theory .
cardinal numbers offer the best insight into to the study of fractal strings in n - dimensions , these interpretations in fact have direct physical consequences .
most notably it can explain the situation unleashed by the infamous epr - bell paradox , where faster than light ( ftl ) communication appears possible ( depending on planckian scaling ) . for recent theoretical implications and interpretation of the epr - bell paradox see . in each scaling the laws of physics
would be very different , and hence superluminal velocities would seem to appear in lower branes .
thus the traditional interaction of strings should not be limited on a specified dimension , but behave as a set of cardinal numbers .
thus a slight revision of the nambu - goto action should be given which yields s=_0 ^2 where @xmath15 $ ] .
_ to divide a cube into two or other cubes , a fourth power , or , in general , any power whatever into two powers of the same denomination above the second is impossible _
" + fermat fermat s last theorem can be associated with fractal geometry in one respect , there is no general real solution to fractal geometry above dimension 2 .
this may be a simple coincidence and may have no deeper meaning , however , this is contrary to the recently created holographic principle ( hp ) .
the hp relates that the universe may exist in dimensions of infinitum status .
however , the laws of physics are best projected onto a pseudo two - dimensional screen , and our three - dimensional world is only a pseudo manifestation of an @xmath1-dimensional continuum . in string theory
we can view our universe as made up of two two - dimensional branes ( described by type iia d2 membranes ) .
thus any other dimension outside the holographic conjecture yields no physical meaning and no solution .
just as what is suggested by fermat s last theorem , therefore our four dimensional slice of the brane is a pseudo physical manifestation of the holographic screen . there is only one explanation for this result , there must exist a physical constant for specified energy scales , i.e. the planck scale . here
another coincidence appears to arrive , the two dimensional wave equation for string theory @xmath16 .
it seems that both mathematically and physically there is a special importance with dimension 2 .
this discussion is largely philosophical , however it is interesting to note that _ cardinal strings are given by complex numbers .
in fact a cardinal string in four - dimensions is remarkably similar to the two - dimensional form , and seems to correspond to a torus : s^2=_0 d^2 + _ 0 d^2 _ 0 d^4 ( x1 ^2_iiat^2_iiam ) from this complex structuring , and properties of cardinal numbers it can be seen why quaternions were alluded to in section [ fg ] . _
einstein s theory of gr transformed newton s theory of a gravitational force , to a direct consequence of geometry .
however , although the idea of a force was replaced with a geometry , the geometry still yields a force when explained in riemannian geometry . an even more radical approach to gravitation as a geometry was produced by roger penrose in his theory of twistor spaces .
the geometry itself is more important than physical masses , in fact masses only come important when one expands this theory .
this is true for a fractal revision of string theory , it exists as a pure geometry , the interacting geometry in fact produces mass .
this seems almost a radical stance from the point of view of gr , however geometry remains a key factor as the ideal of a force to gr .
the vacuum exist from a state of virtual particles being produced via cardinal string fragments .
since virtual particles are a pure construction of particles " in @xmath1-dimensions , they are not true strings ( i.e. they violate the hp ) .
these particles thus carry no mass - energy equivalent within our universe , never the less they still posses a geometry .
this situation only holds true when the system is localized , however , when interacting with non cardinal strings can induce an energy exchange . by the well known casimir effect
the energy of the vacuum should be given by : @xmath17 and when interacting with an inertial mass system we have : @xmath18 this relates the fact that as mass is accelerated it pushes the quantum vacuum energy ( which is analogous to the assumption made be postulate one ) . in other words
it reacts in the same fashion as air molecules do when inertial masses accelerate on earth ( producing pressure on the system ) .
furthermore , it can be assumed that a material body increases its rest mass by absorbing this zero - point - energy ( this assumption must be given in order to satisfy conservation laws ) .
moreover , since they are cardinal strings they are unified in a manner , thus the vacuum is a geometrical manifestation of string particles .
that is the geometrical patterns formed through string interactions is what we call a gravitational field , i.e. a virtual gravitational field .
since these string interactions are only virtual there is no reason to modify the einstein field equation , unless one wishes to discuss quantum string effects .
furthermore @xmath1-dimensional spacetime metrics have shown to be very similar to the structure of four - dimensional spacetimes .
therefore the classical gravitational field is removed from quantum mechanics as it exist in a virtual sense , thus quantum mechanics is a property of matter , i.e. interacting geometries .
the vacuum however is not currently treated as the geometry of a fractal spacetime system , and hence is incompatible with other vacuum theories .
however a fractal model for quantum mechanics appears to agree with at least one interpretation of the quantum vacuum .
in fact this interpretation goes right along with loop quantum gravity , and string theory see ref .
the hausdorff ( or fractal ) dimension suggest that dimensions maybe confined to a @xmath19 spacetime , with fractal string scaling .
this principle was postulated earlier in this letter as the consequence of the
planckian scaling , " and is the leading postulate in new relativity .
here the importance now becomes what is the meaning of de - broglie phenomenon .
the einstein de - broglie equation @xmath20 can be seen as a representation of the relativistic wave equation , i.e. mass is a quantifiable measure of energy . which can be applied directly to string theory , the vibrations of the string are given in a fractal frequency comparable to the compton wavelength .
> from the above consideration it can be equally applied the the origin of a bodies mass intern comes from the gravitational field itself .
this requires the use of feynman diagrams , and believe it or not this approach is indeed correct , if strings are represented by cardinal numbers ( and if quantum mechanics is considered to be correct ) .
since this allows for ftl communication at the classical level it can be interpreted at least at the quantum level that mass originates from spacetime ( when measured at the planckian scale ) .
in fact at the quantum level the production of virtual particles may be responsible for a light paths geodesic curvature , yielding a quantum gravity theory .
more correctly it may be viewed that the quantum theory is in reality a classical approximation of string theory .
( brussels approximation ) .
quantum mechanics exist as a classical approximation of string interactions which possesses an apparent time reversed symmetry . deriving this theorem
let us consider the following thought experiment : if mass is composed of vibrating stings , and intern these strings produce gravitational fields in n - dimensions then space is vibrating .
however , such an approach would imply that geometrically speaking the two systems are unaware of their own vibrations under gauge invariance . on the other hand ,
the interaction of fractal strings are given in a complex field , which itself is anti - communcating .
thus spacetime , or string space is subjected to uncertainty principles as shown in ref . .
since at the classical level , the fractal space can give way to real solutions i now make the assertion that the quantum particles are at flux , and not the space itself ( for argumentative purposes only ) .
therefore when a quantum is observed , it is the space which becomes fuzzy , " not the particle , and when not observed the inverse follows .
thus the theorem leads to two possible out comes during an observation sequence .
i ) the fuzzy quantum particle becomes a point particle , when time symmetries are reversed .
ii ) spacetime is fuzzy , however when collapsed by a point particle elucidates to a natural " state .
thus it is seen that only when time symmetries are reversed does one obtain the laws commonly associate with quantum mechanics . making the only valid approximation of quantum mechanics the brussels interpretation
, this may also explain the epr paradox .
however our thought experiment does yield one solution which interpretation i and ii are consistent .
when a particle is observed by a frame , it is in reality observed by the local states ( brane ) of the cardinal string , which may be in any number of states . as a point particle ( non local string )
enters the system it begins to collapse the wave function of the local state .
this makes it appear that a fuzzy quantum particle has entered the system , much as a star appears to twinkle in the night sky .
this collapsing of the state i will call the observational s frame present , " before the action the observational state was fuzzy .
however , after the event a self organization took place , an event occurred , producing a present state .
after the quanta is observed by the observational frame , its present state then becomes certain in the terminology of classical quantum mechanics .
therefore local brane string interactions can not take place until a non local ( cardinal ) string collapses the wave function of the system .
> from the feynman interpretation of the time reversed symmetries of the gravitational field , new conclusions about the nature of spacetime can be made .
it is here postulated that classical mechanics is in reality a description of a quantum system given under an approximation of a time reversed symmetry .
such that the following statements become true : * reversal of bohmian mechanics ( bm ) yields classical mechanics * reversal of brussels interpretation ( bi ) yields standard quantum mechanics ( sqm ) with the fractalization of spacetime given in section [ fg ] , we may conclude that ( with the use of bohmian mechanics ) that inertial mass yields an expansion of spacetime .
thus as a body gains mass as it accelerates in classical spacetime it causes the fractalization of the bohmian system to increase ( which is analogous to a lorentz transformation ) . which thus gives the allusion that spacetime is contracting in the classical real frame .
the gravitational force , thus is an inertial acceleration which radiates a pseudo center of gravity vector in terms of newtonian mechanics .
however , this is how we interpret the events classically , in reality it is the expansion of the fractal bohmian space ( the @xmath21 term in section [ fg ] , e.g. the time dimension ) which yields inertial acceleration . therefore , it is the brussels interpretation of quantum mechanics which yields the epr paradoxes , the connection of the particles is created by the ( incorrect ) approximation of time reversed symmetries . that is to say the epr paradox only includes simple ( non fractal ) states , which by time reversal appears to yield ftl communication ( see figure [ fig ] ) .
this appearance of ftl communication should nt be taken to seriously since recent experiments ( cfr .
wang , et al ) appear to yield ftl communication .
however , it is the string interactions which yield quantum mechanics , in fact it yields the same interpretations as bohmian mechanics .
thus bohmian mechanics adequately describes the behavior of particles " while , the brussels interpretations yields standard quantum mechanics [ meaning that this system is only an approximation ] .
since complex spaces have been presented as a solution to the singularity problem , it is natural for a formulation of a complex spacetime . to proceed in this manner
one must neglect the cherished einstein - hilbert action @xmath22 and replace it with the tucker - wang action : @xmath23 therefore we can now discuss a complex gravitational field , without the traditional riemannian geometry .
the classical einsteinian relativity gives the generic field for a spacetime geometry as @xmath24 .
such that i now wish to make the generalized statement @xmath25 , or in canonical terms @xmath26 .
as such a generalization of a purely idealistic spacetime governed by perfect fluid becomes : g^ab=16gt_ab where under ideal cases one can have the geometry @xmath27 .
the reason the field takes on the term @xmath28 , as opposed to @xmath29 , can be seen with the use riemannian metrics .
first , let us begin with a ricci symmetric tensor of the form : @xmath30 which within a constant field becomes @xmath31 .
this field can thus transpose to @xmath32 , and couple to an opposing electromagnetic field by the connection @xmath33 which therefore leads to the following antisymmetric riemannian field @xmath34 whence therefore means that there must be an equivocal orthonormal action taking place , such that one has @xmath35 . in which the generic geometry for a perfect fluid becomes that of g^ab=16 gt^_ab which translates to the field equation : r^a_b -1 4g(e_(a ) ) , e_(b))r=-16 g c^4t^_ab this equation must be modified when given in an @xmath1-dimensional system such that on has : r^a_b -1 2(n)g(e_(a),e_(b))r=-8 ( n)g c^2(n)t^_ab [ wfe ] the above equation is in essence a canonical gravitational field equation , which appears to be a good candidate for a quantum gravity theory . where the geodesic equations become - _ [ ] ^ dx^ dx^ ( 1 + _
ab(n ) ) 0 however , this interpretation suggest that spacetime is quantitized by a canonical action however , quantum particles are given as classical particles
. hence this interpretation would be an inverse of understood quantum mechanics .
however , here a paradox opens up , when time is reversed particles remain in one quantum state , thus sqm is not retrieved .
thus , it is seen that there exist no true
quantum gravity " theory .
in fact if one applies this formulation with the planck length it destroys the principle of _ planckian invariance and gives an allusion to the existence of an ther . _ therefore the planck length no longer remains a constant but becomes a dynamical function .
first let us write the planck length in terms of @xmath1-dimensions and apply it to the above field equation such that we have : l_p = ( ^n /m^n _ p c^3n)^1/2n.momentum must be reevaluated from @xmath36 , such that @xmath37 .
thereby the planck length , and mass are actually given by a particles rest momentum .
such that the planck length is in reality given by the function : l_p = ( ^n /m_p_0 ^n c^3n)^1/2n .
this thereby has major implications , that the planck length is not really a constant at all but a function of momentum . such that as an object increase in speed with respect to its rest momentum , its planck energy becomes larger .
that is as a material body is subjected to length contraction , its planckian energy is modified to compensate for the effect . since the momentum is measured at rest m remains a constant , it is the velocity of the system which changes . thereby meaning that length contraction in special relativity is not given by lorentz transformations , but by the local rest momentum of the planck barrier .
this results when we interpreted this action canonically however under bm it yields expected results .
therefore it is seen that a canonical formulation fails to keep planckian invariance " which represents a failed attempt at a quantum gravity theory .
finally i bring light to an alternative explanation for the acceleration of spacecraft . since the findings of the anomalous acceleration towards the sun , " there have been a number of possible explanations given . with the construction of a fractal @xmath1-dimensional spacetime ,
i view this as a quantum gravity effect . as an object accelerates its fractal geometry changes [ by means of bm ] , thus resulting in pressure on the system .
pressures as the source of a gravitational field were pioneered long ago by einstein : @xmath38 this method is not ad hoc , gravitational pressures for atomic gases and radiation can be given by : p_gas = hqtp_rad=13at^4 which lends the general results @xmath39^{1/3}\\q^{4/3}=c(\beta)q^{4/3}\end{aligned}\ ] ] thus after a slight modification of eq.([wfe ] ) , one can obtain the following gravitational pressure : r^ca_b=-(t^ca_b-12g^c_abt)+2a^2^c_ab=(2-p ) [ rfe ] therefore the flat field equations can be given by : ^c _ ab(z)0,r^ca _ b(z)0 thus a line elements trajectory would be given by ( , z^2)=g_caz_cz_a therefore a metric in a complex fractal spacetime can be given by : _
ab=1,2^c = iz_a^c z_b^c0 since this quantum gravity effect originates from the planck length it is very unlikely that the yukawa interaction : @xmath40\ ] ] will take place ( unless special conditions arise ) .
however , if such an effect does arise , it may yield peculiar motion for an obejects geodesic path .
first let us begin with the two - dimensional lagrangian hamiltonian , so that we have an equation of motion from the simple action = h p_i;=h q_i in canonical terms motion is given by q_i - h p_i;p_i = h q_i lending a four - vector of the form @xmath41 .
in such that a hamiltonian wave within a gravitational field would be in motion according to the geodesic path : -^ _ [ ] dx^ dsdx^ ds0 this geodesic unlike the prior for a classical particle , will not differentiate and thus its motion need not transverse through classical euclidean space . therefore it can be seen that complex spaces could impose unseen forces which would effect a geodesic path for a body ( or wave ) in motion . a proposal made in ref .
, made a like was case in the relativistic sense so that one would have : f^=m_0w wdy ddx _
d. although there is no direct physical evidence of this , it is still however an intriguing explanation .
after all the so called anomalous acceleration " is only experienced by small bodies , not massive ones such as planets .
thus a quantum interpretation of this effect seems to fit the observed data better than any other approach .
alternatively modanese has also predicted a macroscopic quantum gravity effect , however it is limited to the podkletnov experiment .
the formulation of this theory was based on a desire for a reformulation of gr in order to describe a singularity free theory ; in which a fractal formulation of the field equations were derived .
the second desire for this theory was the formulation of a quantum construction of gr , however the end result is a gravity theory which describes quantum mechanics . therefore the gravitational field and matter can be considered to be molded into the following form .
matter exist as a pseudo point particle who s field of movement is restricted onto a two - dimensional ( complex ) frame .
this two - dimensional frame s movement is governed by bm , and in part by the hp .
matter , is thus in reality a fractal vibrating string fragment which continues on into @xmath1-dimensions .
the fractalization of this cardinal string " produces virtual particles which posses a geometry , it is this ( virtual ) fractal geometry that is responsible for the gravitational field . in light of future studies
it is likely that an adequate formulation for an alternative to gr be given in the following forms .
one the acceptance of a fractal ( even if only quasi fractal ) structure of matter and space as an adequate formulation for the geometry of spacetime .
two the acceptance of complex systems into the equations , e.g. quaternions , octonions , c * algebras , etc . and finally three , the acceptance of physical conditions which may not be popular , " but yield results that are not contradictory to known data .
the mathematical conditions are the most intriguing to author because there seems to be a hidden mechanism in the mathematics .
however , my advanced mathematics skills are mediocre at best so these avenues are still left open in this letter .
several physical arguments against the existence of singularities have been given by loinger , as well as einstein s classic objections .
thus one may inquire what happens at the planck length , i.e. what are the laws of physics ? here
i now quote kip thorne , on our current understanding of singularites and ` quantum foam . '
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ how probable is that a black hole s singularity will give birth to ` new universes ? '
we do nt know .
it might well never happen , or it might be quite common or we might be on completly the wrong track in believing that singularites are made of quantum foam .
_ " thorne ( 1994 ) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ this now leads a discussion to recent attempts to model gravity in terms of @xmath1-dimesnional spaces ( arkani - hamed , et al 62 - 69 ) , in which the planck length varies with the number of dimensions .
of course the planck length could be infinitely small in an infinite system , clearly a challenge to the principle of planck invariance .
however , we note that with mach s principle ( mp ) the planck length must be observed by an external mechanism to remain invariant .
thus the planck length exist as a fractalization of bm , which becomes an observational frame in classical real mechanics .
furthermore , from this it may be seen that the laws of physics as we understand them are in direct consequence of the planck length .
we may also assume the chosen string field is quantitized ( i.e. given by bm ) , because its mass is attributed to a complex pseudo oscillation ( vibration ) .
where i now quote david bohm ( cfr .
bohm , 22 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ we may conclude that all systems which oscillate are quantitized with @xmath42 whether these systems be mathematical oscillators , sound waves , or electromagnetic waves . _ " _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ so what does physics look like beyond the planck length , remember the ( local ) laws of physics are given by two complex d2-branes . when we interpret these interactions we receive the traditional gr effects at the macroscopic level .
however , at the planck scale singularities do nt exist such that the frame interacts via cardinal strings " and not classical gr .
thus interactions on local branes cease , and supersymmetry takes over .
only strings which are connected to a form of the d2-brane will have observable physical manifestations , this deals with planckian invariance " . in fact each dimension may have its own unique planck length which governs its own local laws ( explaining the limitation of classical string theory to a set number of dimensions ) . which leaves open several areas in @xmath1-dimensional black hole mechanics , and planck length physics .
i have shown that there is enough evidence at present to challenge gr as the correct theory for gravitation .
i have also introduced the study of a complex fractal spacetime system and its possible relationship to the planck length .
the given formulation for a canonical gravitational field resulted in contradictory conclusions , thus ruling out a canonical approach to quantum gravity . "
finally if my hypothesizes hold valid then sqm will begin to make invalid predictions for the behavior of particles near the planck length .
thus a fractal correction for sqm will be needed under certain gravitational fields , which may be comparable to bm .
a new weak equivalence principle ( wep ) for the gravitational field can be postulated utilizing a complex fractal minkowski spacetime ( cfms ) system ( see eq.([k ] ) ) .
since there is no spatial acceleration for the gravitational field ( in respect to mp ) , it is the acceleration of the pseudo time dimension in the cfms which produces a gravitational curvature .
therefore a material body would have the traditional minkowski spacetime , acting as a lorentz frame .
since gravitational fields extend indefinitely , this should cause time to continually progress within an inertial acceleration frame .
this therefore means that as an object enters a gravitational field it becomes less massive , in terms of a lorentz transformation .
equivocally it can be stated that energy is lost in curved spacetime .
a similar effect is all ready known , known as a gravitational time delay , " i.e. the _ shapiro effect . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ according to the general theory , the speed of a light wave depends on the strength of the gravitational potential along its path . _
" shapiro ( 1964 ) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ this is however contradictory to sr , because it fails to describe inertial acceleration within a gravitational field correctly .
however , the _ principle of relativity is still preserved , because the curvature of spacetime corrects for the cfms .
therefore the reason the _ equivalence principle is fundamental in gr is because it is the only priori condition which satisfies the _ principle of relativity .
thus without an equivalence principle , there would be no relativistic theory for the gravitational field .
_ _ _ the logarithm gravitational time delay , may also be responsible for the apparent anomalous acceleration " of spacecraft .
david crawford has offered a similar explanation , where the gravitational term arises from interplanetary dust .
here i now hit upon a topic hinted upon in section [ qedf ] ; converting fractal geometry in the terminology of qcd .
let us now rewrite eq.([rfe ] ) , so that we have an equation of the form : -(r_b ^ca-12g^c_abr^c ) = 8t^c_ab with this equation a yang - mills gravitational pressure can arise under the following field : s_e=14g^2 ^ 4 zf^_f^_+1_ok^ia _
b k^ib _ a ^2 which must be given in a conformal field , i.e. @xmath43 , thus we have : s=12_0+_0_p ^2 from this it is now seen that a cardinal string , " is in fact an @xmath1-dimensional world line . which can communacate with other world lines , where we have a self - organization of the system by v(s)=14 ^ 2 g_ab^jklm_a t_ab_jklm it is these interactions which generate a spinor space , which attributes mass to the geometry .
therefore the yang - mills field is added to the gravitational field , by means of a gravitational pressure . here
a less restrictive form of the strong equivalence principle ( sep ) can be applied : thus the interaction of two or more
cardinal strings , " produces a twistor like action , represented by @xmath45 .
this also means that certain gravitational anomalies may not only arise at the planck length , and may result in experimental verification .
10 yilmaz h. did the apple fall ?
in m. barone and f. selleri , editors , frontiers of fundamental physics .
115 - 124 ( 1994 ) fackerell e. remarks on the yilmaz and alley papers .
proceedings of the first australaian conference on general relativity and gravitation , ed .
d. l. wiltshire , university of adelaide ( 1996 ) , 117 available url : http://www.physics.adelaide.edu.au/asgrc/acgrg1/fackerell.html castro c. on m theory , quantum paradoxes and the new relativity physics/0002019 castro c. why we live in 3 dimensions hep - th/0004152 ya.kobelev l. do gravitational and electromagnetic fields have rest masses in the fractal universe ?
physics/0006043 ya.kobelev l. the theory of gravitation in the space - time with fractal dimensions and modified lorents transformations physics/0006029 ansoldi s. loop quantum mechanics and the fractal structure of quantum spacetime , chaos solitons fractals 10 ( 1999 ) 197 hep - th/9803229 anderson j , _ et al . _ indication from pioneer 10/11 , galileo , and ulysses data , of an apparent anomalous , weak , long - range acceleration , phys.rev.lett .
81 ( 1998 ) 2858 - 2861 gr - qc/9808081 podkletnov e. weak gravitation shielding properites of composite bulk @xmath46 superconductor below 70 k under e.m .
field cond - mat/9701074 overduin j. and wesson p. kalzua - klien gravity .
phys.rept . 283
( 1997 ) 303 - 380 gr - qc/9805018 wallace d. the quantization of gravity an introduction gr - qc/0004005 rovelli c. loop quantum gravity .
living reviews in relativity vol 1 ( 1998 ) max - planck - gesellschaft . /1998
- 1rovelli available url http://www.livingreviews.org/articles/volume1/1998-1rovelli rovelli c. notes for a brief history of quantum gravity gr - qc/0006061 motz l. and weaver j. the story of mathematics , 1993 ( new york : avon books ) szab l. removing the last obstacle to the einstein - fine resolution of the epr - bell problem quant - ph/0002030 t hooft g. dimensional reduction in quantum gravity gr - qc/9310026 casadio r. and harms b. black hole evaporation and large extra dimensions hep - th/0004004 krasnoholovets v. and ivanosky d. motion of a particle and the vacuum , physics essays , vol .
4 , pp . 554 - 563 ( 1993 ) quant - ph/9910023 jaekel m. and reynand s. movement and flutuations of the vacuum , rept.prog.phys . 60
( 1997 ) 863 - 887 quant - ph/9706035 haisch b. on the relation between zero - point - field - induced inertial effect and the einstein - de broglie formula , phys.lett .
a268 ( 2000 ) 224 - 227 gr - qc/9906084 ansoldi s. aurilia a. and spallucci e. loop quantum mechanics and the fractal structure of quantum spacetime , chaos solitons fractals 10 ( 1999 ) 197 hep - th/9803229 sasakura n. an uncertainty relation of space - time , prog.theor.phys . 102 ( 1999 ) 169 - 179 hep - th/9903146 bostm k. regaining time symmetry in the generalized quantum mechaincs of the brussels school quant - ph/0005024 berndl k , et al . a survey of bohmian mechanics , nuovo cim .
b110 ( 1995 ) 737 - 750 quant - ph/9504010 loinger a. on the concept of mass point in general relativity gr - qc/0006033 loninger a. on the propagation speed of wavy metric tensors gr - qc/0007048 crawford d. a possible explanation for the anomalous acceleration of pioneer 10 astro - ph/9904150 tureyshev s , et al .
the apparent anomalous long - rang acceleration of pioneer 10 and 11 gr - qc/9903024 stvang d. assumption of static gravitational field resulting in an apparently anomalous force gr - qc/9910054 einstein a. the meaning of relativity , 1921 ( new york : mjf books ) long j. experimental status of gravitational - strength forces in the sub - centimeter , regime nucl.phys .
b539 ( 1999 ) 23 - 34 hep - ph/9805217 modanese g. theoretical analysis of a reported weak gravitational shielding effect .
hep - th/9505094 wang l. kuxmich a. and dogarlu a. `` gain - assisted superluminal light propagation . ''
nature 406 ( 2000 ) 277 arkani - hamed n. dimopoulos s. and dvali g. `` the universe s unseen dimensions . '' scientific american 283 ( 2000 ) 62 - 69 youm d. extra force in brane worlds hep - th/0004144 bohm d. quantum theory , 1951 ( toronto : dover books ) tuaber g. albert einstein s theory of general relativity : 60 years of its influence on man and the universe , 1979 ( new york : crown publishers , inc . )
thorne k. black holes & time warps : einstein s outrageous legacy , 1994 ( new york : norton ) pg . 478 | in this letter recent developments are shown in experimental and theoretical physics which brings into question the validity of general relativity .
this letter emphasizes the construction of a fractal @xmath0 spacetime , in @xmath1-dimensions in order to formalize a physical and consistent theory of ` quantum gravity . '
it is then shown that a ` quantum gravity ' effect could arise by means of the strong equivalence principle . which is made possible through a pressure of the form @xmath2 . where it is seen that nuclear pressures can be added to the gravitational field equations by means of twistor spaces .
* keywords : * fractal geometry , new relativity , quantum vacuum , epr , zero - point field , quantum gravity , mach s principle , holographic principle , fermat s last theorem , alternative gravity , bohmian mechanics .
pacs numbers .
4.50 , 4.60 |
grb 990510 was detected by batse on - board the compton gamma ray observatory and by the _
beppo_sax gamma ray burst monitor and wide field camera on 1999 may 10.36743 ut ( @xcite , dadina et al .
its fluence ( 2.5@xmath3 erg @xmath4 above 20 kev ) was relatively high ( @xcite ) .
follow up optical observations started @xmath5 hr later and revealed an @xmath6 ( @xcite ) optical transient , ot ( vreeswijk et al .
1999a ) , at the coordinates ( j2000 ) @xmath7 , @xmath8 ( @xcite ) ( galactic coordinates @xmath9 , @xmath10 ) .
1 shows the digital sky survey ii image of the field of grb 990510 , together with the european southern observatory ( eso ) very large telescope ( vlt ) image we obtained ( see below ) : the ot is clearly visible in the latter .
the ot showed initially a fairly slow flux decay @xmath11 ( galama et al .
1999 ) , which gradually steepened , @xmath12 after @xmath13 d ( @xcite ) , @xmath14 after @xmath15 d ( @xcite , bloom et al 1999 ) , @xmath16 after @xmath17 d ( @xcite , 1999b ) ) .
vreeswijk et al .
( 1999b ) detected fe ii and mg ii absorption lines in the optical spectrum of the afterglow .
this provides a lower limit of @xmath18 to the redshift , and a @xmath19ray energy of @xmath20 erg , in the case of isotropic emission .
polarization is one of the clearest signatures of synchrotron radiation , if this is produced by electrons gyrating in a magnetic field that is at least in part ordered .
polarization measurements can provide a crucial test of the synchrotron shock model ( @xcite ) .
an earlier attempt to measure the linear polarization of the optical afterglow of grb 990123 yielded only an upper limit ( @xcite ) of @xmath21% .
our observations of grb 990510 were obtained at eso s vlt
antu ( ut1 ) , equipped with the focal reducer / low dispersion spectrometer ( fors ) and bessel filter @xmath1 .
the ot associated with grb 990510 was observed @xmath0 hr after the burst , when the @xmath1-band magnitude was @xmath22 .
observations were performed in standard resolution mode with a scale of @xmath23/pixel ; the seeing was @xmath24 .
the observation log is reported in table1 .
.observation log for the polarimetric observation of the grb990510 field .
[ cols="^,^,^,^ " , ] -0.85truecm imaging polarimetry is achieved by the use of a wollaston prism splitting the image of each object in the field into the two orthogonal polarization components which appear in adjacent areas of the ccd image . for each position angle @xmath25
of the half wave plate rotator , we obtain two simultaneous images of cross polarization , at angles @xmath26 and @xmath27 .
relative photometry with respect to all the stars in the field was performed and each couple of simultaneous measurements at orthogonal angles was used to compute the points in fig . 2 ( see eq .
[ eq : sphi ] ) .
this technique removes any difference between the two optical paths ( ordinary and extraordinary ray ) and the polarization component introduced by galactic interstellar grains along the line of sight .
moreover , being based on relative photometry in simultaneous images , our measurements are insensitive to intrinsic variations in the optical transient flux ( @xmath28 magnitudes during the time span of our observations ) . with the same procedure , we observed also two polarimetric standard stars , bd135073 and bd125133 , in order to fix the offset between the polarization and the instrumental angles .
the data reduction was carried out with the eso
midas ( version 97nov ) system .
after bias subtraction , non uniformities were corrected using flat
fields obtained with the wollaston prism .
the flux of each point source in the field of view was derived by means of both aperture and profile fitting photometry by the daophotii package ( @xcite ) , as implemented in midas .
for relatively isolated stars the two techniques differ only by a few parts in a thousand . in order to evaluate the parameters describing the linear polarization of the objects , we compute , for each instrumental position angle @xmath26 , the quantity : @xmath29 where @xmath30 and @xmath31 are the intensities of the object measured in the two beams produced by the wollaston prism , and @xmath32 are the average ratios of the intensities of the stars in the field .
this corrects directly for the small instrumental polarization ( and , at least in part , for the possible interstellar polarization ) .
these field stars ( see fig.3 ) have been selected over a range in magnitude ( @xmath33 ) to check for possible non linearities .
since the interstellar polarization of any star in the field may be related to the patchy dust structure and/or to the star distance , we have verified that the result does not depend on which stars are chosen for the analysis . the parameter @xmath34 is related to the degree of linear polarization @xmath35 and to the position angle of the electric field vector @xmath36 by : @xmath37 @xmath35 and @xmath36 are evaluated by fitting a cosine curve to the observed values of @xmath34 .
the derived linear polarization of the ot of grb 990510 is @xmath38% ( 1@xmath39 error ) , at a position angle of @xmath40 .
the errors for the polarization level and position angle are computed propagating the photon noise of the observations and the contribution of the normalization to the stars in the field and of the calibration of the position angle .
the latter quantities , however , amounts to only a minor fraction of the quoted 1@xmath39 uncertainties . fig .
2 shows the data points and the best fit @xmath41 curve .
the statistical significance of this measurement is very high .
a potential problem is represented by a `` spurious '' polarization introduced by dust grains interposed along the line of sight , which may be preferentially aligned in one direction .
stanek et al .
( 1999b ) , using dust infrared emission maps ( @xcite ) , reported a substantial galactic absorption ( @xmath42 ) in the direction of grb 990510 .
the maps by dickey & lockman ( 1990 ) and by burstein & heiles ( 1982 ) give instead a somewhat lower value , @xmath43 and @xmath44 , respectively . applying an empirical relation ( @xcite )
this polarization can amount to @xmath45 , _
i.e. _ @xmath46% .
these are only statistical estimates and large variations on the main trend may be expected .
then a fraction , or even all the polarization of the ot could be caused by the passage of its light through the galactic ism .
however the normalization of the ot measurements to the stars in the field already corrects for the average interstellar polarization of these stars , even if this does not necessarily account for all the effects of the galactic ism along the line of sight to the ot ( e.g. the ism could be more distant than the stars , not inducing any polarization of their light ) . to check this possibility , we plot in fig .
3 the degree of polarization vs. the instrumental position angle for each star and for the ot .
all points in this figure have been derived avoiding to normalize with respect to other objects .
it is apparent that , while the position angle of all stars are consistent with being the same ( within 10 degrees ) , the ot clearly stands out .
the polarization position angle of stars close to the ot differs by @xmath47 from the position angle of the ot ( see fig .
this is contrary to what one would expect if the polarization of the ot were due to the galactic ism .
indeed , the higher polarization level measured for the ot when normalized to the stars in the same field implies that the ism actually _ somewhat de - polarizes _ the ot .
we therefore conclude that the ot , even if contaminated by interstellar polarization , must be intrinsically polarized to give the observed orientation .
we can place tight limits on the amount of absorption , and hence the associated polarization , that could be produced by interstellar material in the host galaxy of grb 990510 .
assuming that the intrinsic spectrum is a power law @xmath48 , we require that the fluxes measured simultaneously in the @xmath49 , @xmath50 , @xmath1 and @xmath51 band ( pietrzynski & udalski 1999 ; kaluzny et al .
1999 ; hjorth et al .
1999b ) lie on a power law curve .
this strongly limits the amount of the local extinction , affecting the flux at rest
frame frequencies of @xmath52 , i.e. in the uv , where extinction is more severe .
we find a maximum allowed value @xmath53 , corresponding to a maximum induced polarization level of @xmath54 .
incidentally , the best fit power law is obtained for @xmath55 , a galactic @xmath56 and @xmath57 .
this value of @xmath58 matches the predictions of the standard model for the decaying afterglow flux ( @xcite ) , which gives @xmath59 . for @xmath60 ,
the expected flux decay is in agreement with that measured at the time of the observations .
relativistic fireball models do explain the main properties of grbs and their afterglows ( @xcite )
. polarized optical synchrotron emission may be observable if : ( i ) the coherence length of the magnetic field in the fireball grows at a sizeable fraction of the speed of light ( gruzinov & waxman 1999 ; gruzinov 1999 ) or , ( ii ) the fireball is collimated ( hjorth et al .
1999a ) ( i.e. it is beamed ) .
therefore , measurements of optical polarization can provide constraints on the geometry of the emitting source .
additional information come from the afterglow light curve which shows a gradual steepening in the bands @xmath50 , @xmath1 and @xmath51 , which was never observed before ( marconi et al .
the observed steepening is almost wavelength independent , thus excluding that it could be entirely caused by a curved spectrum shifting in time rigidly to lower frequencies , in which case we ought to see the highest frequencies steepening first .
in addition , the @xmath61 and @xmath62 colors are changing very slowly during the evolution , indicating that the spectral slope is changing slowly with time .
these information suggest that the fireball is collimated in a jet .
the solid angle of the jet visible to the observer is limited to those regions making an angle smaller than @xmath63 with the line of sight .
as @xmath64 decreases , the visible solid angle increases as @xmath65 , until @xmath66 ( with @xmath67 being the cone of semi aperture angle and @xmath68 the angle between the cone axis and the line of sight ) .
when @xmath69 the observed solid angle remains constant , since the entire jet is visible . for @xmath64factors between @xmath70 and @xmath71 the observed solid angle increases somewhat slower than @xmath65 .
since the flux at the earth is proportional to the observed solid angle , we have two well defined behaviors of the light curve , corresponding to @xmath72 and @xmath73 , and a transition period of gradual steepening in between .
photons produced in regions at an angle @xmath63 with respect to the line of sight are emitted , in the comoving frame , at @xmath74 from the velocity vector .
a comoving observer at this angle can then see a compressed emitting region and a projected magnetic field structure with a preferred orientation .
if the gradual steepening of the light curve is due to the mechanism just mentioned , we would observe only some regions at a viewing angle @xmath63 , not all those we would see in axis
symmetric situation , and this asymmetry can be the cause of the observed linear polarization .
the above arguments suggest that we are observing , slightly off
axis , a collimated beam .
if this is the case , we would have a link between the flux decay behavior , the presence of polarization , and the degree of collimation , opening a new perspective for measuring the intrinsic power of grbs .
deeper understanding of polarization in grbs may come from future multi
filter observations and from spectropolarimetry .
frequency dependent polarization can in fact easily disentangle different components of polarization .
in addition , variability in the degree of polarization and its position angle is expected in such fastly evolving sources : therefore repeated observations of the same afterglow will also be important .
we thank the eso vlt service team , and in particular h. boehnhardt , f. bresolin , p. mller and g. rupprecht .
fh thanks the kind hospitality of the department of physics of the university of milan .
we also thank the referee j. hjorth for his helpful comments .
amati , l. , frontera , f. , costa , e. & feroci , m. , 1999 , gcn 317 axelrod , t. , mould , j & shmidt , b. , 1999 , gcn 315 bloom , j. s. , kulkarni , s. r. , djorgovski , s. , frail , d a. , axelrod , t.s . ,
mould , j.r . , & shmidt , b.p . , 1999 ,
gcn 323 burstein , d. , heiles , c. , 1987 , aj , 87 , 1165 dadina , m. , di ciolo , l. , coletta , a. , et al . , 1999 ,
iauc 7160 dickey , j. m. , lockman , f.j .
, 1990 , ara&a , 28 , 215 galama , t. j. , vreeswijk , p.m. , rol , e. et al . , 1999 , gcn 313 hiltner , w.a . , 1956 , apj , 2 , 389 hjorth , j. et al .
, 1999a , science , 283 , 2073 hjorth , j. , burud , i. , pizzella , a. , pedersen , h. , jaunsen , a.o .
& lindgren , b. , 1999b , gcn 320 kaluzny , j , garnavich , p.m. , stanek , k.z .
, pych w. & thompson , i. , 1999 , gcn 314 kippen , r. m. , 1999 , gcn 322 gruzinov , a. , 1999 , submitted to apj ( astro - ph/9905276 ) gruzinov , a. & waxman , e. , 1999 , apj , 511 , 852 marconi , g. , israel , g.l .
, lazzati , d. , covino , s. & ghisellini , g. , 1999a , gcn 329 marconi , g. , israel , g.l . ,
lazzati , d. , covino , s. & ghisellini , g. , 1999b , gcn 332 mszros , p. & rees , m. j. , 1997 , apj , 476 , 232 pietrzynski , g. & udalski , a. , 1999 , gcn 316 rees , m.j . & mszros , p. , 1992 , mnras , 258 , l41 sari , r. , piran , t. & narayan , r. , 1998 , apj , 497 , l17 schlegel , d. j. , finkbeiner , d.p .
& davis , m. , 1998 , apj , 500 , 525 serkowski , k. , mathewson , d.l . &
ford , v.l . , 1975 ,
apj , 196 , 261 stanek , k.z . , garnavich , p.m. , kaluzny , j. , pych , w. & thompson , i. , 1999b , gcn 318 stanek , k.z . , garnavich , p.m. , kaluzny , j. , et al
. 1999 , astro - ph/990534 stetson , p. b. , 1987 , pasp , 99 , 191 vietri , m. , 1997 , apj , 478 , l9 vreeswijk , p. m. , galama , t.j . ,
rol , e. et al .
, 1999a , gcn 310 vreeswijk , p. m. , galama , t.j .
, rol , e. et al .
, 1999b , gcn 324 waxman , e. , 1997 , apj , 485 , l5 | models for gamma ray burst afterglows envisage an hyper relativistic fireball that is decelerated in the ambient medium around the explosion site .
this interaction produces a shock wave which amplifies the magnetic field and accelerates electrons to relativistic energies , setting the conditions for an efficient production of synchrotron photons . if produced in a region of large
scale ordered magnetic field , synchrotron radiation can be highly polarized .
the optical transient associated with grb 990510 was observed @xmath0 hr after the event and linear polarization in the @xmath1 band was measured at a level of @xmath2 .
this is the first detection of linear polarization in the optical afterglow of a gamma ray burst .
we exclude that this polarization is due to dust in the interstellar material , either in our galaxy or in the host galaxy of the gamma ray burst .
these results provide important new evidence in favor of the synchrotron origin of the afterglow emission , and constrains the geometry of the fireball and/or magnetic field lines . |
recently andrei teleman considered moduli spaces of projectively anti - selfdual instantons in certain hermitian rank-2 bundles over a closed oriented 4-manifold with negative definite intersection form @xcite .
these play a role in his classification program on class vii surfaces @xcite@xcite
. however , in certain situations the instanton moduli spaces involved consist of projectively flat connections and therefore have very interesting topological implications . in this article
we will study these ` casson - type ' moduli spaces .
suppose @xmath8 is a hermitian rank-2 bundle with first chern - class a ( minimal ) characteristic vector @xmath9 of the intersection form .
in other words , it is the sum of elements @xmath10 in @xmath11 which induce a basis of @xmath12 diagonalising the intersection form ( because of donaldson s theorem @xcite )
. then for one possible value of a strictly negative second chern class @xmath13 the moduli space is compact ( independently of the riemannian metric ) . in particular ,
if the manifold has second betti - number @xmath2 divisible by 4 and first betti - number @xmath14 the instanton moduli space consists of projectively flat connections and has expected dimension zero .
this should be thought of as a ` casson - type ' moduli space because the holonomy yields a surjection onto the space of @xmath15 representations of @xmath16 with fixed stiefel - whitney class @xmath17 .
non - emptiness of the casson - type moduli space implies that none of the elements @xmath18 can be poincar dual to an element representable by a sphere , i.e. to an element in the image of the hurewicz homomorphism .
prasad and yeung @xcite constructed aspherical manifolds @xmath19 which are rational - cohomology complex projective planes , generalisations of mumford s fake projective plane @xcite .
if @xmath20 denotes this manifold with the opposite orientation , a natural candidate of a manifold for which the moduli space might be non - empty is given by the connected sum @xmath21 of 4 copies of @xmath20 , and a candidate of a manifold for which the casson - invariant can be defined is given by a ` ring of 4 copies of @xmath20 ' ( the last summand in the 4-fold connected sum @xmath22 is taken a connected sum with the first ) .
after recalling the gauge - theoretical situation considered in @xcite we show that if the casson - type moduli space is non - empty , then we can not have a connected sum decomposition @xmath4 unless both @xmath5 and @xmath6 are divisible by four .
in particular the moduli space for the above mentioned @xmath21 - ring is empty .
this result still leaves open the question of whether there is any @xmath0 with a non - empty casson - type moduli space .
we give therefore in some detail a possible construction of suitable @xmath7-manifolds @xmath0 ( along with the correct representations of @xmath16 ) .
we would like to point out that even though recent investigation leads us to believe that the casson - type invariant is vanishing @xcite , the casson - type moduli space may still be non - empty and is interesting from a topological perspective .
our construction also suggests the possibility of considering casson - type moduli spaces for manifolds with boundary .
a similar moduli space and invariant has been defined by ruberman and saveliev for @xmath23$]-homology hopf surfaces , going back to work of furuta and ohta @xcite , and for @xmath23$]-homology 4-tori @xcite .
our situation is simpler than their first mentioned situation because of the absence of reducibles in the moduli space due to the condition on @xmath2 .
the first author thanks simon donaldson for useful conversations .
the second author is grateful to andrei teleman for turning his interest to low - energy instantons and for a stimulating conversation on them , and also wishes to express his gratitude to stefan bauer for helpful conversations .
both authors thank kim fryshov profusely for invaluable advice and ideas .
we are also grateful to the referee for the care taken in helping us substantially improve the article .
after briefly recalling some general instanton gauge theory @xcite , and introducing our notations , we shall quickly turn to the special situation of ` low - energy instantons ' over negative definite 4-manifolds mentioned in the introduction .
we show that the gauge - theoretical situation is indeed relatively simple , indicate a definition of an invariant , and set up the correspondance of the moduli space to representation spaces of the fundamental group in @xmath15 .
let @xmath0 be a smooth riemannian four - manifold and @xmath8 a hermitian rank-2 bundle on @xmath0 .
let further @xmath24 be a fixed unitary connection in the associated determinant line bundle @xmath25 .
we define @xmath26 to be the affine space of unitary connections on @xmath27 which induce the fixed connection @xmath24 in @xmath28 .
this is an affine space over @xmath29 , the vector space of @xmath30-valued one - forms on @xmath0 .
let us denote by @xmath31 the principal @xmath32 bundle of frames in @xmath27 , and let @xmath33 be the bundle that is associated to @xmath31 via the projection @xmath34 , @xmath35
. the space @xmath36 of connections in the @xmath37 principal bundle @xmath33 and the space @xmath26 are naturally isomorphic .
if we interpret a connection @xmath38 as a @xmath37 connection via this isomorphism it is commonly called a projective connection .
the adjoint representation @xmath39 descends to a group isomorphim @xmath40 .
the associated real rank-3 bundle @xmath41 is just the bundle @xmath30 of traceless skew - symmetric endomorphisms of @xmath27 .
thus the space @xmath26 is also isomorphic to the space @xmath42 of linear connections in @xmath30 compatible with the metric .
we shall write @xmath43 for connections in the @xmath37 principal bundle and denote the associated connection in @xmath30 by the same symbol . should we mean the unitary connection which induces the connection @xmath24 in @xmath44 we will write @xmath45 instead
. let @xmath46 denote the group of automorphisms of @xmath27 of determinant @xmath47 .
it is called the ` gauge group ' .
this group equals the group of sections @xmath48 , where @xmath49 is given by conjugation .
we shall write @xmath50 for the quotient space @xmath51 .
a connection is called _ reducible _ if its stabiliser under the gauge group action equals the subgroup given by the centre @xmath52 which always operates trivially , otherwise _
irreducible_. equivalently , a connection @xmath45 is reducible if and only if there is a @xmath45 - parallel splitting of @xmath27 into two proper subbundles .
let us point out that the characteristic classes of the bundle @xmath30 are given by @xmath53 for a connection @xmath43 we consider the anti - selfduality equation @xmath54 where @xmath55 denotes the curvature form of the connection @xmath56 , and @xmath57 its self - dual part with respect to the hodge - star operator defined by the riemannian metric on @xmath0 .
the moduli space @xmath58 of antiself - dual connections , @xmath59 is the central object of study in instanton gauge theory .
this space is in general non - compact and there is a canonical `` uhlenbeck - compactification '' of it .
the anti - selfduality equations are elliptic , so fredholm theory provides finite dimensional local models for the moduli space . the often problematic aspect of donaldson theory is the need to deal with reducible connections and with a non - trivial compactification
. we will consider special situations where these problems do not occur .
we restrict now our attention to smooth riemannian four - manifolds @xmath0 with @xmath60 and @xmath61 .
according to donaldson s theorem @xcite the intersection form of such a four - manifold is diagonal .
let @xmath62 be a set of elements in @xmath11 which induce a basis of @xmath12 diagonalising the intersection form .
[ no reductions ] @xcite suppose the hermitian rank-2 bundle @xmath8 has first chern class @xmath63 and its second chern class is strictly negative , @xmath64
. then @xmath8 does not admit any topological decomposition @xmath65 into the sum of two complex line bundles . _
proof : _ suppose @xmath66 . then @xmath67 and @xmath68 for some @xmath69 .
therefore , @xmath70
@xmath71 let @xmath8 be as in the previous lemma .
then the moduli space @xmath72 does not admit reducibles .
for a connection @xmath73 chern - weil theory gives the following formula : @xmath74 in particular , for anti - selfdual connections the left hand side of this equation is always non - negative , and we can draw the following observation from the formula : @xcite 1 . for @xmath75 the moduli space
@xmath72 is always compact , independently of the chosen metric or any genericity argument .
in fact , the lower strata in the uhlenbeck - compactification consist of anti - selfdual connections in bundles @xmath76 with @xmath77 and @xmath78 for @xmath79 .
for @xmath63 we have @xmath80 .
thus , if @xmath81 and @xmath82 the moduli space @xmath72 will consist of projectively flat connections only . we recall the expected dimension of the moduli space @xmath72 .
it is given by the formula @xmath83 in particular it can happen that @xmath84 in the situation we consider , namely , @xmath60 , @xmath85 , and @xmath64 , the latter condition assuring that we are in the favorable situation of lemma [ no reductions ] .
+ interesting is the following special case of ` casson - type ' moduli spaces that we consider from now on : [ flat ] let @xmath0 be a negative definite riemannian four - manifold with strictly positive second betti - number @xmath2 divisible by four , and @xmath14 .
let @xmath8 be a hermitian rank-2 bundle with @xmath63 and with @xmath86 .
then the moduli space @xmath72 of projectively anti - selfdual connections in @xmath27 is compact and consists of irreducible projectively flat connections only , and is of expected dimension zero .
after suitable perturbations a gauge theoretic invariant can be defined in this situation : it is an algebraic count of a perturbed moduli space which consists of a finite number of points , the sign of each point is obtained by a natural orientation determined by the determinant line bundle of a family of elliptic operators .
this has been done in the meantime in @xcite , where it is shown that this invariant is actually zero .
we would like to emphasise that the vanishing of this invariant does nt imply emptiness of the unperturbed moduli space that we shall investigate further here .
suppose we are in the situation that our moduli space @xmath72 consists of flat connections in @xmath87 , as for instance in the last proposition .
then we must have @xmath88 by chern - weil theory .
the holonomy establishes a correspondance between flat connections in the oriented real rank-3 bundle @xmath89 and representations of the fundamental group @xmath16 in @xmath15 with a prescribed stiefel - whitney class .
more precisely , let @xmath90 be a representation of the fundamental group .
let @xmath91 be the universal covering of @xmath0 ; it is a @xmath16 principal bundle over @xmath0 .
we can form the associated oriented rank-3-bundle @xmath92 it admits a flat connection as it is a bundle associated to a principal bundle with discrete structure group .
therefore it has vanishing first pontryagin class , @xmath93 , by chern - weil theory .
its only other characteristic class @xcite is its second stiefel - whitney class @xmath94 .
therefore we will say that the representation @xmath95 has stiefel - whitney class @xmath96 if @xmath97 .
on the other hand , let @xmath98 be an oriented real rank-3 bundle with a flat connection @xmath56 .
then the holonomy of @xmath56 along a path only depends up to homotopy on the path , and therefore induces a representation @xmath99 .
in particular , the holonomy defines a reduction of the structure group to @xmath16 , and the bundle can therefore be reconstructed as @xmath100 .
in particular the representation @xmath101 has stiefel - whitney class @xmath102 .
the moduli space @xmath72 has been obtained by quotienting the space of antiself - dual connections in @xmath103 by the gauge group @xmath46 . from the perspective of the @xmath37 connections in @xmath30 this gauge group is not the most natural one . instead
, the group @xmath104 is the natural group of automorphisms of connections in @xmath30 .
not every element @xmath105 admits a lift to @xmath46 ; instead , there is a natural exact sequence @xmath106 quotienting by @xmath107 has the advantage of a simpler discussion of reducibles , as discussed above .
let us denote by @xmath108 the moduli space of anti - self dual connections in @xmath30 modulo the full gauge group @xmath109 .
then there is a branched covering @xmath110 with ` covering group ' @xmath111 .
let us denote by @xmath112 the space of representations of @xmath16 in @xmath15 up to conjugation and of stiefel - whitney class @xmath96 .
the above discussion implies that there is a homeomorphism @xmath113 where @xmath114 .
in particular , @xmath72 surjects onto @xmath112 .
we will use the above derived relation of the ` casson - type moduli space ' @xmath72 to the representation space @xmath112 to obtain the vanishing result which is mentioned in the introduction .
the above construction of the bundle @xmath115 associated to a representation @xmath116 is functorial in the following sense : [ naturality ] suppose we have a map @xmath117 between topological spaces , and @xmath90 a representation of the fundamental group of @xmath0 .
then there is a natural isomorphism @xmath118 between the pull - back of the bundle @xmath115 via @xmath119 and the bundle @xmath120 , where @xmath121 is the map induced by @xmath119 on the fundamental groups . _
proof : _ we have a commutative diagram @xmath122{\widetilde{x } } \arrow{s } \\ \node{w } \arrow{e , t}{f } \node[1]{x , } \end{diagram } \end{split}\ ] ] where the vertical maps are the universal coverings , and where @xmath123 is the unique map turning the diagram commutative ( we work in the category of pointed topological spaces here ) .
it is elementary to check that the map @xmath123 is equivariant with respect to the action of @xmath124 , where this group acts on @xmath91 via @xmath125 and the deck transformation group of @xmath91 .
the claimed isomorphism follows then from naturality of the associated bundle construction .
@xmath126 + [ restrictions ] suppose the four - manifold @xmath0 splits along the connected 3-manifold @xmath127 as @xmath128 into two four - manifolds @xmath129 and @xmath130
. then any representation @xmath131 induces representations @xmath132 via @xmath133 where the map @xmath134 is the inclusion . for these representations we have @xmath135 conversely , given representations @xmath132 such that @xmath136 , where @xmath137 denotes the inclusion , there is a representation @xmath116 inducing @xmath138 and @xmath139 via the respective restrictions . _
proof : _ this follows from the theorem of seifert and van kampen and the lemma above or , equivalently , by gluing connections .
@xmath126 + [ hurewicz ] let @xmath0 be a four - manifold with @xmath60 , and let @xmath96 be @xmath140 .
suppose there is a representation @xmath90 with fixed second stiefel - whitney class @xmath9 . then none of the poincar dual of the basis elements @xmath18 is in the image of the hurewicz - homomorphism @xmath141 . _ proof @xcite : _ suppose we have a map @xmath142 such that @xmath143 $ ] , where @xmath144 \in h_2(s^2;{\mathbb{z}})$ ] denotes the fundamental cycle of @xmath145 , and @xmath146 denotes the poincar dual of @xmath18
. then we have @xmath147 \rangle \equiv \langle \sum e_j , pd(e_i ) \rangle = e_i^2 = -1 \ ( mod \ 2 ) .\ ] ] on the other hand , by naturality of the cohomology - homology pairing , we get @xmath148 \rangle & = \langle w_2(v_\rho ) , f_*[s^2 ] \rangle = \langle f^ * w_2(v_\rho ) , [ s^2 ] \rangle \ . \end{split}\ ] ] but the above lemma [ naturality ] implies that @xmath149 .
as @xmath145 has trivial fundamental group the bundle @xmath150 is clearly the trivial bundle , so the left hand side of equation ( [ 0 ] ) must be zero modulo 2 , a contradiction to equation ( [ -1 ] ) .
@xmath126 + by hopf s theorem on the cokernel of the hurewicz - homomorphism , expressed in the exact sequence @xmath151 the fundamental group has to have non - trivial second homology in order to obtain a non - empty casson - type moduli space .
this proposition gives a topological significance of the zero - energy instantons : if the moduli space is non - empty then the elements @xmath146 are not representable by spheres ! one might wonder whether there exists any four - manifold where the elements @xmath146 are not representable by spheres .
certainly this can not be a simply connected four - manifold because of the hurewicz - isomorphism theorem .
interestingly , the answer is affirmative .
generalising mumford s fake projective plane @xcite , prasad and yeung have constructed manifolds with the rational cohomology of the complex projective space @xmath152 whose universal cover is the unit ball in @xmath153 @xcite .
such a manifold @xmath19 is therefore an eilenberg - maclane space @xmath154 .
now let @xmath155 be the four - manifold that we obtain from the connected sum of four @xmath20 , where we do again a connected sum of the last summand with the first .
the so obtained `` 4-@xmath20-ring '' is diffeomorphic to @xmath156 this manifold has negative definite intersection form and has betti - numbers @xmath157 and @xmath158 .
in addition , no element of @xmath159 is representable by a 2-sphere , so we get no obstruction to non - emptiness from proposition [ hurewicz ] . thus the four - manifold @xmath155 is a prototype of a four - manifold on which to consider the moduli space of @xmath37 instantons associated to the bundle @xmath160 with @xmath63 and @xmath161 ( and therefore of representations of @xmath162 with fixed stiefel - whitney class @xmath163 ) .
however , as we will see , there are no such instantons .
[ doldwhitneyapplication ] let @xmath0 be a smooth closed negative definite four - manifold . if there is a representation @xmath164 with stiefel - whitney class @xmath165 , then the second betti - number @xmath2 must be divisible by four .
_ proof : _ the bundle @xmath115 has @xmath166 and vanishing first pontryagin - class @xmath93 because this bundle admits a flat connection .
now the dold - whitney theorem @xcite states that the second stiefel - whitney class @xmath167 and the first pontryagin class @xmath168 of any oriented real rank-3 bundle satisfy the equation @xmath169 here @xmath170 denotes the pontryagin square , a lift of the cup - product squaring @xmath171 to the coefficient group @xmath172 .
if the class @xmath173 is the mod-2 reduction of an integral class @xmath174 then the pontryagin square is simply the mod-4 reduction of the square of @xmath175 , i.e. @xmath176 in our case the dold - whitney theorem thus implies that @xmath177 @xmath126 + hence we obtain the following let @xmath0 be a four - manifold with negative definite intersection form and suppose it admits a connected sum decomposition @xmath178 .
suppose @xmath116 is a representation of the fundamental group of @xmath0 with fixed stiefel - whitney class @xmath179 .
then both @xmath5 and @xmath6 must be divisible by four .
_ proof : _ note first that the intersection form of both @xmath129 and @xmath130 must be diagonal .
this follows from eichler s theorem on unique decomposition of symmetric definite forms over @xmath180 , see @xcite .
therefore the basis vectors @xmath62 of @xmath11 are simply given by the union of basis vectors @xmath181 of @xmath182 , diagonalising the intersection form of @xmath129 , and basis vectors @xmath183 of @xmath184 , diagonalising the intersection form of @xmath130 .
note that @xmath185 .
the above proposition [ restrictions ] now applies yielding representations @xmath186 .
its second stiefel - whitney class computes , using the above equation ( [ restriction ] ) , @xmath187 and likewise for @xmath188 .
the above theorem therefore concludes the proof .
@xmath126 + this implies that the above considered manifold @xmath189 does not admit a representation @xmath116 with stiefel - whitney class being the mod-2 reduction of the sum of basis elements diagonalising the intersection form . as a ` converse ' to the above vanishing theorem ,
suppose we are given a connected sum @xmath190 and representations @xmath132 with the desired stiefel - whitney classes on @xmath191 . according to proposition [ restrictions ] ,
we obtain the representation @xmath192 which has the desired stiefel - whitney class .
this is in contrast to well - known vanishing theorems for connected sums of manifolds with @xmath193 as in ( * ? ? ?
* theorem 9.3.4 and , in particular , proposition 9.3.7 ) .
there is much interest in the relationship between the fundamental group of a @xmath7-manifold and its intersection form .
the casson - type invariant considered in this paper gives rise to the natural question of whether there exists _ any _
@xmath7-manifold @xmath0 with non - empty casson - type moduli space . in this section
we describe a construction that we hope will provide the first examples of such manifolds , by indicating how to construct non - empty representation spaces @xmath194 .
let @xmath196 be a smooth immersion of @xmath197 @xmath195-spheres such that any points of self - intersection of @xmath198 occur with negative sign and between two branches of the same component of @xmath198 .
suppose there are @xmath199 self - intersections .
blowing up @xmath199 times and taking the proper transform we obtain an @xmath197-component embedded link @xmath200 .
each component of @xmath201 intersects each exceptional sphere of @xmath202 either at no points or at one point positively and at one point negatively ( this is because each intersection point of @xmath198 occurred within a single component and with negative sign ) .
hence each component of @xmath201 is trivial homologically and so the embedding of @xmath201 extends to a @xmath203-neighbourhood .
we do surgery on @xmath201 by removing @xmath204 and gluing in @xmath205 .
call the resulting @xmath7-manifold @xmath0 .
the construction of @xmath0 was suggested by kim fryshov .
it turns out to be very suited to our purposes ; we have so we shall be done if we can find @xmath199 embedded tori in @xmath0 which are pairwise disjoint and which each have self - intersection @xmath216 .
figure [ torusbasis ] shows how to find these tori .
working inside @xmath215 , each exceptional sphere @xmath27 intersects @xmath201 transversely in two points .
connect these two points by a path on @xmath201 .
the @xmath203-neighbourhood of @xmath201 pulls back to a trivial @xmath203-bundle over the path .
the fibres over the two endpoints can be identified with neighbourhoods of these two points in @xmath27 . removing these neighbourhoods from @xmath27 we get a sphere with two discs removed and we take the union of this with the @xmath217 boundaries of all the fibres of the @xmath203-bundle over the path .
we have shown how to associate to a given immersed @xmath195-link @xmath218 with only negative self - intersections and disjoint components , a smooth @xmath7-manifold @xmath219 which is diagonal and negative definite , with basis elements of @xmath220 represented by embedded tori .
[ [ so3-representations - of - pi_1-and - presentations - of-2-links ] ] @xmath15 representations of @xmath221 and presentations of @xmath195-links ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ using the same notation as in the previous subsection , we give a method to describe links @xmath198 that come with representations @xmath222 with the correct stiefel - whitney class @xmath223 .
this method may not at first appear entirely general , but we show that if there is such a link @xmath198 then it must admit a description of this form .
1 . @xmath226-handles ( circle creation ) .
2 . simple crossing changes ( see figure [ crossing ] ) .
ribbon - type reidemeister moves of type ii ( see figure [ ribbon ] ) .
ribbon - type @xmath47-handle addition ( see figure [ ribbon ] ) .
[ [ representations - of - the - fundamental - group - and - ribbon - presentations - of-2-links ] ] representations of the fundamental group and ribbon presentations of @xmath195-links ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ for notation , let @xmath228 be a height function corresponding to lemma [ pres_lem ] with exactly @xmath195 critical points that restricts to a morse function on @xmath198 such that all the @xmath229-handles of @xmath198 occur in @xmath230 and the self - intersections of @xmath198 occur in @xmath231 . then there is a morse function on the blow - up @xmath232 with one maximum and one minimum , and @xmath199 index @xmath195 critical points .
these index 2 critical points all occur at @xmath233 , and h restricts to a morse function on the proper transform @xmath201 with the @xmath229-handles of @xmath201 occurring in @xmath234 .
we can use the same movie of @xmath198 to describe the embedding of @xmath201 .
recall that @xmath235 .
we compute @xmath236 using the van kampen theorem .
first note that @xmath237 is the boundary connect sum of @xmath199 copies of the complement of @xmath195 fibres in the @xmath203-bundle over @xmath145 of euler class @xmath216 , and @xmath238 copies of @xmath239 where the @xmath203 with @xmath240 is trivially embedded .
here @xmath199 is the number of self - intersections of @xmath198 ( and hence the number of blow - ups required on the way to constructing @xmath219 ) and @xmath238 is the number of extra @xmath226-handles used in the movie presentation of @xmath198 satisfying lemma [ pres_lem ] . since by assumption
@xmath219 has a non - empty casson - type moduli space and dim@xmath241 , we can write @xmath242 by theorem [ doldwhitneyapplication ] .
it is easy to compute that @xmath243 is the free ( non - abelian ) group on @xmath244 generators .
we fix representatives of a basis for this group as simple loops coming down from infinity , linking the relevant circle by small meridians and heading back up again . for each of the @xmath224 generators coming from the blowups
we allow ourselves two representatives - one for each circle .
note that our representatives live in the _ boundary _ of @xmath237 . to get the space @xmath245 we attach the complements of some @xmath47-handles to @xmath246 .
what this means is that for every @xmath47-handle of @xmath201 , we glue a @xmath247 to @xmath237 , via a homeomorphism of @xmath248 with a subset of @xmath249 .
( all discs in this discussion are trivially embedded ) . since @xmath250 , @xmath251 , and
the map on @xmath221 induced by inclusion is onto , the van kampen theorem tells us that adding the complement of a @xmath47-handle adds a single , possibly trivial , relation to @xmath221 . in other words ,
we obtain a presentation of @xmath252 with @xmath244 generators and as many relators as there are @xmath47-handles .
since we obtain @xmath253 from @xmath245 by gluing on the complement of some trivially embedded @xmath203 s ( one for each @xmath195-handle of @xmath201 ) in @xmath254 , it follows that @xmath255 .
hence we have a presentation of @xmath256 .
now by assumption , @xmath219 has a non - empty casson - type moduli space , so we choose some representation @xmath257 that has the correct associated characteristic classes .
each generator of the presentation is associated to some circle or hopf link in figure [ 0-handles ] .
we decorate each circle or hopf link with the image of the associated generator under @xmath95 .
we call these images @xmath258 .
each @xmath47-handle complement that we attach appears in the movie of @xmath198 as a ribbon - type @xmath47-handle addition as illustrated in figure [ ribbon ] .
once we have added each ribbon - type handle then by assumption we have an unlink .
[ [ representations - of - the - fundamental - group - and - a - singular - link - diagram ] ] representations of the fundamental group and a singular link diagram ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 1 .
each component of @xmath198 has genus @xmath226 .
2 . suppose two circles of figure [ 0-handles ] are joined by three paths of @xmath47-handle cores @xmath262 in the singular diagram @xmath261 .
if @xmath262 meet the first circle in three points that go clockwise ( respectively anticlockwise ) around the circle , then @xmath262 must meet the second circle in three points that go anticlockwise ( respectively clockwise ) around that circle . 1 .
self - intersections of @xmath198 only occur within a component and not between two components of the preimage of @xmath198 .
the singular diagram @xmath261 describes an obvious singular link in @xmath263 .
given a hopf link in figure [ 0-handles ] , we require that the two circles comprising it are part of the same component in this singular link . 1 .
the representation + @xmath264 + determined by the labelling @xmath258 factors through + @xmath265 2 .
each circle in figure [ 0-handles ] bounds an obvious oriented disc which has no double points when projected to the plane of the diagram .
consider a core of a @xmath47-handle @xmath56 in the singular link diagram @xmath261 .
suppose @xmath56 connects circles decorated by @xmath15 elements @xmath266 and @xmath267 , and that the arc , given the orientation from @xmath266 to @xmath267 , intersects discs bounded by circles which are decorated by elements @xmath268 .
define the element @xmath269 , where the @xmath270 index is the sign of the intersection of the arc with the disc .
we require + @xmath271 -handle , which will take place within the dotted circle .
we have indicated @xmath7 generators of @xmath273 . by the van kampen theorem , adding the @xmath47-handle imposes the relation that the rightmost generator is a conjugate of the leftmost generator as in item @xmath7 of our checklist .
thinking of the ribbon as a thickened arc , we note that this calculation does not depend on whether the arc is locally knotted , but only on the order in and parity with which it intersects the discs bounded by the @xmath226-handles .
also , since a small loop encircling both strands of a ribbon clearly bounds a disc in @xmath245 , it is also immaterial how the arcs link each other.,title="fig:",width=384,height=240 ] 1 .
the representation @xmath274 has the correct stiefel - whitney class @xmath223 .
2 . * for each hopf link of figure [ 0-handles ] , choose a path of cores of @xmath47-handles in the singular diagram @xmath261 which connects the circles of the hopf link .
( such a path exists by lemma [ selfint ] ) .
+ say it consists of cores @xmath275 .
we order and orient these cores so that the start point of @xmath276 and the end point of @xmath277 are on different components of the hopf link and the end point of @xmath278 is on the same circle of figure [ 0-handles ] as the start point of @xmath279 for @xmath280 .
write @xmath266 for the element decorating the hopf link
. then we require that + @xmath281 * the elements @xmath282 are each conjugate to the element @xmath283 ( in other words each element @xmath284 is a rotation by @xmath285 radians ) . by naturality
, this means that the map @xmath288 has to give the non - trivial flat bundle over @xmath289 .
given a basis for @xmath290 this is equivalent to asking that @xmath291 sends each of the two basis elements to rotations by @xmath285 , but around orthogonal axes .
for each basis element of @xmath292 , there is an associated hopf link in figure [ 0-handles ] .
say the hopf link is decorated by @xmath293 and there is a path connecting the two components of the hopf link as in lemma [ selfint ] .
consider the loop which we gave as a generator of @xmath294 corresponding to the hopf link , and a loop based at @xmath295 which goes down to the path and follows it around until returning to the hopf link and then returns back up to @xmath295 .
this gives two basis elements for @xmath221 of a @xmath289 representing the basis element of @xmath296 .
the former is sent to @xmath266 by @xmath291 and the latter is sent to @xmath297 . since necessarily @xmath297 commutes with @xmath266 , the requirement that @xmath298 , ensures that @xmath297 is a rotation by @xmath285 around an axis orthogonal to that of @xmath266 .
if we can find a presentation of some @xmath198 with decoration by some @xmath260 satisfying the conditions of lemmas [ genus0 ] , [ selfint ] , [ relatorsgoto0 ] , and [ swclass ] , then we have seen that we can construct a negative definite @xmath7-manifold @xmath219 with non - empty casson - type moduli space . in particular
we have exhibited a particular representation giving such a presentation of @xmath198 is equivalent to giving first the singular link diagram @xmath261 and then giving a framing to the cores of each @xmath47-handle .
therefore we have the following : [ constructionresult ] suppose we give a singular link diagram @xmath261 in the sense of definition [ g ] , starting with figure [ 0-handles ] and then adding arcs which begin and end at points of figure [ 0-handles ] .
further suppose that there is a decoration of @xmath261 by @xmath260 that satisfies the conditions on the singular link diagrams given as the latter statements of lemmas [ genus0 ] , [ selfint ] , [ relatorsgoto0 ] , and [ swclass ] .
then , if there exists a framing of the arcs of @xmath261 such that the corresponding @xmath47-handle additions to figure [ 0-handles ] gives a diagram of a trivial link , there exists a @xmath7-manifold with non - empty casson - type moduli space .
@xmath126 + the symmetry group on @xmath7 elements @xmath300 can be embedded in @xmath15 as the rotational symmetry group of a cube . under this embedding ,
all elements of @xmath301 are taken to rotations by @xmath285 around some axis . in figure
[ example ] we have given an example of a diagram ( of labelled hopf links , simple circles , and arcs ) satisfying all the conditions of theorem [ constructionresult ] .
the group element decorations of the simple circles and the hopf links are given in the cycle notation for @xmath302 . if we can find a way to add more arcs , each satisfying the conditions of theorem [ constructionresult ] ( the steifel - whitney condition of lemma [ swclass ] has already been satisfied in the diagram )
such that when we replace each arc by a ribbon we get the unlink , then we will have described an immersion @xmath303 such that @xmath219 has non - empty casson - type moduli space .
[ onepoint ] if figure [ example ] is an intermediary diagram of a movie presentation of an immersed @xmath198 satisfying theorem [ constructionresult ] , then @xmath219 has exactly @xmath47 point in the representation space @xmath112 .
recent discoveries @xcite have indicated that the invariant defined as the signed count of the casson moduli space may always be @xmath226 .
results such as proposition [ onepoint ] are still valuable as they may be useful in showing that links are not slice ( for more in this direction see @xcite ) .
suppose that we have some new decoration satisfying theorem [ constructionresult ] .
call the decorating elements of @xmath15 @xmath305 where the initials stand for _ _ t__op , _ _
b__ottom , _ _
l__eft , _ _ r__ight .
each of the four decorations is a rotation by @xmath285 around some axis , so each element is equivalent to a choice of axis , and we use the same labels for these axes . by lemma [ swclass ]
, we must have @xmath306 perpendicular to @xmath307 and @xmath308 perpendicular to @xmath309 .
there is an arc connecting @xmath306 to @xmath307 . by condition lemma [ relatorsgoto0 ]
we can interpret this as meaning that the unique axis perpendicular to both @xmath308 and @xmath309 lies in the same plane as @xmath306 and @xmath307 and is at an angle of @xmath310 to both of them .
similarly , there is an arc connecting @xmath308 and @xmath309 , which implies that the axis perpendicular to @xmath306 and @xmath307 is in the same plane as @xmath308 and @xmath309 and at an angle of @xmath311 to both of them .
it is a simple matter to convince oneself that any two ordered pairs of ordered pairs of perpendicular axes satisfying the conditions of the previous paragraph must be equivalent via the action of an element of @xmath15 .
hence , up to conjugation , there is exactly @xmath47 representation @xmath304 of the correct characteristic class . @xmath126
+ 99999 a. dold , h. whitney , _ classification of oriented sphere bundles over a 4-complex _ , the annals of mathematics vol .
3 , ( 1959 ) , 667 - 677 .
s. donaldson , _ an application of gauge theory to four - dimensional topology _ , j. differential geom .
18 , no . 2 ( 1983 ) , 279 - 315 . s. donaldson , _ the orientation of yang - mills moduli spaces and 4-manifold topology _ , j. differential geom . 26 , no . 3 ( 1987 ) , 397 - 426
s. donaldson , p.b .
kronheimer , _ the geometry of four - manifolds _ , oxford mathematical monographs , ( 1990 ) .
m. furuta , h. ohta , _
differentiable structures on punctured 4-manifolds _ , topology and its applications , no .
51 ( 1993 ) , 291 - 301 .
d. husemoller , j. milnor , _ symmetric bilinear forms _ , ergebnisse der mathematik und ihrer grenzgebiete , band 73 ( 1973 ) .
p. b. kronheimer , _ four - manifold invariants from higher - rank bundles _ , j. differential geom .
70 , no . 1 ( 2005 ) , 59 - 112 .
a. lobb , _ the dold - whitney theorem and slicing 2-component links _ , preprint d. mumford , _ an algebraic surface with @xmath312 ample , @xmath313 , @xmath314 .
_ , amer . j. math . 101 ( 1979 ) , 233 - 244 .
g. prasad ; s .- k .
yeung , _ fake projective planes _ , invent .
168 , no . 2 ( 2007 ) , 321 - 370 .
d. ruberman , n. saveliev , _ casson - type invariants in dimension four _ , geometry and topology of manifolds , fields inst . commun .
47 , ams ( 2005 ) , 281 - 306 .
a. teleman , _ harmonic sections in sphere bundles , normal neighborhoods of reduction loci , and instanton moduli spaces on definite 4-manifolds _ , geometry & topology 11 ( 2007 ) , 1681 - 1730 .
a. teleman , _ donaldson theory on non - khlerian surfaces and class vii surfaces with @xmath315 _ , invent .
162 ( 2006 ) , 493 - 521 .
a. teleman , _ gauge theoretical methods in the classification of non - khlerian surfaces _ ,
preprint 2007 , to appear in postnikov memorial volume , banach center publications .
r. zentner , _ a vanishing result for a casson - type instanton invariant _ , in preparation ( 2009 ) . | recently andrei teleman considered instanton moduli spaces over negative definite four - manifolds @xmath0 with @xmath1 .
if @xmath2 is divisible by four and @xmath3 a gauge - theoretic invariant can be defined ; it is a count of flat connections modulo the gauge group .
our first result shows that if such a moduli space is non - empty and the manifold admits a connected sum decomposition @xmath4 then both @xmath5 and @xmath6 are divisible by four ; this rules out a previously naturally appearing source of @xmath7-manifolds with non - empty moduli space .
we give in some detail a construction of negative definite @xmath7-manifolds which we expect will eventually provide examples of manifolds with non - empty moduli space . |
advances in metrology are pivotal to improve measurement standards , to develop ultrasensitive technologies for defence and healthcare , and to push the boundaries of science , as demonstrated by the detection of gravitational waves @xcite . in a typical metrological setting
, an unknown parameter @xmath0 is dynamically imprinted on a suitably prepared probe .
we can think e.g. of a two - level spin undergoing a unitary phase shift @xmath3 . by subsequently interrogating the probe
one builds an estimate @xmath4 for the parameter @xcite .
the corresponding mean - square error @xmath5 can be reduced , for instance , by using @xmath1 uncorrelated identical probes . in that case
, @xmath6 scales asymptotically as @xmath7 , which is referred to as the standard quantum limit @xcite .
however , if those @xmath1 probes were prepared in an entangled state , the resulting uncertainty could be further reduced by an additional factor of @xmath8 , leading to @xmath9 . this ultimate quantum enhancement in resolution
is termed heisenberg limit and incarnates the _ holy grail _ of quantum metrology @xcite . in practice
, the unitary dynamics of the probe will be distorted by noise , due to unavoidable interactions with its surroundings .
unfortunately , the metrological advantage of entangled probes over separable ones vanishes for most types of uncorrelated noise , such as spontaneous emission , depolarizing noise @xcite , or phase damping @xcite .
entanglement may remain advantageous though , provided one gains precise control over the noise strength , and only for limited cases such as time - inhomogeneous phase - covariant noise @xcite , transversal noise @xcite , or when error - correction protocols may be used @xcite .
creating entangled states with a large number of particles is anyhow a costly process , limited by technological constraints @xcite .
furthermore , to fully harness the metrological power of entanglement in presence of noise , collective measurements on all @xmath1 probes at the output would be generally required @xcite .
this contrasts with the noiseless scenario , in which separable measurements ( i.e. , performed locally on each probe ) suffice to attain the heisenberg scaling @xcite .
one can try to circumvent these problems by devising an alternative _ sequential _ or ` multi - round ' strategy , in which the parameter - imprinting unitary acts @xmath1 consecutive times on a single probe before performing the final measurement . in absence of noise , this sequential setting is formally equivalent to the parallel one @xcite , the only difference being that quantum _ coherence _
@xcite takes over the instrumental role of entanglement .
the sequential scheme seems more appealing from a practical viewpoint , as only a single probe needs to be addressed in both state preparation and final interrogation @xcite .
however , the heisenberg scaling of the precision can not be maintained asymptotically in the sequential scenario either , once again due to the detrimental effects of noise .
given the severe limitations that environmental disturbance places on quantum - enhanced metrology , for practical purposes it seems best to give up the prospect of super - classical _ asymptotic _ scaling of the resolution and to concentrate instead in using the _ finite _ resources available as efficiently as possible .
in this paper , we explore the optimization of phase estimation with a two - level probe , in the presence of _ unital phase - covariant _ noise . to that end , in sec .
[ sec : noise ] we introduce a simple versatile model in which the noise is intrinsically accounted for : we take the generator @xmath10 of the phase shift to be partly unknown and sample instances of it from some probability distribution .
the ensuing average mimics the environmental effects . in sec . [
sec : sens ] we calculate the _ quantum fisher information _ ( qfi ) @xmath11 @xcite , which can be meaningfully regarded as a quantitative benchmark for the optimal estimation sensitivity , and derive a close - fitting lower bound @xmath12 to it .
both quantities grow quadratically for small @xmath1 , reach a maximum at some @xmath2 , and decay to zero as @xmath1 increases further .
in particular , we obtain @xmath2 from @xmath13 in terms of parameters directly accessible via process tomography , giving a useful prescription for practical phase estimation with a large guaranteed sensitivity .
we do this for any unital phase - covariant qubit channel , hence delivering results widely applicable to a broad range of relevant physical processes , including those in which noise effects of the depolarizing type are dominant , such as spin - lattice relaxation at room temperature . in sec .
[ sec : ex ] we then illustrate our results by choosing a specific distribution for the stochastic generator .
we compare the qfi in the sequential setting ( with and without passive correlated ancillas ) with the actual phase sensitivity of given feasible measurements . for completeness
, we also compute the qfi analytically in a parallel - entangled setting starting from an @xmath1-qubit ghz state .
although the qfi exhibits an asymptotic linear scaling in @xmath8 in such setting , we find that entangled probes may provide no practical advantage when their interrogation is restricted to measurements of local observables on each individual qubit .
in fact , in such case the sensitivity for the parallel - entangled strategy reduces to that of the sequential one , where the ` number of probes ' comes to play the role of the ` number of rounds ' .
our analysis , summarized in sec .
[ sec : d ] , reveals feasible solutions for quantum metrology based on the little - studied sequential paradigm ( possibly supplemented by a passive ancilla ) , robust even under sizeable levels of noise .
let us start by introducing our model for ease of illustration . in the sequential scenario ,
a two - level probe undergoes a sequence of @xmath1 phase shifts @xmath14 before being interrogated .
the generator can be written as @xmath15 , where the axis @xmath16 is sampled from some normalized probability distribution @xmath17 and @xmath18 is the vector of the three pauli matrices .
thus the phase - imprinting operation @xmath19 , which transforms the probe state at each step , is @xmath20 the resulting qubit channel is completely positive , trace preserving , _ unital _
( i.e. @xmath21 ) , and contractive @xcite , i.e. any state will be asymptotically mapped to @xmath22 ; note also that eq .
( [ eq : channel ] ) is akin to the classical simulation method of ref .
@xcite . without loss of generality
we may take the average rotation axis @xmath23 proportional to @xmath24 , hence restricting to probability distributions @xmath25 with axial symmetry on the bloch sphere , so that our qubit channel @xmath26 is _ phase - covariant _ , as it commutes with @xmath27 @xcite .
physically , we may think , for instance , of the free evolution of a nuclear spin with gyromagnetic ratio @xmath28 in an external magnetic field @xmath29 pointing along @xmath30 . in that case @xmath31 , so that @xmath32 , where @xmath33 is some coarse - grained time resolution , of the same order as the thermal fluctuations of the environment .
the interactions with the surrounding nuclei at large temperatures result in random changes of the net direction of the magnetic field on our spin .
this gives rise to a relaxation process towards the thermal equilibrium state @xmath34 .
if the direction of @xmath29 were kept fixed and the environmental effects were accounted for by a fluctuating magnetic field intensity @xmath35 , the model would realize pure dephasing instead @xcite , which is often dominant at short time scales . while the model in eq .
( [ eq : channel ] ) can be conveniently adopted as a physical example to focus our analysis , the results derived in this paper will hold for a more general class of channels , namely all the unital phase - covariant qubit channels , whose description is recalled here . given a single - qubit state @xmath36 , any single - qubit channel maps the bloch vector @xmath37 into @xmath38 , where @xmath39 is a @xmath40 real distortion matrix , and @xmath41 is a displacement vector . for the most general unital phase - covariant channel @xcite , @xmath42 and @xmath43
these channels encode information about @xmath0 not only in the rotation of the bloch ball by a function @xmath44 , but also in its deformation , through the singular values of @xmath45 , namely @xmath46 and the doubly - degenerate @xmath47 , which must satisfy @xmath48 and @xmath49 ( implying @xmath50 ) for the map to be completely positive @xcite .
( [ eq : phase_covariant ] ) thus generalizes the canonical phase - covariant channels like phase damping , for which @xmath51 @xcite . in what follows
we shall work with the @xmath52 liouville representation @xmath53 , which acts on vectorizations @xmath54 of any density matrix @xmath55 in the computational basis as @xmath56 , where @xmath53 writes as @xmath57
we will now analyze the sensitivity for estimating a phase imprinted by any unital phase - covariant qubit channel , starting with the sequential estimation setting . recall that the estimate @xmath58 is built by measuring some observable @xmath59 on the final state of the probe @xmath60 after @xmath8 successive applications of the channel .
we can then gauge the phase sensitivity of @xmath59 as @xmath61 where @xmath62 and @xmath63 @xcite .
in general , @xmath64 , where @xmath65 is the _ classical _ fisher information of a projective measurement onto the eigenstates @xmath66 of the observable @xmath67 ( assumed non - degenerate ) @xcite . in particular , one can verify that equality holds for all the explicit examples presented later in the paper , henceforth we shall simply refer to @xmath68 as the phase sensitivity associated to the ( generally suboptimal ) observable @xmath69 .
the qfi @xmath70 , which captures geometrically the rate of change of the evolved probe state @xmath71 under an infinitesimal variation of the parameter @xmath72 , corresponds to the classical fisher information of an _ optimal _ observable ( i.e. @xmath73 ) and is thus a meaningful benchmark for the best estimation protocol .
such an optimal observable is diagonal in the eigenbasis of the symmetric logarithmic derivative ( sld ) @xmath74 , defined implicitly by @xcite @xmath75 note that the prominent role of the qfi @xmath11 in quantum estimation theory is well established in an asymptotic setting , by virtue of its appearance in the quantum cramr - rao bound , @xcite , @xmath76 , which becomes tight in the limit of a large number @xmath77 of independent repetitions .
however , the qfi also enters in both the van trees inequality @xcite and the ziv - zakai bound @xcite , which can provide tighter and more versatile bounds on the mean - square error @xmath78 in the relevant case of finite @xmath79 ( including bayesian settings ) .
therefore , we shall adopt the qfi as our main figure of merit , in keeping with the quantum metrology bulk literature ( see e.g. @xcite for a recent discussion ) and in compliance with the spirit of this paper , which focuses on the use of finite resources to retain quantum enhancements in the estimation sensitivity .
we will also assume maximization of @xmath70 over the initial state of the probe unless stated otherwise . to calculate the qfi @xmath80 we make use of the formula @xmath81 where @xmath82 and @xmath83 are , respectively , the eigenvectors and eigenvalues of @xmath84 @xcite ( excluding terms with @xmath85 from the sum ) , and we have made the dependence of the qfi on @xmath72 and @xmath8 explicit . by preparing the probe in the optimal , maximally coherent state @xmath86 @xcite
, we obtain exactly @xmath87 .
\label{eq : qfi_formula_general}\ ] ] the qfi thus grows as @xmath88 for small @xmath8 , reaches a peak at some @xmath2 , and then decays asymptotically to zero .
a similar qualitative behaviour has been recently reported under other types of ( non - unital ) noise , such as erasure , spontaneous emission , and phase damping @xcite . in order to optimize our sequential protocol
, we need a practical way to determine @xmath2 . to that end , we aim to bound the qfi from below . in general , one may use @xmath89 @xcite , which is not necessarily tight . luckily , for the family of channels encompassed by our @xmath90 , a tighter bound @xmath91 stemming from @xmath92 , where @xmath93 stands for the operator norm , can be established ( see appendix [ app : bound ] ) .
concretely , we will thus define @xmath94 \leq f_n(\varphi ) , \label{eq : bound_result}\ ] ] which closely follows the behaviour of the qfi , as confirmed by our numerical analysis ( cf .
[ fig1 ] ) . maximizing @xmath95 yields @xmath96 ( rounded to the nearest integer and using natural logarithm ) .
this approximation to the optimal ` sampling time ' @xmath2 only depends on @xmath97 , which is experimentally accessible through process tomography @xcite .
( solid ) and lower bound @xmath95 ( dashed ) versus number of rounds @xmath1 under the channel of eq . for the von mises - fisher distribution @xmath98 , with @xmath99 and @xmath100 , and
probe initialized in the @xmath101 state .
the maximum of @xmath95 is at @xmath102 , only one round further the actual maximum of @xmath103 .
@xmath1 is discrete and the continuous lines are a guide to the eye .
all the quantities plotted are dimensionless . ]
it follows that there exist observables with a phase sensitivity of _ at least _
@xmath95 , which could be made grow above @xmath104 by just running the sequential protocol for @xmath105 iterations .
once the precision has been optimized at the single - probe level , it can be scaled - up `` classically '' by increasing the number of independent probes .
we emphasize that eqs . and apply _ generally _ to any phase - covariant channel preserving the identity .
therefore , we have provided useful guidelines for precise phase estimation under a wide class of physically motivated noise models .
to illustrate our results , we shall pick the von mises - fisher distribution @xcite @xmath106 for our random generator @xmath107 in eq . .
this distribution can be seen as the counterpart of a gaussian over the bloch sphere : it becomes uniform for @xmath108 ( i.e. @xmath109 is equally likely to point in any direction , as in a black - box scenario @xcite ) , whereas it localizes sharply around the @xmath110axis for @xmath111 . in appendices [ app : kraus ] and
[ app : sequential ] , we provide an explicit operator - sum representation for this choice of @xmath90 , as well as expressions for the resultant @xmath53 , @xmath103 , and @xmath95 .
recall that saturating the precision bound set by the qfi requires interrogating the probe in the eigenbasis of the sld .
however , the sld basis usually depends explicitly on the unknown parameter @xmath0 ( and , in this case , on the noise parameter @xmath112 ) so that , in practice , one would need to implement adaptive feedforward estimation procedures @xcite , or could have to resort to sub - optimal phase - independent estimators @xmath59 .
it is therefore particularly interesting to see how the sensitivity of _ accessible _ observables compares with the qfi .
we can study , for instance , the phase sensitivity of @xmath113 or , equivalently , that of @xmath114 for any @xmath115 in the equatorial plane .
explicit formulas are provided in appendix [ app : sequential ] , and the resulting @xmath116 is plotted alongside @xmath103 in fig .
[ fig2 ] ( dashed and solid curves , respectively ) .
interestingly , as @xmath1 increases , the sensitivity of @xmath117 is seen to _ oscillate _ regularly between zero and the qfi .
this can be intuitively understood if one thinks of the ` dynamics ' of @xmath84 in the bloch sphere : the maximally coherent preparation @xmath118 lies along the equator , on the bloch sphere surface .
then , the iterative application of @xmath90 gives rise to a trajectory which inspirals on the equatorial plane as @xmath84 approaches its fixed point @xmath119 at the center of the sphere .
this is just a combination of the unitary rotation around the @xmath110axis and the loss of purity that results from the average in eq . .
the eigenstates of the sld follow the rotation of @xmath84 , whereas our actual measurement basis remains fixed . as a result , the sensitivity of @xmath120 oscillates between @xmath121 and @xmath103 , with the latter saturated when the two bases coincide ( see appendix [ app : sequential ] for a visual description ) . as a comparison ,
let us now consider the parallel - entangled strategy , i.e. , a single - round protocol starting from an entangled state of the @xmath1 qubits .
we will choose the maximally - entangled ghz state @xmath122 @xcite as initial preparation .
although this may not be optimal for noisy parameter estimation @xcite , it comes with the advantage that it keeps a simple form under the local application of our channel on all @xmath1 probes .
it also has the same degree of quantum coherence as the single - qubit @xmath123 state , as measured by the @xmath124norm @xcite .
this choice allows us to obtain an analytic expression for the qfi ( see appendix [ app : parallel ] ) , which is plotted in fig .
[ fig2 ] ( dotted curve ) along with the qfi of the sequential case .
note that , while we use the same notation ` @xmath1 ' for the number of rounds in the sequential case and for the number of probes in the parallel one , these are essentially different resources .
one can nonetheless make sense of the comparison between the two metrological settings by invoking their formal equivalence in absence of noise @xcite , and by recalling that @xmath8 equals the overall number of interactions with the phase - imprinting channel in both cases . for : the @xmath8-round sequential setting with a single - qubit initial state @xmath101 ( solid ) , the @xmath8-round sequential setting with a passive ancilla and two - qubit initial bell state ( dot - dashed ) , and the parallel - entangled setting with @xmath8-qubit initial ghz state ( dotted ) .
the dashed line amounts to the phase sensitivity of @xmath125 , @xmath126 , and @xmath127 in each of the three settings , respectively .
the model parameters are the same as in fig .
all the quantities plotted are dimensionless . ]
the resulting parallel qfi exhibits a linear asymptotic scaling with @xmath128 , unlike the sequential setting .
however , even if such a large @xmath8-qubit entangled probe could be prepared , its maximum sensitivity would only be saturated by some phase - dependent _ collective _ measurement on all @xmath1 probes @xcite . indeed , in appendix
[ app : parallel ] we show that this parallel qfi may be split into two contributions : one , with a profile similar to that of the sequential @xmath103 , stemming from the matrix elements of the output state in the subspace of total angular momentum @xmath129 , and another one , related to the complementary subspace , which depends on the phase @xmath0 through the longitudinal deformation parameter @xmath46 .
it is precisely this second contribution which endows the probe with a linearly increasing sensitivity at large @xmath1 .
it seems intuitively clear that singling out the relevant information contained in the subspace @xmath130 requires a collective estimator , such as @xmath131 itself .
note that such coherent manipulations may be demanding to implement , or even unavailable in case the probes are transmitted to @xmath1 remote stations during the process .
alternatively , one could ask about the performance of a collection of accessible _ separable _ measurements , such as @xmath132 , implemented locally on each probe and supplemented by classical communication in the data analysis stage @xcite .
the corresponding phase sensitivity can also be computed analytically ( see appendix [ app : parallel ] ) , and quite remarkably it turns out to _ coincide _ with that of @xmath120 in the @xmath1-round setting ( i.e. , the dashed line in fig .
[ fig2 ] ) .
this behaviour is generic and does not depend on the specific sub - optimal observable measured on each probe .
that is , although the parallel - entangled setting can , in principle , outperform the sequential one asymptotically @xcite , we find that they may become metrologically equivalent when the probe readout at the output is limited to measuring local observables : this restriction _ de facto _ banishes the asymptotic linear scaling of the precision .
it is important to remark that we assumed a particular probe preparation ( ghz states ) .
therefore , our observations should not be understood as a general ` no - go ' result , advocating against parallel - entangled estimation strategies in the presence of noise . the general question of whether the gap between sequential and parallel - entangled settings @xcite persists when optimal input states and more general separable measurements are considered is definitely worthy of further investigation , although it lies beyond the scope of this paper .
finally , as an example of the usefulness of entanglement in a practical sequential scenario , let us supplement the probe with a passive two - level ancilla .
specifically , we can prepare the probe - ancilla pair in a bell state @xmath133 ( which has the same @xmath134norm of coherence as the single - qubit @xmath101 and @xmath8-qubit ghz states ) and apply the noisy channel only on the first qubit , yielding @xmath135 .
we find that , although interrogating probe and ancilla by a separable measurement like @xmath136 reduces once more to the same sensitivity as the single - qubit unassisted scenario ( dashed curve in fig . [ fig2 ] ) , performing instead a non - separable ( yet manageable ) measurement such as @xmath137 does provide a sizeable increase in phase sensitivity ( see appendix [ app : passive ] ) .
in general , one may conclude that using independent probes in a sequential scheme , possibly supplemented by correlated passive ancillas , offers a _ practical advantage _ in noisy parameter estimation , in spite of the potential superiority of parallel - entangled strategies @xcite . as we illustrated , acquiring partial information about the geometry of the parameter - imprinting process allows one to optimize the estimation protocol at the single - probe level , by simply adjusting the sampling time or number of rounds .
such a sequential estimation protocol relies on the initial amount of `` unspeakable '' coherence @xcite , which is a genuinely quantum feature @xcite , and is here confirmed as the key resource for estimating parameters encoded in incoherent operations , which include all phase - covariant channels .
however , the estimation performance only scales linearly or `` classically '' in the probe size , whereby scaling up the probe size is intended as repeating the optimized sequential procedure @xmath77 times using independent probe qubits , all initialized in a maximally coherent state .
nonetheless , at the single - probe level , the sensitivity does scale quadratically in the number of rounds @xmath8 , provided @xmath1 is well below @xmath138 .
notably , in the technologically relevant limit of @xmath139 ( e.g. magnetometry in a very weak magnetic field ) , the optimal number of rounds @xmath2 stays fairly large even for relatively low @xmath140 ( see appendix [ app : kraus ] ) , which translates into a very uncertain phase generator . as a result , a quadratic - like scaling of the precision for each individual probe
can be maintained up to many iterations , although definitely not asymptotically .
an interesting next step could be to extend our analysis to _ multiparameter metrology _ ,
e.g. considering the simultaneous estimation of the phase @xmath72 and the noise parameter @xmath112 , or the actual generator @xmath141 @xcite . to conclude ,
let us remark that , in general , the comparison between different metrological settings is a particularly tricky subject , since all the _ resources _ must be identified and properly accounted for @xcite .
for instance , in spite of the formal equivalence of sequential and parallel settings in absence of noise , promoting the ` number of rounds ' to the status of resource , at the same level as the ` number of probes ' is probably not fair , since the actual costs of preparation , control and measurement of an additional quantum probe are not comparable to the costs of increasing the sampling time for one additional round .
it is surely worthwhile to put different metrological settings under a unified set - up , also including feedback control protocols , so as to carry out an objective bookkeeping of the associated costs
. this will be the subject of future work .
we acknowledge the european research council ( erc stg gqcop , grant no .
637352 ) and the royal society ( grant no .
ie150570 ) for financial support .
we thank a. datta , r. demkowicz - dobrzaski , a. farace , c. gogolin , m. gu , l. maccone , m. mehboudi , k. macieszczak , k. modi , and m. oszmaniec for useful discussions , and especially j. koodyski for helping us improve the manuscript with his feedback .
below we will give further details about the tight - fitting lower bound @xmath95 to the qfi @xmath103 .
as stated in the main text , @xmath89 @xcite does hold in general although it is not necessarily a tight bound .
if the maximization is not restricted to vectorized physical states @xmath142 , but extended to all normalized four - dimensional vectors in liouville space , we would end up calculating the operator norm @xmath143 .
this equals the largest eigenvalue of the enclosed matrix .
in particular , for the family of channels considered in the main text and represented by @xmath53 , the eigenvalues of @xmath144 are @xmath145 , @xmath146 , and @xmath147 $ ] , which is doubly degenerate .
now , considering the expression given in the main text for the qfi of any phase - covariant unital channel and recalling that @xmath148 one can readily see that @xmath149 . furthermore , whenever @xmath150 one could elegantly lower - bound the channel qfi as @xmath151 .
this is the case , for instance , in the example considered in the main text , based on the von mises - fisher distribution . in the most general case , @xmath95 still remains a close - fitting lower bound to the qfi for the whole class of channels represented by @xmath53 .
in this section we shall give explicit expressions for a set of kraus operators @xmath152 realizing the noisy phase - imprinting channel @xmath153 in eq .
( 1 ) of the main text , when the generic distribution @xmath17 for the random generator @xmath154 is chosen to be the von mises - fisher distribution ( vmf ) distribution @xmath98 .
recall from the main text that @xmath98 reads @xmath155 where the concentration parameter @xmath156 gives an idea of how narrow the distribution of @xmath109 is around the @xmath110axis . for @xmath157 ,
the vmf distribution is uniform on the bloch sphere , while it increases in concentration for greater @xmath112 .
this is illustrated in fig .
[ fig : distspheres ] . .
the probability density runs from indigo at its lowest to red at its highest .
note that the setting @xmath158 , used to generate the illustrations of the main text , actually corresponds to a very broad distribution for the random generator @xmath154 .
all the quantities plotted are dimensionless . ]
the kraus operators @xmath159 may be readily obtained from the eigenvalues and eigenvectors of the corresponding choi matrix of the map @xmath160 .
this is the matrix resulting from @xmath161 , where @xmath162 is the two - qubit bell state @xmath163 @xcite .
after a tedious but otherwise straightforward calculation one finds that the only non - zero matrix elements of the corresponding kraus operators @xmath164 ( with @xmath165 and @xmath166 ) are @xmath167 where @xmath168 ( @xmath169 ) , @xmath170 ( @xmath169 ) and @xmath171 ( @xmath172 ) are the following functions of @xmath112 and @xmath72 , @xmath173 ^ 2}}{\kappa[\kappa\sinh{(\kappa+2i\varphi)}-\cosh{(\kappa+2i\varphi)}+\cosh{\kappa}]-2\sin^2{\varphi}\sinh{\kappa}}. \label{eq : kraus_abc } \end{split}\ ] ] this operator - sum representation then allows simple calculation of @xmath174 by bypassing the difficult integration that defines the channel .
the liouville representation @xmath53 of the channel @xmath90 takes the form @xmath175 in terms of the operator - sum representation of the channel .
in particular , we have that @xmath176 where @xmath177 ( @xmath172 ) and @xmath178 denotes the complex argument of @xmath179 .
we will now provide a closed formula for the qfi @xmath103 for a single two - level probe undergoing @xmath1 sequential applications of @xmath90 .
we shall take the optimal ` plus ' state @xmath180 as the probe preparation @xmath181 .
recall from the main text that @xmath103 of @xmath84 simply writes as @xmath182 where @xmath183 and @xmath82 are the eigenvalues and eigenvectors of @xmath60 , and any terms for which @xmath184 are excluded from the sum .
the output state of the probe @xmath185 can be obtained by repeatedly applying the kraus operators defined in eqs . and .
after a lengthy calculation , one may cast @xmath103 in compact form as @xmath186 where the phases @xmath187 and @xmath188 are @xmath189 and @xmath190 . as expected from the discussion in the main text , the qfi of eq .
does grow as @xmath191 before a peak is reached at some optimal @xmath192 , after which there is a noise - dependent exponential decay , asymptotically approaching zero in the limit of large @xmath8 . as obtained from the maximization of @xmath193 , as a function of the phase @xmath0 and the concentration parameter @xmath140 .
all the quantities plotted are dimensionless.,width=283 ] we will now comment on the expression of the lower bound @xmath193 on the qfi for our particular noisy channel , in the sequential setting . in this particular case @xmath194
= n^2 \vert s \vert^{2n } \vert s'/s \vert^2 $ ] . note that @xmath195 is thus just the prefactor in the expression for @xmath103 in of eq . , which essentially modulates the amplitude of the optimal phase sensitivity . in fig
[ fig : contour ] , we plot @xmath196 as a function of the concentration parameter @xmath140 and the phase @xmath0 .
note that for @xmath0 small enough , the optimal number of applications remains on the order of @xmath197 even for very low @xmath140 ( i.e. quasi - uniform distribution for the rotation axis @xmath109 ) . as a result , the sequential setting may exhibit a super - classical scaling in the sensitivity , up to a significantly large number of rounds .
the maximum phase sensitivity , given by the qfi , may only be reached when an optimal estimator is measured on the output state of the probe .
recall that such optimal estimator must be diagonal in the eigenbasis of the sld @xmath198 , which reads @xcite @xmath199 we can instead calculate the phase sensitivity @xmath200 of some sub - optimal observable @xmath59 defined as in eq .
( [ eq : cfi_general ] ) . choosing @xmath120 as our estimator yields @xmath201
the qfi and the phase sensitivity @xmath202 are depicted in fig .
[ fig : blocha ] for particular values of @xmath72 and @xmath112 .
it can be seen that @xmath203 displays the aforementioned quadratic behaviour followed by an exponential tail - off , whereas @xmath202 oscillates between zero and @xmath203 .
this curious behaviour can be understood by visualizing the evolved probe state @xmath60 and the measurement eigenbases of both the sld @xmath74 and the sub - optimal estimator @xmath204 on the equatorial plane of the bloch sphere ( see panels a - d in fig .
[ fig : blocha ] ) . here
, the initial probe state @xmath205 begins on the @xmath206axis at the surface of the sphere . as stated in the main text , the bloch vector of the evolved state @xmath60 _ inspirals _ towards the normalized identity
this is a result of the rotation around the @xmath30axis due to the parameter - encoding unitary , and the loss of purity due to the noise .
the optimal measurement basis begins on the @xmath206axis , parallel to the probe state , and rotates with @xmath8 .
meanwhile , the fixed measurement basis lies on the @xmath207axis , so that it periodically coincides with the optimal one : when they are parallel , @xmath208 , whereas when they become perpendicular , @xmath209 .
the frequency of these oscillations is given approximately by @xmath210 .
( dashed line ) for @xmath211 and @xmath212 .
representations of the evolved probe state ( purple arrow ) , optimal measurement basis ( red arrow ) , and sub - optimal measurement basis ( blue arrow ) , as bloch vectors on the equatorial plane are shown in the bottom panels .
( a ) initially , the probe state and both measurements are aligned .
( b ) the phase sensitivity of @xmath213 meets the qfi when the measurement bases realign .
( c ) the phase sensitivity of @xmath204 vanishes whenever the measurement bases become perpendicular .
( d ) otherwise , the phase sensitivity of @xmath204 oscillates between zero and the qfi .
note that , as @xmath1 grows , the optimal basis vectors become perpendicular to the probe state vector as they rotate .
note as well that the probe state vector gradually shortens due to the loss of purity .
all the quantities plotted are dimensionless.,title="fig : " ] + ( dashed line ) for @xmath211 and @xmath212 .
representations of the evolved probe state ( purple arrow ) , optimal measurement basis ( red arrow ) , and sub - optimal measurement basis ( blue arrow ) , as bloch vectors on the equatorial plane are shown in the bottom panels .
( a ) initially , the probe state and both measurements are aligned .
( b ) the phase sensitivity of @xmath213 meets the qfi when the measurement bases realign .
( c ) the phase sensitivity of @xmath204 vanishes whenever the measurement bases become perpendicular .
( d ) otherwise , the phase sensitivity of @xmath204 oscillates between zero and the qfi .
note that , as @xmath1 grows , the optimal basis vectors become perpendicular to the probe state vector as they rotate .
note as well that the probe state vector gradually shortens due to the loss of purity .
all the quantities plotted are dimensionless.,title="fig : " ] ( dashed line ) for @xmath211 and @xmath212 .
representations of the evolved probe state ( purple arrow ) , optimal measurement basis ( red arrow ) , and sub - optimal measurement basis ( blue arrow ) , as bloch vectors on the equatorial plane are shown in the bottom panels .
( a ) initially , the probe state and both measurements are aligned .
( b ) the phase sensitivity of @xmath213 meets the qfi when the measurement bases realign .
( c ) the phase sensitivity of @xmath204 vanishes whenever the measurement bases become perpendicular .
( d ) otherwise , the phase sensitivity of @xmath204 oscillates between zero and the qfi .
note that , as @xmath1 grows , the optimal basis vectors become perpendicular to the probe state vector as they rotate .
note as well that the probe state vector gradually shortens due to the loss of purity .
all the quantities plotted are dimensionless.,title="fig : " ] ( dashed line ) for @xmath211 and @xmath212 .
representations of the evolved probe state ( purple arrow ) , optimal measurement basis ( red arrow ) , and sub - optimal measurement basis ( blue arrow ) , as bloch vectors on the equatorial plane are shown in the bottom panels .
( a ) initially , the probe state and both measurements are aligned .
( b ) the phase sensitivity of @xmath213 meets the qfi when the measurement bases realign .
( c ) the phase sensitivity of @xmath204 vanishes whenever the measurement bases become perpendicular .
( d ) otherwise , the phase sensitivity of @xmath204 oscillates between zero and the qfi .
note that , as @xmath1 grows , the optimal basis vectors become perpendicular to the probe state vector as they rotate .
note as well that the probe state vector gradually shortens due to the loss of purity .
all the quantities plotted are dimensionless.,title="fig : " ] ( dashed line ) for @xmath211 and @xmath212 .
representations of the evolved probe state ( purple arrow ) , optimal measurement basis ( red arrow ) , and sub - optimal measurement basis ( blue arrow ) , as bloch vectors on the equatorial plane are shown in the bottom panels .
( a ) initially , the probe state and both measurements are aligned .
( b ) the phase sensitivity of @xmath213 meets the qfi when the measurement bases realign .
( c ) the phase sensitivity of @xmath204 vanishes whenever the measurement bases become perpendicular .
( d ) otherwise , the phase sensitivity of @xmath204 oscillates between zero and the qfi .
note that , as @xmath1 grows , the optimal basis vectors become perpendicular to the probe state vector as they rotate .
note as well that the probe state vector gradually shortens due to the loss of purity .
all the quantities plotted are dimensionless.,title="fig : " ]
in this section we will give further details about the performance of the sequential estimation setting when the probe is supplemented with a passive two - level ancilla . recall from the main text that ,
in this case , we prepare the two qubits in a bell state @xmath214 and proceed to apply sequentially @xmath1 times the phase - imprinting channel as @xmath135 .
the resulting ` evolved ' two - qubit state has maximally mixed marginals at any @xmath1 .
the corresponding qfi @xmath215 may be readily evaluated to @xmath216 where @xmath217 once again , we shall particularize our results choosing the vmf distribution for the stochastic generator ( see fig .
[ fig : seq_anc ] ) . in the first place , note that @xmath215 notably outperforms the sensitivity of a single probe . in particular , when a separable estimator such as @xmath136 is considered , the phase sensitivity of the probe - ancilla pair remains upper - bounded by the qfi of a single probe .
however , as we can see , when a bell measurement is performed _
jointly _ on the probe and the ancilla , the ensuing sensitivity can come much closer to the ultimate limit set by @xmath215 . .
the qfi for a single two - level probe ( without ancilla ) is included for comparison ( dot - dashed ) .
the phase sensitivity of @xmath136 ( dotted ) and @xmath218 ( dashed ) have also been plotted .
the vmf distribution is assumed for the stochastic generator of the phase rotations , with @xmath219 and @xmath158 .
all the quantities plotted are dimensionless . ]
we will now compute the qfi in the parallel setting for a maximally entangled greenberger - horne - zeilinger ( ghz ) state of the @xmath1 probes , @xmath220 .
the output state @xmath221 is a type of _ x state _ , that is , a state represented by a @xmath222 matrix with non - zero elements @xmath223 ( with @xmath224 ) only along the main diagonal and in the extreme off - diagonal corners .
in particular , we find @xmath225 where @xmath226 is the hamming weight of @xmath227 , i.e. the number of @xmath228 s in the binary representation of @xmath227 .
the over - bar denotes complex conjugation and @xmath229 is given by @xmath230 then , the qfi ( @xmath231 ) may be calculated using eq . and is given by @xmath232 the contribution of the first term in eq .
to the total sensitivity is qualitatively similar to the qfi in the sequential setting , i.e. it scales quadratically for low @xmath1 and , after peaking , it decays exponentially to zero .
however , the second term , contributes with an classical - like increase in @xmath1 , which ultimately yields a linear asymptotic scaling for the overall qfi . as stated in the main text ,
while the first contribution relates to the outermost `` corners '' of the density operator ( i.e. the subspace spanned by @xmath233 and @xmath234 , with total angular momentum @xmath235 ) , the second one is determined by all the matrix elements along the diagonal of the state , such that @xmath236 . in particular , accessing all the relevant information giving rise to the classical - like asymptotic scaling of the sensitivity thus requires one to perform a non - separable measurement capable of telling the subspace @xmath235 from its complementary @xmath236 , such as a measurement of the total angular momentum @xmath131 .
analogously to the sequential setting , one can consider the sensitivity of a sub - optimal estimator .
in particular , we can compute the phase sensitivity of the separable observable @xmath237 by resorting to eq . and .
as pointed out in the main text this yields exactly the same formula of eq . , thus implying that the parallel - entangled and unentangled sequential settings are metrologically equivalent so long as the estimation is constrained to separable measurements .
note furthermore that , in both sequential and parallel settings , one could consider alternative observables to @xmath204 with eigenbases on the equatorial plane of the bloch sphere , such as @xmath238 or even some suboptimal yet phase - dependent observable .
generically , the oscillatory behavior will remain , but the periodicity and locations of the maxima will change depending on the chosen observable . | the problem of estimating an unknown phase @xmath0 using two - level probes in the presence of unital phase - covariant noise and using finite resources is investigated .
we introduce a simple model in which the phase - imprinting operation on the probes is realized by a unitary transformation with a randomly sampled generator .
we determine the optimal phase sensitivity in a sequential estimation protocol , and derive a general ( tight - fitting ) lower bound .
the sensitivity grows quadratically with the number of applications @xmath1 of the phase - imprinting operation , then attains a maximum at some @xmath2 , and eventually decays to zero .
we provide an estimate of @xmath2 in terms of accessible geometric properties of the noise and illustrate its usefulness as a guideline for optimizing the estimation protocol .
the use of passive ancillas and of entangled probes in parallel to improve the phase sensitivity is also considered .
we find that multi - probe entanglement may offer no practical advantage over single - probe coherence if the interrogation at the output is restricted to measuring local observables . |
over the past few years , cosmological models based on the paradigm of an inflationary universe dominated by cold dark matter ( cdm ) have proved remarkably successful at explaining the origin and evolution of structure in the universe .
the free parameters of astrophysical relevance in this modeling are surprisingly few : the current rate of universal expansion , @xmath8 ; the mass density parameter , @xmath9 ; the primordial baryon abundance , @xmath4 ; and the overall normalization of the power spectrum of initial density fluctuations , @xmath3 . over the past few years
, limits on the values allowed for these parameters have been consistently refined by improved observational techniques and theoretical insight , and it is widely accepted that a new `` standard '' model has emerged as the clear front - runner amongst competing models of structure formation .
this model , which we shall call `` standard '' @xmath7cdm , or s@xmath7cdm for short , envisions an eternally expanding universe with the following properties ( bahcall et al 1999 ) : ( i ) matter makes up at present less than about a third of the critical density for closure ( @xmath0 ) ; ( ii ) a non - zero cosmological constant restores the flat geometry predicted by most inflationary models of the early universe ( @xmath10 ) ; ( iii ) the present rate of universal expansion is @xmath11 km s@xmath12 mpc@xmath12 ( @xmath13 km s@xmath12 mpc@xmath14 ) ; ( iv ) baryons make up a very small fraction of the mass of the universe ( @xmath15 ) ; and ( v ) the present - day rms mass fluctuations on spheres of radius 8 @xmath16 mpc is of order unity ( @xmath17 ) .
the s@xmath7cdm model is consistent with an impressive array of well - established fundamental observations : the age of the universe as measured from the oldest stars ( e.g. , chaboyer et al 1998 ) , the extragalactic distance scale as measured by distant cepheids ( e.g. , madore et al 1998 ) ; the primordial abundance of the light elements ( e.g. , schramm & turner 1998 ) , the baryonic mass fraction of galaxy clusters ( e.g. , white et al 1993 ) , the amplitude of the cosmic microwave background fluctuations measured by cobe ( e.g. , lawrence , scott & white 1999 ) , the present - day abundance of massive galaxy clusters ( e.g. , eke , cole & frenk 1996 ) , the shape and amplitude of galaxy clustering patterns ( e.g. , wu , lahav & rees 1999 ) , the magnitude of large - scale coherent motions of galaxy systems ( e.g. , strauss & willick 1995 , zaroubi et al 1997 ) , and the world geometry inferred from observations of distant type ia supernovae ( e.g. , perlmutter et al 1999 , garnavich et al 1998 ) , among others . because its major parameters are fixed , s@xmath7cdm is an eminently falsifiable model whose predictive power may be used to ascertain its validity on scales different from those used to tune the model .
one scale of particular interest is that of individual galaxies , since few observational constraints on these small scales have been used to adjust the parameters of s@xmath7cdm .
this exercise is especially compelling because the dark halo structure found in cosmological n - body simulations of cold dark matter universes seems at odds with dynamical studies of disk galaxies that assign a substantial gravitational role to the disk component ( see , e.g. , the `` maximal disk '' solutions of debattista & sellwood 1998 , and references therein ) as well as with rotation curve studies of dark matter - dominated galaxies ( moore 1994 , flores & primack 1994 , mcgaugh & de block 1998 , moore et al 1999 , navarro 1999 ) . these claims are based largely on comparisons of the detailed _ shape _ of the rotation curve of very low surface brightness dwarfs with the innermost _ density profile _ of simulated dark halos .
unfortunately , the scales where deviations are most pronounced ( the inner few kpc ) are also the most compromised by numerical uncertainties ( most simulations relevant to this problem published to date have gravitational softening scales of order @xmath18-@xmath19 kpc ) .
the comparison is thus rather uncertain .
for example , kravtsov et al ( 1998 ) have argued , on the basis of simulations similar to those used by the other authors , that cdm halos are actually _ consistent _ with the rotation curves of dark matter - dominated disks , a somewhat surprising result that illustrates well , nonetheless , the vulnerability of numerical techniques on scales close to the numerical resolution of the simulations . from a strictly pragmatic numerical standpoint , it would be desirable to circumvent these uncertainties by adopting comparison criteria that are less sensitive to the effects of numerical shortcomings .
one possible choice is to use , rather than the dark matter density profile near the center , the _ total amount _ of dark mass within the main body of individual galaxies . for spiral galaxies ,
this criterion would imply that simulations that can estimate reliably the amount of dark mass within a couple of exponential scalelengths may be safely used for comparison with observations . for bright spirals like the milky way this corresponds to radii of about @xmath20-@xmath21 kpc , well outside the region that may be compromised by numerical artifacts in the current generation of n - body experiments . in this paper
we follow this proposal and compare the results of recent very high - resolution simulations of the formation of dark halos in the s@xmath7cdm model with observational constraints on the total dark mass within spiral disks derived from the tully - fisher relation and from observations of galactic dynamics .
the numerical setup of the simulations is identical to that described by navarro , frenk & white ( 1997 , hereafter nfw97 ) but the number of particles has been increased more than tenfold . as a result ,
each simulated s@xmath7cdm halo has of order @xmath22 dark matter particles within the virial radius and several thousands within radii comparable to the sun s distance from the galactic center ( the `` solar circle '' @xmath23 ) so our numerical uncertainties are for all practical purposes negligible .
we show below that comparison between these simulations and observational constraints reveals a severe inconsistency : s@xmath7cdm halos have substantially more mass near the center than the maximum inferred from observations .
we argue that this presents a serious challenge to the s@xmath7cdm cosmogony and to many of its likely variants and that possible solutions involve fundamental revision of the basic premises or parameters of the model .
kinematic observations of stars and gas in the galaxy provide tight constraints on the total amount of dark matter within the solar circle , @xmath23 .
a direct estimate can be made by assuming that the halo is spherically symmetric , @xmath24 , where @xmath25 is the contribution of the dark halo to the circular velocity at @xmath23 .
this may be obtained by subtracting the disk contribution from the total circular velocity , @xmath26 . for the iau - sanctioned values of @xmath27 kpc and @xmath28 km s@xmath12 , and assuming that the disk potential is well approximated by an exponential disk with scalelength @xmath29 kpc and total mass @xmath30 ( binney & tremaine 1987 ) , we find @xmath31 the uncertainty in this determination is hard to assess , although the evidence suggests that the mass in eq .
1 is actually an _
upper _ limit to the dark mass inside the solar circle .
this is in good agreement with the recent milky way mass models of dehnen & binney ( 1998 ) , who find that , within the slightly larger radius of @xmath21 kpc the dark halo accounts for less than about @xmath20-@xmath32 , and perhaps as little as @xmath33 . the disk contribution to the circular velocity increases with our estimated value of @xmath34 , which in turn depends on : ( i ) the local density of the disk derived from the vertical kinematics of stars in the solar neighborhood ( i.e. , from `` oort limit '' analysis , @xmath35 ) , ( ii ) on the exponential scalelength of the disk , and ( iii ) on the solar circle itself , through @xmath36 .
recent reviews ( reid 1993 , sackett 1997 ) of available data suggest that the exponential scalelength assumed above may need to be revised downwards by up to @xmath37 , but otherwise leave @xmath38 and @xmath23 largely unchanged kpc upon analysis of the role of interstellar gas on the local values of oort s constants . since the circular velocity at @xmath23 decreases correspondingly , this modification has little effect on our conclusions . ] from the values assumed above .
such revision would _ increase _ the disk mass , leading to values of @xmath39 lower than quoted in eq . 1 .
indeed , sackett ( 1997 ) concludes that the milky way disk may very well be `` maximal '' once this revision is taken into account .
finally , we note that our procedure neglects the contribution of the galaxy s bulge , lending further support to our interpretation of eq . 1 as an upper limit to the dark mass inside @xmath23 .
figure 1 compares , as a function of halo mass , the dark mass estimate in eq . 1 with the results of simulations of several s@xmath7cdm halos .
halo masses ( @xmath40 ) are measured inside the radius , @xmath41 , of a sphere of mean density 200 times the critical density for closure , and are typically characterized by the circular velocity at that radius , @xmath42 .
the reason for this choice is that at @xmath41 the circular orbit timescale is approximately equal to the age of the universe ; @xmath41 thus separates the `` virialized '' region of the halo from the region where mass shells are infalling into the system for the first time .
the relevance of this definition stems from the latter property : only baryons inside @xmath41 may contribute to the baryonic mass of the galaxy , since those beyond this radius have yet to reach the center of the halo ( white et al 1993 ) .
this property may be used to derive a firm _ lower limit _ to the mass of the halo that surrounds the milky way , corresponding to the case where the baryonic mass of the disk equals the total baryonic mass inside @xmath41 : i.e. , @xmath43 , implying that @xmath44^{1/3}$ ] . for the disk mass
adopted above and the cosmological parameters appropriate to s@xmath7cdm discussed in 1 , we find that @xmath45 we emphasize that this is a _ strict _ lower limit to the mass of the halo that surrounds the milky way , since experiments show that typically not more than @xmath46 of the baryons within @xmath41 are actually accreted into the central disk ( navarro & white 1994 , navarro & steinmetz 1997 ) . a similar , albeit more stringent , constraint may be derived by comparing the angular momentum of the milky way with those of s@xmath7cdm halos .
halos with @xmath47 km s@xmath12 typically have specific angular momenta lower than that of the milky way disk ( mo , mao & white 1998 , syer et al 1999 ) and are therefore unlikely hosts of the galaxy .
indeed , baryons typically lose a significant fraction of their angular momentum as they collapse to the disk ( navarro & white 1994 , navarro & steinmetz 1997 ) so we conclude that almost certainly the circular velocity of the milky way halo must exceed @xmath48-@xmath49 km s@xmath12 .
figure 1 compares the constraints from eqs . 1 and 2 with the results of our s@xmath7cdm numerical simulations .
the comparison shows clearly a major discrepancy between the maximum dark matter inside @xmath23 allowed by observations and the results of the numerical experiments .
for example , s@xmath7cdm halos with circular velocities similar to that of the milky way disk ( @xmath50 km s@xmath12 ) have about _ three times _ more dark mass inside the solar circle than inferred from observations . even for the extreme case where the halo has the strict minimum circular velocity allowed by eq . 2
, the simulations indicate an excess of more than @xmath51 in the dark mass within @xmath23 .
this serious discrepancy only worsens if we take into account that some extra dark material may be drawn inside @xmath23 by the formation of the disk .
a rough estimate of the magnitude of this correction can be made by assuming that the halo responds adiabatically to the assembly of the disk ; the discrepancy then increases from @xmath51 to almost @xmath46 for the least massive halo allowed by eq .
we conclude that halos formed in the s@xmath7cdm scenario are too centrally concentrated to be consistent with observations of the dynamics of the galaxy .
it is possible to extend the analysis of the previous subsection to a large fraction of disk galaxies by examining the tight correlation between the total luminosity of galaxy disks and the rotation speed of their gas and stars ( the tully - fisher relation , tully & fisher 1977 ) .
provided that disk mass - to - light ratios and exponential scalelengths can be estimated reliably , it is possible to evaluate the disk contribution to the circular velocity and to apply the same analysis of the previous subsection to derive constraints on the total dark mass contained within the optical radius of the galaxy .
we choose to carry out the analysis at @xmath52 exponential scalelengths ( @xmath53 ) from the center , since the contribution of exponential disks to the circular velocity peaks there and it is at that radius that tully - fisher velocities are typically measured ( courteau 1997 ) .
( we note that @xmath54 in the case of the milky way , so this choice of radius is similar to that adopted in 2.1 . )
constraints on dark masses inside @xmath53 depend sensitively on estimates of the total mass associated with galaxy disks . because of the lack of an oort - limit analog in external galaxies , we resort to broad - band colors as estimators of the mass - to - light ratio of the stellar disk .
late type spirals such as those that make up the majority of galaxies in tully - fisher samples have @xmath55 colors in the range @xmath56 ( courteau 1999 ) which , for a galaxy that has been steadily forming stars for @xmath57 gyrs , imply @xmath5-band mass - to - light ratios of order @xmath58 .
this estimate assumes a salpeter initial mass function and exponential star formation histories with timescales which vary from @xmath59 gyr to star formation rates that are constant over the age of the universe .
similar values are obtained using the gissel96 models of bruzual & charlot ( 1996 ) and the pegase models of fioc & rocca - volmerange ( 1997 ) .
uncertainties in this estimate of @xmath60 are hard to assess , but are unlikely to be larger than about a factor of two , which is the value of the `` error '' we assume here .- band mass - to - light ratios of about @xmath61 for galaxies in the same color range , more than a factor of @xmath19 lower than that derived for the milky way disk from the oort limit ( binney & tremaine 1987 ) .
this reflects the well known galactic `` disk dark matter '' content . assuming that tully - fisher galaxies have similar amounts of ( presumably baryonic ) `` disk dark matter '' would make all the arguments we develop here even stronger . ] with this caveat , we compute the rotation speeds ( @xmath62 ) of hypothetical exponential disk galaxies with @xmath63 assembled at the center of the simulated dark halos shown in figure 1 .
disk radii are chosen so as to satisfy the empirical relation , @xmath64 km s@xmath65 kpc ( courteau 1999 , mo , mao & white , 1998 , navarro 1999 ) . since disk radii
depend on @xmath66 , an iterative procedure is needed , which we implement as follows .
given a halo of circular velocity @xmath67 we assign to it a disk with exponential scalelength derived assuming that @xmath68 . the simulated halo structure and the disk potential are then used to compute a new @xmath69 estimate , taking into account the `` adiabatic '' response of the halo to the disk potential ( see navarro , frenk & white 1996 , hereafter nfw96 , for details ) . this new velocity estimate is used to recompute @xmath70 and the procedure is then iterated until convergence .
the solid - line curves in figure 2 illustrate the result of applying this procedure to three representative dark halos of different mass ( or luminosity , since we assume a constant mass - to - light ratio ) , as a function of the total mass adopted for the disk of the galaxy . in each case
we vary the total disk mass from zero to @xmath71 , the maximum value compatible with the baryonic content of the halo . as the disk mass increases , each hypothetical galaxy moves from left to right across the plot .
when the disk mass becomes comparable to the dark mass inside @xmath72 the curve inches upwards and becomes essentially parallel to the observed tully - fisher relation .
the increased potential due to the disk mass and the extra dark mass drawn inside @xmath72 under the assumption that the halo responds adiabatically to the assembly of the disk contribute similarly to the overall increase in rotation speed over and above that of the original dark halo .
the rightmost point of each curve is reached when the mass of the disk equals @xmath71 .
it is clear from figure 2 that , even under the extreme assumption that galaxies contain _ all _ available baryons in each halo , simulated disks are almost two magnitudes fainter than observed . increasing the baryonic mass of a halo has virtually no effect on this conclusion , since in this case model galaxies would just move further along paths approximately parallel to the tully - fisher relation , as shown in figure 2 .
disk galaxies assembled inside s@xmath7cdm halos therefore can not match the observed tully - fisher relation , unless one or more of the assumptions in our procedure are grossly in error .
perhaps the most uncertain step in our argument is the stellar mass - to - light ratio adopted for the analysis .
the horizontal `` error bar '' shown on the starred symbols in figure 2 indicates the effect on our results of varying the @xmath5-band mass - to - light ratio by a factor of two from our fiducial value of @xmath19 in solar units .
this is not enough to restore agreement with observations , which would require @xmath73 , a value much too low to be consistent with standard population synthesis models .
the vertical `` error bars '' illustrate the effect of varying the `` concentration '' of each halo by a factor of two .
, where the scale radius , @xmath74 is one of the parameters of the density profile model proposed by nfw96 and nfw97 , @xmath75 , see those papers for details . ] even with this large variation in halo structure , our hypothetical disks fail to reproduce the observations .
a second uncertainty comes from the `` adiabatic contraction '' correction applied to dark halos in order to mimic the halo response to the disk assembly , and one may wonder whether our `` adiabatic '' contraction assumption is at all correct .
this possibility may be checked , however , through direct numerical simulation .
we have therefore included gas in the simulations and have evolved them again using with the simplified treatment of radiative cooling and star formation described in detail in steinmetz & navarro ( 1999 ) .
the results are shown as open squares in figure 2 , and are in good agreement with the results of the simple modeling proposed above .
as in the case of the milky way discussed in 2.1 , the problem can be traced to the large central concentration of s@xmath7cdm halos . indeed , one way to solve the problem would be to reduce substantially ( by a factor of two or three ) the dark mass in the innermost few kpc of galaxy - sized dark halos .
this would significantly reduce @xmath69 , bridging the gap between model and observations . in terms of the halo density profile model proposed by nfw96 and nfw97
, this would be equivalent to reducing the `` concentration parameter '' , @xmath76 , by a factor of about five .
this is the same conclusion reached by navarro ( 1999 ) , who advocated that halos with very low `` concentration parameters '' , @xmath77 , and stellar mass - to - light ratios as low as @xmath78 in low surface brightness galaxies were required in order to match the shapes of disk galaxy rotation curves .
we emphasize , however , that the present conclusion is independent of assumptions about the detailed shape of the dark matter density profile , and depends largely on the total dark mass on scales of individual galaxies .
the analysis of the previous subsection demonstrates conclusively that the central mass concentration of s@xmath7cdm halos is inconsistent with observations of the dynamics of spiral galaxies .
does this argument rule out all models based on the cdm+inflation paradigm or are there plausible modifications to the s@xmath7cdm parameters that may bring the model into agreement with observations ? in principle , there are parameter choices that may help make @xmath7cdm models consistent with observations of the internal dynamics of galaxies , but the magnitude of the modifications required are , however , uncomfortably large , and come at the expense of other major successes of the model . as discussed by nfw97 , the central concentration of dark halos is directly proportional to the mean matter density of the universe at a suitably defined collapse time . for halos of fixed mass , these collapse times may depend sensitively on the values adopted for the cosmological parameters . however , as discussed by navarro ( 1999 , see his figure 8) , the combination of parameters needed to reproduce the present - day abundance of galaxy clusters is such that the characteristic densities of galaxy - sized dark halos is approximately independent of @xmath9 and of the value of @xmath8 .
indeed , halos in the former `` standard '' cdm model ( @xmath79 , @xmath80 , @xmath81 ) have very similar concentrations as the s@xmath7cdm halos we discuss here since both models are normalized to match the abundance of massive clusters at @xmath82 .
this implies that all cdm models that match the abundance of clusters are likely to have difficulty reproducing the dynamics of spiral galaxies .
even if cluster normalization is dropped from the list of relevant constraints and only the cobe measurements are used to normalize the power spectrum of initial density fluctuations , reducing substantially the central concentrations of dark halos still implies uncomfortable parameter choices .
for example , if @xmath9 is taken as a free parameter , reducing the dark mass of a @xmath83 km s@xmath12 halo inside @xmath84 kpc by a factor of three relative to s@xmath7cdm requires @xmath85 , a very low number that would call into question the need for a dominant non - baryonic dark matter component in the universe .
similarly , if rather than @xmath9 , @xmath8 is allowed to vary , values as low as @xmath86 km s@xmath12 mpc@xmath12 are required to obtain the desired effect , in gross disagreement with current observational estimates .
if we are to preserve the successes of the s@xmath7cdm model , what seems to be required is a substantial change in the _ shape _ of the power spectrum relative to that predicted by the cdm+inflation paradigm .
a `` designer '' power spectrum that would go some way towards reconciling the model with observations would suppress power on galactic and subgalactic scales while keeping the large scale properties of the model virtually unchanged , as envisioned , for example , in models that introduce a `` tilt '' in the primordial power spectrum relative to the standard harrison - zeldovich value .
this would in principle allow galaxy - sized dark halos to collapse later and thus become less centrally concentrated , although the magnitude of the tilt required is still unclear
. other alternative modifications that may in principle reproduce the desired trend would involve the existence of a sizeable `` hot dark matter '' component ( although much higher than derived from current estimates of the neutrino mass , see , e.g. , primack & gross 1998 ) , or possibly even dark matter candidates that may annihilate without trace in dense regions such as the centers of galaxies .
one problem that afflicts all these proposed modifications is that they may hinder the formation of massive galaxies at high redshift , at odds with the mounting evidence that such galaxies are fairly common at @xmath87 ( see , e.g. , steidel et al 1998 ) .
a thorough investigation of these possibilities will be needed in order to assess just how radically the new `` standard '' model of structure formation must be revised in order to bring spiral galaxies into the realm of observations that are consistent with the current paradigm of structure formation .
this work has been supported by the national aeronautics and space administration under nasa grant nag 5 - 7151 and by nserc research grant 203263 - 98 .
ms acknowledges useful discussions with walter dehnen and rob olling .
jfn acknowledges enlightening discussions with simon white , carlos frenk , mike balogh , and mike hudson , as well as the hospitality of the max - planck institut fr astrophysik , where this manuscript was written .
bruzual , g. , & charlot , s. 1996 , unpublished .
bahcall , n. , ostriker , j.p .
, perlmutter , s. , steinhardt , p. 1999 , _ science _ , in press ( astro - ph/9906463 ) .
binney , j. , tremaine , s. 1987 , _ galactic dynamics _ , princeton university press .
chaboyer , b. , demarque , p. , kernan , p.j . , krauss , l.m .
1998 , , 494 , 96 .
courteau , s. 1997 , , 114 , 2402 courteau , s. , 1999 , astro - ph/9903297 .
debattista , v.p . ,
sellwood , j.a .
1998 , , 493 , l5 .
dehnen , w. , binney , j. 1998 , , 294 , 429 .
eke , v.r . ,
cole , s. , frenk , c.s .
1996 , mnras , 282 , 263 .
fioc , m. , rocca - volmerange , b. , 1997 , , 326 , 950 .
flores , r.a . , primack , j.r .
1994 , , 427 , l1 giovanelli , r. , haynes , m.p . ,
herter , t. , & vogt , n.p .
1997 , , 113 , 22 han , m. , & mould , j.r . 1992 , , 396 , 453 garnavich , p.m. , et al 1998 , , 509 , 74 .
kravtsov , a.v . ,
klypin , a.a . ,
bullock , j.s .
, primack , j.r .
1998 , , 502 , 48 lawrence , c.r . ,
scott , d. , white , m. 1999 , pasp , 111 , 525 .
madore , b.f.,freedman , w. l. ; silbermann , n. ; harding , p. ; huchra , j. ; mould , j. r. ; graham , j. a. ; ferrarese , l. ; gibson , b. k. ; han , m. ; hoessel , j. g. ; hughes , s. m. ; illingworth , g. d. ; phelps , r. ; sakai , s. ; stetson , p. 1998 , nature , 395 , 47 .
mathewson , d.s .
, ford , v.l . , & buchhorn , m. 1992 , , 81 , 413 mcgaugh , s.s .
, de block , w.j.g .
1998 , , 499 , 41 .
mo , h.j . ,
mao , s. , & white , s.d.m .
1998 , mnras , 295 , 319 moore , b. 1994 , , 370 , 629 moore , b. , quinn , t. , governato , f. , stadel , j. , lake , g. , 1999 , , submitted ( astro - ph 9903164 ) navarro , j.f . 1999 , , submitted ( astro - ph 9807084 ) navarro , j.f . , frenk , c.s .
, & white , s.d.m .
1996 , , 462 , 563 ( nfw96 ) .
navarro , j.f . ,
frenk , c.s .
, & white , s.d.m .
1997 , , 490 , 493 ( nfw97 ) .
navarro , j.f . , & steinmetz , m. 1997 , , 478 , 13 navarro , j.f .
, & white , s.d.m .
1994 , , 267 , 401 olling , r.n . ,
merrifield , m.r .
1998 , , 297 , 943 perlmutter , s. et al 1999 , , 517 , 565 .
primack , j.r . , gross , m.a.k . , 1998 , in `` the birth of galaxies '' , xth rencontres de blois , astro - ph/9810204 .
reid , m.j .
1993 , , 31 , 345 sackett , p.d .
1997 , , 483 , 103 schramm , d.n . ,
turner , m.s .
1998 , rev.mod.ph .
steidel , c.c . ,
adelberger , k.l . ,
dickinson , m. , giavalisco , m. , pettini , m. , kellogg , m. 1998 , , 492 , 428 .
steinmetz , m. , & navarro , j.f .
1999 , , 513 , 555 strauss , m.a . ,
willick , j.a . , 1995 , phys .
261 , 271 .
syer , d. , mao , s. , & mo , h.j . 1999 , , 405 , 357 tully , r.b . , & fisher , j.r .
1977 , , 54 , 661 white s. d. m. , navarro j. f. , evrard a. e. , & frenk c. s. 1993 , , 366 , 429 wu , k. , lahav , o. , rees , m. , 1999 , nature , 397 , 225 .
zaroubi , s. , zehavi , i. , dekel , a. , hoffman , y. , kolatt , t. , 1997 , , 486 , 21 . | we compare the central mass concentration of cold dark matter halos found in cosmological n - body simulations with constraints derived from the milky way disk dynamics and from the tully - fisher relation . for currently favored values of the cosmological parameters ( @xmath0 ; @xmath1 ;
@xmath2 ; cobe- and cluster abundance - normalized @xmath3 ; big - bang nucleosynthesis @xmath4 ) , we find that halos with circular velocities comparable to the rotation speed of the galaxy have typically _ three times _ more dark matter inside the solar circle than inferred from observations of galactic dynamics .
such high central concentrations of dark matter on the scale of galaxy disks also imply that stellar mass - to - light ratios much lower than expected from population synthesis models must be assumed in order to reproduce the zero - point of the tully - fisher relation . indeed , even under the extreme assumption that _ all _ baryons in a dark halo are turned into stars , disks with conventional @xmath5-band stellar mass - to - light ratios ( @xmath6 ) are about two magnitudes fainter than observed at a given rotation speed .
we examine several modifications to the @xmath7cdm model that may account for these discrepancies and conclude that agreement can only be accomplished at the expense of renouncing other major successes of the model . reproducing
the observed properties of disk galaxies thus appears to demand substantial revision to the currently most successful model of structure formation . |
a _ robot network _ is a collection of robots working together to achieve a common goal . in order for the robots in such a network to cooperate effectively ,
the ability to observe each other s movements is critical . in many applications , distinguishing between the robots is necessary , but is usually difficult because the robots are identical .
for example , in a _ formation control _ system , robots collectively arrange themselves in some fixed geometric configuration @xcite .
each robot controls its position relative to its neighbours . to achieve this
, the robot must continuously measure the position and determine the identity of each neighbour .
some formation control systems may also benefit from knowledge of the relative orientation of its neighbours , since this information can be used to coordinate views and improve the stability of the system .
we present a novel camera - based method for robot identification , orientation estimation , and approximate distance / angle measurements .
the system uses a camera to observe sequences of coloured lights mounted on the robots .
the lights are mounted in a circle around each robot ( in a plane parallel to the ground ) , such that a camera may see only some of the lights .
the sequences of colours are chosen so that any consecutive subsequence of sufficient length corresponds uniquely to a particular robot in a particular orientation .
this system was implemented in an existing network of _ ebugs_. the _ ebug _
@xcite is a robotics platform designed at monash university s wireless sensor and robot networks laboratory @xcite .
it is equipped with sixteen rgb leds ( red , green and blue light - emitting diodes ) on its perimeter , which can be programmed to display a sequence of colours .
a photo of an ebug may be seen in figure [ fig : ebug ] .
figure [ fig : example colouring ] shows an example of a colouring of four ebugs .
these ebugs have only eight leds , and use only two different colours for illustrative purposes .
any subsequence of five leds is coloured with a unique pattern .
for example , the sequence appears ( counter - clockwise ) only on the right side of the second ebug .
/in 0/1,0,1,1,1,1,1,0 , 1/0,1,0,0,0,0,0,1 , 2/0,0,1,1,0,1,1,0 , 3/0,0,1,1,1,0,0,1 ( 0,0 ) circle [ radius=1 cm ] ; in at ( 360/8*i:1 cm ) ; in a real system , there are limits on the number of colours a camera may reliably distinguish .
similarly , spatial resolution of the camera limits the number of detectable leds around each ebug .
therefore , for a given set of parameters , we want to maximise the number of ebugs that we can use in the system .
this maximum , the _ ebug number _ , is formally defined below .
[ def : ebug number ] suppose every ebug has @xmath1 leds , each of which can be illuminated in one of @xmath0 colours , and suppose further that a camera can reliably detect @xmath8 consecutive leds .
an assignment of colours to the leds of all ebugs is _
@xmath2-valid _ if the camera can distinguish each ebug in each of the @xmath1 orientations . the _ ebug number _
@xmath9 is the maximum number of ebugs for which there exists an @xmath2-valid assignment of colours .
as well as modelling an actual problem that arises in robot networks , determining ebug numbers is a natural combinatorial problem of independent interest .
each of the @xmath10 possible sequences the camera can see can not appear more than once , and each ebug uses @xmath1 distinct sequences .
this gives the following upper bound for the ebug number : @xmath11 colourings that achieve the upper bound in are called _
optimal_. in such colourings , each sequence of @xmath2 colours appears on some ebug . note that when @xmath12 , no @xmath2-sequence of a constant colour can appear on an ebug since all orientations would be identical .
thus optimal colourings can only exist for @xmath13 .
a lower bound may be obtained by applying the lovsz local lemma @xcite : consider a random colouring of @xmath14 ebugs , with each of the @xmath15 leds coloured independently and uniformly at random . for each pair @xmath16 of led sequences ( of length @xmath2 ) ,
let @xmath17 be the event that the same colour sequence has been assigned to @xmath18 and @xmath19 .
thus the colouring is @xmath2-valid exactly when none of the events @xmath17 occurs .
since there are exactly @xmath15 led sequences , and each sequence overlaps with at most @xmath20 other sequences , each event @xmath17 depends on at most @xmath21 other events .
the probability of each @xmath17 is at most @xmath22 ( less if @xmath18 and @xmath19 are overlapping ) .
therefore , by the local lemma , there is an @xmath2-valid colouring whenever @xmath23 , where @xmath24 is euler s number .
hence we obtain the following lower bound : @xmath25 for a fixed value of @xmath2 , this bound is within a constant factor of the upper bound in . in actual camera systems , however , it is reasonable to assume that @xmath2 is proportional to @xmath1 , since a camera can usually detect a fixed arc of the led circle .
thus the lower bound is rather crude , and ultimately we would like to solve the following problem .
[ prob : determine e ] determine @xmath9 exactly . for small values of @xmath0 and @xmath2 ,
a computer search was performed to find large @xmath2-valid colourings .
surprisingly , optimal colourings were found in many cases .
these experiments confirm the following conjecture for all @xmath0 and @xmath2 with @xmath26 . while problem [ prob : determine e ] is likely to be very difficult to solve in general , a mathematically interesting problem is to characterise when optimal colourings exist ( hopefully by proving conjecture [ conj : e always optimal ] ) .
[ conj : e always optimal ] @xmath27 whenever @xmath1 divides @xmath10 and @xmath13 .
this paper provides constructions for some infinite families of optimal colourings , and as such gives evidence to support this conjecture . in section [ sec :
preliminaries ] , problem [ prob : determine e ] is shown to be equivalent to finding many cycles in a _
de bruijn graph _
, with conjecture [ conj : e always optimal ] corresponding to a partition into cycles ( see proposition [ prop : partition conditions ] ) .
existing results about _ de bruijn sequences _ are also discussed .
a well - known algebraic construction of de bruijn sequences is given in section [ sec : lfsrs ] ; we extend this construction to prove some existence results for ebug colourings .
the major result of this section is theorem [ thm : lfsr translation ] , which proves conjecture [ conj : e always optimal ] for infinitely many values .
theoremlfsrtranslation[thm : lfsr translation ] if @xmath0 is a prime power and @xmath28 , then @xmath29 . in section [ sec : necklaces ] , we introduce _ necklaces _ and how they relate to de bruijn graphs .
we then prove the following theorems , both of which yield large ebug colourings from smaller ones .
theoremproductcolouring[thm : product colouring ] fix a value of @xmath2 and set @xmath30 and @xmath31
. then @xmath32 theoreminterleaving[thm : interleaving ] @xmath33 whenever @xmath7 divides @xmath1 .
these theorems preserve optimality , so we may use them to find optimal colourings for large numbers of ebugs .
a valid colouring of ebugs has an interesting interpretation as cycles in a de bruijn graph .
these graphs were discovered independently by de bruijn @xcite and good @xcite in 1946 . the @xmath2-th order @xmath0-ary _ de bruijn graph _
@xmath3 is the digraph @xmath34 , where @xmath35 and @xmath36 .
the vertices of @xmath3 are words of length @xmath2 over an alphabet of size @xmath0 .
there is an edge from @xmath37 to @xmath38 if shifting @xmath37 left and appending any letter gives @xmath38 .
an example of such a graph is shown in figure [ fig : example db ] .
( 000,001,010,100,001,011,111,110,101,011,110,100,000 ) ( 101,010,101 ) ( 000 ) ( 111 ) there is also an alternative , equivalent definition of de bruijn graphs that involves iteratively taking line digraphs @xcite . in this construction , @xmath39 is defined as the complete digraph on @xmath0 vertices with loops .
higher order de bruijn graphs are defined as follows : @xmath40 is the line digraph of @xmath3 .
the vertices in @xmath40 correspond to edges in @xmath3 .
note that while cycles in @xmath3 map directly to cycles in @xmath40 , the converse is not always true : there may be repeated vertices when a cycle from @xmath40 is projected down to @xmath3 .
the objects in @xmath3 that correspond to cycles in @xmath40 are called _ circuits _ , which are closed walks with no repeated edges ( vertex repetition is allowed ) .
[ prop : partition conditions ] the following are equivalent : 1 .
there is a partition of the vertex set of @xmath3 into pairwise disjoint @xmath1-cycles .
3 . there is a partition of the edge set of @xmath41 into pairwise edge - disjoint @xmath1-circuits .
( @xmath42 ) suppose that each vertex of @xmath3 corresponds to a particular camera view of @xmath2 consecutive leds on some ebug .
rotating the ebug to the left corresponds to following an edge in the graph , since the leds shift to the right and one new led is visible .
hence a cycle of length @xmath1 in @xmath3 corresponds to the colouring of a single ebug with @xmath1 leds .
a set of multiple disjoint cycles gives an @xmath2-valid colouring of multiple ebugs ( because vertices are not repeated , each orientation is uniquely identifiable ) , so the ebug number @xmath9 equals the maximum number of _ disjoint @xmath1-cycles _ in @xmath3 .
if every vertex is in one of the @xmath1-cycles , then each colour sequence appears on some ebug .
conversely , if any given colour sequence can be found on some ebug , then the corresponding vertex is in one of the @xmath1-cycles .
thus optimal colourings exist exactly when the whole graph can be partitioned into disjoint @xmath1-cycles .
( @xmath43 ) the equivalence follows immediately from the line digraph construction .
bryant studied edge decompositions of complete directed graphs with loops @xcite , which correspond to the first order de bruijn graphs @xmath39 .
the main result of @xcite was that @xmath39 can be decomposed into @xmath1-circuits if and only if @xmath44 and @xmath1 divides @xmath45 . by proposition [ prop : partition conditions ] , this solves conjecture [ conj : e always optimal ] for @xmath46 .
a similar problem was also posed by dudeney in 1917 @xcite , now commonly known as `` dudeney s round table problem '' .
this problem is equivalent to finding a set of hamiltonian cycles in the complete graph @xmath47 , such that every path of two edges appears in exactly one of the cycles .
dudeney s problem was solved for even @xmath14 @xcite , and also some other cases such as when @xmath48 is a prime power @xcite .
a generalisation of dudeney s problem was studied in @xcite ; here @xmath47 is covered by @xmath1-cycles with the same property ( instead of @xmath14-cycles ) . the main difference between these problems and problem [ prob : determine e ] is that we are concerned with directed circuits in digraphs , and that we allow loops on the vertices .
there is a body of research on cycle decompositions of complete graphs ( see @xcite for an introduction and @xcite for recent results ) , and also some work relating to decompositions into fixed - length directed cycles @xcite .
the methods used , however , are very specific to the special structure of complete graphs , and can not be applied to de bruijn graphs .
there are also results about decomposing de bruijn graphs into _
variable - length _ cycles , using techniques like splitting and merging existing cycles @xcite .
golomb s conjecture , which was proven by mykkeltviet @xcite , states that the decomposition of binary de bruijn graphs into the largest number of disjoint cycles is the decomposition into _ necklaces _ ( see section [ sec : necklaces ] for a definition ) .
these results , unfortunately , can not easily be applied to help with conjecture [ conj : e always optimal ] , since the specific requirement of fixed - length cycles is quite restrictive .
note that in the de bruijn graph @xmath3 , every vertex has in - degree and out - degree @xmath0 .
also , a path can be found from any vertex @xmath37 to any vertex @xmath38 by shifting in letters of @xmath38 one at a time , so the graph is connected .
hence @xmath3 is eulerian , and the next de bruijn graph @xmath40 is hamiltonian ( since an eulerian circuit in @xmath3 is equivalent to a hamiltonian cycle in @xmath40 ) .
this simple fact gives us a starting point for conjecture [ conj : e always optimal ] in the @xmath49 case : it shows that @xmath50 for every @xmath0 and @xmath2 .
hamiltonian cycles in de bruijn graphs are called _ de bruijn sequences_.
the number of @xmath51-de bruijn sequences is @xmath52 this result is due to van aardenne - ehrenfest and de bruijn @xcite , and uses an equivalence between spanning arborescences and eulerian circuits in eulerian digraphs .
there are several known methods for generating de bruijn sequences .
one construction @xcite gives the lexicographically smallest sequence for any given values of @xmath0 and @xmath2 through clever concatenation of necklaces .
this method is described in section [ sec : necklaces intro ] .
another construction involves calculations in finite fields @xcite .
this only works when @xmath0 is a prime power , but has a very simple implementation , which is described in section [ sec : db sequence from lfsr ] .
in this section , algebraic properties of finite fields are exploited to find interesting structures in de bruijn graphs .
section [ sec : db sequence from lfsr ] describes a well - known construction of de bruijn sequences ; we extend this construction further in sections [ sec : lfsr splitting ] and [ sec : lfsr translations ] to find multiple @xmath1-cycles in de bruijn graphs .
we assume that the reader is familiar with elementary group and field theory ; see @xcite for example .
let @xmath0 be a prime power , and choose a primitive element @xmath53 from the finite field @xmath54 .
that is , @xmath53 generates the multiplicative group @xmath55 .
we may consider @xmath56 to be an @xmath2-dimensional vector space over @xmath57 , in which case @xmath58 is a basis .
in particular , @xmath59 can be written as a linear combination of these basis vectors : @xmath60 ( this is called the _ minimal polynomial _ of @xmath53 over @xmath57 ) . a _ linear feedback shift register _ ( lfsr ) is a digital circuit that generates elements of @xmath61 by successive multiplication by @xmath53 .
the simplest implementation , a _
galois lfsr _ , represents the field elements as vectors in @xmath62 with respect to the basis @xmath58 .
multiplication of a vector @xmath63 by @xmath53 is simply a shift of the vector to the right , except that an @xmath59 term is produced . but @xmath59 can be rewritten in terms of the basis vectors , so the multiplication corresponds to the function @xmath64 . since the new state is a linear transformation of the previous state , this function can be expressed as the matrix equation in @xmath65 ( over @xmath57 ) , where the _ _ state change matrix__. ] @xmath66 is given by @xmath67 the constants @xmath68 depend on @xmath53 , and are called the _
feedback coefficients _ for the lfsr .
note that since @xmath61 is generated by @xmath53 , repeatedly applying this operation to some non - zero initial vector generates every non - zero vector in @xmath62 .
a _ fibonacci lfsr _ is a similar construction that uses the transposed state change matrix @xmath69 . in this configuration ,
the next state is given by @xmath70 .
in fact , a fibonacci lfsr performs the same operation as the corresponding galois lfsr when the vectors are represented in a different basis . to see this ,
we must find a matrix @xmath71 that satisfies @xmath72 .
let @xmath71 be defined by @xmath73 for @xmath74 and @xmath75 ( that is , the @xmath16-th entry of @xmath71 is the @xmath76-th entry of @xmath77 ) . the entries of powers of a companion matrix are explicitly known @xcite ,
so we can observe that @xmath71 is a symmetric matrix .
similarly , since @xmath78 , @xmath79 is also symmetric .
hence @xmath80 , so @xmath71 is a change of basis matrix from @xmath66 to @xmath69 .
note that the first row of @xmath71 is @xmath81 , so the first basis vector for the fibonacci lfsr is also @xmath82 ( as in the galois lfsr ) .
we now show that the fibonacci lfsr follows edges in the corresponding de bruijn graph . from here
onwards , we do not use galois lfsrs and instead only represent field elements in the fibonacci basis .
[ prop : edges in fibonacci basis ] let @xmath54 and fix a primitive @xmath83 . if the elements of @xmath56 are identified with the vertices of @xmath3 by expressing them in the fibonacci basis over @xmath84 , then @xmath85 is an edge in @xmath3 for each @xmath86 .
observe that the state change operation @xmath87 in a fibonacci lfsr corresponds to a left shift of the state vector and an extra term on the end .
this is exactly what is required for an edge in @xmath3 .
proposition [ prop : edges in fibonacci basis ] can be used to describe all edges of @xmath3 in terms of field operations .
[ lem : adding scalar preserves edges ] if @xmath86 and @xmath88 , then @xmath89 is an edge in @xmath3 . by proposition [ prop : edges in fibonacci basis ]
, @xmath85 is an edge in @xmath3 . recall that in the construction of the fibonacci basis , the first ( or leftmost ) component corresponds to the basis vector @xmath90 ( the multiplicative identity of @xmath56 ) . thus adding a scalar @xmath24 to a vector only changes the first component , which is shifted out when following an edge in @xmath3 .
hence @xmath91 and @xmath92 have the same out - neighbours ( including @xmath93 ) .
now consider repeatedly applying the state change operation @xmath87 to some initial non - zero field element ( @xmath90 for example ) . since @xmath53 generates @xmath61 , the fibonacci lfsr traverses a cycle of length @xmath94 in @xmath3 .
the missing vertex is @xmath95 , and can always be inserted into this cycle by replacing the edge @xmath96 with two edges @xmath97 and @xmath98 .
note that @xmath97 and @xmath98 are edges by lemma [ lem : adding scalar preserves edges ] ( with @xmath99 , @xmath100 and @xmath101 , @xmath102 , respectively ) .
thus we have found a hamiltonian cycle in @xmath3 , which is a de bruijn sequence . due to the inherently algebraic construction of linear feedback shift registers ,
the symmetry properties of such sequences may be exploited to produce many cycles of the same length . for this section
, we identify vertices of @xmath3 with the elements of @xmath103 via the fibonacci basis described above ( with respect to a fixed primitive @xmath83 ) .
fix a value of @xmath104 , and let @xmath105 for each non - zero scalar @xmath106 .
since @xmath107 , there is an edge from @xmath108 to @xmath109 in @xmath3 ( by lemma [ lem : adding scalar preserves edges ] with @xmath110 ) .
we also have @xmath111 other edges @xmath112 for @xmath113 , so we can form a @xmath1-cycle @xmath114 for each of the @xmath115 values of @xmath106 . in general ,
these @xmath115 cycles are not necessarily pairwise disjoint , but we now show that they are if @xmath1 is small .
[ thm : lfsr splits evenly ] @xmath116 for every prime power @xmath0 and every @xmath117 . let @xmath118 be the value of @xmath18 for which @xmath119 ; that is , the discrete logarithm of @xmath91 with base @xmath53 .
this is well - defined on all of @xmath61 because @xmath53 is a generator .
note that @xmath120 is also the position of @xmath91 in the lfsr sequence ( if the initial state is 1 ) . now consider the relative position of the starting points of two different cycles @xmath121 and @xmath122 as described above , with @xmath123 .
the distance along the lfsr sequence between these starting points is @xmath124 note that @xmath125 , which is a subgroup of @xmath61 of order @xmath115 .
also note that @xmath126 , and that @xmath127 for @xmath128 .
thus @xmath129 has order @xmath115 .
but there is only one subgroup of @xmath61 of order @xmath115 ( since @xmath61 is cyclic ) , so @xmath130 .
hence @xmath131 for some @xmath19 , so @xmath132 is an integer multiple of @xmath133 .
since @xmath134 , the @xmath1 consecutive vertices of @xmath121 can not be in @xmath122 , whose starting vertex is at least @xmath133 places past the start of @xmath121 .
hence these @xmath115 @xmath1-cycles are pairwise disjoint . fix a scalar @xmath88 , and let @xmath135 .
note that @xmath136 has exactly one fixed point , namely @xmath137 .
hence @xmath138 by lemma [ lem : adding scalar preserves edges ] , the @xmath0 out - neighbours of a vertex @xmath91 are @xmath139 .
thus we may partition the edges of @xmath3 into the @xmath0 parts @xmath140 , where @xmath88 .
note that ensures that @xmath141 implies @xmath142 . hence if @xmath143 is a circuit contained in @xmath144 , then @xmath145 is a circuit contained in @xmath146 .
we call this operation _ translating _ the circuit by @xmath24 .
note that the @xmath0 translations of a circuit contained in @xmath144 are pairwise edge - disjoint because the @xmath146 are disjoint .
we are now ready to prove theorem [ thm : lfsr translation ] , which we restate here : if @xmath147 , then the result is trivial since @xmath39 contains @xmath0 loops .
hence we may assume that @xmath148 .
let @xmath4 . by proposition [ prop : partition conditions ] , it is sufficient to find a partition of @xmath41 into @xmath0 edge - disjoint @xmath1-circuits .
recall from section [ sec : db sequence from lfsr ] that the lfsr sequence is constructed using edges solely of the form @xmath149 , and forms a cycle @xmath150 of length @xmath111 .
this cycle ( of vertices ) is also a circuit of @xmath111 edges , and we can construct a translated @xmath151-circuit @xmath152 in @xmath146 for each scalar @xmath88 .
the circuit @xmath150 contains every edge from @xmath144 except the loop @xmath153 , which translates to another loop @xmath154 in @xmath146 .
since @xmath150 contains every non - zero vertex , we may insert , say , the loop @xmath155 into @xmath150 to obtain a @xmath1-circuit @xmath156 ( note that @xmath156 is no longer a cycle since it contains the vertex @xmath157 twice ) . similarly
, we may insert the loop @xmath158 into @xmath152 to generate a @xmath1-circuit @xmath159 for each @xmath88 .
observe that each edge @xmath160 of @xmath41 appears in a unique circuit : if @xmath161 , then the edge is a loop in @xmath162 ; otherwise it is in @xmath159 .
thus we have a partition of @xmath41 into @xmath0 edge - disjoint @xmath1-circuits @xmath163 .
hence @xmath164 .
for example , we may apply theorem [ thm : lfsr translation ] with @xmath165 and @xmath166 to obtain three cycles of length 9 .
suppose we choose a primitive @xmath167 whose minimal polynomial over @xmath168 is @xmath169 .
we now construct a 26-cycle @xmath150 in @xmath170 by iterating the lfsr starting from vertex @xmath171 , which corresponds to the field element @xmath90 ( the multiplicative identity of @xmath56 ) .
this produces the following cycle of symbols : @xmath172 ( the corresponding cycle of vertices is @xmath173 ) .
now we may construct the 27-circuit @xmath156 by inserting a loop , say @xmath174 , into @xmath150 .
the three translations of @xmath156 , shown below in , partition the edge set of @xmath170 , and hence the corresponding cycles in @xmath175 partition the vertex set of @xmath175 .
thus we have shown that @xmath176 .
@xmath177 suppose that in the construction of the lfsr , we chose a non - primitive element @xmath91 with multiplicative order @xmath178 . if @xmath179 is still a basis of @xmath103 over @xmath57 , then vectors with respect to this basis are still in correspondence with field elements .
repeated multiplication by @xmath91 , however , no longer generates every element of @xmath61 ; instead this process traverses the cyclic subgroup of order @xmath1 generated by @xmath91 .
thus the action of the lfsr traces out this subgroup of @xmath61 if the initial state is the identity @xmath90 .
this corresponds to a @xmath1-cycle in @xmath3 . choosing a different starting state for the lfsr translates the whole sequence , but does not change the length of the cycle .
this gives a partition of the non - zero vertices into @xmath1-cycles .
the number of these cycles is @xmath180 , giving @xmath181 .
[ thm : non - primitive lfsrs ] let @xmath0 be a prime power , and @xmath1 a factor of @xmath94 .
if @xmath1 does not divide @xmath182 for each @xmath183 , then @xmath184 since @xmath1 divides @xmath94 , there is an element @xmath91 of multiplicative order @xmath1 in @xmath103 .
since @xmath1 does not divide @xmath182 for @xmath183 , @xmath91 is not in any subfield @xmath185 of @xmath56 .
hence @xmath91 is not the root of any polynomial over @xmath57 of degree @xmath186 . therefore @xmath179 is linearly independent over @xmath57 , and hence a basis of @xmath56 .
therefore , the lfsr generated by @xmath91 traces out a distinct @xmath1-cycle in @xmath3 for each equivalence class of @xmath187 .
we have found @xmath188 disjoint @xmath1-cycles in @xmath3 . since the @xmath1-cycles cover all but one vertex in @xmath3 and @xmath189 , this is the best possible bound . by zsigmondy
s theorem @xcite , there is a prime @xmath190 that divides @xmath94 but not @xmath182 for @xmath183 for any @xmath0 and @xmath2 , except when @xmath191 or @xmath46 and @xmath0 is a mersenne prime ( that is , @xmath192 for some prime @xmath193 ) .
thus theorem [ thm : non - primitive lfsrs ] can be applied with @xmath194 to obtain an almost optimal ebug colouring with @xmath190 leds on each ebug ( only one colour sequence is unused ) .
furthermore , if larger ebugs are desired for the same values of @xmath0 and @xmath2 , any multiple of @xmath190 that divides @xmath94 can also be used for @xmath1 .
this section focuses on the combinatorics of words to find and combine cycles in a de bruijn graph .
word _ of length @xmath1 over an alphabet @xmath195 is a sequence of @xmath1 _ letters _ , each of which is an element of @xmath195 .
we often use the _ left rotation _ operation @xmath196 , which cyclically permutes the order of letters in a word : @xmath197 .
we may rotate a word by any amount by repeatedly applying @xmath196 ; @xmath198 rotates a word by @xmath18 places to the left .
usually , a _ factor _ @xmath199 of a word @xmath200 is defined as any block of consecutive letters in @xmath200 , and @xmath199 is a _ prefix _ if it appears at the start of @xmath200 . in this case , if @xmath200 has length @xmath1 , there are at most @xmath201 factors of length @xmath2 ( or @xmath2-factors ) of @xmath200 . for this section ,
we allow factors to `` wrap around '' , so that @xmath199 is a factor of @xmath200 if and only if it is a prefix of some rotation @xmath202 .
this way , it is possible to have @xmath1 different @xmath2-factors of a word of length @xmath1 .
a @xmath0-ary _ necklace _ is an equivalence class of words over the alphabet @xmath203 under cyclic rotation @xmath198 .
the _ length _ of a necklace is the length of any word in the class , while the _ size _ of a necklace is the number of words in the class .
a necklace with equal length and size is called _
every word in a @xmath0-ary length @xmath2 necklace corresponds to a vertex in the de bruijn graph @xmath3 , and a cyclic rotation corresponds to following an edge in this graph .
thus a size @xmath7 necklace can be thought of as a @xmath7-cycle in @xmath3 .
note that every vertex is part of some necklace , so the vertex set of @xmath3 can be partitioned into necklaces . for a fixed @xmath0 and @xmath2 ,
the possible necklace sizes in @xmath3 are the divisors of @xmath2 .
moreau s necklace counting function @xcite , shown below in , gives the number of @xmath0-ary size @xmath7 necklaces , and is defined in terms of the mbius function @xmath204 .
@xmath205 the total number of length @xmath2 necklaces is more easily calculated using euler s totient function , @xmath206 : @xmath207 when @xmath2 is prime , there are exactly @xmath0 necklaces of size @xmath90 ( the constant words ) ; the remainder of the necklaces have size @xmath2 .
thus there are @xmath208 disjoint @xmath2-cycles in @xmath3 .
hence @xmath209 for any prime @xmath2 .
note that when @xmath210 , this lower bound is tight since there are less than @xmath2 remaining vertices in @xmath3 . a _ lyndon word _ is the lexicographically smallest representative of an aperiodic necklace .
it is possible to construct a de bruijn sequence for @xmath3 by concatenating all @xmath0-ary lyndon words whose length divides @xmath2 in lexicographic order .
in fact , the sequence that is generated is the lexicographically smallest de bruijn sequence of the given order @xcite .
suppose we have two systems of coloured ebugs , where each colouring is @xmath2-valid .
in this section , we describe a type of direct product that yields many more ebugs at the expense of using more colours .
the result is summarised in theorem [ thm : product colouring ] . instead of modelling the colouring problem with de bruijn graphs
, we find a set of necklaces of length @xmath1 that correspond to the disjoint @xmath1-cycles in @xmath3 . the definition of an _
@xmath2-valid _ set of necklaces translates directly from definition [ def : ebug number ] .
we first demonstrate this proof for the special case of @xmath211 ( so @xmath212 ) , and then show that the construction can be extended to the general case .
the construction describes a one - to-@xmath1 mapping from pairs of necklaces to necklaces with @xmath213 colours . in order to construct necklaces over a larger alphabet ,
we use pairs of letters ( colours ) as the letters in the resulting necklaces .
we define a _ merging _ operation @xmath214 that pairs corresponding letters from two words of the same length : if @xmath215 and @xmath216 , then @xmath217 .
let @xmath218 be an @xmath2-valid set of @xmath219 @xmath220-ary necklaces of length @xmath1 , for @xmath221 .
for each necklace @xmath222 , choose a representative word @xmath223 .
now , for a pair of necklaces @xmath224 , we construct @xmath1 new words @xmath225 over @xmath226 , where @xmath227 .
note that we only rotate one of the words , since rotating both by the same amount creates a word that is equivalent under cyclic rotation ( it is only the relative rotation that matters ) .
an example of this process with @xmath228 , @xmath229 and @xmath230 is illustrated in figure [ fig : product colouring ] .
this process can be performed for every pair @xmath224 , generating @xmath1 new words every time .
hence it is clear that @xmath231 words are produced , and that @xmath213 colours are used .
thus it remains only to show that the set of corresponding necklaces is @xmath2-valid .
in each of the original necklaces @xmath222 , there are @xmath1 different @xmath2-factors ( since the @xmath218 were @xmath2-valid ) .
thus the total number of distinct @xmath2-factors in @xmath218 is @xmath232 .
suppose the word @xmath233 occurs in the necklace @xmath234 , and @xmath235 occurs in @xmath236 .
there is a unique @xmath18 such that @xmath237 and @xmath238 have @xmath239 and @xmath240 aligned , so @xmath241 must contain the factor @xmath242 .
since there are @xmath243 pairs of @xmath2-factors from the original necklaces , there must be at least @xmath243 distinct @xmath2-factors in the set of merged words .
but there are only @xmath1 possible @xmath2-factors in each of the @xmath231 merged words , so each @xmath2-factor must appear exactly once .
therefore the set of necklaces corresponding to the merged words is @xmath2-valid . to generalise to the case where @xmath244
, we can traverse the original necklaces multiple times to obtain words of length @xmath245 .
for a necklace @xmath246 , pick a representative word of @xmath14 and repeat it @xmath247 times to obtain @xmath223 . this way we may still merge words from @xmath248 and @xmath249 using @xmath214 ( since they are the same length ) . to model rotations of @xmath250 and @xmath251 , we act on the pair with elements of the group @xmath252 .
simultaneous rotation of @xmath250 and @xmath251 ( by the same amount ) only rotates the merged word that is produced , so we may identify unique merged words with elements of the quotient group @xmath253 .
hence each pair @xmath254 produces @xmath255 unique words , for a total of @xmath256 words . it remains to show that the set of merged words is @xmath2-valid .
the number of @xmath2-factors that appear in @xmath218 is @xmath257 .
as before , suppose the @xmath2-factor @xmath239 occurs in @xmath234 and @xmath240 occurs in @xmath236 .
define @xmath18 and @xmath19 so that @xmath239 is the @xmath2-prefix of @xmath258 , and @xmath240 is the @xmath2-prefix of @xmath259 .
note that rotating both words together keeps @xmath239 and @xmath240 aligned , so the pair @xmath16 corresponds to a unique element of @xmath260 , and thus @xmath261 appears in one of the merged words .
hence there are at least @xmath262 distinct @xmath2-factors in the set of merged words .
the number of possible positions for these @xmath2-factors is the product of the number of merged words and the length of each word .
this number is @xmath263 , so all @xmath2-factors are unique , and the set of corresponding necklaces is @xmath2-valid .
hence @xmath264 .
the conditions in theorem [ thm : product colouring ] guarantee that if the original colourings are optimal , then the resulting colouring is also optimal .
this allows a result for prime powers , such as theorem [ thm : lfsr translation ] , to be extended to any integer by repeated application of theorem [ thm : product colouring ] ( after applying theorem [ thm : lfsr translation ] for each prime power factor in the prime decomposition of @xmath0 ) .
[ cor : db(q , l ) always decomposes into q cycles ] @xmath29 for all @xmath0 and @xmath2 . on the other hand ,
if the original colourings are not optimal , the product colouring may be even `` less '' optimal .
for example , the construction in section [ sec : non - primitive lfsr ] is almost optimal since it uses all but one vertex of @xmath265 .
if we use theorem [ thm : product colouring ] to multiply this with an optimal colouring ( where the @xmath1-cycles use every vertex of @xmath266 ) , there would be @xmath267 vertices of @xmath268 not used by a cycle in the product colouring .
let @xmath269 be an @xmath2-valid set of @xmath9 necklaces of length @xmath1 over the alphabet @xmath203 .
for each necklace @xmath270 , fix a specific representative word @xmath223 .
we now construct the set @xmath271 of all words that appear in some necklace from @xmath269 : @xmath272 since @xmath269 is @xmath2-valid , the necklaces of @xmath269 are aperiodic
. thus we may uniquely identify how far each word in @xmath273 is rotated from its representative @xmath223 .
for each word @xmath274 , define @xmath275 .
we now define a function @xmath276 that interleaves the letters of multiple words to construct a single long word .
@xmath277 note that @xmath276 is injective , since we can deinterleave the resulting word to recover the original words .
we write @xmath278 for this deinterleaving function , which gives a @xmath7-tuple of @xmath1 words from a single word of length @xmath6 . for brevity , we write @xmath7-tuples of words in boldface as @xmath279
. we also extend @xmath280 to operate on @xmath7-tuples , so we may write @xmath281 .
now consider the set @xmath282 .
note that the @xmath283 are in @xmath284 and @xmath7 divides @xmath1 , so taking the sum modulo @xmath7 is well - defined .
simple arithmetic shows that @xmath285 .
we claim that @xmath286 is a @xmath287-valid set of words .
take any @xmath288 from @xmath286 , and observe the following property of the interleaving function : @xmath289 thus if @xmath290 , then @xmath291 . hence rotating @xmath38 to the left increases @xmath292 by 1 . by iterating
, we can find all @xmath6 rotations of @xmath38 .
of these @xmath6 rotations , only @xmath1 have @xmath293 ( every @xmath7-th rotation ) , and only one of these @xmath1 has the first word not rotated ( @xmath294 ) .
hence no other @xmath295 is a rotation of @xmath38 .
now suppose there are @xmath296 that share a @xmath287-factor @xmath37 ( in any position ) .
rotate @xmath38 and @xmath297 , respectively , to find @xmath298 and @xmath299 , both of which have @xmath37 as their @xmath287-prefix .
note that @xmath298 and @xmath299 are not necessarily in @xmath286 , but are still obtained by interleaving a @xmath7-tuple of words from @xmath273 ( see ) .
we may deinterleave @xmath37 into a @xmath7-tuple of length @xmath2 words @xmath300 .
each of these words @xmath301 is the @xmath2-prefix of a unique word @xmath302 due to the @xmath2-validity of @xmath269 .
thus the only interleaved word with @xmath37 as its @xmath287-prefix is @xmath303 , and hence @xmath304 . but @xmath297 is not a rotation of @xmath38 , so @xmath305 .
therefore , @xmath286 is a @xmath287-valid set of length @xmath6 words and @xmath306 . as an example
, we now apply theorem [ thm : interleaving ] with @xmath307 , @xmath228 , @xmath229 and @xmath308 to the eight necklaces of length 8 in figure [ fig : product colouring ] , which in turn were constructed using theorem [ thm : product colouring ] . these necklaces may be more compactly written as strings : @xmath309 .
suppose we wish to interleave the first and second necklaces in all allowable rotations .
we must keep the first necklace fixed and only rotate the second necklace by even amounts to satisfy the conditions of the set @xmath286 : @xmath310 we may now interleave these necklaces to obtain the following four necklaces of length 16 : @xmath311 but this is for just one pair of necklaces ; we can repeat this procedure for every ordered pair of necklaces from our original list of eight . for each of the 64 possible pairs , we produce four new necklaces of length 16 , yielding a total of 256 necklaces .
these are listed in table [ tab : interleaving example ] , with each line corresponding to a particular pair of necklaces .
hence @xmath312 .
@xmath313 as with theorem [ thm : product colouring ] , if the original colouring is optimal , then the colouring obtained by interleaving the necklaces is also optimal .
this allows us to extend existing results by recursively applying theorem [ thm : interleaving ] .
if every prime factor of @xmath7 divides @xmath0 , then @xmath314 recall that @xmath50 and @xmath29 ( see corollary [ cor : db(q , l ) always decomposes into q cycles ] ) .
now apply theorem [ thm : interleaving ] repeatedly using each prime factor of @xmath7 .
note that since each prime factor of @xmath7 divides @xmath0 , and each value of @xmath1 in this process is a multiple of @xmath0 , theorem [ thm : interleaving ] is applicable at each step . we also have a partial extension of theorem [ thm : interleaving ] , which allows us to interleave a pair of necklaces of odd length : [ thm : interleaving 2 ] @xmath315 .
the proof of theorem [ thm : interleaving 2 ] is analogous to that of theorem [ thm : interleaving ] , except that a slightly different condition is used in the construction of the set @xmath286 .
here , @xmath316 .
whenever two length @xmath2 necklaces share an @xmath317-factor , the corresponding cycles in @xmath3 can be concatenated .
this is because the corresponding edge circuits in @xmath41 have a common vertex , and can thus be joined to create a larger circuit , which in turn gives a larger cycle in @xmath3 .
this relationship between necklaces turns out to be very useful , so we construct a _ necklace adjacency graph _ @xmath318 .
the @xmath0-ary length @xmath2 necklaces form the vertex set of @xmath318 , while pairs of necklaces that share an @xmath317-factor are joined by an edge .
consider any ( connected ) subtree @xmath319 in @xmath318 . by applying the above operation for each edge in @xmath319
, the cycles for each necklace in @xmath319 can be concatenated together to produce one long cycle , whose length is the sum of the sizes of the individual necklaces .
hence if we find a spanning forest in @xmath318 where each component subtree has @xmath1 as the total size of its necklaces , we can partition @xmath3 into @xmath1-cycles .
if the forest does not span @xmath318 , this still gives a lower bound on the ebug number : if there are @xmath133 component subtrees in the forest , each with total necklace size @xmath1 , then @xmath320 . in the case when @xmath2 is prime , recall that most necklaces have size @xmath2 .
let @xmath321 be the subgraph of @xmath318 induced by the size @xmath2 necklaces .
suppose that there is a perfect matching in @xmath321 : this is a spanning forest , and each component subtree has total size @xmath322 . hence concatenating the cycles for each pair of necklaces in the matching produces @xmath323 cycles in @xmath3 , each of length @xmath322 .
similarly , we may generalise this to larger multiples of @xmath2 : [ prop : necklace joining ] let @xmath2 be a prime . if there is a spanning forest of @xmath321 in which each component subtree has @xmath7 vertices , then @xmath324 , with equality if @xmath325 .
in particular , the existence of a hamiltonian path in @xmath321 is sufficient to apply proposition [ prop : necklace joining ] for any @xmath7 that divides @xmath208 ( removing every @xmath7-th edge from the path produces the required forest ) . a hamiltonian path in @xmath318
is called a _ gray code _ for necklaces , and is conjectured to exist whenever @xmath0 or @xmath2 is odd @xcite ( if both @xmath0 and @xmath2 are even , a simple parity argument reveals that @xmath318 is bipartite with unequal parts ) .
it is a trivial matter to transform the hamiltonian path in @xmath318 to one in @xmath321 , since all neighbours of a constant necklace are adjacent to each other . for @xmath326
, there is also an existing construction of a _
2-gray code _ for fixed density necklaces @xcite , which lists every necklace with a fixed number of @xmath95s such that consecutive necklaces differ in exactly two places ( a @xmath95 and @xmath90 are exchanged ) . in some applications of robot identification ,
orientation information is either not required or can be obtained through different means @xcite .
this relaxes the conditions necessary to obtain a valid colouring of ebugs , since a particular colour sequence may appear more than once on a single ebug . in the de bruijn graph , this amounts to finding the maximum number of disjoint closed @xmath1-walks ( instead of @xmath1-cycles ) .
when @xmath2 is a divisor of @xmath1 , there is an easy solution that turns out to be the best possible ; this is a direct corollary of golomb s conjecture .
[ prop : necklaces for identification ] if @xmath2 is a divisor of @xmath1 , then the maximum number of pairwise disjoint closed @xmath1-walks in @xmath3 is @xmath327 , the number of @xmath0-ary length @xmath2 necklaces .
golomb s conjecture , which was proved by mykkeltveit @xcite , states that the maximum number of pairwise disjoint cycles ( of any length ) in @xmath3 is @xmath327 .
since each closed walk contains a cycle , there are at most @xmath327 pairwise disjoint closed @xmath1-walks in @xmath3 .
now consider any length @xmath2 necklace .
the size @xmath7 of this necklace must divide @xmath2 and hence @xmath1 , so the corresponding @xmath7-cycle in @xmath3 can be traversed multiple times to obtain a closed @xmath1-walk .
thus there are @xmath327 pairwise disjoint closed @xmath1-walks in the graph .
the theorems presented in sections [ sec : lfsrs ] and [ sec : necklaces ] focus on finding large ebug colourings to obtain bounds on the ebug number @xmath9 .
in particular , we concentrated on constructions that yield optimal colourings to support conjecture [ conj : e always optimal ] .
the algebraic construction in theorem [ thm : lfsr translation ] , and its corresponding extension in corollary [ cor : db(q , l ) always decomposes into q cycles ] , produces @xmath0 ebugs of maximal size for any @xmath0 and @xmath2 .
the combinatorial results in section [ sec : necklaces ] increase the number of ebugs by moving to a larger de bruijn graph .
these results can even produce ebug colourings for practical applications : for example , we have @xmath328 by theorem [ thm : lfsr translation ] , so @xmath329 by theorem [ thm : product colouring ] .
the current ebugs have 16 leds , a camera can distinguish four colours in an image quite easily , and five consecutive leds are visible in practice , so it is possible to construct a network of 64 uniquely identifiable ebugs .
also , theorem [ thm : interleaving ] can be used to significantly increase the number of ebugs without increasing the number of colours , @xmath0 .
moreover , it keeps the ratio @xmath330 constant , which is a reasonable assumption when designing such a network ( since the camera can see a fixed arc of the circle of leds ) .
unfortunately , the only way to produce many ebugs ( more than @xmath0 ) with our results is by applying theorem [ thm : product colouring ] or theorem [ thm : interleaving ] , which necessarily increase either @xmath0 , the number of colours , or @xmath2 , the number of consecutive visible leds .
the major remaining gap appears to be for prime @xmath0 and @xmath2 , with @xmath331 , since the multiplying / interleaving constructions can not produce them . ideally , we would like to be able to have many small ebugs ( small number of leds ) with a small number of colours .
for example , we would like to show that @xmath332 when @xmath2 is large enough ( for each @xmath18 ) .
probabilistic arguments may be useful in finding such colourings , but a constructive approach is preferable for applications to robot networks ( the search space becomes too large even for small practical examples , and in many cases the colourings appear to be quite rare ) .
solving problem [ prob : determine e ] is much harder , since there are no optimal colourings in the cases not covered by conjecture [ conj : e always optimal ] . improving the lower bounds in these cases
, however , is likely to be a much easier task .
we have shown that near - optimal colourings exist in many cases .
we would like to thank heiko dietrich for introducing us to zsigmondy s theorem , and suggesting that it can be used to find examples to which theorem [ thm : non - primitive lfsrs ] can be applied .
we also thank the referees for pointing out some related references and for their helpful suggestions .
10 [ 1]#1 url@samestyle [ 2]#2 [ 2 ] l@#1 = l@#1 # 2 b. alspach , h. gavlas , m. ajna , and h. verrall , `` cycle decompositions .
iv . complete directed graphs and fixed length directed cycles , '' _
j. combin .
theory ser .
a _ , vol .
103 , no . 1 , pp . 165208 , 2003 .
b. d. o. anderson , c. yu , b. fidan , and j. m. hendrickx , `` rigid graph control architectures for autonomous formations : applying classical graph theory to the control of multiagent systems , '' _ ieee control systems magazine _ , vol .
28 , no . 6 , pp .
4863 , 2008 .
d. bryant , `` cycle decompositions of complete graphs , '' in _ surveys in combinatorics 2007 _ , ser .
london math .
lecture note ser.1em plus 0.5em minus 0.4emcambridge univ . press , 2007 , vol .
6797 .
n. dademo , w. l. d. lui , w. h. li , y. a. ekerciolu , and t. drummond , `` ebug an open robotics platform for teaching and research , '' in _ australasian conference on robotics and automation_.1em plus 0.5em minus 0.4emaustralian robotics & automation association , 2011 .
a. k. das , r. fierro , v. kumar , j. p. ostrowski , j. spletzer , and c. j. taylor , `` a vision - based formation control framework , '' _ ieee transactions on robotics and automation _ ,
18 , no . 5 , pp . 813825 , 2002 .
p. erds and l. lovsz , `` problems and results on @xmath333-chromatic hypergraphs and some related questions , '' in _ infinite and finite sets _ , ser .
colloq . math .
jnos bolyai.1em plus 0.5em minus 0.4emnorth - holland , 1975 , vol . 10 , pp . 609627
. x. wang , y. a. ekerciolu , and t. drummond , `` vision - based cooperative pose estimation for localization in multi - robot systems equipped with rgb - d cameras , '' _ robotics _ , vol . 4 , no . 1 , pp
. 122 , 2015 , special issue on coordination of robotic systems .
x. wang , y. a. ekerciolu , and t. drummond , `` a real - time distributed relative pose estimation algorithm for rgb - d camera equipped visual sensor networks , '' in _ proceedings of the seventh acm / ieee international conference on distributed smart cameras ( icdsc 2013 ) _ , palm springs , california , usa , oct . 2013 . | we propose a new camera - based method of robot identification , tracking and orientation estimation .
the system utilises coloured lights mounted in a circle around each robot to create unique colour sequences that are observed by a camera .
the number of robots that can be uniquely identified is limited by the number of colours available , @xmath0 , the number of lights on each robot , @xmath1 , and the number of consecutive lights the camera can see , @xmath2 . for a given set of parameters
, we would like to maximise the number of robots that we can use .
we model this as a combinatorial problem and show that it is equivalent to finding the maximum number of disjoint @xmath1-cycles in the de bruijn graph @xmath3 .
we provide several existence results that give the maximum number of cycles in @xmath3 in various cases .
for example , we give an optimal solution when @xmath4 .
another construction yields many cycles in larger de bruijn graphs using cycles from smaller de bruijn graphs : if @xmath3 can be partitioned into @xmath1-cycles , then @xmath5 can be partitioned into @xmath6-cycles for any divisor @xmath7 of @xmath1 .
the methods used are based on finite field algebra and the combinatorics of words .
= 1 graph theory , robot network , de bruijn graph , graph decomposition , pose estimation , linear feedback shift register |
in recent years our understanding of the x - ray cooling flow phenomenon in galaxy cluster cores has been revolutionised . throughout most of the 1980s and 1990s
, x - ray observations suggested that gas in the central 100 is cooling out at rates of up to several hundred solar masses per year , but the lack of evidence for a reservoir of cooled gas led to heated debate ( summarised by fabian 1994 ) over this interpretation of the x - ray data .
results from _ xmm - newton _ and _ chandra _ have since led to a sharp downward revision in x - ray cooling rates ( e.g. schmidt , allen & fabian 2001 ) and also reveal a strong deficit of line emission from gas cooling below @xmath4 ( peterson et al .
the implication is that x - ray cooling is quenched , for which numerous mechanisms have been proposed , including : rapid mixing of hot and cold phases , inhomogeneously distributed metals in the intracluster medium ( fabian et al .
2001,2002 ) ; active galactic nucleus ( agn ) heating by jets ( brggen & kaiser 2003 ) and sound waves ( fabian et al .
2003 ) ; thermal conduction of heat from the hotter outer parts of the cluster into the cooler core ( voigt et al . 2002 ) ; a significant relativistic cosmic ray component frozen into the thermal gas ( cen 2005 ) ; the release of gravitational energy from blobs of gas which detach from the bulk flow and fall directly into the core ( fabian 2003 ) . concurrently
, significant progress has been made in identifying cool gas and dust in cluster cores .
edge ( 2001 ) detected co emission in the centres of 16 cooling flows , consistent with @xmath5 of 2 at 2040 for a standard co:2 conversion ( see also salom & combes 2003 ) .
these are roughly the masses expected , given the revised cooling rates and likely ages .
interferometry shows further that the co emission is localised within the central few arcsec of the cluster ( edge & frayer 2003 ; salom & combes 2004 ) .
the frequent occurrence of smaller masses ( @xmath6 ) of hot 2 has also been established ( e.g. edge et al .
2002 ; jaffe , bremer & van der werf 2001 ) , and excitation analysis suggests that this hot 2 is a high pressure , transiently - heated component ( wilman et al .
2002 ) . both co and 2 emissions correlate well with the strength of the h@xmath0 emission from ionized gas at @xmath7 , whose prevalence in these environments , often in the form of spectacular filaments , has long been known ( e.g. hu et al . 1983 ;
crawford et al . 1999 ) .
despite the clear association between optical line emission and short central x - ray cooling times ( peres et al .
1998 ; bauer et al . 2005 ) , their physical relationship is ill - understood .
photoionisation by the observed excess population of hot massive stars can energetically account for the h@xmath0 luminosities in the most luminous systems ( allen 1995 ; crawford et al .
_ spitzer _ mips photometry of 11 ccgs by egami et al .
( 2006 ) also shows that the most h@xmath0-luminous in their sample ( a1835 , a2390 and zw3146 ) have prominent far - infrared thermal dust emission plausibly powered by star formation , two of them with @xmath8 . at lower h@xmath0 luminosities
the picture is less clear : the tapping of energy from the intracluster medium ( icm ) through turbulence ( crawford & fabian 1992 ) and heat ( sparks et al .
2004 ) are just two mechanisms which have been invoked to explain the optical nebulosity in such systems . in this paper
we present integral field spectroscopy of the ionized gas in the cores of four such clusters , a1664 , a1835 , a2204 and zw8193 .
the principal aim is to obtain a full two dimensional view of the distribution and kinematics of the gas through high resolution spectroscopy of the h@xmath0+[nii ] emission line , with additional ionization information being gleaned from the [ sii]@xmath96717,6731 and [ oi]@xmath96300,6363 lines where possible .
these four central cluster galaxies ( ccgs ) all have h@xmath0 luminosities exceeding @xmath10 , making them 4 of the top 6 most - h@xmath0 luminous systems in the extensive ccg spectroscopic survey by crawford et al .
( 1999 ) . in this regime of h@xmath0 luminosity , photoionisation by a young stellar population can account energetically for the luminosity of the h@xmath0 nebulosity ( allen 1995 ; crawford et al .
in addition to an analysis of the ccgs , we also present spectroscopy of other sources within the ifu field of view , including other cluster galaxies and ( in the case of a2204 ) a serendipitous gravitationally - lensed background galaxy .
we first present results for the clusters individually and then summarise and interpret their generic features . throughout the paper
we assume a cosmology with @xmath11 , @xmath12 and @xmath13 and all physical quantities quoted from other papers have been converted accordingly .
the observations of a1664 , a1835 and a2204 were taken in service mode on 2003 april 11 with the integral field unit ( ifu ) of the visible multiobject spectrograph ( vimos ) on ut3 of the 8.2 m very large telescope ( vlt ) at eso paranal in chile ( for further information on vimos see lefevre et al .
they were among the first taken with the vimos ifu in open time .
the ifu was operated in hr - red mode , offering a @xmath14 field of view covered by 1600 optical fibres of 0.67 diameter .
the fibres are coupled to a microlenses to ensure near - continuous sky coverage .
the field of view in this ifu mode is split into 4 quadrants of 400 fibres , three of which disperse the light with the hr - red grism over 62508700 , while the fourth quadrant employs the hr - orange grism spanning 55007450 .
the dispersion and spectral resolution are approximately 0.6 per pixel and 1.8 fwhm , respectively . for each of the three targets a pair of 1670s exposures
was taken , with a pointing dither of @xmath15 between them .
the seeing was in the range 0.51 throughout .
table 1 provides an observation log .
the data were reduced on site at the national institute for astrophysics ( inaf ) in milan using the vimos interactive pipeline and graphical interface ( vipgi ) ( scodeggio et al .
2005 ) developed by the virmos consortium .
the raw two - dimensional frames for each quadrant were bias - subtracted , and the spectra for the individual fibres were traced , extracted and flat - fielded .
the pipeline wavelength calibration was affected by instrumental flexure in the time interval between science and calibration observations .
the wavelength calibration was thus performed manually in iraf using night - sky emission features , rebinning to a common dispersion of 0.6 per pixel . using the mapping from detector to aperture fibre position ,
the four quadrants were united into a single @xmath16 fibre data cube for each pointing , covering the wavelength range 55008740 ( i.e. large enough to accommodate the wavelength range of the hr - red and hr - orange grisms ) .
the raw data format is such that the centres of adjacent fibres are separated by 5 pixels on the detector whilst the light from each fibre covers 3 pixels fwhm .
there is thus a small amount of fibre - to - fibre light leakage implying that the spectra of individual fibres will not be completely independent .
however , since the seeing is comparable to the fibre size and we are chiefly concerned with diffuse extended emission , the effect is not considered important and no corrections for it were made .
subtraction of the night - sky emission background was performed by fitting gaussian profiles to a triplet of sky features at 7316 , 7329 and 7341 in each fibre .
this intensity information was used to correct for fibre - to - fibre transmission variations across the field of view , and then to create a night - sky emission spectrum from a source - free region which was then subtracted across the whole cube . due to the use of different grisms ,
sky subtraction was performed separately for quadrants 13 and 4 , respectively . for a1664 and a2204 the two dithered sky - subtracted datacubes for each target
were then combined into a single datacube @xmath17 fibres in size ( @xmath18 ) .
for a1835 , only the second of the two pointings contains significant line emission , since the target was inadvertently placed too near the edge of the field of view .
flux calibration of the spectra was not performed , but comparison with published emission line slit fluxes ( crawford et al .
1999 ) suggets that h@xmath0 emission is detected down to minimum surface brightness levels in the range 0.51.0 @xmath19 .
following the reduction of the raw data , all subsequent cube manipulation and emission line analysis were performed within idl .
observations of the redshifted h@xmath0 emisson in zw8193 were made with the oasis ifu on the william herschel telescope ( wht ) on la palma on 2005 september 2 in 0.8 seeing and photometric conditions .
we used the mr807 filter which covers a wavelength range of 7690 - 8460 at a spectral resolution of @xmath20/@xmath21=2020 ( 2.2 per pixel ) .
the [email protected] field of view was sampled contigiously with 0.26 fibres .
the total integration time was 7200s , split into 4x30mins and each dithered by one ifu lenslet to account for bad pixels .
for further information about the oasis ifu at the wht see benn et al .
( 2003 ) .
the oasis data were reduced using the xoasis data - reduction package ( rousset 1992 ) developed at cral ( lyon ) .
the reduction steps include bias and dark subtraction , extraction of the spectra using an instrument model , wavelength calibration , cosmic - ray rejection , low frequency flatfielding and sky - subtraction .
the individual data - cubes were then mosaiced ( taking into account the small spatial offsets between exposures ) and resampled onto a common grid of 0.26 @xmath220.26 to create the final datacube .
the limiting emission line surface brightness is comparable to that reached in the vimos observations .
line fitting to the oasis data was performed by binning up @xmath23 pixels in order to better match the seeing .
.summary of the observations [ cols= " < , < , < , < , < , < , < , < " , ]
with a extinction - corrected h@xmath0 slit luminosity of @xmath24 , the central cluster galaxy of a1664 is among the brightest h@xmath0 emitters in the optical spectroscopic survey of the rosat brightest cluster sample ( bcs ; crawford et al .
1999 ) ; fits to the excess blue continuum light imply a visible star formation rate of 23 ( see also allen 1995 ) .
the complexity of the line emission in a1664 is evident in the reduced two - dimensional frame of one of the quadrants , shown in figure 1 .
the emission spans a large range in velocity and at many spatial locations two velocity components are required to satisfactorily model the h@xmath0+[nii]@xmath256548,6584 profiles . within each spatial pixel ,
the lines were fitted with zero , one or two h@xmath0+[nii]@xmath256548,6584 gaussian velocity components atop a flat continuum level ; within each velocity component the h@xmath0/[nii ] was allowed to vary but the [ nii]@xmath256548,6584 double ratio was fixed at the theoretical 1:3 ratio .
a simple chi - squared fitting criterion was used to decide whether one , two or zero velocity components were fitted at each position .
the complex was fitted within the wavelength range 73307450 ( i.e. 200 dispersion bins ) and the noise - level was estimated from the spectrum in the wavelength range 64907500 .
chi - squared values were evaluated for fits with 0 , 1 or 2 velocity components , @xmath26 , @xmath27 and @xmath28 , respectively . for @xmath29 ,
1 velocity component was fitted ; for @xmath30 , two components were deemed necessary .
all fits were examined by eye to confirm whether they were reasonable and to remove spurious fits to obvious noise spikes .
[ fig : a1664flux ] shows the h@xmath0 and continuum maps derived from these emission line fits . for clarity ,
three separate h@xmath0 maps are shown : the total flux ( i.e. all the h@xmath0 emission ) , and the separate red and blue component fluxes for those pixels fitted with two h@xmath0 velocity components .
the velocity and fwhm fields derived from these fits are shown in fig .
[ fig : a1664vel ] , and for clarity are again shown separately for the red and blue components where two components are required , and for those regions where only a single velocity component is needed .
line profile fits are shown in fig .
[ fig : a1664profs ] for selected regions of h@xmath0 emission in fig .
[ fig : a1664flux ] . from fig .
[ fig : a1664flux ] , we see that the h@xmath0 emission has a linear filamentary structure spanning approximately 13 arcsec ( 31 kpc ) and passing through the nucleus of the central cluster galaxy along a north - east south - west axis .
two velocity components are detected along its whole length : the red component peaks on the nucleus of the ccg , but the blue component exhibits two maxima ( labelled b1 and b2 ) located approximately 1.5 arcsec from the nucleus , along an axis perpendicular to the general direction of the filament .
this filament and nuclear bipolar structure are discussed further below .
the presence of two velocity components along the length of the filament ( most clearly at its ne end where complications from the nuclear emission of the ccg are minimal ) suggests that it might be due to a disturbed secondary emission component superimposed on a quiescent background , or a stream of gas wrapped around the ccg , or possibly related to some form of outflow .
it is plausibly related to the galaxy ( obj # 1 ) situated just beyond the sw extremity of the filament in fig .
[ fig : a1664flux ] , which is confirmed to be at the cluster redshift ( see fig . [
fig : a1664vel ] ) with a line of sight velocity of @xmath31 in the velocity system adopted in fig .
[ fig : a1664vel ] .
this gas may have been stripped from or disturbed by obj # 1 .
indeed , the velocity of the blue emission component at position b2 ( @xmath32 ) is similar to that of obj # 1 ( @xmath31 )
. the bipolar structure may instead be an outflow from the ccg driven by starburst activity .
the line of sight velocity difference between the blue emission components at b1 and b2 is @xmath33 .
this suggests a minimum deprojected outflow velocity of 175 , but the symmetrical location of b1 and b2 about the nucleus suggests that the outflow may be principally directed into the plane of the sky with a much higher velocity . in the latter case
, the observed radial extent of the outflow ( 1.5=3.5 ) can be used to constrain its age .
assuming a deprojected outflow speed of 600 as for the h@xmath0 filaments in m82 ( shopbell & bland - hawthorn 1998 ) , the implied age is @xmath34 myr . by contrast , the orbital timescale of galaxy # 1 responsible for the creation of the filament is @xmath35 myr ( assuming a circular orbit of radius 15 traversed at 500 ) .
considering the mis - match in these timescales , it is too coincidental that an outflow from a nuclear starburst triggered by the passage of galaxy # 1 should be observed to have a similar extent to the overall stream of gas .
we thus favour an interpretation in which the complex h@xmath0 structure is due to a disturbance induced by , or the stripping of gas from , galaxy # 1 .
to shed further light on the origin of the complex h@xmath0 distribution , we retrieved archival vlt - fors1 imaging of a1664 .
the dataset comprises images three 330 second exposures taken through the v , r , and i filters ( bessel photometric system ) on 2001 may 29 in photometric conditions and 0.8 seeing . following bias - subtraction and flat - fielding , we merged them into a true - colour image which is displayed in fig .
[ fig : a1664fors1 ] .
h@xmath0 contours from the vimos ifu observations were aligned and overlaid on this image .
the main features apparent from fig .
[ fig : a1664fors1 ] are that : ( i ) the north - eastern quadrant of the ccg is redder than other parts of the ccg , suggesting dust obscuration ; ( ii ) the south - western quadrant of the ccg is bluer than other cluster galaxies , suggesting more recent star formation ; ( iii ) the red h@xmath0 velocity component peaks on a nuclear dust lane which bisects the two central peaks of the blue h@xmath0 velocity component ; ( iv ) the southern extension of the h@xmath0 emission ( labelled b4 in fig .
[ fig : a1664flux ] ) coincides with a blue knot of continuum emission .
v r and r i colours ( expressed relative to the colours of the nearest a1664 cluster elliptical ) were extracted for regions b1,2,4 and object # 1 in fig .
[ fig : a1664flux ] . these show that ( i ) galaxy # 1 has the same colour as other cluster ellipticals at this redshift ; ( ii ) the ccg is @xmath36 mag bluer in @xmath37 and @xmath36 mag redder in @xmath38 and that the intrinsic dust extinction increases across it from south - west to north - east . from long - slit spectra ,
the intrinsic extinction inferred from the balmer decrement is in the range 0.460.63 mag ( crawford et al .
1999 ; allen 1995 ) .
corrected for this extinction , the same authors show that approximately 90 per cent of the ( rest - frame ) 4500 continuum is due to o5 and b5 stars giving rise to the 23 starburst .
the gradient of increasing dust extinction from sout - west to north - east could be due to the gradient in time elapsed since the passage of galaxy # 1 , which may have triggered star formation and dust production .
u or b - band imaging would more clearly show the distribution of recent star formation .
although the vimos data contain the emission lines of [ sii]@xmath96717,6731 and [ oi]@xmath256300,6363 in addition to [ nii]+h@xmath0 , we do not present an exhaustive study of the ionization properties of the gas .
instead , we impose the kinematics found for [ nii]+h@xmath0 and accordingly fit the [ sii ] and [ oi ] doublets with zero , one or two velocity components with only the line intensities allowed to vary , i.e. we do not search for emission in these lines where there is no [ nii]+h@xmath0 detection . in so doing
, we discover that the ionisation properties of the h@xmath0 gas are quite uniform , with no notable spatial gradients . in fig .
[ fig : a1664ions ] we show scatter plots of line ratios versus h@xmath0 flux for the combined sample of lenslets .
there is essentially no variation in [ nii]@xmath39/h@xmath0 and [ oi]@xmath40/h@xmath0 over more than one decade in h@xmath0 surface brightness .
these line ratios are consistent with the integrated slit values found by crawford et al .
the uniformity of the ionization state across the source supports our interpretation that all the gas ( the filament and the b1/b2 nuclear structure ) has the same origin . for comparison ,
the emission line filaments in the kiloparsec - scale superwind outflow in m82 have [ nii]@xmath39/h@xmath41 ( shopbell & bland - hawthorn 1998 ) . in the co line survey of cooling flow ccgs by edge ( 2001 ) ,
a1664 exhibits unusually broad co(1 - 0 ) emission spanning 620 full - width at zero intensity , with a total inferred mass of cool molecular hydrogen is @xmath43 m. considering the 23.5 arcsec beam size for these iram 30-m observations , its kinematics are a close match to those of h@xmath0 ; in particular , the skewness of the co line towards negative velocities suggests that the blue h@xmath0 velocity components may have associated co. a direct comparison is shown in fig .
[ fig : a1664coha ] : the co(1 - 0 ) profile is taken from edge ( 2001 ) , and the h@xmath0 profile is formed by integrating over the whole source and adding synthetic gaussians for the h@xmath0 emission only , using the parameters derived from the multi - component fits to h@xmath0+[nii ] .
this procedure is used in order to compare the kinematics without complications from the blended [ nii ] lines .
[ fig : a1664coha ] demonstrates that the widths and centroids of the h@xmath0 and co(1 - 0 ) emission are indeed comparable , and that much of the breadth of these lines can be attributed to the regions of kinematically disturbed h@xmath0 emission with two velocity components . for comparison , in the k - band ukirt cgs4 spectrum of edge et al .
( 2002 ) , the width of the pa@xmath0 line is @xmath44 fwhm , for a n - s slit of @xmath45 arcsec .
this is consistent with the observed h@xmath0 kinematics when the vimos data are extracted over the same area and fitted with a single velocity component . in contrast , the h@xmath42 v=1 - 0 s(3 ) line is not resolved , suggesting an intrinsic width much below the instrumental broadening of 570 fwhm .
this suggests that the hot h@xmath42 is distinct from the co and h@xmath0-emitting gas in the case of a1664 .
a1664 has been observed with the acis - s instrument on the _ chandra _ x - ray observatory for 10 . in fig .
[ fig : a1664chandra ] we show the smoothed 0.5 - 5 image ( courtesy of a.c .
this shows that the x - ray emission is also elongated along a ne - sw axis with a possible filamentary structure , resembling the h@xmath0 emission but on a much larger spatial scale .
a full analysis of the x - ray dataset will be presented elsewhere .
the observed similarity in h@xmath0 and x - ray morphology is analogous to that seen in the core of a1795 ( crawford , fabian & sanders 2005 ) , although in that cluster the emission line gas in the filament is very quiescent , in stark contrast to the case of a1664 . as fig .
[ fig : a1664chandra ] demonstrates , the ccg s small neighbour galaxy ( labelled object # 1 in fig . [ fig : a1664flux ] ) appears to coincide with a local minimum in the x - ray emission . . the absorption feature due to the na d @xmath46 blend is labelled , from which a redshift of @xmath47 is determined .
the feature at 68606880 is telluric absorption.,width=283 ] flux for individual fibres in a1664 .
results are shown separately for the red and blue velocity components ( where fitted ) , and also for fibres where a single velocity component was fitted.,title="fig:",width=359 ] flux for individual fibres in a1664 .
results are shown separately for the red and blue velocity components ( where fitted ) , and also for fibres where a single velocity component was fitted.,title="fig:",width=359 ] flux for individual fibres in a1664 .
results are shown separately for the red and blue velocity components ( where fitted ) , and also for fibres where a single velocity component was fitted.,title="fig:",width=359 ] ) close to the ccg.,width=321 ] and co(1 - 0 ) emission line profiles integrated over comparable areas of a1664 . the co(1 - 0 )
is derived from the iram 30 m observation by edge ( 2001 ) with a 23.5 beam size ( the solid and dashed lines show measurements with different correlators ) .
the h@xmath0 profile was constructed by summing over the whole source and using the multi - component gaussian fits to the h@xmath0+[nii ] complex the solid shows all the emission , whilst the dotted and dashed line show the contributions from fibres requiring one and two velocity components , respectively .
velocities are with respect to the nominal cluster redshift of @xmath48.,width=283 ]
at a redshift of @xmath49 , a1835 is the most distant cluster presented in this study .
it is also our most luminous h@xmath0 emitter , with an extinction - corrected luminosity of @xmath50 and a visible star formation rate of @xmath35 ( crawford et al .
1999 ; allen 1995 ) .
the _ spitzer _
mips photometry of egami et al .
( 2006 ) reveal far infrared thermal dust emission with @xmath51 , plausibly powered by star formation . in the x - rays ,
the cluster is the most luminous in the rosat bcs and subsequent _ chandra _ data are consistent with a young ( @xmath52 ) cooling flow operating within @xmath53 ( schmidt , allen & fabian 2001 ) . in comparison with
a1664 the kinematics are simple and relatively quiescent , requiring a single velocity component at each position .
the h@xmath0 map and velocity field are shown in fig .
[ fig : a1835flux ] .
the h@xmath0 emission has a maximum extent of 8 ( 31 ) and is elongated in a nw - se direction ; the reconstructed continuum image shows a small companion galaxy 5 ( 20 ) north - west of the ccg .
the emission velocity exhibits a shear of @xmath54 along the same axis , possible due to rotation , with linewidths of @xmath55 fwhm . as for a1664
, the [ nii]+h@xmath0 kinematic fits were used to search for associated emission in [ sii]@xmath56 .
unfortunately , the emission in [ oi]@xmath57 is badly affected by night sky emission line features and was not analysed . line ratios as a function of local h@xmath0 flux are shown in fig .
[ fig : a1835ions ] , but the only firm conclusion that can be drawn is that the [ nii]@xmath58/h@xmath0 is once again remarkably constant with luminosity and consistent with the integrated slit value measured by crawford et al .
the [ sii]@xmath59 ratio shows marginal evidence for an increase with h@xmath0 surface brightness , implying higher electron density as hinted at in a1664 .
a comparison with the co emission is once again valuable .
the iram 30 m single dish observation of edge ( 2001 ) shows co(1 - 0 ) from an inferred cool molecular mass of @xmath60 .
the co emission peaks at a velocity of @xmath61 ( relative to an assumed redshift of @xmath49 ) , identical to that of the h@xmath0 emission peak .
the co linewidth of @xmath62 fwhm is again a close match to the h@xmath0 .
subsequent owen valley millimetre array interferometry of the co(1 - 0 ) emission constrains the angular extent of the co to @xmath63 ( 36 kpc ) .
it thus appears that , as in a1664 , the co and h@xmath0 emission share the same kinematics and morphology .
flux for lenslets in a1835.,width=359 ]
abell 2204 lies at @xmath64 and has an observed h@xmath0 slit luminosity of @xmath10 ( uncorrected for reddening ) , although the inferred visible star formation rate is modest ( @xmath65 crawford et al .
the system is also a strong emitter of near - infrared h@xmath42 ro - vibrational emission ( edge et al .
2002 ) , and the co(1 - 0 ) emission from _ iram _
30-m observations implies a molecular gas mass of @xmath66 ( edge 2001 ) .
_ chandra _ x - ray observations reveal that the cluster has a disturbed core morphology , suggestive of a recent merger , with cold fronts at radii of @xmath67 kpc and 54.5 , and the central radio source is also disturbed with three components within 10 arcsec , aligned roughly n - s ( sanders , fabian & taylor 2005 ) .
results from the vimos data in h@xmath0 and the continuum are shown in fig .
[ fig : a2204flux ] , along with an hst wfpc2 f606w ( wide v - band ) continuum image .
the latter shows that the ccg is dusty and possibly interacting with galaxy # 1 , which is 15 away in projection at a relative velocity of 250 ( see jenner 1974 and section 5.1 ) .
the h@xmath0 emission also has an irregular morphology with a number of filaments emanating from the nuclear regions , possibly coincident with the dust obscuration .
the h@xmath0 velocity field has a tri - partite structure with three distinct clumps at velocities of @xmath68200 , 50 and 100 ( relative to @xmath64 ) .
the velocity width of the h@xmath0 is in most locations @xmath69 fwhm , except in the nuclear region where it rises to @xmath70 .
this increased velocity width is unlikely to be due to an agn because the emission line ratios are uncharacteristic of agn and show no spatial variation across the source .
chandra _ x - ray data show no evidence for a nuclear point source .
a similarly high velocity width in a spur to the north - east of the nucleus may be due to interaction between the emission line clouds and the radio source , whose core structure extends 5 arcsec from the nucleus in this direction .
the high h@xmath0 line width may explain the apparent faintness of the co(1 - 0 ) line ( edge 2001 ) .
vlt - isaac long - slit spectroscopy by jaffe , bremer & baker ( 2005 ) of a2204 shows that the pa@xmath0 and 1 - 0 s(1 ) trace each other very closely , spatially and kinematically .
narrow - band h@xmath0 imaging by the same authors shows that the line emission drops off quite sharply , being characterised by a gaussian of fwhm @xmath71 .
other clusters , in contrast , show a power - law emission extending out to radii of 40 and beyond .
as remarked above , the line ratios exhibit no spatial variations so are plotted as a function of local h@xmath0 flux in fig .
[ fig : a2204ions ] . the constancy of the [ nii]@xmath39/h@xmath0 and [ oi]@xmath40/h@xmath0 is striking . as in the other targets ,
there is marginal evidence for an increase in [ sii]@xmath72/h@xmath0 and [ sii]@xmath73/6717 with increasing h@xmath0 surface brightness .
flux for lenslets in a2204.,width=359 ] in fig .
[ fig : a2204contobj ] we show spectra of the three continuum sources visible in the vimos field of view ( fig .
[ fig : a2204flux ] ) alongside the ccg .
the hst wfpc2 image of the same field shows that objects # 1 and 2 are galaxies and object # 3 stellar . from the na d@xmath20 5892.5 absorption blend we determined redshifts of z=0.15096 ( # 1 ) and z=0.13908 ( # 2 ) ; the presence of z=0
h@xmath0 absorption confirms the stellarity of object # 3 .
galaxy # 1 is thus close in projection ( @xmath74 ) and velocity ( @xmath75 ) to the ccg of a2204 .
it is thus possible that the two are interacting .
galaxy # 2 is twice as far away in projection and offset by 3200 in velocity . ) .
redshifts were determined from the na [email protected] absorption blend ( # 1 and 2 ) and confirm their cluster membership .
object # 3 is a star .
these spectra have not been flux calibrated in either an absolute or relative sense .
the features between 68606880 and 76007700 are telluric absorption.,title="fig:",width=283 ] ) .
redshifts were determined from the na [email protected] absorption blend ( # 1 and 2 ) and confirm their cluster membership .
object # 3 is a star .
these spectra have not been flux calibrated in either an absolute or relative sense .
the features between 68606880 and 76007700 are telluric absorption.,title="fig:",width=283 ] gravitationally - lensed background galaxy in the field of a2204 .
( a ) part of the hst wfpc2 f606w frame showing the arc in the continuum ; objects 24 are labelled as in fig .
[ fig : a2204flux ] .
( b ) reconstructed image in the [ oii]@xmath93726,3729 doublet using the vimos data ; the surface brightness is in units of @xmath76 erg @xmath77 s@xmath78 arcsec@xmath79 .
( c ) a portion of the wavelength - calibrated 2d vimos spectral frame of a2204 in which the arc was discovered ; boxes higlight the [ oii]@xmath93726,3729 emission.,width=245 ] the _ hst _ image of the cluster in fig .
[ fig : a2204flux ] exhibits a gravitationally lensed arc which terminates on a bright knot of emission which we refer to as object # 4 .
the vimos data show [ oii]@xmath80 emission in 11 fibres around this position which , with a redshift of @xmath81 , we identify as the lensed background galaxy .
[ fig : a2204gravlens ] shows the portion of the calibrated 2d spectral frame in which this emission was discovered , together with the reconstructed emission line image .
the line is just at the edge of the 76007700 telluric absorption , for which we correct with the aid of the stellar continuum spectrum of object # 3 .
line emission is present along the full length of the arc as defined by the _
hst _ image .
the total [ oii ] flux from it is estimated to be @xmath82 , a figure obtained by comparing our h@xmath0 measurments of the ccg with the slit flux in crawford et al .
the velocity of the emission varies by less than @xmath83 , and the line width is the range 100150 fwhm .
this lensed arc would be suitable for a follow - up with a longer ifu observation at higher spatial resolution , possibly with adaptive optics due to the proximity of the moderately bright star # 3 ( r=15.3 ; b=16.5 mag ) .
the core of zw 8193 is complex in both space and velocity and initial optical spectroscopy of this cluster yielded discrepant redshifts for the emission ( @xmath84 ) and absorption line ( @xmath85 ) components ( allen et al .
[ fig : zw8193 ] shows that there are three galaxies within 23 arcsec of the ccg ; ukirt long - slit spectroscopy revealed pa@xmath0 emission in and between the ccg and its neighbour to the north , with an apparent velocity difference of @xmath86 ( edge et al .
the benefit of higher spatial and spectral resolution oasis ifu data now reveals the full complexity of the system .
the h@xmath0 emission consists of three distinct ` blobs ' , one centred on the ccg , one on a galaxy to the north and one slightly offset from the latter to the west .
the velocity ranges over @xmath87 ( relative to @xmath88 ) and reconstructing the placement of the 1.2-wide n - s slit used for the ukirt observations shows the consistency of the pa@xmath0 and h@xmath0 velocity fields .
the [ nii]/h@xmath0 ratio varies from 0.9 on the ccg to 0.7 at the position of the northern blobs .
the ratio [ sii]@xmath72/h@xmath89 and is also relatively constant across the emission .
the [ sii]@xmath73 line is badly affected by a night - sky emission line .
the extinction - corrected h@xmath0 slit luminosity quoted by crawford et al .
( 1999 ) is @xmath90 ( converted to the cosmology used here ) .
the co(1 - 0 ) observations of zw 8193 have yielded seemingly inconsistent results ( edge 2001 ) .
one observation revealed a broad plateau of emission from @xmath91 to @xmath92 but a subsequent observation failed to confirm it .
as remarked by edge ( 2001 ) , the presence of a flat - spectrum radio source in zw 8193 complicates the baseline subtraction .
the quoted upper limit on the molecular gas mass is @xmath93 . as in a1664 , and to a lesser extent a2204 ,
the h@xmath0 emission morphology and kinematics appear to be influenced by the close passage of a small cluster galaxy into the core .
despite the diversity and complexity of these individual clusters and the small sample size , a number of generic features are apparent .
they are : _ @xmath94 a possible association between disturbed h@xmath0 emission and secondary galaxies within 1020_. in a1664 it appears plausible on both morphological and kinematic grounds that the h@xmath0 filament and complex nuclear structure in the ccg have been produced by the infall of a small cluster galaxy .
this galaxy is now visible at the extremity of the filament .
since the mass of h@xmath0-emitting ionized gas is relatively small ( a few @xmath95 , as inferred from the h@xmath0 luminosity and the measured electron density ) , it is possible that a significant fraction of it may have been stripped from the infalling galaxy .
our preferred interpretation , discussed in section 8 , is that gas distribution has been disturbed by the infalling galaxy . in zw8193
there is extended h@xmath0 emission associated with a galaxy which lies within 6 ( projected ) of the ccg . in a2204 , the irregular h@xmath0 kinematics may result from a recent interaction with a nearby galaxy ( # 1 in fig .
[ fig : a2204flux ] ) , or disturbance from the irregular radio source . in a1835 ,
the h@xmath0 emission is elongated in the direction of a small companion galaxy 20 away .
another well - studied ccg showing strong evidence for the influence of a nearby secondary galaxy on the h@xmath0 emission is rxj 0820.9 + 0752 at @xmath96 ( bayer - kim et al .
2002 ) . _
@xmath94 uniformity of the ionization state .
_ despite the morphological and kinematic complexity of the h@xmath0 emission , the ionisation state of the gas is surprisingly uniform , as a function of position , h@xmath0 surface brightness and between clusters .
the [ nii]@xmath58/h@xmath0 and [ oi]@xmath97/h@xmath0 ratios are particularly tight , but there may be hints of a weak increase of [ sii]@xmath98/h@xmath0 and [ sii]@xmath59 with h@xmath0 surface brightness .
these results suggest that the same mechanism(s ) excite all the optical line emission in these systems , and that they are quite independent of the processes which disturb the gas morphologically and kinematically on the kiloparsec scales we observe , i.e. an infalling galaxy or a weak radio source may stir up the gas , but they seem not to impact on its ionization state .
for this reason , it seems unlikely that large - scale shocks are involved .
the line ratios we measure are naturally consistent with the integrated slit values found by crawford et al .
( 1999 ) and characteristic of high h@xmath0 luminosity ( @xmath99 ) cooling flow ccgs .
the latter show strong agn - like [ nii]@xmath58/h@xmath0 and [ oi]@xmath97/h@xmath0 , but weak [ sii]@xmath98/h@xmath0 and [ oiii]@xmath100/h@xmath101 similar to starbursts .
as mentioned in section 1 , it was shown by crawford et al .
( 1999 ) that the observed h@xmath0 luminosities of such systems can be powered by photoionisation by the o - star populations which explain their excess blue light .
typically @xmath102 0 stars and larger populations of b , a and f stars are required in excess of the continuum expected from a cluster elliptical . however , the ro - vibrational h@xmath42:pa@xmath0 line ratios of ccgs are much higher than typical starburst galaxies ( jaffe , bremer & van der werf 2001 ; edge et al .
this may require a distinct population of much denser clouds ( wilman et al .
2002 ) or excitation by a spectrum harder than that of o - type stars ( jaffe , bremer & baker 2005 ) .
this may also explain why the optical emission line ratios differ from those of starburst galaxies .
for the ensemble of line - emitting ccgs , crawford et al .
( 1999 ) discovered a continuous trend of decreasing [ nii]@xmath58/h@xmath0 with increasing h@xmath0 luminosity , and similar weaker trends in [ oiii]@xmath100/h@xmath101 and [ sii]@xmath72/h@xmath0 .
our results show that similar trends do not extend to _ local _
h@xmath0 surface brightness within the most luminous systems .
sparks et al .
( 1989 ) proposed that cool gas ( e.g. from a galaxy merger ) falling into the core of a galaxy cluster could be energised by thermal conduction from the x - ray - emitting icm .
this would lead to optical line emission from the cooler gas and a local enhancement in x - ray emission .
such a model was used to account for the close correspondence between the x - ray and h@xmath0 filaments in the virgo and perseus clusters ( sparks et al .
it is not known whether this mechanism can reproduce the observed emission line ratios . _ @xmath94 co and h@xmath0 trace the same gas , at least kinematically and probably also spatially .
_ two of the four clusters , a1664 and a1835 , have robust detections of co(1 - 0 ) emission in the survey of edge ( 2001 ) . in both these clusters the co emission profile is in very good agreement with the h@xmath0 kinematics , in terms of its centroid velocity and line width , strongly suggesting that the co and h@xmath0 emission arise from the same ensemble of clouds . given this strong co : h@xmath0 link , the weak ( a2204 ) or ambiguous ( zw8193 ) co detections in edge ( 2001 ) most likely stem from a high h@xmath0 velocity width or zero - point offset ( relative to the ccg redshift then assumed ) , respectively . in the aforementionned rxj 0820.9 + 0752 at @xmath96 ( bayer - kim et al .
2002 ) the co and extended h@xmath0 kinematics are also well - matched .
this is in line with the correlation between h@xmath0 luminosity and molecular gas mass discovered by edge ( 2001 ) .
in contrast , the ro - vibrational h@xmath42 emission and kinematics appear to be less strongly coupled to the h@xmath0 . in a1664
, the h@xmath42 has a much lower velocity width than pa@xmath0 and co , although in a2204 the match is much closer .
rxj 0820.9 + 0752 has weak or non - existent h@xmath42 emission in the edge et al .
( 2002 ) survey , despite its strong extended co and h@xmath0 emission .
we note , however , that jaffe , baker & bremer ( 2005 ) discovered that the pa@xmath0 and hot 2 trace each other spatially and kinematically out to radii beyond 20 , but a2204 is the only cluster common to our two samples .
we now interpret the above findings , taking as our starting point the co observations .
as reviewed in section 1 , these have revealed substantial masses of cool molecular gas ( @xmath103 , at 2040 ) in the cores of cooling flows , with a mass which correlates with the h@xmath0 luminosity ( edge 2001 ) .
subsequent interferometry has shown that the co emission is confined to scales @xmath3 ( edge & frayer 2003 ) .
the lack of double - peaked co line profiles suggests that the molecular cloud distribution is quasi - spherical , rather than disk - like . such a dense , compact molecular gas reservoir
will most likely be disturbed by the infall of a small galaxy into the cluster core , as we believe is happening most clearly in a1664 , or by the onset of a nuclear radio source .
we consider these possibilities in more detail below .
first we consider some physical processes in the unperturbed molecular gas reservoir and their observable consequences .
it is plausible that the co and h@xmath0 emission arise within the same clouds , because they appear to share the same kinematics and are closely correlated in luminosity .
the ionised surfaces of such clouds are likely to be in pressure equilibrium with the hot x - ray gas for which @xmath104 , @xmath105 ; for optical emission line regions @xmath106 , @xmath107 , such that @xmath108 is the same for both phases .
pressure equality may also extend to the cool molecular gas if @xmath109 , although it is possible that the cloud cores are self - gravitating .
the clouds may be warmed / ionised by a combination of young stars and x - ray emission / conduction from the intracluster medium , but the precise details do not concern us here . the rovibrational h@xmath42 emission may be a distinct , transiently - heated high - pressure component , since line ratio analysis points to thermal excitation in dense gas with @xmath110 and @xmath111 ( jaffe , bremer , van der werf 2001 ; wilman et al .
wilman et al .
( 2002 ) proposed that shocks between a population of low - density ( @xmath112 ) h@xmath0-emitting clouds could be used to transiently create such high - density molecular gas , although an external stellar radiation field was still used to excite the observed emission .
we now modify this model and investigate collisions amongst the denser co clouds .
consider that a mass @xmath113 of co - emitting molecular gas is confined within a radius @xmath114 of the ccg , distributed within an ensemble of identical clouds of radius @xmath115 , mass @xmath116 , space density @xmath117 and hydrogen number density @xmath118 .
interferometry shows that the co is centrally concentrated so we take @xmath119@xmath78 , as in the inner part of an nfw profile ( navarro , frenk & white 1997 ) .
for a typical relative cloud velocity of @xmath120 , the rate of cloud collisions per unit volume is @xmath117@xmath121 @xmath122 , where @xmath123 is the collision cross - section , which we approximate with the geometrical cross - section @xmath124 .
it then follows that the rate at which mass is processed through such collisions is : @xmath125 for @xmath126 , @xmath127 , @xmath128 and @xmath129 , eqn .
( 1 ) evaluates to @xmath130 for @xmath131 ( m@xmath132 is the proton mass ) .
such supersonic collisions will shock the clouds to temperatures of several @xmath133 . for the assumed densities
the cooling time from these temperatures is very short , @xmath65 month , so the gas will quickly return to the cool phase , i.e. collisions will merely recycle the cool gas and not deplete it .
it is , however , possible the shock compression will act as a trigger for star formation .
they may also create the dense molecular gas and a locally strong , hard ( e.g. black body at @xmath134 ) radiation field required for the production of the rovibrational h@xmath42 emission .
the euv luminosity produced by such shocks is @xmath135 , or : @xmath136 for the same parameters as before , this amounts to @xmath137 .
although this is several orders of magnitude below the inferred luminosity of the young stellar component in these systems , the cubic dependence on @xmath120 and @xmath114 and uncertainty over the actual values of @xmath115 and @xmath118 means that such shocks could make a non - negligible contribution to the excitation of the line emission in ccgs , at least locally if the gas clouds are not uniformly distributed .
we note , however , that limits on the strength of the [ oiii]@xmath138 emission line imply that shocks are unlikely to play a strong role in the production of the optical line emission in h@xmath0-luminous ccgs ( e.g. voit & donahue 1997 ) .
the mechanism may , however , be a relatively more important source of excitation in lower - h@xmath0 luminosity systems .
such shocks will steadily drain energy from the molecular gas reservoir , forcing the distribution to smaller radii and thus increasing the collision rate still further .
we now consider the consequences of disturbing this reservoir with the infall of a small galaxy , as appears to be happening most dramatically in a1664 . from the constant excitation state of the ionised gas
, it seems that this does not cause any large scale shocks or cloud destruction .
most likely , the cloud distribution will be dynamically perturbed by the infalling galaxy , producing the observed streams and filaments .
such interactions are likely to enhance the star - formation rate , either by increasing the cloud - cloud collision rate ( as above ) , or via some other means . with reference to the relationship between [ nii]/h@xmath0 and h@xmath0 luminosity discovered by crawford et al .
( 1999 ) , the observed uniformity of the line ratios we find in our sample may be due to the fact that , at high h@xmath0 luminosity , all parts of the ccg are saturated at the lowest [ nii]/h@xmath0 values characteristic of massive star formation . in less h@xmath0-luminous ccgs , we would thus predict more variations in the line ratio , with [ nii]/h@xmath0 increasing away from the less - abundant star - forming clumps .
the possibility that galaxy interactions may trigger star - formation in these molecular gas - rich environments highlights the similarities between these ccgs and luminous infrared galaxies , as discussed by edge ( 2001 ) .
the _ spitzer _
observations of egami et al .
( 2006 ) further strengthen the connection , revealing large thermal dust luminosities @xmath8 in two of the most h@xmath0-luminous ccgs ( zw3146 and a1835 ) .
the results presented in this paper have shed new light on the kinematic and morphological complexity of the emission line gas in the cores of these h@xmath0-luminous clusters .
our principal findings are that : @xmath94 in many cases the h@xmath0 emission appears to have been disturbed by interaction with a small companion galaxy to the central galaxy . @xmath94
the co and h@xmath0 emission share the same kinematics and by implication trace each other closely .
@xmath94 the emission line ratios of the gas are uniform , showing no strong variations in [ nii]/h@xmath0 , [ sii]/h@xmath0 or [ oi]/h@xmath0 , as a function of position or h@xmath0 surface brightness , suggesting that its excitation state is independent of the processes which disturb it on the observed kiloparsec scales .
our interpretation is that the h@xmath0 and co emission are produced in a population of molecular clouds warmed by a starburst , which has itself been triggered by the passage of a secondary galaxy through the gas - rich cluster core . in these most luminous h@xmath0 ccgs , this galaxy - wide starburst leads to the saturation of the optical emission line ratios at a uniform level characteristic of massive star formation . in the absence of such triggering
, the cooled gas reservoir will be in a more quiescent state with a much lower level of h@xmath0 and co emission ; optical emission line ratios will be governed by other processes , e.g. excitation through cloud cloud collisions ( as described above ) or mixing layers ( crawford & fabian 1992 ) , to mention just two possibilities .
starting from this low h@xmath0 luminosity quiescent state , adding some star - forming clumps will increase the overall h@xmath0 and co luminosity and give rise to spatial variations in the optical line ratios . at the highest h@xmath0 luminosities
, star - formation will be widespread , with little spatial variation in optical emission line ratios . to test this interpretation ,
future ifu observations should be carried out on ccgs covering the [ oii]@xmath139 and [ oiii]+h@xmath101 emission lines and the sub-4000 continuum in order to probe current star formation , and with near - infrared ifus to study the h@xmath42 and pa@xmath0 emission .
observations should also be performed on lower h@xmath0 luminosity ccgs in order to examine whether spatial variations in optical emission line ratios are indeed larger , due to less widespread star formation . on a longer timescale
, mm - interferometers such as alma will reveal the spatial distribution of the co emission ( and dust ) for a full comparison with the h@xmath0 .
our findings also highlight the need for numerical modelling of these dense co reservoirs , with and without external perturbations .
rjw and ams are supported by pparc .
ace acknowledges the royal society for 10 years of support during which much of the groundwork for the present work was laid .
the data published in this paper have been reduced using vipgi designed by the virmos consortium and developed by inaf milano , in connection with which we thank bianca garilli and marco scodeggio for their hospitality and assistance .
we would also like to thank wht staff for the oasis ifu service observations , in particular chris benn and danny lennon , and richard mcdermid for data reduction advice . the william herschel telescope is operated on the island of la palma by the isaac newton group in the spanish observatorio del roque de los muchachos of the instituto de astrofisica de canarias .
we are grateful to andy fabian for the _ chandra _ image of a1664 .
allen s.w . , et al . , 1992 , mnras , 259 , 67 allen s.w . , 1995 , mnras , 276 , 947 bauer f. , fabian a.c . , sanders j.s . , allen s.w
. , johnstone r.m . , 2005 , mnras , 359 , 1481 bayer - kim c.m . , crawford c.s
. , allen s.w . , edge a.c . , fabian a.c . , 2002 , mnras , 337 , 938 benn c. , talbot g. , bacon r. , 2003 , ing newsletter , 7 , 21 brggen m. , kaiser c.r , 2002 , nature , 418 , 301 cen r. , 2005 , apj , 620 , 191 crawford c.s . , allen s.w . , ebeling h. , edge a.c .
, fabian a.c
. , 1999 , mnras , 306 , 857 crawford c.s .
, fabian a.c . , 1992 , mnras , 259 , 265 edge a.c . , 2001 ,
mnras , 328 , 762 edge a.c . , frayer d.t . , 2003 ,
apj , 594 , l13 edge a.c . ,
wilman r.j . ,
johnstone r.m . , crawford , c.s .
, fabian a.c . , allen s.w . , 2002 , mnras , 337 , 49 egami e. , et al . , 2006 ,
accepted for publication in apj ( astro - ph/0603656 ) fabian a.c . , 1994 , ara&a , 32 , 277 fabian a.c . , mushotzky r.f .
, nulsen p.e.j .
, peterson j.r . , 2001 , mnras , 321 , l20 fabian a.c . ,
allen s.w .
, crawford c.s . , johnstone r.m . ,
morris r.g . , sanders j.s . ,
schmidt r.w . , 2002 , mnras , 332 , l50 fabian a.c . , 2003 , mnras , 344 , l27 fabian a.c . , sanders j.s
, crawford c.s . , conselice c.j . ,
gallagher j.s . ,
wyse r.f.g . , 2003 , mnras , 344 , l48 hu e.m . , cowie l.l .
, kaaret p. , jenkins e.b .
, york d.g . ,
roesler f.l . , 1983 , apj , 275 , 27 jaffe w. , bremer m.n . , baker k. , 2005 , mnras , 360 , 748 jaffe w. , bremer m.n .
, van der werf p.p . , 2001 ,
mnras , 324 , 443 jenner d.c . , 1974 , apj , 191 , 55 lefevre o. , et al .
, 2003 , spie , 4841 , 1670 navarro j.f .
, frenk c.s .
, white s.d.m . , 1997 ,
apj , 490 , 493 peres c.b .
, fabian a.c . ,
edge a.c . ,
allen s.w . , johnstone r.m . ,
white d.a . , 1998 , mnras , 298 , 416 peterson j.r . , kahn s.m . , paerels f.b.s . ,
kaastra j.s . , tamura t. , bleeker j.a.m . , ferrigno c. , jernigan j.g . , 2003 , apj , 590 , 207 rousset a. , 1992 , phd thesis , univ . j. monnet de saint - etienne sanders j.s . , fabian a.c . ,
taylor g.b . , 2005 , mnras , 356 , 1022 salom p. , combes f. , 2003 , a&a , 412 , 657 salom p. , combes f. , 2004 , a&a , 415 , l1 schmidt r.w . , allen s.w . , fabian a.c
. , 2001 , mnras , 327 , 1057 scodeggio a. , et al . , 2005 , pasp , 117 , 1284 shopbell p.l . , bland - hawthorn j. , 1998 , apj , 493 , 129 sparks w.b . , donahue m. , jordn a. , ferrarese l. , ct p. , 2004 ,
apj , 607 , 294 voigt l.m .
, schmidt r.w .
, fabian a.c . , allen s.w .
, johnstone r.m . , 2002 , mnras , 335 , l7 voit g.m . , donahue m. , 1997 , apj , 486 , 242 wilman r.j . , edge a.c . ,
johnstone r.m .
, fabian a.c . , allen s.w . , crawford c.s . , 2002 , mnras , 337 , 63 | we present optical integral field spectroscopy of the h@xmath0-luminous ( @xmath1 ) central cluster galaxies in the cores of the cooling flow clusters a1664 , a1835 , a2204 and zw8193 . from the [ nii]+h@xmath0 complex in these moderate resolution ( 70150 ) spectra we derive two - dimensional views of the distribution and kinematics of the emission line gas , and further diagnostics from the [ sii ] and [ oi ] lines
. the h@xmath0 emission shows a variety of disturbed morphologies , ranging from smooth but distorted to clumpy and filamentary , with velocity gradients and splittings of several hundred on spatial scales of 20 or more . despite the small sample size
, there are some generic features .
the most disturbed h@xmath0 emission appears to be associated with secondary galaxies within 1020 ( projected ) of the central galaxy and close in velocity to the h@xmath0 .
the global h@xmath0 kinematics match those of the co(1 - 0 ) emission in single - dish data .
the [ nii]/h@xmath0 , [ sii]/h@xmath0 and [ oi]/h@xmath0 ratios vary little with position , local h@xmath0 surface brightness or between clusters .
we propose that the h@xmath0 and co emission arise in molecular clouds heated by a starburst , and that the latter has been triggered by interaction with a secondary galaxy .
such co emission is known to trace massive ( @xmath2 ) compact ( @xmath3 ) reservoirs of cool molecular gas , and it is plausible that an infalling galaxy would disturb this gas , distorting the h@xmath0 morphology and initiating widespread star formation .
we also examine the role of cloud - cloud collisions in the undisturbed molecular gas reservoir , and suggest that they might be an important source of excitation for the emission line gas in the cores of lower h@xmath0 luminosity clusters with less intense star formation . #
1to 0pt#1 # 1 # 1 to size # 1 to size = cmr10 scaled2 = cmbx10 scaled2 = cmti10 scaled2 ps .
2 = 6truein galaxies : clusters : individual : a1664 , a1835 , a2204 , zw8193 - cooling flows intergalactic medium ionized gas |
one of the main physics goals of lep2 and future @xmath3 colliders is to directly test the gauge nature of couplings among the electroweak gauge bosons .
the process with largest cross section at lep2 involving these couplings is the @xmath1pair production , @xmath4 , which is sensitive to the trilinear @xmath5 and @xmath6 couplings .
the measurement of these couplings and the sensitivity to possible deviations from the standard model ( sm ) predictions have been extensively studied in the recent years @xcite .
the most general phenomenological parametrization for these couplings @xcite can be achieved by means of an effective lagrangian @xcite that involves operators with dimension higher than four , containing the relevant fields at low energies and respecting the symmetries of the standard model .
the effective lagrangian approach is a model independent way to describe new physics that can occur at an energy scale @xmath7 much larger than the scale where the experiments are performed .
the effective lagrangian depends on the particle content at low energies and since the higgs boson has not yet been found , there are two logical possibilities to describe the new physics effect at low energies . in one of them ,
the higgs boson can be light , being present in the higher dimensional operators , in addition to the electroweak gauge bosons , and the sm symmetries are linearly realized @xcite .
alternatively , the higgs boson can be very heavy and it must be integrated out at low energies . in this case , the relevant fields at low energies are only electroweak gauge bosons and the sm symmetries are realized non - linearly @xcite . here
we focus on a linearly realized @xmath8 invariant effective lagrangian to describe the bosonic sector of the standard model , keeping the fermionic couplings unchanged .
the same effective lagrangian used to describe anomalous trilinear gauge couplings can , in general , lead to anomalous quartic interaction among gauge bosons and also to anomalous couplings of these particles with the higgs field .
all these interactions should also be investigated at lep2 and at the next linear colliders ( nlc ) in order to search for hints about the nature of the new physics described by these higher dimensional operators .
new quartic gauge boson couplings have been studied before in many different processes at future @xmath3 , @xmath9 , @xmath10 , @xmath11 and @xmath12 colliders @xcite . however ,
most of these previous works have focused on the so called genuinely quartic operators , _
i.e. _ operators that give rise only to quartic gauge boson interactions without altering the trilinear couplings @xcite .
since these operators do not appear in a dimension6 linearly realized @xmath13 invariant effective lagrangian @xcite , they will not be considered here .
anomalous higgs boson couplings have also been studied before in higgs and z boson decays @xcite , in @xmath3 @xcite and @xmath14 colliders @xcite .
the process with largest cross section in @xmath3 colliders that also involves quartic couplings , and possibly anomalous higgs couplings , besides the trilinear couplings , is @xmath15 .
therefore , it is the most promising channel to look for possible deviations from the standard model predictions .
this process has been considered by blanger and boudjema @xcite and by leil and stirling @xcite in the context of genuinely quartic operators , where the higgs and trilinear couplings were set to the standard model values and @xmath16 deviations in the total cross section were used to determine the reach of this reaction .
knetter and schildknecht @xcite have considered the effect of a _ single _ higher dimensional operator usually denoted by @xmath17 ( see below ) in the above process , taking into account modifications on both trilinear and quartic couplings .
however , they assumed that the higgs boson mass lies above the energy region to be investigated and therefore they disregarded its contribution .
the purpose of this work is to study the sensitivity to these anomalous couplings of the process @xmath0 at lep2 and the nlc .
we consistently include in our calculations _ all _ new couplings introduced by the effective lagrangian that has become widely adopted to describe new physics beyond the standard model . in particular ,
this process is sensitive to operators related to anomalous higgs boson couplings that do _ not _ affect the self coupling of gauge bosons and hence are not constrained by the lep2 measurements of @xmath18 .
therefore , the process @xmath0 may provide important information about these operators at the nlc .
this paper is organized as follows . in section [ eff :
lag ] , we review the framework of effective lagrangians that we use to parametrize anomalous couplings and explain the methodology used to study the @xmath19 production . in section [ lep ] , we analyze the sensitivity at lep2 based on the total cross section . in section [ nlc ] , we study the improvements arising from going to nlc energies , the effects of having a polarized electron beam , and the impact of being able to measure the @xmath1 boson polarization .
we then concentrate on the analysis of operators which will not be probed by the @xmath4 process , obtaining limits based on the photon energy spectrum .
we present our conclusions in section [ con ] .
in order to write down the most general dimension6 effective lagrangian containing all sm bosonic fields , _ i.e. _ @xmath23 , @xmath24 , @xmath25 , and @xmath26 , we adopt the notation of hagiwara _ et al .
this lagrangian has eleven independent operators in the linear representation that are locally @xmath8 invariant , @xmath27 and @xmath28 even .
we discard the four operators which affect the gauge boson two point functions at tree level and therefore are strongly constrained by lep1 measurements .
we also do not consider the two operators that modify only the higgs boson self
interactions , since they are not relevant for our calculations .
we are then left with five independent operators , and the lagrangian is written as , @xmath29 with each operator @xmath30 defined as , @xmath31 \\ { \cal o}_{ww } & = & \phi^{\dagger } \hat{w}_{\mu \nu } \hat{w}^{\mu \nu } \phi \\ { \cal o}_{bb } & = & \phi^{\dagger } \hat{b}_{\mu \nu } \hat{b}^{\mu \nu } \phi \\ { \cal o}_{w } & = & ( d_{\mu } \phi)^{\dagger } \hat{w}^{\mu \nu } ( d_{\nu } \phi ) \\ { \cal o}_{b } & = & ( d_{\mu } \phi)^{\dagger } \hat{b}^{\mu \nu } ( d_{\nu } \phi ) \ ; , \end{aligned}\ ] ] where @xmath32 is the higgs field doublet , which in the unitary gauge assumes the form , @xmath33 and @xmath34 with @xmath35 and @xmath36 being the field strength tensors of the @xmath37 and @xmath38 gauge fields respectively .
the operator @xmath39 contributes only to anomalous gauge couplings , @xmath40 and @xmath41 contribute only to anomalous higgs couplings , @xmath42 and @xmath43 , whereas @xmath44 and @xmath45 give rise to both types of new couplings .
therefore , the existence of anomalous trilinear gauge couplings could be related to the anomalous quartic gauge couplings and higgs interaction , which are the subject of our investigation .
studies of anomalous trilinear gauge boson couplings from @xmath1pair production will significantly constrain combinations of the parameters @xmath46 , @xmath47 and @xmath48 .
however they are blind " with respect to @xmath49 and @xmath50 .
we chose to study the reaction @xmath0 since it is the process with the largest cross section involving triple , quartic gauge boson couplings and also anomalous higgs gauge boson couplings .
therefore , it is also sensitive to @xmath49 and @xmath50 , offering an excellent possibility for a detailed study of these couplings . the standard model cross section for the process @xmath51 was evaluated in ref .
when we neglect the electron mass , higgs contributions for this reaction do not appear at tree level since the couplings @xmath52 and the @xmath53 are generated only at one loop @xcite . taking into account these contributions , there are 16 feynman diagrams involved in the reaction @xmath0 , which are represented in fig .
[ fig:1 ] ( the crossed diagrams are not shown ) which yields , @xmath54 where we have required that the angle between any two particles is larger than @xmath55 .
the cross section peaks at roughly @xmath56 gev and is typically two orders of magnitude smaller than the two body process @xmath4 , used to constrain anomalous trilinear couplings . in order to compute the contribution from all possible anomalous couplings ,
we have developed a mathematica code to automatically generate the feynman rules for the lagrangian ( [ lagrangian ] ) that were then incorporated in helas
type @xcite subroutines .
these new subroutines were used to extend a madgraph @xcite generated code to include all the anomalous contributions and to numerically evaluate the helicity amplitudes and the squared matrix element . in our calculations , we have taken into account the standard loop higgs contributions besides all the relevant anomalous couplings , which give rise to the 42 contributions shown in fig.[fig:1 ] , [ fig:2 ] , and [ fig:3 ] .
we have checked that our code passed the non trivial test of electromagnetic gauge invariance .
we employed vegas @xcite to perform the monte carlo phase space integration with the appropriate cuts to obtain the differential and total cross sections .
moreover , we have studied the angular variables in order to find optimal cuts to improve the anomalous contribution over the sm signal .
we studied the reaction @xmath0 at lep2 assuming a center of mass energy of @xmath57 and an integrated luminosity of @xmath58 .
we applied a cut in the photon energy ( @xmath59 ) , and we required the angle between any two particles to be larger than @xmath60 . our results for the sensitivity of lep2 to the operators appearing in the effective lagrangian ( [ lagrangian ] ) , from an analyses of the total cross section , are summarized in fig.[fig:4 ] for a fixed value of the higgs boson mass , @xmath61 gev .
we plot the contributions of the 5 different operators separately , assuming that only one operator contributes each time .
we also show the result for an extension of the so called hisz scenario @xcite , where all the coefficient are considered equal , _
i.e. _ @xmath62 , in order to reduce the number of free parameters to only one ( @xmath63 ) .
the standard model cross section and its value with 1 , 2 and 3 @xmath64 deviations are depicted as horizontal lines . the most sensitive contribution comes from @xmath39 , @xmath44 , and @xmath45 .
a 1@xmath64 deviation in the total cross section would be observed for the following ranges of the coefficients of these operators , for @xmath65 tev , @xmath66 whereas for the extended hisz scenario , we have , @xmath67 of course , the operators that also give rise to changes in the triple vector boson couplings can also be constrained at lep2 via the reaction @xmath4 .
a recent analyses of @xmath1boson pair production based on a log likelihood fit to a five fold differential cross section obtained the 1@xmath64 limits @xcite , @xmath68 , @xmath69 , and @xmath70 .
however , one should keep in mind that this reaction is insensitive to @xmath49 and @xmath50 , and therefore the study of the process @xmath71 can provide further information on these operators , as we show in this paper . the contribution of the anomalous couplings involving only the higgs boson , _
i.e. _ @xmath49 and @xmath50 ( see fig.[fig:3 ] ) , is dominated by on mass shell higgs production with the subsequent @xmath72 decay , @xmath73 for large values of the operator coefficients , the total higgs width is dominated by the anomalous decay @xmath74 @xcite , which is also proportional to @xmath49 and @xmath50 . on the other hand , the anomalous width @xmath75 depends only on @xmath49 .
therefore , the contribution from the anomalous coupling @xmath50 is much less sensitive than the contributions from the other operators since @xmath76 becomes almost independent of this coefficient .
fortunately , this is not the case if one is sensitive to small values of the coefficients , as will occur at the nlc study in the next section .
we have investigated various distributions to try to improve the lep2 sensitivity .
the most promising distribution is the angular distribution of the @xmath1 bosons with respect to the beam direction ( see fig .
[ fig:5 ] ) .
we computed the total cross section with the extra cut @xmath77 , as suggested by this distribution , and found an increase in sensitivity from @xmath78 to @xmath79 .
however , due to the small deviations in the shape of the kinematical distributions and small statistics , no further improvement seems to be possible .
the effect of the anomalous operators becomes more evident with the increase of energy , and we are able to put tighter constraints on the coefficients by studying their contribution to different processes at the next linear collider .
we studied the sensitivity of nlc to the process @xmath0 assuming @xmath80 and an integrated luminosity @xmath81 .
we adopted a cut in the photon energy of @xmath82 and required the angle between any two particles to be larger than @xmath55 .
we have analyzed this process for different values of the higgs boson mass . in fig .
[ fig:6 ] , we show the results for the total cross section , for @xmath83 , including the effects of the anomalous operators .
the values of the coefficients @xmath63 s for which a @xmath84 deviation is obtained are shown in table [ tab:1 ] , being typically of the order of @xmath85 tev@xmath86 .
as we could expect , the @xmath1pair production at nlc is able to put a limit that is one order of magnitude better for the coefficients @xmath87 @xcite .
however this latter reaction is _ not _ able to constraint @xmath88 . in an attempt to increase the sensitivity
, we looked at the effects of a @xmath89 polarized electron beam in order to reduce the sm background , mainly the one coming from diagrams of fig.[fig:1]a and [ fig:1]b were just left handed electrons are present .
we have considered both left handed ( lh ) and right handed ( rh ) polarizations , expecting a larger anomalous sensitivity for rh electrons . in fig .
[ fig:7 ] , we show the results for the total cross section , for a @xmath89 right handed ( rh ) polarized electron , for @xmath90 gev . comparing fig .
[ fig:6 ] and fig.[fig:7 ] , we notice that the effect of the anomalous contributions in the total cross section are larger for the polarized case
. however , the small absolute value of the cross section for the polarized case reduces the statistics and leads to no improvement in the established limits , as shown in table [ tab:2 ] .
since we expect the new interactions to involve mainly longitudinally polarized gauge bosons , we studied the sensitivity for different combinations of the polarizations of the @xmath91pair . in fig .
[ fig:8 ] , we show the analogous of fig .
[ fig:6 ] for the @xmath92 case .
again , the effect of the anomalous contributions to the total cross section is increased , but no further improvements are found due to the small statistics . the results for the bounds on the anomalous coefficients for the @xmath93 , @xmath94 , and @xmath95 cases can be seen in table [ tab:3 ] .
these bounds were obtained requiring a 2@xmath64 effect on the total cross section .
it is important to notice that the kinematical distributions of the longitudinally polarized @xmath1 s are quite different from the sm results .
as we could expect , the new physics effects becomes more evident for longitudinal @xmath1 s since the decay @xmath96 is dominated by this state of polarization . in fig.[fig:9 ] , we present the angular distribution of the longitudinal @xmath97 boson with the initial positron and with the final photon , the energy and the transverse momentum distributions .
we can see , for instance , that the @xmath1 energy distribution is very different from the sm prediction .
its characteristic behavior for @xmath98 gev is due to the presence of the higgs boson , which decays into the @xmath1 pair giving rise , at the same time , to a monochromatic photon .
we present in fig .
[ fig:10 ] the percent deviation of the sm prediction in the photon transverse momentum distribution , _
i.e. _ , @xmath99 for the different polarization of the @xmath1 s
. once again the relevance of the @xmath92 case is evident : @xmath100 for @xmath101 gev .
when a cut of @xmath102 gev is implemented , the background is drastically reduced and the ratio of anomalous over sm events per year goes from @xmath103 to @xmath104 , for @xmath105 tev@xmath86 . using the reaction @xmath0 , we are also able to establish bounds on the values of the coefficients @xmath49 and @xmath50 , for which the @xmath1pair process is insensitive , since they only affect the higgs boson couplings . in fig.[fig:11 ] , we present the results of a combined sensitivity analysis in the form of a contour plot for the two free parameter , @xmath50 and @xmath49 , for @xmath90 gev .
these are the most relevant coefficients for the anomalous higgs boson phenomenology and they are not constrained by the @xmath91pair production .
we should keep in mind that the @xmath5 production at lep2 can put a 1@xmath64 bound on @xmath49 ( [ bound : lep ] ) while it is not possible to impose a limit on @xmath50 since the cross section is quite insensitive to this coefficient .
if the higgs boson is found with a mass in the range from 170 to 300 gev , one would have a large sensitivity for the anomalous higgs couplings @xmath49 and @xmath50 in the photon energy distribution of the process @xmath0 .
this increased sensitivity comes about because the existence of a peak in the photon energy spectrum due to the @xmath106body nature of the dominant contribution , _ i.e. _ @xmath107 followed by the subsequent decay @xmath72 ( see fig.[fig:3 ] ) . in fig .
[ fig:12 ] , we illustrate this effect with a typical photon energy distribution , for @xmath108 tev@xmath86 and @xmath90 gev , where the higgs peak appears very clearly in the photon spectrum of the anomalous contribution . in order to analyse the significance of the signal based on the photon energy spectrum
, we took different energy bins of @xmath109 , @xmath110 and @xmath111 gev .
the reason is to roughly mimic the effects of a realistic simulation including the finite energy resolution of the detector and the small spread in the real center of mass energy due to initial state radiation .
we have not considered the experimental efficiency , @xmath112 , for @xmath1 reconstruction .
it can be easily incorporated by multiplying the obtained significances by @xmath113 .
table [ tab:4 ] shows the improvement on the sensitivity compared to the total cross section analysis for the @xmath114 tev@xmath86 and @xmath115 case . in table
[ tab:5 ] , we present our results for the sensitivity on @xmath116 and @xmath117 , assuming @xmath50 = @xmath49 , for the three energy bins above .
we obtained a sensitivity of the order of a tev@xmath86 for @xmath118 , decreasing by a factor of roughly four for @xmath119 , which does not depend in a significant way of the bin size . for larger higgs boson masses ,
the cross section is reduced due to phase space suppression . for smaller higgs boson masses ,
the cross section is reduced since the higgs boson is off mass shell , and in this case it would be better to study processes like @xmath120 or @xmath121 @xcite .
the search for the effect of higher dimensional operators that give rise to anomalous bosonic couplings should be pursued in all possible processes since the results may provide important information on physics beyond the standard model .
we have studied here the production of a @xmath1pair plus a photon in @xmath3 colliders in order to analyse the contributions of anomalous couplings arising from dimension6 operators of a linearly realized @xmath8 invariant effective lagrangian .
we have included _ all _ the anomalous trilinear and quartic gauge couplings , as well as the anomalous higgs couplings with gauge bosons .
we present the limits attainable at lep2 and at nlc , including the standard model irreducible background .
polarization of the electron beam and of the @xmath1pair are found to be insufficient to improve the limits obtained from the total cross section .
we also focused on the operators @xmath40 and @xmath122 , which can not be tested in the @xmath91pair production process .
we showed , in particular , that for higgs boson masses in the range @xmath123@xmath124 , the photon energy spectrum provides a sensitive signature for the anomalous higgs couplings .
typical sensitivities of a few tev@xmath86 at the nlc are obtained for these coefficients , providing complementary information on different higher dimensional operators .
h. ahihara _ et al .
_ , summary of the working subgroup on anomalous gauge boson interactions of the dpf long - range planning study , to be published in _ electroweak symmetry breaking and beyond the standard model _ , edited by t. barklow , s. dawson , h. haber and j. siegrist , hep - ph/9503425 ; + z. ajaltuoni _ et al .
_ , triple gauge boson couplings " , in _ proceedings of the cern workshop on lepii physics _ , edited by g. altarelli _ et al .
_ , cern 96 - 01 , vol . 1 , p. 525
( 1996 ) , hep - ph/9601233 ; + t. barklow _ et al .
_ , summary of the snowmass subgroup on anomalous gauge boson couplings , to appear in the _ proceedings of the 1996 dpf / dpb summer study on new directions in high - energy physics _ ,
june 25 july 12 ( 1996 ) , snowmass , co , usa , hep - ph/9611454 .
s. weinberg , _ physica _ * 96a * , 327 ( 1979 ) ; see also h. georgi , _ weak interactions and modern particle theory _
( benjamim / cummings , menlo park , 1984 ) and j. f. donoghue , e. golowich and b. r. holstein , _ dynamics of the standard model _ ( cambridge university press , 1992 ) .
w. buchmller and d. wyler , nucl .
phys . , 621 ( 1986 ) ; c. j. c. burgess and h. j.schnitzer , nucl .
* b228 * , 454 ( 1983 ) ; c. n. leung , s. t. love and s. rao , z. phys . * 31 * , 433 ( 1986 ) ; a. de rjula , m. b. gavela , p. hernndez and e.mass , nucl .
* b384 * , 3 ( 1992 ) .
s. coleman , j. wess and b. zumino , phys.rev .
* 177 * , 2239 ( 1969 ) ; c. callan , s. coleman , j. wess and b. zumino , phys .
rev . * 177 * , 2247 ( 1969 ) ; t.appelquist and c. bernard , phys . rev .
* d22 * , 200 ( 1980 ) ; a. longhitano , nucl .
* b188 * , 118 ( 1981 ) ; t.appelquist and g .- h .
wu , phys .
rev . * d48 * , 3235 ( 1993 )
. for a review , see s. godfrey , contributed paper to the _ international symposium on vector boson self - interactions _ , los angeles , ca , 13 february 1995 , preprint ocip / c-95 - 5 , hep - ph/9505252 and references therein .
g. blanger and f. boudjema , phys .
lett . , 201 ( 1992 ) , phys .
* b288 * , 210 ( 1992 ) ; o. j. p. boli , m. c. gonzalez garcia and s. f. novaes , nucl .
* b411 * , 381 ( 1994 ) ; o. j. p. boli , m.b .
magro , p. g. mercadante , and s. f. novaes phys .
, 15 ( 1995 ) k. hagiwara and m. l.stong , z. phys . * 62 * , 99 ( 1994 ) ; b. grzadowski and j. wudka , phys .
lett . , 49 ( 1995 ) ; g. j. gounaris , f. m. renard and n.d .
vlachos , nucl .
phys . * b459 * , 51 ( 1996 ) ; w. killian , m. krmer and p. m. zerwas , phys .
* b381 * , 243 ( 1996 ) . | we present sensitivity limits on the coefficients of a dimension6 effective lagrangian that parametrizes the possible effects of new physics beyond the standard model .
our results are based on the study of the process @xmath0 at lep2 and nlc energies . in our calculations , we include _ all _ the new anomalous interactions , involving vector and higgs bosons , and take into account the standard model irreducible background .
we analyse the impact of these new interactions on the total cross section , including the effects of the initial electron and final @xmath1 polarizations .
we then focus on the operators that will not be constrained by the @xmath2 process , obtaining limits based on the photon energy distribution . |
recent developments in semiconductor technology ( e.g. mbe and electron lithography ) opened the possibility to create totally confined electron systems , the so called artificial atoms or quantum dots @xcite .
this is one of the examples of present - day - physics interest in confined finite systems , among which one can mention also the atomic and electronic traps@xcite , and the condensation of confined bosons@xcite .
quantum dots exhibit very interesting properties like the possibility of varying their parameters ( number of electrons , applied fields , dot s geometry , temperature ) over a wide range , the observation of conductance oscillations@xcite and kohn s theorem@xcite , etc .
theoretically , these systems have been studied mainly with the help of numerical methods . however
, analytic approaches have proven to work extremely well providing , at the same time , a qualitative understanding of the quantum dynamics . among these approaches one can mention the semiclassical quantisation@xcite , regularised perturbation theory@xcite , pade approximant techniques@xcite , and the @xmath4-expansion@xcite .
in the present paper , we continue the analytic - qualitative line of research and apply the @xmath4-expansion ( @xmath5 is the absolute value of the angular momentum ) to compute the energy levels of two and three anyons in a model parabolic dot .
the particles interact through a coulombic ( @xmath2 ) repulsive potential .
a magnetic field is applied perpendicularly to the plane of motion .
numerical results for the two - anyon system were obtained by myrheim et al@xcite .
exact analytic solutions at particular values of the coupling constants were found in@xcite .
bohr - sommerfeld quantisation was applied to this system at low magnetic fields@xcite .
we shall see that our method provides extremely accurate solutions for states with angular momentum @xmath6 . a picture for the `` geometry '' of the states ( the spatial distribution of probability )
is obtained also . in the three - anyon problem , however , to our knowledge there are no numerical or approximate calculations .
we start from the hamiltonian of @xmath7 anyons moving in a two dimensional quantum dot in the presence of a perpendicular homogenous magnetic field . in the bosonic gauge
, it is given by the expression @xcite @xmath8 in which the vector potential is taken in the symmetric gauge , @xmath9 @xmath10 is the statistical vector potential , @xmath11 @xmath12 is the unit vector perpendicular to the plane of motion of the anyons , @xmath13 is the anyon s charge , @xmath14 is the anyonic parameter , @xmath15 is the frequency of the parabolic potential needed to confine the anyons in the dot and @xmath16 is the dielectric constant of the medium . a dimensionless hamiltonian is obtained by means of the change of variables @xmath17 @xmath18 where @xmath19 is the cyclotronic frequency , @xmath20 , and @xmath21 is the square root of the ratio between the coulombic and oscillator characteristic energies .
the problem has one exactly solvable limit : a low - density limit , which we call the wigner limit , reached when @xmath22 . in the @xmath23 ( oscillator )
limit , the two - anyon problem is exactly solvable @xcite , whereas the three - anyon system has an infinite family of exact linear states @xcite . in real semiconductors , @xmath24 . introducing jacobi coordinates , @xmath25
the centre - of - mass and relative motions are separated @xmath26 where @xmath27 is the centre of mass hamiltonian and @xmath28 is the hamiltonian of the relative motion .
we introduced the following notation : @xmath29 , and @xmath30 .
we will obtain approximate expressions for the energy eigenvalues of @xmath31 for @xmath32 and @xmath33 as a function of @xmath34 by means of the @xmath0-expansion .
in the two - anyon problem , we have only one jacobi coordinate , @xmath35 , and the hamiltonian of the internal motion reads @xmath36 notice that @xmath37 . after the scaling transformation @xmath38 @xmath39 , we get @xmath40 = \frac { ( \tilde\nu+1)^2}{\rho _ 1 ^ 2}+\frac 14 \rho _ 1 ^ 2 + \frac { \tilde
\beta ^3}{\rho _ 1}\nonumber\\ & - & \frac 1{j^2 } \left ( \frac{\partial ^2}{\partial \rho _ 1 ^ 2}+ \frac 1{\rho _
1}\frac \partial { \partial \rho _ 1 } \right ) , \end{aligned}\ ] ] where @xmath41 .
we have renormalised " the coupling constant , @xmath42 , and the statistical parameter , @xmath43 , in order to take account of the coulomb repulsion and the statistical interaction in a nonperturbative way when taking the formal limit @xmath44 .
we shall look for symmetric eigenfunctions of @xmath45 , i.e. @xmath46 shall be even . in the @xmath47 limit ,
the only term surviving in the hamiltonian is the effective ( classical ) potential energy @xmath48 minimising @xmath49 , we obtain the leading contribution to the energy , @xmath50 , where the radius of the `` bohr orbit '' is obtained from @xmath51 substituting @xmath52 in the r.h.s . of ( 11 ) and expanding , we get @xmath53 where the operator coefficients are given by @xmath54 similar series are written for the wave function and the scaled energy , that is , @xmath55=\sum \limits_{i=0}^ { \infty } \frac{\epsilon_i}{|j|^{i/2}}.\end{aligned}\ ] ] inserting ( 19 ) , ( 20 ) and the series expansion for the hamiltonian into the schdinger equation , we may compute the coefficients @xmath56 and @xmath57 in a systematic way .
up to second order , for example , the system is described in terms of small oscillations around the equilibrium orbit , i.e. the wave function is @xmath58 where @xmath59 is the angle associated to the vector @xmath60 , the @xmath61 are two - dimensional harmonic oscillator radial states with frequency @xmath62 , and the first two coefficients for the energy are @xmath63 afterward , we may take account of anharmonicities .
the results for the next two coefficients are the following @xmath64 -\frac { ( 2n+1)}{2 \omega_1 ^ 5 \rho_{01}^4}\left [ 87n^2 \right . \nonumber\\ & + & 87n+86 + \frac { 12(\tilde\nu+1)^2 \left ( 87n^2 + 87n+86\right ) } { \rho_{01}^4}\nonumber\\ & + & \left.\frac { 4(\tilde\nu+1)^4 \left ( 713n^2 + 713n+709\right ) } { \rho_{01}^8 } \right]+ \frac { 9(2n+1)}{2 \omega_1 ^ 7 \rho_{01}^4 } \nonumber\\ & \times & \left ( 25n^2 + 25n+19 \right ) \left ( 1+\frac { 6(\tilde\nu+1)^2}{\rho_{01}^4}\right ) \nonumber\\ & \times & \left ( 1+\frac { 4(\tilde\nu+1)^2}{\rho_{01}^4}\right)^2 - \frac { 15(2n+1)}{8 \omega_1 ^ 9 \rho_{01}^4}\left
( 47n^2 + 47n+31 \right ) \nonumber\\ & \times & \left ( 1+\frac { 4(\tilde\nu+1)^2}{\rho_{01}^4}\right)^4.\end{aligned}\ ] ] notice that in both wigner ( @xmath65 ) and oscillator ( @xmath66 ) limits the corrections @xmath67 and @xmath68 go to zero .
the expressions found in @xcite are reproduced if we take @xmath69 .
we show in fig .
1 the relative weight of @xmath68 in @xmath70 for the first states with @xmath71 and 6 .
the parameter @xmath72 was fixed to 1/2 ( semions ) .
the relative contribution of @xmath68 is never greater than @xmath73 or @xmath74 for @xmath71 or 6 respectively .
this shows that the @xmath0-series is extremely well behaved .
a comparison with the exact solutions found in @xcite is carried on in fig .
2 , where the relative difference @xmath75 is plotted against @xmath72 . the state with @xmath76 , @xmath77 is shown .
it may be easily verified that @xmath78 , @xmath79 .
the comparison shows that the relative error of our estimate is not greater than @xmath80 .
the internal hamiltonian of the system of three anyons in jacobi coordinates @xmath81 and @xmath82 is written as @xmath83 + \frac 14 \sum\limits_{i=1}^2\rho _
i^2 \nonumber\\ & + & 9\nu ^2\frac{\left ( \rho _ 1 ^ 2 + \rho _ 2 ^ 2\right ) ^2 + 4\left ( \rho _ 1 ^ 2\rho _ 2 ^ 2-\left ( \vec \rho _ 1\cdot \vec \rho _ 2\right ) ^2\right ) } { \rho _ 1 ^ 2\left| \vec \rho _ 1+\sqrt{3}\vec \rho _
2\right| ^2\left| \vec \rho _
1- \sqrt{3}\vec \rho_2\right| ^2}+ \nonumber\\ & & \beta ^3\left [ \frac 1{\rho _ 1}+\frac 2{\left| \vec \rho _ 1 + \sqrt{3}\vec \rho _ 2\right| } + \frac 2{\left| \vec \rho _ 1-\sqrt{3 } \vec \rho _ 2\right|}\right].\end{aligned}\ ] ] @xmath85 = \frac 14\left ( \frac 1{\rho _ 1 ^ 2}+\frac 1{\rho _ 2 ^ 2}\right ) + \frac 14\left ( \rho _ 1 ^ 2+\rho _ 2 ^ 2\right ) \nonumber\\ & + & \tilde \beta ^3\left [ \frac 1{\rho _ 1}+\frac
2{\left| \vec \rho _
1+\sqrt{3}\vec \rho _ 2\right| } + \frac 2{\left| \vec \rho
_ 1-\sqrt{3}\vec \rho _ 2\right| } \right ] \nonumber \\ & + & 4\tilde\nu \frac{\rho _ 1 ^ 2\left ( 2 - 3\cos ^2\theta \right ) + 3\rho _ 2 ^ 2\left ( 2-\cos ^2\theta \right ) } { \left| \vec \rho _ 1+\sqrt{3 } \vec \rho_2\right| ^2\left| \vec \rho _ 1-\sqrt{3}\vec \rho _ 2 \right| ^2}+\frac { \tilde\nu}{\rho _ 1 ^ 2 } \nonumber\\ & + & 9\tilde\nu ^2\frac{\left ( \rho _ 1 ^ 2+\rho_2 ^ 2\right ) ^2 + 4\rho _ 1 ^ 2\rho _
2 ^ 2\sin ^2\theta } { \rho _ 1 ^ 2\left| \vec \rho_1+\sqrt{3}\vec \rho _
2\right| ^2\left| \vec \rho _
1-\sqrt{3}\vec \rho_2\right| ^2 } + \frac 1j \left [ \imath\left ( \frac 1{\rho _ 1 ^ 2 } - \frac 1{\rho _ 2 ^ 2}\right ) \frac \partial { \partial \theta } \right .
\nonumber\\ & + & \tilde\nu \left ( 12\imath \frac{\rho _ 1\rho _ 2\sin 2\theta } { \left| \vec \rho _
1+\sqrt{3}\vec \rho _
2\right| ^2\left| \vec \rho _ 1-\sqrt{3}\vec \rho_2\right| ^2}\left ( \rho _ 1 \frac \partial { \partial \rho _ 2}-\rho _ 2\frac \partial { \partial \rho _
1}\right ) \right.\nonumber\\ & + & \imath \frac 2{\rho _ 1 ^ 2}\frac \partial { \partial \theta } - \left . \left .
8\imath \frac{\rho _ 1 ^ 2\left ( 1 - 3\cos ^2\theta \right)+3\rho _ 2 ^ 2\left ( 1+\cos ^2\theta \right ) } { \left| \vec \rho _ 1+\sqrt{3}\vec \rho _ 2\right| ^2\left| \vec \rho _
1- \sqrt{3}\vec \rho _
2\right| ^2}\frac \partial { \partial \theta } \right)\right ] \nonumber\\ & + & \frac 1{j^2}\left [ -\left ( \frac{\partial ^2 } { \partial \rho _
1 ^ 2}+\frac 1{\rho _ 1 } \frac \partial { \partial \rho _
1}+\frac{\partial ^2 } { \partial \rho_2 ^ 2}+\frac 1{\rho _ 2 } \frac \partial { \partial \rho _ 2}+\right .
\nonumber\\ & & \left .
\left ( \frac 1{\rho_1 ^ 2}+\frac 1{\rho _ 2 ^ 2}\right ) \frac{\partial ^2}{\partial \theta ^2}\right)\right ] , \end{aligned}\ ] ] the minimum of the classical potential entering ( 27 ) is reached in the configuration of an equilateral triangle ( @xmath89 , @xmath90 ) .
we choose , for example , @xmath89 , @xmath91 .
@xmath92 is obtained as the solution of the equation @xmath97 , \\ h_3 & = & -\frac 1{\rho _ { 01}}\left ( \frac{\partial}{\partial y_1}+ \frac{\partial}{\partial y_2}\right ) + \frac 2{\rho _ { 01}^3}\left ( y_1+y_2\right ) \nonumber\\ & + & \frac{3\imath~{\rm sign}(j)}{\rho _ { 01}^4}\left ( y_1 ^ 2-y_2 ^ 2\right ) \frac \partial { \partial z}-\frac 1{\rho _ { 01}^5 } \left(y_1 ^ 3+y_2 ^ 3\right ) \nonumber\\ & - & \frac 1{64\rho _ { 01}}\left ( 1-\frac 1{\rho _ { 01}^4}\right ) \left ( 19y_1 ^ 3 + 3y_1 ^ 2y_2 + 33y_1y_2 ^ 2 + 9y_2 ^ 2 \right . \nonumber\\ & - & \left
. 9\rho _ { 01}^2y_1z^2 + 21\rho_{01}^2y_2z^2\right ) - \frac { 9\tilde \nu^2}{64 \rho_{01}^5}\left ( y_2 ^ 3 - 5y_1 ^ 3 -21y_1 ^ 2 y_2 \right.\nonumber\\ & - & \left .
39 y_1 y_2 ^ 2 + 7\rho_{01}^2 y_1 z^2 - 11\rho_{01}^2 y_2 z^2 \right ) \nonumber\\ & + & \tilde\nu\left [ \frac { 3\imath~{\rm sign}(j)}{2 \rho_{01}^4 } \left \ { \left ( 4y_1 ^ 2 - 2y_2 ^ 2 - 2y_1y_2-\rho_{01}^2z^2 \right ) \frac { \partial}{\partial z } \right.\right.\nonumber\\ & - & \left .
\rho_{01}^2 z\left ( y_1 \frac{\partial } { \partial y_2}+y_2 \frac{\partial}{\partial y_1 }
\right ) + 2\rho_{01}^2 y_2 z \frac{\partial}{\partial y_2 } \right\ } \nonumber\\ & - & \frac{3}{32 \rho_{01}^5 } \left ( 21 y_1 ^ 3 + 15 y_2 ^ 3 + 5 y_1 ^ 2 y_2 + 23 y_1 y_2 ^ 2 \right.\nonumber\\ & + & \left.\left .
\rho_{01}^2 y_1 z^2 - 13 \rho_{01}^2y_2 z^2\right ) \right ] , \\ h_4 & = & \frac 1{\rho _ { 01}^2}\left ( y_1\frac{\partial } { \partial y_1}+y_2\frac{\partial}{\partial y_2}\right ) - \frac 3{\rho _ { 01}^4}\left(y_1 ^ 2+y_2 ^ 2\right ) \frac{\partial ^2 } { \partial z^2 } \nonumber\\ & - & \frac{4\imath~{\rm sign}(j)}{\rho _ { 01}^5}\left ( y_1 ^ 3-y_2 ^ 3\right ) \frac \partial { \partial z}+\frac 5{4\rho _ { 01}^6 } \left ( y_1 ^ 4+y_2 ^ 4\right ) \nonumber\\ & + & \frac 1{256\rho _ { 01}^2}\left ( 1-\frac 1{\rho _ { 01}^4}\right ) \left ( \frac{329}4y_1 ^ 4 - 25y_1 ^ 3y_2 \right.\nonumber\\ & + & \frac{123}2y_1 ^ 2y_2 ^ 2 + 135y_1y_2 ^ 3 + \frac 94y_2 ^ 4-\frac{99}2\rho _ { 01}^2y_1 ^ 2z^2 \nonumber \\ & + & 27\rho _ { 01}^2y_1y_2z^2 + \left . \frac{141}2\rho _ { 01}^2y_2 ^ 2z^2+\frac{41}4\rho _ { 01}^4z^4
\right ) \nonumber\\ & + & \tilde\nu \left [ -\frac { \imath~{\rm sign}(j)}{8 \rho_{01}^5 } \left\ { 6\left(11y_1 ^ 3-y_1 ^ 2y_2 - 7y_1y_2
^ 2 - 3y_2 ^ 3 \right.\right.\right .
\nonumber\\ & - & \left
. 3 \rho_{01}^2y_1z^2-\rho_{01}^2 y_2z^2\right)\frac{\partial}{\partial z } \nonumber\\ & + & \rho_{01}^2 z\left(9y_1 ^ 2 - 3y_2 ^ 2 - 18y_1y_2-\rho_{01}^2z^2 \right)\frac{\partial}{\partial y_1 } \nonumber\\ & - & \left . \rho_{01}^2
z \left(9y_1 ^ 2 - 27y_2 ^ 2 + 6y_1y_2-\rho_{01}^2z^2 \right ) \frac{\partial}{\partial y_2 } \right\ } \nonumber\\ & + & \frac{3}{512 \rho_{01}^6 } \left ( 503y_1 ^ 4 - 28y_1 ^ 3y_2 + 266y_1 ^ 2y_2 ^ 2 + 612y_1y_2 ^ 3 \right .
\nonumber\\ & + & 183y_2 ^ 4 - 58 \rho_{01}^2y_1 ^ 2z^2 + 20 \rho_{01}^2y_1y_2z^2 - 154 \rho_{01}^2y_2 ^
2z^2 \nonumber\\ & - & \left.\left .
41 \rho_{01}^4z^4 \right ) \right ] + \frac{3\tilde\nu ^2}{1024\rho _ { 01}^6}\left ( 549y_1 ^ 4 - 411y_2 ^ 4- 84y_1 ^ 3y_2 \right.\nonumber\\ & + & 2718y_1 ^ 2y_2 ^ 2 + 1836y_1y_2 ^ 3 - 750\rho_{01}^2 y_1 ^ 2z^2 + 1266\rho_{01}^2y_2 ^ 2z^2 \nonumber\\ & + & 60\rho_{01}^2y_1y_2z^2 + \left .
37\rho_{01}^4z^4\right ) . \end{aligned}\ ] ] these are the `` linear '' three - anyon states . we stress that they are obtained as harmonic excitations against the equilateral triangle configuration , and are not necessarily related to a cigar - like shape of the wave function ( @xmath122 ) as described in @xcite .
the set of numbers \{@xmath123 } compatible with the symmetry constraints ( the wave function shall be symmetric ) are obtained upon comparison with harmonic - oscillator wave functions at @xmath69 , @xmath124 .
details may be found in @xcite .
an additional requirement is that the state at @xmath124 should be a linear state .
for example , the lowest linear states are the following : @xmath125 ( the g.s . , starting from @xmath126 at the bosonic end ) , @xmath127 , and @xmath128 ( starting from @xmath129 ) , @xmath130 and @xmath131 ( starting from @xmath132 ) , @xmath133 , @xmath134 , and @xmath135 ( starting from @xmath136 ) , etc .
the lowest state with @xmath137 is @xmath138 , which starts at @xmath139 .
of course , states with small values of @xmath46 can not be described within our method . in what follows
, we restrict the analysis to levels with quantum numbers @xmath140 .
this leaves only the linear anyonic states with @xmath141 , where @xmath142 is an integer @xcite .
the geometry of the state is an equilateral triangle .
it can be seen from ( 29 ) that the side of the triangle increases with @xmath72 when @xmath143 , and decreases when @xmath137 .
thus , the coulomb repulsion is much more stronger for @xmath137 states , and the ordering of levels may dramatically change as @xmath34 is increased . on the other hand , for @xmath65
the side grows like @xmath144 and becomes independent of the anyonic parameter , as one expects .
a strong coupling expansion @xcite shows that the leading contribution to the energy ( potential energy ) is @xmath145 , the next corrections ( quantum fluctuations ) are @xmath146 , the angular momentum and the statistical parameter enter the second order corrections , which are @xmath147 .
we show in fig .
3 the relative weight of @xmath67 in @xmath70 for three semions in states with @xmath162 and @xmath76 .
the numbers are similar to those appearing in the two - anyon problem .
thus , we expect a similar accuracy to this order , i.e. one part in @xmath163 or better . in fig .
4 , the levels with @xmath164 are drawn .
@xmath34 is increased from 0.5 to 8 .
notice that the coulomb effects are stronger for the state with negative @xmath1 , and that the levels become flatter ( as a function of @xmath72 ) as @xmath34 rises . in conclusion ,
the energy levels of two and three anyons in a model parabolic dot were computed by means of the @xmath0- expansion .
the qualitative picture emerging from the @xmath0- expansion is that of a rigid structure ( an orbit in the two - anyon system , an equilateral triangle for three anyons ) against which harmonic and anharmonic oscillations are developed .
the coulomb repulsion is much stronger for negative- @xmath1 states .
comparison with exact particular solutions for two anyons shows excellent agreement .
9 electronic mail : [email protected] electronic mail : [email protected] n. f. johnson , j. phys . : condens .
matter * 7 * , 965 ( 1995 ) .
j. h. jefferson and w. husler , cond - mat/9705012 .
s. becker et al . ,
* 66 * , 4902 ( 1995 ) ; h. j. kluge , nucl .
. res . * b 98 * , 500 ( 1995 ) .
c. c. bradley , c. a. sackett , j. j. tollet and r. g. hulet , phys .
lett . * 75 * , 1687 ( 1995 ) ; k. b. davis , m. o. mewes , m. r. andrews et al , phys .
lett . * 75 * , 3969 ( 1995 ) .
m. a. kastner , rev .
* 64 * , 849 ( 1992 ) .
b. meurer , d. heitmann and k. ploog , phys .
lett . * 68 * , 1371 ( 1992 ) .
p. a. maksym , phys . rev . * b 53 * , 10871 ( 1996 ) .
a. matulis and f. m. peeters , j. phys .
: condens .
matter * 6 * , 7751 ( 1994 ) .
a. gonzalez , i. mikhailov , cond - mat/9706013 ; int .
* b 11 * ( 1997 ) , to appear .
a. gonzalez , j. phys .
: condens .
matter * 9 * , 4643 ( 1997 ) .
a. gonzalez , b. partoens , f. m. peeters , cond - mat/9707258 ; phys .
* 56 * ( 1997 ) , to appear .
a. belov , yu .
lozovik , a. gonzalez , phys . lett . * a 142 * , 389 ( 1989 )
. l. quiroga , a. camacho , a. gonzalez , j. phys : condens .
matter * 7 * , 7517 ( 1995 ) .
a. gonzalez , r. prez and p. fileviez , j. phys .
: condens .
matter * 9 * , 8465 ( 1997 ) .
j. myrheim , e. halvorsen and a. vercin , phys .
* b 278 * , 171 ( 1992 ) .
a. vercin , phys . lett .
* b 260 * , 120 ( 1991 ) . l. salasnich , mod .
* b 11 * , 269 ( 1997 ) . j. myrheim , anyons , in course on geometric phases , ictp , trieste , 1993 . j. m. leinaas and j. myrheim , nuovo cimento * 37 b * , 1 ( 1977 )
. y. s. wu , phys .
53 * , 111 ( 1984 ) ; m. d. jhonson and g. s. canright , phys . rev . *
d 41 * , 6870 ( 1990 ) ; a. p. polychronakos , phys .
* b 264 * , 362 ( 1991 ) .
m. sporre , j. verbaarschot and i. zahed , nucl . phys .
* b 389 * , 645 ( 1993 ) .
a. gonzalez , l. quiroga and b. rodriguez , few - body systems * 21 * , 47 ( 1996 ) . | the energy levels of two and three anyons in a two - dimensional parabolic quantum dot and a perpendicular magnetic field are computed as power series in @xmath0 , where @xmath1 is the angular momentum .
the particles interact repulsively through a coulombic ( @xmath2 ) potential . in the two - anyon problem , the reached accuracy is better than one part in @xmath3 . for three anyons ,
we study the combined effects of anyon statistics and coulomb repulsion in the `` linear '' anyonic states . |
the high temporal variability was one of the first properties to be attributed to the gamma - ray burst ( grb ) prompt emission in the @xmath0-ray energy band ( @xcite ) .
the advent of _ swift _ ( @xcite ) revealed that a highly variable emission characterises also the early time x - ray afterglows in the form of erratic flares .
this established the temporal variability as one of the key features in interpreting the grb phenomena .
grb050502b and the x - ray flash 050406 ( @xcite ; @xcite ; @xcite ) provided the first examples of dramatic flaring activity superimposed to a smooth decay : in particular , grb050502b demonstrated that flares can be considerably energetic , with a 0.3 - 10 kev energy release comparable to the observed prompt fluence in the 15 - 150 kev band . thanks to the rapid re - pointing _ swift _ capability , it was later shown that flares are a common feature of the early x - ray afterglows , being present in the @xmath10 of x - ray light - curves ( @xcite , hereafter c07 ; @xcite , hereafter f07 ) . on the contrary , a convincing optical flare , counterpart to a detected x - ray flare is still lacking , suggesting that the detected optical afterglow contemporaneous to the high - energy flares is dominated by a different emission component ( see e.g. grb060904b , @xcite but see also @xcite where an optical flare was probably detected but , unfortunately , contemporaneous x - ray coverage is lacking ) . based on the temporal and spectral study of a statistical sample of x - ray flares within grbs , c07 and
f07 showed that the flares share common properties and that the flare phenomenology can be described using averaged properties ( see c07 and f07 and references therein ) : * the same grb can show multiple flares ( see e.g. grb051117a which contains a minimum of 11 structures in the first 1 ks of observation ) ; * the underlying continuum is consistent with having the same slope before and after the flare , suggesting that flares constitute a separate component in addition to the observed continuum ; * each flare determines a flux enhancement evaluated at the peak time @xmath11 between @xmath12 and @xmath13 , with a fluence that competes in some cases ( e.g. grb050502b ) with the prompt @xmath0-ray fluence .
the average flare fluence is @xmath14% the 15 - 150 kev prompt fluence ; * flares are sharp structures , with @xmath15 , a fast rise and a slower decay ; * each flare determines a hardening during the rise time and a softening during the decay time ( f07 ) , reminiscent of the prompt emission ( e.g. @xcite ) : the result is a hardness ratio curve that mimics the flare profile ( see e.g. grb051117a , @xcite , their figure 9 ) . in this sense flares are spectrally harder than the underlying continuum ; * the spectrum of a consistent fraction of flares is better fitted by a band ( @xcite ) model , similarly to prompt emission pulses ( see e.g. @xcite ) .
the flare spectral peak energy is likely to be in the soft x - ray range ( a few kev ) .
the spectrum evolves with time as testified by the hardness ratio curve and by accurate spectral modelling . during the decay time
a clear softening is detected ( e.g. @xcite ; @xcite ) ; * there is no correlation between the number of flares and the number of prompt emission pulses ; * the vast majority of flares are concentrated in the first 1 ks after trigger . however ,
late - time flares ( @xmath16 s ) are present as well : flares are not confined to the steep decay phase , but can happen during the plateau and the normal decay phases .
their temporal properties are consistent with those of early flares ( @xcite ) , even if their lower brightness prevents a detailed comparison with the entire set of early time flare properties ( this is especially true from the spectral point of view ) ; * flares happen both in low - z and high - z environments : the record holder grb090423 at z@xmath17 ( @xcite ; @xcite ) shows a prominent flare with standard properties when compared to the sample of x - ray flares of @xcite ( c10 , hereafter ) ;
* flares have been detected both in hard and soft events such as x - ray flashes ( e.g. xrf050406 ) ; * variability has also been detected in the x - ray afterglows of _ short _ grbs ( grb with a prompt emission duration @xmath18 s , @xcite ) .
however , given the lower brightness associated to these events it is still unclear if what is currently identified as a short grb flare emission , quantitatively shares the very same properties as the population of flares detected in _
long _ grbs .
grb050724 ( @xcite ) constitutes a good example of short grb with late - time variability . from the systematic study of 113 flares in the xrt 0.3 - 10 kev energy band , as well as in 4 sub - energy bands
, c10 demonstrated that : * flares are asymmetric with an average asymmetry parameter similar to the prompt emission value ; no flare is found rising slower than decaying ; * the flare width evolves linearly with time @xmath19 .
this is a key point which clearly distinguishes the flares from the prompt emission , where no evolution of the pulse - width has ever been found ( see e.g. @xcite ) ; * the width evolution is the result of the linear evolution of both the rise and the decay times : @xmath20 ; @xmath21 .
* the asymmetry does not evolve with time .
instead the rise over decay time ratio is constant with time , implying that both time scales are stretched of the same factor .
furthermore @xmath22 .
flares are _ self - similar _ in time . * at high energy the flares are sharper with shorter duration : @xmath23 .
pulses share the same property , with a similar dependence on the energy band ( @xcite ; @xcite ) ; * the flare peak luminosity decreases with time .
accounting for the sample variance the best fit relation reads : @xmath24 .
the average flare luminosity declines as a power - law in time @xmath25 ( @xcite ) ; * the isotropic 0.3 - 10 kev flare energy distribution is a log normal peaked at @xmath26 erg ( this can be viewed as the typical flare isotropic energy ) .
the distribution further shows a hint of bimodality . * in multiple - flare grbs ,
the flares follow a softening trend which causes later time flares to be softer and softer : the emitting mechanism keeps track of the previous episodes of emission . starting from these pieces of evidence , c07 , f07 and c10 concluded that in spite of the softness and width evolution , the x - ray flares and the prompt pulses are likely to share a common origin .
however important questions are still to be addressed : do flares follow the entire set of temporal and spectral relations found from the analysis of prompt emission pulses ? in particular : is it possible to define a flare peak - lag ?
do flares follow a lag - luminosity relation ?
is the lag - time linked to other temporal properties ( e.g. the flare duration , asymmetry , etc . ) ?
what can be said about the pulse start conjecture for flares ( @xcite ) ? is it possible to quantify the evolution of the flare temporal properties as a function of the energy band ?
do the rise and decay times evolve differently with energy band ?
is it possible to track and quantify the evolution of the spectral peak energy @xmath27 during the flare emission ? is there any connection between the temporal and spectral properties of the flares ?
what is the position and the track of the flares in the @xmath28 ( spectral peak energy - isotropic luminosity ) plane ?
is there any link between the flares and the underlying x - ray afterglow morphology ?
this set of still open questions constitutes the major motivation for undertaking the present investigation .
the primary goal of this paper is to model the x - ray flare profiles and constrain their evolution with energy to obtain parameters that uniquely qualify the shape and spectrum of the flares and compare those values to the well known signatures of the prompt emission pulses .
since this is often difficult because of low statistics , overlap or non trivial estimate of the underlying continuum , the present work concentrates on very bright and isolated flares for which the underlying continuum does not play a major role neither from the temporal point of view nor from the spectral point of view .
this work is observationally driven : a critical review of theoretical models in the light of the present results is in preparation .
this work is organised as follows : the sample selection and data reduction are described in sect .
[ sec : datared ] . in sect .
[ sec : tempan ] we perform the flare temporal analysis , while to the spectral properties of the sample is dedicated sect .
[ sec : spec ] .
the discussion follows ( sect .
[ sec : disc ] ) .
conclusions are drawn in sect .
[ sec : conc ] .
the phenomenology of the different grbs is presented in the observer frame unless otherwise stated . the 0.3 - 10 kev energy band
is adopted unless specified .
the zero time is assumed to be the trigger time .
the convention @xmath29 is followed , where @xmath30 is the spectral index , related to the photon index @xmath31 by @xmath32 .
all the quoted uncertainties are given at 68% confidence level ( c.l . ) : a warning is added if it is not the case .
standard cosmological quantities have been adopted : @xmath33 , @xmath34 , @xmath35 .
we select the brightest , isolated flares detected by the x - ray telescope ( xrt , @xcite ) on board _ swift _ ( @xcite ) in the time period april 2005 - january 2010 .
the sample comprises 9 flares detected in 8 different grbs : grb050822 , grb060418 , grb060526 , grb060904b , grb060929 , grb070520b , grb070704 and grb090621a ( see table [ tab : bestfit ] ) .
the required isolation allows us to constrain the flares temporal and spectral properties with high accuracy ; their brightness assures the possibility to analyse in detail the evolution of their temporal properties in different energy bands inside the 0.3 - 10 kev of the xrt ; at the same time , this requirement guarantees that the partial ignorance of the underlying continuum properties has a negligible impact on our conclusions , both from the spectral and from the temporal point of view .
xrt data have been processed with the heasoft package v. 6.6.1 and corresponding calibration files : standard filtering and screening criteria have been applied .
_ swift_-xrt is designed to acquire data using different observing modes to minimise the presence of pile - up . when the source is brighter than a few @xmath36 , the data are acquired in windowed timing ( wt ) mode . in this case
we applied the standard pile - up corrections following the prescriptions of @xcite when necessary . for lower count - rates the spacecraft automatically switches to the photon counting ( pc ) mode to follow the fading of the source .
the events are extracted from circular regions centred at the afterglow position with progressively smaller radii to assure the best signal to noise ratio ( sn ) .
when the pc data suffered from pile - up , we extracted the source events in an annulus whose inner radius is derived comparing the observed to the nominal point spread function ( psf , @xcite ; @xcite ) .
the background is estimated from a source - free portion of the sky and then subtracted .
the background subtracted , psf and vignetting corrected light - curves were then re - binned so as to assure a minimum signal - to - noise ( sn ) equals to 4 for wt and pc mode data .
data coming from different orbits of observation were merged to build a unique data point , when necessary .
this procedure was applied to extract grb light - curves in the nominal xrt energy range ( 0.3 - 10 kev ) as well as count - rate light - curves of the same event in 4 different sub - energy bands : 0.3 - 1 kev ; 1 - 2 kev ; 2 - 3 kev and 3 - 10 kev .
-0.0 true cm s , peak time of the flare in the 0.3 - 10 kev energy band as derived from the best fit parameters of table [ tab : bestfit ] .
softest profiles are wider .
, title="fig : " ] -0.0 true cm each flare profile in each energy band is modelled using the @xcite ( hereafter n05 ) profile .
this choice allows a direct comparison between the properties of the x - ray flares and of the wide , long - lag prompt pulses analysed by n05 ( see sec .
[ sec : disc ] ) .
the n05 profile reads : @xmath37\,\,\,\rm{for}\,\,\,t > t_{\rm{s}}\ ] ] where @xmath38 and @xmath39 .
@xmath40 is the pulse onset time .
the parameters are defined following n05 . in particular , the intensity peaks at : @xmath41 the pulse width is measured between the two @xmath42 intensity points and consequently reads : @xmath43 the asymmetry is defined as : @xmath44 where @xmath45 and @xmath46 are the @xmath42 decaying and rising times , respectively : @xmath47\ ] ] we account for the entire covariance matrix during the error propagation procedure .
the continuum underlying the flare emission is estimated from data acquired before and after the flare and is modelled using a power - law or a broken power - law : the parameters related to the continuum have been left free to vary in a first fit and then frozen to their best fit values in a second fit .
the best fit parameters and related quantities are reported in table [ tab : bestfit ] .
figure [ fig:060904bprofile ] shows the best fit profiles in the different xrt band - passes obtained for the flare detected in grb060904b , taken as an example .
a general trend can be seen for the flare to be wider and peak later at lower energies .
a complete summary of the results obtained from the fitting procedure in different energy bands is portrayed in fig .
[ fig:060904b ] : the detected evolution of the different parameters with energy ( panels _ f _ to _ m _ ) will be extensively treated in subsec .
[ subsec : parevolfreq ] and [ subsec : lag ] .
the flare detected in grb060904b is here shown as an example .
figures [ fig : obsbestfitpar1 ] and [ fig : obsbestfitpar2 ] show the correlations or the lack thereof , for the quantities derived from the best fit parameters of table [ tab : bestfit ] superimposed to the results obtained by n05 for a sample of long - lag wide prompt grb pulses and by c10 for a sample of 113 early x - ray flares . in these plots
we see that bright flares obey the set of correlations found for the complete x - ray flare catalogue : the rise time is linearly correlated with the decay time , and both linearly evolve with time .
this gives rise to a flare width that linearly grows with time ( see fig .
[ fig : obsbestfitpar1 ] ) .
figure [ fig : obsbestfitpar2 ] depicts instead the relation between the asymmetry @xmath48 and the temporal parameters describing the flare profiles and demonstrates that it is not possible to distinguish the bright from the main sample on the basis of their asymmetry .
furthermore , the asymmetry seems to be independent of the other parameters ( fig .
[ fig : obsbestfitpar2 ] ) .
-0.0 true cm -0.0 true cm the well known trend of narrower pulses at higher energy is apparent in fig .
[ fig:060904bprofile ] for the flare detected in grb060904b .
this trend is quantified in fig .
[ fig : par_energy ] for the entire sample : the narrowing with energy follows a power - law behaviour where @xmath49 stands for the xrt response matrix , @xmath50 , as found by c10 for the average flare spectral energy distribution . for the hardest band
the best fit spectrum of each flare is used . ] : @xmath51 .
the same is true for the rise and decay times : @xmath52 ; @xmath53 .
the width evolution is the combined result of the evolution of these two time scales and its evolution is consequently associated to a power - law index which is between 0.29 and 0.37 .
notably the rise time shows a steeper dependence on the energy band than the decay time . as a result ,
the asymmetry is slightly dependent on energy : flares are on average more asymmetric at higher energies .
flare profiles peak later at lower energies ( fig .
[ fig:060904b ] , panel _ m _ ) ; on the contrary , it seems that high energy profiles start their rise later ( fig .
[ fig:060904b ] , panel _
l _ ) : while the flare in grb060904b is here portrayed as an example , the same is true for the entire sample as testified by the results of table [ tab : energyevol ] .
figure [ fig : par_energy ] also demonstrates that the energy dependence of the various parameters varies from one flare to another : the average relations describe the general behaviour of the parameters with energy band , but the evolution in a particular flare can be markedly different .
this statement is quantified in table [ tab : energyevol ] , where the best fit power - law index describing the behaviour of the temporal parameters with energy band is listed .
the best fit relations are portrayed in fig .
[ fig:060904b ] , panels _ ( f ) _ through _ ( m ) _ for grb060904b .
how this different evolution rate relates to the spectral properties of the flares is investigated in sec .
[ sec : spec ] .
-0.0 true cm ; _ ( b ) _ @xmath54 ; _ ( c ) _ @xmath55 ; _ ( d ) _ @xmath56.,title="fig : " ] -0.0 true cm .
darker points are associated to spectrally harder flares.,title="fig : " ] [ cols="<,<,<,<,<,<",options="header " , ] through the analysis of 9 bright x - ray flares we proved the existence of a set of correlations which link both the temporal and spectral flare parameters .
the choice of this sample introduces an obvious caveat in our analysis , namely , that we make the explicit assumption that the results we derive from the bright flares hold for all the flares .
however , apart from the brightness , there is no evidence that our sample of bright flares is not representative of the entire population at the 80% c.l . as obtained by a kolmogorov - smirnov ( ks ) test .
we therefore conclude that our results are likely to be extended to the entire population of x - ray flares .
flares have different temporal profiles in different energy bandpasses : in strict analogy to prompt emission pulses , flares are broader and peak later at lower energy .
c10 found @xmath23 .
our sample of bright flares shows @xmath57 .
for the prompt pulses the relation reads : @xmath58 ( @xcite , but see also @xcite , n05 , @xcite and references therein ) .
the width evolution results from the joint evolution of two time scales : the rise and the decay times .
the evolution with energy of @xmath59 and @xmath60 is well described by a power - law , with power - law indices which are found to be very similar ( see table [ tab : energyevol ] ) : however , the very good statistics of our sample allows us to conclude that the _ rise _ time is more sensitive to the energy bandpass than the decay time .
this is evident from fig .
[ fig : trisetdecayslope ] .
since the evolution of the temporal parameters with energy band is strictly linked to the spectral evolution during the flare emission , this would point to some spectral differences between the rise and decay portions of the light - curve .
in particular , the different relative position of the evolving @xmath61 with respect to the fixed energy bands during the @xmath59 and @xmath60 is likely to play a major role : this can be in turn a footprint of two different physical mechanisms dominating the rise and the decay phases .
the first consequence of this finding is that each energy channel is only nearly an exact time - stretched version of the others : flares can be treated as self - similar in energy only at the 0th level of approximation . the second direct consequence is the presence of a slight trend pointing to more asymmetric flares at higher energy ( table [ tab : summaryenergy ] , fig .
[ fig : par_energy ] , panel _ d _ ) .
@xcite defining the asymmetry as @xmath62 found no evolution of this ratio with energy for grb prompt pulses : however , the separate evolution @xmath63 and @xmath64 was not calculated . while the soft energy flare profiles are clearly broader and peak later , it is unclear if the flares start simultaneously at all energies .
the pulse start conjecture was first proposed by @xcite and then tested by @xcite : these authors studied a sample of 199 prompt batse pulses in 75 different grbs using an automatic pulse fitting methodology .
each pulse is modelled by a n05 profile ( the same profile used for flares ) .
we caution that a strong coupling is built in the n05 profile between @xmath40 and @xmath65 : the two parameters are clearly related since both fit the flares before the peak .
this translates in a non - negligible probability of obtaining a good fit with perfect @xmath66 but with non - physical @xmath40 and @xmath65 : any slight but systematic variation @xmath67 would be destroyed .
the other 2 major problems are connected to the structure overlap and low statistics that affect part of the @xcite sample as discussed by the authors . in spite of these limitations
the pulse start conjecture was concluded to hold for prompt gamma - ray pulses with an uncertainty of @xmath68 s. for the x - ray flares the situation is different : the selection of 7 isolated flares with very good statistics led to the conclusion that a slight trend for flare profiles at _ high _ energy to start _ later _ is present .
an example is shown in fig .
[ fig:060904b ] , panel _ l_. table [ tab : bestfit ] reports the start time values for the entire sample of bright flares : the high energy profile @xmath69 is always larger than the low energy @xmath69 : in 5 cases out of 7 the difference between the two values is significant at more than @xmath70 c.l .
the flare emission preferentially builds up at lower energies at the beginning of the grb flares .
figure [ fig : par_energy ] demonstrates that the rate of evolution of the temporal parameters with energy in a single flare can be markedly different from the average sample behaviour ( see the values reported in table [ tab : energyevol ] compared to those of table [ tab : summaryenergy ] ) .
notably , this is directly linked to the observed flare spectral properties as proven by fig .
[ fig : hr_evol_slope ] : the harder the flare , the lower is the detected rate of evolution of the flare profile in different energy bands . since _
hard _ flares also display a _ limited _ spectral evolution in the 0.3 - 10 kev energy range ( insets of fig .
[ fig : hr_rise_decay ] ) , ( likely because their @xmath61 lies outside the xrt energy band during the majority of the emission time ) , the result above implies a strict link between the spectral properties and the evolution of the temporal properties .
the same is true when the flare peak lag is considered : in the case of the prompt emission , the cross - correlation lag between energy bands was shown to anti - correlate with the batse spectral hardness ratio by @xcite ; when _ pulses _ properties are considered instead of time integrated _ burst _ properties , the result is that the pulse spectral hardness anti - correlates with pulse lag and duration , and correlates with pulse intensity ( @xcite ) .
figure [ fig : lag_hr ] demonstrates that x - ray flares share the same property : a trend is evident for flares with larger @xmath71 to have shorter peak lags .
the lag - hardness correlation of this figure , together with fig .
[ fig : hr_rise_decay ] again implies a robust connection between the flares temporal properties and their spectral evolution : a stronger spectral evolution directly translates into a higher lag .
this means that the lag characterising the flares profiles is the direct result of their spectral evolution .
the same conclusion was reached by @xcite for the prompt emission : these authors concluded that the fundamental origin of the observed lag is the evolution of the grb prompt spectra to lower energies .
our results of fig . [ fig : hr_evol_slope ] extend this finding to other flare temporal properties : the rate of evolution with energy of the rise time , decay time and width is inversely correlated to the spectral hardness and directly linked to the spectral evolution .
finally we mention that only soft flares have been found to be highly asymmetric ( fig .
[ fig : par_hr ] , panel _ c _ ) .
similarly @xcite found that soft prompt pulses are on average more asymmetric .
-0.0 true cm the flare peak lag is directly related to the flare duration ( table [ tab : summary ] , fig . [
fig : obslag_corr ] ) and since @xmath72 has been found to vary inversely with the flare peak luminosity ( c10 , but see also the relation reported in table [ tab : summary ] ) , an inverse correlation between peak lag and luminosity is expected . since only three flares in our sample have measured redshift , we selected the flares with the best statistics , measured redshift and well defined peak times in both the hardest and softest xrt energy bands from the sample of c10 .
figure [ fig : laglum ] confirms the expectation above , showing that the x - ray flares define a rather tight lag - luminosity relation : @xmath73 where @xmath74 is the rest frame peak lag calculated as the difference in time between the flare peak times in the 0.3 - 1 kev and the 3 - 10 kev energy bands .
the subscript @xmath75 reminds that the time lag is calculated in the x - ray regime .
we properly account for the sample variance following the method outlined by @xcite .
errors are provided at 90% c.l .
it is not possible to constrain the spectral peak energy and the spectral slopes of the majority of the flares of c10 : for this reason we fitted the spectrum extracted around the peak time with an absorbed spl within xspec and conservatively quote the isotropic peak luminosity as obtained in the observed 0.3 - 10 kev bandpass ( black bullets in fig .
[ fig : laglum ] ) . in two cases , for grb060418 and
grb060904b the band parameters were determined thanks to the bat+xrt coverage ( table [ tab : spectot ] ) : this allows us to plot the 1 - 10000 kev ( rest frame ) isotropic peak luminosity .
however , the best fit relation was derived using their 0.3 - 10 kev values for homogeneity ( the two red bullets in fig .
[ fig : laglum ] consistently lie above the expectation ) . in the cases of the grb060418 and grb090904b flares , the isotropic peak luminosity in the 0.3 - 10
kev observed band underestimates the 1 - 10000 kev rest frame value of a factor 2 - 3 .
the lag - luminosity is one of the key relations which connects the grb _ prompt _ temporal and spectral properties : discovered by @xcite as a time integrated property of each particular burst , the relation was conclusively demonstrated to reflect _
pulses _ rather than _ bursts _ properties by @xcite .
figure [ fig : laglum ] shows a direct comparison between the flare and the prompt properties in the lag - luminosity diagram : this is of particular interest since long and short bursts are known to occupy different regions of the plane ( see e.g. @xcite ) .
flares in long grb are consistent with the long grbs lag - luminosity relation . however
, we should consider that : first , in fig .
[ fig : laglum ] the lag of the prompt data is calculated using the cross correlation function ( ccf ) to the entire bat light - curve and consequently reflects a time integrated property ( see @xcite for any detail ) .
second : for the prompt data the lag is defined as the time difference between light - curve structures in the 50 - 100 kev and 100 - 200 kev channels . for the flares we calculated the _ peak _ lag between x - ray energy bands . in both cases
the lag has been computed between band - passes around the event spectral peak energy .
third : the prompt peak luminosity is calculated in the 1 - 10000 kev rest frame energy band , while for the flares the peak luminosity is calculated from the 0.3 - 10 kev ( observed ) bandpass ( which is expected to be a factor of 2 - 3 below the 1 - 10000 kev value ) . with these caveats in mind , it is remarkable that the best fit slope of the flare lag - luminosity is consistent with the @xcite results based on bat data ( see table [ tab : laglum ] ) : @xmath76 ( red dot dashed line in fig .
[ fig : laglum ] ) .
it is not surprising that this relation overestimates the flare luminosity which is computed in a narrower energy window by a factor @xmath77 .
only a marginal consistency can be quoted with @xcite who reported an index @xmath78 .
our findings are instead fully consistent with the @xcite law ( both in normalisation and index ) , with a power - law index 1.14 ( blue dashed line in fig . [
fig : laglum ] ) : these authors reported lags between batse energy bands 100 - 300 kev and 25 - 50 kev .
the same is true if we consider the lag - luminosity power - law index by @xcite who reported a value of 1.01 .
the prompt lag - luminosity relations are summarised in table [ tab : laglum ] .
this result strongly suggests a common physical mechanism producing both the grb prompt emission and the x - ray flare emission hundreds of seconds later .
the lag - luminosity relation has been proven to be a fundamental law extending @xmath79 decades in energy and @xmath79 decades in time . at the 0th level of approximation
, the flare phenomenology summarised in table [ tab : summary ] reduces to a set of 4 independent relations ( namely eq . [ eq : trtd ] , [ eq : wtpeak ] , [ eq : laglum ] and [ eq : softening ] ) .
this consistently limits the available parameters space for any theoretical model aiming at explaining the flare phenomena .
the first fundamental relation expresses the concept that flares profiles tend to be self - similar both in energy and in time ( c10 ) : @xmath80 which implies : @xmath81 : the n05 profile contains a degree of freedom which is not required by the data , since the rising and decaying time scales are not independent from one another ; @xmath82 , @xmath83 and @xmath84 . the pulse peak lag between two energy bands centered on @xmath85 and @xmath86 reads : @xmath87 since @xmath88 at the first level of approximation .
@xmath89 directly inherits the energy dependence of @xmath46 and @xmath45 ( table [ tab : summaryenergy ] ) : @xmath90 where @xmath91 .
this implies @xmath92 where @xmath93 is a factor of proportionality which depends on @xmath85 and @xmath86 . a lag - width correlation is consequently expected since @xmath94 ( or alternatively a @xmath95 or @xmath96 relations ) .
the width - peak time linear correlation discussed by c10 , introduces the concept of temporal @xmath97 , which is the key to interpret the grb x - ray flare phenomenology : @xmath98 which automatically implies @xmath99 and @xmath100 ; considering the lag - width correlation of the previous paragraph , eq . [ eq : wtpeak ] translates into a lag - tpeak correlation .
the third fundamental relation links the flare peak - lags to their peak - luminosities , in strict analogy to the prompt emission : @xmath101 consequently , @xmath46 , @xmath45 and @xmath72 will be linked to the flare peak luminosity . in turn , the peak luminosity is expected to evolve with time to lower values , as found : @xmath102 although with a large scatter .
the fourth relation again describes the temporal _ evolution _ of the flare spectral properties : @xmath103 flares become softer and softer as time proceeds . as a result , an inverse @xmath104 , @xmath105
, @xmath106 and @xmath107 relation is automatically built .
this completes the set of 28 correlations ( or lack thereof ) of table [ tab : summary ] .
an important implication is that at the 0th level of approximation , the physics underlying the flare emission determines that out of 3 time scales describing the flare temporal profile ( namely @xmath59 , @xmath60 and @xmath108 ) , only one degree of freedom survives . as a second level approximation
, one should consider that @xmath109 ; @xmath110 which likely reflects the relative position of the spectral peak energy with respect to @xmath85 and @xmath86 ; @xmath111 since the rise and decay times evolution with energy band slightly differs from one another(fig .
[ fig : trisetdecayslope ] ) . any theoretical model aiming at explaining the flare emission
is asked to be consistent with these set of findings , together with the flare spectral properties discussed in the following paragraph .
a critical revision of the existing models is in preparation .
for the two flares with the highest statistics , we find that the flare spectrum is well described by an evolving band function . in both cases
the spectral peak energy evolves with time to lower values following an exponential decay which tracks the decay of the flare flux .
after @xmath108 the n05 profile is progressively dominated by the @xmath113 factor , with the relative strength of the @xmath114 dropping quickly .
remarkably , in both flares , the temporal evolution of the spectral peak energy and the flux seem to share the same e - folding time : for the flare detected in grb060418 we have @xmath115 s with @xmath116 s ; for the grb060904b flare we find @xmath117 s in good agreement with @xmath118 s. the evolution of @xmath112 to lower values during the pulse decay time is one of the signatures of the prompt emission as demonstrated by a number of studies ( see e.g. @xcite and references therein for a recent study on single prompt pulses ) .
the spectral peak energy of prompt pulses is found to evolve following a power - law @xmath119 with an evolutionary slope @xmath120 ( see @xcite and references therein ) .
the observed evolution of the flare @xmath112 is instead much faster and inconsistent with the @xmath121 behaviour after rescaling the time to the beginning of the flare emission .
taking the grb060904b flare which starts at @xmath122 s as an example , @xmath123 kev would imply @xmath124 kev if the @xmath121 behaviour is assumed : the observations instead constrain the @xmath125 to be well below the xrt bandpass , so that @xmath126 kev .
while this could be the result of a wrong choice of the power - law zero time ( the relevant time could be the ejection time instead of @xmath127 , see @xcite for details ) , further investigations are required to establish if the faster @xmath112 evolution is a peculiar feature of the flare emission when compared to the prompt phenomenology .
the spectral analysis of subsec . [ subsec : specmodeling ] allows us to constrain the properties of two x - ray flares in the @xmath130 and @xmath131 planes for the first time .
figure [ fig : epeisoliso ] shows the result : when compared to the spectral properties of the prompt emission of 83 grbs with measured redshift and well constrained spectra ( @xcite ; @xcite ) , the two flares show higher @xmath2 and @xmath3 than expected starting from their rest frame spectral peak energy @xmath1 .
if the flares are part of the prompt emission , then they are not expected to share the time integrated @xmath132 best fit normalisation : however , @xcite demonstrated that individual sequences of the same burst do follow the @xmath128 relation with a _ higher _ normalisation ( reflecting the fact that the total energy budget is distributed over the sequence , see @xcite , their fig .
the opposite is observed for the two flares .
the @xmath131 plane does not suffer from this effect , and time integrated properties can be here directly compared to their time resolved counterparts . in this plane
the two flares are barely inside the @xmath133 best fit area : however , the peak luminosity is poorly sampled below @xmath134 and a deviation to lower spectral peak energies of the main sample can not be excluded since the best fit relation is mainly established by high luminosity data points . in this case
the x - ray flares would belong to the same @xmath131 as the prompt emission .
alternatively , and equally interesting , the flares do not follow the @xmath135 relation of the prompt data .
this possibility is suggested by the fact that the two flares are outliers of the @xmath136 relation even when the effect of the time resolved spectral analysis is taken into account . whether this depends on the chosen integration energy band ( 1 - 10@xmath137 kev ) is currently under investigation .
for grb060418 the spectral properties during the prompt emission were measured ( see @xcite ; @xcite ) : @xmath138 kev ; @xmath139 erg ; @xmath140 erg .
the flare isotropic input comprises @xmath77% of the isotropic gamma ray prompt energy while the flare isotropic peak luminosity is @xmath141% the @xmath142 .
the contribution of the flare emission is therefore well within the uncertainties estimated for the prompt parameters : this results helps to understand how much the unaccounted flare emission contributes to the scatter affecting the @xmath4 relation ( @xcite ) .
the time resolved analysis of the two flares with redshift and bat detection reveals that the rest frame spectral peak energy @xmath1 correlates with the isotropic luminosity @xmath6 within single flares .
the same has been demonstrated to be true for the prompt pulses : the _ time integrated _ correlations are the result of the existence of similar _ time resolved _ correlations of the same parameters .
individual grb prompt pulses are consistent with the @xmath5 as proved by @xcite , and @xcite .
our work proves that 2 x - ray flares share this property and that flares in general are likely to be consistent with this behaviour .
the best fit slopes of the two relations of fig .
[ fig : epeisolisotime ] are however steeper than the @xmath144 slope found for the prompt emission ( see e.g. @xcite ) : whether this is something peculiar of the flare emission or not , needs to be understood with a larger sample . at the moment we note that the best fit time resolved @xmath5 slope within single prompt pulses
have been found to be different from the 0.5 value in some cases : see e.g. grb090323 and grb090328 in @xcite their fig .
3 , where the @xmath5 track seems to be steeper than @xmath145 . intriguingly , the spectrum extracted at the onset of the flare in grb060904b is a clear outlier , being characterised by an @xmath1 higher than expected : a similar result was found by @xcite for the prompt emission .
these authors performed a time resolved spectral analysis of the grb061007 prompt emission and concluded that the _ initial _ rising phase of each pulse is an outlier of the @xmath5 relation , with @xmath1 around twice the value expected from the spectral correlation .
however , no difference between the rise and decay portions of the prompt pulses has been reported by @xcite from the analysis of 2 _ fermi _ grbs .
the presence of these spectral correlations during the flare emission tightly links the x - ray flare emission to the prompt phase ; as a by - product , these results strengthen the interpretation of the spectral energy correlations as manifestation of the physics of the grbs ( see @xcite and references therein for a detailed discussion of this topic ) .
-0.0 true cm the entire set of findings presented in this work establishes a strong link between early x - ray flares and prompt pulses .
flares are consistent to be independent episodes of emission superimposed to the contemporary steep decay or afterglow components : while the first has been proven to be generated by the tail of the previous pulses ( @xcite ; see e.g. @xcite ) and is therefore related to the physical mechanism which gives rise to the prompt emission , the latter is likely to be a completely independent component ( e.g. @xcite ) .
x - ray afterglows with no breaks are likely to be dominated by the afterglow component from the beginning of the xrt observations ( see @xcite for a dedicated study ) .
we therefore expect the spl x - ray afterglows to show a limited flaring component if flares really constitute an independent component ( flares would be perhaps present but hidden by the contemporaneous afterglow emission ) .
figure [ fig : flares_lctype ] confirms this expectation : the no - break class is clearly under - represented in the flare sample , while the vast majority of the detected early time flares reside in one - break or canonical x - ray afterglows . the probability that the two samples are drawn from the same population
is evaluated to be as low as @xmath146% using a ks test .
if this is due to the brightness of the afterglow emission , than the conclusion is that flares have an independent origin : in the standard scenario this means that flares are not produced by the external shock .
another equally interesting possibility is that flaring emission is quenched ( instead of being hidden ) in spl - afterglow grbs .
distinguishing between these two possibilities is beyond the scope of this work and will be the subject of a forthcoming investigation .
-0.0 true cm ) . we stress that this is an _ average _ behaviour .
the thick solid arrow underlines the dependence of the flare properties on the _ time _ , in strong opposition to prompt pulses .
, title="fig : " ] the grb x - ray flare phenomenology shows a list of properties strictly analogous to the gamma - ray prompt emission ( see @xcite for a recent study on individual prompt pulses ) .
however , it differs in one crucial aspect : flares _ evolve _ with _ _
time__. flares evolve with time to lower peak intensities , larger widths , larger lags and softer emission .
[ fig : flareparadigma ] illustrates the flare paradigm . as the prompt emission pulses ,
flares have correlated properties : short - lag flares have shorter duration , are more luminous and harder than long - lag flares . in particular the following properties add to the lists of sec .
[ sec : introduction ] : * flares define a lag - luminosity relation : + @xmath147 .
the best fit slope is remarkably consistent with the prompt findings ( see @xcite and references therein for a recent study ) ; * the lag is found to correlate with the flares width while it is inversely correlated to the flare spectral hardness ; * the flares temporal profiles in different band - passes are only a nearly exact time stretched version of one another : the rise and decay times evolve following slightly different power - laws in energy : @xmath148 , @xmath149 with @xmath150 .
the result is that flares are on average more asymmetric at high energy ; * the rate of evolution of a flare profile in different band - passes anti - correlates with its spectral hardness : the harder the flare the lower is the rate of evolution from one energy band to the other ; * the flare spectral peak energy @xmath151 evolves to lower values following an exponential decay which tracks the decay of the flare flux .
the detected evolution is faster and inconsistent with the @xmath152 behaviour even when the zero time is re - set to the beginning of the flare emission ; * the two flares with best statistics show higher than expected @xmath2 and @xmath3 values when compared to the @xmath4 and @xmath5 prompt correlations ; * the rest frame spectral peak energy @xmath1 correlates with the isotropic luminosity @xmath6 within single flares , giving rise to a time resolved @xmath5 correlation ; * the flare emission preferentially builds up at lower energies : flares do not seem to be consistent with the pulse start conjecture ; * among the different types of x - ray afterglow light - curves , the simple power - law afterglows are under - represented in the flare sample .
flares are preferentially detected superimposed to one - break or canonical light - curves .
the strict analogy between the prompt pulses and flares phenomenology strongly suggests a common origin of the two phenomena .
rm is grateful to l. nava for providing the comparison data portrayed in fig .
[ fig : epeisoliso ] . this work is supported by asi grant swift i/011/07/0 , by the ministry of university and research of italy ( prin miur 2007tnyzxl ) , by mae and by the university of milano bicocca , italy .
99 , l. , guidorzi , c. , frontera , f. , della valle , m. , finelli , f. , landi , r. & montanari , e. , 2008 , mnras , 391,577 , d. , matteson , j. , ford , l. et al . , 1993 , apj , 413 , 281 , s. d. , barbier , l. m. , cummings , j. r. et al . , 2005 , ssrv , 120 , 143 , s. d. , chincarini , g. , burrows , d. n , et al . , 2005b , nature , 438 , 994 , l. , frontera , f. , guidorzi , c. , montanari , e. , vetere , l. , & soffitta , p. , 2007 ,
a&a , 465 , 765 , d. n. , hill , j. e. , nousek , j. a. et al . , 2005 ,
ssrv , 120 , 165 , d. n. , romano , p. , falcone , a. , 2005b , science , 309 , 1833 , g. , moretti , a. , romano , p. , et al .
, apj , 671 , 1903 ( c07 ) , g. , mao , j. , margutti , r. , et al .
, 2010 , mnras submitted ( c10 ) , p. a , starling , r. l. c. , o brien , p. t. , godet , o. , van der horst , a. j. & wijers , r. a. m. j. , 2008 , a&a , 487 , 533 , p. a , page , k. l. , cummings , j. r. , barthelmy , s. d. & marshall f. e. , 2009 , gcn rep . 223.1 , g , arxiv 0511182 , p. a. , beardmore , a. p. , page , k. l. et al . , 2009 , mnras , 397 , 1177 , e. e. , zand , j. j. m. , norris , j. p. , bonnell , j. t. , & nemiroff , r. j. , 1995 , apjl , 448 , 101l , a. d. , burrows , d. n. , lazzati , d. , et al . , 2006 ,
apj , 641 , 1010 , a. d. , morris , d. , racusin , j. , et al . , 2007 , apj , 671 , 1921 , ( f07 ) , l. a. , band , d. l. , matteson , j. l. , et al , 1995 , apj , 439 , 307 , n.,chincarini , g. , giommi , p. et al . , 2004 ,
apj , 611 , 1005 , n. , norris , j. p. , barthelmy , s. d. , et al . , 2006 , nature , 444 , 1044 , g , nava , l. , ghisellini , g. , 2009 , arxiv:0908.2807 , m. r. , page , k. l. , godet , o. et al .
, 2007 , a&a , 468 , 103 , o. , page , k. l. , osborne , j. , et al . , 2007 , a&a , 471 , 385 , j. , kruhler , t. , mcbreen , s. , et al . , 2009 ,
apj , 693 , 1912 , j. , giblin , t. w. , norris , j. p. , fragile , p. c. & bonnell , j. t. , 2008 , apjl , 677 , 81 , j. , nemiroff , r. j. , 2009 , apj , 705 , 372 , p. m. w. , burton , w. b. , hartmann , d. et al . , 2005 , a&a , 440 , 775 , y. , preece , r. d. , briggs , m. s. , paciesas , w. s. , meegan , c. a. & band , d. , 2006 , apjs , 166 , 298 , r. w. , strong , i. b. & olson , r. a. , 1973 , apj , 182 , 85l , a. , gendre , b. , stratta , g. , et al . , 2008 , a&a , 483 , 847 , d. & liang , e. , 2003 , apj , 594 , 385 , c. , meegan , c. a. , fishman , g. j. , et al . , 1993 ,
apjl , 413 , l101 , h. a. , granot , j. , marshall , f. e. et al . , 2007 , apj , 665 , 554 , h. a. , yamaoka , k. , sugita , s. , et al . , 2009 , apj , 704 , 1405 , p. & panaitescu , a. , 2000 , apj , 541l , 51 , d. , perna , r. , begelman , m. c. , 2008 , mnras , 388 , l15 , e. w. , lu , h. j. , hou , s.j . ,
zhang , b. b. , zhang , b. , 2009 , apj , 707 , 328 , r. , moretti , a. , pasotti , f. et al . , 2008 ,
a&a , 480 , 677 , r. , genet , f. , granot , j. et al . , 2010 , mnras , 402
, 46 moretti et al .
, 2005 , proceeding of spie , vol . 5898 , pp .
360 - 368 . ,
a. , margutti , r. , pasotti , f. , et al . , 2008 , a&a , 478 , 409 , l. , ghirlanda , g. , ghisellini , g. & firmani , c. , 2008 , mnras , 391 , 639 , r. j. , 2000 , apj , 544 , 805 , j. p. , nemiroff , r. j. , bonnell , j. t. , et al .
, 1996 , apj , 459 , 393 , j. p. , marani , g. f. & bonnell , j. t. , 2000 , apj , 534 , 248 , j. p. , bonnell , j. t. , kazanas , d. et al . , 2005 ,
apj , 627 , 324 , n05 , m. , ioka , k. , yamaoka , k , et al . , 2009 , pasj , 61 , 201 , z. y. , ma , l. , zhao , x. h. , yin , y. , fang , l. m. , bao , y. y. , 2009 , apj , 698 , 417 , e. , fenimore , e. e. , 2000 , apj , 539 , 712 , p. , campana , s. , chincarini , g. et al . ,
2006a , a&a , 456 , 917 , p. , moretti , a. , banat , p. l. , et al .
, 2006b , a&a , 450 , 59 , r. , della valle , m. , campana , s. , et al .
, 2009 , nature , 461,1258 , e. b. , 2007 , apj , 660 , 16 , n. r. , fox , d. b. , levan , a. j. , et al .
, 2009 , nature , 461 , 1254 , t. n. , stamatikos , m. , dhuga , k. s. , et al . , apj accepted , arxiv 0908.2370 , s. , goad , m. r. , beardmore , a. p. , et al . , 2006 , apj , 638 , 920 , r. , genet , f. , granot , j. & obrien , p. t. , 2010 , mnras , accepted , d. , murakami , t. , nakamura , t. , yamazaki , r. , inoue , a. k. & ioka , k. , 2004 , apj , 609 , 935
[ lastpage ] | the temporal and spectral analysis of 9 bright x - ray flares out of a sample of 113 flares observed by _ swift _ reveals that the flare phenomenology is strictly analogous to the prompt @xmath0-ray emission : high energy flare profiles rise faster , decay faster and peak before the low energy emission .
however , flares and prompt pulses differ in one crucial aspect : flares evolve with time .
as time proceeds flares become wider , with larger peak lag , lower luminosities and softer emission .
the flare spectral peak energy @xmath1 evolves to lower values following an exponential decay which tracks the decay of the flare flux .
the two flares with best statistics show higher than expected isotropic energy @xmath2 and peak luminosity @xmath3 when compared to the @xmath4 and @xmath5 prompt correlations .
@xmath1 is found to correlate with @xmath6 within single flares , giving rise to a time resolved @xmath7 .
like prompt pulses , flares define a lag - luminosity relation : @xmath8 .
the lag - luminosity is proven to be a fundamental law extending @xmath95 decades in time and @xmath95 in energy .
moreover , this is direct evidence that grb x - ray flares and prompt gamma - ray pulses are produced by the same mechanism .
finally we establish a flare- afterglow morphology connection : flares are preferentially detected superimposed to one - break or canonical x - ray afterglows .
[ firstpage ] gamma - ray : bursts radiation mechanism : non - thermal x - rays |
at high temperature @xmath2 and low momenta @xmath3 the effective degrees of freedom in any field theory are collective modes , bosonic plasmons " @xcite and fermionic `` plasminos '' @xcite .
even if the fundamental field theoretical degrees of freedom were massless , the collective modes possess a finite thermal " mass which is generated dynamically by the interactions among the fundamental degrees of freedom .
the thermal masses regulate ( some of ) the severe infrared divergences which the massless theory would otherwise develop at finite temperature due to the singularity of the bose distribution at zero momentum @xcite .
the strong rise of the bose distribution at small momenta causes , for example , the transport properties ( viscosity , heat conduction ) of a hot plasma of massless fields to be dominated by the interaction between low - momentum collective modes @xcite .
the determination of their dispersion relation and their collisional width is therefore a necessary prerequisite for any microscopic calculation of the transport coefficients , for example using the kubo formulae @xcite .
it is known that a consistent determination of the plasmon dispersion relation and width in a massless field theory requires the resummation of certain loop contributions to the propagator , the so - called hard thermal loops " ( htl ) @xcite . in massless gauge theories , this resummation becomes non - trivial beyond leading order , due to the complicated analytical structure of the htls ( i.e. their momentum dependence and logarithmic branch cuts from landau damping @xcite ) . in massless
scalar @xmath4 theory , the resummation is much easier @xcite , since the leading htl ( the tadpole diagram ) is just a momentum - independent real constant , and it can therefore be relatively easily be carried beyond leading order @xcite .
still , when we started this work , no calculation of the collisional _ width _ of finite momentum plasmons existed . since the latter is needed , however , for the calculation of transport coefficients from the kubo formulae ( previous calculations @xcite used , for lack of better knowledge but without justification , the zero momentum limit of the plasmon width ) , we present in this paper a calculation of the resummed scalar field propagator in @xmath0 theory at the 2-loop level , including real _ and _ imaginary parts in 4-dimensional momentum space . while this work was in progress
, jeon @xcite presented a calculation of the plasmon width at zero and finite momentum on the plasmon mass shell .
our results agree with his and extend them into the off - shell domain .
this paper is organized as follows . in section [ s2 ]
we will give a brief review of results from ref .
@xcite on resummation in hot @xmath4 theory which we will need later . in section [ s3 ]
we present the calculation of the imaginary part of the 2-loop self - energy and investigate the analytical structure of the full propagator at two - loop order .
the properties of plasmons at rest and at finite momentum are studied analytically and numerically in sections [ s3a ] and [ s3b ] , respectively .
our conclusions are summarized in section [ s4 ] .
appendix [ appa ] contains some technical steps of the calculations in section [ s3 ] .
for a hot scalar field , we consider the following lagrangian with @xmath5 : @xmath6 at the tree level it describes massless scalar fields with a quartic self interaction .
the induced thermal mass resulting from the single htl ( tadpole ) is of order @xmath1 and reads @xcite @xmath7 the effects from this thermal mass can be resummed by defining an effective lagrangian through @xmath8 and treating the last term as an additional interaction .
this effective lagrangian defines an effective propagator with thermal mass @xmath9 @xmath10 where in the imaginary time formalism @xmath11 , @xmath12 , @xmath13 .
it represents the infinite sum of iterated htls .
since @xmath9 leads to a qualitative modification of the resummed effective propagator for soft momenta , @xmath14 , its inclusion via resummation is necessary to avoid infrared divergences at finite temperature @xcite .
vertex corrections from htls are down by a factor @xmath1 relative to the bare vertex @xmath15 and do nt qualitatively change the latter . in a perturbative expansion of @xmath4 theory we can thus keep using the bare vertex @xcite .
one can now recalculate the 1-loop self - energy using this resummed effective propagator .
this amounts to computing the contribution of all cactus diagrams " or superdaisies " in the original lagrangian @xcite each of which is infrared divergent . using dimensional regularization in the ms renormalization scheme , the diagram in fig .
[ f1]a gives @xmath16 \right ) + { \cal o}(g^5 \ln g)\ , , \ ] ] where @xmath17 is the scale parameter in dimensional regularization and @xmath18 is euler s constant .
the resummation of an infinite series of infrared diagrams reflects itself through a @xmath19 correction to the leading term @xmath20 ; it is non - analytic in the coupling constant @xmath21 and for @xmath22 dominates any finite genuinely perturbative corrections of order @xmath23 . in this sense resummation
is essential . including also the contribution from the new 2-point interaction in ( [ lag ] ) , fig .
[ f1]b , the full 1-loop self - energy can be written as @xmath24 .
it is purely real .
thus , to order @xmath19 , the full propagator has only two poles at @xmath25 where @xmath26 is the 1-loop resummed plasmon mass : @xmath27 following ref .
@xcite , we can now improve the resummation procedure by including into the effective lagrangian all 1-loop effects which are independent of the regularization scheme , i.e. by writing @xmath28 and expanding into effective scalar propagators with an effective plasmon mass @xmath26 .
then all factors @xmath9 in eq .
( [ sigma1p ] ) are replaced by @xmath26 , and from the full 1-loop self energy @xmath29 all terms up to order @xmath19 cancel , leaving a leading contribution of order @xmath23 which is real . at 2-loop order ,
the feynman diagrams shown in fig .
[ f2 ] contribute .
their leading contributions are also of order @xmath23 .
diagram [ f2]c arises from the new 2-point interaction of the lagrangian ( [ lag1 ] ) and contributes at the same order . the contribution @xmath30 of the diagrams [ f2]a and [ f2]c is easily evaluated .
it is purely real . as a consequence of resummation ,
all leading @xmath23 terms cancel , and the result is found to be @xmath31 a convenient way to calculate diagram [ f2]b is to use the saclay method @xcite . its contribution can be expressed as @xmath32 \ , \big\ { s(e_k , e_q , e_r)\lbrack ( 1+f_k)(1+f_q)(1+f_r)-f_kf_qf_r\rbrack \nonumber\\ & & \phantom{\int d[k , q ] } + s(e_k , e_q ,- e_r)\lbrack(1+f_k)(1+f_q)f_r - f_kf_q(1+f_r)\rbrack \nonumber\\ & & \phantom{\int d[k , q ] } + s(e_k ,- e_q , e_r)\lbrack(1+f_k)f_q(1+f_r)-f_k(1+f_q)f_r\rbrack \nonumber\\ & & \phantom{\int d[k , q ] } + s(-e_k , e_q , e_r)\lbrack f_k ( 1+f_q ) ( 1+f_r ) -(1+f_k ) f_q f_r \rbrack \big\ } \ , , \end{aligned}\ ] ] with the same notation as in ref .
@xcite , namely @xmath33 & = & { g^4 \mu^{4\varepsilon } \over 6}\ , { d^{d-1}k \over ( 2\pi)^{d-1 } } \ , { d^{d-1}q \over ( 2\pi)^{d-1 } } \ , { 1 \over 8 e_k e_q e_r } \
, , \\
s(e_k , e_q , e_r ) & = & { 1 \over i p_0 + e_k + e_q + e_r } - { 1 \over i p_0 - e_k - e_q - e_r } \ , , \end{aligned}\ ] ] where @xmath34 @xmath35 is the the number of space - time dimension introduced in dimensional regularization .
please notice that now the one - loop resummed mass @xmath26 occurs in all on - shell energies .
the real part of @xmath36 contains an ultraviolet divergence
. a detailed discussion of the subtraction of this ultraviolet divergence by a counter - term in the lagrangian and the final expression for the real part of the renormalized self - energy can be found in ref .
we can see from eq .
( [ dkq ] ) that @xmath37 is of the order @xmath23 and depends on the mass parameter @xmath17 of the dimensional regularization .
the complete 2-loop expression for the real part of the self - energy is given by @xmath38 with the last term taken from ref .
a detailed calculation of the imaginary part of the self - energy will be given in the following section .
the leading contribution to the imaginary part of the self energy comes from the 2-loop diagram [ f2]b . making an analytical continuation @xmath39 , @xmath40 , and using @xmath41
, we obtain from eq .
( [ sigma2pp ] ) @xmath42 } \\ & & \big\{\lbrack\delta(\omega+e_k+e_q+e_r)-\delta(\omega - e_k - e_q- e_r)\rbrack \lbrack(1+f_k)(1+f_q)(1+f_r)- f_kf_qf_r\rbrack \nonumber\\ & & \
! + \lbrack\delta(\omega+e_k+e_q - e_r)-\delta(\omega - e_k- e_q+e_r)\rbrack \lbrack(1+f_k)(1+f_q)f_r- f_kf_q(1+f_r)\rbrack \nonumber\\ & & \
! + \lbrack\delta(\omega+e_k - e_q+e_r)-\delta(\omega - e_k+e_q- e_r)\rbrack \lbrack(1+f_k)f_q(1+f_r)-f_k(1+f_q)f_r\rbrack \nonumber\\ & & \
! + \lbrack\delta(\omega - e_k+e_q+e_r)-\delta(\omega+e_k - e_q- e_r)\rbrack \lbrack f_k(1+f_q)(1+f_r)-(1+f_k)f_qf_r\rbrack\big\}. \nonumber \end{aligned}\ ] ] this gives the rates for physical decay and scattering processes @xcite : the first line corresponds to the decay of one off - shell into three on - shell @xmath43 mesons and ( with opposite sign ) the inverse process ; the two @xmath44-functions represent the decay of particles with negative and positive energy , respectively .
the remaining lines in eq .
( [ imsigma ] ) correspond to landau damping via scattering of the off - shell meson with on - shell particles from the heat bath , forward and backward scattering process contributing with opposite signs .
landau damping requires the presence of thermally excited states and vanishes at zero temperature . in each term
every ingoing meson from the heat bath receives a thermal weight @xmath45 and each outgoing one a bose - enhancement factor @xmath46 .
it is easy to show that @xmath47 so we restrict ourselves to positive energy @xmath48 . after suitably relabeling the integration variables ,
the imaginary part of the self - energy ( [ imsigma ] ) for @xmath48 can be reduced to @xmath49 with the 3-body decay and landau damping contributions , respectively : @xmath50 f_k f_q f_r \delta(\omega - e_k - e_q - e_r)\ , , \\
\label{img+2 } { \rm im\ , } g_2(\omega,{\bf p } ) & = & 3 \pi ( e^{\omega / t}-1 ) \int d[k , q ] ( 1+f_k ) f_q f_r \delta(\omega+e_k - e_q - e_r)\ , .
\end{aligned}\ ] ] the @xmath44-functions in eqs .
( [ img+1 ] ) and ( [ img+2 ] ) arise from energy conservation and constrain the integration over @xmath51 and @xmath52 to the kinematically allowed domain .
the evaluation of the resulting kinematic limits of the integration variables is the most demanding part of the following calculation and is shortly discussed in appendix [ appa ] . before going into the technical details of evaluating the integrals , let us summarize some relevant analytical properties of the propagators and self - energies . in the imaginary time formalism , the full propagator @xmath53 has the spectral representation @xmath54 the spectral density satisfies @xmath55 \nonumber\\ & = & { 1\over \pi}\,{{{\rm im\,}\sigma(\omega,{\bf p } ) } \over { [ \omega^2-{\bf p}^2-{m_{\rm p}}^2 + { \rm re}\sigma(\omega,{\bf p})]^2 + [ { \rm im\,}\sigma(\omega,{\bf p})]^2 } } \ , .
\end{aligned}\ ] ] it is an odd function of @xmath56 , see eqs .
( [ resigma ] ) : @xmath57 where @xmath58 .
thus the full propagator @xmath59 has the following analytical structure : if @xmath60 , the full propagator has a cut along the real axis in the complex @xmath56 plane , with a discontinuity across this cut given by the spectral function .
if @xmath61 goes to zero , the identity @xmath62 shows that the spectral density can be expressed as a @xmath44-function at @xmath63 ; the corresponding analytical structure of the full propagator @xmath59 are poles in the complex @xmath56 plane given by the zeroes of the @xmath44-function . from @xmath59 the full retarded propagator @xmath64
is obtained by analytical continuation @xmath65 .
the full propagator in the real time formalism can be written as @xcite @xmath66 all of these full propagators have the same analytical structure . defining the damping rate as @xmath67 and substituting this into eq .
( [ rho ] ) , we see that if @xmath68 , then @xmath69 is strongly peaked at @xmath70 .
the solution of this dispersion relation defines a weakly damped single - particle state which is interpreted as a quasiparticle , the so - called plasmon . denoting the solution for @xmath56 by @xmath71 and defining @xmath72 , the spectral density can in this limit
be expressed approximately in the form of a relativistic breit - wigner function with full width @xmath73 : @xmath74 finally , the spectral function fulfills the sum rule @xcite @xmath75 in the following two subsections we now investigate in detail the properties of plasmons in @xmath4 theory , both at rest and at finite momentum @xmath76 relative to the heat bath . for @xmath77 the evaluation of the imaginary part of the 2-loop self - energy ( [ img ] )
can be carried out in the following manner : for the @xmath78-integration we choose the @xmath79-axis along @xmath80 and define @xmath81 .
the integrals in eqs .
( [ img+1 ] ) and ( [ img+2 ] ) thus take the form @xmath82
\longrightarrow { g^4 \over 48(2\pi)^3 } \int_0^{\infty } dk\ , k^2 \int_0^{\infty } dq\ , q^2 \int_{-1}^1 dt\ , { 1\over e_k e_q e_r}\ , , \ ] ] where @xmath83 .
the energy conserving @xmath44-functions in eqs .
( [ img+1],[img+2 ] ) will be used to perform the @xmath84-integration by using @xmath85 the condition that the zeroes of the @xmath44-function fall into the interval @xmath86 , i.e. @xmath87 , leads to restrictions of the integration regions in @xmath51 and @xmath52 .
these restrictions are worked out in appendix [ appa ] . for the 3-body decay contribution
we find @xmath88 where @xmath89 the specific values @xmath90 and @xmath91 are given by eqs .
( [ k1star ] ) and ( [ qi],[qii ] ) for @xmath77 , respectively , divided @xmath2 .
similarly , the landau damping contribution @xmath92 is given by @xmath93 \nonumber\\ & & \times \bigl(1+f(v)\bigr ) f(w+v ) \ , \ln\left ( { f\left(\varepsilon(y_3)\right ) f\left(\varepsilon(y_4)-v - w\right ) \over f\left(\varepsilon(y_4)\right ) f\left(\varepsilon(y_3)-v - w\right ) } \right ) \ , , \end{aligned}\ ] ] with @xmath94 and @xmath95 given by eqs .
( [ k4star ] ) and ( [ qiii],[qiv ] ) for @xmath77 , respectively , divided by @xmath2 . while @xmath96 vanishes below the 3-particle threshold @xmath97 ( see eq .
( [ epstar ] ) ) , the landau damping contribution @xmath98 ( which arises only at @xmath99 ) is nonzero for _ all _ positive values of @xmath56 . due to the symmetry ( [ resigma ] ) , at 2-loop order
the full propagator thus has a cut along the whole real @xmath56 axis .
this is qualitatively different from the simple double pole structure at 1-loop order . on the plasmon mass shell @xmath100
only landau damping contributes to the imaginary part . from appendix [ appa ]
we find that in this case @xmath101 and thus @xmath102 , @xmath103 as well as @xmath104 .
substituting this into eq .
( [ imgg2 ] ) and using eqs .
( [ img ] ) and ( [ gamma ] ) , we obtain the on - shell plasmon damping rate ( see also @xcite ) @xmath105 \ , , \ ] ] where @xmath106 is the spence function defined as @xmath107 the leading term in the last expression of ( [ gamma1 ] ) was also found in ref .
@xcite .
@xmath108 is finite and does not require any regularization . for dimensional grounds the width @xmath109 for plasmons at rest is thus proportional to the temperature @xmath2 . from ( [ gamma1 ] )
we further see that for small @xmath1 the ratio @xmath110 is proportional to @xmath19 . using eq .
( [ rho ] ) we can also evaluate the spectral function .
we notice that , as discussed in section [ s2 ] , the @xmath111 term in @xmath112 contains the renormalization scale @xmath17 which should be absorbed into the renormalized coupling constant .
but since the @xmath111 corrections have only a minimal effect on the position of the peak in the spectral function , and all @xmath113 terms have already been included in the plasmon mass , we can to very good approximation neglect @xmath114 . in this approximation
the plasma frequency @xmath115 is strictly proportional to @xmath2 .
an additional logarithmic temperature dependence through the running of the coupling constant can only enter in higher order . in the following we comment on some further properties of the plasmon based on numerical results . in fig .
[ f4]a we show for @xmath116 the 2-loop spectral function together with a line indicating the pole position at one - loop order ( including htl resummation ) .
note that there is no visible shift of the peak due to the higher order corrections ; the shift of the peak below the value of @xmath9 is entirely due to the 1-htl - resummation , see eq .
( [ plfreq ] ) .
the major new feature at 2-loop order is just the finite width of the plasmon .
although the 2-loop spectral density is strictly positive for all values of @xmath56 , it approaches zero very fast as one moves away from the plasmon peak .
only in the neighborhood of the 3-body decay threshold @xmath117 it gains some weight again as shown in fig .
[ f4]b ; the reader should notice , however , the tiny vertical scale of this figure . in fig .
[ f5 ] we show the plasmon peak in the spectral function for three different values of @xmath118 . since both the plasmon frequency and its width grow linearly with @xmath2 , and according to eq .
( [ rhobw ] ) @xmath119 , the height of the peak decreases with rising temperature like @xmath120 for fixed @xmath1 .
this is seen in fig .
@xmath121 decreases accordingly with increasing coupling constant like @xmath122 for small @xmath1 . for small @xmath1
the ratio of the damping rate to the plasma frequency is @xmath123 from eq .
( [ plfreq ] ) we see that our resummation procedure is only applicable for @xmath124 . in this domain
the ratio ( [ ratio ] ) is always smaller than about 1/10 , i.e. the plasmon is a well - defined quasiparticle which is not overdamped .
the evaluation of the imaginary part of the self - energy for @xmath126 is a little more complicated because now two non - trivial angular integrations occur . for the @xmath52-integration we choose the @xmath79-axis in the direction of @xmath127 and define @xmath128 . for the @xmath51-integration we choose the @xmath79-axis along @xmath76 and define @xmath129
in this way the integrals in eqs .
( [ img+1 ] ) and ( [ img+2 ] ) can be expressed as @xmath130 \longrightarrow { g^4 \over 96 ( 2\pi)^3 } \int_0^{\infty } dk\ , k^2 \int_{-1}^1 du \int_0^{\infty } dq \
, q^2 \int_{-1}^1 dt\ , { 1\over e_k e_q e_r } \ , .\ ] ] the occurrence of two angles in the integrand renders the determination of the integration limits from the energy - conserving @xmath44-functions more complicated
. however , as shown in appendix [ appa ] , similar methods can be applied as in the case @xmath77 .
the @xmath84-integration in eq .
( [ intukqt ] ) can be performed by using @xmath131 the condition that the zero of the energy @xmath44-function lies inside integration domain leads to kinematic limits for the integrations over @xmath52 , @xmath51 , and @xmath132 .
we find ( see appendix [ appa ] ) @xmath133 with the same dimensionless variables ( [ xya ] ) as before and @xmath134 the limits @xmath90 , @xmath135 , and @xmath136 are now given by eqs . ( [ kistar ] ) , ( [ qi ] ) , and ( [ qii ] ) ,
respectively , divided by @xmath2 . in the limit @xmath137 ( @xmath138 ) eq .
( [ img1p ] ) correctly reduces to eq .
( [ img1 ] ) . for the landau damping contribution @xmath139
we obtain @xmath140 where @xmath141 { x^2 \over |{\bf x}-{\bf z}|\ , \varepsilon(x ) } \nonumber\\ & & + \theta(\omega - e_p ) { 1\over z } \int_a^{\infty } d\varepsilon(x ) \int_{|x - z|}^{|x+z| } d|{\bf x}-{\bf z}| \end{aligned}\ ] ] if @xmath142 and @xmath143 { x^2 \over |{\bf
x}-{\bf z}|\ , \varepsilon(x ) } \nonumber\\ & & + \theta(\omega - e_p ) { 1\over z } \int_a^{\infty } d\varepsilon(x ) \int_{|x - z|}^{|x+z| } d|{\bf x}-{\bf z}| \end{aligned}\ ] ] if @xmath144 , respectively . in deriving these expressions from the results ( [ intku1 ] ) and
( [ intku2 ] ) in appendix [ appa ] , we used for the region @xmath145 the relation @xmath146 to rewrite the @xmath132-integration . since the integrand in ( [ imgg2p ] ) does not depend on @xmath132 , the resulting integrals @xmath147 in ( [ intku11 ] ) and ( [ intku22 ] ) can be evaluated trivially .
again @xmath148 is non - zero for all @xmath48 , and the full propagator has a cut along the whole real @xmath56 axis , arising from landau damping as for the case @xmath149 .
one easily checks that in the zero - momentum limit eq .
( [ imgg2p ] ) reduces to ( [ imgg2 ] ) . on the plasmon energy shell @xmath150
the plasmon damping rate can be deduced as @xmath151 \
, , \end{aligned}\ ] ] where @xmath152 this on - shell result has also recently been obtained by jeon @xcite by a different method . for the numerical evaluation of the spectral function we substitute eqs .
( [ img1p ] ) and ( [ imgg2p ] ) into eq .
( [ img ] ) and then use eq .
( [ rho ] ) . in fig .
[ f6 ] we show the spectral density at fixed @xmath153 for three values of the momentum @xmath154 . as @xmath154 increases , the width of the spectral function decreases as illustrated in fig . [ f7 ] .
for large @xmath154 the width goes to zero , and at the same time the height of the spectral function approaches a constant .
this can been seen explicitly in figs .
[ f7 ] and [ f8 ] .
this means that for large @xmath154 the plasmon looses the features of a collective mode and again behaves like a non - interacting free particle ! at fixed ,
non - zero momentum @xmath76 the dependences on the temperature @xmath2 and the coupling constant @xmath1 , on the other hand , are qualitatively the same as at @xmath77 . from the analysis of htl resummation in gauge theories
@xcite it is known that there the summation scheme can only be guaranteed to work in the soft momentum region @xmath155 .
a partial reason for this is that the thermal loops are momentum dependent , and the particular momentum dependence of the hard " thermal loop arises as the leading term in a high - temperature expansion of the 1-loop self energy @xcite .
correspondingly , one might be inclined to distrust the above results for momenta @xmath156 . in @xmath4 theory
the situation is , however , different in that the htl which we resum is momentum independent to order @xmath15 and @xmath19 ; momentum dependence enters only at order @xmath23 , and for weak coupling @xmath157 as in figs .
[ f6 ] , [ f7 ] it is weak on the scale @xmath26 .
therefore the essential criterium for the validity of the resummation procedure is the smallness of the coupling constant and not that of @xmath158 or @xmath154 .
this argument is also supported by a numerical investigation of the sum rule eq .
( [ sumrule ] ) . in table [ t1 ] we give values for the integral up to the 3-particle threshold , @xmath159 for different values of @xmath160 ( or of the coupling constant @xmath21 ) at @xmath161 and for different values of @xmath154 at fixed @xmath162 . from the given numbers we see that in all cases the sum rule is nearly exhausted by the collective plasmon mode , and that the 3-body decay contribution at @xmath163 is negligible . for large coupling constant ( @xmath164 ) , where the resummation of hard thermal loops is no longer expected to be a good approximation , we begin to see appreciable violations of the sum rule . on the other hand , for @xmath165 ,
the sum rule is very well satisfied even for very large momenta , @xmath166 . as a consequence , the shape of the spectral function ( in particular the interplay of the @xmath167-dependence of its width and height at the plasmon pole shown in figs .
[ f7 ] and [ f8 ] ) is completely controlled by this sum rule .
in this paper we studied the analytical structure of the full propagator and the properties of the plasmon in hot @xmath4 theory at 2-loop order , using the hard thermal loop ( htl ) resummation scheme .
since in @xmath4 theory the one - loop self - energy is completely real , damping of the plasmon occurs only at the 2-loop level .
its physical origin is landau damping by scattering with thermal scalar field quanta .
this results in a cut of the full propagator along the entire real frequency axis .
however , for small coupling constant @xmath1 the spectral function which is given by the discontinuity across this cut shows a strong and rather sharp plasmon peak located at the position expected from the 1-htl - resummation .
its width @xmath168 is approximately given by @xmath169 for plasmons at rest in the heat bath , and this width decreases even further if the plasmon is moving relative to the heat bath . for very fast plasmons
the width goes to zero while the height of the plasmon peak approaches to a constant : the plasmon behaves again like a non - interacting free particle , i.e. it looses the features of a collective mode . within the range of validity of our resummation scheme , which is defined by the smallness of the coupling constant @xmath170
compared to 1 , the plasmon remains a well - defined , weakly - damped quasiparticle excitation for all values of the temperature .
the sum rule eq .
( [ sumrule ] ) for the spectral density is very well satisfied by the 2-loop result even for momenta much larger than @xmath2 , and it is essentially saturated by the plasmon peak .
we are grateful to u. wiedemann for helpful discussions , and we would like to thank the referee for a careful reading of the manuscript and for pointing out an error in the original version of the paper . this work was supported in part by the deutsche forschungsgemeinschaft ( dfg ) , the bundesministerium fr bildung und forschung ( bmbf ) , the national natural science foundation of china ( nsfc ) and the gesellschaft fr schwerionenforschung ( gsi ) .
here we give some technical details on how to solve for the kinematic limits of the integral ( [ intukqt ] ) resulting from the energy conserving @xmath44-functions .
we perform the calculation for arbitrary external momentum @xmath76 ; the expressions needed in section [ s3a ] for plasmons at rest are easily obtained by setting @xmath77 . in this limit
the integrand is independent of @xmath132 , and the @xmath132-integration gives a trivial factor of 2 .
we start by doing the @xmath84-integration in ( [ intukqt ] ) using ( [ dert ] ) and exploiting the energy @xmath44-function . in order to obtain a non - zero contribution ,
the solution @xmath171 of the @xmath44-function constraint must lie in the interval @xmath172 defined by the limits of the @xmath84 integration , @xmath86 .
for fixed @xmath173 ( or fixed @xmath51 and @xmath132 ) these limits on @xmath171 generate upper and lower integration limits for @xmath52 .
if the latter are complex , the corresponding values of @xmath51 and @xmath132 must be excluded from the exterior integrations . for the 3-body decay ( [ img+1 ] ) we must solve @xmath174 , with the notation of eqs .
( [ ref1 ] ) , ( [ ref2 ] ) .
this has no solutions unless @xmath56 exceeds the 3-particle threshold @xmath175 with the help of the auxiliary function @xmath176 the lower and upper integration limits for @xmath52 which result from ( [ eint ] ) can be expressed as @xmath177 \qquad { \rm with } \qquad q'_1 = { 1\over 2 } \left ( \sqrt { a(\omega , |{\bf k}-{\bf p}| ) } - |{\bf k}-{\bf p}| \right)\ ] ] and @xmath178 for @xmath179 , @xmath180 to be real , @xmath181 must be positive .
the function @xmath182 is shown in fig .
@xmath183 has a single zero at @xmath184 which for @xmath77 reduces to @xmath185 a @xmath76-independent double zero at @xmath186 and a singularity in between at @xmath187 which for @xmath77 becomes simply @xmath188 one checks that for @xmath189 the energy constraint can not be satisfied ; for @xmath190 the function @xmath191 is negative , giving rise to complex @xmath52 integration limits .
hence the allowed region for @xmath51 is @xmath192 .
there is no constraint on @xmath132 .
the @xmath52-integration can be done analytically , giving the results presented in section [ s3 ] . for the landau damping term ( [ img+2 ] )
the calculation is slightly more involved .
we must solve @xmath193 with the restriction ( [ eint ] ) .
for fixed @xmath194 the resulting lower and upper limits for @xmath52 can now be expressed in terms of another auxiliary function @xmath195 as @xmath196 \qquad { \rm with } \qquad q'_3 = { 1\over 2 } \left ( \sqrt { b(\omega , |{\bf k}-{\bf p}| ) } - |{\bf k}-{\bf p}| \right ) \ , , \ ] ] and @xmath197 the function @xmath198 is illustrated in fig .
[ f3]b . for @xmath199 ,
@xmath200 is always positive , so @xmath201 are real and the @xmath44-function can be satisfied for all @xmath202 $ ] and @xmath203 $ ] . for @xmath204 , @xmath205 has a positive and real zero at @xmath206 if simultaneously @xmath207 with @xmath208 @xmath209 itself is real only for @xmath210 where @xmath211 for @xmath212 , positivity of @xmath213 thus restricts the integration range of @xmath51 to @xmath214 if @xmath207 , while all @xmath51 are allowed for @xmath215 . for @xmath216
there is no upper constraint on @xmath132 , and @xmath51 is restricted to the region @xmath214
. we can summarize the results so far by writing @xmath219 \nonumber\\ & & \!\ !
+ \theta(\omega - e_p ) \int_{-1}^1 du \int_0^{\infty } dk \ , .
\end{aligned}\ ] ] the missing _
lower _ limits of the @xmath132-integration in two of the terms will be determined now .
( [ kiiistar ] ) shows that for @xmath220 , @xmath217 has a singularity at @xmath221 .
( at @xmath222 , @xmath217 remains finite because the numerator vanishes , too . )
this implies an additional restriction on the lower limit of the @xmath132 integration domain for the two @xmath217-dependent terms in eq .
( [ intku ] ) whenever @xmath220 .
which of these two terms is affected depends on the magnitude of @xmath167 : for @xmath142 we find also @xmath223 , so in this case the second term in eq .
( [ intku ] ) is not affected by this problem because it receives contributions only from @xmath224 .
for @xmath144 we get also @xmath225 , and then both the first and the second term in eq .
( [ intku ] ) are affected .
it turns out that the region @xmath226 is always unphysical because either the function @xmath227 is negative or the @xmath44-function constraint can not be satisfied .
one thus finds , using the dimensionless quantities ( [ xya ] ) , @xmath228 \nonumber\\ & & + \theta(\omega - e_p ) \int_{-1}^1 du \int_0^{\infty } dx \end{aligned}\ ] ] and @xmath229 \nonumber\\ & & + \theta(\omega - e_p ) \int_{-1}^1 du \int_0^{\infty } dx \ , .
\end{aligned}\ ] ] | we study the 2-loop resummed propagator in hot @xmath0 theory .
the propagator has a cut along the whole real axis in the complex energy plane , but for small @xmath1 , the spectral density is sharply peaked around the plasmon . the dispersion relation and the width of the plasmon
are calculated at zero _ and _ finite momentum . at large momenta
the spectral width vanishes , and the plasmon looses its collectivity and behaves like a non - interacting free particle . -1 cm |
a catalog of proper motions in the * international * celestial reference system ( icrs ) , with precise positions and proper motions as well as deep * limiting * magnitude , is important for * research * on the galactic structure , galactic kinematics and stellar * populations*. there are two methods to determine the absolute proper motions of stars based on the hypothesis that the proper motions of background galaxies are zero .
one is that all the observations obtained in different epochs are transformed into one reference position system using galaxies , and absolute proper motions are calculated directly .
for example , munn et al . (
2004 ) took this method to construct the improved sdss proper motion catalog combining usno - b ( monet et al .
2003 ) and sdss ( gunn et al . 1998
; york et al . 2000 ; lupton et al . 2001 ; stoughton et al .
the basic procedure in this work is as follows . for each object from usno - b ,
the nearest 100 galaxies ( classified as @xmath11 by sdss morphological classification ) detected in both sdss and usno - b were selected .
then the mean offsets in right ascension and declination between the sdss and usno - b positions were measured from the 100 nearest galaxies , and were added to the usno - b position of this object .
finally the proper motions of objects in sdss were computed using a linear fit with the re - calibrated usno - b positions and the sdss positions at different epochs .
the other method to determine absolute proper motions is to take a novel calibration method which iteratively uses the stellar sources and galaxies on each position system to eliminate all systematic errors between different position systems .
absolute proper motions out side the plane ( apop ) adopted this technique with the guide star catalog ( gsc- ) schmidt plate data ( qi et al . 2012 ; lasker et al .
meanwhile , the apop used the measured coordinates ( @xmath12 , @xmath13 ) of objects instead of equatorial coordinates to calculate the absolute proper motions .
the principal calibration steps for apop are as follows . 1 . removing the position - dependent systematic errors ( pde ) with a moving - mean filter using good - quality stellar objects ; 2 .
selecting galaxies from non - point - like sources ( non - stars ) via their common null - motion characteristics ; 3 . calibrating magnitude- and color- dependent ( mde and cde ) errors and the residual pde of all objects with respect to the galaxies selected above ; 4 . calculating absolute proper motions from all - epoch plate data with a linear fit .
the construction of the latest version of apop catalog and evaluation of the catalog are described in detail by qi et al .
( 2014 ) . in this paper , we take almost the same method and pipeline as apop to derive absolute proper motions .
the scuss ( zou et al . 2014 ; zhou et al . 2014 ; jia et al . 2014 ) is a @xmath1-band ( 3538 ) imaging survey using the @xmath14 bok telescope .
the survey covers about 5000 square degrees .
most of the scuss fields were observed twice in the course of survey from 2010 to 2013 .
it is difficult to measure accurate proper motions for the scuss objects from the scuss data only . in this paper
, we construct a scuss proper motion catalog by combining the scuss data with the gsc- schmidt plate data .
the gsc- schmidt plate data is an all - sky astro - photometric data base which were derived from the digitization of the palomar and uk schmidt surveys .
the gsc- contains astrometry , photometry and classification for objects down to the limiting magnitude of the plates .
this paper is organized as follows . in section 2 , we summarize the scuss program and gsc- project that are used to derive the scuss proper motion catalog .
section 3 describes the construction of the scuss proper motion catalog . in section 4 , quasars are used to perform internal test of the proper motions .
we compare the catalog with the sdss proper motion catalog in section 5 . we summarize the paper in section 6 .
the scuss is a @xmath1-band imaging survey with an effective wavelength of @xmath0 3538 and a bandwidth @xmath0 345 .
this @xmath1-band filter is slightly bluer than the sdss @xmath1-band filter .
the imaging depth is @xmath0 @xmath15 , roughly @xmath16 deeper than that of the sdss imaging survey .
the scuss project began its observations in september 2010 and ended in october 2013 .
it covers a total 5000 square degrees .
proper motions are measured with respect to a reference frame of galaxies .
hence , the accuracy of star / galaxy classification affects systematic error of the proper motions .
the scuss proper motion catalog covers 3700 square degrees ( galactic latitude @xmath17 and * equatorial latitude @xmath18 * ) . in this area ,
the accuracy of star / galaxy classification is reliable .
the spatial distribution of the scuss proper motion catalog is shown in figure 1 .
most of the images in this area were taken with seeing @xmath19 , sky background @xmath20 adu .
the description of reduction of the scuss data can be found in zou et al.(2014 ) .
photometry was done using sextractor ( bertin & arnouts 1996 ) . the measured @xmath12 and @xmath13 positions were calculated by applying sextractor to sources brighter than 1@xmath21 of the background level .
external astrometric errors were determined by crossing - matching the scuss sources to the ucac4 sources ( zacharias et al .
2013 ) . over the whole survey ,
the mean offsets of right ascension and declination between the scuss and ucac4 are @xmath22 , @xmath23 , respectively . in this work
, we will * construct * a proper motion catalog using the measured coordinates ( @xmath12 , @xmath13 ) , the celestial coordinates ( @xmath24 , @xmath25 ) , the @xmath1-band magnitudes from scuss and the star / galaxy morphological classifications from sdss for objects on scuss images . the basic information for each frame including plate size and scale as well as observation time is also needed in the process of calculation of proper motions .
gsc- ( lasker et al . 2008 ) , as first epoch data to obtain the proper motions of scuss , was originally taken from the survey plates observed by the palomar schmidt telescope and the uk schmidt telescope .
each individual plate was digitized by the guide star automatic measuring machine ( gamma ) and was processed through image processing and plate calibration .
the final products were stored in the compass database ( lasker et al .
gsc- contains astrometry , multi - band photometry ( @xmath26 , @xmath27 , @xmath28 ) and star / non - star classification for objects down to the limiting magnitude of the survey plates .
for stellar sources , the all - sky average absolute position error per coordinate ranges from @xmath29 to @xmath30 depending on magnitude .
stellar photometry is determined to @xmath31 as a function of magnitude and photographic passbands ( @xmath26 , @xmath27 , @xmath28 ) . outside of the galactic plane ,
the catalogue is complete to @xmath32 .
there are about 1210 schmidt plates from different surveys used in this paper .
the main characteristics of the surveys used in this paper including the epoch , band , limiting magnitude and sky coverage as well as the number of plates used in this work are summarized in table 1 . the basic information for objects on each plate including the measured coordinates , the equatorial coordinates , magnitude and classification ( star / non - star ) is needed in this paper . in addition , the basic information for each plate from gsc- data base including telescope name , plate size and scale , observation time is also needed .
two major aspects determine the final accuracy of absolute proper motions .
one is the measured uncertainties of object positions on the images of all epochs , and the other one is the systematic error between different images .
the latter one requires unifying images in different epochs and different bands to one reference system ( qi et al .
2012 ) .
the general steps of the construction of the scuss proper motion catalog are similar to * those * of apop .
relevant equations and details are recommend to see the paper of qi et al .
a few steps are slightly different . the steps are summarized as follows . 1 .
extract all the gsc- plates and scuss images in the same sky field ; 2 .
compare gsc- reference plates with the sdss , update the star / non - star classification in gsc with star / galaxy in the sdss ; 3 .
reject the synthetic double stars ( stars with the same i d ) in each gsc- plate ; 4 .
find common objects between the program images and the reference plate ; 5 .
smooth out position - dependent difference between each image pair using stars with small measuring error ; 6 .
regenerate the @xmath12 , @xmath13 of all the objects on program plates using galaxies and remove the mde , cde and residual pde of proper motions ; 7 .
calculate the absolute proper motions and new positions in a given epoch using all the images .
the plates , which are from gsc- with good quality and overlap with the scuss , were chosen as reference plates .
the star / galaxy classifications of sources play an important role in the calculation of the proper motions .
however , in gsc- , the probability of stars correctly classified drops with magnitude from better than @xmath33 at @xmath34 to less than @xmath35 at @xmath36 , while the percentage that a galaxy is classified as non - star remains above @xmath33 for all the magnitude range . outside of the galactic plane
, stellar classification is reliable to at least 90% confidence for magnitudes brighter than @xmath37 ( lasker et al .
the low accuracy of star / non - star classification of the fainter objects of the gsc- reference plates eventually affect the error of the proper motions . in our work ,
most of the scuss survey are covered by the sdss , star / galaxy separation of sdss is about @xmath38 correct at 21.0 mag in @xmath39 band ( stoughton et al .
so , we update the gsc- star / non - star classification with the star / galaxy classification from sdss morphological classification .
the scuss proper motion catalog contains the equatorial coordinates ( @xmath24 , @xmath25 ) , the absolute proper motions ( @xmath40 , @xmath5 ) , and the scuss @xmath1-band magnitudes for about 7.7 million objects .
the proper motions of quasars can not be detected by present astrometric observation in the optical band , so the intrinsic proper motions of quasars can be considered as zero ( wu et al .
therefore , the measured proper motions of quasars can be used to derive the systematic and random error of the proper motion .
we choose the spectroscopically confirmed quasars from sdss dr9 ( ahn et al .
a total of 22831 common quasars among the sdss and scuss catalog are selected as a quasar sample .
the gaussian function is used to fit the distribution of the quasar proper motions to derive the systematic and random error .
the systematic error of proper motions is derived from the mean proper motion of the quasar sample .
the dispersion of the proper motions of quasars can be considered as the random error of proper motions .
figure 2 displays the histograms of the measured proper motions in right ascension @xmath40 ( upper ) and in declination @xmath5 ( lower ) for all sdss spectroscopically confirmed quasars in scuss proper motion catalog .
the random errors of the proper motions are 4.2 and @xmath41 in @xmath40 and @xmath5 , respectively .
the systematic error of @xmath40 is about @xmath42 and the systematic error of @xmath5 is about @xmath3 .
next , all the 22831 quasars are used to test the position- , magnitude- and color- dependent errors of the scuss proper motions .
the scatters of @xmath40 ( upper ) and @xmath5 ( lower ) of the selected quasar sample as a function of @xmath24 and @xmath25 are shown in figures 3 and 4 , respectively . in order to derive the position - dependent error of the proper motions
, quasars are divided into different position groups with a group size of @xmath43 ( in declination ) and @xmath44 ( in right ascension ) . in figures 3 and 4 ,
red points and red error bars show the systematic and random errors of the quasar proper motions in different groups . as mentioned above
, the systematic and random errors correspond to the mean and dispersion of the best fit gaussian function of the distribution of proper motions for quasar groups .
figure 3 shows that , for @xmath45cos@xmath25 , there is no significant dependence of the systematic or random errors in the @xmath24 direction .
the systematic errors of @xmath45cos@xmath25 for quasar groups are small and fluctuate around zero .
the random errors of @xmath45cos@xmath25 are about @xmath46 .
in addition , there is also no significant dependence of the errors of @xmath5 in @xmath24 direction .
the systematic errors of @xmath5 fluctuate around zero .
the random errors of @xmath5 are around @xmath46 and do not significantly change with @xmath24 . from figure 4 , we can see that there is no significant relation between error ( systematic and random error ) of proper motions and @xmath25 .
the absolute values of systematic errors of @xmath45cos@xmath25 and @xmath5 for most quasar groups are less than @xmath47 and fluctuate around zero .
the random errors of proper motions are about @xmath46 .
so , from figures 3 and 4 , we can find that the pde of scuss proper motion catalog is well eliminated and the random errors of the proper motions do not significantly change with position .
figure 5 shows the histogram of scuss @xmath1-band magnitude distribution of quasar sample .
the quasar sample covers the scuss @xmath1-band magnitude from 16.0 to @xmath48 .
quasars are divided into different magnitude groups with a group size of 1 mag .
figure 6 shows the scatters of @xmath45cos@xmath25 ( upper ) and @xmath5 ( lower ) of quasar sample as a function of scuss @xmath1-band magnitude . in figure 6 ,
the red points and red error bars show the systematic and random errors of quasar proper motions in different groups . from figure 6
, we can see that no mde of the proper motions is found .
the systematic errors of @xmath45cos@xmath25 and @xmath5 are both small .
the random error of proper motions increases from the brightest end of the magnitude group to the faintest end of the magnitude group . for @xmath45cos@xmath25 ,
the random error goes up from @xmath49 for quasars brighter than @xmath50 to @xmath51 for quasars fainter than @xmath52 . for @xmath53 ,
the random error goes up from @xmath54 for quasars brighter than @xmath55 to @xmath56 for quasars fainter than @xmath57 .
the increasing of the random error may be caused by the large measuring errors of the fainter sources on image .
@xmath58 is used to describe the random error of proper motions as a function of magnitude , where @xmath59 is the random error of proper motions , @xmath60 , @xmath61 , and @xmath62 are the unknown parameters to be fitted , and @xmath63 is the scuss @xmath1-band magnitude .
table 2 lists the best fit parameters for proper motions in right ascension and declination .
the scatters of @xmath45cos@xmath25 ( upper ) and @xmath5 ( lower ) of quasars as a function of sdss @xmath64 color are shown in figure 7 .
quasars with @xmath64 color of between @xmath65 and 1.75 are divided into different color groups with a group size of 0.3 .
there is no relation between error of proper motions and color .
the absolute values of systematic errors of @xmath45cos@xmath25 and @xmath53 are around @xmath47 .
the random errors of @xmath45cos@xmath25 and @xmath5 are both about @xmath66 .
the cde of scuss proper motion catalog is well eliminated .
the random error of scuss proper motion catalog does not significantly change with color .
comparing the scuss proper motion catalog with other catalogs can estimate the external error of proper motions .
the sdss proper motion catalog ( munn et al . 2004 , 2008 ; ahn et al .
2012 ) covers most of the scuss fields .
we compare the proper motions between scuss and sdss catalog .
the sdss proper motion catalog is derived by combining sdss and re - calibrated usno - b astrometry . for sdss proper motion catalog ,
the random errors in right ascension and declination are about @xmath67 and @xmath68 , respectively ( munn et al .
2004 ) . the median proper motion for all the quasar sample is about @xmath69 , but
the systematic errors can be larger by a factor of 2 - 3 in small sky patches ( bond et al . 2010 ; ahn et al .
2012 ) .
the common stars between scuss catalog and sdss catalog are eslected and divided into different position ( @xmath24 , @xmath25 ) and magnitude groups .
the means and the dispersions of differences of proper motions in different common star groups are derived .
the mean differences of proper motions as a function of @xmath24 are plotted in figure 8 .
the upper panel is the mean differences of @xmath45 of stars and the lower panel is the mean differences of @xmath5 of stars .
figure 8 shows that , for many regions , the mean differences of proper motions between sdss and scuss are small . from the upper panel of figure 8
, we can see that , in the region @xmath70 , the absolute values of mean differences of @xmath45 of common stars are about @xmath71 .
the lower panel shows that , in the region @xmath72 , @xmath73 , @xmath74 and @xmath75 , the absolute values of mean differences of @xmath5 of common stars are about @xmath76 .
however , in the region @xmath77 , the absolute value of mean difference of @xmath5 of stars reaches about @xmath78 . in most regions , the mean differences of the proper motions are smaller than the random errors of sdss and scuss proper motions .
the means and dispersions of differences of proper motions as a function of @xmath25 are plotted in figure 9 . from figure 9
, we can see that the mean differences of proper motions between sdss and scuss are small .
the absolute values of mean differences are less than @xmath79 .
the means and dispersions of differences of proper motions as a function of scuss @xmath1-band magnitude are plotted in figure 10 .
figure 10 shows that , in the magnitude range @xmath80 , the mean differences of proper motions between sdss and scuss are small and the dispersion of the differences of proper motions increases with magnitude .
the absolute proper motions of scuss program are computed by combining the scuss and gsc- .
the basic method of calculation of absolute proper motions follows that of apop . the measured coordinates ( @xmath12 , @xmath13 ) of objects instead of
equatorial coordinates are used to calculate the absolute proper motions .
the stellar sources with good imaging quality are used to eliminate the pde between the reference and program image .
the galaxies with good imaging quality are used to eliminate the mde , cde and the residual pde .
the 22831 common quasars among scuss and sdss are selected to test the properties of scuss proper motion catalog .
the systematic errors of the proper motions are @xmath2 and @xmath3 in @xmath45cos@xmath25 and @xmath5 .
in addition , the random errors of the proper motions are 4.2 and @xmath41 in @xmath45cos@xmath25 and @xmath5 .
we made a detailed analysis of the obtained proper motions in order to determine the position- , magnitude- and color- dependent errors of the proper motions .
finally , we find that the pde , mde and cde of scuss proper motion catalog are small .
the position- and color- dependent random errors remain about the same .
the random errors of @xmath45cos@xmath25 and @xmath5 increase with magnitude . beyond that , no significant relations between errors ( systematic and random ) and the position , magnitude and color are found .
the common stars between scuss and sdss catalog are selected . the means and dispersions of differences of proper motions between scuss and sdss
are derived .
the mean differences of proper motions between sdss and scuss as a function of @xmath24 are small except for some regions .
there is no relation between the mean differences of proper motions and @xmath25 .
the mean differences of proper motions are small in magnitude region @xmath81 .
the scuss proper motion catalog with small systematic error and low random error can be used for many astronomical studies such as the galaxy disk kinematics , substructure of the galaxy and galaxy rotation curve .
the release of the scuss proper motion catalog is in preparation .
the final version of the scuss proper motion catalog will be available when the scuss data becomes public .
if you need a preliminary version of scuss proper motion catalog , you can email xiyan peng ( [email protected] ) .
we would like to thank the anonymous referee for providing rapid and thoughtful report that helped improve the original manuscript greatly .
this work is supported by national basic research program of china ( 973 program ) , no.2014cb845702 , no .
2014cb845704 , and no .
this work is also supported by the national natural science foundation of china ( nsfc , nos.11433005 , 11373035 , 11373033 , 11373003 , 11273003 , 11203034 , 11203031 , 11303038 , 11303043 ) .
the project is managed by the national astronomical observatory of china and shanghai astronomical observatory .
this work is also supported by the joint fund of astronomy of the national nature science foundation of china and the chinese academy of science , under grants u1231113 .
the scuss is funded by the main direction program of knowledge innovation of chinese academy of sciences ( no .
kjcx2-ew - t06 ) .
it is also an international cooperative project between national astronomical observatories , chinese academy of sciences and steward observatory , university of arizona , usa .
technical supports and observational assistances of the bok telescope are provided by steward observatory .
lupton , r. , gunn , j. e. , ivezi , . ,
knapp , g. r. , kent , s. , @xmath82 yasuda , n. 2001 , in asp conf .
238 , astronomical data analysis software and systems x , ed .
f. r. harnden , jr .
, f. a. primini , @xmath82 h. e. payne ( san francisco : asp ) , 269 | absolute proper motions for @xmath0 7.7 million objects were derived based on data from the south galactic cap @xmath1-band sky survey ( scuss ) and astrometric data derived from uncompressed digitized sky surveys that the space telescope science institute ( stsci ) created from the palomar and uk schmidt survey plates .
we put a great deal of effort into correcting the position- , magnitude- and color- dependent systematic errors in the derived absolute proper motions .
the spectroscopically confirmed quasars were used to test the internal systematic and random error of the proper motions .
the systematic errors of the overall proper motions in the scuss catalog are estimated as @xmath2 and @xmath3 for @xmath4 and @xmath5 , respectively .
the random errors of the proper motions in the scuss catalog are estimated independently as 4.2 and @xmath6 for @xmath4 and @xmath5 .
there is no obvious position- , magnitude- and color- dependent systematic errors of the scuss proper motions .
the random error of the proper motions goes up with the magnitude from about @xmath7 at @xmath8 to about @xmath9 at @xmath10 .
the proper motions of stars in scuss catalog are compared with those in the sdss catalog , and they are consistent well . |
model categories were introduced in @xcite , in which a main example is @xmath0 , the category of simplicial sets .
the proof given is combinatorial but uses the theory of _ minimal fibrations _ , thus relies essentially on the axiom of choice and so is very much non - constructive .
another popular approach can be found in @xcite , where the category of simplicial sets is related to a category of spaces via a nerve - realization adjunction .
we begin by introducing a new viewpoint on the trivial cofibrations , which amounts to the observation that many examples of trivial cofibrations are in fact _ strong anodyne extensions _ ,
i.e. relative cell complexes of horn inclusions ( not featuring retracts ) .
we introduce the notion of _ p - structure _ , a convenient tool for exhibiting strong anodyne extensions , and use it for several examples .
the main example is the natural embedding of a simplicial set @xmath1 into its kan fibrant replacement @xmath2 , whose properties we review within this framework . from this work , and a handful of elementary facts about simplicial homotopy
, we can quickly deduce the existence of the kan - quillen model structure .
finally , in an appendix , we apply this technique to prove a mild strengthening of the right - properness of this model structure .
specifically , we show that the pullback of a horn inclusion along a fibration is a _
strong _ anodyne extension . viewed in dual form ,
this describes how to equip the dependent product of a fibration along a fibration with the _ structure _ of a fibration when given choices of such structure for the two original fibrations .
throughout this section @xmath3 will denote a monomorphism in @xmath0 .
[ def - sae ] we say @xmath4 is a _ strong anodyne extension _ ( sae ) if it admits an anodyne presentation .
an _ anodyne presentation _ for @xmath4 consists of an ordinal @xmath5 , a @xmath5-indexed increasing family of subcomplexes @xmath6 of @xmath7 , satisfying : * @xmath8 , @xmath9 . * for every non - zero limit ordinal @xmath10 , @xmath11 , * for every @xmath12 , the inclusion @xmath13 is a pushout of a coproduct of horn inclusions .
[ rem - sae ] the definition of the class of strong anodyne extensions given above is the same as the original ` weak ' version given in @xcite , minus the condition of closure under retracts .
the usual factorization of a simplicial map as an anodyne extension followed by a kan fibration using quillen s small object argument ( introduced in ii.3 of @xcite ) gives us a factorization as a _
strong _ anodyne extension followed by a kan fibration .
it follows easily that the usual notion of anodyne extension is recovered by closing the strong anodyne extensions under retracts .
observe that , in the final clause of [ def - sae ] , if @xmath14 is somehow a pushout of a coproduct of horn inclusions then it is so in a unique way .
non - degenerate simplices in @xmath15 but not in @xmath16 come in pairs : one simplex which is maximal with respect to being absent from @xmath16 paired with one of its maximal faces . from such data
one can infer which horns must feature in the description .
in fact , one can essentially describe the anodyne presentation using just such information .
[ not - nond ] let @xmath1 be any simplicial set , then by @xmath17 we denote the set of non - degenerate simplices of @xmath1 .
[ def - parent - child ] suppose @xmath3 is equipped with some fixed anodyne presentation .
let @xmath18 , then there exists a least @xmath19 such that @xmath20 . by ` continuity ' at limit ordinals
, @xmath19 must be @xmath21 for some @xmath22 with @xmath23 . as described above
, @xmath24 is either a maximal simplex such that @xmath25 , or if not it is a maximal face of some @xmath26 maximal with @xmath27 . in the former case
, we say that @xmath24 is a _
type i simplex _ and in the latter we say that @xmath24 is a _ type ii simplex_. in the latter case we write @xmath28 . [ def - ancestral - pre ] suppose again that @xmath3 comes with some fixed anodyne presentation . the _ ancestral preorder _
@xmath29 on @xmath30 is the smallest transitive relation such that 1 . whenever @xmath31 and @xmath32 is a face of @xmath33 , then @xmath34 , 2 . whenever @xmath28 then we have @xmath35 ( and by ( a ) we have @xmath36 ) .
[ def - ancestral ] in the _ ancestral relation _
@xmath37 on @xmath38 , the elements of @xmath39 are indiscernible minimal elements , and @xmath37 agrees with @xmath29 on @xmath30 , except that now for each type ii simplex @xmath24 , @xmath24 and @xmath40 are not related .
[ prop - anc - wf ] suppose that @xmath3 comes with some anodyne presentation .
then the ancestral relation is well - founded .
define a map @xmath41 valued in ordinals by sending @xmath24 to the least @xmath22 such that @xmath42 . then @xmath43 is relation - preserving : @xmath44 for every type ii simplex @xmath24 , every other face of @xmath40 must be in @xmath45 for some @xmath46 , and @xmath47 . but the existence of a relation - preserving map into @xmath48 is equivalent to well - foundedness .
we could have defined the ancestral relation directly via the map @xmath49 in [ prop - anc - wf ] , but observe that definition [ def - ancestral ] relies on no aspect of the anodyne presentation other than the division of @xmath30 into type i and type ii simplices .
[ def - structure ] a _ p - structure _ on @xmath3 consists of * a partition @xmath50 into type i and type ii simplices . * a bijection @xmath51 , such that * @xmath52 , * @xmath24 is a face of @xmath40 in a unique way , i.e. @xmath53 for a unique @xmath54 , * the ancestral relation , defined exactly as in [ def - ancestral ] is well - founded . [ lem - finite - pred ] given a p - structure for @xmath3 and @xmath18 , then @xmath55 is finite .
define @xmath56 . given @xmath57 ,
define @xmath58 to be the set of proper faces of elements of @xmath57 together with the proper faces of @xmath59 for every @xmath60 .
choose a relation preserving map @xmath61 .
then since each @xmath57 is finite , the sequence @xmath62 is strictly decreasing while it is non - zero , and hence is eventually 0 , i.e. eventually @xmath63 .
but then @xmath57 is eventually constant .
hence @xmath64 is finite , but @xmath65 .
[ prop - structure - to - presentation ] a p - structure on @xmath3 gives rise to an anodyne presentation for @xmath4 ( of ` length ' at most @xmath66 ) .
define @xmath67 by [ lem - finite - pred ] , if @xmath68 then @xmath69 is a positive integer .
now we can define @xmath70 by @xmath71 it is now easy to check the conditions of [ def - sae ] .
the following is immediate .
[ cor - length - omega ] any strong anodyne extension admits an anodyne presentation of length @xmath66 , i.e. for which @xmath72 .
compare this with the following consequence of the small object argument : every ( not necessarily strong ) anodyne extension is a retract of a monomorphism which admits a length @xmath66 anodyne presentation . in many examples
it turns out to be considerably simpler to give the data of a p - structure than the full anodyne presentation . as a first demonstration
, we give a proof of a classical result ( which appears as i.4.6 in @xcite ) . in it
its usual form , the word `` strong '' is omitted : [ lem - horn - bdy ] let @xmath73 , @xmath74 and @xmath75 .
then @xmath76 is a strong anodyne extension . for convenience assume @xmath77 .
the non - degenerate simplices of @xmath78 may be identified with `` @xmath79-lattice walks on @xmath80 '' , that is , sequences @xmath81 of some length @xmath82 , such that for @xmath83 we have @xmath84 and @xmath85 with at least one of these inequalities strict .
if we think of @xmath86 as labelling rows and @xmath87 as labelling columns , then a walk in the grid represents a simplex : * in @xmath88 if and only if some column is skipped , * in @xmath89 if and only if some row other than @xmath90 is skipped .
consider a non - degenerate simplex @xmath91 , where @xmath92 . now
every ` move ' in the walk is either @xmath93 , @xmath94 , @xmath95 , or one of @xmath96 and @xmath97 where the @xmath90^th^ row is skipped .
let us declare that @xmath24 is type ii if and only if it contains a move of the form @xmath97 ( skipping the @xmath90^th^ ) , or if the last point of the walk on the @xmath90^th^ row is followed by a move of the form @xmath95 . in the first case , define @xmath40 to be the same walk but with that move split into @xmath95 , @xmath93 . in the second case ,
define @xmath40 to be the same walk but with that move split into @xmath94,@xmath93 .
this is enough to determine the p - structure .
the conditions of [ def - structure ] are easy to check ( well - foundedness is simply due to finiteness ) .
in fact , the proof works for @xmath98 with no modification .
the case @xmath99 follows by symmetry ( or consider the ` dual ' decomposition @xmath100 , which looks at the _ first _ point on the @xmath90^th^ row ) .
[ cor - anodyne - mono ]
let @xmath3 be a strong anodyne extension , and @xmath101 any monomorphism .
then @xmath102 is a strong anodyne extension .
( sketch ) . without the word `` strong ''
, this appears as i.4.6 in @xcite . express the displayed map as a transfinite composite of strong anodyne extensions , by using the given anodyne presentation for @xmath4 and the skeleton decomposition of @xmath103 .
we present a second example in the use of p - structures , which makes use of definitions recalled below in [ def - exinf ] .
[ prop - subd - sae ] the subdivision functor @xmath104 preserves strong anodyne extensions .
it suffices to show that any horn inclusion is sent to a strong anodyne extension .
fix some @xmath105 ( otherwise the proposition is trivial ) and for notational convenience assume @xmath106 , so we consider @xmath107 .
thinking of the non - degenerate simplices of @xmath108 as strictly increasing sequences of subsets of @xmath109 , they come in six forms : * @xmath110 , where @xmath111 and @xmath112 and @xmath113 for each @xmath54 , * @xmath114 , where @xmath111 and @xmath112 and @xmath113 for each @xmath54 , * @xmath115 , where @xmath116 , @xmath111 and @xmath117 , @xmath118 for each @xmath119 , and @xmath120 where @xmath121 in case @xmath122 , * @xmath123 , where @xmath116 , @xmath111 , and @xmath117 , @xmath118 for each @xmath119 , * @xmath124 , where @xmath125 , * @xmath126 , where @xmath125 .
then there is a p - structure for which the type ii simplices are precisely those @xmath24 of form ( a ) , ( c ) or ( e ) , where @xmath40 is the obvious simplex of form ( b ) , ( d ) or ( f ) respectively . the conditions of [ def - structure ] are easy to check .
let us recall some definitions .
an account of @xmath127 can be found iii.4 of @xcite .
[ def - exinf ] let @xmath128 be the simplex category with objects @xmath129 : n \geq 0\}$ ] .
recall that there is a functor _ nerve _ @xmath130 .
recall also that for any standard simplex @xmath131 , the set @xmath132 is naturally a poset and is isomorphic to the poset of non - empty subsets of @xmath109 .
we define _ subdivision _ @xmath133 by @xmath134 ) = n ( { ( \triangle^n)_{\textrm{\scriptsize n.d.}}}).\ ] ] we may extend subdivision to a functor @xmath135 by left kan extension .
this functor has a right adjoint , called _ extension _ , given by @xmath136 we think of this as the collection of all ` binary pasting diagrams ' in @xmath1 , and may thus refer to elements of @xmath137 as ` diagrams ' .
we also need the _ last - vertex map _ , @xmath138 : @xmath139 the family @xmath140 is easily seen to be natural , and hence gives rise to a map @xmath141 , where an @xmath103-simplex @xmath142 is sent to @xmath143 .
finally , we define @xmath2 to be the colimit of the sequential diagram @xmath144 ( a ) -- node[auto ] { $ j_x$ } ( b ) ; \draw[->,font=\footnotesize ] ( b ) -- node[auto ] { $ j_{{\mathop{\mathsf{ex}}}x}$ } ( c ) ; \draw[->,font=\footnotesize ] ( c ) -- node[auto ] { $ j_{{\mathop{\mathsf{ex}}}({\mathop{\mathsf{ex}}}x)}$ } ( d ) ; \end{tikzpicture}\ ] ] there is a map @xmath145 which comes from the colimit , which we also denote this by @xmath146 .
the main contribution in this section is a new proof that the map @xmath147 is a weak homotopy equivalence , indeed a strong anodyne extension .
combined with the following classical result , this means that @xmath127 is a functorial fibrant replacement for the kan - quillen model structure on simplicial sets .
[ prop - exinf - kan ] @xmath2 is a kan complex .
see , for example , iii.4.8 in @xcite . since @xmath2 is the result of freely adjoining ` composites ' for diagrams of shape @xmath108 in @xmath1
, we might interpret the proof of [ prop - exinf - kan ] as telling us how to construct fillers for horns in situations where we initially know how to paste together subdivision diagrams .
we may interpret the proof of [ thm - ex - sae ] below as telling us how to express the pasting of subdivision diagrams in terms of horn - filling operations .
observe that filling a horn of shape @xmath148 takes @xmath103 @xmath103-simplices as input to produce another @xmath103-simplex ( and an @xmath149-simplex ) , whereas pasting together a diagram of shape @xmath108 takes @xmath150 @xmath103-simplices as input to produce another @xmath103-simplex .
thus we should expect that modelling subdivision via horn - filling will take an increasingly large number of steps as dimension increases . to cope with this , we need to introduce a carefully chosen notion of complexity on the elements of @xmath151 .
[ def - jk ] let @xmath152 with @xmath153 .
define @xmath154 is the nerve of the unique map @xmath155 that preserves binary joins and satisfies : @xmath156 for example , @xmath157 is the identity , and @xmath157 is the last vertex map @xmath138 composed with the appropriate inclusion @xmath158 .
it is easily seen that @xmath159 is idempotent , i.e. @xmath160 , so that @xmath161 is idempotent for any @xmath1 .
splitting these idempotents gives us a ` filtration ' of @xmath151 .
[ notn - filtration - ex - x ] fix some simplicial set @xmath1 .
let @xmath162 it is easily checked that @xmath163 for @xmath164 , so that we may write @xmath165 roughly speaking , each @xmath166 is the collection of those diagrams of shape @xmath108 in @xmath1 which consist of only @xmath167 non - degenerate simplices .
now we introduce a generalization of the codegeneracy maps whose role , approximately , is to reduce the construction of one element of @xmath168 to the construction of @xmath169 elements of @xmath166 followed by a horn - filling .
[ def - rk ]
let @xmath170 with @xmath153 , define @xmath171 as the nerve of the unique map @xmath172 that preserves binary joins and satisfies : @xmath173 these maps are almost the subdivided codegeneracy maps @xmath174 , except for their behaviour at @xmath175 .
indeed , @xmath176 .
we now have all the definitions needed to prove : [ thm - ex - sae ] the map @xmath147 is a strong anodyne extension . we can give a p - structure explicitly .
let us say that a non - degenerate @xmath103-simplex @xmath177 is has type ii if and only if it is _ not _ of the form @xmath178 with @xmath179 , for any @xmath180 . in this case ,
let us define @xmath181 , where @xmath90 is determined by @xmath182 .
the rest of the proof is simply to check that this works even though this will be difficult , the specification of the p - structure is very simple and , as we shall see in [ prop - reflects ] , useful too . in checking
that this p - structure is well - defined , we will need the following collection of equations .
[ lem - jr - eqns ] under the individually given assumptions , the following equations hold : @xmath183 this is an easy calculation at the level of posets prior to taking the nerve and using the fact that all the maps involved preserve non - empty joins .
we begin by showing that the operation @xmath184 is injective .
[ lem - i - in - a - unique - way ] there is at most one way to write any @xmath185 as @xmath186 where @xmath187 and @xmath188 .
suppose @xmath189 with @xmath190 for @xmath191 .
then by , we have @xmath192 for @xmath191 .
by we have @xmath193 , so in particular it is enough to show that @xmath194 .
now @xmath195 using .
hence @xmath196 , hence @xmath197 .
but by a similar argument @xmath198 .
the next lemma shows that the image of the operation @xmath184 does indeed contain all type i simplices .
[ lem - all - have - a - type ] if @xmath199 is of the form @xmath200 where @xmath187 and @xmath182 , then either @xmath201 is type ii or @xmath202 is degenerate .
if @xmath201 is degenerate , then it follows immediately from and that @xmath202 is as well .
if @xmath201 is non - degenerate yet not of type ii , then there is an expression @xmath203 , with @xmath180 and @xmath204 .
since @xmath205 by , it follows by that @xmath206 , hence @xmath207 . now applying we see that @xmath208 hence @xmath209 , so indeed @xmath210 .
it now follows from that @xmath211 is degenerate .
it is clear from the definition of @xmath184 that for any type ii @xmath103-simplex @xmath201 , @xmath212 is type i unless it is degenerate or in @xmath213 .
let @xmath185 be type ii .
then there is a unique @xmath214 such that @xmath215 . in particular
, @xmath212 is non - degenerate .
suppose @xmath182 . *
if @xmath216 , we have @xmath217 by . * if @xmath218 , then by , @xmath219 , so it can not be equal to @xmath201 . * if @xmath220 , then by , @xmath221 . by ,
+ if we also had @xmath223 , then putting @xmath224 , we can exhibit @xmath225 in the form @xmath226 with @xmath187 and @xmath227 .
that is , @xmath225 would fail to be of type ii even if it were non - degenerate , so it can not be equal to @xmath201 .
+ thus we may assume that @xmath228 .
but now by , @xmath225 is degenerate and so not equal to @xmath201 .
observe that for any type ii @xmath103-simplex @xmath201 , the equation @xmath229 would imply ( by and ) that @xmath230 , so indeed @xmath231 .
combining this with the previous lemmas gives us that the operation @xmath184 is indeed a bijection between the type ii and type i simplices .
all that remains is to check that the induced ancestral relation is well - founded .
rank _ of @xmath185 be the pair @xmath232 with @xmath90 minimal such that @xmath233 .
ranks are ordered lexicographically .
we show by induction on rank that each simplex has only finitely many ancestors . without loss of generality ,
let @xmath201 be a type ii simplex of rank @xmath232 , @xmath187 .
consider the other maximal faces of @xmath212 .
if @xmath218 , @xmath225 has rank at most @xmath234 by .
thus , by induction , each of the finitely many faces @xmath225 with @xmath218 contributes only finitely many ancestors to @xmath212 . if @xmath220 , then @xmath221 by . now if @xmath235 is in @xmath236 , it will follow from that @xmath225 is degenerate , so all of its ( finitely many ) non - degenerate faces have rank less than @xmath232 .
thus we may assume that @xmath237 .
there are three cases to consider . *
if @xmath235 is degenerate then , by either or as appropriate , @xmath225 is also degenerate , so as before all of its non - degenerate faces have rank less than @xmath232 . *
suppose @xmath235 is type i , say @xmath238 where @xmath239 .
then @xmath207 , since either @xmath240 or @xmath241 by and .
now by either or as appropriate , we again find that @xmath225 is degenerate .
* finally , suppose @xmath235 is type ii .
then since @xmath242 and @xmath237 , we see that @xmath225 is type i. but each of its ( finitely many ) faces has dimension less than @xmath103 , therefore rank less than @xmath232 , so we are done by induction .
the map @xmath243 is a strong anodyne extension .
the following useful properties of @xmath127 are well - known ( except for strong anodyne extensions ) .
[ prop - properties ] @xmath127 preserves : finite limits , monomorphisms , strong anodyne extensions , fibrations , trivial fibrations , simplicial homotopies , homotopy equivalences .
finite limits follows from the fact that @xmath127 is a filtered colimit of right adjoints monomorphisms , homotopies and homotopy equivalences now follow easily . to see that @xmath127 preserves fibrations
it is enough to see that @xmath244 sends the boundary inclusions to monomorphisms , but this is obvious .
preservation of trivial fibrations is similar , using [ prop - subd - sae ] . finally , let @xmath245 be a strong anodyne extension and consider the diagram : @xmath246 ( x ) edge ( ex ) edge node[auto , swap ] { $ m$ } ( y ) ( ex ) edge ( un ) edge[bend left=20 ] node[auto ] { $ { \mathop{{\mathop{\mathsf{ex}}}^\infty}}m$}(ey ) ( y ) edge ( un ) edge[bend right=20 ] ( ey ) ; \draw[->,dotted ] ( un ) edge ( ey ) ; \draw ( un ) + + ( -.5,.2 ) -- + + ( 0,0.3 ) -- + + ( 0.3,0 ) ; \end{tikzpicture}\ ] ] it is clear how every arrow in the diagram admits a p - structure ( transferring across pushouts in the obvious way ) , except possibly for @xmath247 and the dotted arrow
. however , the p - structure on @xmath248 is given by restriction from the one on @xmath249 .
it is easy to check now that in this situation , the p - structure on @xmath249 also induces a p - structure on the dotted arrow .
hence @xmath247 is the composite of two strong anodyne extensions .
the following result contains all remaining difficulty in establishing the model structure axioms .
[ prop - reflects ] @xmath127 reflects the triviality of fibrations .
that is , if @xmath250 is a fibration and @xmath251 ( which is automatically a fibration ) is moreover a _ trivial _
fibration , then @xmath252 is also a trivial fibration . consider the lifting problem @xmath253 ( b ) edge ( t ) edge ( x ) ( x ) edge ( y ) ( t ) edge ( y ) ; \draw[->,dotted ] ( t ) -- ( x ) ; \end{tikzpicture}.\ ] ] there is a diagonal filler if we compose this square with the naturality square for @xmath254 : @xmath255 ( b ) edge ( t ) edge ( x ) ( x ) edge ( y ) edge ( eix ) ( t ) edge ( y ) ( y ) edge ( eiy ) ( eix ) edge ( eiy ) ; \draw[->,preaction={draw = white , line width=6pt } ] ( t ) -- ( eix ) ; \end{tikzpicture}.\ ] ] by finiteness of @xmath131 , we can replace @xmath127 with @xmath256 for some @xmath103 . by an inductive argument
, we can reduce to the case with @xmath257 .
@xmath258 ( b ) edge ( t ) edge node[auto ] { $ b$ } ( x ) ( x ) edge ( y ) edge ( eix ) ( t ) edge node[auto , swap ] { $ \sigma$ } ( y ) ( y ) edge ( eiy ) ( eix ) edge ( eiy ) ; \draw[->,preaction={draw = white , line width=6pt } ] ( t ) -- node[auto , pos=0.3 ] { $ \tau$ } ( eix ) ; \end{tikzpicture}.\ ] ] assuming @xmath259 does not factor through @xmath1 , let @xmath260 $ ] be the ` minimal anodyne subextension ' of @xmath261 containing @xmath259 . letting @xmath262 , where @xmath263 is a non - degenerate
@xmath4-simplex of @xmath137 and @xmath264 \to [ m]$ ] is a surjection in @xmath128 , this is the subcomplex whose non - degenerate simplices are @xmath265)_{\textrm{\scriptsize n.d . } } } = \{\sigma \in { ( { \mathop{\mathsf{ex}}}x)_{\textrm{\scriptsize n.d . } } } \mid \sigma { \mathrel{\preceq_p}}\upsilon \},\ ] ] i.e. @xmath263 and all predecessors of @xmath263 , together with @xmath266 if @xmath263 has type ii . clearly @xmath267 $ ] is a strong anodyne extension .
the crucial point now is that @xmath260 \to { \mathop{\mathsf{ex}}}x \to { \mathop{\mathsf{ex}}}y$ ] factorizes through @xmath268 .
we show by induction on @xmath37 that for any @xmath269)_p$ ] , its image in @xmath270 is in @xmath271 .
this is true for any simplex of @xmath1 .
we must also check that any face of a map in @xmath272 is in @xmath271 and that for any @xmath273 , @xmath274 for any @xmath275 .
these are just special cases of and .
now consider the lifting problem @xmath276 $ } ; \node ( x2 ) at ( 2,0 ) { $ x$ } ; \node ( y ) at ( 2,-2 ) { $ y$ } ; \draw[- > ] ( x1 ) edge ( xt ) ( xt ) edge ( y ) ( x2 ) edge node[auto ] { $ f$ } ( y ) ; \draw[- > ] ( x1 ) edge ( x2 ) ; \draw[->,dotted ] ( xt ) -- ( x2 ) ; \end{tikzpicture}.\ ] ] since @xmath252 is a fibration , there is a certainly a solution .
but the image of @xmath259 under this map is a solution to the original lifting problem .
we begin with the definition of the classes of morphisms .
it is easiest to describe all five classes and then subsequently to show that they are interrelated correctly .
[ def - classes ] the class @xmath277 of _ weak equivalences _ is the class of all morphisms @xmath252 such that @xmath251 is a simplicial homotopy equivalence . the class @xmath278 of _ cofibrations _ is the class of monomorphisms .
the class of _ trivial cofibrations _ is the class of morphisms which are retracts of strong anodyne extensions .
the class @xmath279 of _ fibrations _ is the class of morphisms with the right lifting property with respect to all horn inclusions .
the class of _ trivial fibrations _ is the class of morphisms with the right lifting property with respect to all monomorphisms .
[ lem - weq ] the weak equivalences have the 2-out - of-3 property and are closed under retracts .
it is an easy exercise that homotopy equivalences have these properties .
[ lem - factorization ] any map @xmath250 can be factorized as a cofibration followed by a fibration , and we may choose either map to be trivial .
this is quillen s small object argument introduced in @xcite .
note that in the ( trivial cofibration)fibration factorization we in fact get a strong anodyne extension for the first factor .
[ lem - saturation ] the saturation axiom holds for these classes .
that is , the class of trivial cofibrations is precisely the class of maps with the left - lifting property with respect to the fibrations , as is the class of cofibrations to the trivial fibrations ; and the class of trivial fibrations is precisely the class of maps with the right - lifting property with respect to the cofibrations , as is the class of fibrations to the trivial cofibrations .
the third assertion is nothing more than definition [ def - classes ] , and the fourth follows easily from the definitions by standard colimit closure properties of left classes .
the former two assertions are now easy consequences of [ lem - factorization ] and the fact that monomorphisms are stable under retracts .
lemma [ lem - saturation ] is the only place in this argument where we essentially require the axiom of choice .
even so , it can be avoided by changing definition [ def - classes ] so that the fibrations are defined to have the lifting property with respect to all anodyne extensions .
this is in practice no great inconvenience since it amounts only to being able to make a simultaneous choice of lifting for each horn inclusion : this is actually what we get from constructions like the small object argument .
results like [ prop - properties ] can be seen as transporting the _ structure _ of a kan fibration on @xmath252 to the structure of one on @xmath251 . to complete the proof , we need a few basic results which have elementary proofs in the literature .
[ lemb - anod ] an anodyne extension between kan complexes is part of a simplicial homotopy equivalence .
proposition 3.2.3 in @xcite [ lemb - triv ] any trivial fibration is part of a simplicial homotopy equivalence .
proposition 3.2.5 in @xcite .
[ lemb - fib ] a fibration between kan complexes which is part of a simplicial homotopy equivalence is a trivial fibration .
proposition 3.2.6 in @xcite .
the classes defined in [ def - classes ] give rise to a model structure on @xmath0 .
it remains only to show that @xmath280 is indeed the class of trivial cofibrations as described in [ def - classes ] and that @xmath281 is the class of trivial fibrations .
clearly every trivial fibration is a fibration and every trivial cofibration is a cofibration .
we just have four inclusions to prove .
let @xmath282 be a strong anodyne extension . then by [ prop - properties ] , @xmath247 is a strong anodyne extension .
now by [ lemb - anod ] , @xmath247 is a simplicial homotopy equivalence , so @xmath4 is a weak equivalence .
any retract of @xmath4 is a weak equivalence by [ lem - weq ] .
let @xmath250 be a trivial fibration . by [ lemb - triv ]
@xmath252 is a simplicial homotopy equivalence , so by [ prop - properties ] @xmath251 is a simplicial homotopy equivalence and so @xmath252 is a weak equivalence .
now suppose @xmath250 is a fibration which is a weak equivalence .
then @xmath251 is a simplicial homotopy equivalence and by [ prop - properties ] is a fibration .
then by [ lemb - fib ] @xmath251 is a trivial fibration and by [ prop - reflects ] @xmath252 is a trivial fibration .
finally suppose that @xmath282 is a cofibration which is a weak equivalence .
then @xmath4 can be factorized as @xmath283 where @xmath54 is a strong anodyne extension and @xmath284 is a fibration . by [ lem - weq ] , @xmath284 is a weak equivalence and thus a trivial fibration .
hence @xmath4 is a retract of @xmath54 .
a model category is said to be _ right - proper _ when the class of weak equivalences is stable under pullback along fibrations .
@xmath0 equipped with the kan - quillen model structure is well known satisfy both this condition and its dual , left - properness . as may be found in the account @xcite , left - properness follows rather formally from the fact that every simplicial set is cofibrant , and right - properness follows formally from the excellent properties of the @xmath127 functor . in this appendix
i propose a new elementary and direct proof of right - properness , which in fact leads us to a slightly stronger formulation .
since trivial fibrations are stable under pullback along any map , it suffices to show that strong anodyne extensions are stable under pullback along fibrations .
pullbacks in @xmath0 commute with colimits , so it will suffice to check : [ thm - proper ] the pullback of a horn inclusion along a fibration is a _ strong _ anodyne extension .
the inclusion of the word `` strong '' is the strengthened result not afforded to us by the traditional proofs of right - properness .
for the rest of the paper , we fix the following pullback square @xmath285 ( a ) -- node[auto , swap ] { $ m$ } ( b ) ; \draw[->,font=\footnotesize ] ( b ) -- node[auto , swap ] { $ f$ } ( t ) ; \draw[- > ] ( a ) -- ( h ) ; \draw[- > ] ( h ) -- ( t ) ; \end{tikzpicture}\ ] ] in which @xmath252 is a fibration . we shall exhibit a p - structure for @xmath4 .
we introduce some notational conventions that will allow us to greatly simplify the simplicial identities .
let @xmath286 be some fixed countable dense totally ordered set without greatest or least elements .
we shall refer to the elements of @xmath286 as _ names_. given a non - empty finite subset @xmath287 considered as a poset , let @xmath288 denote the nerve of @xmath289 .
note that @xmath290 in a unique way .
given such an @xmath289 and a simplicial set @xmath1 , an _
@xmath289-simplex _ of @xmath1 is a map @xmath291 .
the set of @xmath289-simplices is denoted @xmath292 clearly @xmath293 .
given such an @xmath289 and @xmath1 , if @xmath294 denote by @xmath295 that function @xmath296 induced by precomposition with the nerve of the poset inclusion @xmath297 .
if we have such an @xmath289 and @xmath1 , then for any @xmath298 such that the unique isomorphisms @xmath299 map @xmath300 , @xmath301 and @xmath302 to each other , and furthermore @xmath303 in @xmath286 , denote by @xmath304 that function @xmath305 induced by precomposing with the nerve of the unique poset surjection @xmath306 mapping @xmath301 and @xmath302 to @xmath300 . finally , given an @xmath289-simplex @xmath291 , for @xmath294 we give the evident meaning to the _
@xmath300^th^ vertex _ @xmath307 of @xmath24 .
now observe that for any @xmath308 ( with @xmath309 ) and @xmath310 , we have @xmath311 . as we go through the proof of [ thm - proper ] we shall silently assume that simplices come with a suitable choice of names . given a simplex @xmath312 , define the _ profile _ of @xmath24 to be the pair @xmath313 where @xmath314 is the dimension of @xmath24 and @xmath315 is the number of vertices ( counted with multiplicity ) over the vertex @xmath90 of @xmath131 , i.e. the number of names @xmath294 such that @xmath316 .
( recall that we have fixed a horn inclusion @xmath317 ) .
we consider profiles to be ordered lexicographically . in order to define a p - structure on @xmath4
, we will need to define something bigger first . for each @xmath313-simplex @xmath24 in @xmath7 but not in @xmath318
, we will assign an @xmath319-simplex @xmath320 which has @xmath24 as a face .
if we assume that @xmath24 is given as an @xmath289-simplex , then for some fresh name @xmath33 we will give @xmath320 as an @xmath321 simplex , where @xmath322 , the @xmath33^th^ vertex @xmath323 lies over @xmath90 and moreover @xmath33 is the greater than all of the names in @xmath289 whose corresponding vertices in @xmath24 lie over @xmath90 .
we will indicate the choice of name @xmath33 by writing @xmath324 .
it is also important that the following holds : if @xmath300 is any name in @xmath24 whose vertex lies over @xmath325 , then @xmath326 .
[ def - qarent ] the following clauses define @xmath324 by recursion over profiles , where @xmath24 is a ( not necessarily non - degenerate ) @xmath289-simplex of @xmath7 not in @xmath318 : * if @xmath327 , then @xmath328 * if @xmath329 , then @xmath330 * and otherwise , if @xmath24 is non - degenerate and not already in the image of @xmath331 : let @xmath332 be the set of names in @xmath24 whose vertices lie over @xmath90 ; then @xmath333 is a compatible family in @xmath7 ; since @xmath252 is a fibration there is a filler ; let this be @xmath334 .
[ lem - qarent - faces ] any partially defined @xmath331 satisfying the recurrence relations in [ def - qarent ] has the property that for any @xmath289-simplex @xmath24 in @xmath7 but not in @xmath318 , if @xmath300 is a name in @xmath24 whose vertex lies over @xmath325 then @xmath335 easy inspection . to complete definition [ def - qarent ]
we need to check that the rules do indeed give rise to a well - defined ( up to choice of fillers for the fibration @xmath252 ) function .
thus we must check that the rules give the same output for any @xmath24 which is degenerate in two different ways , or which is both degenerate and in the image of @xmath331 .
the following lemma deals with the first case .
let @xmath1 be a simplicial set , @xmath287 , @xmath336 with @xmath337 and @xmath338 , @xmath24 an @xmath339-simplex and @xmath26 an @xmath340-simplex .
suppose @xmath341 .
then either @xmath342 and @xmath343 , or else there exists a unique @xmath344-simplex @xmath32 with @xmath345 and @xmath346 . easy calculation . the following lemma deals with the second case .
suppose we have defined @xmath331 for some initial segment of all profiles .
if @xmath347 are two presentations for the same @xmath289-simplex , then @xmath348 , i.e. @xmath349 of that simplex is well - defined .
we can not have @xmath350 since , in @xmath324 , @xmath33 is the greatest name whose vertex lies over @xmath325 . if @xmath351 , then @xmath352 , so @xmath353 , so @xmath354 .
hence @xmath355 , from which the result follows .
alternatively , if @xmath356 , then @xmath357 , and the result follows easily .
we are now in a position to give the p - structure on @xmath4 .
a non - degenerate simplex @xmath358 is of type i if and only if it is in the image of @xmath331 . for a type ii simplex @xmath26
, we define @xmath359 .
we need to check the conditions of [ def - structure ] .
we clearly have a partition and a bijection .
if @xmath24 is type ii then @xmath360 .
suppose @xmath361 for some name @xmath362 .
then @xmath300 is a name in @xmath24 lying over @xmath325 .
so by [ lem - qarent - faces ] , @xmath363 , a contradiction , since @xmath24 is type ii and therefore not in the image of @xmath331 .
the construction clearly gives us that @xmath364 .
the ancestral relation is well - founded since indeed every type ii simplex has only finitely many ancestors .
we proceed by induction on the profile of a type ii simplex @xmath24 .
if @xmath300 is a name in @xmath24 lying over @xmath325 , then @xmath365 has the same ancestors as @xmath366 , which has profile smaller than @xmath24 .
if @xmath301 is a name in @xmath24 not lying over @xmath325 , then @xmath367 has smaller profile than @xmath24 .
hence by [ prop - structure - to - presentation ] , @xmath4 is a strong anodyne extension .
thanks go to my supervisor , martin hyland , for his support and encouragement , and to zhen lin low for many helpful conversations . | by careful analysis of the comparison map from a simplicial set to its image under kan s ex - infinity functor we obtain a new and combinatorial proof that it is a weak homotopy equivalence .
moreover , we obtain a presentation of it as a _
strong anodyne extension_. from this description we are able to quickly deduce some basic facts about ex - infinity and hence provide a new construction of the kan - quillen model structure on simplicial sets , which avoids the use of topological spaces or minimal fibrations . |
supervised classification concerns the task of assigning an object ( or a number of objects ) to one of two or more groups , based on a sample of labelled training data .
the problem was first studied in generality in the famous work of @xcite , where he introduced some of the ideas of linear discriminant analysis ( lda ) , and applied them to his iris data set . nowadays ,
classification problems arise in a plethora of applications , including spam filtering , fraud detection , medical diagnoses , market research , natural language processing and many others .
in fact , lda is still widely used today , and underpins many other modern classifiers ; see , for example , @xcite and @xcite .
alternative techniques include support vector machines @xcite , tree classifiers @xcite , kernel methods @xcite and nearest neighbour classifiers @xcite .
more substantial overviews and in - depth discussion of these techniques , and others , can be found in @xcite and @xcite .
an increasing number of modern classification problems are _ high - dimensional _ , in the sense that the dimension @xmath0 of the feature vectors may be comparable to or even greater than the number of training data points , @xmath1 . in such settings , classical methods such as those mentioned in the previous paragraph tend to perform poorly @xcite , and may even be intractable ; for example , this is the case for lda , where the problems are caused by the fact that the sample covariance matrix is not invertible when @xmath2 .
many methods proposed to overcome such problems assume that the optimal decision boundary between the classes is linear , e.g. @xcite and @xcite .
another common approach assumes that only a small subset of features are relevant for classification .
examples of works that impose such a sparsity condition include @xcite , where it is also assumed that the features are independent , as well as @xcite and @xcite , where soft thresholding is used to obtain a sparse boundary .
more recently , @xcite and @xcite both solve an optimisation problem similar to fisher s linear discriminant , with the addition of an @xmath3 penalty term to encourage sparsity .
in this paper we attempt to avoid the curse of dimensionality by projecting the feature vectors at random into a lower - dimensional space .
the use of random projections in high - dimensional statistical problems is motivated by the celebrated johnson
lindenstrauss lemma ( e.g. * ? ? ?
this lemma states that , given @xmath4 , @xmath5 and @xmath6 , there exists a linear map @xmath7 such that @xmath8 for all @xmath9 . in fact , the function @xmath10 that nearly preserves the pairwise distances can be found in randomised polynomial time using random projections distributed according to haar measure as described in section [ sec chooserp ] below .
it is interesting to note that the lower bound on @xmath11 in the johnson
lindenstrauss lemma does not depend on @xmath0 .
as a result , random projections have been used successfully as a computational time saver : when @xmath0 is large compared to @xmath12 , one may project the data at random into a lower - dimensional space and run the statistical procedure on the projected data , potentially making great computational savings , while achieving comparable or even improved statistical performance .
as one example of the above strategy , @xcite obtained vapnik
chervonenkis type bounds on the generalisation error of a linear classifier trained on a single random projection of the data .
see also @xcite , @xcite and @xcite for other instances .
other works have sought to reap the benefits of aggregating over many random projections .
for instance , @xcite considered estimating a @xmath13 population inverse covariance ( precision ) matrix using @xmath14 , where @xmath15 denotes the sample covariance matrix and @xmath16 are random projections from @xmath17 to @xmath18 .
@xcite used this estimate when testing for a difference between two gaussian population means in high dimensions , while @xcite applied the same technique in fisher s linear discriminant for a high - dimensional classification problem .
our proposed methodology for high - dimensional classification has some similarities with the techniques described above , in the sense that we consider many random projections of the data , but is also closely related to _ bagging _
@xcite , since the ultimate assignment of each test point is made by aggregation and a vote .
bagging has proved to be an effective tool for improving unstable classifiers ; indeed , a bagged version of the ( generally inconsistent ) @xmath19-nearest neighbour classifier is universally consistent as long as the resample size is carefully chosen ; see @xcite .
more generally , bagging has been shown to be particularly effective in high - dimensional problems such as variable selection @xcite .
another related approach to ours is @xcite , who consider ensembles of random rotations , as opposed to projections . from model 2 in section
[ sec tsims ] with @xmath20 dimensions and prior probability @xmath21 .
top row : three projections drawn from haar measure ; bottom row : the projections with smallest estimate of test error out of 100 haar projections with lda ( left ) , quadratic discriminant analysis ( middle ) and @xmath22-nearest neighbours ( right ) . ] one of the basic but fundamental observations that underpins our proposal is the fact that aggregating the classifications of all random projections is not sensible , since most of these projections will typically destroy the class structure in the data ; see the top row of figure [ fig : useless ] . for this reason
, we advocate partitioning the projections into non - overlapping blocks , and within each block we retain only the projection yielding the smallest estimate of the test error .
the attraction of this strategy is illustrated in the bottom row of figure [ fig : useless ] , where we see a much clearer partition of the classes .
another key feature of our proposal is the realisation that a simple majority vote of the classifications based on the retained projections can be highly suboptimal ; instead , we argue that the voting threshold should be chosen in a data - driven fashion in an attempt to minimise the test error of the infinite - simulation version of our random projection ensemble classifier .
in fact , this estimate of the optimal threshold turns out to be remarkably effective in practice ; see section [ sec alpha ] for further details .
we emphasise that our methodology can be used in conjunction with any base classifier , though we particularly have in mind classifiers designed for use in low - dimensional settings .
the random projection ensemble classifier can therefore be regarded as a general technique for either extending the applicability of an existing classifier to high dimensions , or improving its performance .
our theoretical results are divided into three parts . in the first
, we consider a generic base classifier and a generic method for generating the random projections into @xmath18 and quantify the difference between the test error of the random projection ensemble classifier and its infinite - simulation counterpart as the number of projections increases .
we then consider selecting random projections from non - overlapping blocks by initially drawing them according to haar measure , and then within each block retaining the projection that minimises an estimate of the test error . under a condition implied by the widely - used sufficient dimension reduction assumption @xcite , we can then control the difference between the test error of the random projection classifier and the bayes risk as a function of terms that depend on the performance of the base classifier based on projected data and our method for estimating the test error , as well as terms that become negligible as the number of projections increases .
the final part of our theory gives risk bounds on the first two of these terms for specific choices of base classifier , namely fisher s linear discriminant and the @xmath22-nearest neighbour classifier .
the key point here is that these bounds only depend on @xmath11 , the sample size @xmath1 and the number of projections , and not on the original data dimension @xmath0 .
the remainder of the paper is organised as follows .
our methodology and general theory are developed in sections [ sec rp ] and [ sec chooserp ] .
specific choices of base classifier are discussed in section [ sec base ] , while section [ sec prac ] is devoted to a consideration of the practical issues of the choice of voting threshold , projected dimension and the number of projections used . in section [ sec
empirical ] we present results from an extensive empirical analysis on both simulated and real data where we compare the performance of the random projection ensemble classifier with several popular techniques for high - dimensional classification .
the outcomes are extremely encouraging , and suggest that the random projection ensemble classifier has excellent finite - sample performance in a variety of different high - dimensional classification settings .
we conclude with a discussion of various extensions and open problems .
all proofs are deferred to the appendix . finally in this section , we introduce the following general notation used throughout the paper . for a sufficiently smooth real - valued function @xmath23 defined on a neighbourhood of @xmath24 , let @xmath25 and @xmath26 denote its first and second derivatives at @xmath27 , and
let @xmath28 and @xmath29 denote the integer and fractional part of @xmath27 respectively .
we start by describing our setting and defining the relevant notation .
suppose that the pair @xmath30 takes values in @xmath31 , with joint distribution @xmath32 , characterised by @xmath33 , and @xmath34 , the conditional distribution of @xmath35 , for @xmath36 . for convenience ,
we let @xmath37 . in the alternative characterisation of @xmath32 ,
we let @xmath38 denote the marginal distribution of @xmath39 and write @xmath40 for the regression function . recall that a _ classifier _ on @xmath17 is a borel measurable function @xmath41 , with the interpretation that we assign a point @xmath42 to class @xmath43 .
we let @xmath44 denote the set of all such classifiers . the misclassification rate , or _ risk _ , of a classifier @xmath45 is @xmath46 , and is minimised by the _ bayes _
classifier @xmath47 ( e.g. * ? ? ?
its risk is @xmath48 $ ] . of course
, we can not use the bayes classifier in practice , since @xmath49 is unknown . nevertheless , we often have access to a sample of training data that we can use to construct an approximation to the bayes classifier . throughout this section and section [ sec chooserp ] ,
it is convenient to consider the training sample @xmath50 to be fixed points in @xmath31 .
our methodology will be applied to a base classifier @xmath51 , which we assume can be constructed from an arbitrary training sample @xmath52 of size @xmath1 in @xmath53 ; thus @xmath54 is a measurable function from @xmath55 to @xmath56 .
now assume that @xmath57 .
we say a matrix @xmath58 is a _ projection _ if @xmath59 , the @xmath11-dimensional identity matrix .
let @xmath60 be the set of all such matrices . given a projection @xmath61 ,
define projected data @xmath62 and @xmath63 for @xmath64 , and let @xmath65 .
the projected data base classifier corresponding to @xmath54 is @xmath66 , given by @xmath67 note that although @xmath68 is a classifier on @xmath17 , the value of @xmath69 only depends on @xmath70 through its @xmath11-dimensional projection @xmath71 .
we now define a generic ensemble classifier based on random projections . for @xmath72 ,
let @xmath73 denote independent and identically distributed projections in @xmath74 , independent of @xmath30 .
the distribution on @xmath74 is left unspecified at this stage , and in fact our proposed method ultimately involves choosing this distribution depending on @xmath75 .
now set @xmath76 for @xmath77 , the _ random projection ensemble _ classifier is defined to be @xmath78 we emphasise again here the additional flexibility afforded by not pre - specifying the voting threshold @xmath79 to be @xmath80 .
our analysis of the random projection ensemble classifier will require some further definitions .
let @xmath81 where the randomness here comes from the random projections .
let @xmath82 and @xmath83 denote the distribution functions of @xmath84 and @xmath85 , respectively .
we will make use of the following assumption : * ( a.1 ) * : : @xmath82 and @xmath83 are twice differentiable at @xmath79 .
the first derivatives of @xmath82 and @xmath83 , when they exist , are denoted as @xmath86 and @xmath87 respectively ; under * ( a.1 ) * , these derivatives are well - defined in a neighbourhood of @xmath79 .
our first main result below gives an asymptotic expansion for the test error @xmath88 of our generic random projection ensemble classifier as the number of projections increases .
in particular , we show that this test error can be well approximated by the test error of the infinite - simulation random projection classifier @xmath89 this infinite - simulation classifier turns out to be easier to analyse in subsequent results .
note that under * ( a.1 ) * , @xmath90 [ thm bvar ] assume * ( a.1)*. then @xmath91 as @xmath92 , where @xmath93 the proof of theorem [ thm bvar ] in the appendix is lengthy , and involves a one - term edgeworth approximation to the distribution function of a standardised binomial random variable .
one of the technical challenges is to show that the error in this approximation holds uniformly in the binomial proportion .
define the test error of @xmath68 by through an integral rather than defining @xmath94 to make it clear that when @xmath95 is a random projection , it should be conditioned on when computing @xmath96 . ]
@xmath97 our next result controls the test excess risk , i.e. the difference between the test error and the bayes risk , of the infinite - simulation random projection classifier in terms of the expected test excess risk of the classifier based on a single random projection .
an attractive feature of this result is its generality : no assumptions are placed on the configuration of the training data @xmath75 , the distribution @xmath32 of the test point @xmath30 or on the distribution of the individual projections .
[ thm : baggingbound ] we have @xmath98
in this section , we study a special case of the generic random projection ensemble classifier introduced in section [ sec rp ] , where we propose a screening method for choosing the random projections .
let @xmath99 be an estimator of @xmath96 , based on @xmath100 , that takes values in the set @xmath101 .
examples of such estimators include resubstitution and leave - one - out estimates ; we discuss these choices in greater detail in section [ sec base ] . for @xmath102 ,
let @xmath103 denote independent projections , independent of @xmath30 , distributed according to haar measure on @xmath74 .
one way to simulate from haar measure on the set @xmath74 is to first generate a matrix @xmath104 , where each entry is drawn independently from a standard normal distribution , and then take @xmath105 to be the matrix of left singular vectors in the singular value decomposition of @xmath106 . for @xmath107 , let @xmath108 where @xmath109 denotes the smallest index where the minimum is attained in the case of a tie .
we now set @xmath110 , and consider the random projection ensemble classifier from section [ sec rp ] constructed using the independent projections @xmath111 .
let @xmath112 denote the optimal test error estimate over all projections .
the minimum is attained here , since @xmath99 takes only finitely many values . for @xmath113 ,
let @xmath114 where @xmath95 is distributed according to haar measure on @xmath74 .
we assume the following : * ( a.2 ) * : : there exist @xmath115 and @xmath116 such that @xmath117 for @xmath118 . condition * ( a.2 ) * asks for a certain growth rate of the distribution function of @xmath99 close to its minimum value @xmath119 ; observe that the strength of the condition decreases as @xmath120 increases . under this condition , the following result is a starting point for controlling the expected test excess risk of the classifier based on a single projection chosen according to the scheme described above .
[ thm ela * ] assume * ( a.2)*. then @xmath121 where @xmath122 .
the form of the bound in proposition [ thm ela * ] motivates us to seek to control @xmath123 in terms of the test excess risk of a classifier based on the projected data , in the hope that we will be able to show this does not depend on @xmath0 . to this end , define the regression function on @xmath18 induced by the projection @xmath61 to be @xmath124 .
the corresponding induced bayes classifier , which is the optimal classifier knowing only the distribution of @xmath125 , is given by @xmath126 its risk is @xmath127 in order to ensure that @xmath119 will be close to the bayes risk , we will invoke an additional assumption on the form of the bayes classifier : * ( a.3 ) * : : there exists a projection @xmath128 such that @xmath129 where @xmath130 denotes the symmetric difference of two sets @xmath131 and @xmath45 .
condition * ( a.3 ) * requires that the set of points @xmath132 assigned by the bayes classifier to class 1 can be expressed as a function of a @xmath11-dimensional projection of @xmath70 .
note that if the bayes decision boundary is a hyperplane , then * ( a.3 ) * holds with @xmath133 .
moreover , proposition [ prop : cond ] below shows that , in fact , * ( a.3 ) * holds under the sufficient dimension reduction condition , which states that @xmath134 is independent of @xmath39 given @xmath135 ; see @xcite for many statistical settings where such an assumption is natural .
[ prop : cond ] if @xmath134 is independent of @xmath39 given @xmath135 , then * ( a.3 ) * holds . finally , then , we are in a position to control the test excess risk of our random projection ensemble classifier in terms of the test excess risk of a classifier based on @xmath11-dimensional data , as well as terms that reflect our ability to estimate the test error of classifiers based on projected data and terms that depend on @xmath136 and @xmath120 .
[ thm : main ] assume * ( a.1 ) * , * ( a.2 ) * and * ( a.3)*. then @xmath137 as @xmath138 , where @xmath139 is defined in theorem [ thm bvar ] , @xmath140 is defined in proposition [ thm ela * ] and @xmath141 . regarding the bound in theorem [ thm : main ] as a sum of four terms , we see that the last two of these can be seen as the price we have to pay for the fact that we do not have access to an infinite sample of random projections .
these terms can be made negligible by choosing @xmath136 and @xmath120 to be sufficiently large , but it should be noted that @xmath140 may increase with @xmath120 .
this is a reflection of the fact that minimising an estimate of test error may lead to overfitting .
the behaviour of this term , together with that of @xmath142 and @xmath143 , depends on the choice of base classifier , but in the next section below we describe settings where these terms can be bounded by expressions that do not depend on @xmath0 .
in this section , we change our previous perspective and regard the training data @xmath144 as independent random pairs with distribution @xmath32 , so our earlier statements are interpreted conditionally on @xmath75 .
we consider particular choices of base classifier , and study the first two terms in the bound in theorem [ thm : main ] .
linear discriminant analysis ( lda ) , introduced by @xcite , is arguably the simplest classification technique .
recall that in the special case where @xmath145 , we have @xmath146 so * ( a.3 ) * holds with @xmath133 and @xmath147 , a @xmath148 matrix . in lda , @xmath149 , @xmath150 and @xmath151
are estimated by their sample versions , using a pooled estimate of @xmath151 .
although lda can not be applied directly when @xmath2 since the sample covariance matrix is singular , we can still use it as the base classifier for a random projection ensemble , provided that @xmath152 . indeed , noting that for any @xmath61 , we have @xmath153 , where @xmath154 and @xmath155 , we can define@xmath156 here , @xmath157 , @xmath158 , @xmath159 and @xmath160 .
write @xmath161 for the standard normal distribution function . under the normal model specified above
, the test error of the lda classifier can be written as @xmath162 where @xmath163 and @xmath164 .
@xcite studied the excess risk of the lda classifier in an asymptotic regime in which @xmath11 is fixed as @xmath1 diverges .
in fact , he considered a very slightly different data generating model , in which the training sample sizes from each population are assumed to be known in advance , so that without loss of generality , we may assume that @xmath165 and @xmath166 , while @xmath167 , as before .
specialising his results for simplicity to the case where @xmath1 is even and @xmath168 , @xcite showed that using the lda classifier with @xmath169 , @xmath170 and @xmath171 yields @xmath172 as @xmath173 , where @xmath174 .
it remains to control the errors @xmath140 and @xmath143 in theorem [ thm : main ] .
for the lda classifier , we consider the resubstitution estimator @xmath175 @xcite provided a vapnik chervonenkis bound for @xmath99 under no assumptions on the underlying data generating mechanism : for every @xmath176 and @xmath177 , @xmath178 see also ( * ? ? ?
* theorem 23.1 ) . we can then conclude that @xmath179 the more difficult term to deal with is @xmath180 . in this case , the bound can not be applied directly , because @xmath181 are no longer independent conditional on @xmath182 .
nevertheless , since @xmath183 are independent of @xmath75 , we still have that @xmath184 we can therefore conclude by almost the same argument as that leading to that @xmath185 note that none of the bounds , and depend on the original data dimension @xmath0 .
moreover , substituting the bound into theorem [ thm : main ] reveals a trade - off in the choice of @xmath120 when using lda as the base classifier .
choosing @xmath120 to be large gives us a good chance of finding a projection with a small estimate of test error , but we may incur a small overfitting penalty as reflected by .
quadratic discriminant analysis ( qda ) is designed to handle situations where the class - conditional covariance matrices are unequal .
recall that when @xmath186 , and @xmath187 , for @xmath188 , the bayes decision boundary is given by @xmath189 , where @xmath190 in qda , @xmath149 , @xmath150 and @xmath191 are estimated by their sample versions . if @xmath192 , where @xmath193 is the number of training sample observations from the @xmath194th class , then at least one of the sample covariance matrix estimates is singular , and qda can not be used directly .
nevertheless , we can still choose @xmath195 and use qda as the base classifier in a random projection ensemble .
specifically , we can set @xmath196 where @xmath197 , @xmath198 and @xmath199 were defined in section [ sec lda ] , and where @xmath200 for @xmath188 .
unfortunately , analogous theory to that presented in section [ sec lda ] does not appear to exist for the qda classifier ( unlike for lda , the risk does not have a closed form ) .
nevertheless , we found in our simulations presented in section [ sec empirical ] that the qda random projection ensemble classifier can perform very well in practice . in this case
, we estimate the test errors using the leave - one - out method given by @xmath201 where @xmath202 denotes the classifier @xmath203 , trained without the @xmath204th pair , i.e. based on @xmath205 . for a method like qda that involves estimating more parameters than lda
, we found that the leave - one - out estimator was less susceptible to overfitting than the resubstitution estimator .
the @xmath22-nearest neighbour classifier ( @xmath22nn ) , first proposed by @xcite , is a nonparametric method that classifies the test point @xmath132 according to a majority vote over the classes of the @xmath22 nearest training data points to @xmath70 .
the enormous popularity of the @xmath22nn classifier can be attributed partly due to its simplicity and intuitive appeal ; however , it also has the attractive property of being universally consistent : for every distribution @xmath32 , as long as @xmath206 and @xmath207 , the risk of the @xmath22nn classifier converges to the bayes risk ( * ? ? ?
* theorem 6.4 ) .
@xcite derived the rate of convergence of the excess risk of the @xmath22-nearest neighbour classifier . under regularity conditions ,
the optimal choice of @xmath22 , in terms of minimising the excess risk , is @xmath208 , and the rate of convergence of the excess risk with this choice is @xmath209 .
thus , in common with other nonparametric methods , there is a ` curse of dimensionality ' effect that means the classifier typically performs poorly in high dimensions .
@xcite found the optimal way of assigning decreasing weights to increasingly distant neighbours , and quantified the asymptotic improvement in risk over the unweighted version , but the rate of convergence remains the same .
we can use the @xmath22nn classifier as the base classifier for a random projection ensemble as follows : let @xmath210 , where @xmath211 and @xmath212 .
given @xmath213 , let @xmath214 be a re - ordering of the training data such that @xmath215 , with ties split at random .
now define @xmath216 where @xmath217 . the theory described in the previous paragraph
can be applied to show that , under regularity conditions , @xmath218 .
once again , a natural estimate of the test error in this case is the leave - one - out estimator defined in , where we use the same value of @xmath22 on the leave - one - out datasets as for the base classifier , and where distance ties are split in the same way as for the base classifier . for this estimator
, @xcite showed that for every @xmath176 , @xmath219 see also ( * ? ? ? * chapter 24 ) .
it follows that @xmath220 @xcite also provided a tail bound analogous to for the leave - one - out estimator .
arguing very similarly to section [ sec lda ] , we can deduce that under no conditions on the data generating mechanism , @xmath221
we now discuss the choice of the voting threshold @xmath79 in ( [ eq rp ] ) .
the expression for the test error of the infinite - simulation random projection ensemble classifier given in suggests the ` oracle ' choice @xmath222 } \bigl[\pi_1 g_{n,1}(\alpha ' ) + \pi_2\{1 - g_{n,2}(\alpha')\}\bigr].\ ] ] note that , if assumption * ( a.1 ) * holds and @xmath223 then @xmath224 and in theorem [ thm bvar ] , we have @xmath225 of course , @xmath226 can not be used directly , because we do not know @xmath82 and @xmath83 ( and we may not know @xmath227 and @xmath228 either ) . nevertheless ,
for the lda base classifier we can estimate @xmath229 using @xmath230 for @xmath188 .
for the qda and @xmath22-nearest neighbour base classifiers , we use the leave - one - out - based estimate @xmath231 in place of @xmath232 .
we also estimate @xmath149 by @xmath233 , and then set the cut - off in ( [ eq rp ] ) as @xmath234 } \bigl[\hat{\pi}_1 \hat{g}_{n,1}(\alpha ' ) + \hat{\pi}_2\{1 - \hat{g}_{n,2}(\alpha')\}\bigr].\ ] ] since empirical distribution functions are piecewise constant , the objective function in does not have a unique minimum , so we choose @xmath235 to be the average of the smallest and largest minimisers . an attractive feature of the method is that @xmath236 ( or @xmath237 in the case of qda and @xmath22nn ) are already calculated in order to choose the projections , so calculating @xmath238 and @xmath239 carries negligible extra computational cost
. figures [ fig gplots0.5 ] and [ fig gplots0.33 ] illustrate @xmath240 as an estimator of @xmath241 , for different base classifiers as well as different values of @xmath1 and @xmath227 . here , a very good approximation to the estimand was obtained using an independent data set of size 5000 . unsurprisingly
, the performance of the estimator improves as @xmath1 increases , but the most notable feature of these plots is the fact that for all classifiers and even for small sample sizes , @xmath235 is an excellent estimator of @xmath226 , and may be a substantial improvement on the naive choice @xmath242 ( which may result in a classifier that assigns every point to a single class ) .
in ( black ) and @xmath240 ( red ) for the lda ( left ) , qda ( middle ) and @xmath22nn ( right ) base classifiers after projecting for one training data set of size @xmath243 50 ( top ) , 100 ( middle ) and 1000 ( bottom ) from model 1 . here , @xmath244 , @xmath245 and @xmath246 .
] in ( black ) and @xmath240 ( red ) for the lda ( left ) , qda ( middle ) and @xmath22nn ( right ) base classifiers after projecting for one training data set of size @xmath243 50 ( top ) , 100 ( middle ) and 1000 ( bottom ) from model 1 .
here , @xmath247 , @xmath245 and @xmath246 . ] we want to choose @xmath11 as small as possible in order to obtain the best possible performance bounds as described in section [ sec base ] above .
this also reduces the computational cost .
the performance bounds rely on assumption * ( a.3 ) * , whose strength decreases as @xmath11 increases , so we want to choose @xmath11 large enough that this condition holds ( at least approximately ) .
in section [ sec empirical ] we see that the random ensemble projection method is quite robust to the choice of @xmath11
. nevertheless , in some circumstances it may be desirable to have an automatic choice .
as one way to do this , suppose that we wish to choose @xmath11 from a set @xmath248 .
for each @xmath249 , generate independent and identically distributed projections @xmath250 according to haar measure on @xmath251 .
for each @xmath249 and for @xmath252 , we can then set @xmath253 where @xmath254 . finally , we can select @xmath255 in figures [ fig empiricaldist1d ] and [ fig empiricaldist2d ] we present the empirical distribution functions of @xmath256 , where @xmath257 , for one training dataset from model 1 ( described in section [ sec independent ] ) , and model 3 ( described in section [ sec multisims ] ) . in each case
, we set @xmath258 , @xmath259 , @xmath20 and @xmath260 . for the lda ( left ) , qda ( middle ) and @xmath22nn ( right ) base classifiers after projecting for
model 1 , @xmath258 , @xmath259 , @xmath245 , and @xmath246 ( black ) , 3 ( red ) , 4 ( blue ) and 5 ( green).,scaledwidth=100.0% ] for the lda ( left ) , qda ( middle ) and @xmath22nn ( right ) base classifiers after projecting for model 2 , @xmath258 , @xmath259 , @xmath245 , and @xmath246 ( black ) , 3 ( red ) , 4 ( blue ) and 5 ( green).,scaledwidth=100.0% ] figures [ fig empiricaldist1d ] and [ fig empiricaldist2d ] do not suggest great differences in performance for different choices of @xmath11 , especially for the qda and @xmath22nn base classifiers .
for the lda classifier , it appears , particularly for model 2 , that projecting into a slightly larger dimensional space is preferable , and indeed this appears to be the case from the relevant entry of table [ tab tdist ] below .
the ideas presented here may also be used to decide between two different base classifiers .
for example , comparing the green lines across different panels of figure [ fig empiricaldist1d ] , we see that for model 1 and @xmath261 , we might expect the best results with the qda base classifier , and indeed this is confirmed by the simulation results in table [ tab indep ] below . in order to minimise the third term in the bound in theorem [ thm : main ]
, we should choose @xmath136 to be as large as possible .
the constraint , of course , is that the computational cost of the random projection classifier scales linearly with @xmath136 .
the choice of @xmath120 is more subtle ; while the fourth term in the bound in theorem [ thm : main ] decreases as @xmath120 increases , we saw in section [ sec base ] that upper bounds on @xmath262 may increase with @xmath120 . in principle , we could try to use the expressions given in theorem [ thm : main ] and section [ sec base ] to choose @xmath120 to minimise the overall upper bound on @xmath263 . although @xmath264 and @xmath265 are unknown to the practitioner , these could be estimated based on the empirical distribution of @xmath266 , where @xmath267 are independent projections drawn according to haar measure . in practice
, however , we found that an involved approach such as this was unnecessary , and that the ensemble method was robust to the choice of @xmath136 and @xmath120 . in all of our simulations , we set @xmath260 .
in this section , we assess the empirical performance of the random projection ensemble classifier . throughout this section , rp-@xmath22nn@xmath268 , rp - lda@xmath268 and rp - qda@xmath268 denote the random projection classifier with lda , qda and @xmath22nn base classifiers , respectively ; the subscript @xmath11 refers to the dimension of the image space of the projections . for comparison , we also present results for several state - of - the - art methods for high - dimensional classification , namely penalized lda @xcite , nearest shrunken centroids @xcite , shrunken centroids regularized discriminant analysis @xcite , and independence rules ( ir ) @xcite , as well as for the base classifier applied in the original space . for the standard @xmath22nn classifier
, we chose @xmath22 via leave - one - out cross validation .
the tuning parameters for the other methods were chosen using the default settings in the corresponding ` r ` packages ` penlda ` @xcite , ` nsc ` @xcite and ` scrda ` @xcite , namely 6-fold , 10-fold and 10-fold cross validation , respectively . in each of the simulated examples below
, we used @xmath269 , @xmath20 , two different values of the prior probability @xmath227 and @xmath270 .
each experiment was repeated 100 times and the risk was estimated on an independent test set of size 1000 .
model 1 : here , @xmath271 is the distribution of @xmath0 independent components , each with a standard laplace distribution , while @xmath272 , with @xmath273 .
.misclassification rates multiplied by 100 ( with standard errors as subscripts ) for model 1 , with smallest in bold .
* n / a : not available due to singular covariance estimates . [ cols=">,^,^,^,^,^,^ " , ] [ tab ozone ] in this example , all of the random projection ensemble classifiers performed very well , though scrda was also competitive , particularly for the largest sample size .
we have introduced a general framework for high - dimensional classification via the combination of the results of applying a base classifier on carefully selected low - dimensional random projections of the data .
one of its attractive features is its generality : the approach can be used in conjunction with any base classifier . moreover , although we explored in detail one method for combining the random projections ( partly because it facilitates rigorous statistical analysis ) , there are many other options available here .
for instance , instead of only retaining the projection within each block yielding the smallest estimate of test error , one might give weights to the different projections , where the weights decrease as the estimate of test error increases .
another interesting avenue to explore would be alternative methods for estimating the test error , such as sample splitting .
the idea here would be to split the sample @xmath75 into @xmath274 and @xmath275 , say , where @xmath276 and @xmath277 .
we then use @xmath278 to estimate the test error @xmath279 based on the training data @xmath274 .
since @xmath274 and @xmath275 are independent , we can apply hoeffding s inequality to deduce that @xmath280 it then follows by very similar arguments to those given in section [ sec lda ] that @xmath281 the advantages of this approach are twofold : first , the bounds hold for any choice of base classifier ( and still without any assumptions on the data generating mechanism ) ; second , the bounds on the terms in merely rely on hoeffding s inequality as opposed to vapnik chervonenkis theory , so are typically sharper .
the disadvantage is that the other terms in the bound in theorem [ thm : main ] will tend to be larger due to the reduced effective sample size .
the choice of @xmath282 and @xmath283 in such an approach therefore becomes an interesting question .
many practical classification problems involve @xmath284 classes .
the main issue in extending our methodology to such settings is the definition of @xmath285 analogous to . to outline one approach ,
let @xmath286 for @xmath287 . given @xmath288 with @xmath289
, we can then define @xmath290 the choice of @xmath291 is analogous to the choice of @xmath79 in the case @xmath292 .
it is therefore natural to seek to minimise the the test error of the corresponding infinite - simulation random projection ensemble classifier as before . in other situations
, it may be advantageous to consider alternative types of projection , perhaps because of additional structure in the problem .
one particularly interesting issue concerns ultrahigh - dimensional settings , say @xmath0 in the thousands . here
, it may be too time - consuming to generate enough random projections to explore adequately the space @xmath251 . as a mathematical quantification of this , the cardinality of an @xmath293-net in the euclidean norm of the surface of the euclidean ball in @xmath17 increases exponentially in @xmath0 ( e.g. * ? ? ? * ; * ? ? ?
in such challenging problems , one might restrict the projections @xmath95 to be axis - aligned , so that each row of @xmath95 consists of a single non - zero component , equal to 1 , and @xmath294 zero components .
there are then only @xmath295 choices for the projections , and if @xmath11 is small , it may be feasible even to carry out an exhaustive search .
of course , this approach loses one of the attractive features of our original proposal , namely the fact that it is equivariant to orthogonal transformations . nevertheless , corresponding theory
can be obtained provided that the projection @xmath296 in * ( a.3 ) * is axis - aligned .
this is a much stronger requirement , but it seems that imposing greater structure is inevitable to obtain good classification in such settings .
recall that the training data @xmath297 are fixed and the projections @xmath298 , are independent and identically distributed in @xmath74 , independent of the pair @xmath30 .
the test error of the random projection ensemble classifier has the following representation : @xmath299 where @xmath300 is defined in .
let @xmath301 , for @xmath302 .
then , conditional on @xmath303 $ ] , the random variables @xmath304 are independent , each having a bernoulli(@xmath305 ) distribution .
recall that @xmath82 and @xmath83 are the distribution functions of @xmath306 and @xmath307 , respectively .
we can therefore write @xmath308 where here and throughout the proof , @xmath309 denotes a @xmath310 random variable .
similarly , @xmath311 it follows that @xmath312 where @xmath313 .
writing @xmath314 , we now show that @xmath315 as @xmath92 .
our proof involves a one - term edgeworth expansion to the binomial distribution function in , where the error term is controlled uniformly in the parameter .
the expansion relies on the following version of esseen s smoothing lemma .
* chapter 2 , theorem 2b ) let @xmath316 , @xmath317 , @xmath318 , let @xmath319 be a non - decreasing function and let @xmath320 be a function of bounded variation .
let @xmath321 and @xmath322 be the fourier
stieltjes transforms of @xmath323 and @xmath324 , respectively .
suppose that * @xmath325 and @xmath326 ; * @xmath327 ; * the set of discontinuities of @xmath323 and @xmath324 is contained in @xmath328 , where @xmath329 is a strictly increasing sequence with @xmath330 ; moreover @xmath323 is constant on the intervals @xmath331 for all @xmath332 ; * @xmath333 for all @xmath334 .
then there exist constants @xmath335 such that @xmath336 provided that @xmath337 .
[ thm esseen ] let @xmath338 , and let @xmath161 and @xmath339 denote the standard normal distribution and density functions , respectively . moreover , for @xmath24 , let @xmath340 and @xmath341 in proposition [ cor edgeworth ] below we apply theorem [ thm esseen ] to the following functions : @xmath342 and @xmath343 [ cor edgeworth ] let @xmath344 and @xmath345 be as in ( [ eq fb1 ] ) and ( [ eq gb1 ] ) .
there exists a constant @xmath346 such that , for all @xmath72 , @xmath347 proposition [ cor edgeworth ] , whose proof is given after the proof of theorem [ thm : main ] , bounds uniformly in @xmath305 the error in the one - term edgeworth expansion @xmath345 of the distribution function @xmath344 . returning to the proof of theorem [ thm bvar ]
, we will argue that the dominant contribution to the integral in ( [ eq rtp ] ) arises from the interval @xmath348,@xmath349 , where @xmath350 .
for the remainder of the proof we assume @xmath136 is large enough that @xmath351 \subseteq ( 0,1)$ ] . for the region @xmath352 , by hoeffding s inequality , we have that @xmath353 for each @xmath354 , as @xmath92 .
it follows that @xmath355 for each @xmath354 , as @xmath92 . for the region @xmath356 , by proposition [ cor edgeworth ]
, there exists @xmath357 such that , for all @xmath136 sufficiently large , @xmath358 where @xmath359 . hence , using the fact that for large @xmath136 , @xmath360 under * ( a.1 ) * , we have @xmath361 as @xmath138 . to aid exposition ,
we will henceforth concentrate on the dominant terms in our expansions , denoting the remainder terms as @xmath362 .
these remainders are then controlled at the end of the argument .
for the first term in ( [ eq main1 ] ) , we write @xmath363 now , for the first term in ( [ eq main2 ] ) , @xmath364 for the second term in ( [ eq main2 ] ) , write @xmath365 returning to the second term in ( [ eq main1 ] ) , observe that @xmath366 the claim will now follow from ( [ eq : hoeffding ] ) , ( [ eq main1 ] ) , ( [ eq main2 ] ) , ( [ eq main3 ] ) , ( [ eq dom3 ] ) and ( [ eq main4 ] ) , once we have shown that @xmath367 as @xmath138 . _ to bound @xmath368 _ : for @xmath369 , let @xmath370 .
observe that , by a taylor expansion about @xmath371 , there exists @xmath372 , such that , for all @xmath373 and all @xmath374 , @xmath375 } |\ddot{h}_\theta(\zeta')| \leq ( \zeta-\alpha)^2\frac{\log^{3}{b_1}}{2\sqrt{2\pi } \{\alpha(1-\alpha)\}^{7/2}}.\end{aligned}\ ] ] using this bound with @xmath376 , we deduce that , for all @xmath136 sufficiently large , @xmath377 as @xmath92 .
_ to bound @xmath378 _
: since @xmath379 is differentiable at @xmath79 , given @xmath380 , there exists @xmath381 such that @xmath382 for all @xmath383 .
it follows that , for all @xmath136 sufficiently large , @xmath384 we deduce that @xmath385 as @xmath92 . _ to bound @xmath386 _ :
for large @xmath136 , we have @xmath387 as @xmath92 . _ to bound @xmath388 _ : by the bound in ( [ eq cont ] ) , we have that , given @xmath177 , for all @xmath136 sufficiently large , @xmath389 _ to bound @xmath390 _ : for all @xmath136 sufficiently large , @xmath391 as @xmath92 . _ to bound @xmath392 _ :
we write @xmath393 where @xmath394 and @xmath395 since @xmath379 is continuous at @xmath79 , given @xmath380 , there exists @xmath396 such that , for all @xmath397 , @xmath398 it follows that , for @xmath397 , @xmath399 for all sufficiently large @xmath136 .
we deduce that @xmath400 as @xmath92 . to control @xmath401 , by the mean value theorem , we have that for all @xmath136 sufficiently large and all @xmath402 $ ] , @xmath403 thus , for large @xmath136 , @xmath404 we deduce that @xmath405 as @xmath92 .
_ to bound @xmath406
_ : write @xmath407 , where @xmath408 and @xmath409 by the bound in ( [ eq cont ] ) , given @xmath410 , for all @xmath136 sufficiently large , latexmath:[\ ] ] similarly , @xmath475 but , for @xmath476 $ ] , @xmath477 it follows that @xmath478 now @xmath479 where @xmath480 . by (
* theorem 2 , section 40 ) , we have that @xmath481 moreover , @xmath482 finally , @xmath483 by ( [ eq ij1 ] ) , ( [ eq : ij ] ) , ( [ eq ij2 ] ) , ( [ eq ij3 ] ) , ( [ eq ij4 ] ) , ( [ eq ij5 ] ) and ( [ eq ij6 ] ) , it follows that @xmath484 by ( [ eq interval ] ) , ( [ eq s12 ] ) , ( [ eq : q * ] ) , ( [ eq : f*middle ] ) , ( [ eq : phi*middle ] ) , ( [ eq : p*middle ] ) , ( [ eq s1s2 ] ) , ( [ eq : phipend ] ) and ( [ eq s2b1 ] ) , we conclude that ( [ eq fourierbound ] ) holds .
the result now follows from theorem [ thm esseen ] , by taking @xmath485 , @xmath486 and @xmath487 in that result .
the research of the second author is supported by an engineering and physical sciences research council fellowship .
the authors thank rajen shah and ming yuan for helpful comments .
bickel , p. j. and levina , e. ( 2004 ) . some theory for fisher s linear discriminant function , ` naive bayes ' , and some alternatives when there are more variables than observations .
_ bernoulli _ , * 10 * , 9891010 .
hastie , t. , tibshirani , r. , and friedman , j. ( 2009 ) . _ the elements of statistical learning : the elements of statistical learning : data mining , inference , and prediction._. springer series in statistics ( 2nd ed . ) .
springer , new york .
tibshirani , r. , hastie , t. , narisimhan , b. and chu , g. ( 2002 ) .
diagnosis of multiple cancer types by shrunken centroids of gene expression .
_ procedures of the natural academy of science , usa _ , * 99 * , 65676572 . | we introduce a very general method for high - dimensional classification , based on careful combination of the results of applying an arbitrary base classifier to random projections of the feature vectors into a lower - dimensional space .
in one special case that we study in detail , the random projections are divided into non - overlapping blocks , and within each block we select the projection yielding the smallest estimate of the test error .
our random projection ensemble classifier then aggregates the results of applying the base classifier on the selected projections , with a data - driven voting threshold to determine the final assignment .
our theoretical results elucidate the effect on performance of increasing the number of projections .
moreover , under a boundary condition implied by the sufficient dimension reduction assumption , we show that the test excess risk of the random projection ensemble classifier can be controlled by terms that do not depend on the original data dimension .
the classifier is also compared empirically with several other popular high - dimensional classifiers via an extensive simulation study , which reveals its excellent finite - sample performance . |
the link between a certain type iip supernova ( sn iip ) and its main - sequence progenitor is poorly known despite the widely accepted view that these supernovae arise from the initial mass range of @xmath5 ( e.g. , * ? ? ?
the primary reasons for the uncertainty in this field are the small number of presupernova ( pre - sn ) mass determinations and the uncertain amount of matter lost through winds , presumably dominated by the red supergiant ( rsg ) wind .
while the former is becoming well - studied , the latter uncertainty is related to our poor knowledge of the complicated physics of mass loss and the unsatisfactory situation with the empirical measurements of mass loss from rsgs . even for the most studied close massive rsg , betelgeuse ,
the range of observational estimates is large : @xmath6 yr@xmath2 @xcite .
the situation for sne iip has some promise because the dominant mass loss at the rsg stage may be observed by the detection of circumstellar ( cs ) interaction at x - ray and radio wavelengths .
the type iip sn 1999em was detected in x - rays with _ chandra _
@xcite , leading to the mass loss estimate @xmath7 yr@xmath2 ( assuming a wind velocity of 10 km s@xmath2 ) . a recent study of available x - ray and radio data for sne iip produced a range of mass loss rates of pre - sn iip @xmath8 yr@xmath2 @xcite .
the application of these rates to the full rsg stage ( @xmath9 yr ) suggests the loss of @xmath10 for pre - sne iip .
the wide range of estimates emphasizes the need for the individual determination of the wind density for each particular sn iip under consideration .
unfortunately , this is not always possible since x - ray and radio observations of sne iip are often not available . here ,
we propose two new diagnostics for the wind density in sne iip that could help .
both rely on spectroscopic observations of h@xmath0 and hei 10830 at the photospheric epoch .
the first one is based on the fact that the interaction of sn ejecta with the wind results in the emission of x - rays from both forward and reverse shocks .
the x - rays cause ionization and excitation of sn ejecta that may be revealed , e.g. , through specific emission lines @xcite .
unfortunately , in sne iip the wind density is low and emission lines caused by cs interaction are extremely weak and can not be detected .
we find , however , that the excitation of h and he produced by x - rays in sn iip ejecta turns out to be sufficient to be detected as high velocity ( hv ) absorption features in h@xmath0 and hei 10830 lines against the bright sn iip photosphere .
this is the core of our proposed diagnostic for the wind in sne iip .
the second proposed probe for cs interaction exploits the possibility that a cool dense shell ( cds ) might form at the sn / cs interface because of radiative cooling .
the cds excited by x - rays could become visible as narrow hv absorption in h@xmath0 .
the velocity of this absorption would provide a direct measure of the expansion velocity of the sn / cs interface , a valuable dynamical characteristic of the cs interaction .
the identification of the expected hv lines in observed spectra is complicated by the presence of weak metal lines @xcite .
however , we argue that hv lines of h@xmath0 and h@xmath11 have been observed in spectra of sne iip , as previously discussed by @xcite for sn 1999em .
the paper is organized as follows .
we start with a description of the interaction model and cs interaction effects in the h@xmath0 and hei 10830 absorption lines formed in the unshocked ejecta during the photospheric epoch ( [ sec - mod ] ) .
we then compare our cs interaction models with the available spectra of sne iip in the h@xmath0 and hei 10830 lines and estimate the wind density for particular sne iip ( [ sec - inter ] ) . in
[ sec - narrow ] we address the issue of the h@xmath0 absorption in the cool dense shell at the sn / cs interface of sne iip .
we discuss implications of our models in the last section .
the model for ejecta - wind interaction effects in h@xmath0 and he i 10830 lines formed in the sn ejecta consists of three major parts : ( i ) an interaction model that provides the dynamical evolution of the sn / wind interface and the x - ray emission from the reverse and forward shocks ; ( ii ) a model for the ionization and excitation of h and he in the unshocked sn ejecta irradiated by x - rays ; and ( iii ) the calculation of line profiles . we perform ( iii ) using either a standard sobolev approximation or the direct integration of the radiation transfer equation , depending on the validity of the sobolev approximation .
the interaction of sn ejecta with cs wind leads to a canonical double - shock structure @xcite with the forward shock propagating in the cs gas and the reverse shock in the sn ejecta .
we treat the cs interaction of ejecta in the thin shell approximation @xcite in which the double - shock layer is reduced to an infinitely thin shell .
we assume that the freely expanding ( @xmath12 ) sn envelope has a sharp boundary at the velocity @xmath13 and begins to interact with a wind starting some moment ( @xmath14 day ) which corresponds roughly to the shock breakout phase .
free expansion is expected to take several doubling times of the initial radius , or several days , to be set up , but the observations we are modeling are at later times
. the maximum velocity @xmath13 is set by the escape of radiation from the shock wave at shock breakout . for the density distribution @xmath15 in the sn iip envelope we use an analytical expression , @xmath16 that closely approximates the combination of an inner plateau and outer power law tail found in hydrodynamic models ( e.g. , * ? ? ?
the parameters @xmath17 and @xmath18 are determined by the kinetic energy @xmath19 , ejecta mass @xmath20 , and @xmath21 : @xmath22 where @xmath23 the power law index @xmath21 lies in the range @xmath24 and is henceforth set to be @xmath25 .
we adopt the boundary velocity for a typical sn iip of @xmath26 km s@xmath2 , which is consistent with models of shock breakout @xcite .
also , the blue edge of the h@xmath11 emission in sn 1999gi on day 1 was observed to be at 15,000 km s@xmath2 @xcite .
hydrodynamical modeling predicts that at the shock breakout phase a thin dense shell forms at the outer boundary @xcite .
the physics of the ` boundary shell ' formation is in the transition from the adiabatic to the radiative regime of shock wave propagation in the outermost layers of the exploding star .
simple considerations of radiative diffusion @xcite give an estimate for the shell mass : @xmath27 where @xmath28 is the progenitor star radius and @xmath29 is the velocity at shock breakout , before free expansion has been established . during the acceleration phase
the boundary shell is subject to the rayleigh - taylor ( rt ) instability @xcite and therefore could be corrugated or even fragmented .
however , it is not clear whether the rt instability results in full fragmentation ; 2d or 3d hydrodynamic modeling of the phenomenon is lacking .
we will treat the boundary shell in our model as an intact thin spherical shell which is already in place at the initial moment .
the mass of the boundary shell is presumably equal to the mass of the outer envelope ( @xmath30 ) with an extrapolated power law @xmath31 .
this provides a good estimate of the shell mass for sn 1999em , @xmath32 , in agreement with hydrodynamical modeling @xcite .
the boundary shell is considered below as a seed for the cds that may , or may not , grow further due to radiative cooling at the reverse shock . to calculate the x - ray emission from the reverse and forward shocks
we assume that the postshock layers are uniform and their densities are 4 times larger than the corresponding preshock density .
we allow the electron temperature to be partially equilibrated , assuming the following interpolation : @xmath33 where @xmath34 is the electron - ion temperature equilibration time @xcite and @xmath35 is the equilibrium shock temperature with @xmath36 .
the lower limit @xmath37 is indicated by observational data on supernova remnants with shock velocities of @xmath38 km s@xmath2 @xcite .
the reverse shock is equilibrated during the photospheric phase for a mass loss rate of @xmath39 yr@xmath2 with a wind velocity of @xmath40 km s@xmath2 , while the forward shock is not .
since the reverse shock is equilibrated and dominates the x - ray luminosity , the uncertainty in @xmath41 does not affect our results .
the shock x - ray luminosity is expressed in terms of the kinetic luminosity as @xmath42 .
the efficiency parameter @xmath43 is smoothly interpolated between extreme cases , @xmath44 , where @xmath45 is the cooling time and is calculated for each shock using the cooling function of @xcite .
the wind is assumed to be steady with the density @xmath46 .
since all the interaction effects for low wind velocities ( @xmath47 ) are determined by the wind density @xmath48 , we use hereafter the dimensionless wind density parameter @xmath49 , where @xmath50 is the mass loss rate in units of @xmath3 yr@xmath2 and @xmath51 is the wind velocity in units of @xmath40 km s@xmath2 . the interaction model for a sn iip with the wind density @xmath52
is illustrated in fig .
[ f - dyn ] .
the two cases shown have the same ejecta mass @xmath53 but different kinetic energy : @xmath54 erg and @xmath55 erg .
the high energy case corresponds presumably to sn 1999em @xcite , while the low energy case illustrates sn 2004dj , as we will see below .
accordingly , we adopt @xmath56 km s@xmath2 in the high energy case and @xmath57 km s@xmath2 in the low energy case . the model with parameters for sn
1999em reproduces the observed ( absorbed ) _ chandra _ x - ray luminosity of sn 1999em in the @xmath58 kev band @xcite for a distance @xmath59 mpc @xcite quite satisfactorily ( fig . [ f - dyn ] ) , which indicates that @xmath52 is a reasonable choice for the wind density . the estimated galactic absorption
, @xmath60 @xmath61 @xcite , is small and does not significantly affect the observed x - ray luminosity .
the x - ray luminosity from the reverse shock dominates in both cases .
the typical electron temperature for the reverse shock at @xmath62 d is @xmath63 kev , while for the forward shock it is @xmath64 kev .
the early behavior of the velocities differs from the self - similar evolution @xcite because of the ejecta density cut - off in our model and the presence of the boundary shell at the initial moment .
the mass of the boundary shell is @xmath65 and @xmath66 in the high and low energy models , respectively .
the total mass of the cds attained at the time of the termination of the radiative regime in the reverse shock ( @xmath67 ) is @xmath68 and @xmath69 with @xmath70 d and @xmath71 d for the high and low energy models , respectively .
remarkably , the reverse shock is adiabatic in the self - similar models for the same parameters through this epoch .
this fact is related to the different dynamics of the self - similar model at an early stage , which results in a lower preshock density at the reverse shock .
the spectrum of x - rays is described below using a standard approximation @xmath72 .
we adopt @xmath73 for the forward shock , a reasonable approximation for temperature @xmath64 kev @xcite , while for the low temperature reverse shock we take @xmath74 .
the latter choice qualitatively describes the calculated energy distribution of x - rays with a significant contribution of emission lines for the electron temperature @xmath75 kev @xcite .
the hydrogen and helium ionization in a sn iip atmosphere at the photospheric epoch is time - dependent @xcite . to take this effect into account , we use a hydrogen atom represented by four levels and continuum .
all the essential radiative and collisional processes for hydrogen are taken into account , including non - thermal ionization and excitation induced by x - ray absorption .
the excitation of the helium @xmath76s level is determined by a simple model with two excited levels ( 2@xmath77s and 2@xmath78s ) .
we consider only major processes for the population of the 2@xmath78s level : nonthermal excitation and ionization with subsequent recombination into triplet states , depopulation by the collisional deexcitation transition @xmath76s@xmath79s , penning ionization @xcite and continuum absorption in the 10830 line ( 2@xmath78s2@xmath78p transition ) with the subsequent spontaneous transition 2@xmath78p1@xmath77s terminated by photon escape .
the energy balance includes the energy deposition due to the absorption of x - rays and the hydrogen photoionization from excited levels by photospheric radiation .
the cooling includes adiabatic loss ( @xmath80 work ) , free - free , free - bound , bound - bound emission of hydrogen , and cooling in the [ oi ] 6300 , 6364 and mgii 2800 lines . the fe ii line cooling is approximately included by multiplying the mgii 2800 cooling rate by a factor of two .
given the possibility of a high temperature regime ( @xmath81 k ) in the outer layers , the cooling at this range of temperatures is treated using the cooling function of @xcite .
the distribution of the ionization and excitation of h and hei in the atmosphere at one time is calculated by tracking the evolution of a lagrangian zone after its emergence from the photosphere .
the initial ionization and excitation at the photosphere is set assuming saha - boltzmann equations for the effective temperature .
subsequent ionization and excitation are computed by solving the system of time - dependent kinetic equations together with the time - dependent energy balance .
the average intensity of the photospheric radiation is assumed to be @xmath82 , where @xmath83 is the dilution factor and @xmath84 is the photospheric brightness .
the luminosity evolution at the photospheric epoch ( @xmath85 d ) is described by the expression @xmath86\,.\ ] ] with @xmath87 erg s@xmath2 and @xmath88 d , this corresponds to the bolometric luminosity of sn 1999em @xcite , assuming a distance of 11.7 mpc .
the photospheric radius is determined by the velocity at the photosphere and the age ( @xmath89 ) , where @xmath90 is deduced from observational data of @xcite .
the spectrum of the photospheric continuum is taken to be blackbody with @xmath91 in the wavelength region longward of the balmer jump , while the ultraviolet ( uv ) radiation shortward of the balmer continuum edge is suppressed by a factor @xmath92 which was assumed to monotonically drop after the explosion from unity to 0.03 on a timescale of 10 d. this evolution was deduced from the behavior observed in sn 1987a @xcite .
we found that the results are only weakly sensitive to the specific form of the variation of @xmath93 .
the energy deposition rate in the atmosphere due to the absorption of x - rays is calculated with an absorption coefficient @xmath94 @xmath95 g@xmath2 ( where @xmath19 is in kev ) .
the deposited energy is shared between heating , ionization , and excitation according to the recipes of @xcite and @xcite .
the solution of the kinetic equations for each lagrangian zone provides populations of the hydrogen levels @xmath96 and @xmath97 , and the @xmath76s level of hei in the atmosphere .
these values are then used to calculate line profiles of h@xmath0 and hei 10830 .
we compared our model in the case of zero wind density with the detailed model of @xcite and found that the present model results in slightly weaker h@xmath0 absorption ; at the relative intensity of 0.9 , the radial velocity in the simple model is lower by @xmath98 km s@xmath2 .
this drawback of the simple model does not noticeably affect the amplitude of the calculated hv absorption formed in the external layers , where the ionization unaffected by x - ray absorption is small .
the model line profiles of h@xmath0 and hei 10830 are computed for different wind densities , ages , ejecta mass and energy ( table 1 ) .
the standard model is characterized by @xmath53 , @xmath54 erg , @xmath25 . in all cases @xmath56 km s@xmath2 with the exception of the very low energy model ( ew ) for which @xmath99 km s@xmath2 .
we designate a model at a particular time by adding the age in days to the model name ; e.g. , w50 stands for model w on day 50 .
the effect of the wind density on h@xmath0 on day 50 is shown in fig .
[ f - modha]_a _ for the wind density parameters @xmath100 , 1 , 2 , and 4 . without a wind , the h@xmath0 absorption on day 50 forms because of the time - dependent ionization effect ;
a high ionization is maintained due to ly@xmath0 trapping and reionization from excited levels of hydrogen .
the cs interaction results in the ionization and excitation of the outer recombined layers of unshocked ejecta and , as a result , a hv absorption feature appears in the blue wing of the undisturbed profile . fig .
[ f - cart1 ] illustrates how h@xmath0 absorption is modified by the cs interaction . for @xmath100 ,
the cs interaction effect is small ( fig .
[ f - modha]_a _ ) and the wind density @xmath100 seems to be the least that could be detected by hv absorption in h@xmath0 . for @xmath101 , the strength of the hv absorption increases with the wind density roughly as @xmath102 and for @xmath103 becomes saturated .
the decrease of the velocity of the blue absorption edge is due to the deceleration caused by the dense wind .
the fact that for @xmath104 the hv absorption merges with undisturbed h@xmath0 absorption implies that in the case of a very dense wind it would be difficult to infer the density parameter @xmath105 using only the h@xmath0 absorption . the evolution of the hv absorption ( fig .
[ f - modha]_b _ ) shows that there is an optimal phase at about the middle of the plateau ( @xmath106 d ) when the cs interaction effect is most pronounced . at an early stage ( e.g. , on day 20 ) the cs interaction effect is not clearly seen because of the merging of hv absorption with the strong undisturbed absorption , while at a late stage ( e.g. , 80 d ) hv absorption becomes very faint .
the sensitivity of hv absorption to the ejecta density is shown by the model mw50 with a mass @xmath107 and energy @xmath108 erg .
the energy is chosen to match the blue wings of the main absorption of models mw50 and w50 .
the hv absorption in the low mass case turns out the same as in the model w50 , i.e. , the effect of reducing the ejecta density by a factor of 2 is negligible .
we also consider the cs interaction effect for sne iip with a very low kinetic energy , @xmath109 erg , which presumably applies to the low - luminosity sne iip , e.g. , sn 1999br @xcite .
this model with @xmath99 km s@xmath2 and @xmath52 predicts strong hv absorption which is merged with the undisturbed absorption ( fig .
[ f - modha]_d _ ) .
for all the models when hv absorption is unsaturated , the sobolev optical depth of this feature for @xmath110 is @xmath111 .
this implies that in the h@xmath11 line the optical depth @xmath112 , so we do not expect that hv absorption in this line will be pronounced ( but see [ sec - narrow ] ) .
we examined the sensitivity of hv absorption to the power law index @xmath113 of the x - ray spectrum of the reverse shock .
we found that a softer spectrum ( @xmath114 ) and harder spectrum ( @xmath115 ) produce slightly deeper hv absorption , so the preferred case @xmath74 corresponds approximately to a minimal effect of the cs interaction on h@xmath0 .
we checked the sensitivity of hv absorption to the adopted sn bolometric luminosity as well .
the hv absorption strength anticorrelates with the bolometric luminosity , but the dependence is weak .
the mechanism for this dependence is the depopulation of excited levels by photoionization .
the cs interaction effect in the hei 10830 line substantially differs from that in h@xmath0 in one important respect : in the absence of a wind the model does not predict noticeable absorption in hei 10830 (fig .
[ f - modhe]_a _ ) , except for a very early stage ( @xmath116 d ) .
the wind density @xmath100 produces weak hv absorption , on the verge of detectability , while the case @xmath52 yields hv absorption with a relative depth of @xmath117 and could be easily detected .
the absorption gets markedly stronger for larger @xmath105 , showing the sensitivity of hei absorption to the wind density in this particular range of the @xmath105 parameter .
remarkably , while in h@xmath0 the contribution of the interaction effect for @xmath104 is difficult to discern ( fig .
[ f - modha]_a _ ) , the hei 10830 hv absorption provides a probe for such a dense wind .
the strength of hv absorption in hei 10830 increases between days 20 and 50 but gets weak on day 80 ( fig . [ f - modhe]_b _ ) , resembling the behavior of hv absorption in h@xmath0 .
the effects of the ejecta density ( fig .
[ f - modhe]_c _ ) are similar to those in h@xmath0 .
the hv absorption for the low energy case is very strong and easily discernable , unlike in h@xmath0 .
spectra of sn 1999em @xcite on days 41 and 54 ( assuming the explosion date jd 2451476 ) show a depression in the blue wing of the h@xmath0 line ( fig .
[ f - ha99em ] ) at the predicted position of hv absorption ( @xmath118 to @xmath119 km s@xmath2 ) . to model the h@xmath0 profiles we adopted the standard parameters @xmath53 , @xmath54 erg , @xmath25 , and @xmath56 km s@xmath2 .
we find the best value of the wind density parameter is @xmath52 with an uncertainty of @xmath120 , assuming the other parameters to be fixed .
the indicated error reflects only the accuracy of estimating this parameter from the best fit and does not include errors in the observed spectrum and uncertainties in the model . also on day 54
, the model reproduces the general strength of the observed hv absorption ( fig .
[ f - ha99em ] ) .
yet we find that the observed hv absorption on day 54 seems to contain a notch component that is deeper than in the model .
the question arises of why the model fails to describe in detail the undisturbed h@xmath0 .
the model emission is weaker , while the absorption is stronger , than is observed ( fig .
[ f - ha99em ] and * ? ? ?
the weak model emission is probably related to our crude description of the uv continuum in the sn atmosphere ( cf . * ? ? ?
however , this can not explain why the observed absorption is shallow .
in fact , a check of available spectra of sne iip indicates that the h@xmath0 absorption seems to be shallow in any normal sn iip ( e.g. , sn 2004dj , see below ) .
in contrast , for the peculiar sn 1987a related to the explosion of a blue supergiant , the h@xmath0 absorption is deep and well reproduced by the time - dependent model @xcite .
we speculate that this is a strong indication that the shallowness of the h@xmath0 absorption in normal sne iip is related to the red supergiant structure of the presupernova .
for instance , convection in the red supergiant atmosphere could produce large amplitude density perturbations that are significantly amplified during the blast wave propagation . as a result , the outer layers of ejecta would acquire a clumpy structure responsible for the shallowness of the h@xmath0 absorption . returning to the hv absorption , in fig .
[ f - ha04dj ] we show the model fit to the observed h@xmath0 in sn 2004dj @xcite on days 55 and 64 ( for an explosion date jd 2453185 ) .
the model parameter set is @xmath53 , @xmath25 , as for sn 1999em , but @xmath121 erg .
one could retain the energy @xmath122 erg used for sn 1999em and vary @xmath20 .
however , in this case one requires @xmath123 to attain a comparable fit .
since the ejecta mass unlikely exceeds @xmath124 , the lower energy case for sn 2004dj is preferred . in accord with the lower energy we take @xmath57 km s@xmath2 and lower velocities at the photosphere by a factor 0.9 compared to sn 1999em . the derived wind density ( @xmath125 )
is determined by the strength of the hv absorption and its position , while the lower energy is needed to reproduce the correct position of the hv absorption . to check the effect of the adopted boundary velocity @xmath13 , we modeled the case @xmath126 km s@xmath2 . in this case
the wind density is lower , @xmath127 .
we prefer @xmath57 km s@xmath2 on the basis of the arguments presented below in sec .
[ sec - conduct ] . on day 55
the hv absorption in the observed spectrum has the appearance of a relatively broad shallow depression similar to sn 1999em on day 41 .
a bit later , on day 64 , a notch ( `` v - shaped '' narrow absorption ) appears at a radial velocity @xmath128 km s@xmath2 ( fig .
[ f - ha04dj ] ) .
the feature dominates the hv absorption until at least day 102 in spectra of sn 2004dj reported by @xcite .
neither the strength nor the shape of the notch at the late photospheric stage @xmath129 d can be produced by a model in which the hv absorption forms only in the unshocked ejecta .
this suggests that the notch in h@xmath0 has a _ different origin _ which is addressed in
[ sec - narrow ] .
the hv absorption in sn 2004dj has a markedly lower velocity compared to sn 1999em at a similar epoch ( 8200 vs. 11,500 km s@xmath2 ) .
this fact is significant for the origin of the depression in the blue wing of h@xmath0 since it rules out the possibility that the feature might be produced by some unidentified metal line .
photospheric velocities at this epoch are similar in both sne within several @xmath130 km s@xmath2 , so the expected positions of a metal line absorption should coincide in both sne within the same range of velocities . since this is not the case ,
we conclude that hv absorption in these supernovae _ can not be _ produced by an unidentified line .
an additional argument against the ` metal line ' option is that sn 1987a does not show hv absorption in h@xmath0 at all .
this would be difficult to understand in the context of the metal line option , but is easily explained by the very low density of the pre - sn wind ( @xmath131 , * ? ? ?
* ) that is incapable of producing detectable interaction effects .
the hv absorption of hei 10830 in sn 1999em on day 20 was identified by @xcite and attributed originally to radioactive excitation by external @xmath132ni .
here we abandon this explanation and propose excitation of he in the external layers by cs interaction .
the hei 10830 line of sn 1999em on day 20 @xcite is shown in fig .
[ f - he99em ] together with models for @xmath52 and the best value @xmath133 , which has an uncertainty of @xmath134 .
the wind density parameter derived from the hv absorption strength in hei 10830 agrees well with the value @xmath52 found from the h@xmath0 line . the clear - cut observed profile and
the sensitivity of hv absorption of hei 10830 to the wind density makes this line a valuable diagnostic tool for probing the cs wind in sne iip .
interestingly , based on a time - dependent model for a sn iip spectrum without cs interaction , @xcite predict strong hv absorption in hei 10830 at a relatively late epoch ; on day 50 , the relative depth of hei 10830 is @xmath135 for an enhanced he abundance @xmath136 .
this suggests a relative depth of @xmath137 for a solar he abundance .
the absorption lies in the range of @xmath119 to @xmath138 km s@xmath2 , which corresponds to the blue wing of hei 10830 in sn 1999em on day 20 ( fig .
[ f - he99em ] ) .
the hei 10830 line in sn 1999em on day 20 does not show evidence for the additional he absorption in the blue wing .
yet it is possible that in a more detailed model the time - dependent effects for he lines might be larger than in our simple model and would substantially contribute to the hv absorption of the hei 10830 line at an age @xmath62 d. however , for sn 1999em on day 20 , we note the close correspondence between the @xmath105 values deduced from the h@xmath0 and hei 10830 lines .
spectra of sn 1999em covering the hei 10830 line are also reported by @xcite for three epochs : 8 , 24 , and 34 d. the first spectrum shows a strong hei 10830 line with a p cygni profile .
this line , related to the early high temperature phase , rapidly disappears and is not seen in the next two spectra .
instead , hv absorption appears in this line . on day 24 ,
the hv absorption of hei 10830 is much the same as on day 20 in fig .
[ f - he99em ] , and it also appears on day 34 . a steady hv absorption feature with time is expected in our model ( [ f - modhe]b ) . on day 34
, an additional redder absorption at @xmath139 appears separate from the hv absorption .
the transformation of the hei 10830 line from a single hv absorption into a double absorption structure is also observed between days 22 and 44 in another type iip supernova , sn 1995v @xcite .
we believe that the low velocity absorption is the result of he excitation by gamma - ray photons leaking from the inner radioactive decays of @xmath132ni and @xmath132co @xcite . to produce the required excitation of he about @xmath140 of @xmath132ni should be mixed out to a velocity @xmath141 km s@xmath2 , which does not contradict the conclusion that the majority of the @xmath132ni resides in deeper layers with velocity @xmath142 km s@xmath2 @xcite .
we now address the issue of the notch that emerged in h@xmath0 in both sn 1999em and sn 2004dj at about day @xmath143 .
we propose that the notch is produced by h@xmath0 absorption of the photospheric radiation in the cds .
we also consider an alternative mechanism : additional excitation of hydrogen in the unshocked ejecta by ly@xmath0 photons emitted from the cds .
we first study the preferred explanation for the notch , h@xmath0 absorption in the cds , and consider the clearly observed case of sn 2004dj . in modeling the h@xmath0 absorption produced by the cds ,
the principal parameters are the column density of the cool gas , and its excitation temperature and turbulent velocity .
the undisturbed or weakly disturbed cds is characterized by a turbulent velocity of the order of the thermal velocity of the cool gas , @xmath144 km s@xmath2 . however , with this turbulent velocity the equivalent width of the hv absorption turns out to be too small .
a possible solution is to increase the turbulent velocity , as expected by the action of the rayleigh - taylor ( rt ) instability of the decelerating thin shell .
a significant deceleration of the model cds of sn 2004dj , about 20% during the first 10 days ( see low energy case in fig . [ f - dyn ] ) , favors the development of the rt instability of the cds followed by mixing of the cool dense gas with hot forward shock gas @xcite .
the mixed cool gas thus can acquire a significant velocity dispersion and turbulent velocities up to @xmath145 of the shell velocity .
both components could contribute to the notch formation : absorption in the weakly disturbed cds and in the mixed cds material .
we therefore consider two components : ( 1 ) a weakly perturbed spherical shell with a turbulent velocity of @xmath146 km s@xmath2 and ( 2 ) mixed cds gas with a large turbulent velocity @xmath147 km s@xmath2 .
the mass fraction of the second component is denoted @xmath11 .
the corresponding spectral components of the hv absorption related to the cds can be dubbed narrow notch and broad notch .
these components should be distinguished from the ` ejecta component ' a broader shallow hv absorption formed in the unshocked ejecta and addressed in the previous section . a cartoon ( fig .
[ f - cart ] ) illustrates the origin the hv absorption produced by the ejecta and the cds .
hydrogen excitation of the cds gas is presumably maintained by the absorption of x - rays emitted from the reverse shock .
the absorbed energy of x - rays is assumed to be homogeneously distributed through the cds material , as expected for the modest shell column densities for our situation .
the number density in the cds is determined by pressure equilibrium @xmath148 ( where @xmath149 is the pre - shock wind density ) for the typical temperature of the cool gas of @xmath150 k implied by the energy balance . on day 64
, the cds mass in our model for sn 2004dj is @xmath69 .
the other input parameters provided by the interaction model are : cds radius , @xmath151 cm , velocity @xmath152 km s@xmath2 , x - ray luminosity of the reverse shock @xmath153 erg s@xmath2 , and electron temperature in the reverse shock @xmath154 kev .
the population of the second level of hydrogen is computed using a steady - state approximation ( justified since @xmath155 @xmath156 ) , and assuming a two level plus continuum model for the hydrogen atom .
both non - thermal excitation and ionization by secondary electrons lead to the excitation of the second level , while depopulation proceeds by the two - photon transition , ly@xmath0 escape and collisional deexcitation .
the calculated profiles of hv absorption related to the cds are overplotted on the model h@xmath0 profile with ejecta hv absorption and displayed in fig .
[ f - nhva ] along with the observed spectrum of sn 2004dj @xcite on day 64 . shown
are three cases : one without the cds absorption and two cases with the cds absorption that differ by the mass fraction of the broad cds component ( @xmath157 and 0.5 ) . for the quoted values , the optical depths of the narrow components are 520 and 175 , while these values for the broad component are 0.2 and 0.45 , respectively .
the absolute value of the radial velocity of the hv absorption produced by the cds ( @xmath158 ) is slightly lower than the velocity of the cds , @xmath159 , owing to the sphericity effect . assuming uniform brightness of the photosphere ( lambert s law ) it is straightforward to show that the average absolute value of the radial velocity of the absorption is @xmath160\ , , \label{eq - hva}\ ] ] where @xmath161 is the ratio of the photospheric to the cds velocity .
according to this expression @xmath162 and , in the limit @xmath163 , @xmath164 with high accuracy .
the sphericity effects are responsible also for the additional broadening of the absorption over the radial velocity range @xmath165 , which is @xmath166 km s@xmath2 for the model of sn 2004dj on day 64 .
the alternative mechanism is that the notch could form as a result of an enhancement of hydrogen excitation in the unshocked sn ejecta due to the scattering of ly@xmath0 photons emitted from the cds that in turn is irradiated by x - rays from the reverse shock .
a resonance condition implies that ly@xmath0 photons can scatter in the unshocked ejecta only in the narrow velocity range @xmath167 , where @xmath168 is the boundary velocity of the unshocked sn ejecta .
the expected flux of ly@xmath0 from the cds is @xmath169 where @xmath170 is the efficiency of ly@xmath0 emission and @xmath171 is the x - ray luminosity absorbed by the cds .
the average excitation rate in the ejecta within the velocity width @xmath172 is balanced by ly@xmath0 escape ; other depopulation mechanisms are negligible .
the population of the second level ( @xmath96 ) thus is approximately determined by the rate equation @xmath173 where @xmath174 is the sn age , @xmath175 is the ly@xmath0 photon energy , and @xmath176 is the sobolev escape probability for ly@xmath0 . for the model of sn 2004dj ,
the notch in h@xmath0 produced by ly@xmath0 scattering in the ejecta is calculated on day 64 ( fig .
[ f - nhva ] ) .
two cases , with efficiencies @xmath177 and 0.5 , are plotted .
one sees that even for the unacceptably high efficiency @xmath178 the notch produced by the ly@xmath0 scattering is significantly weaker than that predicted by the absorption in the cds .
we conclude that the latter mechanism is the most likely explanation of the notch in h@xmath0 .
we examined whether the absorption in the cds might produce some effect in the hei 10830 line . for the above model
, the optical depth in the hei 10830 line turns out to be a factor of @xmath179 lower than in h@xmath0 .
therefore , it is unlikely that the notch can be seen in hei 10830 .
the main reason is the efficient depopulation of the 2@xmath78s level in the cds by penning ionization of hydrogen .
given the relatively high h@xmath0 optical depth of the cds , a notch at the same velocity could be seen in h@xmath11 as well .
we inspected spectra of sn 1999em @xcite and found the signature of a notch in h@xmath11 at the right position ( @xmath180 km s@xmath2 ) between days 39 and 81 .
in fact , @xcite already discussed the presence of the hv absorption in h@xmath0 and h@xmath11 in spectra of sn 1999em and concluded that these features can not be caused by metal line absorptions .
we also find a notch in h@xmath11 in spectra of sn 2004dj obtained by @xcite .
since the hv absorption components of h@xmath0 and h@xmath11 have not been previously mentioned for this supernova , we show parts of the sn 2004dj spectra @xcite retrieved from the sn archive suspect ( fig .
[ f - obs04dj ] ) .
the spectra show conspicuous notches at about @xmath181 km s@xmath2 both in h@xmath0 and h@xmath11 between days 67 and 102 . below we will model hv absorption in both lines in more detail . here
we present a model that can account for the strength and shape of the notch in h@xmath0 of sn 2004dj at the late photospheric epoch .
preliminary modeling suggests that the simple two - component model of the cds ( narrow and broad ) introduced above is not able to describe the strength of the notch at late times ( @xmath182 d ) : the excitation produced by the x - ray absorption is insufficient .
this suggests the presence of a more efficient excitation mechanism or an additional source of heat .
another problem is the relative depth of notches in the h@xmath0 and h@xmath11 lines : they are comparable and shallow .
this indicates that some veiling is present , possibly due to clumpiness .
both reasons force us to modify the two - component model of the cds absorption .
an important outcome of the rt mixing of the cds material in the forward shock is the growth of the area of the contact surface ( @xmath183 ) between cool and hot gas .
the development of mixing may be thought as the progressive decrease of the minimum linear scale of fragments ( @xmath184 ) .
this process is demonstrated in numerous experiments on the rt instability and in the case of sn / cs interaction is clearly seen in 2d and 3d hydrodynamical computations @xcite .
a fragmentation cascade of a fixed amount of cool gas leads to the scaling @xmath185 , while for a stationary mixing process @xmath186 @xcite .
the growth of @xmath183 and , accordingly , of the surface - to - volume ratio for the mixed cool gas , should lead to a greater role of thermal conduction in the heating of the cool gas at late times .
we speculate that this is the missing factor that could resolve the excitation problem . to describe the effect of thermal conduction , we introduce a third cds component : an additional broad cds component powered predominantly by thermal conduction .
the mass fraction of the third component is @xmath187 , while the cumulative area of the contact surface for the third component , @xmath188 , can be large , with an area ratio @xmath189 .
we assume that the third component is homogeneously heated by thermal conduction , which is a sensible approximation for a small thickness of sheets of cds gas , @xmath190 cm , comparable to the mean free pass length of hot electrons and protons in the cool dense gas .
the heat flux is presumably a small fraction of the saturated flux @xmath191 ; we adopt @xmath192 to allow qualitatively for the possible suppression of thermal conduction by a magnetic field .
the saturated flux is taken of the form @xmath193 @xcite .
the forward shock is not equilibrated ( @xmath194 ) , so the ion ( proton ) conduction could be significant . in our case ,
the ratio of proton - to - electron thermal flux is @xmath195 and this fact is taken into account . ionization and thermal balance are solved for the dense hydrogen cooled by ly@xmath0 and two - photon radiation .
the rt mixing generally produces an inhomogeneous ( lumpy ) distribution of mixing regions .
this is the result of the existence of a dominant angular scale of the rt instability , about @xmath196 @xcite and of the intermittence of the turbulent mixing .
we describe the distribution of mixed regions as an ensemble of isolated lumps which occupy a fraction @xmath197 of the volume ; @xmath198 corresponds to a homogeneous mixture of cool and hot phases .
the fraction @xmath199 should not be confused with the filling factor of the mixed cool dense gas , which is of the order of @xmath200 .
observationally , the lumpiness of mixing zones could be manifested as a veiling effect , i.e. , a comparable moderate depth of absorption lines with different optical depth ( viz . ,
h@xmath0 and h@xmath11 as in our case ) . to facilitate the treatment of the radiation transfer in the lumpy medium we use an occultation optical depth , @xmath201 , i.e. , the average number of lumps along the radius .
this value is defined by @xmath199 and a ratio of the total thickness of the mixing layer ( forward post - shock layer @xmath202 ) to the mean intercepted length of the lump @xmath203 ( for a sphere , @xmath204radius ) : @xmath205 .
generally , @xmath201 could be different for each of the three components .
however , we assume the same value of @xmath201 for all three components .
the effective optical depth in the line ( @xmath206 ) then can be expressed through the average optical depth of homogeneously distributed cds material @xmath207 and @xmath201 : @xmath208 $ ] @xcite .
we thus consider a three component model of the cool dense gas which includes : ( i ) a narrow cds component with turbulent velocity @xmath209 km s@xmath2 , ( ii ) a broad cds component with @xmath210 km s@xmath2 , both powered by x - rays , and ( iii ) a broad cds component powered by the thermal conduction .
the cds notches in h@xmath0 and h@xmath11 in sn 2004dj are calculated on the basis of the previous sn 2004dj model assuming @xmath211 km s@xmath2 for narrow and @xmath212 km s@xmath2 for broad cds components . the adopted fraction of the broad cds component is @xmath213 at the early epoch ( @xmath214 d ) , with a subsequent increase proportional to the age , while the fraction of the third component is @xmath215 ( table 2 ) .
the cds absorption in h@xmath0 is superimposed on the model line profile with the ejecta hv absorption . for the h@xmath11 line ,
the cds hv absorption is superimposed on the background spectrum taken as a linear interpolation between the blue and red wings of hv absorption in the observed profile .
the occultation optical depth @xmath216 ( table 2 ) is chosen to satisfy the ratio h@xmath0/h@xmath11 , while the area ratio @xmath217 is determined from the strength of the absorption .
( [ f - hvaev ] ) shows that a three - component model with the area ratio @xmath218 ( table 2 ) describes the hv absorption of both h@xmath0 and h@xmath11 lines satisfactorily . on day 102
the ejecta hv absorption is somewhat stronger , which is responsible for the red - blue asymmetry in the model profile .
however , this does not affect the intensity of the cds absorption feature .
the comparison to the two - component model and the behavior of @xmath217 show that the role of thermal conduction increases with time , which is the expected result .
interestingly , the result depends rather weakly on @xmath219 .
this is related to the canceling of effects of the variation of @xmath219 in the thermal excitation of hydrogen and the column density of the third component .
we performed modeling of the cds hv absorption adopting the lower boundary velocity , @xmath126 km s@xmath2 , and found that in this case deceleration of the cds is too slow , which leads to a larger velocity of the cds hv absorption on day 102 than is observed . the value @xmath57 km s@xmath2 is about the least that provides the required deceleration rate of the cds .
although the parameter values used for the third component are reasonable , the modeling can not be considered as proof that thermal conduction actually dominates the excitation of the cds material at a late epoch
the model contains many assumptions .
yet the need for an additional source of excitation of hydrogen responsible for the hv absorption is an established result .
thermal conduction is a natural mechanism strongly suggested by the inevitable rt instability and mixing of the cds matter with hot gas from the forward shock . on the other hand
, the additional excitation might at least partially be related to enhanced soft x - ray radiation from hot dense inhomogeneities produced by the rt mixing and heated by thermal conduction up to @xmath220 k. the identification of the features produced by h@xmath0 and h@xmath11 absorption in the cds provides an efficient tool for the direct determination of the velocity of the sn / wind interface in sne iip . with this value in hand we are able to obtain the wind parameter @xmath105 using the thin shell model of the deceleration .
of course , the method assumes that the ejecta parameters @xmath20 , @xmath19 , and @xmath21 are known . to facilitate the application of this method we use the self - similar solution .
this description is valid at the late photospheric epoch , when the initial conditions related to the structure of the outermost ejecta layers are already ` forgotten . '
the self - similar evolution of the cds radius suggests @xmath221 , with @xmath222 @xcite .
the factor @xmath223 for the adopted density distribution ( eq . [ [ eq - den ] ] ) is @xmath224^{1/(k-2)}\,,\ ] ] were @xmath225 is defined by equation ( [ eq - cmce ] ) and @xmath18 by equation ( [ eq - v0 ] ) .
the parameters for sn 1999em are as before : @xmath53 , @xmath54 erg , and @xmath25 .
the radial velocity of hv absorption is calculated both for the lambert photosphere ( eq . [ [ eq - hva ] ] ) and with limb darkening described by the first chandrasekhar approximation .
the evolution of the photospheric velocity is set according to @xcite .
the observational estimates of the radial velocity of hv absorption of h@xmath0 are obtained using the spectra reported by @xcite .
the best fit for the wind density parameter is found to be @xmath226 ( fig .
[ f - cds ] ) .
this estimate is in a good agreement with the value derived from the ejecta hv absorption of h@xmath0 ( @xmath227 ) and the hei 10830 line ( @xmath228 ) .
uncertainties in the limb darkening law and in the photospheric velocity do not affect the derived value of the wind density . in the case of sn
2004dj we use the same model as before : @xmath53 , @xmath229 erg , and @xmath25 .
the radial velocities of hv absorption are measured from spectra obtained by @xcite .
the best fit wind parameter is @xmath230 ( fig .
[ f - cds ] ) .
this agrees with the value @xmath231 obtained for sn 2004dj from the ejecta hv absorption of h@xmath0 .
the coincidence of @xmath105 values found by diagnostics based on the hv absorption in ejecta and in the cds is remarkable because the first probe is based on the model of x - ray emission by the reverse shock and the model of the ionization and excitation of ejecta , while the second is based only on the self - similar dynamics of the thin shell .
our goal was to study signs of cs interaction in the h@xmath0 and hei 10830 absorption lines of sn iip during the photospheric stage .
we developed a model for the ionization and excitation of h and he in the unshocked ejecta taking into account time - dependent effects and the irradiation of ejecta by x - rays .
the principal result is a prediction that for a typical rsg wind ( @xmath101 ) , the effects of cs interaction in a normal sn iip are significant and can be detectable at the photospheric stage in both lines .
these effects consist of the emergence of : ( i ) hv absorption in the blue wing of the undisturbed h@xmath0 line ( absorption ` shoulder ' ) at @xmath232 d , and ( ii ) hv absorption in hei 10830 at an age @xmath233 days at a radial velocity of @xmath234 km s@xmath2 .
we identify the hv absorption in sn 1999em in both h@xmath0 and hei 10830 and derive a wind density parameter @xmath235 .
hv absorption in h@xmath0 is identified in sn 2004dj as well and the value of the wind density parameter is again found to be @xmath235 .
the second important finding is that , in addition to the ejecta , absorption in the cds formed at the sn / wind interface plays an important role in the hv feature , especially at the late photospheric phase .
this component of the hv absorption is manifested as a notch in the hv absorption of h@xmath0 in sn 1999em and sn 2004dj after about day 60 .
a similar notch is seen in the h@xmath11 line of these sne .
we developed a model that produces the notch in both h@xmath0 and h@xmath11 at the different epochs of sn 2004dj which suggests a growing role for thermal conduction in the heating of the mixed cds gas .
the wind densities estimated from the observed velocities of hv absorption produced by the cds in sn 1999em and sn 2004dj are consistent with the densities derived from the hv absorption produced by unshocked ejecta .
the identification of hv absorption in h@xmath0 and h@xmath11 in sn 1999em has been discussed by @xcite . here , we identified hv absorption in h@xmath0 and h@xmath11 in sn 2004dj at a velocity @xmath236 km s@xmath2 lower than in sn 1999em .
the difference provides an additional argument against a possible metal line origin for these features .
@xcite claim that a similar component is also seen in the nai doublet .
indeed , there is a weak dip with a depth of @xmath237 at the right velocity , @xmath238 km s@xmath2 , if it is associated with the nai 5890 doublet .
we calculated the optical depth of the nai 5890 doublet for sn 1999em on day 60 in the model of absorption by the cds and found that the narrow cds component in this line has an optical depth of @xmath239 , which means that after broadening by about 10 times the depth of absorption should become equal to @xmath240 , comparable to the observed notch .
however , we did not find a similar notch in the nai line of sn 2004dj , or in any other sn iip . at present
, the reality of the hv absorption in the nai 5890 doublet remains uncertain .
we examined other published spectra of sne iip at the photospheric epoch and found evidence for hv absorption in the h@xmath0 blue wing of sn 1985p @xcite , in both h@xmath0 and hei 10830 of sn 1995v @xcite , in h@xmath0 of sn 2003gd , and sn 2006ov @xcite .
available spectra of sn 1999gi do not cover the range @xmath241 days when the most pronounced cs interaction effects should be observed . in noisy spectra on day 39 and 89
reported by @xcite , we do not find unequivocal evidence for hv absorption in h@xmath0 .
we checked spectra of sn 2004et @xcite and did not find convincing evidence for hv absorption in h@xmath0 or h@xmath11 .
early stage ( @xmath242 d ) spectra of sn 2004et show a shallow absorption in the h@xmath0 blue wing rapidly evolving towards the red .
however the siii 6347 , 6371 doublet may be responsible for this absorption . at a later epoch , a weak notch at 6282 is present but unfortunately it coincides with the telluric absorption 6281.7 line and the identification with hv absorption is doubtful .
@xcite consider the notch in h@xmath0 of sn 2004et to be real and remark that it might be related to cs interaction .
if the hv absorption in sn 2004et is absent , the question arises why the dense wind suggested by a relatively high radio luminosity of this sn compared to sn 2004dj and sn 1999em @xcite is not revealed optically .
the answer may be that the wind is so dense that the hv absorption is strong and gets merged with the main absorption as in the model case of @xmath104 ( fig .
[ f - modha ] ) .
this conjecture is consistent with the strong h@xmath11 absorption observed in sn 2004et @xcite .
a direct test of the proposed explanation might be the detection of strong hei 10830 absorption in sn 2004et .
unfortunately , to our knowledge , observations of the hei 10830 region in sn 2004et are lacking .
interestingly , sn 2006my @xcite , which does not show hv absorption , also has strong h@xmath11 like sn 2004et . if the same mechanism ( dense wind , @xmath243 ) explains the absence of the hv absorption in this case , then sn 2006my should demonstrate high radio and x - ray luminosities for a sn iip .
we did not find signatures of hv absorption in h@xmath0 observations of the low - luminosity sn 1999br in spectra taken during the first 42 days after the explosion @xcite . in the spectrum on day 42
, the h@xmath11 line is rather weak and the explanation that has been invoked for sn 2004et ( strong cs interaction effect ) is not applicable in this case .
the likely reason for the absence of hv absorption in sn 1999br is a low wind density , @xmath244 .
this conjecture has an interesting implication . since the mass loss rate increases with the stellar mass according to the scaling law of @xcite
, we expect that the main - sequence mass of pre - sn 1999br should be lower than that of sn 1999em .
an interesting result , although model dependent , is that at the late photospheric stage of sn 2004dj ( @xmath245 d ) thermal conduction may be the dominant heating mechanism of the mixed cds gas in the forward shock . in this regard , we recall the h@xmath0 problem in sn 1979c : at an age of @xmath63 yr the h@xmath0 luminosity was @xmath246 of the total reverse shock luminosity @xcite , which is beyond reasonable values of the h@xmath0 emission efficiency .
a natural solution to the problem of the high h@xmath0 luminosity in sn 1979c could be thermal conduction in the mixing layer of the forward shock @xcite .
another possibility , however , is that the mixing process could bring about enhanced x - ray emission from the hot dense mixed component , which may be another source of additional excitation of hydrogen in the cds .
this soft x - ray component might explain the softening of the x - ray emission observed in sn 1999em @xcite .
using _ chandra _ x - ray data , @xcite conclude that the wind density parameter for sn 1999em is in the range of @xmath247 ( or @xmath248 for the distance 11.7 mpc ) .
our present value ( @xmath235 ) is within a factor of 1.5 of the lower limit .
this should be considered as good agreement for independent estimates . for sn
2004dj the x - ray data are missing .
the interpretation of radio data for sn 1999em and sn 2004dj in the model of wind
free - free absorption indicates that the wind density in sn 2004dj is twice as low as in sn 1999em @xcite .
this is somewhat discrepant with the present results which indicate similar wind densities for these supernovae .
the wind density of sn 1999em may be overestimated compared to sn 2004dj from radio data because of the uncertain maximum in the 8.47 ghz light curve or a possible contribution of synchrotron self - absorption in sn 1999em @xcite .
the wind density around sn 1999em and sn 2004dj suggests that the amount of the pre - sn material lost at the rsg phase ( @xmath9 yr ) is moderate in these cases . assuming the wind velocity of a typical rsg , @xmath249 km s@xmath2
, we find with @xmath52 that at the rsg stage the mass lost by the rsg wind is @xmath250 . with the ejecta mass of @xmath251 @xcite , @xmath252 enclosed in the neutron star , and @xmath253 lost by the main - sequence wind
, the initial mass of sn 1999em progenitor turns out to be about @xmath254 . the mass loss rate by massive stars is often estimated using the formula of @xcite in which @xmath255 is defined through the stellar luminosity ( @xmath256 ) , mass ( @xmath20 ) , and radius ( @xmath28 ) .
for the rsg stage , this formula can be modified using the relation between the mass and luminosity @xmath257 obtained from evolution calculations of @xcite in the range @xmath258 for moderate / no rotation . with this relation ,
the mass loss rate by @xcite becomes @xmath259 yr@xmath2 .
since the luminosity at the rsg stage is determined by the mass of the he core which in turn is determined by the initial mass , we should use the initial mass to estimate @xmath255 .
for sn 1999em with @xmath260 and @xmath261 @xcite this formula predicts a mass loss rate @xmath262 yr@xmath2 , i.e. , eight times larger than our value .
a similar result is expected for sn 2004dj .
the disparity between our estimate of the pre - sn mass loss rate and the popular scaling law requires an explanation .
the disparity can not be attributed to a lower metallicity , because the metallicities of ngc 1637 ( host galaxy for sn 1999em ) and ngc 2403 ( host galaxy for sn 2004dj ) are approximately solar @xcite .
the difference can be reduced by a factor of 1.5 if we adopt a higher velocity for the pre - sn wind , @xmath263 km s@xmath2 instead of 10 km s@xmath2 .
the choice of @xmath264 km s@xmath2 is supported by the wind velocity of the massive rsg betelgeuse , @xmath265 km s@xmath2 @xcite .
this , however , does not completely resolve the problem .
we admit the possibility that the hydrodynamical model of the light curve @xcite might significantly overestimate the mass of ejecta for some unknown reason . in that case , to obtain the wind @xmath266 ( for @xmath263 km s@xmath2 ) the initial mass of the presupernova must be @xmath267 with an ejecta mass of @xmath268 instead of @xmath269 . with the sn energy
reduced accordingly , the depth and position of the hv absorption remain the same ( fig . [ f - modha]_c _ ) . to conclude , we have found three new probes of the wind density in sne iip that rely on spectroscopic observations during the photospheric epoch 20 - 100 d of ( 1 ) the ejecta hv absorption in h@xmath0 ; ( 2 ) the ejecta hv absorption in hei 10830 ; and ( 3 ) the hv absorption ( notch ) produced by the cds .
the third method is more practical than the other two because it does not require modeling of the absorption lines ; it uses velocity measurements and the self - similar model for the velocity fit . of course ,
this method , as well as the other two , assumes that we know the ejecta parameters ( @xmath20 , @xmath19 , @xmath21 ) .
however , the application of all three methods may be used to constrain ejecta parameters as well .
the first two methods are sensitive to winds with a density parameter @xmath270 . however in h@xmath0 cs interaction effects saturate for large density @xmath271 .
therefore , only hei 10830 remains a useful diagnostic tool for a high density wind with @xmath103 .
for the third method the range of accessible wind density is less certain , because a cds formed by the initial boundary shell could exist in principle even in the case of a low density wind , although the hydrogen excitation and shell density depend on the wind density .
further understanding of the relation between the strength of deep hv absorption and the wind density requires 3d hydrodynamical modeling of the shock breakout phase in the presence of a wind and new spectroscopic observations in both h@xmath0 and hei 10830 lines .
we thank daniela korcakova for sending us spectra of sn 2004dj , jozsef vinko for informing us about sn 2004dj spectra in the archive suspect , d. k. sahu for sending us spectra of sn 2004et , and the referee for helpful comments on the manuscript .
this research was supported in part by nsf grant ast-0307366 . | we propose new diagnostics for circumstellar interaction in type iip supernovae ( sne iip ) by the detection of high velocity ( hv ) absorption features in h@xmath0 and hei 10830 lines during the photospheric stage . to demonstrate the method
, we compute the ionization and excitation of h and he in supernova ejecta taking into account time - dependent effects and x - ray irradiation .
we find that the interaction with a typical red supergiant wind should result in the enhanced excitation of the outer layers of unshocked ejecta and the emergence of corresponding hv absorption , i.e. a depression in the blue absorption wing of h@xmath0 and a pronounced absorption of hei 10830 at a radial velocity of about @xmath1 km s@xmath2 .
we identify hv absorption in h@xmath0 and hei 10830 lines of sn 1999em and in h@xmath0 of sn 2004dj as being due to this effect .
the derived mass loss rate is close to @xmath3 yr@xmath2 for both supernovae , assuming a wind velocity 10 km s@xmath2 .
we argue that , in addition to the hv absorption formed in the unshocked ejecta , spectra of sn 2004dj and sn 1999em show a hv notch feature that is formed in the cool dense shell ( cds ) modified by the rayleigh - taylor instability . the cds results from both shock breakout and radiative cooling of gas that has passed through the reverse shock wave .
the notch becomes dominant in the hv absorption during the late photospheric phase , @xmath4 d. the wind density deduced from the velocity of the cds is consistent with the wind density found from the hv absorption produced by unshocked ejecta . |
large scale structures like galaxies and clusters of galaxies are believed to have formed by gravitational amplification of small perturbations . for an overview and original references , see , e.g. , @xcite .
density perturbations are present at all scales that have been observed @xcite .
understanding the evolution of density perturbations for systems that have fluctuations at all scales is essential for the study of galaxy formation and large scale structures .
the equations that describe the evolution of density perturbations in an expanding universe have been known for a long time and these are easy to solve when the amplitude of perturbations is much smaller than unity .
these equations describe the evolution of density contrast defined as @xmath1 . here
@xmath2 is the density at @xmath3 at time @xmath4 , and @xmath5 is the average density in the universe at that time .
these are densities of non - relativistic matter , the component that clusters at all scales and is believed to drive the formation of large scale structures in the universe . once the density contrast at relevant scales becomes large , i.e. , @xmath6 , the perturbation becomes non - linear and coupling with perturbations at other scales can not be ignored .
the equations that describe the evolution of density perturbations can not be solved for generic perturbations in this regime .
n - body simulations are often used to study the evolution in this regime .
alternative approaches can be used if one requires only a limited amount of information and in such a case either quasi - linear approximation schemes or scaling relations @xcite suffice . in cosmological n - body simulations ,
we simulate a representative region of the universe .
this is a large but finite volume and periodic boundary conditions are often used .
almost always , the simulation volume is taken to be a cube .
effect of perturbations at scales smaller than the mass resolution of the simulation , and of perturbations at scales larger than the box is ignored . indeed
, even perturbations at scales comparable to the box are under sampled .
it has been shown that perturbations at small scales do not influence collapse of perturbations at much larger scales in a significant manner .
this is certainly true if the scales of interest are in the non - linear regime @xcite .
therefore we may assume that ignoring perturbations at scales much smaller than the scales of interest does not affect results of n - body simulations .
perturbations at scales larger than the simulation volume can affect the results of n - body simulations .
use of the periodic boundary conditions implies that the average density in the simulation box is same as the average density in the universe , in other words we ignore perturbations at the scale of the simulation volume ( and at larger scales ) .
therefore the size of the simulation volume should be chosen so that the amplitude of fluctuations at the box scale ( and at larger scales ) is ignorable .
if the amplitude of perturbations at larger scales is not ignorable then clearly the simulation is not a faithful representation of the model being studied .
it is not obvious as to when fluctuations at larger scales can be considered ignorable , indeed the answer to this question depends on the physical quantity of interest , the model being studied and the specific length / mass scale of interest as well . the effect of a finite box size has been studied using n - body simulations and the conclusions in this regard may be summarised as follows . *
if the amplitude of density perturbations around the box scale is small ( @xmath7 ) but not much smaller than unity , simulations underestimate the correlation function though the number density of small mass haloes does not change by much @xcite .
in other words , the formation of small haloes is not disturbed but their distribution is affected by non - inclusion of long wave modes . * in the same situation , the number density of the most massive haloes drops significantly @xcite . * effects of a finite box size modify values of physical quantities like the correlation function even at scales much smaller than the simulation volume @xcite . *
the void spectrum is also affected by finite size of the simulation volume if perturbations at large scales are not ignorable @xcite . *
it has been shown that properties of a given halo can change significantly as the contribution of perturbations at large scales is removed to the initial conditions but the distribution of most internal properties remain unchanged @xcite .
* we presented a formalism for estimating the effects of a finite box size in @xcite .
we used the formalism to estimate the effects on the rms amplitude of fluctuations in density , as well as the two point correlation function .
we used these to further estimate the effects on the mass function and the multiplicity function .
* the formalism mentioned above was used to estimate changes in the formation and destruction rates of haloes @xcite .
* it was pointed out that the second order perturbation theory and corrections arising due this can be used to estimate the effects due to a finite box size @xcite .
this study focused specifically on the effects on baryon acoustic oscillations . *
if the objects of interest are collapsed haloes that correspond to rare peaks , as in the study of the early phase of reionisation , we require a fairly large simulation volume to construct a representative sample of the universe @xcite . in some cases , one may be able to devise a method to `` correct '' for the effects of a finite box - size , but such methods can not be generalised to all statistical measures or physical quantities .
effects of a finite box size modify values of physical quantities even at scales much smaller than the simulation volume @xcite .
a workaround for this problem was suggested in the form of an ensemble of simulations to take the effect of convergence due to long wave modes into account @xcite , the effects of shear due to long wave modes are ignored here .
however it is not clear whether the approach where an ensemble of simulations is used has significant advantages over using a sufficiently large simulation volume .
we review the basic formalism we proposed in @xcite in 2 .
we then extend the original formalism to the cases of non - linear amplitude of clustering and also for estimating changes in skewness and other reduced moments of counts in cells .
this is done in 3 . in
4 we confront our analytical models with n - body simulations .
we end with a discussion in 5 .
initial conditions for n - body simulations are often taken to be a realisation of a gaussian random field with a given power spectrum , for details see , e.g. , .
the power spectrum is sampled at discrete points in the @xmath8 space between the scales corresponding to the box size ( fundamental mode ) and the grid size ( nyquist frequency / mode ) . here
@xmath8 is the wave vector . in the approach outlined above ,
the value of @xmath21 at a given scale is expressed as a combination of the expected value @xmath22 and the correction due to the finite box size @xmath23 . here
@xmath22 is independent of the box size and depends only on the power spectrum and the scale of interest .
it is clear than @xmath24 , and , @xmath25 .
it can also be shown that for hierarchical models , @xmath26 , i.e. , @xmath27 increases or saturates to a constant value as we approach small @xmath12 . at large scales @xmath19 and @xmath27 have a similar magnitude and the _ rms _ fluctuations in the simulation become negligible compared to the expected values in the model .
as we approach small @xmath12 the correction term @xmath27 is constant and for most models it becomes insignificant in comparison with @xmath19 . in models where @xmath19 increases very slowly at small scales or saturates to a constant value
, the correction term @xmath23 can be significant at all scales .
this formalism can be used to estimate corrections for other estimators of clustering , for example the two point correlation .
see @xcite for details .
the estimation of the _ rms _ amplitude of density perturbations allows us to use the theory of mass function and estimate a number of quantities of interest . for details
, we again refer the reader to @xcite but we list important points here . *
the fraction of mass in collapsed haloes is under - estimated in n - body simulations .
this under - estimation is most severe near the scale of non - linearity , and falls off on either side .
if we consider fractional under - estimation in the collapsed fraction then this increases monotonically from small scales to large scales . *
the number density of collapsed haloes is under - estimated at scales larger than the scale of non - linearity .
the maximum in collapsed fraction near the scale of non - linearity leads to a change of sign in the effect of a finite box - size for the number density of haloes at this scale : at smaller scales the number density of haloes is over - estimated in simulations .
this can be understood on the basis of a paucity of mergers that otherwise would have led to formation of high mass haloes . *
the above conclusions are generic and do not depend on the specific model for mass function .
indeed , expressions for both the press - schechter @xcite and the sheth - tormen @xcite mass functions are given in @xcite , and we have also checked the veracity of our claims for the @xcite mass function .
in this section we outline how the formalism and results outlined above may be used to estimate the effect of a finite box - size on reduced moments . reduced moments like the skewness and kurtosis can be computed using perturbation theory in the weakly non - linear regime @xcite .
the expected values of the reduced moments are related primarily to the slope of the initial or linearly extrapolated @xmath17 , as all non - gaussianities are generated through evolution of the gaussian initial conditions and the initial @xmath17 characterises this completely .
we can use the expression for @xmath17 as it is realised in simulations with a finite box size to compute the expected values of reduced moments in n - body simulations in the weakly non - linear regime .
@xmath28 @xmath29 is the expected value of @xmath30 for the given mode , i.e. , when there are no box corrections and @xmath31 is the correction term in @xmath30 due to a finite box size .
box size effects lead to a change in slope of @xmath21 , and hence the effective value of @xmath32 changes .
the last term is the offset in skewness in n - body simulations as compared with the expected values in the model being simulated .
we would like to emphasise that this expression is valid only in the weakly non - linear regime .
in general we expect @xmath33 to increase as we go to larger scales .
thus the skewness is under estimated in n - body simulations and the level of under estimation depends on the slope of @xmath33 as compared to the slope of @xmath22 . in the limit of small scales where @xmath23 is almost independent of scale , we find that the correction is : @xmath34 + \mathcal{o}\left ( \frac{\partial}{\partial\ln{r}}\left(\frac{\sigma_1 ^ 2}{\sigma_0 ^ 2 } \right)^2 \right ) \nonumber \\ & & + \mathcal{o}\left(\frac{1}{\sigma_0 ^ 2 } \frac{\partial\sigma_1 ^ 2}{\partial\ln{r } } \right ) \end{aligned}\ ] ] here @xmath32 is the index of the initial spectrum we are simulating . for non - power law models this will also be a function of scale .
the correction becomes more significant at larger scales and the net effect , as noted above , is to under estimate @xmath30 .
similar expressions can be written down for kurtosis and other reduced moments using the approach outlined above .
we give the expression for kurtosis below , but do not compute further moments as the same general principle can be used to compute these as well .
@xmath35 ^ 2 + 2 \frac{\partial^2\ln\sigma^2}{\partial\ln^2 r } \nonumber \\ & \simeq & \frac{6071}{1323 } - \frac{62}{3}(n+3 ) \left[1 + \frac{\sigma_1 ^ 2}{\sigma_0 ^2 } \right ] + \frac{7}{3 } ( n+3)^2 \left[1 - \frac{8}{7 } \frac{\sigma_1 ^ 2}{\sigma_0 ^ 2 } \right ] \nonumber \\ & & + \mathcal{o}\left ( \frac{\partial}{\partial\ln{r}}\left(\frac{\sigma_1 ^ 2}{\sigma_0 ^ 2 } \right)^2 \right ) + \mathcal{o}\left(\frac{1}{\sigma_0 ^ 2 } \frac{\partial\sigma_1 ^ 2}{\partial\ln{r } } \right ) \end{aligned}\ ] ]
in this section we compare the analytical estimates for finite box size effects for various quantities with n - body simulations .
such a comparison is relevant in order to test the effectiveness of approximations made in computing the effects of a finite box size .
we have made the following approximations : * effects of mode coupling between the scales that are taken into account in a simulation and the modes corresponding to scales larger than the simulation box are ignored .
we believe that this should not be important unless the initial power spectrum has a sharp feature at scales comparable with the simulation size . * sampling of modes comparable to the box size is sparse , and the approximation of the sum over wave modes as an integral can be poor if the relative contribution of these scales to @xmath36 is significant .
table 1 gives details of the n - body simulations used in this paper . in order to simulate the effects of a finite box size
, we used the method employed by @xcite where initial perturbations are set to zero for all modes with wave number smaller than a given cutoff @xmath37 .
the initial conditions are exactly the same as the reference simulation in each series in all other respects . for a finite simulation box
, there is a natural cutoff at the fundamental wave number @xmath38 and simulations a1 , b1 and c1 impose no other cutoff .
these are the reference simulations for the two series of simulations .
simulations a2 , b2 and c2 sample perturbations at wave numbers larger than @xmath39 whereas simulations a3 , b3 and c3 are more restrictive with non - zero perturbations above @xmath40 .
the cutoff of @xmath39 and @xmath40 corresponds to scales of @xmath41 and @xmath42 grid lengths , respectively . for the c series of simulations , the cutoff of @xmath39 and @xmath40 corresponds to scales of @xmath43 h@xmath44mpc and @xmath45 h@xmath44mpc , respectively .
the background cosmology was taken to be einstein - desitter for the a and b series simulations .
the best fit @xmath46cdm model from wmap-5 @xcite was used for the c series of simulations . in order to ensure that the initial conditions do not get a rare contribution from a large scale mode , we forced @xmath47 while keeping the phases random for modes @xmath48 .
we have chosen to work with models where box size effects are likely to be significant , particularly with the larger cutoff in wave number .
this has been done to test our analytical model in a severe situation , and also to further illustrate the difficulties in simulating models with large negative indices .
we present results from n - body simulations in the following section .
.this table lists characteristics of n - body simulations used in our study . here
the spectral index gives the slope of the initial power spectrum and the cutoff refers to the wave number below which all perturbations are set to zero : @xmath38 is the fundamental wave mode for the simulation box .
all models were simulated using the treepm code @xcite .
@xmath49 particles were used in each simulation , and the pm calculations were done on a @xmath49 grid .
power spectra for both the a and the b series of simulations were normalised to ensure @xmath50 at the scale of @xmath51 grid lengths at the final epoch if there is no box - size cutoff .
a softening length of @xmath52 grid lengths was used as the evolution of small scale features is not of interest in the present study .
simulations for both the a and the b series were done with the einstein - desitter background and the c series used the wmap-5 best fit ( bf ) model as the cosmological background , as also for the power spectrum and transfer function . [ cols="^,^,^",options="header " , ] figure 5 shows the number density of haloes for the three series of simulations as a function of mass of haloes .
the haloes have been identified using the friends of friends ( fof ) method with a linking length of @xmath52 in units of the grid length . plotted in the same panels
are the expected values computed using the press - schechter mass function with a correction for the finite box size . for each series of simulations , and at each epoch , we fitted the value of @xmath53 to match the simulation with the natural cutoff at the box scale .
the same value of @xmath53 is then used for other simulations of the series .
we find that the features of the mass function are reproduced correctly by the analytical approximation , namely : * the number density of the most massive haloes declines rapidly as the _ effective _ box size is reduced . *
the number density of low mass haloes increases as the _ effective _ box size is reduced .
this feature is apparent only at the late epoch . in our discussion of analytical estimates of the effects of a finite box size on observable quantities
, we have so far omitted any discussion of velocity statistics .
the main reason for this is that the power spectrum for velocity is different as compared to the power spectrum for density and one can get divergences for quantities analogous to the second order estimators analogous to @xmath21 for models with @xmath54 .
this is due to a more significant contribution of long wave modes to the velocity field than is the case for density .
relative velocity statistics are more relevant on physical grounds and we use these for an empirical study of the effects of a finite box size on velocities .
it is also important to check whether considerations related to velocity statistics put a stronger constraint on the box size required for simulations of a given model .
we measure the radial pair velocity and also the pair velocity dispersion in the simulations used in this work .
these quantities are defined as follows : @xmath55 where the averaging is done over all pairs of particles with separation @xmath56 . in practice
this is done in a narrow bin in @xmath12 . here
@xmath57 is the scale factor , @xmath58 is the hubble parameter and @xmath59 is the velocity of the @xmath60th particle .
similarly , the relative pair velocity dispersion is defined as : @xmath61 where @xmath62 is the relative velocity for a pair of particles , and averaging is done over pairs with separation @xmath12 .
dividing by @xmath63 gives us a dimensionless quantity and the usefulness of this is apparent from the following discussion .
we have plotted the radial component of pair velocity as a function of scale @xmath12 in the top two rows of figure 6 .
panels in these rows show the pair velocity for the different models at early and late epochs . in each panel , we find that the dependence of pair velocity on @xmath12 is very sensitive to the small @xmath15 cutoff used in generating the initial conditions for the simulation .
it has been known for some time @xcite that @xmath64 is an almost universal function of @xmath0 .
this is certainly true in the linear regime where @xmath65 for clustering in an einstein - de sitter universe @xcite . in order to exploit this aspect , and also to check whether the relation between @xmath64 and @xmath0 in the weakly non - linear regime is sensitive to the box size , we plot @xmath64 as a function of @xmath0 at the same scale in the last two rows of figure 6 .
we find that all runs of a series fall along the same line and variations induced by the finite box size are small even in the non - linear regime .
figure 7 shows the relative velocity dispersion as a function of scale ( top two rows ) and also as a function of @xmath0 ( lowest two rows ) .
again , we find that although the relative pair velocity dispersion at a given scale is sensitive to the size of the simulation box , its dependence on @xmath0 is not affected by the large scale cutoff .
thus we can use estimates of the correction in @xmath0 to get an estimate of corrections in pair velocity statistics .
the conclusions of this paper may be summarised as follows : * we have extended our formalism for estimating the effects of a finite box size beyond the second moment of the density field .
we have given explicit expressions for estimating the skewness and kurtosis in the weakly non - linear regime when a model is simulated in a finite box size .
* we have tested the predictions of our formalism by comparing these with the values of physical quantities in n - body simulations where the large scale modes are set to zero without changing the small scale modes .
* we find that the formalism makes accurate predictions for the finite box size effects on the averaged two point correlation function @xmath0 and skewness .
* we find that the formalism correctly predicts all the features of the mass function of a model simulated in a finite box size .
* we studied the effects of a finite box size on relative velocities .
we find that the effects on relative velocities mirror the effects on @xmath0 .
it is desirable that in n - body simulations the intended model is reproduced at all scales between the resolution of the simulation and a fairly large fraction of the simulation box .
the outer scale up to which the model can be reproduced fixes the effective dynamical range of simulations .
one would like @xmath30 be within a stated tolerance of the expected value at this scale .
we plot @xmath66 for power law models at the scale @xmath67 , @xmath68 and @xmath69 in the left panel of figure 8 .
these are plotted as a function of @xmath70 .
it can be shown that this ratio , as also @xmath71 are functions of scale only through the ratio @xmath72 .
we find that @xmath66 is large for large negative @xmath32 and decreases monotonically as @xmath32 increases .
this ratio is smaller than @xmath73 only for @xmath74 at @xmath75 .
the corresponding number for @xmath76 is @xmath77 , and for @xmath67 is @xmath78 clearly , the effective dynamic range decreases rapidly as @xmath79 .
this highlights the difficulties associated with simulating such models .
similarly , one would like @xmath21 and @xmath22 to be comparable at the scale of non - linearity . from requirements of self similar evolution of power law models in simulations
, we find that at the scale of non - linearity @xmath80 is required for the effects of a finite box size to be ignorable .
this gives us a lower bound on @xmath81 for any given model .
the middle panel shows the required @xmath81 as a function of @xmath70 for @xmath82 , @xmath83 and @xmath52 . here
@xmath84 is the asymptotic value of @xmath23 at @xmath85 and is a fairly good approximation at small scales .
we find that the required @xmath81 for @xmath86 is more than @xmath87 for @xmath88 at the scale of non - linearity .
thus we need a simulation with @xmath89 if we are to probe the strongly non - linear regime ( @xmath90 ) with some degree of confidence .
requirements for models with @xmath91 are much more stringent , and for models like @xmath92 even the largest simulations can not be used to study the asymptotic regime . to put things in context for the favoured cosmological model , the right panel in figure 8 shows contours of @xmath66 in the @xmath93 plane for the @xmath46cdm model that best fits the wmap-5 data @xcite .
we find that in order to ensure that the error in skewness is less than @xmath73 at a scale of @xmath94 h@xmath44mpc , we need a simulation box of more than @xmath95 h@xmath44mpc .
the required box size is much bigger if the tolerance on error in skewness is smaller .
this is a very stringent requirement for simulations of the epoch of reionization where one would like to get the clustering right at the scales of a few mpc .
given that the formalism we have proposed works well when compared with simulations , and the fact that calculations in this formalism are fairly straightforward , we would like to urge the cosmological n - body simulations community to make use of this formalism .
we would like to request simulators to report the fractional corrections to the linearly extrapolated amplitude of clustering and the fractional correction to skewness across the range of scales of interest .
this will enable users of simulations to assess potential errors arising due to a finite simulation volume .
numerical experiments for this study were carried out at cluster computing facility in the harish - chandra research institute ( http://cluster.mri.ernet.in ) . this research has made use of nasa s astrophysics data system .
we would like to thank t. padmanabhan for useful discussions .
we thank the anonymous referee for useful comments and suggestions . | n - body simulations are an important tool in the study of formation of large scale structures .
much of the progress in understanding the physics of galaxy clustering and comparison with observations would not have been possible without n - body simulations . given the importance of this tool , it is essential to understand its limitations as ignoring these can easily lead to interesting but unreliable results . in this paper
we study the limitations due to the finite size of the simulation volume . in an earlier work we proposed a formalism for estimating the effects of a finite box - size on physical quantities and applied it to estimate the effect on the amplitude of clustering , mass function . here , we extend the same analysis and estimate the effect on skewness and kurtosis in the perturbative regime .
we also test the analytical predictions from the earlier work as well as those presented in this paper .
we find good agreement between the analytical models and simulations for the two point correlation function and skewness .
we also discuss the effect of a finite box size on relative velocity statistics and find the effects for these quantities scale in a manner that retains the dependence on the averaged correlation function @xmath0 .
methods : n - body simulations , numerical
gravitation
cosmology : theory , dark matter , large scale structure of the universe |
magnetic field generation is a key problem in understanding solar variability across wide range of scales .
modern high - resolution observations of the global magnetic field distribution , such as from hmi / sdo @xcite , and magnetic fields in selected areas by nst / bbso @xcite , sot / hinode @xcite and imax / sunrise @xcite demonstrate the complicated dynamics of magnetic fields and a tendency to self - organization . traditionally , the solar dynamo problem is divided according to two scales : the global dynamo , operating on the scale of the 22-year solar cycle and controlling the global field ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and the local dynamo ( e.g. * ? ? ?
* ; * ? ? ?
* ) , which operates on the scale of granulation and super - granulation and is believed to be responsible for the sun s ` magnetic carpet ' @xcite . however , energetic and dynamic connections between the two dynamos are not clear .
recently , it became possible to resolve flows and magnetic fields associated with magnetic flux emergence on very small subgranular scales ( e.g. * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ) , which probably reflects the working of the local dynamo in subsurface layers . resolving magnetic fluxes though observations on smaller and smaller scales has renewed interest in small - scale magnetic field generation and raised a question about the existence of unresolved ( or ` hidden ' ) magnetic flux in the quiet sun ( e.g. , *
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
numerical simulation is an efficient way to investigate properties of magnetic field generation on small scales . in particular , direct numerical simulations of simplified convective flows
demonstrate the existence of the local dynamo and provide its basic characteristics , for instance , the effect of the magnetic prandtl number on dynamo properties in the case of general turbulent flows ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and also for solar - type convection ( e.g. , * ? ? ?
* ; * ? ? ?
in addition , recent ` realistic'-type radiative mhd simulations have reproduced solar conditions with a high degree of realism and demonstrated that the magnetic field can be quickly amplified by local dynamos from a @xmath1 g seed field to @xmath2 g magnetic elements @xcite , and a 1 g uniformly imposed horizontal seed field increased by dynamo action to @xmath3 g @xcite . in this paper
we present new realistic - type 3d radiative mhd simulations of the local solar dynamo for various seed field strengths , from @xmath0 to @xmath1 g ( in 5 simulation runs ) , and investigate the development and properties of the dynamo process . in particular , we find that the magnetic field can be magnified above the equipartition strength ( @xmath4 g ) , reaching more than 1000 g in the photosphere . in the paper
we discuss the initial stage of the local dynamo , formation of magnetic field strictures of different scales , location of the dynamo process , turbulent properties of magnetoconvection , as well as mechanisms and properties of spontaneously formed magnetic patches below and above the photosphere .
we use the 3d radiative mhd code , ` solarbox ' , developed at nasa ames research center by a. wray and n. mansour for performing local dynamo simulations .
the code is based on a les formulation for compressible flow , and includes a fully coupled radiation solver , in which local thermodynamic equilibrium is assumed .
radiative transfer between fluid elements is calculated using a 3d multi - spectral - bin method using long characteristics . for initial conditions we use a standard solar model of the interior structure and the lower atmosphere .
the code has been carefully tested and was previously used for studying the excitation of solar acoustic oscillations by turbulent convection in the upper convection zone @xcite and other problems ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
we solve the grid - cell averaged equations for the conservation of mass ( [ mass ] ) , momentum ( [ mom ] ) , energy ( [ energy ] ) , and magnetic flux ( [ eqb ] ) : @xmath5 @xmath6 @xmath7_{,i}=0,\ ] ] @xmath8_{,j}=0,\ ] ] where @xmath9 is the average mass density , @xmath10 is the favre - averaged ( density - weighted ) velocity , @xmath11 is the magnetic field , and @xmath12 is the average total energy density @xmath13 , where @xmath14 is the gravitational potential and @xmath15 is the favre - averaged internal energy density per unit mass .
@xmath16 is the radiative flux , which is calculated by solving the radiative transfer equation , and @xmath17 is the averaged stress tensor @xmath18 , where @xmath19 is the viscosity .
the gas pressure @xmath20 is a function of @xmath15 and @xmath9 calculated through a tabulated equation of state @xcite ; @xmath21 is the reynolds stress , @xmath22 is the molecular thermal conductivity , @xmath23 is the turbulent thermal conductivity , @xmath24 is the molecular electrical conductivity , and @xmath25 is the turbulent electrical conductivity . the simulation results are obtained for a computational domain of @xmath26 mm , in which a 1-mm layer of the low atmosphere is included .
the grid - size is 12.5 km in the horizontal and 12 km in the vertical direction . above the solar surface
the vertical grid size is constant and stretches out in deeper layers .
the lateral boundary conditions are periodic .
the top boundary is open to mass , momentum and energy fluxes , and also to the radiation flux .
the bottom boundary is open only for radiation , and simulates the energy input from the interior of the sun .
currently it is impossible to achieve a realistic reynolds number in numerical simulations ; therefore the modeling of dynamical properties of solar convection is achieved through implementation of subgrid - scale les turbulence models .
these can effectively increase the reynolds number and provide better representation of small - scale motions .
here we used a smagorinsky eddy - viscosity model @xcite , in which the compressible reynolds stresses were calculated in the form @xcite : @xmath27 , where the smagorinsky coefficients are @xmath28 , @xmath29 is the large - scale stress tensor , and @xmath30 with @xmath31 , @xmath32 , and @xmath33 being the grid - cell dimensions .
local dynamo action is a complicated interaction of magnetic fields and highly turbulent flows on small scales .
the dynamo modeling is started by adding a very weak seed field into a hydrodynamic simulation model of fully developed solar convection . to investigate the effects of the initial seed - field properties , we consider 5 cases of magnetic field initialization ( table [ tab : cases ] ) . in three cases
the initial magnetic field is @xmath1 g , and has various initial distributions : ( @xmath34 ) uniform vertical field , ( @xmath35 ) checkerboard - like , alternating polarity patterns , and ( @xmath36 ) random white noise . in case
@xmath35 the checkerboard structure has a period of magnetic field variations of 100 km , in order to mix opposite - polarity patches in the intergranular lanes . in two other cases , @xmath37 and @xmath12 , with random seed fields and strength @xmath38 and @xmath0 g
, we test the sensitivity of the dynamo action to the initial field strength .
the hydrodynamic conditions at the time of magnetic field initialization are exactly the same for the cases @xmath39 .
.properties of the seed magnetic field .
[ tab : cases ] [ cols="^,^,^ " , ]
the initial evolution of the seed - field elements is determined by turbulent flows and demonstrates properties similar to uniformly distributed corks that tend to collect in the intergranular lanes , where the granulation flows converge . during this stage , lasting about 1 min , no dynamo action is present , and the field amplification is a result of simple compression by converging flows .
figure [ fig : checker ] shows an example of this effect for an initially regular checkerboard structure of magnetic field ( case @xmath35 , see table [ tab : cases ] ) , where the deformation of the field clearly reflects a flow dynamics in which convective flow drags the magnetic field lines and helical motions cause polarity reversals on small scales .
magnetic field amplification in turbulent solar convection can be can be roughly divided into two basic mechanisms : 1 ) magnetic field concentration due to converging flows , and 2 ) dynamo processes driven by helical or shear motions .
the time delay for the natural appearance of ` new ' opposite - polarity patches can be estimated as a half of the overturning time on the smallest resolved scales , which for a resolution of 12 km is approximately @xmath40 sec .
the time - lag for the appearance of a dynamo - like behavior on the smallest resolved scales due to the helical twisting of magnetic field lines can be estimated as a double overturning time on these scales , @xmath41 sec .
weak magnetic fields do not significantly affect dynamic properties of the quiet sun , in terms of the life - time and size distribution of granules .
nevertheless , the dynamo - generated magnetic fields are highly inhomogeneous , and , in small - scale patches , the field strength can be high enough to act on surrounding turbulent flows ( the so - called ` back reaction ' ) . in the current high - resolution simulations , which account for turbulent dynamics on sub - grid scales , we are able to capture the complicated interaction and energy exchange between the small - scale fields and flows @xcite . figure [ fig : power]@xmath42 shows power spectra ( multiplied by the @xcite law function @xmath43 ) for the turbulent kinetic energy in the photosphere for the pure hydrodynamic case ( black curve ) , and 4 ( green ) and 5 ( red ) hours after seed - field initialization .
scaling of the spectra with @xmath43 shows the difference of the inertial range obtained from the kolmogorov law @xcite .
the spectra show that , during development of the dynamo process , the growing magnetic field on small scales ( @xmath44 ) suppresses the turbulent motions that lead to kinetic energy transport to larger scales ( green curve in fig .
[ fig : power]@xmath42 ) .
the kinetic energy redistribution makes properties of turbulent flows on the solar surface closer to the theoretical kolmogorov slope , @xmath45 .
the increased magnetic flux after 5 h is strong enough to affect the turbulent flow through all scales present in the domain , which is reflected by the decreasing kinetic energy in the turbulence spectra ( fig .
[ fig : power]@xmath42 , red curve ) .
during the first half - hour of magnetic field generation , the shape of the magnetic energy spectrum varies during an exponential growth .
later in time , the shape of the spectrum is nearly constant , showing a steady energy increase in all scales with a slope of @xmath46 for the large scales and @xmath47 in the inertial range before reaching the steep dissipation range .
this illustrates a second amplification phase in the dynamo process ( e.g. * ? ? ?
* ) with a tendency toward an inertial - range slope of @xmath47 at @xmath48 h in our simulations ( figure [ fig : power]@xmath49 ) .
it is known that magnetic field can be amplified by swirling turbulent motions , shearing and converging flows ( e.g. * ? ? ? * ; * ? ? ?
. magnetic field amplification due to the ` stretch - twisting - fold ' mechanism @xcite is expected to work very efficiently in regions with the strongest helical motions .
figure [ fig : time ] shows the vertical distribution of the rms velocity and kinetic helicity and the evolution of the mean kinetic helicity for different depths for different moments of time for the initial seed field of @xmath0 g ( case @xmath12 ) .
these distributions show that the primary layers of the local dynamo action are where the turbulent flows are strongest .
the strongest helical motions occupy the top 1-mm subsurface layer , and the helicity is transported by downdrafts from the subsurface into deeper layers . magnetic field generated in the subsurface layer
is also transported by convective downflows into deeper layers , where the field can be further compressed and amplified .
understanding the local small - scale magnetic field generation process is critical for studying complex solar mhd problems , such as the interaction of magnetic field and flow through different scales . to investigate field generation in detail
we consider different episodes in the small - scale dynamo action . the first case is demonstrated in fig .
[ fig : case1 ] , showing the evolution of ( @xmath42 ) the vertical velocity , ( @xmath49 ) the vertical magnetic field ( color background ) and the horizontal velocity field ( arrows ) , and ( @xmath50 ) the electric current density ( color map ) and enstrophy ( contour lines ) in a selected @xmath51 km @xmath52 km region of the photosphere with 15 sec cadence .
the development of a local bipolar magnetic structure is associated with strong ( @xmath53 km / s ) swirling motions , on scales from @xmath54 km to @xmath55 km , stretching and twisting the magnetic field , which is also compressed by converging flows in the downdraft .
such such small - scale swirling motions in the intergranular lanes are typically associated with strong downflows , @xmath56 km / s @xcite .
the appearance of a bipolar magnetic structure ( with a prominent negative ( blue ) polarity in fig .
[ fig : case1]b ) is also a result of swirling flows driven by a vortex tube oriented along the solar surface ( indicated by the white arrow at fig . [ fig : case1 ] @xmath42 ) , similar to previously described by @xcite .
the horizontally oriented vortex tube captures magnetic field lines and drags them into the subsurface layers .
the dynamics of the positive polarity patch is mostly related to the vertically oriented helical motions similar to the process described by @xcite .
such complicated dynamics of flows is caused by interacting , differently oriented vortex tubes , which , in fact , have variable orientation in space , and only locally can be regarded as having ` vertical ' or horizontal orientation @xcite .
the generated electric current ( fig .
[ fig : case1 ] @xmath50 ) shows a clear correlation with the swirling dynamics of convective flows .
the strongest current density corresponds to areas on the periphery of the helical motions or between vortices , where stretching of the magnetic field lines is strongest .
another example illustrates development of a bipolar magnetic structure due to helical flows and shearing flows along the intergranular lane .
figure [ fig : case3 ] shows a time sequence of : ( @xmath42 ) the vertical velocity , ( @xmath49 ) the vertical magnetic field , ( @xmath50 ) the magnitude of the electric current density , ( @xmath57 ) the time - derivative of the vertical magnetic field with overlaid contour lines of enstrophy , and ( @xmath15 ) the vertical component of the electric current density ( dashed curves correspond to negative values )
. an interesting feature of this example is a different scenario of the development of the magnetic elements of opposite polarity ( determined by the sign of the vertical field ) : the positive - polarity magnetic patch evolves following the ` classical ' scenario of the field amplification due to helical flows described in the first example , whereas the negative - polarity patch starts forming in the intergranular lane mostly due to converging flows ( magnetic collapse ) .
the evolution of the magnetic elements of this bipolar structure is accompanied by locally growing electric current ( fig .
[ fig : case3 ] @xmath58 ) . comparison of the velocity , pattern , and time derivative of the vertical magnetic field shows that some swirling motions cause magnetic field dissipation , probably due to scattering of the field lines , and often the field amplification takes place on the periphery of swirls or regions with shearing flows . the distribution of vertical electric current ( fig .
[ fig : case32de - current ] ) shows a strong correlation with the horizontal velocity of the swirling flows in the subsurface layers , and also interconnections with other subsurface current structures . in deeper layers , the magnitude of the electric current increases , but its distribution becomes diffuse and does not show a clear association with the near - surface dynamics .
the kinetic helicity patterns show that the scale of the swirling motions increases with depth , from @xmath59 km at the photospheric layer ( fig .
[ fig : case32dkinhel ] ) up to @xmath60 km at a depth of 300 km below the photosphere ( panel @xmath57 ) . in the deeper layers
, the scale of helical motions continues to increase ( to larger than 150 km ) , but the distribution of kinetic helicity becomes complicated and consists of opposite - sign helical flows , which however continue swirling together ( fig .
[ fig : case32dkinhel ] @xmath15 ) , and disappear at a depth of about 500 km below the solar surface ( panel @xmath61 ) . cross - correlation of the kinetic helicity and the squared magnetic field strength ( fig .
[ fig : cross - corrhelu - b2 ] ) shows the best correlation in subsurface layers , from the photosphere to a depth of about 500 km , indicating the primary layers where the helicity is the strongest ( see also fig .
[ fig : time ] @xmath49 ) and also where the generated magnetic field is transported by downflows into the deeper layers ( fig .
[ fig : time ] @xmath50 ) .
from previous numerical studies it is known that important dynamical and energetic links between subsurface turbulent convective flows and the low atmosphere are established through small - scale vortex tubes . in the presence of magnetic field ,
the vortex tubes represent channels of energy exchange between the convective layers and the chromosphere , can result in heating of chromospheric layers , and be a source of small - scale spicule - like eruptions @xcite .
in addition to these effects , our simulations of the local dynamo show that the recently debated ( e.g * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ?
* ) anisotropy between the vertical and horizontal magnetic field components exists only in the photosphere and above .
the anisotropy changes with height but there is no dependence of these topological properties from the character of the seed field . the vertical distribution of the rms magnetic field ( fig .
[ fig : brms ] ) shows a slowly increasing magnetic field strength with depth . in the convection zone ,
the distribution of the vertical and horizontal magnetic fields are similar . at the surface ,
the vertical magnetic field becomes dominant , and then sharply decreases above the photosphere , where the magnetic field is mostly represented by the horizontal component .
this predominance of the vertical or horizontal fields in the different layers reflects the topological properties of the dynamo - generated magnetic field , which are characterized by magnetic field lines forming loops above the photosphere ( fig .
[ fig : b063dm - lines ] ) .
such topological structure resembles the magnetic canopy suggested from observations @xcite .
it is interesting to note that the closest opposite - polarity patches may not even be connected by magnetic field lines above the solar surface ( by magnetic loops ) , but instead interact through electric currents above and below the photosphere .
figure [ fig : case33de - current ] illustrates the topological structure by electric current density streamlines above and below the photosphere ( shown as a horizontal semi - transparent plane ) .
each streamline is tracked from a point in the region of positive - polarity ( yellow - orange streamlines ) and negative polarity ( dark blue ) patches .
the topological structure of the electric current above the photosphere is often characterized by spirals , arcs , and large swirls . below the solar surface
such topology can represent highly turbulent flows , as in the case of the positive patch ( yellow streamlines , in fig .
[ fig : case33de - current ] ) , or as a very regular spiral structure , as in the case of the negative patch ( blue lines , fig .
[ fig : case33de - current ] ) .
for instance , in figure [ fig : case33de - current ] a current streamline originating in the positive polarity patch ( yellow ) is strongly twisted around the negative patch .
observations of the sun demonstrate the complicated dynamics of turbulent convection and its interaction with magnetic field in a wide range of temporal and spatial scales . in particular
, the dynamo problem is a key to understanding variations of solar activity and various processes of self - organization . in this paper
, we addressed the problem of small - scale ( local ) dynamos responsible for the quiet - sun magnetic field . to investigate this problem we used a 3d radiative mhd code and performed several simulation runs for different strengths ( @xmath0 to @xmath1 g ) and spatial distributions of the initial seed field ( table [ tab : cases ] ) .
after the initialization of the seed field , the magnetic elements first behave as corks in a fluid and start concentrating in the intergranular lanes . shortly thereafter , overturning turbulent flows drag the field lines , and the elementary magnetic elements start changing polarity . later , this process expands to larger scales . because the strength of the seed field is negligible in all simulation cases , the magnetic field amplification varies among our simulation cases only during first few minutes after initialization field due to the different initial topological distributions .
our simulation results show that the magnetic field amplification is driven by four primary mechanisms : 1 ) concentration by converging flows into the intergranular lanes , 2 ) shearing flows ( shear dynamo ) , 3 ) helical motions ( small - scale dynamo driven by the kinetic helicity , @xmath62-effect ) and 4 ) turbulent collapse of magnetic field .
all four mechanisms of magnetic field amplification are present in our numerical model and linked to each other .
thus , the local dynamo process represents a complicated interplay of these mechanisms .
we presented two characteristic examples : one in which a bipolar magnetic structure was generated by the interaction of vertical and horizontal vortex tubes , and another in which the field was generated by a combination of a vertical vortex tube and shearing flows .
the flow topology of the local dynamo patches is very complicated and requires further investigation .
however , the primary topology of the dynamo - generated magnetic field is represented by compact magnetic loops appearing as bipolar structures in the intergranular lanes and forming an ` internetwork ' .
development of such a magnetic network causes redistribution of the turbulent kinetic energy by transferring kinetic energy from small to large scales , bringing the turbulent spectrum of the photosphere closer to the kolmogorov power law , @xmath63 ( fig .
[ fig : power ] ) .
the process of amplification of magnetic field occupies the upper 500 km - deep subsurface layer , where the helical flows are strongest ( fig .
[ fig : time]@xmath49 , [ fig : cross - corrhelu - b2 ] ) .
the dynamo - generated magnetic field is then transported by convective downflows into deeper layers ( fig .
[ fig : time ] @xmath50 ) .
our simulations show that the relative ratio between the vertical and horizontal field components ( field anisotropy ) changes with height ( fig .
[ fig : brms ] ) .
the vertical fields are dominant near the photosphere , but the horizontal fields become much stronger than the vertical fields in the atmosphere .
this can be explained by the loop - like topology of magnetic field ( fig .
[ fig : b063dm - lines ] ) .
perhaps , this change of the anisotropy of the vertical and horizontal fields can explain discrepancies among different observations ( e.g * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and resolve the controversy . , s. , & gilbert , a. d. 1995 , stretch , twist , fold : the fast dynamo : , lecture notes in physics monographs , volume 37 .
isbn 978 - 3 - 540 - 60258 - 3 .
springer - verlag berlin heidelberg , 1995 , 37 , doi:10.1007/978 - 3 - 540 - 44778 - 8 , r. f. , bercik , d. , & nordlund , .
2003 , in astronomical society of the pacific conference series , vol .
286 , current theoretical models and future high resolution solar observations : preparing for atst , ed .
a. a. pevtsov & h. uitenbroek , 121 ) for the case @xmath35 ( @xmath64 g ) due to the surrounding turbulent convection ( panel @xmath42 ) for @xmath65 sec after the field initialization .
panel @xmath42 ) illustrates the distribution of the vertical velocity at the photosphere .
panel @xmath49 ) shows the deformation of the initial checkerboard structure of the vertical magnetic field .
black - white patterns correspond to opposite polarity magnetic fields , saturated in this image at @xmath66 g. [ fig : checker ] ] ) and kinetic helicity ( @xmath49 ) with 5 min cadence 5 h after the field initialization .
panel @xmath50 shows the time evolution of the mean helicity at different depths for case @xmath12 ( table 1 ) with the initial seed field of @xmath0 g. [ fig : time ] ] ) and magnetic energy density ( panel @xmath49 ) in the photosphere layer ( case @xmath34 ) .
each spectrum is averaged over 20 min .
the kinetic energy spectra are compensated by a factor @xmath43 .
each curve corresponds to different moment of time : 0 , 4 , and 5 hours after the magnetic field initialization for the kinetic energy spectra ( panel @xmath42 ) ; and after 1 , 2 , 4 , and 5 hours for the magnetic energy ( panel @xmath49 ) .
[ fig : power ] ] km @xmath52 km region , where magnetic field is generated by swirling turbulent flows : @xmath42 ) vertical velocity in the photospheric layer ; @xmath49 ) the vertical magnetic field evolution showing the development of small - scale magnetic elements with opposite polarity ( bipolar magnetic structure ) ; black arrows represent the horizontal velocity field ; and c ) the electric current density ( background image ) and the squared magnitude of vorticity ( @xmath67 , contour lines ) .
this example corresponds to case @xmath34 , with the initial @xmath1 g seed field .
the white arrow points to a horizontal vortex tube discussed in the text .
[ fig : case1 ] ] ) vertical velocity ; @xmath49 ) vertical magnetic field ; @xmath50 ) magnitude of the electric current density ; @xmath57 ) and @xmath15 ) show the vertical magnetic field growth rate as a background red - blue image ; in panel @xmath57 ) contours correspond to enstrophy and in panel @xmath15 ) contour lines show the vertical component of electric current ( dashed curves for negative values ) . [ fig : case3 ] ] ( color background ) , and the horizontal velocity field ( arrows ) at different depths from the photosphere ( panel @xmath42 ) to 500 km below ( panel @xmath61 ) for the same moment of time as the first snapshot in fig . 5 .
[ fig : case32dkinhel ] ] , and squared magnetic field , @xmath68 , as a function of depth .
gray thin curves correspond to different moments of time for the case illustrated in fig . 5 , with 30 sec cadence .
black curve is the time - averaged cross - correlation function .
[ fig : cross - corrhelu - b2 ] ] | in the quiet sun , magnetic fields are usually observed as small - scale magnetic elements , ` salt and pepper ' , covering the entire solar surface . by using 3d radiative mhd numerical simulations
we demonstrate that these fields are a result of local dynamo action in the top layers of the convection zone , where extremely weak ` seed ' magnetic fields can locally grow above the mean equipartition field ( e.g. , from a @xmath0 g ` seed ' field to more than 1000 g magnetic structures ) .
we find that the local dynamo action takes place only in a shallow , about 500 km deep , subsurface layer , from which the generated field is transported into deeper layers by convection downdrafts .
we demonstrate that the observed dominance of vertical magnetic fields at the photosphere and the horizontal fields above the photosphere can be explained by multi - scale magnetic loops produced by the dynamo . |
the last few years , cooperative transmission has become widely prominent with the increases in the size of communication networks . in wireless networks , the transmitted message from a node is heard not only by its intended receiver , but also by other neighbour nodes and those neighbour nodes can use the received signals to help transmission .
they bring a cooperative transmission by acting as relays .
the relay channel first introduced by van der meulen in @xcite and it consists of a source aiming to communicate with a destination with the help of a relay . in this case
, we call the relay channel _ one - way relay channel _ or _ single relay channel_. in @xcite , cover and el gamal proposed the fundamental _ decode - forward _ ( df ) and _ compress - forward _ ( cf )
schemes for the one - way relay channels which achieve near capacity rates . in df scheme ,
the relay decodes the messages from the source and forwards them to the destination . in cf scheme
, the relay compresses received signals and forwards the compression indices .
it is proved that the df scheme is optimal for these types of channels : for physically degraded relay channels @xcite in which the output observed at the receiver is a degraded version of the channel output at the relay , for semi - deterministic channels @xcite in which the channel output at the relay is a deterministic function of the channel input at the transmitter and the channel input at the relay . the exact capacity of general relay channels is not known to date , although , there exist tight capacity approximations for a large class of networks @xcite , and schemes like df and cf achieve near - capacity rates .
the upper bound on capacity is given by the cut - set upper bound @xcite and the lower bound is given by chong et al . in @xcite .
the scheme in @xcite is a block - markov transmission scheme that is a combination of the df scheme and the cf scheme .
the one - way relay channel can be extended to the _ two - way relay channel _
, where a relay helps two users exchange their messages .
two types of two - way relay channels can be considered , that is , without a direct link between the two users , and with a direct link between the two users .
the former is a suitable model for wired communication and the latter is suitable for wireless communication .
applications of relay cooperation can be seen in increasing the capacity @xcite , combating the fading effect @xcite , mitigating the effects of interference @xcite and increasing the physical layer security @xcite .
however , df scheme has been used in numerous applications , it achieves capacity only in special few cases . all of these approaches are using random gaussian coding which is impractical for implementation .
thus , applying df scheme in a practical scenario is interesting .
one of the research areas that has such potential is lattice theory .
an @xmath0 dimensional lattice in @xmath1 is the set of integer linear combinations of a given basis with @xmath0 linearly independent vectors in @xmath1 . using lattices for communication over the real awgn channel ,
has been investigated by poltyrev @xcite .
in such a communication system , instead of the coding rate and capacity , normalized logarithmic density ( nld ) and generalized capacity @xmath2 have been introduced , respectively . using construction d of lattices @xcite , the existence of sphere - bound - achieving and capacity - achieving lattices has been proved by forney et al .
a capacity - achieving lattice code can be obtained from a capacity - achieving lattice together with a proper shaping region @xcite .
lattice codes are the euclidean - space analog of linear codes .
researchers have also studied practical lattice codes .
the search for practical implementable capacity - achieving lattices and lattice codes started by proposing low density parity check ( ldpc ) lattices @xcite . in this class of lattices , a set of nested ldpc codes and construction d of lattices @xcite are used to generate lattices with sparse parity check matrices .
another class of lattices , called low density lattice codes ( ldlc ) , introduced and investigated in @xcite .
turbo lattices employed construction d along with turbo codes to achieve capacity gains @xcite .
low density construction a ( lda ) lattices @xcite are another class of lattices with near - capacity error performance and low - complexity , low - storage decoder
. an lda lattice can be obtained from construction a @xcite with a non - binary ldpc code as its underlying code .
the use of lattice codes in relay networks has received significant attentions in recent years @xcite , @xcite , @xcite @xcite , @xcite , @xcite .
it was shown in @xcite and @xcite that lattice codes can achieve the df rates for the relay channel .
all of these achievable schemes rely on asymptotic code lengths , which is a drawback in practical implementation . recently ,
aazhang et al .
proposed a practical scheme based on ldlcs , for the real - valued , full - duplex one - way and two - way relay channels @xcite . in this work
, we propose another class of practical , efficient lattice codes , based on ldpc lattices , for the real - valued , full - duplex one - way and two - way relay channels .
the rest of this paper is organized as follows .
section [ system_model ] introduces the system models of the one - way and two - way relay channels .
section [ lattice ] presents the preliminaries on lattices and lattice codes . in section [ ldpc lattices ] , we introduce ldpc lattices .
the encoding and decoding of these lattices are also presented in this section . in section [ shaping_sec ]
, we consider the application of the ldpc lattices in the power constrained awgn channels by presenting two efficient shaping methods , based on hypercube shaping and nested lattice shaping . in section [ one_way_channel ] , we adapt our shaping algorithms that enable us to do the decomposition of the ldpc lattices into lower - rate components without loss of shaping efficiency .
then , we present a practical block markov scheme for the real - valued , full - duplex one - way relay channels , based on ldpc lattices . in section [ two_way_channel_sec ] , we present another decomposition method based on doubly nested ldpc lattices , for the two - way relay channels . finally , in section [ numerical results ] , we examine the practical performance of our proposed schemes .
section [ conclusions ] contains the concluding remarks .
the relay channel that we have considered , is a three - terminal relay channel .
first , we present the one - way relay channel , depicted in ( a ) .
the source transmits a message , which is mapped to a codeword , to both relay and destination and in the next time slot relay aids the destination by sending the part of the information of the previous time slot .
we assume a full - duplex relay which can simultaneously transmit and receive the massages . for simplicity
we suppose real - valued channels .
( 5.,-57 . )
rectangle ( 136.,-25 . ) ; ( 67.21748791714167,-32.21948106496881 ) ( 52.21748791714167,-32.21948106496881 ) ( 51.490059965882836,-32.48807198275779 ) ( 51.21748791714167,-33.21948106496881 ) ( 51.21748791714167,-36.219481064968804 ) ( 51.48476550176804,-36.95578284369824 ) ( 52.21748791714167,-37.219481064968804 ) ( 67.21748791714167,-37.219481064968804 ) ( 67.95243514205762,-36.95561431076526 ) ( 68.21748791714167,-36.219481064968804 ) ( 68.21748791714167,-33.21948106496881 ) ( 67.9514822812876,-32.484145066191346 ) cycle ; ( 22.217487917141618,-32.21948106496881 ) ( 7.21748791714162,-32.21948106496881 ) ( 6.490059965882777,-32.48807198275779 ) ( 6.21748791714162,-33.21948106496881 ) ( 6.21748791714162,-36.219481064968804 ) ( 6.484765501767987,-36.95578284369824 ) ( 7.21748791714162,-37.219481064968804 ) ( 22.217487917141618,-37.219481064968804 ) ( 22.952435142057574,-36.95561431076526 ) ( 23.217487917141618,-36.219481064968804 ) ( 23.217487917141618,-33.21948106496881 ) ( 22.951482281287547,-32.484145066191346 ) cycle ; ( 44.69811370147054,-47.22790740491434 ) ( 29.698113701470568,-47.22790740491434 ) ( 28.9706857502117,-47.49649832270332 ) ( 28.698113701470568,-48.22790740491434 ) ( 28.69811370147054,-51.22790740491434 ) ( 28.96539128609691,-51.96420918364377 ) ( 29.69811370147054,-52.22790740491434 ) ( 44.69811370147054,-52.227907404914355 ) ( 45.4330609263865,-51.9640406507108 )
( 45.698113701470554,-51.227907404914355 )
( 45.698113701470554,-48.22790740491434 )
( 45.43210806561647,-47.492571406136875 ) cycle ; ( 88.82519949396965,-32.46041161184824 ) ( 73.82519949396965,-32.46041161184824 )
( 73.09777154271082,-32.72900252963721 )
( 72.82519949396965,-33.46041161184824 )
( 72.82519949396965,-36.46041161184824 )
( 73.09247707859603,-37.19671339057766 ) ( 73.82519949396965,-37.46041161184822 )
( 88.82519949396965,-37.46041161184822 )
( 89.56014671888562,-37.19654485764469 )
( 89.82519949396965,-36.46041161184824 )
( 89.82519949396965,-33.46041161184824 )
( 89.55919385811558,-32.725075613070764 ) cycle ; ( 133.80047395779474,-32.47152777316628 ) ( 118.80047395779472,-32.47152777316628 )
( 118.07304600653588,-32.74011869095526 )
( 117.80047395779472,-33.47152777316628 )
( 117.80047395779472,-36.47152777316628 )
( 118.06775154242108,-37.207829551895706 )
( 118.80047395779472,-37.47152777316628 ) ( 133.80047395779474,-37.47152777316628 )
( 134.53542118271065,-37.207661018962746 )
( 134.80047395779474,-36.47152777316628 )
( 134.80047395779474,-33.47152777316628 )
( 134.53446832194066,-32.73619177438882 ) cycle ; ( 111.28109974212362,-47.479954113111795 ) ( 96.28109974212362,-47.479954113111795 )
( 95.55367179086474,-47.74854503090078 )
( 95.28109974212362,-48.479954113111795 )
( 95.28109974212362,-51.479954113111795 )
( 95.54837732674994,-52.21625589184122 )
( 96.28109974212362,-52.47995411311181 )
( 111.28109974212362,-52.47995411311181 )
( 112.01604696703953,-52.21608735890826 )
( 112.28109974212362,-51.479954113111795 )
( 112.28109974212362,-48.479954113111795 ) ( 112.01509410626954,-47.74461811433434 ) cycle ; ( 7.21748791714162,-32.21948106496881) ( 22.217487917141618,-32.21948106496881 ) ; ( 7.21748791714162,-37.219481064968804) ( 22.217487917141618,-37.219481064968804 ) ; ( 6.21748791714162,-33.21948106496881) ( 6.21748791714162,-36.219481064968804 ) ; plot[domain=1.5707963267948966:3.141592653589793,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 23.217487917141618,-33.21948106496881) ( 23.217487917141618,-36.219481064968804 ) ; plot[domain=3.141592653589793:4.71238898038469,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=0.:1.5707963267948966,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=-1.5707963267948966:0.,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 52.21748791714167,-32.21948106496881) ( 67.21748791714167,-32.21948106496881 ) ; ( 52.21748791714167,-37.219481064968804) ( 67.21748791714167,-37.219481064968804 ) ; ( 51.21748791714167,-33.21948106496881) ( 51.21748791714167,-36.219481064968804 ) ; plot[domain=1.5707963267948966:3.141592653589793,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 68.21748791714167,-33.21948106496881) ( 68.21748791714167,-36.219481064968804 ) ; plot[domain=3.141592653589793:4.71238898038469,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=0.:1.5707963267948966,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=-1.5707963267948966:0.,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 29.698113701470568,-47.22790740491434) ( 44.69811370147054,-47.22790740491434 ) ; ( 29.698113701470568,-52.207120204149255) ( 44.69811370147054,-52.227907404914355 ) ; ( 28.698113701470568,-48.22790740491434) ( 28.698113701470568,-51.207120204149255 ) ; plot[domain=1.5707963267948966:3.141592653589793,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 45.698113701470554,-48.22790740491434) ( 45.698113701470554,-51.227907404914355 ) ; plot[domain=3.141592653589793:4.71238898038469,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=0.:1.5707963267948966,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=-1.5707963267948966:0.,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 23.217487917141618,-34.719481064968804 )
( 51.21748791714167,-34.72874555724628 ) ; ( 13.580557760431486,-31.894304393840127 ) node[anchor = north west ] @xmath3 ; ( 58.151882220748504,-31.74423259431044 ) node[anchor = north west ] @xmath4 ; ( 35.49104049176577,-47.05155614633853 ) node[anchor = north west ] @xmath5 ; ( 22.217487917141618,-37.219481064968804 )
( 35.180406011112964,-47.22790740491434 ) ; ( 39.13768193920707,-47.22790740491434 )
( 52.21748791714167,-37.219481064968804 ) ; ( 35.94125589035484,-28.592724804187007 ) node[anchor = north west ] @xmath6 ; ( 23.03508113080176,-40.14825336797292 ) node[anchor = north west ] @xmath7 ; ( 47.49678445414073,-39.848109768913545 ) node[anchor = north west ] @xmath8 ; ( 35.94125589035484,-53.054428127526016 ) node[anchor = north west ] ( a ) ; ( 73.80047395779467,-32.47152777316628) ( 88.80047395779465,-32.47152777316628 ) ; ( 73.80047395779467,-37.47152777316628) ( 88.80047395779465,-37.47152777316628 ) ; ( 72.80047395779467,-33.47152777316628) ( 72.80047395779467,-36.47152777316628 ) ; plot[domain=1.5707963267948966:3.141592653589793,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 89.80047395779465,-33.47152777316628) ( 89.80047395779465,-36.47152777316628 ) ; plot[domain=3.141592653589793:4.71238898038469,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=0.:1.5707963267948966,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=-1.5707963267948966:0.,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 118.80047395779472,-32.47152777316628) ( 133.80047395779474,-32.47152777316628 ) ; ( 118.80047395779472,-37.47152777316628) ( 133.80047395779474,-37.47152777316628 ) ; ( 117.80047395779472,-33.47152777316628) ( 117.80047395779472,-36.47152777316628 ) ; plot[domain=1.5707963267948966:3.141592653589793,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 134.80047395779474,-33.47152777316628) ( 134.80047395779474,-36.47152777316628 ) ; plot[domain=3.141592653589793:4.71238898038469,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=0.:1.5707963267948966,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=-1.5707963267948966:0.,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 96.28109974212362,-47.479954113111795) ( 111.28109974212362,-47.479954113111795 ) ; ( 96.28109974212362,-52.47995411311181) ( 111.28109974212362,-52.47995411311181 ) ; ( 95.28109974212362,-48.479954113111795) ( 95.28109974212362,-51.479954113111795 ) ; plot[domain=1.5707963267948966:3.141592653589793,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 112.28109974212362,-48.479954113111795) ( 112.28109974212362,-51.479954113111795 ) ; plot[domain=3.141592653589793:4.71238898038469,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=0.:1.5707963267948966,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; plot[domain=-1.5707963267948966:0.,variable=](1.*1.*cos(r)+0.*1.*sin(r),0.*1.*cos(r)+1.*1.*sin(r ) ) ; ( 81.80047395779465,-37.47152777316628 ) ( 95.28109974212362,-49.4315279994261 ) ; ( 95.28109974212362,-50.40898585229985 ) ( 80.80047395779465,-37.47152777316628 ) ; ( 112.28109974212362,-50.449155353102896 ) ( 126.80047395779474,-37.47152777316628 ) ; ( 125.80047395779471,-37.47152777316628 ) ( 112.28109974212362,-49.444917833027134 ) ; ( 89.80047395779465,-34.47152777316628 ) ( 117.80047395779472,-34.47152777316628 ) ; ( 117.80047395779472,-35.47152777316628 ) ( 89.80047395779465,-35.47152777316628 ) ; ( 79.46207775396404,-32.1944479928995 ) node[anchor = north west ] @xmath9 ; ( 124.03340221428105,-31.894304393840127 ) node[anchor = north west ] @xmath10 ; ( 102.42306308200614,-47.3516997453979 ) node[anchor = north west ] @xmath5 ; ( 102.57313488153582,-28.292581205127632 ) node[anchor = north west ] @xmath11 ; ( 102.57313488153582,-34.74566858490418 ) node[anchor = north west ] @xmath12 ; ( 91.3177499168093,-40.14825336797292 ) node[anchor = north west ] @xmath13 ; ( 83.21387274220622,-43.14968935856666 ) node[anchor = north west ] @xmath14 ; ( 111.72751465284674,-39.247822570794796 ) node[anchor = north west ] @xmath15 ; ( 119.9814636269795,-42.54940216044791 ) node[anchor = north west ] @xmath16 ; ( 102.57313488153582,-53.054428127526016 ) node[anchor = north west ] ( b ) ; let @xmath17 and @xmath18 denote the signals transmitted by the source and the relay .
let @xmath19 and @xmath20 denote the signals received at the relay and the destination .
the received signals are @xmath21 where @xmath22 , @xmath23 .
moreover , @xmath24 , @xmath25 , and @xmath26 are the channel gains between source , relay , and destination , which follow the usual path - loss model .
in particular , we take the distance between source and destination to be unity , @xmath27 and @xmath28 to be the distance between source to relay and relay to destination , and @xmath29 and @xmath30 to be the corresponding path - loss exponents .
we constrain the source and relay transmissions to have average power no greater than @xmath31 and @xmath32 . in general
, the capacity of this channel is unknown however , the df scheme which is proposed in @xcite , achieves the following inner bound : @xmath33 this rate can be achieved via _ block markov encoding_. after decoding the message , the relay re - encodes the message and transmits the corresponding codeword in the next block .
the lattice - coding version of block markov encoding is proposed in @xcite .
it is proved theoretically that it can achieve the decode - and - forward rates .
the results of @xcite suggest that structured lattice codes may be used to outperform , random gaussian codes in general gaussian networks .
the authors of @xcite have applied this scheme and designed a family of practically implementable ldlc lattice codes for relay channels . in this paper
we present another family of lattice codes which are amenable to practical implementation in block markov schemes .
@#1@@@(#1)@italiccorr next , we present the full - duplex gaussian two - way relay channel , as depicted in ( b ) .
the two sources @xmath9 and @xmath10 exchange their messages , which are mapped to codewords and transmitted over the wireless medium .
the relay node @xmath5 overhears the noisy superposition of signals transmitted from sources and makes its own transmissions to facilitate communications .
this channel can be modeled as follows @xmath34 where the noise components are @xmath22 , @xmath35 and @xmath36 .
similar to the one - way relay channel model , @xmath37 , @xmath38 , @xmath15 , @xmath13 , @xmath14 and @xmath16 are the channel gains which follow the usual path - loss model .
we constrain the sources @xmath9 , @xmath10 and relay @xmath5 transmissions to have average powers no greater than @xmath39 , @xmath40 and @xmath32 , respectively .
the capacity of this channel is unknown , but a df scheme was presented in @xcite that achieves rate pairs @xmath41 satisfying ( [ rate_bound2 ] ) , @xmath42^{+ } \right\},\end{aligned}\ ] ] in which @xmath43^{+}\triangleq \max \left\{x,0\right\}$ ] , @xmath44 , and @xmath45 .
[ lattice ] a discrete , additive subgroup @xmath46 of the @xmath47-dimensional real space @xmath1 is a lattice .
every lattice @xmath46 has a basis @xmath48 where every @xmath49 can be represented as an integer linear combination of vectors in @xmath50 .
the rank of the lattice is @xmath0 and its dimension is @xmath47 . if @xmath51 , the lattice is called a full - rank lattice . in this paper , we consider full - rank lattices .
the matrix @xmath52 with @xmath53 as rows , is a generator matrix for the lattice .
the matrix @xmath54 is a gram matrix for the lattice .
the determinant of the lattice @xmath55 is defined to be the determinant of the matrix @xmath56 and the volume of the lattice is defined as @xmath57 . a voronoi cell
@xmath58 is the set of those points of @xmath59 that are at least as close to @xmath60 as to any other point in @xmath46 .
we call the voronoi region associated with the origin , the fundamental voronoi region of @xmath46 , denoted by @xmath61 or @xmath62 .
we say that a lattice @xmath63 is nested in @xmath64 if @xmath65 . using nested lattices in @xmath59 ,
define the codebook @xmath66 which has the rate @xmath67 suppose that the points of a lattice @xmath46 are sent over an unconstrained awgn channel with noise variance @xmath68 .
the volume - to - noise ratio ( vnr ) of lattice @xmath46 is defined as @xmath69 for a large @xmath0 , vnr is the ratio of the normalized volume of @xmath46 to the normalized volume of a noise sphere of squared radius @xmath70 which is defined as generalized signal - to - noise ratio ( snr ) in @xcite and @xmath71 in @xcite .
let @xmath72 be the transmitted vector on the unconstrained awgn channel , then the received vector @xmath73 can be written as @xmath74 , where @xmath75 and its components are independently and identically distributed ( i.i.d . )
gaussian random variables with zero mean and variance @xmath68 .
the probability of correct decoding , under maximum likelihood decoding , is given by @xmath76 where @xmath77 is the euclidean norm of @xmath60 .
a lattice quantizer is a map @xmath78 for some lattice @xmath79 .
if we use the nearest - neighbor quantizer @xmath80 , then the quantization error @xmath81 .
let @xmath82 be uniformly distributed over the voronoi region @xmath62 .
then , the _ second moment per dimension _ of @xmath46 is @xmath83=\frac{1}{n}\frac{1}{\textrm{vol}(\lambda)}\int_{\mathcal{v}(\lambda)}\|\mathbf{x}_e\|^2d\mathbf{x}_e.\ ] ] a _ lattice constellation _ @xmath84 is the finite set of points in a lattice translate @xmath85 that lies within a compact bounding region @xmath86 of @xmath0-dimensional real space @xmath59 .
the key geometric properties of the region @xmath86 are its _ volume _ @xmath87 and the _ average energy _ @xmath88 per dimension of a uniform probability density function over @xmath86 ( see , e.g. @xcite and @xcite ) : @xmath89 the _ normalized second moment _ of @xmath86 is defined as @xmath90 the normalized second moment of any @xmath0-cube centered at the origin is @xmath91 . the _ shaping gain _
@xmath92 of @xmath86 , measures the decrease in average energy of @xmath86 relative to a baseline region , namely , an interval @xmath93 $ ] or an @xmath0-cube @xmath93^n$ ] , where @xmath94 is related to the @xmath87 @xcite .
the definition of shaping gain is @xmath95 the optimum @xmath0-dimensional shaping region is an @xmath0-sphere @xcite .
the key geometrical parameters of an @xmath0-sphere @xmath96 of radius @xmath97 for an even @xmath0 are @xcite : @xmath98 the shaping gain of an @xmath0-sphere is a function of the dimension @xmath0 . when @xmath0 approaches infinity the shaping gain approaches the _ ultimate shaping gain _
@xmath99 ( @xmath100db ) .
the _ shaping loss _
@xmath101 of a shaping region @xmath86 with respect to an @xmath0-dimensional sphere , where @xmath0 is even , based on ( [ spherev])-([sphereg ] ) , is @xcite @xmath102 the shaping loss is greater than or equal to @xmath103 . if we form the intersection @xmath104 of a lattice with a shaping region @xmath105 , we would expect to obtain a code with about @xmath106 codewords @xcite .
in fact , by using minkowski - hlawka theorem , it is proved that the value @xmath106 is correct in the average over a suitable set of lattices based on codes @xcite .
the rate of the code @xmath107 is approximately @xmath108
[ ldpc lattices ] there exist many ways to construct lattices based on codes @xcite . here
we mention two of them .
the first one is construction a and the other one is construction d. assume that @xmath109 is a linear code over @xmath110 where @xmath111 is a prime number , i.e. @xmath112 . a lattice @xmath46 based on construction a @xcite can be derived from @xmath109 as follows @xmath113 where @xmath114 is the embedding function .
in this work , we are particularly interested in lattices with @xmath115 .
construction d converts a set of parity checks defined by a family of nested codes @xmath116 , into congruences for a lattice @xcite .
the number @xmath117 is called the level of the construction .
an ldpc lattice @xmath118 can be constructed from construction d and a number of nested binary ldpc codes .
more detail about the structure and decoding of these lattices can be found in @xcite .
if we consider one code as underlying code of construction d , which means @xmath119 , construction a is obtained ( * ? ? ?
* proposition 1 ) . in this case
, construction a ldpc lattices or 1-level ldpc lattices @xcite will be obtained . in this paper
, we refer to them as ldpc lattices without mentioning the level of the construction .
an ldpc lattice @xmath46 is a lattice constructed based on construction a or d along with one binary ldpc code @xmath109 as its underlying code .
equivalently , @xmath120 is in @xmath46 if and only if @xmath121 , where @xmath122 is the parity check matrix of @xmath109 .
the generator matrix of construction a lattice @xmath46 using the underlying code @xmath123 is of the form @xmath124,\end{aligned}\ ] ] where @xmath125 $ ] is the generator matrix of @xmath109 in systematic form , @xmath126 is the rank of @xmath109 , @xmath127 and @xmath128 , are identity and all zero square matrices of size @xmath126 , respectively . the practical
encoding and decoding of ldpc lattices , both with linear complexity in the dimension of the lattice , has been addressed in @xcite . in this paper , we consider a translated and scaled version of the lattice @xmath46 , generated by ( [ eq12 ] ) , as suggested in @xcite and @xcite . in the sequel , we present the decoding of these scaled and translated versions of ldpc lattices , which is proposed in @xcite and it is obtained by combining the suggested decoding method in @xcite , and the decoding of binary ldpc codes .
construction and decoding of these new lattices can be done using the following steps .
first , convert the codewords of @xmath129 $ ] binary code @xmath109 into @xmath130 notation ( convert @xmath131 to @xmath132 and @xmath103 to @xmath103 ) @xcite which produces a set @xmath133 consisting of the vectors of the form @xmath134 the set of points in ( [ newlattice ] ) strictly speaking , is not a lattice , but the translate of a lattice by the vector @xmath135 .
the regular addition of vectors of the form ( [ newlattice ] ) will not be of the same form .
however , we can show that @xmath133 is closed under following addition . for any @xmath136
, we have @xcite @xmath137 the encoding of an integer row vector @xmath138 can be done as follows @xmath139 where @xmath140 is encoding function and @xmath141 is defined as ( [ eq12 ] ) .
let @xmath142 be the transmitted lattice vector and @xmath143 be the received vector from awgn channel @xmath144 where @xmath145 and @xmath109 is a binary ldpc code in @xmath130 notation , @xmath146 and @xmath147 .
first , we decode @xmath148 and next we find @xmath149 .
the proposed algorithm in @xcite is similar to the sum - product algorithm ( spa ) for ldpc codes in message passing structure @xcite .
the inputs are the log likelihood ratios ( llr ) for the a priori message probabilities from each channel .
the estimation of the llr vector @xmath150 for ldpc lattices is proposed in @xcite as follows @xmath151 where @xmath152 is the nearest integer to @xmath153 .
input the llr vector @xmath154 to spa decoder of ldpc codes and consider @xmath155 as the output of this decoder .
convert @xmath155 to @xmath130 notation and call the obtained vector @xmath156 .
estimate @xmath157 as follows @xmath158 then , @xmath159 is the final decoded lattice vector .
the complexity of this decoding algorithm is significantly lower than other lattices with practical decoding algorithm like lda lattices @xcite and ldlcs @xcite .
the decoding algorithm of ldlcs first proposed in @xcite which has complexity @xmath160 , where @xmath161 is the resolution and its typical value is @xmath162 , @xmath0 is the dimension of lattice , @xmath163 is the number of iterations and @xmath164 is the average code degree . then , in @xcite , a new algorithm proposed with lower complexity @xmath165 , where @xmath166 is the number of replications , @xmath0 , @xmath163 and @xmath164 are similar to above .
proposed typical value for @xmath166 is @xmath167 and for @xmath168 it is @xmath169 or @xmath170 .
the computational complexity of the proposed decoding algorithm in @xcite is only @xmath171 , which is significantly lower than the complexity of ldlcs decoding .
the main purpose of this section , is constructing practically implementable lattice constellations . as we mentioned in previous section , definition of lattice constellations
entails finding the intersection of shaping region and a translated infinite lattice .
thus , translating by @xmath172 in the proposed structure has no inconsistency with the definition of lattice codes . in the next section , we present efficient shaping methods to obtain a family of lattice codes based on ldpc lattices .
in practical channels , there exists a power constraint which is needed to be fulfilled .
this entails selecting a group of lattice points with a bounded norm . in theoretical approaches ,
the coding lattice is intersected with a spherical shaping region to produce an efficient , power - constrained lattice code
. however , spherical shaping has high computational complexity both for encoding and decoding . in @xcite , several efficient and practical shaping algorithms proposed for ldlc lattices . here
, we present an efficient and practical shaping algorithm for ldpc lattices based on hypercube shaping method . in order to perform shaping , restrict the integer vector @xmath173 to the following finite constellation @xmath174 shape the lattice codeword @xmath175 by translating each @xmath176 by an integer multiple of @xmath177 , @xmath178 .
this is equivalent to transmitting a lattice point @xmath179 as follows @xmath180 where @xmath181 is a diagonal matrix and the new integer vector is @xmath182 .
the choice of integer vector @xmath183 , depends on the employed shaping method . in hypercube
shaping we choose @xmath183 such that the new lattice codeword components are constrained to lie in the hypercube , i.e. @xmath184 , for @xmath185 .
the authors of @xcite used a triangular structure for @xmath122 , the inverse of their generator matrix @xmath56 , with unit diagonal elements . in this case ,
hypercube shaping is straightforward , and @xmath186 and @xmath187 will be found recursively .
instead , we use the generator matrix of the lattice for shaping . in our case
, the generator matrix beside the triangular structure , has a simpler form which helps us to obtain the components of @xmath183 directly .
since we have used a translated and scaled lattice , we shape the lattice vectors inside the hypercube around the origin of coordinate .
then , we scale them by factor @xmath169 and translate them by the all @xmath132 vector . to this end , we need to solve the following system of linear equations @xmath188,\ ] ] by choosing an integer vector @xmath189 such that @xmath184 for @xmath185 . from ( [ shaping_system ] )
, we have the following equations @xmath190 by choosing @xmath191 , we have @xmath192 and @xmath193 , for @xmath194 . for @xmath195
, we have the following inequalities @xmath196 or equivalently @xmath197 this interval contains only one integer number which is the unique solution @xmath198 note that , after finding the shaped lattice codeword as discussed above , we must scale it by factor @xmath169 and then translate it by @xmath199 .
then , the shaped vectors of @xmath200 will be uniformly distributed over hypercube @xmath201 , where @xmath202 since , the lattice @xmath203 and shaping region @xmath204 are translated by the same vector and @xmath205 , based on ( [ rate_loeliger ] ) , the rate of our scheme is @xmath206 algorithm [ unshaped ] explains the method of obtaining original information @xmath173 from shaped lattice codeword @xmath179 .
the complexity of this algorithm is @xmath207 , where @xmath164 is the average number of nonzero elements in a row of @xmath141 .
@xmath208 @xmath209 .
@xmath210 * return * @xmath211 . despite of its low complexity nature ,
hypercube shaping suffers a performance loss of @xmath100db in high dimensions compared to optimal hypersphere shaping @xcite .
thus , we consider nested lattice shaping , which is suboptimal but it offers more shaping gains comparing to hypercube shaping @xcite .
first , limit the rate of the code by restricting the integer row vector @xmath173 to take values from a finite constellation in which @xmath212 for each @xmath213 .
similar to the hypercube shaping , let @xmath214 . in this case
, we choose the vector @xmath183 as follows @xmath215 choosing @xmath183 that minimizes @xmath216 is equivalent to finding the nearest lattice point of the scaled lattice @xmath217 to the non - shaped lattice point @xmath60 .
therefore , the codewords will be uniformly distributed along the voronoi cell of the coarse lattice @xmath217 .
thus , the rate of the code is @xmath218 the complexity of solving ( [ nested_lattice ] ) is exponential in the dimension of lattice , even by restricting the components of @xmath173 . using the triangular structure of the parity check matrix @xmath122 , the authors of @xcite suggested a tree search with affordable complexity for shaping their lattices .
practically , their tree search can be done with simple sub - optimal sequential decoders such as the @xmath168-algorithm @xcite . following their method
, we present a nested lattice shaping by using the triangular structure of our generator matrix .
our algorithm starts from the first row of @xmath141 , and sequentially goes down along the rows .
the input at row @xmath219 is a list of up to @xmath168 candidate sequences @xmath220 for @xmath221 . for @xmath222 the list
is initialized with a single empty sequence .
each of these @xmath168 sequences will be extended to @xmath223 , for @xmath224 , where @xmath225 takes all possible values for @xmath186 .
the algorithm assigns to each extended sequence a respective score @xmath226 .
the scores are sorted , and the @xmath168 sequences with smallest score are retained as input to the next row . finally , we choose @xmath183 as the sequence with smallest score after processing of the last row @xmath227 .
the storage and the computational complexity of this algorithm are determined by the value of @xmath168 , as @xmath228 . using an @xmath168-algorithm with @xmath229 , nested lattice shaping is equivalent to hypercube shaping , where for @xmath230 , the algorithm is a full exponential tree search that finds the exact solution for @xmath183 . the shaping gain and shaping loss of ldpc lattice codes
are presented in .
the results are obtained by using monte carlo simulation and @xmath231 assumed to be @xmath232 .
an @xmath168-algorithm with @xmath233 has been applied for nested lattice shaping .
the results show that increasing the rate of the code , used as underlying code of the lattice , declines the shaping gain for hypercube shaping but it improves the shaping gain for nested lattice shaping . increasing the dimension of the lattice also improves the shaping gain for both of the applied shaping methods .
thus , the reasonable shaping gains can be obtained by applying nested lattice shaping for high dimensional ldpc lattices , which have high rate ldpc codes as their underlying codes .
[ table1 ] cc|c|c||c|c| & & + & @xmath234 & shaping gain & shaping loss & shaping gain & shaping loss + & @xmath232 & @xmath235 & @xmath236 & @xmath237 & @xmath238 + & @xmath232 & @xmath239 & @xmath240 & @xmath241 & @xmath242 + & @xmath232 & @xmath243 & @xmath244 & @xmath245 & @xmath246 + & @xmath232 & @xmath247 & @xmath248 & @xmath249 & @xmath250 + & @xmath232 & @xmath251 & @xmath252 & @xmath253 & @xmath254 + & @xmath232 & @xmath255 & @xmath256 & @xmath257 & @xmath258 + & @xmath232 & @xmath259 & @xmath260 & @xmath261 & @xmath262 + & @xmath232 & @xmath263 & @xmath264 & @xmath265 & @xmath266 + & @xmath232 & @xmath267 & @xmath268 & @xmath269 & @xmath270 +
in this section , we construct a lattice - coding scheme base on ldpc lattice codes that achieves the decode - and - forward bound of ( [ rate_bound ] ) .
we present our scheme by following the scheme of @xcite for ldlcs .
all steps of this scheme are rephrased due to the inherent differences between ldpc lattice codes and ldlcs . based on block markov encoding and decomposition of the full lattice codebook into lower - rate codebooks
, we can propose an encoding / decoding scheme for implementing ldpc lattice codes on one - way relay channels .
let @xmath138 be the information vector , where each element @xmath176 is drawn from the finite constellation of ( [ constelation ] ) .
then , we define the @xmath271 element of the _ resolution component _ @xmath272 as @xmath273 where @xmath274 is a @xmath126-element subset of @xmath275 , whose members are chosen randomly .
let @xmath276 be the _ vestigial component _ vector .
define the the @xmath271 element of @xmath277 as follows @xmath278 based on ( [ constelation_r ] ) and ( [ constelation_v ] ) , define the resolution lattice codeword @xmath279 and the vestigial lattice codeword @xmath280 , where @xmath140 is defined by ( [ encoding ] ) .
it can be easily checked that the original lattice codeword @xmath281 is the sum of its resolution and vestigial components , that is @xmath282 in preceding subsection , we used unconstrained powers for codewords @xmath60 , @xmath283 and @xmath284 , due to the fact that we still have not enforced shaping for the lattice . linear decomposition is straightforward in the unconstrained power situation , but it is not trivial with power constrained scenario .
we use the altered version of the proposed method in @xcite with the proposed hypercube shaping method of section [ ldpc lattices ] .
the steps of this method are as follows : 1 .
map the information integer vector @xmath173 to @xmath285 such that the @xmath271 element of @xmath285 is @xmath286 and @xmath287 the power - constrained lattice codeword is then given as @xmath288 .
2 . decompose the original integer information vector @xmath173 as in ( [ constelation_r ] ) and ( [ constelation_v ] ) .
3 . map the resolution component @xmath272 to @xmath289 such that the new integer vector results in a power - constrained codeword : @xmath290 where @xmath291 and @xmath292 are the @xmath271 elements of @xmath272 and @xmath289 , respectively .
for hypercube shaping , @xmath293 can be written as , @xmath294 where @xmath295 .
then , the mapped lattice codeword is @xmath296 .
4 . for the sake of preserving the linearity of the lattice decomposition , we map the vestigial information vector @xmath277 to @xmath297 such that the @xmath298 element of @xmath297 is given as , @xmath299 where @xmath186 and @xmath293 are given in ( [ si2 ] ) and ( [ sri ] ) , respectively .
then , the vestigial codeword @xmath300 can be obtained as @xmath301 .
the primary information integers @xmath176 , @xmath291 and @xmath302 can be recovered from @xmath303 , @xmath291 and @xmath302 by modulo @xmath177 operation .
it must be noted that the lattice codeword @xmath179 and the resolution lattice codeword @xmath304 respect the power constraint but , the vestigial lattice codeword @xmath300 may not fulfil the power constraint in general .
however , this is not a problem for the considered relay network since we do not transmit the vestigial information alone . in order to obtain further shaping gain ,
we employ nested lattice shaping in our decomposition . in this case , we find the vectors @xmath183 and @xmath305 such that @xmath306 after finding @xmath183 and @xmath305 , the vestigial codeword can be found as @xmath307 . here
, we explain the implementation of the encoding and the decoding of ldpc lattice codes , for one - way relay network .
let @xmath308 $ ] be the full - rate power - constrained lattice codeword associated with @xmath309 block , and let @xmath310 $ ] be the resolution codeword decoded at the relay from the @xmath311^{th}$ ] block .
the source and the relay transmit @xmath312 $ ] and @xmath313 $ ] during the @xmath309 block . at block
@xmath314 , the source sends @xmath315 $ ] and the relay sends nothing . at blocks @xmath316 , destination receives the superposition of signals from the source and the relay at blocks @xmath163 and @xmath317 , respectively . at block @xmath318 ,
destination only receives the resolution information from the relay since the source has no information to send .
the source transmits @xmath319 symbols over @xmath320 channel use , so the encoding rate is close to @xmath5 for large @xmath321 .
more detail about implementing block markov encoding using lattices can be found in @xcite and @xcite . following the outline that proposed in (
* theorem 1 ) , our decoding contains three stages . at the first stage , the relay decodes @xmath179 , in the next stage , the destination decodes @xmath304
, finally , the destination uses @xmath304 to decode @xmath300 .
first , we consider the decoding at the relay .
the received signal by the relay at the @xmath309 block is @xmath322=h_{sr}\sqrt{p_s}\mathbf{x}'[t]+\mathbf{z}_r[t].\ ] ] we start by scaling @xmath323 $ ] by the factor @xmath324 .
then , we use iterative decoder of ldpc lattices to obtain full - rate lattice codeword @xmath325 $ ] @xmath326=\textrm{dec}_{\textrm{ldpcl}}\left ( \frac{\mathbf{y}_r[t]}{\sqrt{p_s}h_{sr}},\frac{n_r}{\sqrt{p_s}h_{sr } } \right),\ ] ] where @xmath327 is the iterative decoding algorithm of ldpc lattices described in section [ ldpc lattices ] .
we use @xmath143 and @xmath68 to estimate llr vector in ( [ llr ] ) .
the resulting information vector is denoted by @xmath328 $ ] and can be obtained as @xmath328=\textrm{mod}(\hat{\mathbf{x}}'[t],\mathbf{l},\mathbf{g}_{\lambda}^{-1})$ ] , where @xmath329 is @xmath330.\end{aligned}\ ] ] the relay estimates the resolution information @xmath331 $ ] from @xmath328 $ ] and ( [ constelation_r ] ) , from which it finds the estimation of shaped resolution codeword @xmath332 $ ] . in this case , the decoding error is @xmath333=\mathbf{x}_r'[t]-\hat{\mathbf{x}}_r'[t]$ ] .
the received signal at the destination during @xmath334 block is @xmath335=h_{rd}\sqrt{p_r}\hat{\mathbf{x}}_r'[t]+h_{sd}\sqrt{p_s}\mathbf{x}'[t+1]+\mathbf{z}_d[t+1].\end{aligned}\ ] ] replace @xmath332 $ ] by @xmath336-\mathbf{e}_r[t]$ ] in @xmath337 $ ] . in the destination ,
treat @xmath338+\mathbf{z}_d[t+1]-h_{rd}\sqrt{p_r}\mathbf{e}_r[t]$ ] as noise and decode the resolution information vector @xmath339 $ ] @xmath340=\left\lfloor \mathcal{d}\left(\textrm{dec}_{\textrm{ldpcl}}\left ( \frac{\mathbf{y}_d[t+1]}{\sqrt{p_r}h_{rd}},\frac{n_d}{\sqrt{p_r}h_{rd } } \right)\right)\mathbf{g}_{\lambda}^{-1}\right\rceil,\end{aligned}\ ] ] where @xmath341 .
then , the relay obtains @xmath342=\textrm{mod}(\tilde{\mathbf{x}}_r'[t],\mathbf{l},\mathbf{g}_{\lambda}^{-1})$ ] , from which it finds the unshaped resolution codeword @xmath343 $ ] .
the decoding errors are @xmath344 & = & \mathbf{x}_r'[t]- \tilde{\mathbf{x}}_r'[t ] , \nonumber \\
\mathbf{e}_{d_1}[t ] & = & \mathbf{x}_r[t ] -\tilde{\mathbf{x}}_r[t].\end{aligned}\ ] ] now , the destination knows the unshaped resolution codeword @xmath343 $ ] and the shaped resolution codeword @xmath339 $ ] .
next , the receiver turns to @xmath345 $ ] , which can be written as @xmath346&=&h_{rd}\sqrt{p_r}(\mathbf{x}_r'[t-1]-\mathbf{e}_r[t-1])+ \nonumber\\ & & \quad\quad\quad\quad\quad\quad\quad h_{sd}\sqrt{p_s}\mathbf{x}'[t]+\mathbf{z}_d[t].\end{aligned}\ ] ] from @xmath345 $ ] , we know @xmath347 $ ] . by using the linearity property ( [ decomposition ] ) ,
we can rewrite ( [ y_dt ] ) as follows @xmath348&=&h_{rd}\sqrt{p_r}(\mathbf{x}_r'[t-1]-\mathbf{e}_r[t-1])+ \mathbf{z}_d[t]+\nonumber\\ & & h_{sd}\sqrt{p_s}(\mathbf{x}_r'[t]+\mathbf{x}_v'[t]+(1,\ldots , 1)).\end{aligned}\ ] ] now , we subtract the decoded resolution information @xmath349+(1,\ldots , 1))+h_{rd}\sqrt{p_r}\tilde{\mathbf{x}}_r'[t-1]$ ] from ( [ y_dt2 ] ) to obtain @xmath350=h_{sd}\sqrt{p_s}\mathbf{x}_v'[t]+\mathbf{e}_{d_2}[t]+\mathbf{z}_d[t],\end{aligned}\ ] ] where @xmath351=h_{rd}\sqrt{p_r}(\mathbf{e}_{d_1}'[t-1]-\mathbf{e}_r[t-1])+ h_{sd}\sqrt{p_s}\mathbf{e}_{d_1}'[t].\end{aligned}\ ] ] then , we use @xmath352 $ ] in ( [ y_dt3 ] ) to decode the vestigial information as follows @xmath353=\textrm{dec}_{\textrm{ldpcl}}\left ( \frac{y_d'[t]}{\sqrt{p_s}h_{sd}},\frac{n_d}{\sqrt{p_s}h_{sd } } \right).\end{aligned}\ ] ] once we have decoded both the resolution and vestigial lattice codeword , the destination can find the desired lattice codeword by @xmath354=\tilde{\mathbf{x}}_v'[t]\oplus\tilde{\mathbf{x}}_r'[t]=\tilde{\mathbf{x}}_v'[t]+\tilde{\mathbf{x}}_r'[t]+(1,\ldots,1)$ ] . then , the desired integer vector can be obtained as @xmath355= \textrm{mod}(\tilde{\mathbf{x}}'[t],\mathbf{l},\mathbf{g}_{\lambda}^{-1})$ ] .
in this section , we present a practical block markov encoding scheme for the two - way relay channel based on ldpc lattice codes .
this scheme achieves the decode - and - forward bound in ( [ rate_bound2 ] ) and it is proposed in @xcite for ldlcs . due to the structural differences between ldpc lattices and ldlcs , as well as , differences between our shaping methods and theirs , all steps of this scheme are rephrased for ldpc lattice codes .
our applied scheme relies on doubly - nested lattice codes in which the two sources use different codebooks with different transmit powers .
let @xmath356 and @xmath357 be the shaping and coding lattices for @xmath9 and @xmath10 , respectively . consider the codebooks @xmath358 and @xmath359 .
it is also assumed by the authors of @xcite that @xmath360 , @xmath361 , and @xmath362 .
there is another lattice , which is referred to as _ meso - lattice _ , that partitions the lattice codebook into lower - rate constituent codebooks @xcite .
the meso - lattice @xmath363 , is nested according to @xmath364 , where @xmath365 and @xmath366 are also nested lattices .
define the resolution codebooks @xmath367 and @xmath368 .
then , the vestigial codebooks are @xmath369 and @xmath370 .
let @xmath371 be the rates of @xmath372 , respectively . then , as before , @xmath373 and @xmath374 .
thus , each full - rate codeword can be decomposed into a unique modulo sum of resolution and vestigial codewords .
it is proved that the preceding code construction achieves near - capacity rates @xcite .
let @xmath138 be the information vector , where each element @xmath176 of @xmath173 is drawn from finite constellations @xmath375 and @xmath376 for hypercube shaping and nested lattice shaping , respectively .
the @xmath271 element of the resolution component @xmath272 is @xmath377 where @xmath378 , for some @xmath379 .
thus , the @xmath271 element of @xmath380 lies in the finite constellation @xmath381 and the rate of the obtained codebook @xmath382 can be acquired via ( [ rate ] ) or ( [ rate_nested ] ) , depending on the applied shaping method , by replacing @xmath177 s with @xmath383 s .
the vestigial component is defined as follows @xmath384 if the employed shaping method is hypercube shaping , it can be shown that the @xmath271 element of @xmath385 lies in the following finite constellation @xmath386 if the employed shaping method is nested lattice shaping , the @xmath271 element of @xmath385 lies in the finite constellation @xmath387 .
the rate of the vestigial codebook is approximately @xmath388 . due to the aforementioned assumptions
, each codeword @xmath389 decomposes into its resolution component @xmath390 and its vestigial component @xmath391 as @xmath392 . the first step in shaping the lattice codeword @xmath60 is translating @xmath176 by an integer multiple of @xmath177 .
then , following the same procedure as in section [ shaping_sec ] , the shaped lattice point is obtained .
the shaped lattice codeword is @xmath393 , where @xmath183 is given in ( [ si ] ) and ( [ nested_lattice ] ) , for hypercube shaping and nested lattice shaping , respectively .
next , we adapt the shaping methods to the resolution codewords in such a way that the decomposition of the lattice codebook remains linear .
we shape the resolution component @xmath380 to @xmath394 as follows @xmath395 where @xmath176 and @xmath396 are the @xmath271 elements of @xmath173 and @xmath394 , respectively , and @xmath397 we choose the elements of @xmath398 according to @xmath399 for hypercube shaping . for nested lattice shaping , we consider @xmath400 , where @xmath398 is given by ( [ sr_s_nested2 ] ) .
indeed , the @xmath401 function is regular modulo operation , when the employed shaping method is nested lattice shaping .
then , the shaped resolution component is given by @xmath402 . in order to preserve the modulo linearity of the lattice decomposition , we map the vestigial integer vector @xmath385 to @xmath403
as follows @xmath404 where @xmath183 is given in ( [ si ] ) and ( [ nested_lattice ] ) , for hypercube shaping and nested lattice shaping , respectively . according to this decomposition , we have @xmath405 in general .
however , the decomposition preserves componentwise modulo linearity , that is , @xmath406 .
then , the vestigial codeword is @xmath407 .
it can be easily shown that @xmath408 .
it should be noted that the vestigial lattice codeword @xmath300 may not fulfil the power constraint , which is not a problem for the considered relay network , since we do not transmit the vestigial information alone .
the original information vectors @xmath409 can be recovered from @xmath410 by using algorithm [ unshaped ] with @xmath411 .
similarly , the resolution information vector @xmath380 can be recovered from @xmath394 by using algorithm [ unshaped ] with @xmath412 . here , we present the implementation of encoding and decoding for two - way relay network using ldpc lattice codes .
first , we describe the encoding scheme and then we discuss the decoding schemes at each node .
the two sources employ ldpc lattice codes as described in previous subsection .
the full - rate information vector , resolution information vector , vestigial information vector , full - rate codeword , resolution codeword , and vestigial codeword of source @xmath219 are denoted by @xmath413 and @xmath414 , respectively , for @xmath415 .
let the constellation sizes of the @xmath271 element of @xmath416 and @xmath417 be @xmath418 and @xmath419 , respectively .
further , let @xmath383 be the constellation size of the resolution codeword , which is selected such that @xmath383 divides both @xmath418 and @xmath419 @xcite .
it is assumed by the authors of @xcite that the sources use transmit powers such that @xmath420 where @xmath421 and @xmath422 .
furthermore , it is assumed that @xmath423 . in the rest of this paper , we assume @xmath424 .
first , we describe the encoding steps .
the two sources @xmath9 and @xmath10 transmit their signals to the relay and to each other
. meanwhile , the relay transmits its own signal to both @xmath9 and @xmath10 .
thus , @xmath9 receives the superposition of signals from @xmath10 and the relay .
similarly , @xmath10 receives signals from @xmath9 and the relay . during the @xmath309 block , where @xmath425 , the sources @xmath9 and @xmath10 transmit their new codewords @xmath426 $ ] and @xmath427 $ ] , while the relay transmits the resolution component of the decoded sum codeword @xmath428 $ ] . during the block @xmath314 , the relay transmits nothing and at block @xmath318 , the sources receive the resolution information @xmath429 $ ] from the relay , since both sources have no information to send .
next , we describe the decoding steps . since , the decoding at @xmath9 and @xmath10 is similar , we only consider decoding at @xmath10 .
decoding occurs in three phases . in phase one
, the relay decodes the sum codeword @xmath430 . in phase two
, @xmath10 decodes the resolution codeword @xmath431 by treating other codewords as noise . in the last phase , @xmath10 decodes the vestigial codeword @xmath432 .
the received signal at relay in @xmath309 block is @xmath433 & = & \sqrt{p_{s_1}}h_{s_1r}\mathbf{x}_{s_1}'[t]+\sqrt{p_{s_2}}h_{s_2r}\mathbf{x}_{s_2}'[t]+\mathbf{z}_r[t]\nonumber \\ & = & \rho\left(m_1\mathbf{x}_{s_1}'[t]+m_2 \mathbf{x}_{s_2}'[t]\right)+\mathbf{z}_r[t ] .
\label{y_r_two_way}\end{aligned}\ ] ] let @xmath434=\mathbf{y}_r[t]-\rho\left(-m_1-m_2 + 1\right)$ ] . using @xmath434 $ ] , the relay performs ldpc lattice decoding to obtain an estimate of the sum information @xmath435+m_2 \mathbf{b}_{s_2}'[t]$ ] @xmath436\oplus m_2 \hat{\mathbf{x}}_{s_2}'[t ] & = & \textrm{dec}_{\textrm{ldpcl}}\left(\frac{\mathbf{y}_r'[t]}{\rho},\frac{n_r}{\rho}\right ) , \quad\,\,\,\label{sum_information1}\\ m_1\hat{\mathbf{b}}_{s_1}'[t]+m_2 \hat{\mathbf{b}}_{s_2}'[t ] & = & \left\lfloor \mathcal{d}\left ( \hat{\mathbf{x}}_r'[t ] \right ) \mathbf{g}_{\lambda}^{-1}\right\rceil,\label{sum_information2}\end{aligned}\ ] ] where function @xmath437 is defined in section [ encoding_oneway ] and @xmath438=m_1\hat{\mathbf{x}}_{s_1}'[t]+m_2 \hat{\mathbf{x}}_{s_2}'[t]$ ] .
the @xmath271 element of the sum is given by @xmath439 then , the relay takes the result modulo @xmath383 to find the modulo sum of the two resolution information vectors @xmath440,l_i^{(r)}\right ) \\ & = & \mathrm{smod}\left ( \left [ m_1\hat{b}_{s_1,i}'+m_2 \hat{b}_{s_2,i}'\right],l_i^{(r ) } \right).\end{aligned}\ ] ] since @xmath383 and @xmath441 , @xmath442 were selected to be co - primes , the individual codewords can be recovered from a single codeword and the modulo sum . due to the existent constraints on transmit power , the relay employs shaping methods and maps @xmath443 $ ] to another information vector @xmath444 $ ] @xmath445 & = & \hat{\mathbf{b}}_{r}^{(r)}[t]-\mathbf{s}_r\mathbf{l}^{(r)},\end{aligned}\ ] ] where @xmath446 is given in ( [ s_r_twoway ] ) and ( [ sr_s_nested2 ] ) for hypercube shaping and nested lattice shaping , respectively .
then , the shaped lattice codeword is @xmath447=\mathcal{e}\left(\hat{\mathbf{b}}_{r}^{\prime(r)}[t]\right)$ ] . during block @xmath448
, the relay transmits @xmath449 $ ] .
now , we describe the decoding process at @xmath10 .
the received signal at @xmath10 in block @xmath448 is @xmath450 & = & h_{rs_2}\sqrt{p_r}\hat{\mathbf{x}}_{r}^{\prime(r)}[t]\nonumber\\ & + & h_{s_1s_2}\sqrt{p_{s_1}}\mathbf{x}_{s_1}'[t+1]+\mathbf{z}_{s_2}[t+1].\end{aligned}\ ] ] source @xmath169 subtracts its own scaled and translated codeword @xmath451+(1,\ldots,1)\right]$ ] from the received signal to obtain @xmath452 & = & h_{rs_2}\sqrt{p_r}\left[\hat{\mathbf{x}}_{r}^{\prime(r)}[t]-m_2\mathbf{x}_{s_2}[t]-(1,\ldots,1)\right]\nonumber\\ & + & h_{s_1s_2}\sqrt{p_{s_1}}\mathbf{x}_{s_1}'[t+1]+\mathbf{z}_{s_2}[t+1].\end{aligned}\ ] ] then , treating @xmath453+\mathbf{z}_{s_2}[t+1]$ ] as noise , @xmath10 uses ldpc lattice decoding to obtain @xmath454=\left\lfloor \mathcal{d}\left ( \textrm{dec}_{\textrm{ldpcl}}\left(\frac{\mathbf{y}_{s_2}^{\prime\prime}[t+1 ] } { \gamma},\frac{n_{s_2 } } { \gamma}\right ) \right ) \mathbf{g}_{\lambda}^{-1}\right\rceil,\ ] ] where @xmath455 .
it must be noted that the decoded integer vector in ( [ dec_s2_twoway ] ) is @xmath456=\mathrm{smod}\left ( \tilde{\mathbf{b}}_r^{(r)}[t],\mathbf{l}^{(r ) } \right)-m_2\mathbf{b}_{s_2}[t]-\mathbf{s}_r\mathbf{l}^{(r)},\ ] ] where @xmath457= m_1\tilde{\mathbf{b}}_{s_1}[t]+m_2 \tilde{\mathbf{b}}_{s_2}[t]$ ] .
let @xmath458=\mathrm{smod}\left(\left[\tilde{\mathbf{b}}_{s_1}^{\prime ( r)}[t]+m_2\mathbf{b}_{s_2}[t]\right],\mathbf{l}^{(r ) } \right)$ ] and @xmath459=\mathrm{smod2mod}\left(\mathbf{b}_1[t],\mathbf{l}^{(r ) } \right)$ ] , in which the function @xmath460 , that is defined next , is applied componentwise @xmath461 source @xmath169 obtains the resolution information of source @xmath103 as @xmath462 \right . & - & \left .
m_2\mathbf{b}_{s_2}[t]\right ) \pmod*{\mathbf{l}^{(r)}}\nonumber\\ & = & \left[\left(m_1\tilde{\mathbf{b}}_{s_1}[t]+m_2 \tilde{\mathbf{b}}_{s_2}[t]\right)\pmod*{\mathbf{l}^{(r ) } } \right .
\nonumber\\ & & - \left
. m_2\mathbf{b}_{s_2}[t]-\mathbf{s}_r\mathbf{l}^{(r)}\right]\pmod*{\mathbf{l}^{(r)}}\nonumber\\ & = & m_1\tilde{\mathbf{b}}_{s_1}[t]\pmod*{\mathbf{l}^{(r)}}.\end{aligned}\ ] ] then , it recovers the @xmath271 element of unshaped resolution information @xmath463 $ ] by computing @xmath464 where @xmath465 is the unique integer such that @xmath466 .
such a unique @xmath465 always exists since @xmath467 and @xmath441 are coprime @xcite .
next , @xmath10 uses @xmath463 $ ] and @xmath468+m_2\mathbf{b}_{s_2}[t]$ ] to obtain @xmath469=\mathcal{e}(\tilde{\mathbf{b}}_{s_1}^{(r)}[t])$ ] and @xmath470=\mathcal{e}(\tilde{\mathbf{b}}_{s_1}^{\prime ( r)}[t]+m_2\mathbf{b}_{s_2}[t])$ ] .
hence , it has @xmath471 $ ] and @xmath469 $ ] from @xmath472 $ ] and @xmath473 $ ] .
then , @xmath10 subtracts @xmath474+h_{s_1s_2}\sqrt{p_{s_1}}\left(\tilde{\mathbf{x}}_{s_1}^{(r)}[t]+(1,\ldots,1)\right)$ ] from the original received signal at @xmath309 block @xmath472 $ ] .
the simplified form of this subtracted signal is @xmath475=h_{s_1s_2}\sqrt{p_{s_1}}\mathbf{x}_{s_1}^{\prime ( v)}[t]+\mathbf{e}_{s_2}+\mathbf{z}_{s_2}[t],\ ] ] where @xmath476 is the decoding error @xmath477-\tilde{\mathbf{x}}_{s_1}^{(r)}[t]\right ] \nonumber\\ & + & h_{rs_2}\sqrt{p_r}\left[\hat{\mathbf{x}}_r^{\prime(r)}[t-1]-\tilde{\mathbf{x}}_{r}^{\prime(r)}[t-1 ] \right].\end{aligned}\ ] ] then , @xmath10 uses ldpc lattice decoding to find @xmath478=\mathbf{l}^{(r)}\circ\left\lfloor \mathcal{d}\left ( \textrm{dec}_{\textrm{ldpcl}}\left(\mathbf{y}_{s_2},\boldsymbol{\sigma}_{s_2}\right ) \right ) \mathbf{g}_{\lambda}^{-1}\right\rceil,\ ] ] where @xmath479 denotes the hadamard product or entrywise product of matrices , @xmath480 and @xmath481 . to obtain the vestigial information
, @xmath10 computes @xmath482=\mathrm{smod}(\tilde{\mathbf{b}}_{s_1}^{\prime ( v)}[t],\mathbf{l}_{s_1})$ ] . finally , @xmath10 obtains the shaped and unshaped full - rate information by taking the sum of the resolution and vestigial information as @xmath483=\tilde{\mathbf{b}}_{s_1}^{(r)}[t]+\tilde{\mathbf{b}}_{s_1}^{\prime(v)}[t]$ ] and @xmath484=\tilde{\mathbf{b}}_{s_1}^{(r)}[t]+\tilde{\mathbf{b}}_{s_1}^{(v)}[t]$ ] , respectively .
a symbol error occurs at @xmath10 if @xmath485 .
note that the above decoding process is presented for the case that the employed shaping method is hypercube shaping .
when the employed shaping is nested lattice shaping , this decoding steps are still valid by changing @xmath401 function into regular modulo operation .
this is due to the fact that , the components of lattice vectors , given as the inputs for hypercube shaping and nested lattice shaping , are drawn from different sets . for hypercube and nested lattice shading methods we use the sets @xmath486 and @xmath487 , respectively .
in the simulations , @xmath488 , i.e. , we assume @xmath489 of the information integers are zero for resolution and vestigial information vectors .
we have used binary ldpc codes with @xmath490 , where @xmath0 and @xmath126 are the codeword length and the dimension of the code , respectively . symbol error rate ( ser ) performance of ldpc lattice codes are plotted against the sum power at source and the relay , i.e. , @xmath491 .
we have considered @xmath492 , @xmath493 and @xmath494 .
the path loss exponents are @xmath495 , @xmath496 .
the variances of the noise at the relay and destination are @xmath497 .
the maximum number of iterations in each step of the decoding is assumed to be @xmath498 . since , the encoder and the decoder both know the locations of the zeros in resolution and vestigial information , based on ( [ si2])([bvi ] ) , for following locations we have @xmath499 we estimate @xmath500 and @xmath501 for our scheme , for the case that hypercube shaping is applied .
when , all of the elements of the lattice codewords are uniformly distributed over @xmath502 , the average power of @xmath187 is @xmath503 . due to the fact that , the resolution lattice vectors contain more zeros , the average power in this case is less than @xmath504 .
we assume that the members of incoming integer vector are uniformly distributed over @xmath505 , for @xmath185 .
since @xmath506 for @xmath507 and @xmath508 for @xmath509 , we have @xmath510 for @xmath509 , and @xmath511 for @xmath507 , and from above equations we have @xmath512 put @xmath513 , then we have @xmath514 for @xmath515 , @xmath516 , so @xmath517 . for @xmath518 , @xmath519 .
thus , @xmath520 we have considered @xmath521 , for @xmath185 .
thus , for the cases that we have employed hypercube shaping , based on ( [ rate ] ) , the corresponding rate is @xmath522 bits / integer . for employing nested lattice shaping , we consider @xmath523 and @xmath524 . then , based on ( [ rate_nested ] ) , the corresponding rate is @xmath525 bits / integer . in order to achieve these rates , according to ( [ rate_bound ] ) , the total required powers are @xmath526db , and @xmath527db , respectively . in
, we have presented ser variation versus sum of transmit powers for both nested - lattice shaping and hypercube shaping . in @xcite , the implementation of block markov encoding proposed for ldlc lattice codes .
we have considered @xmath6 , @xmath7 , @xmath8 and other parameters similar to their corresponding values in @xcite .
the ser performance of an ldlc lattice code with dimension @xmath528 and rate @xmath529 , which is obtained by employing nested lattice shaping , at @xmath530 is @xmath531 away from its corresponding df inner bound , which is @xmath532db .
we observe that the ser performance of an ldpc lattice code of length @xmath528 at @xmath530 is @xmath533 away from its corresponding df inner bound .
this is a natural result , due to the better ser performance of ldlcs comparing to ldpc lattice codes , over awgn channels .
different decoders have been proposed for ldlcs . as far as we know ,
the best one is proposed in @xcite .
the decoding complexity of ldlcs , by using proposed decoder in @xcite , is at least @xmath534 times more than the decoding complexity of ldpc lattices .
indeed , the decoding complexity of an ldpc lattice of dimension @xmath528 is equivalent to the decoding complexity of an ldlc with dimension @xmath535 .
results of show that the increase in the dimension of the lattice can decrease the gap between df bound and the performance curve . using an ldpc lattice code of dimension @xmath536 instead of dimension @xmath528 makes about @xmath537db improvement in the performance . in , we plot the ser versus the sum of transmit powers @xmath538 , for the two - way relay channel . we suppose that the relay is midway between the sources , that is , @xmath539 and @xmath540 .
we assume @xmath541 , for @xmath542 and @xmath543 , for @xmath544 .
for the resolution lattice we consider @xmath545 .
we choose @xmath546db .
path - loss exponents are @xmath547 and @xmath548 .
the used underlying codes are the same ones that we used for one - way relay channels .
depending on the dimension of the lattice and the applied shaping method , our scheme achieves to within @xmath169db of the achievable rate in ( [ rate_bound2 ] ) .
in this paper , we present the implementation of block markov encoding using ldpc lattice codes over the one - way and two - way relay channels . in order to apply this scheme
, we employ a low complexity decoding method for ldpc lattices .
then , for using these lattices in the power constrained scenarios , we propose two efficient shaping methods based on hypercube shaping and nested lattice shaping .
we apply different decomposition schemes for one - way and two - way relay channels .
the applied decomposition schemes are the altered versions of the applied methods for decomposing ldlcs . due to the lower complexity of decoding ldpc lattices comparing to ldlcs ,
the complexity of the proposed schemes in this paper are significantly lower than the ones proposed for ldlcs .
simulation results show that ldlcs outperform ldpc lattices in general .
however , having lower decoding complexity enables us to increase the dimension of the lattice to compensate the existing gap between the performance of the ldpc lattice codes and the ldlcs .
j. n. laneman , d. n. c. tse , and g. w. wornell , `` cooperative diversity in wireless networks : efficient protocols and outage behavior , '' _ ieee trans . on inform .
50 , no . 12 , pp .
30623080 , dec . 2004 .
n. lee and c. wang , `` aligned interference neutralization and the degrees of freedom of the two - user wireless networks with an instantaneous relay , '' _ ieee trans . on commun .
_ , vol .
61 , no . 9 , pp . 36113619 , sep . | low density parity check ( ldpc ) lattices are obtained from construction d and a family of nested binary ldpc codes .
we consider an special case of these lattices with one binary ldpc code as underlying code . this special case of ldpc lattices
can be obtained by lifting binary ldpc codes using construction a lattices .
the ldpc lattices were the first family of lattices which have efficient decoding in high dimensions .
we employ the encoding and decoding of the ldpc lattices in a cooperative transmission framework .
we establish two efficient shaping methods based on hypercube shaping and voronoi shaping , to obtain ldpc lattice codes .
then , we propose the implementation of block markov encoding for one - way and two - way relay networks using ldpc lattice codes .
this entails owning an efficient method for decomposing full - rate codebook into lower rate codebooks .
we apply different decomposition schemes for one - way and two - way relay channels which are the altered versions of the decomposition methods of low density lattice codes ( ldlcs ) . due to the lower complexity of the decoding for ldpc lattices comparing to ldlcs ,
the complexity of our schemes are significantly lower than the ones proposed for ldlcs .
the efficiency of the proposed schemes are presented using simulation results .
relay networks , lattice codes , block markov encoding . |
the first indications of qso clustering came from the observations of possible physical pairs of quasars around 1980 ( hazard 1979 , oort 1981 , margon 1981 , webster 1982 ) .
although the individual cases were of tantalizing interest , it was difficult to gain an appreciation of their true significance from a - posteriori statistics .
systematic searches began with the pioneering work of osmer ( 1981 ) and the first detection was obtained by shaver ( 1984 ) , using a large , inhomogeneous qso catalog .
the surveys carried out to investigate the qso luminosity function ( lf ) before long provided statistically - well - defined samples with sufficient surface density and size for a study of the clustering on scales of a few tens of mpc ( shanks 1983 , 1987 ) .
the two - point correlation function ( tpcf ) and its fourier transform , the power - spectrum , have been applied as a standard approach to investigate the qso clustering , but other techniques have also been explored , such as counts in cells ( andreani 1994 , carrera 1998 ) , the minimal spanning tree ( graham 1995 ) , fractal analysis ( andreani 1991 ) and the friend - of - friend algorithm ( komberg 1996 ) .
notwithstanding all the caveats due to their exotic nature , qsos display a number of appealing properties when compared to galaxies as cosmological probes of the intermediate - linear regime of clustering , on distances @xmath5 mpc : they show a rather flat redshift distribution , make it possible to define samples which are in practice locally volume limited , their point - like images are less prone to the surface - brightness biases typical of galaxies and they sparse - sample the environment . in recent times complete samples totaling about 2000 qsos have been assembled , providing a @xmath6 detection of the qso clustering on scales of the order of @xmath7 comoving mpc ( andreani & cristiani 1992 , mo & fang 1993 , croom & shanks 1996 ) , an amplitude which appears to be consistent with or slightly larger than what is observed for present - day galaxies and definitely lower than the clustering of clusters .
the evolution of the qso clustering with redshift is still a matter of debate .
an amplitude constant in comoving coordinates or marginally decreasing with increasing redshift has been the standard claim till recently .
however , a number of indications currently point the opposite direction .
the data of the palomar transit grism survey ( kundi 1997 , stephens 1997 ) make it possible to measure the amplitude of the tpcf at redshifts higher than 2.7 and the result , @xmath8 mpc , is suggestively higher than what is observed at lower redshifts .
la franca ( 1998 ) have observed a sample of 388 qsos with @xmath9 and @xmath10 over a _ connected _ area of 25 square degrees in the south galactic pole ( sgp ) .
the tpcf analysis gives an amplitude @xmath11 mpc , in agreement with previous results .
but when the evolution of the clustering with redshift is analyzed , evidence is found for an _ increase _ of the clustering with increasing z , although only at a @xmath12 level .
boyle & mo ( 1993 ) measured the clustering of low - z qsos in the emss and georgantopoulos & shanks ( 1994 ) used the iras point source catalog to measure the clustering of seyferts . altogether a low value of the average tpcf within 15 mpc and z=0.05 is obtained , @xmath13 . on the other hand , carrera ( 1998 ) , on the basis of 235 x - ray selected agns from the rixos and drs surveys , do not support an increase of the clustering amplitude with redshift .
it is also worth noting that the clustering measured by carrera ( 1998 ) for these x - ray selected objects , @xmath14 ( with the non - crucial assumption of stable evolution ) , is significantly smaller than what is observed for optically - uv - selected qsos .
this opens the interesting possibility of having a contribution of agns to the xrb larger than the standardly quoted 50% , without violating the limits imposed by the angular auto - correlation - function ( carrera & barcons 1992 , carrera 1993 , georgantopoulos 1993 , danese 1993 , chen 1994 , sotan & hasinger 1994 ) .
the customary appeal `` more data are needed '' arises naturally in this context : hidden correlations , for example with the absolute luminosity , as observed for galaxies ( park 1994 ) , although undetected or undetectable in the present data , might confuse our perspective on the issue of clustering evolution .
hopefully we will not wait long , thanks to the 2df qso redshift survey ( cf .
croom , these proceedings ) . at the moment , if we limit the analysis to the optical data and parameterize the evolving correlation function in a standard way : @xmath15 where @xmath16 is an arbitrary ( and not very physical ) fitting parameter , we obtain @xmath17 , which appears inconsistent with the @xmath18 observed for faint galaxies at lower redshifts ( le fvre 1996 , carlberg 1997 , villumsen 1997 , postman 1998 ) .
great care should be exercised however when carrying out this comparison .
quasars could be more significantly related to the population of lyman - break galaxies , which show substantial clustering at @xmath19 ( steidel 1998 ) and can be interpreted as progenitors of massive galaxies at the present epoch or precursors of present day cluster galaxies ( governato 1998 ) . as mentioned
, the standard parameterization of the clustering may not be a sensible description .
the observed clustering is the result of an interplay between the clustering of mass , decreasing with increasing redshift , and the bias , increasing with redshift , convolved with the redshift distribution of the objects ( moscardini 1998 , bagla 1998 ) and obviously is critically linked to the physics of qso formation and evolution .
let us consider here as an archetypical scenario the one outlined by cavaliere ( 1997 ) , modeling the _ rise and fall _ of the qso lf as the effect of two components : the newly formed black holes , which are dominant at @xmath20 and the reactivated black holes , which dominate at @xmath21 .
the reactivation is triggered by interactions taking place preferentially in groups of galaxies of typical halo mass @xmath22 m@xmath23 . in this way
the clustering properties of qsos are related to those of transient , short - lived , objects ( matarrese 1997 ) .
this is markedly different from the situation with respect to galaxies , which in turn can be associated with a merging or object - conserving paradigm of long - lived objects . at high redshifts qsos
correspond to larger ( rarer ) mass over - densities collapsing early and cluster very strongly .
then the clustering amplitude decreases till the mass scale typical of a qso reaches the value of the average collapsing peaks , after which clustering may grow again .
if we think of qsos as objects sparsely sampling halos with @xmath24 , an @xmath25 would provide the desired amount _ and _ evolution of the clustering .
the ideas presented in the previous section are corroborated by the study of qso environments . over the years , starting with the pioneering observations of stockton ( 1978 , 1982 ) , evidence has been accumulating that qsos are associated with enhancements in the galaxy distribution . at low redshifts ( @xmath26 )
quasars reside in small to moderate groups of galaxies , rather than in rich clusters ( yee & green 1984 , haymann 1990 , smith & heckman 1990 ) .
the qso / galaxy correlation function has been measured to be @xmath27 times stronger than the galaxy / galaxy correlation function ( fisher et al .
1996 ) . at higher redshifts ,
there is a marked difference in the environments of radio - loud and radio - quiet qsos ( yee & green 1987 , yee & ellingson 1993 ) .
radio - loud qsos are often found in association with rich clusters ( abell richness @xmath28 ) , at least on scales larger than @xmath29 ( hall & green 1998 ) , while radio - quiet quasars appear to remain in smaller groups .
an indirect confirmation comes from the observations of close line - of - sight ( los ) qso pairs . in their absorption spectra lines
are often seen in the spectrum of the more distant qso at a redshift corresponding to that of the foreground qso - the so - called _ associated absorption _ - showing that qsos are located in higher - than - average galaxy density regions ( shaver & robertson 1983 , phillips 1986 , cristiani & shaver 1988 ) .
the study of the qso clustering on scales larger than a few tens of mpc is a much less beaten path . on the one hand
the familiar tpcf is known to be inefficient for detecting structures on scales comparable with the dimension of the sample ( e.g. pando & fang 1996 ) and other approaches , such as the wavelet analysis or counts - in - cells , may represent better tools . on the other hand observational biases
may be extremely dangerous in this situation : for example modulations of the sensitivity on scales of the order of the size of a schmidt plate may introduce spurious power on scales of the order of 100 mpc . in the present absence of proper systematic surveys on these large scales
one should therefore take with the classical grain of salt the several examples of qso super - structures , derived mostly from inhomogeneous catalogues , being aware of the pitfalls of a - posteriori statistics .
for the large scales we are in many respects in a situation similar to that of the early eighties for the small scales .
nonetheless there are at least two known cases which appear as outstanding examples of large quasar groups : 1 .
a group of 23 quasars within @xmath30 at @xmath31 , found by crampton , cowley & hartwick ( 1989 ) 2 .
a group of 14 quasars within @xmath32 at @xmath33 , found by komberg in the sgp . in the sgp , thanks to the large number of deep surveys carried out by various authors , at least a couple of other possible super - structures , although of less compelling evidence , have been identified ( clowes and campusano 1991 , graham 1995 ) .
such large groups appear as possible progenitors of the present - epoch great attractors ( i.e. regions in the local universe with over - densities @xmath34 on scales @xmath35 ) : and their spatial comoving density can be estimated to be comparable to local super - clusters . pinpointing the qso clustering on these large scales would immediately provide another challenge to cosmological models , but for this a systematization of the field is needed .
great advances are expected from the wide - field facilities which have come recently into operation : sdss , the 2df qso survey and also from large - area follow - up spectroscopy of radio surveys like first .
inspired by the paradigm unifying radio - loud qsos and frii radio - sources it is interesting to check whether the observations of radio - galaxies measure clustering properties consistent with those of qsos . historically , the long debate on the spatial distribution of radio galaxies led to the widely - accepted view that radio sources were isotropically distributed , at least on large scales ( webster 1976 ) . + the problem with radio surveys is that in general they lack the redshift information and the clustering analysis has to resort to the angular correlation function .
the large range in intrinsic luminosity of radio sources leads then to a wide convolution that dilutes dramatically any clustering signal .
in fact , when the studies of the radio sources have been limited to nearby objects as seldner & peebles ( 1981 ) did for 4c sources at @xmath36 or shaver & pierre ( 1989 ) for molonglo sources at @xmath37 , structures were found , including the super - galactic plane .
peacock & nicholson in 1991 used an all - sky sample of radio - galaxies at @xmath36 to measure their 3d tpcf , @xmath38mpc . in recent times kooiman ( 1995 ) and loan , wall & lahav ( 1997 ) analyzed the 4.85 ghz green bank and the parkes - mit - nrao surveys with the tpcf , showing that radio objects appear to be more strongly clustered than local optically - selected galaxies .
+ magliocchetti ( this conference ) reports on the studies carried out on the first database ( magliocchetti 1998a , b , cress et al 1996,1998 ) , which provide indeed an estimate of the correlation amplitude , @xmath39mpc , for objects typically at @xmath40 , consistent with the qso result . to gain more robust insight
it is however mandatory to obtain redshifts and radio morphologies at least for a training subsample of these databases in order to remove the basic statistical uncertainties concerning the redshift distribution and the multi - component nature of many sources .
the environments of radio - galaxies at redshifts @xmath41 also appear to be remarkably similar to those of qsos , that is _ on average _ not richer than abell class 0 clusters ( roche 1998 ) , with a correlation function significantly stronger ( @xmath42mpc ) and steeper ( @xmath43 ) than what is observed for `` normal '' galaxies , but consistent with the estimates for local early - type galaxies of guzzo ( 1997 ) .
a complementary view of the lss is provided by the study of qso absorption spectra , which sensitively probe , in a luminosity unbiased way , the spatial distribution of baryons in the universe .
historically , the so - called sharp metal - lines have been the first class of absorbers recognized to cluster significantly .
already in 1982 young , sargent & boksenberg indicated that the peak observed in the tpcf of civ doublets could be due to galaxy - galaxy clustering at @xmath1 rather than to multiple clouds in individual galaxies .
this was reiterated in 1988 by sargent , steidel & boksenberg on the basis of a sample of 55 qsos observed at @xmath44 km / s resolution , measuring a correlation function @xmath45 for los velocity differences @xmath46 km / s . in recent times
petitjean & bergeron ( 1994 ) , songaila & cowie ( 1996 ) and dodorico ( 1998a ) , using higher - resolution , smaller samples , have measured a @xmath47 at @xmath48 km / s .
quashnock & vanden berk ( 1998 ) have used a compilation of heavy - element qso absorbers to estimate the civ tpcf from 200 up to about 3000 km / s .
the various authors unanimously remark that on small scales the velocity structure must to a certain extent reflect relative motions of clouds within individual halos , a circumstance confirmed by the relative weakness of the correlation observed in the transverse direction on the adjacent lines - of - sight of a qso triplet ( crotts 1997 ) .
however , we can see from fig .
2 that , taken at face value , the results by dodorico on small scales and by quashnock & vanden berk together define with remarkable accuracy a civ cf at @xmath49 of the canonical power - law form , with an @xmath50 mpc and @xmath51 . + an evolution of the cf as a function of @xmath52 is also detected and can be described in the standard parameterization of 1 by an @xmath53 .
this evolutionary pattern is apparently followed also by the mgii systems , which , as shown by petitjean & bergeron ( 1990 , 1994 ) and caulet ( 1989 ) , in spite of the different ionization phase are similar to civ systems in terms of cross sections .
there is also evidence for a dependence ( increase ) of the clustering strength on the ( increasing ) column density of the systems ( dodorico 1998a , quashnock & vanden berk 1998 ) , which suggests , together with the large correlation scale found , that the civ absorbers are biased tracers of relatively high density regions .
it also cautions us that the agreement with the standard @xmath54 power - law behavior of the cf shown in fig.2 may be `` too - good - to - be - true '' , since the lines of the sample by dodorico et al .
have on average lower @xmath55 ( @xmath56 ) than the ones in the quashnock & vanden berk sample ( @xmath57 ) .
again , plenty of data are needed to disentangle the various dependences , including the evolutionary rates as a function of the column density .
the search for clustering of the lyman-@xmath58 lines has produced disparate results over the years .
systematic studies of the redshift distribution in the qso lyman-@xmath58 forest began in the early 80 s with the work by sargent ( 1980 ) who concluded that no structures could be identified .
almost all the subsequent results have failed to detect any significant correlation on velocity scales @xmath59 km s@xmath60 ( sargent 1982 , webb & barcons 1991 ) . on smaller scales
( @xmath61 km s@xmath60 ) there have been indications of weak clustering ( webb 1987 , crotts 1989 , rauch 1992 , chernomordik 1995 , cristiani 1995 , kim 1997 ) together with relevant non - detections ( pettini 1990 , stangler & webb 1993 , kirkman & tytler 1997 ) .
the absence of power in the tpcf has been claimed as a striking characteristic of the lyman-@xmath58 forest and has been used , together with the observed low metallicity , as a basic argument to develop a scenario of the lyman-@xmath58 absorbers as a totally distinct population with respect to metal systems and therefore galaxies . in recent times a large database of high - resolution ( @xmath62 km / s ) , high - s / n spectra of the lyman forest has become available , allowing a detailed investigation , especially at high redshift , where the density of lines provides particular sensitivity .
cristiani ( 1997 ) have used more than 1600 lines to detect a weak but significant signal , with @xmath63 on scales of 100 km s@xmath60 at a @xmath64 level . exploring the variations of the clustering as a function of the column density no evidence for clustering
is found for lines with @xmath65 . on the contrary , for lines with @xmath66 , the correlation function at @xmath67 km s@xmath60 shows a remarkable increase in amplitude ( @xmath68 ) and significance ( @xmath69 ) .
no relevant feature other than the peak at small velocity separations is observed .
in particular previous claims for anti - clustering on scales @xmath70 km s@xmath60 are not confirmed .
[ figclustew ] shows more in detail the variation of the amplitude of the tpcf as a function of the column density .
a trend of increasing amplitude with increasing column density is apparent all the way up to the two points in the upper right that represent the clustering of metal systems .
the tricky part in the otherwise trivial computation of the tpcf in the velocity space @xmath71 is the correct estimation of @xmath72 , the number of pairs expected in a given velocity bin from a random distribution in redshift .
the results are not sensitive to the particular cosmological evolution ( @xmath73 ) adopted , but the effects of the intrinsic blending , line blanketing of weak lines due to strong complexes and finite spectral resolution have to be carefully reproduced .
fernndez - soto ( 1996 ) have investigated the clustering properties of the lyman-@xmath58 clouds on the basis of the corresponding civ absorptions .
they measure in this way a strong clustering for lines with @xmath74 @xmath75 , suggesting that civ may resolve better the small - scale velocity structure .
a caveat however has to be filed , about the different behavior of high and low - ionization species and in general of and metal systems ( see also below the observations of qso gravitational lenses , 3.3 )
. of course , rather than just correlating lines regardless of their properties , increasingly more refined approaches are possible : from the correlation of ew or @xmath76 weighting ( webb & barcons 1991 , zuo 1992 ) , to the correlation of optical depths or pixel intensities ( press 1993 , cen 1998 ) , to the full recovery of the density field ( haehnelt & nusser , this conference ) and much is being learnt from simulations ( weinberg , this conference ) .
the study of the evolution of the tpcf with the redshift in the optical range ( @xmath77 ) suggests that for lyman-@xmath58 lines with column densities @xmath78 the amplitude of the correlation at @xmath44 km s@xmath60 decreases with increasing redshift ( cristiani 1997 ) .
unfortunately , hst data are still at too low - resolution or are too scanty ( bahcall 1996 , brandt 1995 ) to allow a meaningful comparison . nonetheless , the result by ulmer ( 1996 ) , who measured with fos data a @xmath79 , on scales of @xmath80 km / s at @xmath81 , suggests that the trend persists at lower redshifts . if we add to this observation the following pieces of evidence : 1 . at low redshift a significant fraction of the lyman-@xmath58 lines has been observed to be associated with luminous galaxies and the local large - scale structure ( le brun 1996 , chen 1998 , 4 ) ; 2 .
metallicities of the order @xmath82 , i.e. similar to the ones derived for the heavy - element absorptions originated in galactic halos , are observed for lyman-@xmath58 clouds with @xmath83 ( cowie 1995 , songaila & cowie 1996 ) ; 3 .
the volume density and cosmological evolution of the same @xmath84 clouds are similar to those of the damped systems ( giallongo 1996 ) ; all together this suggests a physical association between the lyman-@xmath58 clouds with @xmath83 and the halos of protogalactic systems .
the typical column density above which the clustering is observed ( @xmath85 ) corresponds to the position of the break in their column density distribution ( petitjean 1993 , giallongo 1996 ) and we should not be surprised that below that no clustering is observed since the weaker lines are recognized to probe under - dense regions of the universe . on large scales observations show the existence of significant over- and under - densities of lines on scales of a few tens of mpc .
a typical case is the occurrence of the so - called voids for which a few interesting claims exist : crotts ( 1987 , 1989 ) was the first to suggest the presence of a @xmath86 gap of absorption lines in the lyman forest of the qso 0420 - 388 , dobrzycki & bechtold ( 1991a , 1991b ) found a void of @xmath87 in q0302 - 003 and cristiani ( 1995 , 1997 ) two of about @xmath88 in q0055 - 269 .
the systematical analysis of the same sample of lines described in the previous subsection corroborates the idea that voids of a few tens of mpc occur occasionally in qso absorption spectra : the filling factor is low , a few percent of the available los path - length .
regions corresponding to the voids are not completely devoid of lines : weak absorptions are observed within the voids .
this agrees with low - redshift observations ( shull 1996 ) showing that in the local universe voids are not entirely devoid of matter . over densities on scales
again of a few tens of mpc ( never more than that ) have been found at @xmath89 , roughly with the same frequency observed for under - densities .
the most remarkable ones remain q0237 - 233 ( sargent 1988 , heisler , hogan & white 1989 ) and the pair 1037/1038 - 27 ( jakobsen 1986 , 1988 , cristiani 1987 , sargent & steidel 1987 , lespine & petitjean 1997 ) , which display clumps of metal - lines on scales of the order of 10000 km / s .
francis & hewett found an interesting coincidence ( within 300 km / s ) of two pairs damped systems at z=2.38 and 2.85 in the spectra of two qsos whose los are separated by @xmath90 .
great care should be exercised before interpreting these structures in terms of gravitational instabilities since other phenomena , also of great interest , may be acting : proximity effects , i.e. excess ionization caused by a nearby qso , and any ionization inhomogeneity in the uv background which becomes more and more important with increasing redshift ( haardt and madau 1996 , reimers 1998 ) .
an unfortunate shortcoming of clustering studies based on absorption lines is the 2-d nature of the information , typically confined to the analysis along the los .
pairs and groups of qsos offer an invaluable insight in this respect and the advent of 8-m class telescopes will certainly boost the field , allowing a systematic detailed 3-d mapping of absorbers and galaxies at the same time up to the highest redshifts .
historically , quasar pairs have provided information about _ a ) _ the size of the absorbers , which turned out to be much larger than the few tens of kpc scale thought a decade ago , _
b ) _ the existence of superclusters at high redshift ( see 3.3 ) , _ c ) _ the association of qsos with moderate overdensities ( 1.3 ) . where do we stand now ?
on small scales ( @xmath91 kpc ) gravitational lenses have shown that the forest of lines in the spectra of the two images of a lens can be amazingly similar with differences less than a few percent on scales of a few hundred pc ( rauch , this conference ) .
this is very important because it tells us that in the case of - not metal lines - the medium is quite homogeneous on such scales and therefore an empirical limit is provided for the critical issue of the resolution in hydrodynamical simulations ( cf .
theuns , this conference ) . at the intermediate ( @xmath92 ) scales progress
has been slow in the last decade , due to the scarcity of suitably bright pairs and groups for 4m - class telescopes .
dodorico ( 1998b ) have recently re - observed the qso pair 0307 - 195 and re - discussed this and the other 7 pairs in the literature with data of comparable quality .
a typical size of the absorbers - with the assumption of spherical geometry - of @xmath93 kpc is inferred ( fig.4 ) . besides , remarkably quiet velocity fields are observed with differences between coincident lines within 100 km / s ( dinshaw 1994 ) .
the present data are not sufficient to support a correlation of the typical inferred size with the proper separation or with the redshift of the pairs , which was claimed in the past .
it may be argued then whether such dimensions can really be considered cloud radii or are rather correlation lengths , a distinction which tends to be blurred in view of the present hydrodynamic simulations .
it is also debatable whether flattened disk - like geometries tend to be favored ( dinshaw 1997 ) , or spherical clouds with an @xmath94 column density profile and a power - law distribution of radii can give a satisfactory representation of the observations , too . certainly the empirical evidence points to large ionization fractions and consequently large masses .
on the basis of the inferred sizes and the uv background measured at @xmath95 by giallongo ( 1996 ) the optically thin absorbers probed in qso pairs are expected to have a neutral fraction @xmath96 and their contribution to the mean intergalactic density is close to conflict with the baryon limit from nucleosynthesis . on the large scales , after the studies of tol1037/1038 , there has been disappointingly little progress .
the most remarkable recent result has been produced by williger ( 1996 ) who observed metal systems in 25 los spanning @xmath97 , spread over about 1 sq .
degree in the south galactic pole . in spite of the praiseworthy effort ,
the final sample of civ systems collected by williger makes it possible to compute the tpcf by cross - correlating a maximum of 22 absorbers and is just sufficient to reject the null hypothesis of a random distribution .
the reason is again the same : the surface density of qsos accessible to 4-m telescopes for this type of work is simply too small .
future prospects , however , look really exciting . starting at the end of 1999 , the echelle spectrograph uves will be available on the ut2 of vlt , extending , with respect to hires at keck , the possibility of high - resolution ( @xmath98 ) observations down to at least @xmath99 and to a shorter wavelength range ( i.e. to more numerous , lower - redshift qsos ) .
then ( by 2001 ) the flames facility ( see http://http.hq.eso.org/instrumen + ts / flames/ ) will allow observation with uves of up to 8 qsos at the same time in a field of 25 arcmin diameter and/or 130 targets in the same area with an intermediate / high resolution spectrograph , giraffe .
this opens up enormous possibilities : for example the alcock - paczinsky test applied to qso groups to investigate the cosmological geometry ( hui , this conference ) would require few observing nights .
this test , based on the comparison between the clustering properties of the ( ) absorbers , namely the tpcf , observed along the los and the corresponding estimate in the transverse direction , is especially sensitive to @xmath100 and should be able to discriminate between an ( @xmath101 ) and an ( @xmath102 ) universe at a @xmath103 level by observing @xmath104 qso pairs ( hui 1998 ) .
at lower redshift , after the pioneering work on the identification of galaxies giving origin to metal absorbers ( bergeron & boiss 1991 , bergeron 1992 ) , the more recent activity on systems , triggered by the hst observations , is fueling a lively debate . a number of authors ( morris 1993 , lanzetta 1995 , le brun , bergeron & boiss 1996 , van gorkom 1996 , chen 1998 , weymann et al .
1998 ) have argued whether the low - redshift absorption systems have a direct physical connection with galaxies or a significant fraction of them is caused by a truly intergalactic medium .
a central point in this debate is the anti - correlation between the equivalent width and the impact parameter of the los from the center of the galaxies measured by some authors ( lanzetta 1995 , cen 1998 ) and not fully confirmed by others ( le brun 1996 , bowen 1996 ) . at the moment it appears possible to reconcile the various observations _
if _ the presence of well defined galactic envelopes is invoked only for the relatively strong absorption lines .
the covering factor around each galaxy for this class of absorbers is high at impact parameters @xmath105 kpc , dropping rapidly at larger separations
. the weaker lines , which also display a significantly slower redshift evolution ( weymann 1998 ) , occur then in filaments up to several mpc away from the nearest galaxy .
the zoo of interpretations for the low - z ly@xmath58 systems includes : tidal tails ( morris & van den berg 1994 ) , galactic winds ( wang 1995 ) , outer parts of disk and irregular galaxies ( maloney 1992 ) , clouds that are pressure confined by a hotter gas in extended galactic halos ( mo 1994 ) , extended low redshift disk galaxies which , in a short burst of star formation , blow out their denser centers and disappear from view , while their gas cross - section remains ( the `` cheshire cat model '' : salpeter 1993 ) , galaxy clusters ( lanzetta et al 1996 ) .
possibly , more than one population will have to be invoked in order to fully account for all the observations .
99 andreani p. , cristiani s. , la franca f. , 1991 , mnras , 253 , 527 andreani p. , cristiani s. 1992 , apj , 398 , l13 andreani p. , cristiani s. , lucchin f. , matarrese s. , moscardini l. , 1994 , apj , 430 , 458 bagla j.s . , 1998 , mnras , 297 , 251 bahcall j. , bergeron j. , boksenberg a. , , 1996 , apj , 457 , 19 bergeron j. , boiss p. , 1991
, a&a , 243 , 344 bergeron j.,cristiani s. , shaver p.a . , 1992 , a&a , 257 , 417 bowen d.v . , blades j. , pettini m. , 1996 , apj , 464 , 141 boyle b.j . , mo h.j . , 1993 , mnras , 260 , 925 brandt j.c . , heap s.r . ,
beaver e.a . , , 1995 , aj 105 , 831 carlberg r.g . , cowie l.l .
, songaila a. , hu e.m . , 1997 ,
apj , 484 , 538 carrera f.j . ,
barcons f.j . , 1992 , mnras , 257 , 507 carrera f.j . , 1993 , mnras , 260 , 376 carrera f.j . , astro - ph/9804314 caulet a. , 1989 , apj , 340 , 90 cavaliere a. , perri m. , vittorini v. , 1997 , mem.s.a.it .
, 68 , 27 cen r. , phelps s. , miralda - escud/e j. , ostriker j.p . , 1998 , apj , 496 , 577 chen l.w .
, fabian a.c . , warwick r.s .
, branduardi - raymont g. , barber c.r . , 1994 , mnras , 266 , 846 chen h .- w .
, lanzetta k.m .
, webb j.k . ,
barcons x. , 1998 , apj , 498 , 77 chernomordik v.v . , 1995 ,
apj , 440 , 431 clowes r.g . ,
campusano l.e .
, 1991 , mnras , 249 , 218 cowie l.l .
, songaila a. , kim t. , hu e.m .
, 1995 , aj 109 , 1522 crampton d. , cowley a.p . , hartwick f.d.a . , 1989 , apj , 345 , 59 cress c.m . , helfand d.j . ,
becker r.h . ,
gregg m.d . ,
white r.l . , 1996 , apj , 473 , 7 cress c.m . , kamionkowsky m. , 1998 , astro - ph/9801284 cristiani s. , danziger i.j . ,
shaver p.a . , 1987 , mnras , 227 , 639 cristiani s. , shaver p.a . ,
1988 , 3rd iap astrophysical meeting on _ high redshift and primeval galaxies _ , ed .
bergeron , kunth , rocca - volmerange and tran thanh van , pg.383 cristiani s. , dodorico s. , fontana a. , giallongo e. , savaglio s. , 1995 , mnras , 273 , 1016 cristiani s. , dodorico s. , dodorico v. , , 1997 , mnras , 285 , 209 croom s.m . , shanks t. , 1996 , mnras , 281 , 89 crotts a.p.s . , 1987 , mnras , 228 , 41p crotts a.p.s . , 1989 , apj , 336 , 550 crotts a.p.s . , burles s. , tytler d. , 1997 , apj , 489 , l7 d odorico v. , cristiani s. , d odorico s. , fontana a. , giallongo e. , 1998a , a&as , 127 , 217 d odorico v. , cristiani s. , d odorico s. , fontana a. , giallongo e. , shaver p. , 1998b
, a&a , 339 , 678 danese l. , toffolatti l. , franceschini a. , martin - mirones j.m . , de zotti g. , 1993 , apj , 412 , 56 dinshaw n. , impey c.d .
, foltz c.b . , weymann r.j . ,
chaffee f.h . , 1994 ,
apj , 437 , l87 dobrzycki , a. , bechtold , j. 1991a , in crampton ed .
workshop the space distribution of quasars , asp conference series 21 , p. 272 dobrzycki , a. , bechtold , j. 1991b , apj , 377 , l69 fernndez - soto a. , lanzetta k.m . , barcons x. , carswell r.f .
, webb j.k . , yahil a. , 1996 , apj , 460 , l85 fisher k.b . , bahcall j.n . , kirhakos s. , schneider d. p. , 1996
, apj , 468 , 469 francis p.j . ,
hewett p.c.,1993 , aj , 105 , 1633 and erratum 1993 , aj , 106 , 2587 georgantopoulos i. , stewart g.c . , shanks t. , griffiths r.e .
, boyle b.j . , 1993 , mnras , 262 , 619 georgantopoulos i. , shanks t. 1994 , mnras , 271 , 773 giallongo e. , cristiani s. , dodorico s. , fontana a. , savaglio s. , 1996 , apj , 466 , 46 governato , f. , baugh c.m . , frenk c.s . ,
cole s. , lacey c.g . , quinn t. , stadel j. 1998 , nature , 392 , 359 graham m.j . , clowes r.g .
, campusano l.e . , 1995 , mnras , 275 , 790 guzzo l. , strauss m.a . ,
fisher k.b . , giovanelli r. , haynes m. , 1997 , apj , 489 , 37 haardt f. , madau p. , 1996 ,
apj , 461 , 20 hall p.b .
, green r.f . , 1998 ,
apj , astro - ph/9806151 haymann p. g. , 1990 , mnras , 245 , 358 hazard c. , arp h.c . ,
morton d.c . , 1979 , nature , 282 , 271 heisler j. , hogan c. j. , white s. d. m. , 1989 , apj 347 , 52 hui l. , stebbins a. , burles s. , 1998 , astro - ph/9807190 jakobsen p. , perryman m.a.c . ,
ulrich m.h . , macchetto f. , di serego - alighieri s. , 1986 , apj , 303 , l27 jakobsen p. , perryman m.a.c . ,
cristiani s. , 1988 , apj , 326 , 710 kim t .- s . , hu e.m .
, cowie l.l .
, songaila a. , 1997 , aj , 114 , 1 kirkman d. , tytler d. , 1997 , apj,484 , 672 komberg b. v. , kravtsov a. v. , lukash v.n .
, 1996 , mnras , 282 , 713 kooiman l.k . , burns j.o .
, klypin a.a . , 1995 , apj , 448 , 500 kundi t. , 1997 , apj , 482 , 631 la franca f. , andreani p. , cristiani s. 1998 , apj , 497 , 529 lanzetta k.m . , bowen d.b . , tytler d. , webb j.k . , 1995 ,
apj , 442,538 lanzetta k.m .
, webb j.k . , barcons x. , 1996 , apj , 456 , l17 le brun v. , bergeron j. , boiss p. , 1996 , a&a 306 , 691 le fvre o. , hudon d. , lilly s.j . , crampton d. , hammer f. , tresse l. , 1996 , apj , 461 , 534 lespine y. , petitjean p. , 1997
, a&a , 317 , 416 loan a.j . , wall j.v . , lahav o. , 1997 , mnras , 286 , 994 magliocchetti m. , maddox s.j . , lahav o. , wall j.v . ,
1998a , astro - ph/9802269 magliocchetti m. , maddox s.j .
, lahav o. , wall j.v . , 1998b , astro - ph/9806342 maloney p. , 1992 ,
apj , 398 , 89 margon b. , chanan g.a . , downes r.a . , 1981 , nature , 290 , 480 matarrese s. , coles p. , lucchin f. , moscardini , l. , 1997 , mnras , 286 , 115 mo h.j . ,
fang l .- z . , 1993 , apj , 410 , 493 mo h.j . , 1994 , mnras , 269 , 49 morris s.l . , weymann r.j . ,
dressler a. , mccarthy p.j . ,
smith b.a . ,
terrile r.j . , giovanelli r. , irwin m. , 1993 , apj , 419 , 524 morris s.l . , van den berg s. , 1994 , apj , 427 , 696 moscardini l. , coles p. , lucchin f. , matarrese s. , 1998 , mnras in press , astro - ph/9712184 oort j.h .
, de ruiter h. , arp h. , 1981 , a&a , 95 , 70 osmer p.s . , 1981 ,
apj , 247 , 762 pando j. , fang l .- z . , 1996 , apj , 459 , 1 park c. , vogeley m.s . , geller m.j . ,
huchra j.p . , 1994 , apj , 431 , 569 peacock j.a . , nicholson d. , 1991 , mnras , 253 , 307 petitjean p. , bergeron j. , 1990 , a&a , 231 , 309 petitjean p. , webb j.k . , rauch m. , carswell r.f . , lanzetta k. , 1993 , mnras , 262 , 499 petitjean p. , bergeron j. , 1994 , a&a , 283 , 759 pettini m. , hunstead r. w. , smith l. j. , mar d. p. 1990 ,
mnras , 246 , 545 phillips s. , 1986 , mnras , 223 , 173 postman m. , lauer t.r . , szapudi i. , oegerle w. , astro - ph/9804141 press w.h . , rybicki g.b .
, schneider d.p . , 1993 , apj , 414 , 64 quashnock j.m . , vanden berk d.e .
, 1998 , apj , 500 , 28 rauch m. , carswell r. f. , chaffee f. h. , foltz c. b. , webb j. k. , weymann r. j. , bechtold j. , green r. f. 1992 , apj , 390 , 387 reimers d. , khler s. , wisotzki l. , groote d. , rodriguez - pascual p. , wamsteker w. , 1997 , a&a , 327 , 890 roche n. , eales s. , hippelein h. , 1998 , mnras , 295 , 946 salpeter e.e . , 1993 , aj , 106 , 1265 sargent w. l. w. , young p. j. , boksenberg a. , tytler d. 1980 , apjs , 42 , 41 sargent w. l. w. , young p. j. , schneider d. p. , 1982
, apj , 256 , 374 sargent w.l.w . ,
steidel c.c .
, 1987 , apj , 322 , 142 sargent w.l.w .
, steidel c.c .
, boksenberg a. , 1988 , apjs , 68 , 539 seldner m. , peebles p.j.e . , 1981 ,
mnras , 194 , 251 shanks t. , fong r. , green m.r . , clowes r.g . ,
savage a. , 1983 , mnras , 203 , 181 shanks t. , fong r. , boyle b. , peterson b.a . , 1987 , mnras , 227 , 739 shaver p.a . , 1984 ,
a&a , 136 , l9 shaver p.a . , robertson j.g . , 1983 ,
nature , 303 , 155 shaver p. a. , pierre m. , 1989 , a&a , 220 , 35 shull j.m .
, stocke j.t . , penton s. , 1996 , aj 111 , 72 smith e. p. , heckman t.m .
, 1990 , apj , 348 , 38 soltan a. , hasinger g. , 1994 , a&a , 288 , 77 songaila a. , cowie l.l . , 1996 , aj 112 , 335 stengler - larrea e. a. , webb j. k. 1993 , in chincarini g. , iovino a. , maccacaro t. , maccagni d. eds . , observational cosmology , asp conference series , 51 , 591 steidel c.c .
, adelberger k.l
. , dickinson m. , giavalisco m. , pettini m. , kellogg m. , 1998 , apj , 492 , 428 .
stephens a.w . ,
schneider d.p . , schmidt m. , gunn j.e . , weinberg d.h . , 1997 , aj , 114 , 41 stockton a. , 1978 , nature , 274 , 342 stockton a. , 1982 , apj , 257 , 33 ulmer a. , 1996 , apj , 473 , 110 van gorkom j. , carilli c.l .
, stocke j.t . , perlman e.s . ,
shull j.m . , 1996 , aj , 112 , 1397 villumsen j.v . , freudling w. , da costa l.n
, 1997 , apj , 481 , 578 wang b. , 1995 , apj , 444 , 17 webb j. k. , 1987 , in hewett a. , burbidge g. , fang l. z. eds .
124 , observational cosmology .
reidel , dordrecht , p. 803
webb j. k. , barcons x. , 1991 , mnras , 250 , 270 webster a. , 1976 , mnras , 175 , 71 webster a. , 1982 , mnras , 199 , 683 weymann r.j .
, jannuzi b.t . , lu l. , astro - ph/9806123 williger g.m . ,
hazard c. , baldwin j.a . ,
mcmahon r.g . , 1996 , apjs , 104 , 145 yee h.k.c . , green r.f . , 1984 ,
apj , 280 , 79 yee h.k.c . , green r.f . , 1987 , apj , 319 , 28 yee h.k.c . , ellingson e. , apj , 1993 , 411 , 43 young p. , sargent w. l. w. , boksenberg a. , 1982,apjs , 48 , 455 zuo l. , 1992 , mnras , 258 , 45 | qsos and radio - galaxies , together with the cmb , `` normal '' galaxies and clusters , represent the main source of information about the origin and evolution of the large scale structure .
they can be used either directly , as tracers of the density peaks , or as cosmic lighthouses , probing the universe along the line of sight with their conspicuous flow of photons .
observations of the qso and radio - galaxy clustering and its evolution interweave information about their nature and the cosmology that will hopefully be disentangled with the map and planck missions .
the amplitude of the qso clustering has been measured to be @xmath0 at @xmath1 and appears to increase with redshift .
these properties are consistent with a scenario in which the quasars , with a lifetime of @xmath2 , sparsely sample halos of mass greater than @xmath3 m@xmath4 .
radio - galaxies are found to cluster on similar scales .
the study of qso and radio - galaxy environments is in qualitative agreement with their clustering properties .
absorption - line systems observed in high - resolution spectra of qsos show structure , including voids / underdensities and overdensities , from the smallest scales to tens of mpc .
qso pairs , groups and lenses allow us to get a 3-d glimpse of the absorbers , providing invaluable information about their size , ionization , mass .
great progress in this field is expected from the new instrumentation coming into operation at the 8-m class telescopes .
much work remains to be done also at low redshift to clarify whether the absorption systems have a direct physical connection with galaxies or a significant fraction of them is caused by a truly intergalactic medium .
psfig 24.5 cm 16.5 cm -2.5 cm -2.2 cm invited review at the mpa / eso cosmology conference `` evolution of large - scale structure : from recombination to garching '' garching , germany , 2 - 7 august , 1998 . |
the nature of the low energy excitations in the hubbard model has attracted a great deal of attention .
it is well established that at half filling the ground state is an antiferromagnetic ( af ) insulator .
also , there exists conclusive evidence which indicates that antiferromagnetism is rapidly suppressed upon doping @xcite .
close to half filling , a large amount of work suggests the existence of spin polarons , made of dressed holes , which propagate within a given sublattice with kinetic energy which in the strong coupling limit is of the order of @xmath4 @xcite , where @xmath2 is the hopping integral and @xmath5 the on site coulomb repulsion .
these results are consistent with similar calculations in the strong coupling , low doping limit of the hubbard model , the @xmath6 model@xcite .
there is also evidence for an effective attraction between these spin polarons@xcite .
however , recent and extensive monte carlo calculations for 0.85 filling and @xmath7 , have shown that the pairing correlations vanish as the system size or the interaction strength increases @xcite .
we have recently analyzed the dynamics of spin polarons @xcite and the interactions between them @xcite by means of a systematic expansion around mean field calculations of the hubbard model .
two spin polarons in neighboring sites experience an increase in their internal kinetic energy , due to the overlap of the charge cloud .
this repulsion is of the order of @xmath2 .
in addition , a polaron reduces the obstacles for the diffussion of another , leading to an assisted hopping term which is also of same order .
the combination of these effects is an attractive interaction at intermediate values of @xmath8 .
the purpose of this work is to discuss in detail the results and the approach proposed in @xcite .
we present new results which support the validity of our approach , highlighting the physically appealing picture of pairing that it provides .
an alternative scheme to go beyond the unrestricted hartree fock approximation is to supplement it with the gutzwiller projection method , or , equivalently , slave boson techniques @xcite .
these results are in agreement with the existence of significant effects due to the delocalization of the solutions , as reported here .
the rest of the paper is organized as follows . in section
ii we discuss the physical basis of our proposal and the way in which we implement the configuration interaction method .
a discussion of the limit of large @xmath8 in the undoped case is presented in section iii .
it is shown that , contrary to some expectations , the hartree - fock scheme reproduces correctly the mean field solution of the heisenberg model .
the systematic corrections analyzed here can be put in precise correspondence with similar terms discussed for quantum antiferromagnets .
results for the @xmath0 cluster are compared with exact results in section iv .
section v is devoted to discuss our results for a single hole ( spin polaron ) and for two or more holes .
hole correlations are also presented in this section .
the last section is devoted to the conclusions of our work .
we investigate the simplest version of the hubbard hamiltonian used to describe the dynamics of electrons in cuo@xmath9 layers , namely , @xmath10 @xmath11 @xmath12 the hamiltonian includes a single atomic orbital per lattice site with energy @xmath13=0 .
the sums are over all lattice sites @xmath14 of the chosen cluster of the square lattice and/or the @xmath15 component of the spin ( @xmath16 ) .
the operator @xmath17 destroys an electron of spin @xmath18 at site @xmath19 , and @xmath20 is the local density operator .
@xmath21 is the hopping matrix element between sites @xmath19 and @xmath22 ( the symbol @xmath23 denotes that the sum is restricted to all nearest neighbors pairs ) and @xmath24 is the intrasite coulomb repulsion . here
we take @xmath25 and @xmath26 , and the lattice constant as the unit of length .
as we shall only consider uhf solutions having a local magnetization pointing in the same direction everywhere in the cluster , we shall use the most simple version of the uhf approximation @xcite . within this approximation the effective mean field hamiltonian that accounts for the hubbard term
is written as , @xmath27 @xmath28 the full uhf hamiltonian is then written as , @xmath29 use of the unrestricted hartree fock ( uhf ) approximation in finite clusters provides a first order approximation to the spin polaron near half filling . as discussed elsewhere
, the uhf approximation describes well the undoped , insulating state at half filling @xcite ( see also next section ) .
a realistic picture of the spin wave excitations is obtained by adding harmonic fluctuations by means of the time dependent hartree fock approximation ( rpa)@xcite . at intermediate and large values of @xmath8 , the most stable hf solution with a single hole is a spin polaron@xcite . in this solution ,
approximately half of the charge of the hole is located at a given site .
the spin at that site is small and it is reversed with respect to the antiferromagnetic background . the remaining charge is concentrated in the four neighboring sites .
a number of alternative derivations lead to a similar picture of this small spin bag@xcite .
a similar solution is expected to exist in the @xmath6 model . a schematic picture of the initial one hole and two holes hartree fock wavefunctions used in this work is shown in fig .
[ spins ] .
they represent the solutions observed at large values of @xmath8 for the isolated polaron and two spin polarons on neighboring sites .
the electronic spectrum of these configurations show localized states which split from the top of the valence band . as usual in mean field theories , the uhf solutions for an arbitrary number of holes @xcite , such as the spin polaron solution described above , break symmetries which must be restored by quantum fluctuations .
in particular , it breaks spin symmetry and translational invariance ( see fig . [ spins ] ) .
spin isotropy must exist in finite clusters .
however , it is spontaneously broken in the thermodynamic limit , due to the presence of the antiferromagnetic background .
hence , we do not expect that the lack of spin invariance is a serious drawback of the hartree fock solutions ( this point is analyzed , in some detail in@xcite ) .
results obtained for small clusters @xcite show a slight improvement of the energy , which goes to zero as the cluster size is increased . on the other hand ,
translational invariance is expected to be present in the exact solution of clusters of any size . the way we restore translational invariance
is discussed in the following subsection . finally we know how to estimate the effects due to zero point fluctuations around the uhf ground state @xcite . for spin polarons
these corrections do not change appreciably the results , although they are necessary to describe the long range magnon cloud around the spin polaron @xcite .
we have improved the mean field results by following the procedure suggested years ago by some of us @xcite .
we hybridize a given spin uhf solution with all wavefunctions obtained from it by lattice translations . in the case of two or more holes point symmetry
has also to be restored .
this is accomplished by applying rotations to the chosen configuration .
configurations generated from a given one through this procedure are degenerate in energy and interact strongly . here
we have also investigated the effect of extending the basis by including other configurations having different energies . in all cases
we include sets of wavefunctions with the lattice symmetry restored as mentioned . in a path integral formulation
, this procedure would be equivalent to calculating the contribution from instantons which visit different minima .
on the other hand , it is equivalent to the configuration interaction ( ci ) method used in quantum chemistry .
the ci wavefunction for a solution corresponding to @xmath30 electrons is then written as @xmath31 where the set @xmath32 is formed by some chosen uhf wavefunctions ( slater determinants ) plus those obtained from them by all lattice translations and rotations .
the coefficients @xmath33 are obtained through diagonalization of the exact hamiltonian . the same method , using homogeneous paramagnetic solutions as starting point , has been used in @xcite .
the wavefunctions forming this basis set are not in principle orthogonal .
thus , both wavefunctions overlap and non diagonal matrix elements of the hamiltonian need to be taken into account when mixing between configurations is considered .
if only configurations having the same energy and corresponding , thus , to the same uhf hamiltonian , are included , a physically sound decomposition of the exact hamiltonian is the following @xcite , @xmath34 in writing the matrix elements of this hamiltonian we should note that the basis formed by the wavefunctions @xmath35 is not orthogonal .
then , we obtain , @xmath36 where @xmath37 is the uhf energy of a given mean field solution , and the matrix elements of the overlap @xmath38 are given by @xmath39 this factorization is a consequence of the characteristics of the mean field solutions considered in this work ( only one component of the spin different from zero ) .
the specific expression for the matrix elements of the overlap is , @xmath40 where the number of particles for each component of the spin are determined from the usual conditions , @xmath41 and @xmath42 .
the @xmath43 are the monoelectronic wavefunctions corresponding to the slater determinant @xmath19 , @xmath44 @xmath45 being real coefficients obtained through diagonalization of the @xmath46 hamiltonian .
the matrix element of the exchange operator between slater determinants @xmath19 and @xmath22 is , @xmath47 where the matrix elements of @xmath48 between monoelectronic wavefunctions are given by , @xmath49 on the other hand the matrix elements of @xmath50 are , @xmath51 where each @xmath52 is given by an equation similar to eq . ( 10 ) .
the matrix elements of the density operator between monoelectronic wavefunctions are , @xmath53 if the ci basis includes wavefunctions having different uhf energies , the above procedure is not valid and one should calculate the matrix elements of the original exact hamiltonian .
although this is in fact reduced to calculate the matrix elements of the kinetic energy operator t , the procedure is slightly more costly ( in terms of computer time ) than the one described above .
the matrix elements of the exact hamiltonian in the basis of slater determinants are , @xmath54 where @xmath55 are given by an equation similar to eq .
( 10 ) , and the matrix elements of the kinetic energy operator between monoelectronic wavefunctions are @xmath56 the matrix elements involved in the calculation of hole
hole correlations for a given ci wavefunction are similar to those that appeared in the computation of the hubbard term .
in particular the following expectation value has to be computed , @xmath57 where @xmath58 .
terms of four operators in this expectation value are similar to eq .
those requiring more computer time involve @xmath59 with @xmath60 , @xmath61 calculations have been carried out on @xmath62 clusters with periodic boundary conditions ( @xmath63 ) and @xmath64 .
some results for lower values of @xmath5 are also presented .
note that @xmath65 is widely accepted as the most physically meaningful value of coulomb repulsion in these systems ( see for instance @xcite ) .
although larger clusters can be easily reached , no improvement of the results is achieved due to the short range character of the interactions ( see below ) .
the numerical procedure runs as follows .
localized uhf solutions are first obtained and the slater determinants for a given filling constructed .
the full ci basis set is obtained by applying all lattice traslations to the chosen localized uhf slater determinants all having the same @xmath15 component of the spin @xmath66 .
then we calculate the matrix elements of the overlap and of the hamiltonian in that basis set .
this is by far the most time consuming part of the whole calculation .
diagonalization is carried out by means of standard subroutines for non orthogonal bases .
the state of lowest energy corresponds to the ci ground state of the system for a given @xmath66 .
the desired expectation values are calculated by means of this ground state wavefunction .
the procedure is variational and , thus , successive enlargements of the basis set do always improve the description of the ground state .
the hartree fock scheme , for the undoped hubbard model in a square lattice gives an antiferromagnetic ground state , with a charge gap . at large values of @xmath8 ,
the gap is of order @xmath5 .
the simplest correction beyond hartree - fock , the rpa approximation , leads to a continuum of spin waves at low energies@xcite .
thus , the qualitative features of the solution are in good agreement with the expected properties of an antiferromagnetic insulator . there is , however , a great deal of controversy regarding the adequacy of mean field techniques in describing a mott insulator@xcite . in principle , the hubbard model , in the large @xmath5 limit , should describe well such a system . at half filling and large @xmath5 ,
the only low energy degrees of freedom of the hubbard model are the localized spins , which interact antiferromagnetically , with coupling @xmath4 .
it has been argued that , as long range magnetic order is not relevant for the existence of the mott insulator , spin systems with a spin gap are the most generic realization of this phase .
a spin gap is often associated with the formation of an rvb like state , which can not be adiabatically connected to the hartree fock solution of the hubbard model .
so far , the best examples showing these features are two leg spin 1/2 ladders@xcite .
recent work@xcite indicates that , in the presence of magnetic order , the metal insulator transition is of the slater type , that is , coherent quasiparticles can always be defined in the metallic side .
these results seem to favor the scenario suggested in@xcite , and lend support to our mean field plus corrections approach . without entering into the full polemic outlined above ,
we now show that the method used here gives , in full detail , the results which can be obtained from the heisenberg model by expanding around the antiferromagnetic mean field solution@xcite .
such an expansion gives a consistent picture of the physics of the heisenberg model in a square lattice .
the ground state energy of the hartree fock solution in a @xmath0 cluster is compared to the exact value@xcite at large values of @xmath5 in table i. the corresponding heisenberg model is : @xmath67 in a @xmath0 cluster , the exact ground state energy is @xmath68 where @xmath69 @xcite , in good agreement with the results for the hubbard model .
the mean field energy can be parametrized in the same way , except that @xmath70 .
this is the result that one expects for the mean field solution of the heisenberg model , which is given by a staggered configuration of static spins .
this solution can be viewed as the ground state of an anisotropic heisenberg model with @xmath71 and @xmath72 .
we now analyze corrections to the hartree fock solution by hybridizing it with mean field wavefunctions obtained from it by flipping two neighboring spins ( hereafter referred to as sf ) .
these solutions are local extrema of the mean field solutions in the large @xmath5 limit . in table
i we show the energy difference between these states and the antiferromagnetic ( af ) hartree fock ground state , and their overlap and matrix element also with the ground state .
we have checked that these are the only wavefunctions with a non negligible mixing with the ground state .
the overlap goes rapidly to zero , and the energy difference and matrix elements adjust well to the expressions @xmath73 @xmath74 these are the results that one obtains when proceeding from the heisenberg model .
these values , inserted in a perturbative analysis of quantum corrections to the ground state energy of the heisenberg model@xcite , lead to excellent agreement with exact results ( see also below ) .
as already pointed out , in the ci calculation of the ground state energy we only include the mean field wavefunctions with two neighboring spins flipped .
restoring point symmetry gives a total of 4 configurations , while applying lattice translations leads to a set of @xmath75 configurations ( remember that configurations on different sublattices do not interact ) to which the af wavefunction has to be added . in the case of the @xmath0 cluster
the set has a total of 33 configurations .
the ci energy for this cluster is given in table i along with the exact and the uhf energies .
it is noted that the ci calculation reduces in 50% the difference between the exact and the mean field result .
improving this results would require including a very large set , as other configurations do only decrease the ground state energy very slightly . in the large u limit , the largest interaction is @xmath76 .
then , neglecting the overlap between the af and the sf mean field solutions , the ci energy of the ground state can be approximated by , @xmath77\ ] ] for @xmath78 this expression gives @xmath79=-1.421 , in excellent agreement with the ci result given in table i. note that a perturbative calculation of the corrections of the ground state energy in finite clusters is somewhat tricky , as the matrix element scales with @xmath80 , where @xmath81 is the number of sites in the cluster , while the energy difference is independent of @xmath82 .
the first term in a perturbative expansion ( pe ) coincides with the first term in the expansion of the square root in eq .
( 21 ) , @xmath83 in agreement with the result reported in @xcite .
although this correction to the af energy has the expected size dependence for an extensive magnitude ( it is proportional to the number of sites @xmath82 ) and gives an energy already very similar to the exact , it was obtained by inconsistently expanding in terms of a parameter that can be quite large .
for instance in the @xmath0 cluster and @xmath5=50 , @xmath84 , close to the exact result ( table i ) while @xmath85 much larger than 1 .
thus , perturbation theory is doomed to fail even for rather small clusters . on the other hand ,
the ci calculation described above introduce a correction to the af energy which has not the correct size dependence .
this can be easily checked in the large cluster limit in which the ci energy can be approximated by @xmath86 while the correct expression should scale as @xmath82 , because , in large clusters , the difference between the exact and the hartree - fock ground state energies must be proportional to @xmath82 , irrespective of the adequacy of the hartree - fock approximation .
thus , one obtains a better approximation to the ground state energy in the thermodynamic limit , by using the perturbative calculations in small clusters and extrapolating them to large clusters , as in the related @xmath6 model@xcite . in any case
, the problem outlined here does not appear when calculating corrections to localized spin textures , such as the one and two spin polarons analyzed in the next sections .
the relevant properties are associated to the size of the texture , and do not scale with the size of the cluster they are embedded in . from the previous analysis
, we can safely conclude that our scheme gives a reliable approximation to the undoped hubbard model in a square lattice in the strong coupling regime , @xmath87 .
we can not conclude whether the method is adequate or not for the study of models which exhibit a spin gap .
it should be noted , however , that a spin gap needs not only be related to rvb like ground states .
a spin system modelled by the non linear sigma model can also exhibit a gap in the ground state , if quantum fluctuations , due to dimensionality or frustration , are sufficiently large . in this case , a mean field approach plus leading quantum corrections should be qualitatively correct .
in order to evaluate the performance of our approach we have calculated the ground state energy of two holes in the @xmath0 cluster and compared the results with those obtained by means of the lanczos method @xcite . the results are reported in tables ii iv , where the energies for one hole are also given for the sake of completness ( a full discussion of this case can be found in @xcite , see also below ) . in the case of one hole the standard spin polaron solution ( fig .
[ spins ] ) plus those derived from it through lattice translations form the basis set . for two holes
we consider solutions with @xmath88 or 1 . in the first case
we include either the configuration having the two holes at the shortest distance , i.e. , separated by a ( 1,0 ) vector and/or at the largest distance possible , that is separated by a ( 2,1 ) vector , and those obtained from them through rotations .
the basis used for the two polarons at the shortest distance is shown in fig .
[ bipolaronci ] .
the set of these four configuration has the proper point symmetry .
again , lattice translations are applied to these configurations to produce the basis set with full translational symmetry . on the other hand , wavefunctions with @xmath89
can be constructed by including configurations with the two holes separated by vectors ( 1,1 ) and/or ( 2,2 ) .
the results for the energies of wavefunctions with @xmath88 for several values of the interaction parameter @xmath5 are reported in tables ii and iii . as found for a single hole
, the kinetic energy included by restoring the lattice symmetry , improves the wavefunction energies @xcite .
the improvement in the energy is larger for intermediate @xmath5 .
for instance for @xmath90 a 10% gain is noted . within uhf ,
the solution with the holes at the largest distance is more favorable for @xmath91 . instead , restoring the translational and point symmetries favors the solution with the holes at neighboring sites for all @xmath5 shown in the tables .
the results also indicate that the correction introduced by this procedure does no vanish with @xmath5 . a more detailed discussion of the physical basis of this result along with results for larger values of @xmath5 and larger clusters will be presented in the following section . on the other hand
the energies get closer to the exact energies ( see table ii ) .
a further improvement in the energy is obtained by including both uhf configurations , namely , \{1,0 } and \{2,1}. this improvement is larger for intermediate @xmath5 and vanishes as @xmath5 increases ( table iii ) .
other configurations , such as that proposed in @xcite in which the two holes lie on neighboring sites along a diagonal and a neighboring spin is flipped , may contribute to further improve the ci energy of the ground state .
it is interesting to compare these results with those corresponding to wavefunctions with @xmath89 also reported in table iv .
it is noted that for @xmath92 the energy of the solution including all configurations from the set \{1,1 } is smaller than those obtained with all configurations from either the set \{1,0 } or the set \{2,1}. however , the wavefunction constructed with all configurations from the last two sets is more favorable than the best wavefunction with @xmath89 .
the latter is in agreement with exact calculations @xcite which obtained a ground state wavefunction with @xmath88 .
here we only consider the quasiparticle band structure associated to the single polaron , the energy gain induced through restoration of translational symmetry has been considered elsewhere @xcite . the calculated dispersion band of a single polaron is shown in fig . [ polaron ] . because of the antiferromagnetic background , the band has twice the lattice periodicity .
exact calculations in finite clusters do not show this periodicity , as the solutions have a well defined spin and mix different background textures . as cluster sizes are increased , however , exact solutions tend to show the extra periodicity of our results .
we interpret this as a manifestation that spin invariance is broken in the thermodynamic limit , because of the antiferromagnetic background .
hence , the lack of this symmetry in our calculations should not induce spurious effects .
[ polaron ] shows the polaron bandwidth as a function of @xmath5 .
it behaves as @xmath93 , the fitted law being @xmath94 this result indicates that the band width tends to zero as @xmath5 approaches infinite , as observed in the results for the energy gain reported in @xcite .
our scheme allows a straightforward explanation of this scaling . without reversing the spin of the whole background ,
the polaron can only hop within a given sublattice .
this implies an intermediate virtual hop into a site with an almost fully localized electron of the opposite spin .
the amplitude of finding a reversed spin in this new site decays as @xmath93 at large @xmath5 . on the other hand , we find that the dispersion relation can be satisfactorily fitted by the expression : @xmath95 + 4t_{22 } \cos ( 2 k_x ) \cos ( 2 k_y ) + \nonumber \\ 4 t_{31 } [ \cos ( 3 k_x ) \cos ( k_y ) + \cos ( k_x ) \cos ( 3 k_y ) ] .\end{aligned}\ ] ] for @xmath96 , we get @xmath97 , @xmath98 , @xmath99 , and @xmath100 .
all hopping integrals vanish as @xmath93 in the large @xmath5 limit for the reason given above .
also the energy gain with respect to uhf @xcite behaves in this way .
all these features are in good agreement with known results@xcite for both the hubbard and the @xmath6 models .
we now consider solutions with two spin polarons .
the relevant uhf solutions are those with @xmath101 ( solutions with @xmath89 will also be briefly considered ) . in order for the coupling to be finite , the centers of the two spin polarons must be located in different sublattices .
the mean field energy increases as the two polarons are brought closer , although , for intermediate and large values of @xmath5 , a locally stable hartree fock solution can be found with two polarons at arbitrary distances . we have not attempted to do a full ci analysis of all possible combinations of two holes in a finite cluster .
instead , we have chosen a given mean field solution ( uhf ) and hybridized it with all others obtained by all lattice translations and rotations .
some results of calculations in which more than one uhf solution is included will be also presented .
clusters of sizes up to @xmath102 were studied which , as in the case of the polaron , are large enough due to the short range interactions between different configurations . the basis used for the two polarons at the shortest distance
is shown in fig .
[ bipolaronci ] .
this procedure leads to a set of bands , whose number depends on the number of configurations included in the ci calculation . for instance
if the basis set of fig . [ bipolaronci ] is used four bands are obtained ( see also below ) .
like in the single polaron case , we obtain a gain in energy ( with respect to uhf ) , due to the delocalization of the pair .
the numerical results for @xmath103=6 , 8 and 10 and @xmath5 in the range @xmath104 are shown in the inset of fig .
[ difference ] .
they can be fitted by the following straight lines , @xmath105 @xmath106 where ( 26a ) corresponds to holes at the shortest distance and ( 26b ) to holes at the largest distance .
note that , whereas in the case of the holes at the largest distance , the gain goes to zero in the large @xmath5 limit , as for the isolated polaron , when the holes are separated by a @xmath107 vector the gain goes to a finite value .
this result is not surprising , as the following arguments suggest .
the hopping terms in the bipolaron calculation , that are proportional to @xmath2 at large @xmath5 , describe the rotation of a pair around the position of one of the two holes .
each hole is spread between four sites . in order for a rotation to take place
, one hole has to jump from one of these sites into one of the rotated positions .
this process can always take place without a hole moving into a fully polarized site with the wrong spin .
there is a gain in energy , even when @xmath108 . in the single polaron case ,
the motion of a hole involves the inversion of , at least , one spin , which is fully polarized in the large @xmath5 limit .
because of this , hybridization gives a vanishing contribution to the energy as @xmath109 .
the results discussed above are in line with those for the width of the quasiparticle band .
the numerical results can be fitted by , @xmath110 @xmath111 thus , the total bandwidth of the two bands obtained for holes in neighboring sites does not vanish in the infinite @xmath5 limit ( as the energy gain reported in fig .
[ bipolaronci ] ) .
the internal consistency of our calculations is shown comparing the large @xmath5 behavior of the two holes at the largest distance possible with the corresponding results obtained for the isolated polaron ( compare this fitting with that given in eq .
( 24 ) ) .
hole interaction , i.e. , the difference between the energy of a state built up by all configurations with the two holes at the shortest distance ( separated by a vector of the set \{1,0 } ) and the energy of the state having the holes at the largest distance possible at a given cluster is depicted in fig . [ difference ] .
two holes bind for intermediate values of @xmath5 @xcite .
this happens because the delocalization energy tends to be higher than the repulsive contribution obtained within mean field .
the local character of the interactions is illustrated by the almost null dependence of the results shown in fig .
[ difference ] on the cluster size .
the only numerical calculations which discuss the binding of holes in the hubbard model are those reported in @xcite .
energetically , it is favorable to pair electrons in a @xmath0 cluster for values of @xmath8 greater than 50 .
the analysis of the correlation functions suggests that pairing in real space ceases at @xmath112 , leading to the suspicion of large finite size effects .
our results give that pairing dissappears at @xmath113 , which is consistent with the analysis in@xcite .
similar calculations for the t - j model give binding between holes for @xmath114@xcite .
taking @xmath4 , this threshold agrees well with our results . in order to clarify some aspects of the method ,
we have carried out a more detailed analysis of two hole solutions in @xmath115 clusters .
the results are presented in table v. within uhf the most favorable solution is that with the two holes at the largest distance ( 2,3 ) but for the smallest @xmath5 shown in the table .
the solution with the holes at the shortest distance ( 1,0 ) is only favored at small @xmath5 , while for @xmath116 even the solution with @xmath89 has a smaller energy . instead when the lattice symmetry is restored the solution with the holes at the shortest distance is the best for all @xmath5 excluding @xmath117 .
for such a large @xmath5 the wavefunction constructed with all configurations from \{2,3 } has the lowest energy .
the solution with @xmath89 is unfavorable for all @xmath5 shown in table v , in contrast with the results found in the @xmath0 cluster , indicating that size effects were determinant in the results for the smaller cluster .
including all configurations with @xmath66 either 1 or 0 does not change this trend .
the small difference between the results for \{1,0 } and those with all configurations with @xmath88 for large @xmath5 is misleading .
in fact , the weight of the configuration with the holes at the largest distance ( 2,3 ) in the final ci wavefunction increases with @xmath5 .
this will be apparent in the hole
hole correlations discussed in the following paragraph .
we have analysed the symmetry of the ground state wavefunction @xmath118 obtained with all configurations having the holes at the shortest distance .
the numerical results for all @xmath5 show that @xmath119 , where the @xmath120 are the four configurations shown in fig .
this symmetry corresponds to the @xmath121 symmetry , in agreement with previous theoretical studies of the hubbard and @xmath122 models @xcite .
the quasiparticle band structure for two holes has also been investigated .
the main interactions @xmath123 and the overlaps @xmath124 between the configurations are given in table vi ( the meaning of the symbols is specified in fig .
[ bipolaronci ] ) .
the results correspond to a @xmath115 cluster with the two holes at the shortest distance . at finite @xmath5 many interactions
contribute to the band , in table vi we only show the largest ones .
of particular significance is the @xmath125 interactions which accounts for the simultaneous hopping of the two holes .
this term , which clearly favors pairing , vanishes in the infinite @xmath5 limit , in line with the results for the hole hole interaction ( see above ) which indicate that pairing is not favored at large @xmath5 .
also in this limit @xmath126 . including only the interactions given in table vi
, the bands can be easily obtained from , @xmath127 neglecting the overlap , the bands are given by , @xmath128 ^ 2 + 4t_2 ^ 2 ( 1+{\rm cos}k_x)(1+{\rm cos}k_y)}\ ] ] in the infinite @xmath5 limit ( @xmath129 and @xmath130 ) the bands are simply , @xmath131 @xmath132 note that , as in the single hole case and due to the antiferromagnetic background , the bands have twice the lattice periodicity . the dispersionless band has also been reported in @xcite and , in our case , it is a consequence of the absence of two hole hopping in the infinite @xmath5 limit ( @xmath129 ) .
our results , however , disagree with the conclusions reached in @xcite concerning the absence of hole attraction .
we find a finite attraction for holes at intermediate @xmath5 s .
it is interesting to note that our effective hopping is of order @xmath2 , and not of order @xmath93 as in citeck98 .
this effect is due to the delocalized nature of the single polaron texture ( 5 sites , at least ) , and it does not correspond to a formally similar term which can be derived from the mapping from the hubbard to the t - j model@xcite . the results for the hole hole correlation , @xmath133 , as function of the hole hole distance @xmath134 are reported in tables vii and viii . the normalization @xmath135 has been used .
the results correspond to ci wavefunctions with @xmath88 and were obtained including all configurations from either the set \{1,0 } or from the sets \{1,0 } , \{1,2 } , \{3,0 } and \{2,3}. the results are in qualitative agreement with those in @xcite .
when comparing with results obtained for the t - j model , one must take into account that , in the hubbard model , the hole hole correlation , as defined above , can take negative values ( see appendix ) .
this is due to the appearance of configurations with double occupied sites , which are counted as negative holes .
aside from this effect , our results describe well a somewhat puzzling result found in the t - j model ( see for instance @xcite ) : the maximum hole hole correlation occurs when the two holes are in the same sublattice , at a distance equal to @xmath1 times the lattice spacingi @xcite .
this result follows directly from the delocalized nature of the spin polarons , as seen in fig .
[ spins ] .
the center of each spin polaron propagates through one sublattice only , but the electron cloud has a finite weight in the other one , even when @xmath108 .
this effect is noted in all cases but for @xmath117 with all configurations . in that case
there is not a clear maximum and the correlations are appreciable even at rather large distances .
the reason for this behavior is that for large @xmath5 the configuration with the holes at the largest distance , namely , \{2,3 } , have the lowest energy and , thus , a large weight in the ci wavefunction .
this is consistent with the fact that no attraction was observed at large @xmath5 ( see fig . [
difference ] ) .
finally we note that the slower decrease with distance of hole hole correlations , obtained for @xmath136 including configurations from the four sets ( table viii ) may be a consequence of the decrease in the difference between uhf and ci energies as @xmath5 diminishes ( see fig .
an interesting question is whether the holes would tend to segregate when more holes are added to the cluster . in order to investigate this point
, we have calculated total energies for four holes on @xmath102 clusters with the holes either centered on a square , or located on two bipolarons separated by a ( 5,5 ) vector and with the holes at the shortest distance .
two ( four ) configurations ( plus translations ) were included in each case . in the case of two bipolarons only configurations in which the two bipolarons are rotated simultaneously are included .
other possible configurations have different energies and contribute to a less extent to the wavefunction . in any case , increasing the size of the basis set would not have changed the essential conclusion of our analysis ( see below ) .
the results for several values of @xmath5 are shown in table ix .
we note that already at the uhf level the solution with two separated bipolarons has a lower energy .
the coulomb repulsion term in the hamiltonian does not favor the configuration with the aggregated holes but for very small @xmath5 . restoring lattice symmetry decreases the energy in both cases to an amount which in neither case vanishes in the infinite @xmath5 limit .
the decrease is slightly larger in the case of the four holes on a square .
this result can be understood by noting that the holes move more freely ( producing the smallest distortion to the af background ) when the charge is confined to the smallest region possible . in any case , this is not enough to compensate the rather important difference in energy between the two cases at the uhf level .
these results indicate that for large and intermediate @xmath5 no hole segregation takes place and that the most likely configuration is that of separated bipolarons .
as discussed above , in the large @xmath5 limit the bipolaron moves over the whole cluster due to the interactions among the four mean field wavefunctions of fig .
[ bipolaronci ] ( interactions @xmath137 and @xmath138 in fig .
[ bipolaronin ] ) .
this mechanism can be viewed as another manifestation of hole assisted hopping .
the possibility of hole assisted hopping has been already considered in @xcite , although in a different context .
it always leads to superconductivity . in our case
, we find a contribution , in the large @xmath5 limit , of the type : @xmath139 this term admits the bcs decoupling @xmath140 .
it favors superconductivity with either @xmath141 or @xmath142 wave symmetry , depending on the sign of @xmath143 .
since we find @xmath144 , @xmath142wave symmetry follows .
we have analyzed the leading corrections to the hartree fock solution of the hubbard model , with zero , one and two holes .
we show that a mean field approach gives a reasonable picture of the undoped system for the entire range of values of @xmath8 .
the main drawback of mean field solutions in doped systems is their lack of translational invariance .
we overcome this problem by using the configuration interaction method . in the case of one hole ,
the localized spin polaron is replaced by delocalized wavefunctions with a well defined dispersion relation .
the bandwidth , in the large @xmath5 limit , scales as @xmath145 , and the solutions correspond to spin polarons delocalized in a given sublattice only . as in the undoped case , these results are in good agreement with other numerical calculations , for the hubbard and t - j models .
the same approach is used to analyze the interactions between pairs of holes .
we first obtain hartree fock solutions with two holes at different separations . from comparing their respective energies , and also with single hole solutions , we find a short range repulsive interaction between holes .
this picture is significantly changed when building delocalized solutions .
the energy gain from the delocalization is enough to compensate the static , mean field , repulsion .
there is a net attractive interaction for @xmath146 , approximately .
the correlations between the holes which form this bound state are in good agreement with exact calculations , when available .
the state has @xmath121 symmetry . in this range of parameters
, we find no evidence of hole clustering into larger structures .
a further proof of the efficiency ot the present ci approach results from a comparison with the ci approach of @xcite , that is based upon an extended basis ( @xmath147space ) . for a @xmath115 cluster and @xmath148 the uhf localized solution
has an energy of -31.747 .
a ci calculation including 36 localized configurations lowers the energy down to -31.972 .
this has to be compared with the result reported in @xcite obtained by means of 2027860 extended configurations , namely , -30.471 .
the difference between the two approaches should further increase for larger @xmath5 .
we have not applied the same technique to other hartree fock solutions which have been found extensively in the hubbard model : domain walls separating antiferromagnetic regions @xcite .
the breakdown of translational symmetry associated with these solutions is probably real and not just an artifact of the hartree fock solution , as in the previous cases .
hybridization of equivalent solutions can , however , stabilize domain walls with a finite filling , which are not the mean field solutions with the lowest energy .
because of the qualitative differences between spin polarons and domain walls , we expect a sharp transition between the two at low values of @xmath8 .
note , however , that the scheme presented here , based on mean field solutions plus corrections , is equally valid in both cases .
financial support from the cicyt , spain , through grants pb96 - 0875 , pb96 - 0085 , pb95 - 0069 , is gratefully acknowledged .
here we explicitly calculate the hole
hole correlations in the hydrogen molecule described by means of the hubbard model .
let us call @xmath149 and @xmath150 the operators that create a particle with spin @xmath18 at sites @xmath151 and @xmath152 respectively .
the ground state wavefunction has @xmath88 and is given by , @xmath153 where , with @xmath157 the wavefunctions in eq .
( [ h2 ] ) are orthonormalized , namely , @xmath158 .
the result for the hole hole correlations on different sites is , @xmath159 as @xmath160 when @xmath161 , this expectation value varies from -0.5 to 0.0 .
thus it can take negative values as found in the case of clusters of the square lattice . instead , the hole
hole correlation on the same site is given by , @xmath162 which is positive for all values of @xmath5 .
particle correlations are obtained by adding 1 to these results .
kampf , phys .
rep .. * 249 * , 219 ( 1994 ) .
d.j . scalapino , phys .
* 250 * , 329 ( 1995 ) .
n. bulut , d. j. scalapino and s. r. white , phys .
* 72 * , 705 ( 1994 ) .
r. preuss , w. van der linden and w. hanke , phys .
* 75 * , 1344 ( 1995 ) .
e. dagotto , a. nazarenko and m. boninsegni , phys .
lett . * 73 * , 728 ( 1994 ) . p. w. leung and r. j. gooding , phys .
b * 52 * , r15711 ( 1995 ) .
d. duffy , a. nazarenko , s. haas , a. moreo , j. riera and e. dagotto , phys .
b * 56 * , 5597 ( 1997 ) .
g. fano , f. ortolani and a. parola , phys .
b * 42 * , 6877 ( 1990 ) m. boninsegni and e. manousakis , phys .
b * 47 * , 11897 ( 1993 ) .
d. poilblanc , j. riera and e. dagotto , phys .
b * 49 * , 12318 ( 1994 ) .
e. dagotto , rev .
phys . * 66 * , 763 ( 1994 ) .
k. kuroki and h. aoki , preprint , cond - mat/9710019 . c. gazza , g. b. martins , j. riera and e. dagotto , preprint cond - mat/9803314 ) .
s. zhang , j. carlson and j. e. gubernatis , phys .
* 78 * , 4486 ( 1997 ) .
e. louis , g. chiappe , j. galn , f. guinea and j. a. vergs , phys .
b * 48 * , 426 ( 1993 ) .
e. louis , g. chiappe , f. guinea , j. a. vergs and e. v. anda , phys .
b * 48 * , 9581 ( 1993 ) .
e. louis , f. guinea , m.p .
sancho and j.a .
vergs , europhys .
* 44 * , 229 ( 1998 ) .
g. seibold , e. sigmund and v. hizhnykov , phys .
b * 57 * , 6937 ( 1998 ) .
g. seibold , phys .
b , in press ( cond - mat/9809113 ) .
j. a. vergs , e. louis , p.s .
lomdahl , f. guinea and a. r. bishop , phys .
b * 43 * , 6099 ( 1991 ) . f. guinea , e. louis and j. a. vergs , phys . rev .
b * 45 * , 4752 ( 1992 ) .
j. e. hirsch , phys .
lett . * 59 * , 228 ( 1987 ) .
a. kampf and j. r. schrieffer , phys .
b * 41 * , 6399 ( 1990 ) .
e. dagotto and j. r. schrieffer , phys .
b * 43 * , 8705 ( 1990 ) .
a. auerbach , _ interacting electrons and quantum magnetism _ , springer verlag , new york ( 1994 ) .
e. louis , j. galn , f. guinea , j. a. vergs and j. ferrer , phys .
( b ) * 173 * , 715 ( 1992 ) .
a. ramsak and p. horsch , phys .
b * 57 * , 4308 ( 1998 ) . b. friedman and g. levine , phys .
b * 55 * , 9558 ( 1997 ) .
s. r. white and d. j. scalapino , phys .
b * 55 * , 6504 ( 1997 ) .
r. b. laughlin _ a critique of two metals _ , preprint cond - mat/9709195 ( 1997 ) . p. w. anderson and g. baskaran , _ a critique of `` a critique of two metals " _
, preprint cond - mat/9711197 ( 1997 ) .
r. chitra and g. kotliar , _ dynamical mean field theory of the antiferromagnetic metal to antiferromagnetic insulator transition _ , preprint cond - mat/9811144 ( 1998 ) .
e. dagotto and t. m. rice , science * 271 * , 618 ( 1996 ) .
e. w. carlson , s. a. kivelson , z. nussinov and v. j. emery , phys .
b * 57 * , 14704 ( 1998 ) .
a. w. sandvik , phys .
b * 56 * , 11678 ( 1997 ) .
note that at large @xmath5 the ferromagnetic solution sets in . in ref .
@xcite it was shown that the energy of two holes in a ferromagnetic @xmath62 cluster could be fitted by @xmath163 .
comparing the present numerical results with this energy gives a transition to the ferromagnetic solution approximately at @xmath164 , for the cluster sizes investigated here .
e. dagotto , rev .
* 66 * , 763 ( 1994 ) .
m. cyrot and d. pavuna , _ introduction to superconductivity and high t@xmath165 materials _ , world scientific ( 1992 ) .
s. trugman , phys .
b * 37 * , 1597 ( 1988 ) .
j. riera and e. dagotto , phsy .
b * 57 * , 8609 ( 1997 ) . note that this results is obtained whithout including configurations which clearly favor it , such as that proposed in @xcite in which the two holes lie on neighboring sites along a diagonal and a neighboring spin is flipped . j. e. hirsch , phys . rev .
b * 48 * , 3327 ( 1993 ) .
a. v. dotsenko , phys .
b * 57 * , 6917 ( 1998 ) . j. zaanen and o. gunnarsson , phys .
b * 40 * , 7391 ( 1989 ) .
d. poilblanc and t. m. rice , phys .
b * 39 * , 9749 ( 1989 ) . h. j. schulz , phys .
* 64 * , 1445 ( 1990 ) . .
exact @xcite , uhf and ci ( see text ) energies of the hubbard model versus @xmath5 , for the half filled 4 @xmath166 4 cluster .
all energies are given in units of the hopping integral @xmath2 .
the uhf solution of lowest energy corresponds to the antiferromagnetic ( af ) configuration while the ci results were obtained by adding to the af configuration the 32 configurations having two neigboring spins flipped ( referred to as sf ) .
the interaction and overlap between the af and the sf configurations ( @xmath167 and @xmath168 ) are also given . [ cols="<,^,^,^,^,^,^ " , ] | the interactions between holes in the hubbard model , in the low density , intermediate to strong coupling limit , are investigated by systematically improving mean field calculations .
the configuration interaction basis set is constructed by applying to local unrestricted hartree - fock configurations all lattice translations and rotations .
it is shown that this technique reproduces , correctly , the properties of the heisenberg model , in the limit of large u. upon doping , dressed spin polarons in neighboring sites have an increased kinetic energy and an enhanced hopping rate .
both effects are of the order of the hopping integral and lead to an effective attraction at intermediate couplings .
the numerical results also show that when more than two holes are added to the system , they do not tend to cluster , but rather hole pairs remain far appart .
hole
hole correlations are also calculated and shown to be in qualitative agreement with exact calculations for @xmath0 clusters .
in particular our results indicate that for intermediate coupling the hole hole correlation attains a maximum when the holes are in the same sublattice at a distance of @xmath1 times the lattice spacing , in agreement with exact results and the @xmath2@xmath3 model .
the method is also used to derive known properties of the quasiparticle band structure of isolated spin polarons . |
the constrained voter model has been originally introduced in @xcite to understand the opinion dynamics in a spatially structured population of leftists , centrists and rightists . as in the popular voter model @xcite , the individuals are located on the vertex set of a graph and
interact through the edges of the graph at a constant rate .
however , in contrast with the classical voter model where , upon interaction , an individual adopts the opinion of her neighbor , it is now assumed that this imitation rule is suppressed when a leftist and a rightist interact . in particular
, the model includes a social factor called homophily that prevents agents who disagree too much to interact .
* model description * this paper is concerned with a natural generalization of the previous version of the constrained voter model that includes an arbitrary finite number @xmath0 of opinions and a so - called confidence threshold @xmath1 . having a connected graph @xmath3 representing the network of interactions ,
the state at time @xmath4 is a spatial configuration @xmath5 each individual looks at each of her neighbors at rate one that she imitates if and only if the opinion distance between the two neighbors is at most equal to the confidence threshold . formally , the dynamics of the system is described by the markov generator @xmath6 \end{array}\ ] ] where configuration @xmath7 is obtained from @xmath8 by setting @xmath9 and where @xmath10 means that the two vertices are connected by an edge .
note that the basic voter model and the original version of the constrained voter model including the three opinions leftist , centrist and rightist can be recovered from our general model as follows : @xmath11 the main question about the constrained voter model is whether the system fluctuates and evolves to a global consensus or fixates in a highly fragmented configuration . to define this dichotomy rigorously
, we say that * fluctuation * occurs whenever @xmath12 and that * fixation * occurs if there exists a configuration @xmath13 such that @xmath14 in other words , fixation means that the opinion of each individual is only updated a finite number of times , therefore fluctuation and fixation exclude each other .
we define convergence to a global consensus mathematically as a * clustering * of the system , i.e. , @xmath15 note that , whenever @xmath16 , the process reduces to the basic voter model with @xmath0 instead of two different opinions for which the long - term behavior of the process is well known : the system on lattices fluctuates while the system on finite connected graphs fixates to a configuration in which all the individuals share the same opinion .
in particular , the main objective of this paper is to study fluctuation and fixation in the nontrivial case when @xmath17 . * main results * whether the system fluctuates or fixates depends not only on the two parameters but also on the initial distribution .
in particular , we point out that , throughout the paper , it will be assumed that the initial distribution is the product measure with constant densities . to avoid trivialities , we also assume that the initial density of each of the opinions is positive : @xmath18 for the constrained voter model on the one - dimensional torus with @xmath19 vertices , the mean - field analysis in @xcite suggests that , in the presence of three opinions and when the threshold is equal to one , the average domain length at equilibrium is @xmath20 ^ 2 \ \sim \ \frac{2 \rho_2}{\pi}\ ] ] when the initial density of centrists @xmath21 is small and @xmath19 is large .
vzquez et al .
@xcite also showed that these predictions agree with their numerical simulations from which they conclude that , when the initial density of centrists is small , the system fixates with high probability in a frozen mixture of leftists and rightists .
in contrast , it is conjectured in @xcite based on an idea in @xcite that the infinite system fluctuates and clusters whenever @xmath2 , which includes the threshold one model with three opinions introduced in @xcite . to explain this apparent disagreement
, we first observe that , regardless of the parameters , the system on finite graphs always fixate and there is a positive probability that the final configuration consists of a highly fragmented configuration , thus showing that spatial simulations of the necessarily finite system are not symptomatic of the behavior of its infinite counterpart .
our first theorem shows that the conjecture in @xcite is indeed correct .
[ th : fluctuation ] assume and @xmath2 . then , 1 .
the process on @xmath22 fluctuates and clusters . 2 .
the probability of consensus on any finite connected graph satisfies @xmath23 the intuition behind the proof is that , whenever @xmath2 , there is a nonempty set of opinions which are within the confidence threshold of any other opinions .
this simple observation implies the existence of a coupling between the constrained and basic voter models , which is the key to proving fluctuation .
the proof of clustering is more difficult .
it heavily relies on the fact that the system fluctuates but also on an analysis of the interfaces of the process through a coupling with a certain system of charged particles .
in contrast , our lower bound for the probability of consensus on finite connected graphs relies on techniques from martingale theory .
note that this lower bound is in fact equal to the initial density of individuals who are in the confidence threshold of any other individuals in the system .
returning to the relationship between finite and infinite systems , we point out that the simulation pictures of figure [ fig : interface ] , which show two typical realizations of the process on the torus under the assumptions of the theorem , suggest fixation of the infinite counterpart in a highly fragmented configuration , in contradiction with the first part of our theorem , showing again the difficulty to interpret spatial simulations .
note also that , for the system on the one - dimensional torus with @xmath19 vertices , the average domain length at equilibrium is bounded from below by @xmath24 which , together with the second part of the theorem , proves that the average domain length scales like the population size when @xmath2 and that does not hold . while our fluctuation - clustering result holds regardless of the initial densities
provided they are all positive , whether fixation occurs or not seems to be very sensitive to the initial distribution .
also , to state our fixation results and avoid messy calculations later , we strengthen condition and assume that @xmath25 the next theorem looks at the fixation regime in three different contexts .
[ th : fixation ] assume .
then , the process on @xmath22 fixates in the following cases : 1 .
@xmath26 and @xmath27 is small enough .
2 . @xmath0 is large , @xmath28 and @xmath29 where @xmath30 3 .
@xmath31 and @xmath32 and @xmath33 .
the first part of the theorem is the converse of the first part of theorem [ th : fluctuation ] , thus showing that the condition @xmath34 is critical in the sense that * when @xmath2 , the one - dimensional constrained voter model fluctuates when starting from any nondegenerate distributions whereas * when @xmath26 , the one - dimensional constrained voter model can fixate even when starting from a nondegenerate distribution .
the last two parts of the theorem specialize in two particular cases .
the first one looks at uniform initial distributions in which all the opinions are equally likely . for simplicity ,
our statement focuses on the fixation region when the parameters are large but our proof is not limited to large parameters and implies more generally that the system fixates for all pairs of parameters corresponding to the set of white dots in the phase diagram of figure [ fig : diagram ] for the one - dimensional system with up to twenty opinions .
note that the picture suggests that the process starting from a uniform initial distribution fixates whenever @xmath28 even for a small number of opinions .
the second particular case returns to the slightly more general initial distributions but focuses on the threshold one model with four opinions for which fixation is proved when @xmath21 is only slightly less than one over the number of opinions = 0.25 .
this last result suggests that the constrained voter model with four opinions and threshold one fixates when starting from the uniform product measure , although the calculations become too tedious to indeed obtain fixation when starting from this distribution .
* structure of the paper * the rest of the article is devoted to the proof of both theorems . even though our proof of fluctuation - clustering and fixation differ significantly , a common technique we introduce to study these two aspects for the one - dimensional process is a coupling with a certain system of charged particles that keeps track of the discrepancies along the edges of the graph rather than the actual opinion at each vertex .
in contrast , our approach to analyze the process on finite connected graphs is to look at the opinion at each vertex and use , among other things , the optimal stopping theorem for martingales .
the coupling with the system of charged particles is introduced in section [ sec : coupling ] and then used in section [ sec : fluctuation ] to prove theorem [ th : fluctuation ] .
the proof of theorem [ th : fixation ] is more complicated and carried out in the last five sections [ sec : condition][sec : fixation - particular ] .
in addition to the coupling with the system of charged particles introduced in the next section , the proof relies on a characterization of fixation based on so - called active paths proved in section [ sec : condition ] and large deviation estimates for the number of changeovers in a sequence of independent coin flips proved in section [ sec : deviation ] .
to study the one - dimensional system , it is convenient to construct the process from a graphical representation and to introduce a coupling between the process and a certain system of charged particles that keeps track of the discrepancies along the edges of the lattice rather than the opinion at each vertex .
this system of charged particles can also be constructed from the same graphical representation . since the constrained voter model on general finite graphs
will be studied using other techniques , we only define the graphical representation for the process on @xmath22 , which consists of the following collection of independent poisson processes : * for each @xmath35 , we let @xmath36 be a rate one poisson process , * we denote by @xmath37 its @xmath38th arrival time . this collection of independent poisson processes is then turned into a percolation structure by drawing an arrow @xmath39 at time @xmath40 and , given a configuration of the one - dimensional system at time @xmath41 , we say that this arrow is * active * if and only if @xmath42 the configuration at time @xmath4 is then obtained by setting @xmath43 an argument due to harris @xcite implies that the constrained voter model starting from any configuration can indeed be constructed using this percolation structure and rule . from the collection of active arrows , we construct active paths as in percolation theory .
more precisely , we say that there is an * active path * from @xmath44 to @xmath45 , and write @xmath46 , whenever there exist @xmath47 such that the following two conditions hold : 1 . for @xmath48 , there is an active arrow from @xmath49 to @xmath50 at time @xmath51 .
2 . for @xmath52
, there is no active arrow that points at @xmath53 .
note that conditions 1 and 2 above imply that @xmath54 moreover , because of the definition of active arrows , the opinion of vertex @xmath55 at time @xmath4 originates and is therefore equal to the initial opinion of vertex @xmath56 so we call vertex @xmath56 the * ancestor * of vertex @xmath55 at time @xmath4
. one of the key ingredients to studying the one - dimensional system is to look at the following process defined on the edges : identifying each edge with its midpoint , we set @xmath57 and think of edge @xmath58 as being * empty whenever @xmath59 , * occupied by a pile of @xmath60 particles with positive charge whenever @xmath61 , * occupied by a pile of @xmath60 particles with negative charge whenever @xmath62 .
the dynamics of the constrained voter model induces evolution rules which are again markov on this system of charged particles .
assume that there is an arrow @xmath63 at time @xmath4 and @xmath64 indicating in particular that there is a pile of @xmath60 particles with positive charge at @xmath65 .
then , we have the following alternative : * there is no particle at edge @xmath66 or equivalently @xmath67 in which case the individuals at vertices @xmath55 and @xmath68 already agree so nothing happens .
* there is a pile of @xmath69 particles at edge @xmath66 in which case @xmath55 and @xmath68 disagree too much to interact so nothing happens .
* there is a pile of @xmath70 particles at @xmath66 in which case @xmath71 in particular , there is no more particles at edge @xmath66 and a pile of @xmath72 particles all with the common charge @xmath73 at edge @xmath74 .
similar evolution rules are obtained by exchanging the direction of the interaction or by assuming that we have @xmath75 from which we can deduce the following description : * piles with more than @xmath1 particles can not move therefore we call such piles * blockades * and the particles they contain * frozen * particles .
* piles with at most @xmath1 particles jump one step to the left or one step to the right at the same rate one therefore we call the particles they contain * active * particles . * when a pile with positive / negative particles jumps onto a pile with negative / positive particles , positive and negative particles *
annihilate * by pair which results in a smaller pile of particles all with the same charge .
we refer to figure [ fig : particles ] for an illustration of these evolution rules .
note that whether an arrow is active or not can also be characterized from the state of the edge process : @xmath76 in particular , active arrows correspond to active piles of particles .
the key ingredient to proving fluctuation of the one - dimensional system and estimating the probability of consensus on finite connected graphs is to partition the opinion set into two sets that we shall call the set of * centrist * opinions and the set of * extremist * opinions : @xmath77 note that the assumption @xmath2 implies that the set of centrist opinions is nonempty .
note also that both sets are characterized by the properties @xmath78 as shown in figure [ fig : partition ] which gives a schematic illustration of the partition .
fluctuation is proved in the next lemma using this partition and relying on a coupling with the voter model .
[ lem : fluctuation ] the process on @xmath22 fluctuates whenever @xmath2 and @xmath79 .
it follows from that centrist agents are within the confidence threshold of every other individual .
in particular , for each pair @xmath80 we have the transition rates @xmath81 and similarly @xmath82 now , we introduce the process @xmath83 since for all @xmath84 the transition rates @xmath85 are constant over all @xmath86 according to , we have the following local transition rate for this new process : @xmath87 using in place of and some obvious symmetry , we also have @xmath88 this shows that the spin system @xmath89 reduces to the voter model
. in particular , the lemma directly follows from the fact that the one - dimensional voter model itself , when starting with a positive density of each type , fluctuates , a result proved based on duality in @xcite , pp 868869 .
[ lem : cluster ] the process on @xmath22 clusters whenever @xmath2 and @xmath79 .
the proof strongly relies on the coupling with the voter model in the proof of the previous lemma . to begin with
, we define the function @xmath90 which , in view of translation invariance of the initial configuration and the evolution rules , does not depend on the choice of @xmath58 . note that , since the system of charged particles coupled with the process involves deaths of particles but no births , the function @xmath91 is nonincreasing in time , therefore it has a limit : @xmath92 as @xmath93 .
now , on the event that an edge @xmath58 is occupied by a pile of at least one particle at a given time @xmath4 , we have the following alternative : * @xmath65 is a blockade . in this case , since the centrist agents are within the confidence threshold of all the other agents , we must have @xmath94 but since the voter model @xmath89 fluctuates , @xmath95 in particular , at least of one of the frozen particles at @xmath58 is killed eventually .
* @xmath65 is a live edge . in this case , since one - dimensional symmetric random walks are recurrent , the active pile of particles at @xmath58 eventually intersects another pile of particles , and we have the following alternative : * * the two intersecting piles of particles have opposite charge , which results in the simultaneous death of at least two particles . *
* the two intersecting piles have the same charge and merge to form a blockade in which case we are back to the previous case : since the voter model @xmath89 fluctuates , at least one of the frozen particles in this blockade is killed eventually . *
* the two intersecting piles have the same charge and merge to form a larger active pile in which case the pile keeps moving until , after a finite number of collisions , we are back to one of the previous two possibilities : at least two active particles annihilate or there is creation of a blockade with at least one particle that is killed eventually . in either case , as long as there are particles , there are also annihilating events indicating that the density of particles @xmath91 is strictly decreasing as long as it is positive .
in particular , the density of particles decreases to zero so there is extinction of both the active and frozen particles : @xmath96 in particular , for all @xmath97 with @xmath98 , we have @xmath99 which proves clustering . + + the second part of the theorem , which gives a lower bound for the probability of consensus of the process on finite connected graphs , relies on very different techniques , namely techniques related to martingale theory following an idea from @xcite , section 3 . however , the partition of the opinion set into centrist opinions and extremist opinions is again a key to the proof .
[ lem : consensus ] for the process on any finite connected graph , @xmath100 the first step is to prove that the process that keeps track of the number of supporters of any given opinion is a martingale .
then , applying the martingale convergence theorem and optimal stopping theorem , we obtain a lower bound for the probability of extinction of the extremist agents , which is also a lower bound for the probability of consensus . for @xmath101 , we set @xmath102 and we observe that @xmath103 letting @xmath104 denote the natural filtration of the process , we also have @xmath105 indicating that the process @xmath106 is a martingale with respect to the natural filtration of the constrained voter model .
this , together with , implies that @xmath107 also is a martingale .
it is also bounded because of the finiteness of the graph therefore , according to the martingale convergence theorem , there is almost sure convergence to a certain random variable : @xmath108 and we claim that @xmath109 can only take two values : @xmath110 to prove our claim , we note that , invoking again the finiteness of the graph , the process gets trapped in an absorbing state after an almost surely stopping time so we have @xmath111 assuming by contradiction that @xmath112 gives an absorbing state with at least one centrist agent and at least one extremist agent . since the graph is connected , this implies the existence of an edge @xmath113 such that @xmath114 but then we have @xmath115 and @xmath116 showing that @xmath117 is not an absorbing state , in contradiction with the definition of time @xmath118 .
this proves that our claim is true .
now , applying the optimal stopping theorem to the bounded martingale @xmath107 and the almost surely finite stopping time @xmath118 and using , we obtain @xmath119 from which it follows that @xmath120 to conclude , we observe that , on the event that @xmath121 , all the opinions present in the system after the hitting time @xmath118 are within distance @xmath1 of each other therefore the process evolves according to a voter model after that time . since the only absorbing states of the voter model on finite connected graphs are the configurations in which all the agents share the same opinion , we deduce that the system converges to a consensus .
this , together with , implies that @xmath122 this completes the proof of the lemma .
the main objective of this section is to prove a sufficient condition for fixation of the constrained voter model based on certain properties of the active paths .
[ lem : fixation - condition ] for all @xmath123 , let @xmath124 then , the constrained voter model fixates whenever @xmath125 this is similar to the proof of lemma 2 in @xcite and lemma 4 in @xcite . to begin with ,
we define recursively a sequence of stopping times by setting @xmath126 in other words , the @xmath60th stopping time @xmath127 is the @xmath60th time the individual at the origin changes her opinion .
now , we define the following random variables and collection of events : @xmath128 the assumption together with reflection symmetry implies that the event @xmath129 occurs almost surely for some positive integer @xmath130 , which implies that @xmath131 since @xmath132 is the event that the individual at the origin changes her opinion infinitely often , in view of the previous inequality , in order to establish fixation , it suffices to prove that @xmath133 to prove equations , we let @xmath134 be the set of descendants of @xmath55 at time @xmath4 which , due to one - dimensional nearest neighbor interactions , is necessarily an interval and its cardinality , respectively . now , since each interaction between two individuals is equally likely to affect the opinion of each of these two individuals , the number of descendants of any given site is a martingale whose expected value is constantly equal to one .
in particular , the martingale convergence theorem implies that @xmath135 therefore the number of descendants of @xmath55 converges to a finite value .
since in addition the number of descendants is an integer - valued process , @xmath136 which further implies that , with probability one , @xmath137 finally , we note that , on the event @xmath129 , the last time the individual at the origin changes her opinion is at most equal to the largest of the stopping times @xmath138 for @xmath139 therefore @xmath140 according to .
this proves and the lemma .
in order to find later a good upper bound for the probability in and deduce a sufficient condition for fixation of the process , the next step is to prove large deviation estimates for the number of piles with @xmath60 particles with a given charge in a large interval . more precisely , the main objective of this section is to prove that for all @xmath60 and all @xmath141 the probability that @xmath142 decays exponentially with @xmath130 . note that , even though the initial opinions are chosen independently , the states at different edges are not independent .
for instance , a pile of particles with a positive charge is more likely to be surrounded by negative particles .
in particular , the result does not simply follow from large deviation estimates for the binomial distribution .
the main ingredient is to first show large deviation estimates for the number of so - called changeovers in a sequence of independent coin flips .
consider an infinite sequence of independent coin flips such that @xmath143 where @xmath144 is the outcome : heads or tails , at time @xmath60 .
we say that a * changeover * occurs whenever two consecutive coin flips result in two different outcomes .
the expected value of the number of changeovers @xmath145 before time @xmath130 can be easily computed by observing that @xmath146 and by using the linearity of the expected value : @xmath147 then , we have the following large deviation estimates for the number of changeovers .
[ lem : changeover ] for all @xmath141 , there exists @xmath148 such that @xmath149 to begin with , we let @xmath150 be the time to the @xmath151th changeover and notice that , since all the outcomes between two consecutive changeovers are identical , the sequence of coin flips up to this stopping time can be decomposed into @xmath151 strings with an alternation of strings with only heads and strings with only tails followed by one more coin flip .
in addition , since the coin flips are independent , the length distribution of each string is @xmath152 and lengths are independent . in particular , @xmath150 is equal in distribution to the sum of @xmath151 independent geometric random variables with parameters @xmath153 and @xmath154 , namely , we have @xmath155 now , using that , for all @xmath156 , @xmath157 and large deviation estimates for the binomial distribution implies that @xmath158 for a suitable constant @xmath159 and all @xmath130 large .
similarly , @xmath160 for a suitable @xmath161 and all @xmath130 large . combining , we deduce that @xmath162 taking @xmath163 and observing that @xmath164 , we deduce @xmath165 for a suitable @xmath166 and all @xmath130 large . in particular , for all @xmath130 sufficiently large , @xmath167 for suitable constants @xmath168 and @xmath169 and all @xmath130 sufficiently large . using the previous two inequalities and the fact that the event that the number of changeovers is equal to @xmath170 is also the event that the time to the @xmath170th changeover is less than @xmath130 but the time to the next changeover is more than @xmath130 , we conclude that @xmath171 for all @xmath130 sufficiently large .
this completes the proof .
+ + now , we say that an edge is of type @xmath172 if it connects an individual with initial opinion @xmath173 on the left to an individual with initial opinion @xmath60 on the right , and let @xmath174 denote the number of edges of type @xmath172 in the interval @xmath175 . using the large deviation estimates for the number of changeovers established in the previous lemma
, we can now deduce large deviation estimates for the number of edges of each type .
[ lem : edge ] for all @xmath141 , there exists @xmath176 such that @xmath177 for any given @xmath173 , the number of edges @xmath172 and @xmath178 with @xmath179 has the same distribution as the number of changeovers in a sequence of independent coin flips of a coin that lands on heads with probability @xmath180 .
in particular , applying lemma [ lem : changeover ] with @xmath181 gives @xmath182 for all @xmath130 sufficiently large .
in addition , since each @xmath173 preceding a changeover is independently followed by any of the remaining @xmath183 opinions , @xmath184 combining with large deviation estimates for the binomial distribution , conditioning on the number of edges of type @xmath185 for some @xmath186 , and using that @xmath187 @xmath188 we deduce the existence of @xmath189 such that @xmath190 for all @xmath130 large . similarly , there exists @xmath191 such that @xmath192 for all @xmath130 large .
the lemma follows from . + + note that the large deviation estimates for the initial number of piles of particles easily follows from the previous lemma .
finally , from the large deviation estimates for the number of edges of each type , we deduce the analog for a general class of weight functions that will be used in the next section to find a sufficient condition for fixation of the constrained voter model .
[ lem : weight ] let @xmath193 and assume that @xmath194 with @xmath195 for @xmath196 . for all @xmath141 , there exists @xmath197 such that @xmath198 first , we observe that @xmath199 this , together with lemma [ lem : edge ] , implies that @xmath200 for a suitable constant @xmath201 and all @xmath130 large .
in view of lemma [ lem : fixation - condition ] , in order to prove fixation , it suffices to show that the probability of the event in equation , that we denote by @xmath202 , tends to zero as @xmath203 .
let @xmath204 be the first time an active path starting from @xmath205 hits the origin , and observe that @xmath206 from which it follows that @xmath207 denote by @xmath208 the initial position of this active path and by @xmath209 the rightmost source of an active path that reaches the origin by time @xmath204 , i.e. , @xmath210 and define @xmath211 .
now , note that each blockade initially in the interval @xmath212 must have been destroyed , i.e. , turned into a set of active particles through the annihilation of part of the particles that constitute the blockade , by time @xmath204 .
moreover , all the active particles initially outside the interval @xmath212 can not jump inside the space - time region delimited by the two active paths implicitly defined in because the existence of such particles would contradict either the minimality of @xmath213 or the maximality of @xmath214 .
in particular , on the event @xmath202 , all the blockades initially in @xmath212 must have been destroyed before time @xmath204 by either active particles initially in @xmath212 or active particles resulting from the destruction of the blockades initially in @xmath212 . to estimate the probability of this last event
, we first give a weight of @xmath215 to each particle initially active by setting @xmath216 now , since each blockade with @xmath60 frozen particles can induce the annihilation of at least @xmath217 active particles after which there is a set of at most @xmath1 initially frozen particles becoming active , the weight of such an edge is set to @xmath218 , i.e. , @xmath219 the fact that the event @xmath202 occurs only if all the blockades initially in the interval @xmath212 are destroyed by either active particles initially in @xmath212 or active particles resulting from the destruction of the blockades initially in @xmath212 can be expressed as @xmath220 to find an upper bound for the probability of the event on the right - hand side of , we first compute the expected value of the weight function @xmath221 and then use the large deviation estimates proved in lemma [ lem : weight ] .
the next lemma gives an explicit expression of the expected value of the weight function and will be used repeatedly in the proofs of our last three theorems in order to identify sets of parameters in which fixation occurs . to prove this lemma as well as the so - called contribution of additional events later , we make use of the identities @xmath222 [ lem : expected - weight ] assume . then , @xmath223 where @xmath224 to begin with , we note that @xmath225 from which we deduce that @xmath226 for all @xmath227 while @xmath228 for @xmath229 .
it follows that @xmath230 also , decomposing the second sum in depending on whether the number of particles @xmath60 is larger or smaller than @xmath231 and changing variables , we obtain @xmath232 combining , we obtain @xmath233 which , recalling the notation in , becomes @xmath234 to further simplify the expected value , we note that @xmath235 then , plugging into , we get @xmath236 this completes the proof .
+ + to deduce theorem [ th : fixation].a , we observe that @xmath237 using the continuity of @xmath238 together with lemma [ lem : expected - weight ] gives @xmath239 then , using and applying lemma [ lem : weight ] with @xmath240 , we deduce @xmath241 this together with lemma [ lem : fixation - condition ] implies fixation .
we now specialize in the case of a uniform initial distribution : @xmath29 , which forces the initial density of each opinion to be equal to @xmath242 . in this case
, the expected value of the weight function reduces to the expression given in the following lemma .
[ lem : expected - weight - uniform ] assume with @xmath29 .
then , @xmath243 according to lemma [ lem : expected - weight ] , we have @xmath244 in other respects , using @xmath245 with @xmath246 together with gives @xmath247 the lemma follows from combining .
+ + letting @xmath248 and taking @xmath249 , the lemma implies that @xmath250 in particular , @xmath251 whenever @xmath0 is large and @xmath252 using again lemmas [ lem : fixation - condition ] and [ lem : weight ] as well as , we deduce as in the previous section that fixation occurs under the condition above when the parameters are large enough .
this is not exactly the assumption of our theorem .
note however that the weight function is defined based on a worst case scenario in which the active particles _ do their best _ to destroy the blockades and to turn as many frozen particles as possible into active particles . to improve the lower bound for the asymptotic critical slope from @xmath253 to @xmath254 ,
the idea is to take into account additional events that eliminate some active particles and find lower bounds for their contribution defined as the number of active particles they eliminate times their probability .
the events we consider are 1 .
* annihilation of active particles * due to the collision of a pile of active particles with positive charge with a pile of active particles with a negative charge , 2 .
* blockade formation * due to the collision of two piles of active particles with the same charge and total size exceeding the confidence threshold , 3 .
* blockade increase * due to the jump of a pile of active particles with a certain charge onto a blockade with the same charge .
more precisely , we introduce the events @xmath255 as well as the three events @xmath256 the next three lemmas give lower bounds for the contribution of these three events .
[ lem : contribution-1 ] the contribution of the event @xmath257 satisfies @xmath258 by conditioning on the possible values of @xmath259 , we get @xmath260 in particular , when @xmath261 and @xmath75 we have @xmath262 note that on the event @xmath263 the number of particles that are eliminated is twice the size of the smallest of the two active piles therefore the contribution of this event is given by @xmath264 where the factor 1/3 is the probability that there is an arrow pointing at @xmath55 before there is an arrow pointing at @xmath265 . using obvious symmetry , and , we deduce that @xmath266 this completes the proof .
[ lem : contribution-2 ] the contribution of the event @xmath132 satisfies @xmath267 first , we note that taking @xmath261 and @xmath268 in equation gives @xmath269 in addition , the event @xmath270 with @xmath271 replaces a set of @xmath272 active particles by a blockade of size @xmath272 so it induces the annihilation of at least @xmath273 active particles .
the contribution of @xmath270 is therefore @xmath274 using again some obvious symmetry together with and and the fact that the contribution above is a function of @xmath275 , we deduce that @xmath276 this completes the proof .
[ lem : contribution-3 ] the contribution of the event @xmath277 satisfies @xmath278 assume that @xmath279 and @xmath69 .
then , again implies that @xmath280 in addition , the event @xmath281 replaces an active pile of size @xmath173 and a blockade of size @xmath60 with a blockade of size @xmath272 so it induces the annihilation of @xmath282 active particles .
the contribution of @xmath281 is therefore @xmath283 where we use that the conditional probability is larger than @xmath284 using symmetry together with and , we get @xmath285 this completes the proof .
+ + using again lemma [ lem : fixation - condition ] and the large deviation estimates of lemma [ lem : weight ] , we deduce that the one - dimensional constrained voter model fixates whenever @xmath286 which , together with lemmas [ lem : expected - weight][lem : contribution-3 ] , gives the condition for fixation : @xmath287 letting again @xmath248 and taking @xmath249 , we obtain fixation whenever @xmath288 this gives fixation when @xmath0 is large and @xmath289 which completes the proof of theorem [ th : fixation].b .
in this last section , we assume , thus returning to slightly more general initial distributions , but specialize in the system with threshold one and four opinions , which is not covered by part b of the theorem .
note that , applying lemma [ lem : expected - weight ] with @xmath31 and @xmath32 , we obtain @xmath290 then , using that @xmath291 , we get @xmath292 in particular , using the same arguments as in the previous two sections , we deduce that the system with threshold one and four opinions fixates whenever @xmath293 to improve this condition to the one stated in the theorem , we follow the same strategy as in the previous section , namely we compute the contribution of the three events . using the specific value of the parameters allows us to significantly improve the lower bounds for the contribution of the three events .
this is done in the following three lemmas . to simplify the notation
, we define @xmath295 and partition the event @xmath257 into two events distinguishing between two types of initial conditions that result in two different contributions : @xmath296 also , we let @xmath118 be the time of the first active arrow that either starts or points at @xmath68 . since two particles are eliminated on @xmath257 , the contribution of the event @xmath297 is given by @xmath298 now , we let @xmath299 be the distance between vertex @xmath55 and the rightmost agent with either initial opinion 1 or initial opinion 3 to the left of @xmath55 , i.e. , @xmath300 and observe that , in order to have a change of opinion at @xmath55 before time @xmath118 , there must be a sequence of at least @xmath299 arrows all occurring before time @xmath118 . summing over all possible positions of this rightmost agent and using that @xmath118
is exponentially distributed with rate 4 , we obtain @xmath301 where the first term corresponds to the case where @xmath302 and is obtained by further conditioning on whether the arrow @xmath303 and the reverse arrow occur before time @xmath118 or not . combining our lower bound for the contribution of @xmath297 together with , we obtain @xmath304 repeating the same reasoning for the event @xmath305 gives @xmath306 conditioning on all possible values of @xmath299 where @xmath299 now keeps track of the position of the closest agent with opinion 2 to the left of @xmath55 , which is geometric with parameter @xmath21 , we get @xmath307 therefore , the contribution of the event @xmath305 is bounded by @xmath308 the lemma follows from the combination of and . note that the set of initial configurations on the event @xmath132 is @xmath310 again , we let @xmath118 be the time of the first active arrow that either starts or points at @xmath68 .
since two active particles are eliminated on the event @xmath132 , we have @xmath311 using and as in the previous lemma gives @xmath312 this completes the proof .
the set of initial configurations is now @xmath314 note that the time @xmath118 of the first active arrow that either starts or points at @xmath68 is now exponentially distributed with rate two .
since two active particles are eliminated on @xmath277 , we have @xmath315 using the same reasoning as in but recalling that @xmath118 is now exponentially distributed with parameter two instead of four , we also have @xmath316 from which we deduce that @xmath313 this completes the proof . + + we refer the reader to figure [ fig : contribut ] for a plot of the expected value of @xmath317 and the lower bounds proved in lemmas [ lem : contribution-4][lem : contribution-6 ] with respect to the initial density @xmath21 . using again the same arguments as in the previous two sections
, we deduce that the one - dimensional constrained voter model with threshold one and four opinions fixates whenever @xmath318 which , recalling , using lemmas [ lem : contribution-4][lem : contribution-6 ] , and expanding and simplifying the expression above , gives the following sufficient condition for fixation : @xmath319 where @xmath320 to complete the proof of the theorem , we need the following technical lemma . to begin with
, we introduce the polynomials @xmath323 and observe that the derivative of @xmath321 can be written as @xmath324 now , it is clear that @xmath325 in other respects , we have @xmath326 showing that @xmath327 is decreasing in the interval @xmath322 . therefore , the polynomial @xmath328 is concave in this interval , from which it follows that @xmath329 combining , we deduce that @xmath321 is decreasing in @xmath322 . to prove that @xmath321 also has a unique root in this interval , we use its monotonicity and continuity , the fact that @xmath330 and the intermediate value theorem .
adamopoulos , a. , scarlatos , s. ( 2012 ) .
behavior of social dynamical models ii : clustering for some multitype particle systems with confidence threshold .
lecture notes in computer science , vol .
7495 , pp . 151160 , springer , heidelberg 2012 . | the constrained voter model describes the dynamics of opinions in a population of individuals located on a connected graph .
each agent is characterized by her opinion , where the set of opinions is represented by a finite sequence of consecutive integers , and each pair of neighbors , as defined by the edge set of the graph , interact at a constant rate .
the dynamics depends on two parameters : the number of opinions denoted by @xmath0 and a so - called confidence threshold denoted by @xmath1 .
if the opinion distance between two interacting agents exceeds the confidence threshold then nothing happens , otherwise one of the two agents mimics the other one just as in the classical voter model .
our main result shows that the one - dimensional system starting from any product measures with a positive density of each opinion fluctuates and clusters if and only if @xmath2 .
sufficient conditions for fixation in one dimension when the initial distribution is uniform and lower bounds for the probability of consensus for the process on finite connected graphs are also proved . |
third harmonic ( th ) generation and supercontinuum ( sc ) emission are two phenomena which have attracted broad interest in the past years @xcite .
an evident reason is their direct application in atmospheric remote sensing measurements based on lidar ( light detection and ranging ) femtosecond laser setups @xcite . in this context ,
spectral broadening originates from complex mechanisms that drive the long - range propagation of ultrashort pulses , when they form narrow filaments in optically - transparent media .
the physics of isolated femtosecond filaments in air is nowadays rather well understood ( see , e.g. , @xcite and references therein ) .
it involves the competition between kerr self - focusing and plasma defocusing , triggered whenever the input pulse power exceeds the critical power for self - focusing @xmath0 . here
, @xmath1 is the central laser wavelength , @xmath2 and @xmath3 are the linear and nonlinear refraction indices in air , respectively . for high enough powers , multiple filaments nucleated after an early stage of modulational instability
have also been widely investigated @xcite .
they produce spectral patterns mostly analogous to those generated by a single filament , as filamentary cells emerge in phase from the background field and possess the same phase link @xcite . by comparing terawatt ( tw ) multifilamented beams with gigawatt ( gw ) single filaments in air ,
this property was again verified in the uv - visible region ( 230 - 500 nm ) , where femtosecond self - focusing pulses centered at 800 nm generically produce a tremendous plateau of wavelengths @xcite .
this latter phenomenon has recently become a subject of inspiration for several researchers .
two scenarios have been proposed for justifying the build - up of new wavelengths in the uv - visible range .
on the one hand , temporal steepening phenomena undergone by the pump were shown to deeply modify the filament spectrum @xcite .
full chromatic dispersion included in the optical field wave number @xmath4 affects both the diffraction operator and the nonlinearities .
this induces shock - like dynamics at the back edge of the pulse through space - time focusing and self - steepening effects , which strongly blueshift the spectra . on the other hand , spectral broadening
becomes enhanced by harmonic generation .
the coupling of th with an infrared ( ir ) pump produces a two - colored
filament from pump intensities above 10 tw/@xmath5 @xcite .
the amount of pump energy transferred into th radiation depends on the linear wave vector mismatch parameter @xmath6^{-1}$ ] fixing the coherence length @xmath7 .
the smaller the coherence length , the weaker th fields . along
meter - range distances , the th component can stabilize the pump wave with about @xmath8 conversion efficiency @xcite .
experimental and numerical data reported ring structures embarking most of the th energy and having a half - divergence angle of about 0.5 mrad @xcite .
this process contributes to create a continuous spectral band of uv - visible wavelengths @xcite . resembling spectral dynamics
have also been reported from 1-mj infrared pulses propagating in argon at atmospheric pressure , after subsequent compression by chirped mirrors @xcite .
simulations of these experiments @xcite , discarding th emission , revealed that temporal gradients inherent to the steepening operators are sufficient to amplify uv shifts and cover the th bandwidth down to 250 and 210 nm for initial pulse durations of 10 and 6 fs , respectively .
very recently , numerical simulations @xcite refound this tendency for atmospheric propagation , i.e. , th generation , while it affects the pump dynamics to some extent over long ranges , does not change significantly sc spectra , whose variations are mostly induced by the fundamental field in air .
despite these last results , we are still missing a detailed understanding of the key parameters which are supposed to drive sc generation .
a first important parameter is , of course , the laser wavelength itself : how does the supercontinuum evolve when @xmath1 is varied ?
this question was addressed in ref .
@xcite for various laser wavelengths , at which some spectral components were seen to merge . however , the model used a two - envelope approximation ( for the pump and th fields , separately ) .
as emphasized in @xcite , splitting into th and sc pump within envelopes becomes problematic when their respective spectra overlap inside a wide frequency interval where the basic validity condition @xmath9 ( @xmath10 ) may no longer be fulfilled .
actually , th radiation produced through the nonlinear polarization needs to be described self - consistently from a single equation governing the total real optical field .
this model was missing in refs .
@xcite , which made the role of th overestimated in the white light emission .
another important parameter is the length of the self - guiding range : successive cycles of focusing and defocusing events promote the creation of shorter peaks in the pulse temporal profile and lead to a maximal extension of the spectrum .
a third potential player is the input pulse duration . in @xcite ,
this was shown to affect the spectra in noble gases for pulses containing a few optical cycles mainly . clearing this aspect requires several simulations using distinct pulse durations and exploiting different propagation ranges . in connection , we demonstrate that spectral enlargements are directly linked to the level of maximum intensity : steepening operators as well as th radiation broaden all the more the spectra as the intensity in the filament is high . the paper is organized as follows : sec .
[ model_equations ] presents the model equations , namely , a unidirectional propagation equation for the total electric field that generates higher - order harmonics ( mostly th ) through kerr nonlinearities .
results from this equation will be compared with those inferred from the standard nonlinear evolution equation ( nee ) for the pump wave .
the major difference between these two models lies in the production of the th field and its coupling with the pump wave .
[ long - range ] is devoted to the long - range propagation of 127-fs pulses in air described by the previous models .
emphasis is put on the influence of the central wavelength @xmath1 ( 248 , 800 , 1550 nm ) .
we discuss spectral modifications versus the height of @xmath11 , the input duration , together with the temporal steepening dynamics and merging between th and pump spectral bands .
[ short - range ] revisits sc for short - range ( focused ) propagations .
it is shown that @xmath11 becomes closer to analytical evaluations when the beam develops few focusing / defocusing cycles . in this configuration
, a lesser broadening may be achieved .
[ conclusion ] finally summarizes the generic features resulting from our analysis .
our unidirectional pulse propagation equation ( uppe ) assumes scalar and radially - symmetric approximations .
it also supposes negligible backscattering .
these hypotheses hold as long as the beam keeps transverse extensions larger than the central laser wavelength and as the nonlinear responses ( together with their longitudinal variations ) are small compared with the linear refraction index .
straightforward manipulations of maxwell equations allow us to establish the equation for the spectral amplitude of the optical electric field in the forward direction as @xcite @xmath12 where @xmath13 is the fourier transform of the forward electric field component , @xmath14 is the propagation variable , @xmath15 ( @xmath16 ) is the diffraction operator , @xmath17 , @xmath18 is the wavenumber of the optical field depending on the linear susceptibility tensor @xmath19 defined at frequency @xmath20 . in eq .
( [ uppe_1 ] ) , @xmath21 is the fourier transform of the nonlinearities that include the nonlinear optical polarization @xmath22 and the current density @xmath23 created by charged particles .
( [ uppe_1 ] ) restores the earlier uppe formulation proposed by kolesik _
et al . _
@xcite in the limit @xmath24 ( @xmath25 ) . for practical use , it is convenient to introduce the complex version of the electric field @xmath26 where @xmath27 employs the central wavenumber and frequency of the pump wave ( @xmath28 ) and @xmath29 denotes the heaviside function . because @xmath30 satisfies @xmath31 ( @xmath32 means complex conjugate ) , it is then sufficient to treat the uppe model ( [ uppe_1 ] ) in the frequency domain @xmath33 only .
the field intensity can be defined by @xmath34 averaged over an optical period at least .
expressed in w/@xmath5 , it is simply given by the classical relation @xmath35 .
concerning the nonlinearities , we assume a linearly polarized field .
we consider a cubic susceptibility tensor @xmath36 keeping a constant value around @xmath37 , so that @xmath22 contains the instantaneous cubic polarization expressed as @xmath38 .
in addition , the phenomenon of raman scattering comes into play when the laser field interacts with anisotropic molecules , in which vibrational and rotational states are excited .
depending on the transition frequency @xmath39 in three - level molecular systems and related dipole matrix element @xmath40 @xcite , the raman response takes the form @xmath41 where @xmath42 is the inverse of the fundamental rotational frequency and @xmath43 is the dipole dephasing time .
expressed in terms of the rescaled complex field @xmath30 [ eq . ( [ complex ] ) ] and with appropriate normalizations @xcite , eq .
( [ praman ] ) completes the cubic polarization as [ raman_123 ] @xmath44 where @xmath45 is the kerr nonlinear index .
expression ( [ raman_1 ] ) possesses both retarded and instantaneous components in the ratio @xmath46 .
the instantaneous part @xmath47 of eq .
( [ raman_3 ] ) describes the response from the bound electrons .
the retarded part @xmath48 accounts for the raman contribution , in which fast oscillations in @xmath34 give negligible contributions , as @xmath49 fs are assumed to exceed the optical period @xmath50 . when free electrons are created , they induce a current density @xmath51 , which depends on the electron charge @xmath52 , the electron density @xmath53 and the electron velocity @xmath54 .
@xmath23 is computed from fluid equations involving external plasma sources and the electron collision frequency @xmath55 . at moderate intensities ( @xmath56 w/@xmath5 )
, the current density obeys @xmath57 assuming electrons born at rest , the growth of the electron density is only governed by external source terms , i.e. , @xmath58 that include photo - ionization processes with rate @xmath59 and collisional ionization with cross - section @xmath60 here , @xmath61 and @xmath62 are the density of neutral species and the ionization potential , respectively .
electron recombination in gases is efficient over long ( ns ) time scales , and therefore we omit it . in eq .
( [ rho1 ] ) , the rate for photo - ionization @xmath59 follows from the perelomov , popov and terentev ( ppt ) s theory @xcite incorporating ammosov , delone and krainov ( adk ) coefficients @xcite ( see also ref .
optical field ionization theories stress two major limits bounded by the keldysh parameter , @xmath63 namely , the limit for multi - photon ionization ( mpi , @xmath64 ) concerned with rather low intensities and the tunnel limit ( @xmath65 ) concerned with high intensities , from which the coulomb barrier becomes low enough to let the electron tunnel out . here
, @xmath66 denotes the peak optical amplitude . for laser intensities @xmath67 w/@xmath5 , mpi characterized by the limit @xmath68 dominates , where @xmath69 is the number of photons necessary to liberate one electron .
the level of clamped intensity , @xmath11 , depends on the selected ionization rate .
energy lost by the pulse through single ionization processes is determined by a local version of the poynting theorem , yielding the loss current @xmath70 , such that @xmath71 . as a result
, our uppe model reads in fourier space as @xmath72 \widehat{\cal e } + \frac{i \mu_0 \omega^2}{2 k(\omega ) \sqrt{c_1 } } \theta(\omega ) \widehat{p}_{\rm nl } \\
& - \frac{i k_0 ^ 2 \theta(\omega)}{2 \epsilon(\omega_0 ) k(\omega)(1+\frac{\nu_e^2}{\omega^2 } ) } \left(\widehat{\frac{\rho \mathcal{e}}{\rho_c}}\right ) - \frac{\theta(\omega)}{2 } \sqrt{\frac{\epsilon(\omega_0)}{\epsilon(\omega ) } } { \cal l}(\omega ) , \end{split}\ ] ] where @xmath73 \mbox{e}^{i \omega t } dt.\ ] ] in eq .
( [ finaluppe1 ] ) , @xmath22 and the expression containing the electron density @xmath74 [ eq .
( [ rho1 ] ) ] must be transformed to fourier space , from which we retain only positive frequencies for the symmetry reasons given above . alternatively , when a central frequency @xmath37 is imposed , eq .
( [ uppe_1 ] ) restitutes the nonlinear envelope equation ( nee ) , earlier derived by brabec and krausz @xcite .
we can make use of the taylor expansion @xmath75 where @xmath76 , @xmath77 , @xmath78 , and take the inverse fourier transform of eq .
( [ uppe_1 ] ) in which terms with @xmath4 in their denominator are expanded up to first order in @xmath79 only . after introducing the complex - field representation @xmath80
, the new time variable @xmath81 can be utilized to replace the pulse into the frame moving with the group velocity @xmath82 .
furthermore assuming @xmath83 , @xmath84 and ignoring the th component , the nonlinear envelope equation for the forward pump envelope @xmath85 expands as follows : @xmath86 u \\ & \quad - i \frac{k_0}{2n_0 ^ 2 \rho_c } { t}^{-1 } \rho u - \frac{\sigma}{2 } \rho u - ( \rho_{\rm nt } - \rho)\frac{u_i w(i)}{2|u|^2 } u , \end{split } % \end{align}\ ] ] where @xmath87 , @xmath88 and @xmath89 .
the first term of the operator @xmath90 corresponds to group - velocity dispersion with coefficient @xmath91 .
equation ( [ 1 ] ) describes wave diffraction , kerr focusing response , plasma generation , chromatic dispersion with self - consistent deviations from the classical slowly - varying envelope approximation through the space - time focusing and self - steepening operators [ @xmath92 and @xmath93 , respectively ] .
this model will be integrated numerically by using initially gaussian pulses , @xmath94 which involves the input power @xmath95 , the beam waist @xmath96 and @xmath97 pulse half - width @xmath98 .
input pulses can be focused through a lens of focal length @xmath99 and they linearly diffract over the distance @xmath100 where @xmath101 is the rayleigh range of the collimated beam ( @xmath102 ) . in the coming analysis , we shall employ the nonlinear refractive indices @xmath103 , @xmath104 , and @xmath105 @xmath5/w for the wavelengths @xmath106 , 800 and 1550 nm , respectively . at 800 nm
, we consider a fitted mpi formulation for the ionization rate , @xmath107 , where @xmath108 and @xmath109 s@xmath110@xmath111/w@xmath112 . this approximation is known to reproduce experimental data at @xmath113 nm rather faithfully @xcite .
for the two other wavelengths , we lack well established formulations , so we employ ppt ionization rates .
we consider @xmath114 molecules having the lowest gap potential ( @xmath115 ev ) as the main specy undergoing ionization with an effective residual charge @xmath116 @xcite .
all ionization rates used in the present paper are illustrated in fig .
they yield saturation intensity below the threshold of 100 tw/@xmath5 currently claimed in the literature @xcite .
the dispersion relation for air has been parametrized as in ref .
@xcite .
ionization rates for the different wavelengths used throughout this paper : @xmath117 nm ( solid line ) , @xmath118 nm ( dotted line ) , and @xmath119 nm ( dashed line ) .
the dashed - dotted line shows the overestimated ionization rate chosen in sec .
[ imax ] for @xmath119 nm . ] since our outlook is to understand spectral variations versus the propagation dynamics at different wavelengths , we find it instructive to fix the same ratio of input power over critical , e. g. , @xmath120 , at all wavelengths . to locate the kerr - driven filamentation onset upon comparable @xmath14 scales
, we also adjust the ratio of the nonlinear focus @xmath121 and rayleigh range @xmath122 between 2 and 4 , by adapting suitably the beam waist @xmath96 between 1 and 4 mm . in air , dispersion is weak with @xmath123 fs@xmath124/cm in the uv as well as in the mid - ir domains .
self - channeling then mainly relies on the dynamical balance between kerr self - focusing and plasma defocusing , so that estimates for peak intensities ( @xmath11 ) , electron densities ( @xmath125 ) and filament radius ( @xmath126 ) can be deduced from equating diffraction , kerr and ionization responses in eq .
( [ 1 ] ) .
this yields the simple relations [ estimate13 ] @xmath127 where @xmath128 represents the maximal effective kerr index over the initial pulse profile .
for practical use , @xmath129 can be simplified to @xmath130 in mpi - like formulation .
the magnitude of @xmath11 directly impacts spectral broadening , which is basically driven by self - phase modulation ( spm ) .
because the frequency spectrum is expanded by the nonlinearity , spm leads to sc , as the wave intensity strongly increases through the self - focusing process .
noting by @xmath131 the phase of the field envelope , frequency variations are dictated in the limit @xmath132 by @xmath133 which varies with the superimposed actions of the kerr and plasma responses . near the focus point @xmath121 , only the front edge of the pulse survives from this interplay and a redshift is enhanced by plasma generation . at later distances , second focusing / defocusing sequences attenuate this first tendency . in contrast , when accounting for temporal steepening ( @xmath134 ) , shock edges in the back of the pulse are created and a
blue shoulder appears in the spectrum , to the detriment of the early redshift @xcite . in addition
, the cubic polarization generates third - order harmonics , modeled by the last term of eq .
( [ raman_1 ] ) . in self - focusing regimes , the third - harmonic intensity usually contributes by a little percentage to the overall beam fluence @xcite . despite its smallness
, this component may act as a saturable nonlinearity for the carrier wave .
it lowers the peak intensity of the pump and contributes to enhance the blue side of the spectrum after the th and pump bandwidths increase and overlap @xcite .
we numerically analyze supercontinuum generation for the three laser wavelengths of 248 nm , 800 nm and 1550 nm .
results of the nonlinear schrdinger - like equation ( [ 1 ] ) for the pump envelope involving or not space - time focusing and self - steepening operators are compared with those of the unidirectional propagation equation ( [ finaluppe1 ] ) avoiding any taylor expansion in the dispersion relation .
special attention is here given to the long - range propagation , which rather favors several cycles of focusing - defocusing events . before proceeding with the above parameters specifically , we perform three different series of simulations showing sc at 800 nm , whose
results are summarized in fig .
the first one concerns direct integrations of eq .
( [ finaluppe1 ] ) ; the second refers to the same pulse described by eq .
( [ 1 ] ) , which eludes th production ; the third approach relies on eq .
( [ 1 ] ) , in which temporal steepening is omitted , i.e. , @xmath135 .
peak intensities ( solid curves , left - hand side scales ) and peak electron densities ( dashed curves , right - hand side scales ) of 127-fs pulses with ratio of input power over critical equal to 4 and waist @xmath136 at @xmath117 nm . the insets show on - axis spectra : dotted curves represent the initial spectra ; the solid curves the spectra at the propagation distance @xmath137 m where maximal broadening is observed : ( a ) uppe model eq .
( [ finaluppe1 ] ) ; ( b ) nee eq .
( [ 1 ] ) applied to the pump wave ; ( c ) nee eq . ( [ 1 ] ) modified with setting @xmath138.,title="fig : " ] peak intensities ( solid curves , left - hand side scales ) and peak electron densities ( dashed curves , right - hand side scales ) of 127-fs pulses with ratio of input power over critical equal to 4 and waist @xmath136 at @xmath117 nm . the insets show on - axis spectra : dotted curves represent the initial spectra ; the solid curves the spectra at the propagation distance @xmath137 m where maximal broadening is observed : ( a ) uppe model eq .
( [ finaluppe1 ] ) ; ( b ) nee eq .
( [ 1 ] ) applied to the pump wave ; ( c ) nee eq . ( [ 1 ] ) modified with setting @xmath138.,title="fig : " ] peak intensities ( solid curves , left - hand side scales ) and peak electron densities ( dashed curves , right - hand side scales ) of 127-fs pulses with ratio of input power over critical equal to 4 and waist @xmath136 at @xmath117 nm . the insets show on - axis spectra : dotted curves represent the initial spectra ; the solid curves the spectra at the propagation distance @xmath137 m where maximal broadening is observed : ( a ) uppe model eq .
( [ finaluppe1 ] ) ; ( b ) nee eq .
( [ 1 ] ) applied to the pump wave ; ( c ) nee eq . (
[ 1 ] ) modified with setting @xmath138.,title="fig : " ] the insets in fig .
[ fig1 ] show the on - axis spectra when sc is maximal .
following the uppe description , th , which emerges from @xmath139 m , develops a limited redshift , whereas sc of the fundamental is widely extending towards the blue / uv wavelengths [ fig .
[ fig1](a ) ] . note that , although a broad plateau occurs in this domain , the th bandwidth still appears separated from the pump spectrum . following the nee description , there is no th generation .
however , sc is so amplified in the blue region by temporal steepening effects , that it overlaps the th zone and simply hides it [ fig . [ fig1](b ) ] .
finally , when neglecting temporal steepening , the pump instead develops a wide redshift ( overestimated by plasma coupling ) and a much narrower blueshift [ fig .
[ fig1](c ) ] .
a first observation can be drawn from fig .
[ fig1 ] : since th is responsible for lowering the saturation intensity of the pump @xcite , @xmath11 reached in the uppe model is lower and the frequency variations @xmath140 are diminished compared with nee spectra for the pump wave alone .
apart from this difference , no significant other change was detected between both these models , so that nee seems to be nothing else but the uppe description subtracted by the self - generated harmonics @xcite .
importantly , omitting temporal derivatives of the operators @xmath141 imply more serious discrepancies , as can be seen from fig .
[ fig1 ] .
we now examine gaussian pulses at different wavelengths ( 248 , 800 and 1550 nm ) with @xmath142 fs and @xmath143 as initial conditions for the uppe model in parallel geometry @xmath144 .
figures [ fig1](a ) and [ fig2 ] show snapshots of spectra at maximal extent , together with associated peak intensities and electron densities .
we here specify that no th generation was included for @xmath106 nm , because no reliable data of the dispersion relation was available for this wavelength .
we believe , instead , that spectral components below 90 nm should be rapidly absorbed by the medium .
peak intensities ( solid curves , left - hand side scales ) and peak electron densities ( dashed curves , right - hand side scales ) of 127-fs pulses with ratio of input power over critical equal to 4 at different wavelengths @xmath1 .
the insets show on - axis spectra : dotted curves represent the initial spectra ; the solid curves the spectra at the propagation distance @xmath145 where maximal broadening is observed : ( a ) @xmath118 nm , @xmath146 mm , @xmath147 m ; ( b ) @xmath119 nm , @xmath148 mm , @xmath149 m.,title="fig : " ] peak intensities ( solid curves , left - hand side scales ) and peak electron densities ( dashed curves , right - hand side scales ) of 127-fs pulses with ratio of input power over critical equal to 4 at different wavelengths @xmath1 .
the insets show on - axis spectra : dotted curves represent the initial spectra ; the solid curves the spectra at the propagation distance @xmath145 where maximal broadening is observed : ( a ) @xmath118 nm , @xmath146 mm , @xmath147 m ; ( b ) @xmath119 nm , @xmath148 mm , @xmath149 m.,title="fig : " ] temporal evolutions of the pulses shown in figs .
[ fig1](a ) and [ fig2 ] in the @xmath150 plane : ( a ) @xmath118 nm ; ( b ) @xmath117 nm ; ( c ) @xmath119 nm.,title="fig : " ] temporal evolutions of the pulses shown in figs .
[ fig1](a ) and [ fig2 ] in the @xmath150 plane : ( a ) @xmath118 nm ; ( b ) @xmath117 nm ; ( c ) @xmath119 nm.,title="fig : " ] temporal evolutions of the pulses shown in figs .
[ fig1](a ) and [ fig2 ] in the @xmath150 plane : ( a ) @xmath118 nm ; ( b ) @xmath117 nm ; ( c ) @xmath119 nm.,title="fig : " ] by comparing figs . [ fig1](a ) and [ fig2 ] , it is seen right away that supercontinuum generation increases with the wavelength . to quantify this observation
we introduce @xmath151 as the total extension of the on - axis spectra over wavelengths at @xmath152 times the maximal spectral intensity .
then a measurement for the effective broadening is the ratio @xmath153 , which we find close to @xmath154 at 248 nm , @xmath155 at 800 nm and @xmath156 at 1550 nm .
a look at the propagation dynamics reveals that the self - guiding range is noticeably augmented at longer wavelengths .
this can be explained by the transverse size of the filament .
equation ( [ estimate3 ] ) gives an estimate for the beam waist in filamentation regime .
if we assume comparable @xmath11 for all wavelengths , we deduce that the filament diameter at 1550 nm is about one order of magnitude larger than that at 248 nm , which is compatible with our numerical data .
hence , the larger the wavelength , the slower the filament is expected to diffract . moreover , by virtue of the formula for the critical power @xmath157 and since a filament conveys a few @xmath158 @xcite , it is obvious that ir filaments contain much more energy than their uv counterparts .
thus , nonlinear losses along the filamentation range are less dramatic in the ir domain . due to the longer propagation range , more focusing / defocusing events participate in enlarging the spectra at longer wavelengths .
figure 4 details the evolution of the filaments in the plane @xmath150 .
it is seen that the time window in which the pulses disperse occupies the length of the input pulse duration .
although shorter temporal peaks arise through self - focusing / defocusing events , multi - peaked profiles mostly develop patterns having a whole extent close to @xmath98 . before going on
, we find it worth investigating sc at 1550 nm more thoroughly , in relationship with third - harmonic generation .
figure [ fig3 ] plots three of the sc development stages , first when th and pump components are clearly separated ( @xmath159 m ) , second when they start to merge ( @xmath160 m ) . at the last propagation distance ( @xmath161 m ) , we can observe that , unlike fig .
[ fig1 ] , th and pump spectra overlap and make the th bandwidth not distinguishable . for comparison ,
results from the nee model for the pump wave alone have also been plotted .
on - axis spectra for the pulse used in fig .
[ fig2](b ) ( @xmath162 nm ) for the propagation distances ( a ) @xmath163 m ( solid line ) and @xmath164 m ( dash - dotted line ) and ( b ) @xmath165 m ( solid line ) .
the dashed line in ( b ) refers to a spectrum at @xmath165 m computed from the nee model.,title="fig : " ] on - axis spectra for the pulse used in fig .
[ fig2](b ) ( @xmath162 nm ) for the propagation distances ( a ) @xmath163 m ( solid line ) and @xmath164 m ( dash - dotted line ) and ( b ) @xmath165 m ( solid line ) .
the dashed line in ( b ) refers to a spectrum at @xmath165 m computed from the nee model.,title="fig : " ] the impact of the saturation intensity onto sc is investigated by simply changing the ionization model : by decreasing the photo - ionization rate artificially it is possible to increase @xmath11 and the maximal plasma level @xmath125 accordingly .
reversely , increasing @xmath59 reduces these two quantities , which can have a direct influence on the spectral broadening , as inferred from eq .
( [ spmchin ] ) . to study this point
, we concentrate on the wavelength of 1550 nm only , because it yields the broadest spectra explored so far .
figure [ fig4 ] shows the maximal intensity and peak electron density at this wavelength , when the ionization rate is artificially increased ( see fig .
[ fig0 ] ) .
all other parameters are unchanged , compared with the simulation shown in fig .
[ fig2](b ) . with the original ionization rate
, @xmath11 reaches the value of 80 tw/@xmath5 ; with the artificial one , @xmath11 stays below 13 tw/@xmath5 .
the inset in fig .
[ fig4 ] details the corresponding spectrum at @xmath166 m. with low @xmath11 , the th component is reduced to some extent as the pump intensity barely exceeds the th conversion threshold @xcite .
meanwhile , sc of the pump driven by the @xmath141 operators in the blue side decreases , i.e. , a lower @xmath11 for analogous pulse compression implies less sharp temporal gradients and smoother optical shocks , which weakens blueshifted frequency variations .
these features are visible in fig .
[ fig4 ] , where th and pump broadbands remain separated . in figs .
[ fig2](b ) and [ fig3](b ) , in contrast , sc extends beyond the th wavelength and increases more the pulse spectrum .
similar features were observed at the two other wavelengths , when the ionization rate was changed .
peak intensity ( solid curve , left - hand side scale ) and peak electron density ( dashed curve , right - hand side scale ) for the same pulse as in fig .
[ fig2](b ) ( @xmath162 nm ) computed from the uppe model with an overestimated ionization rate ( see fig . [ fig0 ] ) .
the inset shows maximal spectral broadening attained at @xmath166 m. ] we investigate the influence of the initial pulse duration on the propagation dynamics and sc generation .
since we consider transform - limited pulses , the value of @xmath98 is directly linked to the initial spectral width .
moreover , @xmath11 comes into play in sc generation and is expected to scale as @xmath167 [ see eq .
( [ estimate1 ] ) ] .
thus , the initial pulse duration should play a significant role in spectral broadening . to check this assessment
, we performed several simulations using the uppe model , by varying @xmath98 from 20 fs up to 500 fs . because group - velocity dispersion becomes very efficient at short pulse durations and
may even stop the kerr self - focusing at powers too close to critical @xcite , we increased the input power up to 20 @xmath158 for @xmath168 fs . with this
, we ensure to trigger a filamentation regime even for this short input duration . to illustrate the dependency of @xmath11 upon @xmath98 , short wavelengths are preferable because the number of photons for ionization is small .
figure [ fig6 ] shows maximum intensity , peak electron density and maximal spectral extent of a 20-fs pulse at 248 nm . at this wavelength , @xmath169 and , following eq .
( [ estimate1 ] ) , @xmath11 and @xmath125 should increase by a factor @xmath170 compared to the 127-pulse shown in fig . [
fig2](a ) .
indeed , both quantities are increased by a factor of @xmath171 in the simulation .
spectral broadening is augmented from 0.5 to 0.8 in terms of @xmath172 , especially to the blue side . as explained in sec .
[ imax ] , this results from the action of the steepening operators .
the overall propagation dynamics , characterized by the filamentation length and number of focusing / defocusing cycles are , however , comparable for both the 20-fs and the 127-fs pulses .
peak intensity ( solid curve , left - hand side scale ) and peak electron density ( dashed curve , right - hand side scale ) of a 20-fs pulse with ratio of input power over critical equal to 20 , @xmath118 nm , @xmath146 mm .
the inset shows on - axis spectra : the dotted curve represents the initial spectrum ; the solid curve the spectrum at the distance @xmath173 m where maximal broadening is observed . ] on the other hand , if we increase the initial duration @xmath98 towards the ps time scale , the propagation dynamics changes drastically . as an example
, figure [ fig5bis2 ] shows the temporal evolution of a 500-fs pulse at 800 nm .
compared with fig .
[ fig2bis](b ) employing @xmath142 fs , the obvious difference is the huge number of focusing / defocusing cycles . the action of the generated plasma breaks the pulse profile into a larger number of shorter peaks . with a longer pulse duration ,
time slices are available for feeding successive focusing events .
the filamentation range is increased and @xmath174 is maintained over several meters .
inspection of the simulations , however , reveals maximal spectral extent comparable with that dispayed in fig . [ fig1](a ) .
temporal evolution of a 500-fs pulse at 800 nm with 4 @xmath158 and waist @xmath136 in the @xmath150 plane . ] at 1550 nm the influence of @xmath98 on the maximal intensity is much less pronounced , since we have @xmath175 [ see figs .
[ fig2](b ) and [ fig5 ] ] . for all pulse durations , we indeed observe @xmath176 tw/@xmath5 .
this can explain why the maximal spectral extension @xmath172 is always found between 1.5 and 2 , regardless @xmath98 may be .
the major difference lies in the filamentation range , which increases with the initial pulse duration .
so , there is no significant change in sc generation between short and long pulses over large enough propagation scales .
peak intensities ( solid curves , left - hand side scales ) and peak electron densities ( dashed curves , right - hand side scales ) of 4 mm waisted pulses with different durations and ratios @xmath177 at @xmath119 nm .
the insets show on - axis spectra : dotted curves represent the initial spectra ; the solid curves the spectra at the propagation distance @xmath145 where maximal broadening is observed : ( a ) @xmath178 fs , @xmath179 , @xmath180 m ; ( b ) @xmath181 fs , @xmath182 , @xmath166 m.,title="fig : " ] peak intensities ( solid curves , left - hand side scales ) and peak electron densities ( dashed curves , right - hand side scales ) of 4 mm waisted pulses with different durations and ratios @xmath177 at @xmath119 nm .
the insets show on - axis spectra : dotted curves represent the initial spectra ; the solid curves the spectra at the propagation distance @xmath145 where maximal broadening is observed : ( a ) @xmath178 fs , @xmath179 , @xmath180 m ; ( b ) @xmath181 fs , @xmath182 , @xmath166 m.,title="fig : " ] this last observation invites us to look at the temporal profiles upon propagation .
if the spectral extent is comparable , we might also find similar temporal patterns . indeed , the on - axis temporal profiles shown in fig .
[ fig5bis ] all exhibit structures with duration of 10 - 15 fs .
it seems that the initial pulse length @xmath98 just determines how many of these peaks appear , or , in other words , how many focusing / defocusing cycles the pulse is able to develop upon propagation .
another indication for the change in the effective pulse duration upon propagation is provided by the curve of the intensity maximum in fig
. [ fig5](b ) : the first focusing cycle is halted at slightly lower intensities @xmath183 60 tw/@xmath5 , because in this early stage the peak duration remains of the order of @xmath181 fs . at later stages
, @xmath11 increases as the pulse undergoes temporal compression . on - axis
temporal profiles of pulses with different duration at @xmath145 : @xmath178 fs [ solid line , parameters used in fig .
[ fig5](a ) ] ; @xmath184 fs [ dashed line , parameters used in fig .
[ fig2](b ) ] ; @xmath181 fs [ dotted line , parameters used in fig .
[ fig5](b ) ] . ]
so far , we have analyzed free propagation dynamics where long filaments achieve temporal gradients and sc extents similar to those produced by initially much shorter pulses .
now , we force all pulses to cover the same short filamentation range through a focusing optics ( @xmath185 m ) .
results are shown in fig .
[ fig7 ] for @xmath168 and 500 fs at 1550 nm .
we can check that the maximum intensity @xmath11 follows the theoretical expectations ( [ estimate13 ] ) involving @xmath98 .
the short filamentation range prevents the occurrence of several focusing / defocusing cycles , especially for the 500-fs pulse .
hence , although the pulse self - focuses , its temporal extent remains comparable with @xmath98 , as we can see in figs .
[ fig7](c ) and ( d ) .
the 20-fs pulse shows significant spectral broadening with a visible , broadened th peak .
the 500-fs pulse is spectrally too narrow to generate supercontinuum over roughly 1-m filamentation range .
therefore , the fundamental and harmonic peaks clearly stay separated ( note the occurrence of the fifth harmonics at 310 nm , which is self - consistently described by the uppe model )
. thus , the initial pulse duration strongly influences spectral broadening in configurations of short filamentation range mainly , which is consistent with the numerical results of ref .
@xcite .
peak intensities ( solid curves , left - hand side scales ) and peak electron densities ( dashed curves , right - hand side scales ) of 4 mm waisted pulses with different duration and ratio of input power over critical equal to 4 at @xmath119 nm propagating in focused geometry ( @xmath186 m ) .
the insets show on - axis spectra : dotted curves represent the initial spectra ; the solid curves the spectra at the distance @xmath187 m where maximal broadening is observed : ( a ) @xmath178 fs ; ( b ) @xmath181 fs .
( c ) and ( d ) show the respective temporal evolution for the pulses of ( a ) and ( b ) in the @xmath150 plane.,title="fig : " ] peak intensities ( solid curves , left - hand side scales ) and peak electron densities ( dashed curves , right - hand side scales ) of 4 mm waisted pulses with different duration and ratio of input power over critical equal to 4 at @xmath119 nm propagating in focused geometry ( @xmath186 m ) .
the insets show on - axis spectra : dotted curves represent the initial spectra ; the solid curves the spectra at the distance @xmath187 m where maximal broadening is observed : ( a ) @xmath178 fs ; ( b ) @xmath181 fs .
( c ) and ( d ) show the respective temporal evolution for the pulses of ( a ) and ( b ) in the @xmath150 plane.,title="fig : " ] peak intensities ( solid curves , left - hand side scales ) and peak electron densities ( dashed curves , right - hand side scales ) of 4 mm waisted pulses with different duration and ratio of input power over critical equal to 4 at @xmath119 nm propagating in focused geometry ( @xmath186 m ) .
the insets show on - axis spectra : dotted curves represent the initial spectra ; the solid curves the spectra at the distance @xmath187 m where maximal broadening is observed : ( a ) @xmath178 fs ; ( b ) @xmath181 fs .
( c ) and ( d ) show the respective temporal evolution for the pulses of ( a ) and ( b ) in the @xmath150 plane.,title="fig : " ]
in summary , we have revisited recent works on sc generation versus third harmonic emission , by showing from a complete uppe model that spectral enlargements of femtosecond pulses in self - guiding regime are mostly driven by space - time focusing and self - steepening .
th generation , although changing the pump dynamics , affects the spectra to a limited extent only .
this conclusion corrects some previous statements @xcite , based on a propagation model in which temporal steepening terms were analyzed separately from an envelope description for th generation .
going one step beyond , we have demonstrated the important role of the saturation intensity in the frequency variations enlarging both th and pump broadbands .
the input pulse duration becomes a significant player in the spectral extents as long as pulses do not propagate too far , i.e. , they do not let the temporal profiles of the pulse fluctuate so much that many peaks and sharp gradients can develop along the optical path .
this property is mainly enlightened in short - range focused geometry and lost in long - range parallel geometry .
finally , the role of the central wavelength is preeminent : our numerical simulations displayed evidence that sc clearly augments with the laser wavelength .
s. l. chin , a. brodeur , s. petit , o. g. kosareva , v. p. kandidov , filamentation and supercontinuum generation during the propagation of powerful ultrashort laser pulses in optical media ( white light laser ) , j. nonlinear opt .
( 1999 ) 121 .
a. b. fedotov , n. i. koroteev , m. m. t. loy , x. xiao , a. m. zheltikov , saturation of third - harmonic generation in a plasma of self - induced optical breakdown due to the self - action of 80-fs light pulses , opt .
commun . 133
( 1997 ) 587 .
j. kasparian , m. rodriguez , g. mjean , j. yu , e. salmon , h. wille , r. bourayou , s. frey , y. b. andr , a. mysyrowicz , r. sauerbrey , j. p. wolf , l. wste , white - light filaments for atmospheric analysis , science 301 ( 2003 ) 61 .
l. berg , s. skupin , r. nuter , j. kasparian , j. p. wolf , optical ultrashort filaments in weakly - ionized , optically - transparent media , http://arxiv.org/abs/physics/0612063 , submitted to rev .
( 2006 ) .
s. l. chin , s. petit , w. liu , a. iwasaki , m .- c .
nadeau , v. p. kandidov , o. g. kosareva , k. y. andrianov , interference of transverse rings in multifilamentation of powerful femtosecond laser pulses in air , opt .
commun . 210
( 2002 ) 329 .
n. akzbek , a. becker , m. scalora , s. l. chin , c. m. bowden , continuum generation of the third - harmonic pulse generated by an intense femtosecond ir laser pulse in air , appl .
b : lasers & optics 77 ( 2003 ) 177 .
l. berg , s. skupin , g. mjean , j. kasparian , j. yu , s. frey , e. salmon , j. p. wolf , supercontinuum emission and enhanced self - guiding of infrared femtosecond filaments sustained by third - harmonic generation in air , phys . rev .
e 71 ( 2005 ) 016602 .
h. yang , j. zhang , j. zhang , l. z. zhao , y. j. li , h. teng , y. t. li , z. h. wang , z. l. chen , z. y. wei , j. x. ma , w. yu , z. m. sheng , third - order harmonic generation by self - guided femtosecond pulses in air , phys .
e 67 ( 2003 ) 015401(r ) .
n. akzbek , a. iwasaki , a. becker , m. scalora , s. l. chin , c. m. bowden , third - harmonic generation and self - channeling in air using high - power femtosecond laser pulses , phys .
( 2002 ) 143901 .
i. alexeev , a. c. ting , d. f. gordon , e. briscoe , b. hafizi , p. sprangle , characterization of the third - harmonic radiation generated by intense laser self - formed filaments in air , opt .
( 2005 ) 1503 .
g. mjean , j. kasparian , j. yu , s. frey , e. salmon , r. ackermann , j .-
wolf , l. berg , s. skupin , uv - supercontinuum generated by femtosecond pulse filamentation in air : meter - range experiments versus numerical simulations , appl .
b : lasers & optics 82 ( 2006 ) 341 .
s. a. trushin , s. panja , k. kosma , w. e. schmid , w. fuss , supercontinuum extending from @xmath188 to 250 nm , generated by focusing ten - fs laser pulses at 805 nm into ar , appl .
b : lasers & optics 80 ( 2005 ) 399 .
n. akzbek , s. a. trushin , a. baltuka , w. fuss , e. goulielmakis , k. kosma , f. krausz , s. panja , m. uiberacker , w. e. schmid , a. becker , m. scalora , m. bloemer , extending the supercontinuum spectrum down to 200 nm with few - cycle pulses , new j. phys . 8
( 2006 ) 177 .
s. skupin , l. berg , u. peschel , f. lederer , g. mjean , j. yu , j. kasparian , e. salmon , j .-
wolf , m. rodriguez , l. wste , r. bourayou , r. sauerbrey , filamentation of femtosecond light pulses in the air : turbulent cells versus long - range clusters , phys .
e 70 ( 2004 ) 046602 . | supercontinuum generation by femtosecond filaments in air is investigated for different laser wavelengths ranging from ultraviolet to infrared .
particular attention is paid on the role of third - harmonic generation and temporal steepening effects , which enlarge the blue part of the spectrum . a unidirectional pulse propagation model and nonlinear evolution equations are numerically integrated and their results are compared .
apart from the choice of the central wavelength , we emphasize the importance of the saturation intensity reached by self - guided pulses , together with their temporal duration and propagation length as key players acting on both supercontinuum generation of the pump wave and emergence of the third harmonics .
maximal broadening is observed for large wavelengths and long filamentation ranges . , supercontinuum generation , femtosecond filaments , nonlinear schrdinger equation , self - phase modulation 42.65.tg , 52.38.hb , 42.65.jx , 42.68.ay |
dynamical chaos arising from the overlap of orbital resonances is responsible for orbital instabilities in the solar system @xcite .
for the simplest case of the planar circular restricted three body problem , the overlap of first order , @xmath14 , mean motion resonances occurs for @xmath15 , where @xmath16 is an integer , @xmath17 and @xmath18 is the planet - star mass ratio @xcite .
this resonance overlap condition defines an approximately annular `` chaotic zone '' in the vicinity of a planet s circular orbit in which initially circular test particle orbits are rendered strongly chaotic . making use of kepler s third law with eq .
[ eq : pro ] , the half - width of this zone on either side of the planet s orbit is given by @xmath19 where @xmath13 is the semi - major axis of the planet , and @xmath20 is a numerical coefficient .
an alternative analytical derivation by @xcite found @xmath21 . a numerical analysis by @xcite determined @xmath22 for planet mass in the range @xmath23 .
although eq .
[ eq : rowidth ] was derived for a planet on a circular orbit , the relative insensitivity of @xmath24 to planet eccentricities up to about 0.3 @xcite makes eq .
[ eq : rowidth ] a powerful tool for planetary dynamics .
it is the first step in understanding the separations of long term stable planetary orbits in the solar system and in exo - solar systems . in our solar system , the outer edge of the asteroid belt and the inner edge of the kuiper belt are approximately coincident with this estimate of the inner and outer boundaries of the chaotic zones of jupiter and neptune , respectively . for extra - solar planetary systems , this equation has been employed to constrain the locations of planetesimal belts in some systems @xcite , and in estimating the masses of unseen planets that may be responsible for the observed gaps and edges of debris disks @xcite .
@xcite determined that eq .
[ eq : rowidth ] , with a coefficient of @xmath25 , describes the cleared region of the fomalhaut disk perturbed by an eccentric planet . @xcite and
@xcite have studied the dependence of the chaotic zone on the test particles eccentricity , finding that non - zero initial eccentricities lead to wider chaotic zone widths . in this paper
, we answer the following questions : where do chaotic zone particles go , and how long do they take to get there ? a few previous studies have touched on these questions ( e.g. @xcite ) , although none have attempted a systematic study . to this end
, we investigate the orbital evolution of particles in initially circular orbits in a planet s chaotic zone in the framework of the planar circular restricted three - body problem , for planet - star mass ratios in the range @xmath26 ( corresponding to pluto - mass objects to brown dwarf mass objects orbiting solar mass stars ) .
there are four possible final states of such particles : collision with the star , collision with the planet , escape to infinity , or bounded but non - collision orbits for infinite time . using numerical integrations
, we determine particle loss timescales from the chaotic zone and branching ratios for the particle final states as a function of planet mass and planet size .
( as a practical matter , we determined the final states of particles at a finite but long time . )
although we refer to the massive bodies as `` star '' and `` planet '' , our results are applicable to other restricted three - body astronomical contexts with small mass ratios , such as binary minor planets , star / brown - dwarf , or black - hole / star systems .
we describe our methodology in section 2 , and present our results in section 3 .
we summarize and discuss an application of our results in section 4 .
the test particle orbits of interest here are strongly chaotic and must be followed accurately through close planetary encounters . for the numerical integration of the equations of motion
, we used the bulirsch - stoer method @xcite in the swift integration package .
this method utilizes a modified midpoint method for computing integration steps and an adaptive step size for controlling numerical error .
we adopted a fractional truncation error tolerance of @xmath27 . throughout the integrations
, we confirmed that there was no secular drift of the planet s semi - major axis within 1 part in @xmath28 . within the framework of
the planar circular restricted three - body problem , we adopt units such that the total mass ( star and planet ) is unity , the planet star distance ( @xmath13 ) is unity , and the orbital period of the planet about the star is @xmath29 .
we denote by @xmath0 the ratio of the planet mass to that of the host star .
we carried out simulations for twelve values of @xmath30 , @xmath31 for each of these cases , we integrated 7840 test particles in initially circular orbits ( defined as osculating circular orbits about the star ) in an annulus about the planet s orbit . for values of @xmath32 ,
the initial test particle orbits were uniformly spaced within each of 98 annuli in the radial range @xmath33 $ ] ; this is approximately 3050% larger than the resonance overlap zone defined by eq .
[ eq : rowidth ] . in the cases of planet masses
@xmath34 , the half - width was extended out to @xmath35 in order to more effectively determine the extent of clearing .
the initial longitude of the planet was zero and the initial angular positions of the test particles were chosen uniformly over the full range , @xmath36 , with the exception that we excluded initial positions within @xmath37 distance of the planet , where @xmath38 is the planet s hill radius ; the excluded region is approximately the region of stable satellite orbits in the restricted three body problem .
we integrated the test particle orbits for a time period of up to @xmath39 revolutions of the planet to determine their final destinations and removal timescales .
for most simulation results discussed here , integrations of @xmath40 revolutions of the planet were sufficient .
we stopped integrating a test particle under the following conditions : 1 .
_ particle s periastron was within @xmath41_. for @xmath42 au , this distance corresponds to approximately two solar radii .
we adopted this distance as our definition for a collision with the star .
particle s distance from the star exceeded @xmath43_. we refer to this condition as ` scattered ' .
several test runs indicated that many of these particles would eventually travel to distances beyond 100@xmath13 .
this choice is additionally motivated by the observation that the planets in the solar system have separations of @xmath44 ; multiple planet systems amongst the known exo - planets also typically have smaller separations @xcite .
particle approached the planet within a prescribed small distance , @xmath45_. we refer to this condition as ` collision with the planet ' .
we carried out three sets of simulations , one with @xmath46 , the second with @xmath47 and the third with @xmath48 . + our choice of scaling the collision radius to the planet s hill radius
is motivated by the expectation that the ratio of collisions to scatterings is governed by the square of the ratio of the surface escape velocity of the planet to the mean relative velocity of encounters , the so - called safronov number , @xmath49 . for initially circular test particle orbits in the vicinity of the planet s orbit
, we expect that the typical relative velocity of encounter is on the order of the `` hill velocity '' , @xmath50 , where @xmath51 is the planet s circular orbital velocity . in this case
, @xmath52 is inversely proportional to the ratio of the physical radius to the hill radius of the planet , @xmath53 the physical radius of a planet as a fraction of its hill radius can be expressed as , @xmath54 where @xmath55 are the mean density of the planet and star , respectively , @xmath56 is the mean density of earth , @xmath57 is the stellar radius , and @xmath58 and @xmath59 are the mean solar density and radius .
this fraction is only weakly dependent on planet density , and is mainly dependent on planet - star separation .
the currently known planets in the galaxy have @xmath6 concentrated in the range 0.010.1 , as shown in figure [ fig : rph ] , since the majority of planets are discovered at small orbital radii by means of radial velocity and transit observations .
the solar system planets have smaller values ( also indicated in figure [ fig : rph ] ) , as do directly imaged planets , due to their greater distances from their host star .
particles that did not meet the above conditions over the duration of the integrations are referred to as ` survivors ' . , the ratio of the physical radius of a planet as a fraction of its hill radius .
the histograms represent the distribution of _ kepler _ planet candidates ( radii from nasa exoplanet archive , retrieved on 06 march 2014 ) for different assumptions of bulk density .
the black dots indicate the solar system planets . ]
we note that there is one known conserved quantity , the jacobi integral @xcite , that constrains the evolution of test particles in the circular restricted three body problem . for the planar case , the jacobi integral
is given by @xmath60 where @xmath61 , and @xmath62 and @xmath63 are the distances of the test particle from the star and the planet , respectively , and @xmath64 and @xmath65 are the particle s position and velocity vector in an inertial reference frame with origin at the barycenter of the star and the planet . a particle s range in the @xmath66@xmath67 plane is bounded to regions where @xmath68 , therefore eq .
[ eq : jacobi ] ( with @xmath69 , defines the boundary of this region , the so - called zero - velocity curve .
it is pertinent to note that the zero - velocity curves for most of our initial conditions of particles within the chaotic zone allow all possible end - states , including collision with the star , collision with the planet , scatter to infinity , or possibly long - lived bounded orbits that may be either chaotic or quasi - periodic ; i.e. , the value of @xmath70 alone is generally not indicative of the final destinations of our chaotic zone particles . a small minority of initial conditions , near the inner and outer edges of the chaotic zone annulus , have values of @xmath70 that are bounded to remain entirely interior or entirely exterior to the planet s orbit .
in addition to meeting the truncation error tolerance in the numerical integrations , we monitored the value of the jacobi integral , @xmath70 , for each test particle . for the range of initial conditions of interest , @xmath70 is close to 3 , and @xmath71 is small , on the order of @xmath72 .
we found that most of the chaotic zone test particles preserved their values of @xmath70 to better than a part in @xmath73 , but high numerical accuracy of the preservation of the jacobi integral was sometimes more difficult to achieve than meeting the local truncation error tolerance .
unsurprisingly , the cases of small values of @xmath6 pose a particularly stiff computational challenge .
we identified numerical solutions as inaccurate if , over the course of the integrations , the fractional variation of @xmath70 exceeded @xmath74 .
simulations with higher planet masses and lower values of @xmath6 had larger numbers of these cases .
for all cases of @xmath75 and @xmath76 that we simulated , we found that no more than @xmath77 of our test particles failed this jacobi conservation criterion ; the case of @xmath78 for @xmath48 had the highest fraction , 27% , of the test particles failing our jacobi conservation criterion .
we found no obvious pattern in the initial conditions of these particles .
these cases were excluded from the analysis that we report in the following sections .
discarding these cases has little effect on the reported results .
, for fixed @xmath6 . ] , for the case of @xmath47 . for clarity , we scale the initial semimajor axis separation from the planet , @xmath79 , to the half - width of the chaotic zone , defined as @xmath80 .
white patches indicate particles that were excluded due to poor conservation of the jacobi constant . ]
overall , we find that particles do not hit the star in any of our simulations . scattering to large distances is the most frequent outcome at larger planet masses ; impact with the planet becomes relatively more frequent at larger planet sizes .
we find that impact with the planet was the sole fate of non - surviving particles for planetary masses @xmath5 and @xmath81 .
scattering is the dominant loss mechanism when @xmath82 , @xmath83 , and @xmath84 , for @xmath850.001 , 0.01 , and 0.1 , respectively .
figure [ fig : fr ] plots , as a function of @xmath0 , the fraction of non - surviving particles that were scattered or that collided with the planet , for each case of @xmath6 .
the final destination of test particles as a function of initial semimajor axis and longitude are shown in figure [ fig : br ] , for three representative values of @xmath0 for the case of @xmath47 .
( figures for all other simulated cases of @xmath0 and @xmath6 are available by request from the authors . )
we observe a noticeable asymmetry in figure [ fig : br ] ( see also figure [ fig : tvaall ] ) in the fates of the interior versus exterior particles with initial distance from the planet , especially for larger planet masses .
the range of initial semimajor axis of particles that scatter or collide with the planet is smaller and closer to the planet s orbit in the case of interior particles , and is relatively larger and extends farther from the planet s orbit in the case of exterior particles . in figure
[ fig : br ] , we also observe that most particles with initial orbits such that @xmath86 are survivors ; a minor fraction of these collide with the planet .
this surviving population is associated with the triangular lagrange points , l@xmath87 and l@xmath88 .
these particles are stable over the duration of our integrations , and survive in low eccentricity orbits ( @xmath89 ) either as trojan - type orbits librating about l@xmath87 or l@xmath88 ( analogous to the trojan asteroids of jupiter ) , or as `` horseshoe '' orbits librating about both l@xmath87 and l@xmath88 . for @xmath90 , we find that these `` co - orbital '' survivors span a radial range of @xmath91 within the nominal chaotic zone annulus , [ @xmath92 .
larger ranges of initial longitudes in the co - orbital region are stable for lower planet masses .
the stable co - orbital region decreases significantly for planet masses @xmath93 , and it vanishes for @xmath94 .
this behavior is likely related to the well known linear instability of the triangular lagrange points , l@xmath87 and l@xmath88 , for @xmath95 @xcite .
survivors are also found near the outer edges of the chaotic zone .
a few of these particles survive in nearly circular orbits , in libration in a first order mean motion resonance , and exhibit small ( @xmath96 ) fractional changes in their semimajor axes .
but , more typically , the survivors near the boundaries of the chaotic zone are found on eccentric orbits , with eccentricities up to @xmath97 . this population is larger for the chaotic zones of lower mass planets . for our smallest planet mass , @xmath98 ,
the outer @xmath99 of the radial extent of the chaotic zone survives on eccentric orbits for more than @xmath40 revolutions of the planet .
however , these particles spend insignificant time in the `` cleared zone '' defined below . in previous studies ,
the boundaries of the chaotic zone have generally been defined in terms of the initial conditions . for practical astronomical applications ( for example , gaps in protoplanetary or debris disks )
, it is of interest to define the extent of the `` cleared zone '' , the zone in which the density of surviving particles is nearly vanishing . in order to compute the radial boundaries of this zone , we adopted the following procedure .
we first computed the time - averaged astro - centric distance of the surviving particles in their final osculating orbits , i.e. , @xmath100 .
we binned these values into 98 bins in the range @xmath101 to @xmath102 ; these are the same bins that we adopted for the osculating semimajor axes of the test particle initial conditions .
we computed the survivor fraction ( the ratio of the number of survivors to the initial number ) in each annular bin .
we then determined the interior and exterior boundaries of the cleared zone by finding the most distant annulus on either side of the planet that has a survivor fraction less than 5% .
figure [ fig : reso ] shows these inner and outer boundaries of the cleared zone as a function of @xmath0 , for the case of @xmath103 .
we observe a noticeable asymmetry in the interior and exterior cleared zones , which increases with increasing @xmath0 .
for each case of @xmath6 , we computed a least - squares fit to a power law of the form @xmath104 to the interior and exterior boundaries of the cleared zone .
the values of the parameters @xmath105 and @xmath106 are given in table [ tbl : clzone ] .
we find , unsurprisingly , that both the power law index and the numerical coefficient have little dependence on the value of @xmath6 .
the cleared zone boundary interior to the planet s orbit mimics the first order resonance overlap criterion , eq .
[ eq : rowidth ] , with @xmath107 .
the exterior boundary has a slightly steeper power law index as well as a larger numerical coefficient ; it is better described by @xmath108 .
this asymmetry , which increases with increasing @xmath0 , arises because the locations and strengths of interior and exterior overlapping first order mean motion resonances are not symmetric . , for initially circular orbits near a planet with @xmath47 .
numbers under each point correspond to the values of @xmath109 in eq .
[ eq : pro ] , the analytical estimate of the most distant overlapping @xmath14 resonance .
the straight line indicates the chaotic zone width from the first order resonance overlap criterion ( eq . [ eq : rowidth ] ) , with c=1.5 . ]
.best - fit power law function for the interior and exterior boundaries of the cleared zones . [ cols= " < ,
< , < , < , < , > , > " , ] to describe the behavior of @xmath110 , we computed a nonlinear least - squares fitting to a double power law function , @xmath111 where @xmath112 and @xmath113 are the asymptotic power law slopes at @xmath114 and @xmath115 , respectively , and @xmath116 is the clearing timescale at the transition mass , @xmath117 .
the best - fit values of the parameters @xmath118 and @xmath113 , together with their @xmath119 confidence limits , are listed in table [ tbl : tcl ] , for each of the three cases of @xmath120 .
the best - fit double power law models are plotted as dashed lines in figure [ fig : tcl ] .
using numerical integrations , we examined the orbital evolution of particles in the chaotic zone of a planet , within the framework of the circular , planar restricted three body model , in order to determine the final destinations of chaotic particles and their clearing timescales .
our results are summarized as follows . 1 . for the planetary masses
@xmath5 and planet sizes @xmath3 , particles in initially circular orbits within the chaotic zone are lost via impact with the planet . for more massive planets , chaotic particles may either collide with the planet or scatter to large distances ; scattering is the dominant fate of chaotic particles for planet - star mass ratios @xmath82 , @xmath83 , and @xmath84 , for @xmath120 , respectively . for the entire range of parameters that we investigated
, we found no particles colliding with the star .
2 . for planetary masses
@xmath121 , most particles with initially circular orbits in the annulus of half - width approximately one hill radius , @xmath122 , surrounding the planet s orbit remain stable , in low eccentricity orbits in libration about the triangular lagrange points , l@xmath87 and l@xmath88 .
this stable `` co - orbital '' region is approximately @xmath123% of the radial range defined by the classical chaotic zone annulus , eq .
[ eq : rowidth ] .
the stable co - orbital region decreases for @xmath93 and vanishes for @xmath124 .
3 . particles initially near the inner and outer boundaries of the chaotic zone can be long lived , surviving more than @xmath125 revolutions of the planet .
a minor fraction survive on low eccentricity orbits in small pockets of stability associated with the libration centers of first order mean motion resonances
. however , most of these particles survive in eccentric orbits and spend little time in the vicinity of the planet .
this long - lived chaotic population is larger for lower mass planets ; for @xmath98 , nearly 45% of the outer radial extent of the chaotic zone survives in eccentric orbits for more than @xmath40 revolutions of the planet .
the interior and exterior cleared zone boundaries , defined as the radius where the particle survivor fraction is reduced by more than 95% , are not symmetric about the planet s orbit : the outer cleared zone boundary is farther from the planet than the inner cleared zone boundary .
the interior boundary mimics the classical first - order resonance overlap zone ( eq . [ eq : rowidth ] ) , with a slightly smaller numerical coefficient , @xmath126 .
the exterior boundary of the cleared zone has a slightly steeper dependence on @xmath0 , and is better described by @xmath127 .
5 . for fixed ratio of the planet radius to its hill radius ,
the timescale , @xmath110 , to clear the chaotic zone has double power law behavior ( fig .
[ fig : tcl ] , eq .
[ eq : tcl ] , table [ tbl : tcl ] ) . a shallow slope prevails at small @xmath0 , a steeper slope at large @xmath0 .
the asymptotic behavior is approximately @xmath128 for small @xmath0 and @xmath129 for large @xmath0 .
the shallow slope behavior correlates with particle clearing dominated by collisions with the planet ; the steeper slope behavior correlates with scattering being the dominant particle loss mechanism . for @xmath130 ,
the transition between the two power laws occurs at @xmath131 ; the corresponding clearing timescale is approximately @xmath132 planet revolutions .
the transition mass , @xmath117 , increases roughly linearly with @xmath6 .
the lack of collisions with the star is rather surprising , at first sight . it may be due to a hidden dynamical constraint or may be owed to a computational limitation in our work .
this is an interesting topic for a future investigation .
our results show that collisions with the planet become less frequent while scattering becomes more frequent when the planet s mass exceeds a critical mass , @xmath117 .
the relative probabilities of particle loss by collision with the planet or by gravitational scattering are not scale - free , and @xmath117 depends upon the planet size . in our work , we scaled the planet size to its hill radius , @xmath6 , and computed the relative probabilities for values of this parameter spanning a fair range of physically relevant values .
smaller values of this parameter greatly challenge our computational resources .
however , our results indicate that , for @xmath133 , scattering is the dominant outcome . in this limit ,
the clearing timescale is nearly independent of @xmath6 ( as witnessed by the convergence at @xmath115 in fig .
[ fig : tcl ] ) . using the results for @xmath48 as a proxy ,
the clearing timescale in the limit of vanishing @xmath45 can be approximated as @xmath134 ( in units of the planet s orbit period ) .
furthermore , in the parameter regime where clearing is dominated by collisions with the planet , the clearing timescale can be approximated as @xmath135 .
thus , collisions dominate when @xmath136 for @xmath18 .
we can summarize these estimates for the clearing timescale ( in units of the planet s orbital period ) as follows , @xmath137 we remind the reader that these results are obtained for the planar case .
the three dimensional case has two additional free parameters ( inclination and nodal angles ) and is computationally much more demanding : particle clearing timescales will , in general , be longer , and we can expect the collision dominated clearing time to scale as @xmath138 .
we hope to use the present results to develop intuition and to stimulate further numerical and theoretical analysis of this problem , so as to better understand the aggregate chaotic evolution of particles in the vicinity of a planet . despite their limitations ,
our results indicate that particle clearing timescales can be quite long in a planet s chaotic zone , for all but the very massive close - in planets .
distant , high mass planets clear material predominantly by scattering rather than by collisions with the planet . to illustrate
, neptune would require more than @xmath139 myr to clear its chaotic zone at its current semi - major axis ( 30 au ) .
such long clearing timescales are pertinent for the interpretation of recent observations of debris disks of young exo - planetary systems .
the clearing timescale for putative long period planets sculpting these disks can be comparable to the age of the system .
therefore , given the age of a system with a detectable gap in a circumstellar disk , the chaotic zone clearing timescale can be used to derive a lower mass limit of a ( usually undetected ) planet required to clear the gap .
hr 8799 and hd 95086 are two recently detected exo - planetary systems , each hosting a warm inner debris disk and a cold outer disk with inner edges beyond @xmath140 au and @xmath141 au , respectively @xcite .
a putative planet shepherding the inner edge of the cold disk would have orbital period @xmath142 yr and @xmath143 yr , respectively , and @xmath144 .
for the system ages of @xmath14530 myr and 17 myr @xcite , our results imply that a shepherding planet that could clear dust and planetesimals near the inner boundary of the cold disk must have a mass @xmath146 .
( our model uses massless disk particles and a fixed orbit of the planet ; if the debris disk is massive , it could significantly affect the planet s orbit , which may affect the robustness of this planet mass limit . ) indeed , hr 8799 is known to host four planets ( directly imaged ) , each of 57 @xmath147 ( based on planetary thermal models ) ; hd 95086 is known to host a single planet ( directly imaged ) interior to its cold debris disk , with an estimated mass of @xmath148 also based on planetary thermal models @xcite .
the lower mass limit imposed by dynamical clearing timescales offers an independent constraint on the mass of directly imaged planets in debris disk systems ; and it may be used to estimate the masses of undetected planets . this constraint may become of greater value in the future , when higher sensitivity and higher spatial resolution observations of exo - planetary systems hosting fainter debris disks sculpted by lower mass planets become available .
we thank scott tremaine and roman rafikov for helpful discussions .
we also thank an anonymous reviewer for comments that have improved this manuscript . this research was partially supported by nasa ( nessf grant nnx13ao65h ) and nsf ( grant ast-1312498 ) .
this research made use of the nasa astrophysics data system bibliographic services and the nasa exoplanet archive . | we investigate the orbital evolution of particles in a planet s chaotic zone to determine their final destinations and their timescales of clearing .
there are four possible final states of chaotic particles : collision with the planet , collision with the star , escape , or bounded but non - collision orbits . in our investigations , within the framework of the planar circular restricted three body problem for planet - star mass ratio @xmath0 in the range @xmath1 to @xmath2 , we find no particles hitting the star .
the relative frequencies of escape and collision with the planet are not scale - free , as they depend upon the size of the planet . for planet radius @xmath3 where @xmath4 is the planet s hill radius
, we find that most chaotic zone particles collide with the planet for @xmath5 ; particle scattering to large distances is significant only for higher mass planets . for fixed ratio @xmath6 , the particle clearing timescale , @xmath7 , has a broken power - law dependence on @xmath0 .
a shallower power - law , @xmath8 , prevails at small @xmath0 where particles are cleared primarily by collisions with the planet ; a steeper power law , @xmath9 , prevails at larger @xmath0 where scattering dominates the particle loss . in the limit of vanishing planet radius ,
we find @xmath10 .
the interior and exterior boundaries of the annular zone in which chaotic particles are cleared are increasingly asymmetric about the planet s orbit for larger planet masses ; the inner boundary coincides well with the classical first order resonance overlap zone , @xmath11 ; the outer boundary is better described by @xmath12 , where @xmath13 is the planet - star separation . |
it is well established that the statistical properties of energy levels in an isolated metal present universal features characteristic of the chaotic systems@xcite . at low energy scales , in the so - called ergodic regime ,
the correlations functions are well described by the random matrix theory ( rmt)@xcite . in the diffusive regime , i.e. when the electronic mean free path @xmath10 is smaller than the typical size @xmath11 of the system , deviations from rmt occur @xcite when the energy scales become larger than the thouless energy @xmath12 .
this energy is the inverse of the characteristic time @xmath13 for a particle to diffuse through the sample .
@xmath14 is the diffusion coefficient . on the other hand , in the localized regime ,
the correlations between levels are weaker and in the limit of an infinite system the statistics of energy levels becomes poissonnian .
it has been argued that the statistics of energy levels _ at _ the metal insulator transition ( mit ) is characterized by a third universal distribution which is a hybrid between the wigner and the poisson distributions .
this was clearly shown by a numerical study of the scaling of the nearest neighbor level spacing distribution @xmath15@xcite .
several other numerical works have confirmed this idea @xcite .
moreover , by using diagrammatic calculations , semiclassical description and scaling arguments , it was found that long range correlations between levels exhibit a new power
law behavior whose exponent is related to the exponent @xmath16 of the localization length @xcite .
there are several quantities which measure the fluctuations of energy levels . in the rmt
, they depend only on the symmetry of the hamiltonian .
if it is invariant under time reversal symmetry , the fluctuations are described by the gaussian orthogonal ensemble ( goe ) of random matrices ( @xmath17 ) . when time reversal symmetry is broken , the spectrum becomes more rigid ( gaussian unitary ensemble , gue , @xmath18 ) .
these quantities are the following : * the number variance @xmath19 @xmath20 it measures the fluctuation of the number of levels @xmath21 in a strip of width @xmath22 .
the average @xmath23 can be taken either over different regions of the spectrum or over several realizations of disorder . in the rmt
, @xmath19 increases logarithmically with @xmath22 . for @xmath24
, it varies as @xmath25 where @xmath26 , @xmath27 being the average interlevel spacing . * the distribution @xmath28 of the spacing @xmath29 between consecutive levels . in the rmt
, it is well described by the wigner surmise : @xmath30 , while , when there is no correlation between levels , it has a poisson behavior : @xmath31 . * the two - point correlation function of the density of states ( dos)@xcite .
however , we have chosen to follow here the notations of refs.@xcite ] : @xmath32 where @xmath33 is the average dos . using the definition of the dos ,
@xmath34 can be simply rewritten as : @xmath35 where @xmath36 is the distribution of distances @xmath37 between levels separated by @xmath38 other levels@xcite . by definition @xmath39 .
the number variance can obviously be written in terms of this two - point correlation function : @xmath40 * the form factor @xmath41 , fourier transform of @xmath34 : @xmath42 the interest of this quantity is that is can be directly related to some characteristics of the classical motion of the diffusive particles@xcite .
although the number variance is most frequently used in the literature , it is not always the most appropriate quantity to describe the correlations , since it is a double integral of the dos - dos correlation function @xmath34 .
thus the behavior of this quantity at an energy scale @xmath22 depends on the behavior of @xmath43 for _ all _ energies @xmath44 smaller than @xmath22 .
that is why in this paper we have chosen to study @xmath34 as well as its fourier transform @xmath5 directly . although analytical predictions have been given for the asymptotic behavior of these quantities ( large energies or small times ) at the mit @xcite , nothing was known yet about the short range behavior of the correlations .
it is one of the goals of this paper to study these correlations . in section 2
we recall what is their behavior in the metallic regime and compare with numerical experiments . in section 3
, we analyze the two point correlation function at the mit .
section 4 is devoted to a discussion of level spacing distributions , and finally , in section 5 , we present our conclusions .
efetov has shown that at energy scales smaller than the thouless energy @xmath45 , and in the limit where the dimensionless conductance @xmath46 is much larger than @xmath47 , i.e. far from a mit , the two - point correlation function @xmath34 is given by its expression in the gaussian rmt@xcite .
this was confirmed numerically in the framework of the tight - binding anderson model with diagonal disorder : the number variance and the distribution @xmath15 were found to be very well described by the wigner - dyson predictions@xcite .
we first show numerical results for the two - point correlation @xmath34 , fig.[f1 ] , and its fourier transform @xmath41 in the metallic regime , fig.[f2 ] .
it is seen that the rmt result is a very good description of these correlations .
they were obtained by exact diagonalization of a tight binding anderson hamiltonian with on diagonal disorder via a lanczos routine .
the diagonal elements are box
distributed around zero with a width @xmath48 , in units of the transfer integral @xmath49 .
we only use eigenvalues which lie in a central energy strip whose width is chosen such that it contains approximately half the number of all eigenvalues .
even though the average density of states ( dos ) is almost constant in this region , special care has to be taken in unfolding the spectrum .
then we calculate @xmath34 defined in eq.[rdes ] .
thus , @xmath50 is the distribution of differences of eigenvalues in the energy interval @xmath51 used for averaging .
in addition , we average on different disorder configurations . the self - correlation of the levels is not included in the numerical calculation of @xmath34 .
a @xmath52 function has to be added to the result .
using rather small systems but many disorder realizations ( @xmath53 sites and 1000 disorder realizations for @xmath54 ) , we are able to obtain a large amount of eigenvalues and correspondingly rather smooth curves for @xmath55 .
as shown in fig.[f1]b , the numerical precision is even high enough to allow for the observation of the tiny oscillations in @xmath55 in the gue case where @xmath56 .
let us briefly recall some important features of these correlation functions .
it is helpful to use a semiclassical argument developped by argaman _
et al._@xcite in which the function @xmath41 is directly related to the return probability @xmath57 after a time @xmath49 for a classical diffusive particle .
more precisely : @xmath58 the factor @xmath59 in the numerator arises from interference effects between time reversed paths , in the absence of magnetic field or flux .
this relation is valid for classical times @xmath49 much smaller than the heisenberg time @xmath60 . and characteristic times ,
see ref.@xcite ] for time scales larger than @xmath61 , the diffusion is homogeneous so that @xmath57 is constant .
this gives a linear behavior for @xmath41 as observed in fig.[f2 ] .
for @xmath62 , @xmath41 has to saturate . in the energy space
, @xmath63 has first a @xmath52 peak at the origin which describes the self correlation of the levels .
fourier transform of this peak implies that @xmath64 for large @xmath49 . for small energies , there is a relation between the function @xmath34 and the distribution of interlevel spacings @xmath15 : @xmath65 so that indeed @xmath34 starts linearly at small @xmath6 in the absence of a magnetic field . at larger energies
, @xmath34 varies like @xmath66 .
unless otherwise specified , we will now consider the case @xmath67 , i.e. no magnetic field . for time scales smaller than @xmath68 ,
the diffusion is non homogeneous and the classical return probability varies as @xmath69 so that @xmath41 scales like @xmath70 . for @xmath71 , its fourier
transform @xmath34 is given by : @xmath72 and the number variance behaves like @xmath73 for @xmath74 @xcite .
this `` altshuler shklovski '' power - law behavior had not been seen numerically yet @xcite .
this is because there is a long cross over regime around the energy scale @xmath45 , and the power law
is expected to be visible for very large energy scales @xmath75 . in the intermediate regime @xmath76 ,
the quantization of the diffusion modes has to be taken into account : @xmath77 using a large system with small @xmath45 , we have calculated @xmath19 numerically and found the onset of the power law .
@xmath19 starts to be proportional to @xmath78 only for energies @xmath79 ( see fig.[f3 ] ) .
if one takes into account the fact that for a finite system of size @xmath80 the condition @xmath81 should be fullfilled in order to neglect finite size effects , and if one wants to observe the power law over at least one decade of energy , the size of the system should be at least of the order @xmath82 .
on the other hand , the discrete sum describes the numerically calculated @xmath19 very well , and a fit to this sum with @xmath45 as only fitting parameter allows for a precise determination of the thouless energy of the sample ( fig.[f3 ] ) .
it turns out that this contribution of the finite-@xmath83 diffusion modes is hardly visible on the two - point correlation function @xmath84 . for a metal in three dimensions
, @xmath84 varies at large energies positively like @xmath85 instead of @xmath86 for rmt , meaning that the levels tend to attract each other at large energies@xcite .
however , these deviations from rmt are very small , less than a few percent .
we have not been able to see them on the numerically calculated @xmath84 , see fig.[f1 ] . for the same reason ,
the deviations from the linear behavior of @xmath5 for @xmath87 are not visible . moreover , in the metallic regime that we have studied ( @xmath88 ) , we did not see the finite conductance deviations at small energy scales @xmath89 predicted recently @xcite .
we now turn to the description of the correlations at the mit .
the first theoretical predictions of spectral correlations at the mit different from those in the metallic regime were given by altshuler _
et al . _
they predicted a purely linear behavior for the spectral rigidity , @xmath90 for @xmath91 , where the constant @xmath92 was estimated to be of the order of @xmath93 , and thus considerably different from the behavior in the insulating regime where @xmath92 is unity .
later , shklovski _ et al .
_ argued that the level statistics at the mit should be completely universal @xcite .
they predicted a level spacing distribution which should be a hybrid between a wigner dyson type behavior for small @xmath6 , that is @xmath94 for @xmath95 , and a poisson type for @xmath96 : @xmath97 .
then kravtsov _ et al .
_ brought new dynamics into the field by showing that the two - point correlation function @xmath34 has a new long range power - law behavior : instead of varying as @xmath98 like in the wigner - dyson regime , it is predicted to obey@xcite : @xmath99 the non - universal constant @xmath100 depends on the dimension and on the symmetry of the ensemble , and @xmath101 is given by @xmath102 , where @xmath16 is the critical exponent of the localization length @xmath103 .
if one uses the value @xmath104 at the mit @xcite , the exponent @xmath101 is of the order of @xmath105 .
this new power - law behavior can be derived semiclassically from @xmath41 along the same lines as in the metallic regime , but now with anomalous diffusion @xcite .
near the mit , the dimensionless conductance @xmath106 scales with distance @xmath107 like @xmath108 $ ] , where @xmath16 is the exponent of the localization length @xmath109 .
since @xmath27 scales like @xmath110 and @xmath45 like @xmath111 , this implies that the diffusion coefficient has to be size dependent .
for @xmath112 it has to vary as @xmath113\,.\ ] ] to this length dependent @xmath114 , one can associate a time dependent diffusion coefficient @xmath115 , @xmath116\,,\ ] ] so that according to eqs .
[ ais ] and [ diff ] , @xmath41 is constant up to a time dependent correction : @xmath117\,.\ ] ] the semiclassical expression ( [ ais ] ) is known to be equivalent to the two diffuson diagrammatic result of altshuler - shklovski @xcite , now with an energy dependent diffusion coefficient .
et al._@xcite argue that at the transition , there are many more diagramms with the same structure so that the correct prefactor of the power law is not known , even in sign .
this structure is supposed to be correct for @xmath118 . by fourier transform , one has the tail of @xmath34 @xmath119 at the transition @xmath120 so that one finds the size independent power law for the tail of the two - point correlation function , eq.[tail ] .
the integration of the power - law tail gives a contribution to the number variance @xmath19 varying as @xmath121 at large energies .
the existence of a linear term in @xmath122 was first ruled out because of a sum rule@xcite a statement later discussed and contested by aronov and mirlin @xcite .
+ finally , sears and shore @xcite introduced the function @xmath123 , which they computed numerically . with the hypothesis that the scaling behavior of @xmath124 should only depend on the combination of parameters @xmath125 , they were able to reproduce the term @xmath121 in @xmath122 with the same @xmath101 as introduced in @xcite .
however , they indeed found an additional linear term .
their numerical results are in agreement with @xmath126 for a rather large range of disorder and for system sizes up to @xmath127 .
all in all , a consensus now emerges that for @xmath128 , the number variance should contain two terms @xmath129 although the power law originates directly from the long range behavior of the two - point correlation function , the origin of the linear term is more subtle .
it appears as a constant of integration which is a _
global _ measure of the interaction between eigenvalues integrated over energy .
more precisely , since by definition ( eq.[sigmar ] ) , @xmath34 is the second derivative of @xmath130 , the coefficient @xmath92 of the linear term is simply given by @xmath131 .
the absence of a linear term in @xmath19 in the rmt implies that the `` sum rule '' @xmath132 is fulfilled .
the coefficient @xmath92 contains informations about _ all energy scales _ , in particular low energy scales .
however , until now , most studies concentrated on the long range behavior of @xmath133 .
in the present work , we focus on the _ direct _ calculation of the two - point correlation function and of the form factor both for @xmath134 and @xmath135 . using the method described above , we have calculated @xmath34 for rather small system sizes but using many disorder realizations for @xmath136 . in the energy range we considered ,
finite size effects can be neglected , which we checked by using different sizes . in fig.[f6 ] we show the universal results obtained from systems with different sizes in an energy range @xmath137 . in this small energy range
we used system sizes up to @xmath138 sites and up to 1000 disorder realizations .
obviously , the correlations are weaker than in the goe and qualitatively in agreement with the analytical result of kravtsov and mirlin @xcite which predicts a correction to the small energy slope increasing like @xmath139 when @xmath140 decreases .
the linear variation of @xmath34 at small @xmath6 is consistent with the behavior of @xmath15 at small @xmath6 ( fig.[f7 ] ) .
more precisely @xmath141 .
this means that all the @xmath36 for @xmath142 start at least quadratically for @xmath143 , like in the rmt@xcite . for practical purpose ,
we have fitted @xmath34 for @xmath144 with @xmath145 the found coefficients @xmath146 and @xmath147 are shown in table [ tab1 ] .
.numerical coefficients that describe the short range correlations ( see eq .
( [ rsum ] ) ) .
[ tab1 ] [ cols=">,<,<,<,<",options="header " , ] analyzing the central half of the spectra ( totaling about 23 million levles ) , we have been able to extract the behavior of @xmath34 for @xmath148 . in fig.[rslog ] we show the average correlation function obtained from all systems together with the best fit of the tail to a power law , given by @xmath149 in good agreement with the predicted eq.[tail ] .
the comparison of the fitted variables for the five different geometries used allows to estimate the remaining statistical error for the exponent to @xmath150 and for the prefactor to @xmath151 ( simple standard deviations , see table [ tab2 ] ) . rather than by statistical errors , the calculation of @xmath34 for even higher energies
is limited by the precision of the unfolding of the spectra : we were able to obtain @xmath152 constant with a precision of about @xmath153 .
for @xmath154 , @xmath34 becomes comparable with this value and thus starts to fluctuate around zero .
this explains the sharp drop in fig.[rslog ] at these energies and might cause a further systematic error for @xmath6 smaller but close to @xmath155 .
we estimate the total error of the exponent to be less than about 20% .
we now turn to the discussion of the sum rule . as we have seen above
, the coefficient of the linear term in @xmath122 is the integral of @xmath34 .
it can be written as @xmath156 in the case of the wigner - dyson statistics , the weight @xmath157 , exactly compensates the weight 1 of the @xmath52 function at the origin .
in the case of a poisson statistics , @xmath34 is zero for @xmath158 , so that the @xmath52 function is not compensated and there is a linear term with slope @xmath47 in @xmath122 .
the situation at the mit is intermediate .
we see here that the prefactor is modified by the short range behavior of @xmath34 . inserting the numerical values from the fit , we find that @xmath159 is about @xmath160 for @xmath161 and @xmath162 for @xmath163 .
long range correlations reduce @xmath92 further : _ assuming _ that one can extend the obtained power law for @xmath34 up to much higher energies , one can even estimate @xmath164 for very large energies .
it gives @xmath165 for @xmath166 , @xmath167 for @xmath168 and ...
@xmath169 for @xmath170 .
the sum rule seems to be obeyed for energies up to infinity .
this result constradicts the results of refs.@xcite and @xcite who predicted @xmath171 and @xmath172 , respectively .
we finish this study of the two point correlation function by showing in fig.[f5 ] the numerically calculated struture factor @xmath41 as a function of time , including energies up to @xmath173 .
as expected from the above semiclassical argument , it tends to a finite value for small @xmath49 ( but still @xmath49 has to be larger than the collision time ) .
this finite value is related to the slope of @xmath19 for @xmath174 : @xmath175 . in our plot
, @xmath41 seems to converge to the value @xmath176 .
however , in the numerical evaluation of @xmath41 one has to use a cut off in energy .
using @xmath177 , our data for @xmath178 are valid only down to @xmath179 , so that we are not able to calculate properly @xmath8 . for the moment
, we can only conclude that a time dependence of the form @xmath180 is compatible with our data ( @xmath181 ) . for the same reason as in the metallic case ( self correlation of levels )
, @xmath41 has to saturate at @xmath182 for large @xmath49 .
for completeness , we briefly comment on the results concerning @xmath15 . shklovski _
et al._@xcite argued that at large @xmath6 , it must follow a `` modified '' poisson distribution of the form @xmath183 while aronov _
predict a tail of the form @xmath184 .
this tail reflects the behavior of the two - point correlation function at large @xmath6 .
this is found by describing the thermodynamics of a one - dimensional gas with a pairwise interaction consistent with the above two - point correlation function@xcite .
there is no complete consensus yet on the numerical description of @xmath15 at the mit@xcite but it seems that it can be globally well fitted by an expression of the form @xmath185 , the constants @xmath186 and @xmath187 being fixed by normalization of @xmath15@xcite .
quite recently kravtsov and lerner have constructed a plasma model with the appropriate power law interaction in order to get the correct dependence of the two - point correlation function ( [ tail ] ) .
they have been able to relate the coefficient @xmath187 to the prefactor @xmath188 of the power law .
they find a value of @xmath189 in reasonable agreement with our result .
one may wonder if the knowledge of the probability distribution of distances between non - neighboring levels could add more useful information on the long range correlations .
for this purpose , we have studied the function @xmath36 defined above .
these functions are known for the rmt regime@xcite . in the poisson limit
, they simply vary as : @xmath190 it may be much more useful to use these quantities because , for large @xmath6 , @xmath34 and @xmath15 are very small and difficult to calculate . on the other hand , for large distance @xmath6
, @xmath36 is very different from the rmt or poisson cases .
fig.[pns ] shows the successive @xmath36 at the mit , and they are compared to the corresponding functions in the poisson and rmt cases . for @xmath135 , @xmath191 characterizes much better the differences between the statistics at the mit and in the metal than @xmath192 .
unfortunately , we do not know any prediction for the behavior of @xmath36 for large @xmath193 at the mit .
we have presented results of the direct numerical calculation of the spectral two point correlation function in metals and at the metal insulator transition . in the metallic regime
the results are in very good agreement with the rmt predictions .
we have been able to find numerically the altshuler and shklovski regime where the diffusion is power - like . at the metal
insulator transition , the short range behavior of the two point correlation function is clearly modified .
the short range part of the correlation is weaker and is followed by a power law behavior with an exponent which is consistent with the picture of anomalous diffusion . at the transition , @xmath34 is size independent .
this clearly refutes a prediction@xcite that the power law contribution should decrease as @xmath194 .
these short range correlations give rise to a non trivial linear term in the spectral rigidity @xmath19 in addition to a power law term which is the signature of the long range correlations . by integration of @xmath34 , we conclude that the prefactor of this linear term may be much smaller than previously anticipated and may even be zero .
recently , different models of interpolating ensembles of matrices have been proposed @xcite .
our result puts a constraint on the possible choice of ensembles since they should produce a form factor which varies at small times like eq.[kdet ] .
acknowledgments : we have benefitted from useful discussions or correspondance with e. hofstetter , v. kravtsov , i. lerner , m.l . mehta and b. shklovski .
numerical simulations were performed on the cray computer of idris ( institut du dveloppement et des ressources en informatique scientifique , orsay ) .
e. p. wigner , ann . math . * 53 * , 36 ( 1951 ) ; _ ibid _ * 62 * , 548 ( 1955 ) ; _ ibid _ * 65 * , 203 ( 1957 ) ; _ ibid _ * 67 * , 325 ( 1958 ) ; f.j .
dyson , j. math . phys . * 3 * , 140 ( 1962 ) ; f.j .
dyson and m. l. mehta , j. math .
phys . * 4 * , 701 ( 1963 ) ; t. a. brody , j. flores , j. b. french , p. a. mello , a. pandey and s. s. m. wong , rev .
53 * , 385 ( 1981 ) .
o.bohigas , in _ chaos and quantum physics . proceedings of the les - houches summer school .
session lii _ , ed .
by m.j.giannoni et.al ( north holland , amsterdam , 1991 ) , p.91 y.v.fyodorov , phys.rev.lett .
* 73 * , 2688 ( 1994 ) y.v.fyodorov , a.d.mirlin , jetp lett .
* 60 * ( dec 1994 ) , and submitted to phys.rev.b j. l. pichard et b. shapiro , j. phys .
i , france , * 4 * , 623 ( 1994 ) .
m. m. moshe , h. neuberger et b. shapiro , phys .
lett . * 73 * , 1497 ( 1994 ) . c. m. canali , m. wallin , v. e. kravtsov , preprint 1994 . c. blecken , y. chen et k. a. muttalib , j. phys . *
a * : math .
* 27 * , l563 ( 1994 ) . | we have studied numerically the spectral correlations in a metallic phase and at the metal - insulator transition .
we have calculated directly the two - point correlation function of the density of states @xmath0 . in the metallic phase ,
it is well described by the random matrix theory ( rmt ) .
for the first time , we also find numerically the diffusive corrections for the number variance @xmath1 predicted by altshuler and shklovski . at the transition , at small energy scales
, @xmath2 starts linearly , with a slope larger than in a metal . at large separations
@xmath3 , it is found to decrease as a power law @xmath4 , in good agreement with recent microscopic predictions . at the transition
, we have also calculated the form factor @xmath5 , fourier transform of @xmath2 . at large @xmath6 ,
the number variance contains two terms @xmath7 where @xmath8 is the limit of the form factor for @xmath9 .
june 14 , 1995 |
magnetic properties of strongly correlated transition - metal systems and their relation to doping , lattice geometry and band structure are still extensively investigated .
the general result of these investigations is the existence of first - order transition between various commensurate and incommensurate magnetic states which invokes a phase separation ( first discussed by visscher @xcite ) . to describe the properties of such systems one uses many - electron models like the hubbard , @xmath0-@xmath1 exchange ( kondo ) model and anderson lattice model .
these are widely applied , _
e. g. _ , for high-@xmath2 cuprates and rare earth compounds .
there exist some relations between these models in various parameter regions .
the problem of local moments formation , _
e. g. _ within the hubbard model , is very difficult and still under investigation @xcite . on the other hand , in the @xmath0-@xmath1 exchange model
the localized moments ( spins @xmath3 ) are explicitly present in the hamiltonian ( although they are screened in the kondo regime ) . in the present paper
we perform an investigation of the magnetic phase diagram of the anderson - kondo lattice model for the square and simple cubic lattices including the phase separation , as well as non - collinear magnetic ordering , and trace these relations .
we treat the influence of inter - orbital interaction on the spiral state formation , the difference of the hubbard ( one - orbital ) and anderson - kondo lattice ( two - orbital ) model results being considered .
the theoretical investigation of spiral formation in itinerant systems is generally based ( in minimal variant ) on the non - degenerate hubbard and the anderson models . within the hubbard model
@xmath4 the itinerant electrons demonstrate both transport and interaction induced magnetic properties . here
the matrix elements of the electron transfer are @xmath5 for the nearest neighbors ( we assume @xmath6 ) , @xmath7 are the electron creation and annihilation operators , respectively , @xmath8 is the site number , @xmath9 is the spin projection , the last term being responsible for the one - site coulomb interaction of @xmath1-electrons , @xmath10 . in the case of the anderson model transport and magnetic properties
are separated between different systems , @xmath0 and @xmath1 correspondingly : @xmath11 @xmath12 is creation / annihilation electron operator in itinerant ( ` @xmath0-electron ' ) state at site @xmath8 .
@xmath13 is the energy of localized ( ` @xmath1-electron ' ) electron state , @xmath14 is one - site @xmath0-@xmath1 hybridization providing the coupling between these subsystems .
the total electron concentration in the system is @xmath15 , where @xmath16 and @xmath17 are the occupation numbers for itinerant and localized states respectively .
provided that the @xmath1-level is well below the fermi energy and coulomb interaction is sufficiently large ( @xmath18 , @xmath19 ) , this model can be reduced by the schrieffer - wolf transformation @xcite to the @xmath0-@xmath1 exchange model with spin @xmath20 and the exchange parameter @xmath21,\ ] ] where @xmath22 is the fermi level .
the hamiltonian of the latter model reads @xmath23 @xmath24 is localized spin operator , @xmath25 stands for pauli matrices .
we consider ferromagnetic and antiferromagnetic , as well as spiral incommensurate magnetic order , with the magnetization @xmath26 being modulated in the @xmath27-plane with the wave vector @xmath28 @xcite .
after local rotation in spin space matching the average magnetization direction at different sites we have a hopping matrix , non - diagonal with respect to spin , @xmath29 @xcite . the saddle - point expression for the spiral state grand canonical potential ( per site )
@xmath30 has the form @xmath31 where @xmath32 is a contribution from effective fermion hamiltonian @xmath33 describing their motion in the ground state in some effective field , @xmath34 where @xmath35 are eigenvalues of @xmath33 , @xmath36 is the fermi function at @xmath37 , @xmath38 is the chemical potential , @xmath39 is the lattice site number .
@xmath40 is ` inner ' effective field ( ` background ' ) contribution to the grand canonical potential . within the hubbard model @xmath41 where @xmath42 @xmath43 is the bare electron spectrum .
the concrete expressions for the spectrum renormalization factors @xmath44 , @xmath45 and @xmath40 depend on the approximation employed .
the resulting wave vector @xmath46 is determined by minimization of @xmath30 over various spiral states at fixed @xmath38 which allows to take into account the phase separation possibility @xcite .
the generalized hfa for the coulomb interaction in @xmath1-subsystem reads @xmath47 the main shortcoming of hfa is the account of contributions of singly and doubly occupied states to @xmath48 in the equal way , which becomes incorrect at sufficiently large @xmath19 .
correlation - induced band narrowing is absent , @xmath49 and @xmath50 in eq .
( [ eq : hf_def ] ) , @xmath51 in the case of the anderson model we have @xmath52 so that two types of mixing are present : the hybridization @xmath14 of @xmath0- and @xmath1-systems and spin flip terms proportional to @xmath53 .
besides hfa , we apply sba @xcite to the single - band hubbard model .
the idea of this approximation is extension of configuration space .
this duplicates the standard description based on the slater determinant wave functions ( related to operators @xmath54 , @xmath55 ) by using the boson operators @xmath56 , @xmath57 , @xmath58 and their conjugates which correspond to empty , singly occupied , and doubly occupied states respectively . the bosonic space construction is realized by requiring the presence of exactly one boson at any time , @xmath59 any on - site transition operator ( hubbard @xmath60 operator @xcite ) has its counterpart in the slave boson language , _
e.g. _ , @xmath61 for any site @xmath8 , @xmath62 being the projection operator onto the corresponding subspace .
the exact coherence of fermion and boson systems is established by the replacement @xmath63 @xmath64 ( @xmath65 ) being an operator equal to unity on the subspace defined by the equation @xmath66 ( @xmath67 ) needed to reproduce hfa results at small @xmath19 within the saddle - point approximation @xcite .
the consequence of this coherence is the connection of site occupation numbers in terms of fermions and bosons , @xmath68 the hubbard on - site interaction becomes diagonal in the boson representation : @xmath69 . within the saddle - point approximation ,
the bosonic operators are replaced by @xmath54-numbers .
this yields an improvement of the hartree - fock approximation , so that the corresponding effective field can be interpreted as a result of average many - electron site amplitudes satisfying eq .
( [ eq : norm ] ) .
the partial electronic concentrations are parametrized by @xmath70 the subband narrowing is the @xmath54-number function of extremal bosonic fields , @xmath71 the grand canonical background potential has the form @xmath72 the impact of the averaged site states on electron states manifests itself in two types of renormalizations of bare spectrum : ( i ) narrowing of the bare spectrum , similar to the hubbard
i approximation@xcite , which is specified by the factor @xmath73 , ( ii ) the additional energy shift @xmath45 which is an analogue of the harris lange shift @xcite .
both these quantities are essentially spin dependent , which allows one to study the magnetic states formation . unlike hfa , in the slave - boson approach @xmath45
can not be expressed in terms of @xmath74 and @xmath75 only and is obtained from the saddle - point equation as @xmath76,\ ] ] @xmath77 the balance of bosonic @xmath78 and @xmath1 amplitudes is determined by the equation @xmath79 the problem of correlations in the anderson model can be also treated in sba @xcite . however , this yields the exponent in the kondo temperature which is different from the value in the single impurity case . on the other hand , pseudofermion representation by coleman and andrei in the @xmath0-@xmath1 model @xcite
does not have such a drawback and can be successfully applied to the problem of magnetic ordering @xcite .
here we consider the ground state magnetic phase diagram for square and simple cubic lattices for both the hubbard and @xmath0-@xmath1 models calculated in the generalized hfa . for the hubbard model we present also sba results ( in the case of bcc and fcc lattices , they can be found in ref .
@xcite ) . the more general case of intermediate valence within the anderson lattice model will be considered elsewhere , see also @xcite .
we present the results for the square ( fig .
[ fig : sq ] ) and simple cubic ( fig .
[ fig : sc ] ) lattices in the hubbard within the hfa and sba approximations and the @xmath0-@xmath1 exchange model .
one can see that hfa yields the variety of spiral magnetic phases , as well as fm and afm ones at sufficiently large @xmath19 generally and at any @xmath19 in the vicinity of half - filling .
the account of correlations ( sba ) leads to a noticeable suppression of magnetically ordered states in comparison with hfa : the corresponding concentration intervals in the phase diagram decrease strongly , and the variety of the spiral states disappears .
besides that , in sba there occurs a wide region of pm state which is a manifestation of correct treatment of the energy of doubles .
-@xmath1 model ( lower panel ) for the square lattice at @xmath80 .
the spiral phases are denoted according to the form of their wave vector .
filling shows the phase separation regions .
bold ( blue ) lines denote the second - order phase transitions .
solid ( red ) lines correspond to the boundaries between the regions of the homogeneous phase and phase separation . dashed ( red ) lines separate the ps regions corresponding to different phase pairs .
the second - order phase transition boundary produced by the stoner criterion separating the ferromagnetic and paramagnetic regions , is shown by dash dotted ( violet ) line , @xmath81 , title="fig:",scaledwidth=50.0% ] -@xmath1 model ( lower panel ) for the square lattice at @xmath80 .
the spiral phases are denoted according to the form of their wave vector .
filling shows the phase separation regions .
bold ( blue ) lines denote the second - order phase transitions .
solid ( red ) lines correspond to the boundaries between the regions of the homogeneous phase and phase separation . dashed ( red ) lines separate the ps regions corresponding to different phase pairs .
the second - order phase transition boundary produced by the stoner criterion separating the ferromagnetic and paramagnetic regions , is shown by dash dotted ( violet ) line , @xmath81 , title="fig:",scaledwidth=49.0% ] for the hubbard model on the square lattice we have in the vicinity of half - filling only the spiral @xmath82 ( diagonal ) phase , far away from half - filling we have a fm phase for large @xmath83 and the parallel spiral phase ( @xmath84 or @xmath85 ) for moderate @xmath83 . here
@xmath86 is a _ number _ depending on the model parameters .
the physical picture for the cubic lattice is similar to that for the square lattice ( see fig . [
fig : sc]a for a comparison of the results of hfa ( a ) and sba ( b ) approaches ) : the density value @xmath87 corresponding to the perfect afm nesting peculiarity ( but not to a van hove singularity ) retains its crucial role .
the spiral magnetic phase with @xmath88 of the square lattice is replaced by a spiral with @xmath89 .
the density of states has a van hove ( vh ) singularity at the electron concentration @xmath90 .
in contrast to the 2d case , the van hove singularities are bounded and do not much affect the sba diagram the correlation effects strongly suppress magnetism at @xmath74 being far from half - filling , including the regions close to @xmath91 s .
, @xmath92 , title="fig:",scaledwidth=49.0% ] , @xmath92 , title="fig:",scaledwidth=49.0% ] it should be noted that the description of ferromagnetism at small @xmath74 requires the account of @xmath93-matrix renormalizations @xcite .
the ground state phase diagram for the @xmath0-@xmath1 model substantially differs from the results for the hubbard model : the phases are strongly redistributed . the increasing of @xmath94 results in growing of ferromagnetic region . at small @xmath94 the wave vector of magnetic phase
is specified by the position of the maximum of the lindhardt function @xmath95 calculated in the paramagnetic phase .
an important difference of the square and sc lattice ( 2d and 3d ) cases is the form of ferromagnetic region at small @xmath74 . in the former case its width vanishes at @xmath96 , but in the latter case the width is finite and sufficiently large , the transition to the spiral @xmath97 phase being of the first order . for the square lattice
the spiral phases are fully suppressed by fm and afm regions at @xmath98 . for sc lattice
the spiral phases turn out to be more stable . for small number of carriers in the afm matrix ( @xmath99 ) the phase separation between afm and spiral ( @xmath85 for square lattice and @xmath100 for sc lattice ) phases
is present at small interaction parameter . within the mean - field approximation ,
the hubbard model is equivalent to the @xmath0-@xmath1 model with the replacement @xmath101 .
however , the phase separation condition and description of the paramagnetic phase are different in these models . due to existence of localized moments ,
ferromagnetic ordering is favorable already at small @xmath94 , whereas within the hubbard model it occurs at sufficiently large @xmath19 only .
our calculations for the square lattice @xmath0-@xmath1 model are in a qualitative agreement with the monte carlo calculations for the classical @xmath0-@xmath1 model @xcite and slave - fermion treatment of the kondo lattice @xcite ( phase separation and spirals are neglected in both the works ) .
the spiral phases were taken into account in ref .
however , the @xmath82 spiral phase obtained in this work turns out to be unstable with respect to phase separation into the nel afm and @xmath85 spiral phases . with increasing @xmath94 ,
the kondo screening of the localized moments occurs , which is crucial for rare - earth compounds .
this screening can be also included in our phase diagrams and will be considered elsewhere .
the results for 2d and 3d cases are expected to be considerably different too .
the research was carried out within the state assignment of faso of russia ( theme `` electron '' no .
01201463326 ) .
this work was supported in part by the division of physical sciences and ural branch , russian academy of sciences ( project no . 15 - 8 - 2 - 9 , 15 - 8 - 2 - 12 ) and by the russian foundation for basic research ( project no .
16 - 02 - 00995-a ) and act 211 government of the russian federation 02.a03.21.0006 .
the main amount of calculations was performed using the `` uran '' cluster of imm ub ras . | the ground - state magnetic phase diagram is calculated within the hubbard and @xmath0-@xmath1 exchange ( kondo ) for square and simple cubic lattices vs band filling and interaction parameter .
the difference of the results owing to the presence of localized moments in the latter model is discussed .
we employ a generalized hartree - fock approximation ( hfa ) to treat commensurate ferromagnetic ( fm ) , antiferromagnetic ( afm ) , and incommensurate ( spiral ) magnetic phases .
the electron correlations are taken into account within the hubbard model by using the kotliar - ruckenstein slave boson approximation ( sba ) .
the main advantage of this approach is a correct qualitative description of the paramagnetic phase : its energy becomes considerably lower as compared with hfa , and the gain in the energy of magnetic phases is substantially reduced .
magnetic ordering , electron correlations , phase separation |
the design and synthesis of strong organic bases have long been an active field of research @xcite .
infrared spectroscopy is a valuable tool in order to obtain information about the molecular structure and properties of the molecules .
this technique is used widely in qualitative and quantitative molecular analysis .
ir spectrum of interatomic vibrations can be used as structural probes for determining weak changes of structure or chemical bonding in molecules .
cyclohexene-2-ethanamine molecule consists of cyclohexene @xmath2 group attached to the carbon of ethylamine @xmath3 .
there are previous works on the cyclohexene and ethylamine structures .
some studies showed that the lowest energy conformations of cyclohexene are in a half - chair form and a boat structure . basically , the cyclohexene ring can interconvert from one twisted form to the other over the boat conformation with @xmath4 symmetry @xcite . the point symmetry group for trans - ethylamine ion
is @xmath4 whereas there is no such symmetry for gauge - ethylamine @xcite .
cyclohexene-2-ethanamine ( cyhea ) has also important industrial applications , that is used as chemical intermediate in rubber industry .
they demonstrated prototypical non - conjugated olefinic substrate cyhea which was not only a highly active substrate but also a mechanism - based inhibitor for dbm .
cyhea was also used as a substrate and oxidizing agent for ru complex .
sirimanne and may reported that dopamine @xmath5-monooxygenase ( dbm ) catalyzed stereo - selective allylic hydroxylation of cyhea @xcite .
cyhea was first synthesized by izgi et al .
@xcite and some of ir and nmr properties of this compound were reported by them .
density functional theory(dft ) is a widely used and very precise _ ab initio _ technique which is used to provide vibrational frequencies of organic compounds perfectly @xcite .
the vibrational modes and stm images of this molecule have not been investigated by an _ ab initio _ theoretical method . in this study , the molecule has been investigated by using planewave pseudopotential calculation based on dft .
exchange - correlation potential of dft scheme was taken into account within the lda and gga which are the commonly used approximations and both are used in calculation process .
the stable conformation of the molecule is obtained by following a relaxation procedure within the framework of dft under periodic boundary conditions .
the normal modes and stm images of the molecule were calculated with both lda and gga by using the freely available dft program pw - scf ( plane wave self consistent field ) @xcite which uses plane wave basis sets for electronic wavefunctions .
for all calculations , we have used perdew - zunger @xcite and perdew - burke - ernzerhof @xcite exchange - correlation parameterizations for lda and gga , respectively and vanderbilt @xcite ultrasoft pseudopotentials .
the electronic wavefunctions were expanded in terms of plane waves with kinetic energy cut - off up to 25 ry .
the special k - points of the molecule in the cubic cell is selected as @xmath6 gamma point . the lattice constant of cubic cell is 20 bohr(au ) . for experimental work
, the pure cyclohexene-2-ethanamine in liquid form was obtained from aldrich chemical co. , usa and was used without further purification .
the ir spectra of the molecule in liquid form was recorded to be in the range of @xmath0 @xmath1 using perkin elmer ft - ir 2000 spectrometer with a resolution of @xmath7 @xmath1 .
the calculated stable structure of cyhea is shown in fig.[eps1 ] which was drawn by xcrysden ( crystalline structures and densities ) program @xcite . the vibrational assignments and frequencies of cyclohexene-2-ethanamine
was reported experimantally by izgi et al .
the spectral properties of the molecule were evaluated through the calculated vibrational frequencies of the free ligand molecule .
the calculated and experimental infrared spectra data of the molecule are given in table.[table1 ] .
the experimental , gga and lda results are also compared in fig.[eps2 ] .
[ table1 ] [ cols="^,^,^,^,^,^",options="header " , ] the strong n - h asymmetric and symmetric stretch bands seen in table.[table1 ] are due to the contribution of ethylamine ( see fig.[eps4 ] ) .
c - h stretch bands between @xmath8 @xmath1 are attributed to cyclohexene group and the very strong c - h stretch bands at @xmath9 @xmath1 and @xmath10
@xmath1 result from ethylamine .
the very strong bands are attributed to the attachment of ethylamine and cyclohexene and appear between @xmath11 @xmath1 .
most of the modes below the @xmath12 @xmath1 arise from cyclohexen .
if the vibrational assignments of the molecule involving these groups are investigated , it is seen that the assignments obtained for the molecule also involve the group frequencies .
furthermore , the observed medium broad band appears at @xmath13 @xmath1 is an n - h bending band as well as a group frequency .
there is also a good agreement between the experimental and the theoretical vibrational frequencies in the region of @xmath0 @xmath1 except some gga and lda results .
the ground state energy of the molecule was obtained to be -128.66 ryd and -128.53 ryd for gga and lda , respectively .
finally , we examined the electronic properties by using calculated stm images for cyclohexene-2-ethanamine . in fig.[eps4 ] and fig.[eps5 ] which were drawn by using xcrysden , we calculated the stm images at constant current and bias voltage -2.5 ev and 2.5 ev , respectively .
these results supply a microscopic model for stm images and can serve as a source for stm experiments for organic molecules .
the experimental and the theoretical investigation of cyhea molecule have been performed successfully by using ft - ir and density functional theory calculations . for all calculations
, it is shown that the results of gga and lda methods are in excellent agreement with all experimental findings .
thus , density functional theory ( dft ) methods are suitable for the calculation of ground state properties and potential energies .
hence , dft is an excellent method for calculating vibrational spectra and stm images from first principles .
we thank g.gokoglu and t. boz for improving our paper english .
18 f. hibbert , acc .
17 * , 115 ( 1984 ) .
h. a. staab , t. saupe , angew . chem . * 100 * , 895 ( 1988 ) ; angew .
* 27 * , 865 ( 1988 ) .
alder , chem . rev . * 89 * , 1215 ( 1989 ) .
r.w . alder , tetrahedron * 46 * , 683 ( 1990 ) .
s. rodin - bercion , l. lespade , d. cavagnat , j.c .
cornut , j. mol .
( theochem ) * 526 * , 343 ( 2000 ) .
d. zeroka , j.o . jensen and a.c .
samuels , j. mol .
( theochem ) * 465 * , 119 ( 1999 ) .
s.r . sirimanne and s.w .
may , j. am .
soc . * 110 * , 7560 ( 1988 ) .
t. izgi , . alver , c. parlak , m.t .
aytekin and m. enyel , spectrochemica acta a : molecular and biomolecular spectroscopy , accepted ( 2006 ) .
a. kokalj , comp .
* 28 * , 155 ( 2003),www.xcrysden.org/. s. baroni , a. dal corso , s. de gironcoli and p. giannozzi ; http:// www.sissa.it/cm/pwcodes .
n.c . handy , p.e .
maslen , r.d .
amos , j.s .
andrews , c.w .
murray and g.j .
laming , chem .
197 * , 506 ( 1992 ) .
n.c . handy , c.w .
murray and r.d .
amos , j. phys .
* 97 * , 4392 ( 1993 ) .
stephens , f.j .
devlin , c.f .
chabalowski and m.j .
frisch , j. phys .
chem . * 98 * , 11623 ( 1994 ) .
devlin , j.w .
finley , p.j .
stephens and m.j .
frisch , j. phys .
chem . * 99 * , 16883 ( 1995 ) .
lee , b.h .
boo , bull .
korean chem .
* 17 * , 754 ( 1996 ) .
lee , b.h .
boo , bull .
korean chem .
* 17 * , 760 ( 1996 ) .
d. vanderbilt , phys .
b * 41 * , 7892 ( 1990 ) .
perdew and a. zunger , phys .
b * 23 * , 5048 ( 1981 ) .
perdew , k. burke and m. ernzerhof , phys .
* 77 * , 3865 ( 1996 ) . | we have investigated the structure of cyclohexene-2-ethanamine molecule both theoretically and experimantally .
theoretical investigation is based on a first principle technique density functional theory ( dft ) using plane wave basis sets and ultrasoft pseudo - potentials while the experimental technique is infrared ( ir ) spectroscopy .
exchange - correlation potential of dft was approximated in the frame of both local density approximation ( lda ) and generalized gradient approximation ( gga ) schemes .
vibrational properties of this molecule are given by the assignments in the range for wavenumber @xmath0 @xmath1 .
stable equilibrium structure of the molecule was also obtained by using lda and gga .
obtained optimized geometrical structure was used to calculate vibrational properties and stm images .
a remarkable agreement was obtained between theory and experiment , especially in the symmetric and asymmetric vibrations of nh groups . _
keywords : _
cyclohexene-2-ethanamine , first principle , ir spectrum , stm images . |
biological systems are composed of molecular components and the interactions between these components are of an intrinsically stochastic nature . at the same time , living cells perform their tasks reliably , which leads to the question how reliability of a regulatory system can be ensured despite the omnipresent molecular fluctuations in its biochemical interactions .
previously , this question has been investigated mainly on the single gene or molecule species level .
in particular , different mechanisms of noise attenuation and control have been explored , such as the relation of gene activity changes , transcription and translation efficiency , or gene redundancy @xcite .
apart from these mechanisms acting on the level of the individual biochemical reactions , also features of the circuitry of the reaction networks can be identified which aid robust functioning @xcite .
a prime example of such a qualitative feature that leads to an increased stability of a gene s expression level despite fluctuations of the reactants is negative autoregulation @xcite . at a higher level of organization
, the specifics of the linking patterns among groups of genes or proteins can also contribute to the overall robustness .
in comparative computational studies of several different organisms , it has been shown that among those topologies that produce the desired functional behavior only a small number also displays high robustness against parameter variations . indeed , the experimentally observed networks rank high among these robust topologies @xcite . however , these models are based on the deterministic dynamics of differential equations .
modeling of the intrinsic noise associated with the various processes in the network requires an inherently stochastic modeling framework , such as stochastic differential equations or a master equation approach @xcite .
these complex modeling schemes need a large number of parameters such as binding constants and reaction rates and can only be conducted for well - known systems or simple engineered circuits . for generic investigations of such systems ,
coarse - grained modeling schemes have been devised that focus on network features instead of the specifics of the reactions involved @xcite . to incorporate the effects of molecular fluctuations into discrete models ,
a commonly used approach is to allow random flips of the node states .
several biological networks have been investigated in this framework and a robust functioning of the core topologies has been identified @xcite . however , for biological systems , the perturbation by node state flips appears to be a quite harsh form of noise : in real organisms , concentrations and timings fluctuate , while the qualitative state of a gene is often quite stable . a more realistic form of fluctuations than macroscopic ( state flip ) noise should allow for microscopic fluctuations .
this can be implemented in terms of fluctuating timing of switching events @xcite .
the principle idea is to allow for fluctuations of event times and test whether the dynamical behavior of a given network stays ordered despite these fluctuations . in this work
we want to focus on the reliability criterion that has been used to show the robustness of the yeast cell - cycle dynamics against timing perturbations @xcite and investigate the interplay of topological structure and dynamical robustness . using small genetic circuits we explore the concept of reliability and discuss design principles of reliable networks .
however , biological networks have not been engineered with these principles in mind , but instead have emerged from evolutionary procedures .
we want to investigate whether an evolutionary procedure can account for reliability of network dynamics .
a number of studies has focused on the question of evolution towards robustness @xcite .
however , the evolution of reliability against timing fluctuations has not been investigated .
first indications that network architecture can be evolved to display reliable dynamics despite fluctuating transmission times has been obtained in a first study in @xcite . using a deterministic criterion for reliable functioning , introduced in @xcite
, it was found that small networks can be rapidly evolved towards fully reliable attractor landscapes .
also , if a given ( unreliable ) attractor is chosen as the `` correct '' system behavior , it was shown that with a high probability a simple network evolution is able to find a network that reproduces this attractor reliably , i.e. in the presence of noise . here
, we want to use a more biologically plausible definition of timing noise to investigate whether a network evolution procedure can generate robust networks .
we focus on the question whether a predefined network behavior can be implemented in a reliable way , just utilizing mutations of the network structure .
we use a simple dynamical rule to obtain the genes activity states , such that the dynamical behavior of the system is completely determined by the wiring of the network .
a widely accepted computational description of molecular biological systems uses chemical master equations and simulation of trajectories by explicit stochastic modeling , e.g. through the gillespie algorithm @xcite .
however , this method needs a large number of parameters to completely describe the system dynamics .
thus , for gaining qualititative insights into the dynamics of genetic regulatory system it has proven useful to apply strongly coarse - grained models @xcite .
boolean networks , first introduced by kauffman @xcite have emerged as a successful tool for qualitative dynamical modeling and have been successfully employed in models of regulatory circuits in various organisms such as _ d. melanogaster _
@xcite , _ s. cerevisiae _
a. thaliana _
@xcite , and recently _ s. pombe _ @xcite . in this class of dynamical models , genes , proteins and mrna
are modeled as discrete switches which assume one of only two possible states .
here , the active state represents a gene being transcribed or molecular concentrations ( of mrna or proteins ) above a certain threshold level .
thus , at this level , a regulatory network is modeled as a simple network of switches .
time is modeled in discrete steps and the state of all nodes is updated at the same time depending only on the state of all nodes at the previous time step according to the wiring of the network and the given boolean function .
when such a system is initialized with some given set of node states , it will in general follow a series of state changes until it reaches a configuration that has been visited before ( finite number of states ) .
because of the deterministic nature of the dynamics , the system has then entered a limit cycle and repeats the same sequence of states indefinitely ( or keeps the same state in the case of a fixed point attractor ) . in the original boolean model
there are two assumptions that are clearly non - biological and are thus often criticized : the discrete time which implies total synchrony of all components ; and the binary node states which prohibit intermediate levels and gradual effects
. there have been various attempts at loosening these assumptions while keeping the simplicity of the boolean models .
it is a clear advantage of boolean models that they operate on a finite state space .
the synchronous timing , however , does not hold a similar advantage apart from computational simplicity .
models that overcome this synchronous updating scheme have been suggested in a variety of forms . in @xcite
different asynchronous schemes are used in the model of the fruit fly .
the simplest asynchronous model keeps the discrete notion of time but lets events happen sequentially instead of simultaneously .
a continuous - time generalization of boolean models that is inspired by differential equation models has been suggested in @xcite . here , the discreteness of the node states is kept but the dynamics take place in a continuous time . in @xcite the limit of infinitesimally small disturbances from synchronous behavior is investigated .
this concept of allowing variations from the synchronous behavior will also be used in this work .
the principle idea is to use a continuous time description and identify the state of the nodes at certain times with the discrete time steps of the synchronous description @xcite .
further , an internal continuous variable is introduced for every node and the binary value of the node is obtained from this continuous variable using a threshold function .
now a differential equation can be formulated for the continuous variable . and
the corresponding activity state @xmath0 ( @xmath1 , @xmath2).,title="fig:",width=151 ] + and the corresponding activity state @xmath0 ( @xmath1 , @xmath2).,title="fig:",width=340 ] this is pictured in figure [ fcharging ] . here
the internal dynamics and the resulting activity state of a node with just one input are shown for a given input pattern .
the activator @xmath3 of the node @xmath4 is switched on ( say , externally ) at time @xmath5 and stays on until it is switched off at time @xmath6 . in the boolean description
we would say node @xmath3 assumes state @xmath7 at time step 1 and at time step 2 switches to state @xmath8 .
node @xmath4 would react by switching to state @xmath9 at step 2 and to @xmath10 at step @xmath11 . in the continuous version ,
we implement this by a delay time and a `` charging '' behavior of the concentration value of node @xmath4 , driven by the input variable @xmath12 .
as soon as @xmath13 crosses the threshold of @xmath14 , the activity state of @xmath4 switches to @xmath9 .
let us formulate the time evolution of a system of such model genes by the set of delay differential equations @xmath15 here , @xmath16 denotes the transmission function of node @xmath17 and describes the effect of all inputs of node @xmath17 at the current time .
the parameter @xmath18 sets the time scale of the production or decay process .
in general , any boolean functions can be used as transmission function @xmath19 . for simplicity , we choose threshold functions , which have proven useful for the modeling of real regulatory networks @xcite .
let us use the following transmission function @xmath20 where @xmath21 is the transmission delay time that comprises the time taken by processes such as translation or diffusion that cause the concentration buildup of one protein to not immediately affect other proteins .
the interaction weight @xmath22 determines the effect that protein @xmath23 has on protein @xmath17 .
an activating interaction is described by @xmath24 , inhibition by @xmath25 . if the presence of protein @xmath23 does not affect expression of protein @xmath17 , @xmath26 .
the discrete state variable @xmath27 is determined by the continuous concentration variable @xmath28 via a heaviside function @xmath29 $ ] .
the threshold value @xmath30 is given by @xmath31 ( this choice is equivalent to the commonly used threshold value of 0 if the activity states are given by @xmath32 instead of the boolean values used here ) . for the simple transmission function given above , equation ( [ eq : ode ] )
can be easily solved piecewise ( for every time span of constant transmission function ) , leading to the following buildup or decay behavior of the concentration levels @xmath33 this has the effect of a low - pass filter , i.e. a signal has to sustain for a while to affect the discrete activity state . a signal spike , on the other hand ,
will be filtered out . up to now
we have only introduced a continuous , but still deterministic generalization of the synchronous boolean model .
if one now allows noise on the timing delay , the model becomes stochastic and asynchronous . the way we model this stochastic timing is via a signal mechanism .
as soon as one node flips its discrete state at , say , @xmath34 , it sends a signal to each node it regulates .
this signal affects the input of a regulated node at a later time @xmath35 where @xmath36 is a uniformly distributed random number between @xmath37 and @xmath38 .
the random number @xmath36 is chosen for each signal and each link independently , which means that a switching node will affect two regulated nodes at slightly different times .
due to the timing perturbations , the network states at exactly integer time do not hold a special significance any more . to overcome this problem
, we define a new macro step whenever all discrete node states ( not the concentration levels ) are constant for at least a time span of @xmath39 , which amounts to one discrete time step in the synchronous model .
only the system states at these times of extended rest are used in the comparison with the synchronous behavior . this way ,
small fluctuations of the signal events are tolerated , but extended times of inactivity of the system must exist and the state of the network at these times must correspond to the respective states under synchronous dynamics .
we call network dynamics `` reliable '' if , despite the stochastic effects on the signal transmission times , the network follows the synchronous state sequence . although fluctuations in the exact timing are omnipresent , ordered behavior of the sequence of states can still be realized
. an exact definition of the algorithm can be found in the appendix . in @xcite a similar model , but with infinitesimal timing perturbations , was used to identify those attractors in random boolean networks that are reliable .
most attractors in fact are unreliable , but are irrelevant for the system because of very small basins of attraction .
further , it was shown that the number of reliable attractors scales sublinearly with system size , which reconciles the behavior of boolean dynamics with the numbers of observed cell types as was originally proposed in @xcite .
a similar result was also obtained in a sequential updating scheme @xcite .
in this section we want to discuss the principle differences behind reliable and unreliable dynamics and show that the same dynamical sequence can be achieved both in a reliable or in an unreliable fashion , depending on the underlying network that drives the dynamics . , the production time constant , @xmath18 and the noise level @xmath38 are chosen for optimal readability of the figures .
we stress that all our conclusions also hold for the parameter choice from the results part or for any variation of these values within reasonable bounds . ]
we start with the well - known example of two mutually activating genes and model the system according to equations ( [ eq : ode ] ) ( [ solution ] ) .
apart from the trivial fixed points ( both on or both off ) this system displays an unreliable attractor as shown in the left panel of figure [ reliableunreliablecomp ] . in the upper part
, the synchronous boolean attractor is depicted in a simple pictorial form ( black means active , white inactive ) .
below that the continuous variable of both nodes is plotted over time in an example run and it can be seen that because of desynchronization the system can exit the synchronous state sequence .
changing just one link and thus creating an inhibiting self - interaction at the first gene ( see right panel of figure [ reliableunreliablecomp ] ) , the dynamics is now driven by this one node loop .
the synchronous sequence of the attractor is still the same , but now the fixed points of the old network are no longer fixed points but transient states to this attractor . the asynchronous dynamics ,
as shown in the lower part of figure [ reliableunreliablecomp ] now display an ordered behavior that would continue indefinitely . the essential feature that causes these stable oscillations
is the time delay involved . without a time delay , the system would not exhibit stable oscillations in either case but would assume intermediate levels for both nodes .
thus , a direct comparison of these dynamics with a stability analysis of ordinary differential equations without delay is not adequate .
next , we now want to test the reliability of examples of circuits that can be created artificially .
the repressilator is a simple artificially generated genetic circuit implemented in _
e. coli _ @xcite .
consisting of of three genes inhibiting each other in a ring topology ( see upper part of figure [ frep ] ) , this system displays stable oscillations . describing this system using differential equations
it was found that the unique steady state is unstable for certain parameter values and that numerical integration of the differential equations displays oscillatory behavior . also in a stochastic modeling scheme , sustained but irregular oscillations can be observed , which show some resemblance of the experimental time series @xcite . to discuss this model system in our framework
, the synchronous boolean description has to be analyzed first . here
, the three - gene repressilator exhibits two attractors which comprise all eight network states the `` all - active - all - inactive '' ( two states ) and the `` signal - is - running - around '' pattern ( six states ) . in the asynchronous scheme , independently of the initial conditions , the system reaches the second attractor .
once the attractor is reached , the system stays in it forever ( i.e. is reliable in our definition ) see figure [ frep ] .
this is due to the fact that only a single switching happens at a given time .
this is depicted in the lower part of figure [ frep ] by the arrows which are successively active , no two events happening at the same time . , title="fig:",width=151 ] , title="fig:",width=340 ] this picture changes in the case of the four - genes repressilator .
the attractor structure is now much more involved , consisting of the two fixed points ( 1010 ) and ( 0101 ) and three attractors with four states each . using the stochastic scheme as before , we find that only the fixed points emerge as reliable attractors of the system .
if any state of one of the 4-cycles is prepared as initial condition , the system thus always ends up in one of the two fixed points . in figure [ f4rep ] ,
an example run is shown which is initialized with the state ( 1100 ) . without any noise ,
the system would follow a four - state sequence consisting of all states where the two active nodes are adjacent and the two inactive nodes are as well . however , if a small perturbation is allowed , the system can exit this attractor as shown in the lower panel of [ f4rep ] . here ,
we have drawn two arrows showing two causal events happening at the same time .
in fact , there are two independent causal chains in the system dynamics .
if these two chains fluctuate in phase relative to each other , they can extinguish each other and drive the system into a fixed point . for an extended discussion of the concept of multiple causal chains ,
see @xcite . apart from the concept explained in the last section , when a system can accumulate a phase lag through small perturbations in a random walk - like fashion and eventually ends up in a different attractor , there are also examples of systems , in which any small perturbation drives the system away from the current attractor .
this is the case if the concentration variables do not reach their maximal value before their input nodes switch their state .
this can cause the following node to switch even earlier as compared to the unperturbed timing ( in the limit @xmath40 , this behavior is impossible and each perturbation that is not removed from the system is merely `` neutral '' ) .
we show this behavior in figure [ frepunstable ] . here again , the four repressilator is shown , but with the initial state configuration 0000 . without noise
, this state belongs to the `` all - active - all - inactive '' attractor and four independent events are happening at each time step .
the small stochastic asynchrony in the beginning is amplified and leads to a quick loss of the attractor .
the system then enters an intermediate attractor where the neutral perturbation behavior is predominant , because the concentration levels have more time to approach their saturation value .
the opposite behavior is also possible , that the system itself prevents divergence of the phases .
this can happen if the intermediate system state creates a signal spike ( i.e. a short - term status change of a node ) that itself feeds back to the causal chain . even though the causal chains are independent in synchronous mode , they can be connected through such intermediate states .
we want to stress that in the criterion employed in @xcite it can not be identified whether an attractor is marginally stable or exhibits such a self - catching behavior .
this is a limit of the deterministic criterion that is overcome by the explicit modeling used here . in this work
, we consider all attractors as `` unreliable '' that can desynchronize so strongly that the system does not maintain a `` rest phase '' in which no switching events occur for an extended time .
this includes all marginally stable as well as unstable attractors .
we do not distinguish between these in our results as both do not seem suitable for the reliability of a biological system .
now that we have introduced the main concepts and ideas surrounding our definition of reliability , we want to turn to the question , whether such a simple model of regulatory networks can be evolved towards realizations displaying reliable dynamics . for this question ,
we define the notion of a `` functional attractor '' . as we are dealing with random networks , we need a measure of what the system is supposed to do .
thus , we choose one attractor of the starting network as the prototype dynamics that define the desirable dynamical sequence . the functional attractor is determined by running the synchronous model with a randomly chosen initial state until an attractor is found . during the evolution process
we demand each network to reproduce this attractor .
this prescription introduces a bias towards attractors with large basins of attraction .
however , as the basin of attraction is commonly understood as a measure of the significance of an attractor , this appears to be a natural choice .
only unreliable attractors are used as functional attractors , as in the case of reliable attractors , the evolution goal would be achieved before the start of the evolution .
the evolution procedure is chosen as a simple version of a mutation and selection process .
we start by creating a directed random network with the prescribed number of links @xmath41 ( self - links are allowed ) and determine the functional attractor . during the evolution procedure ,
we mutate the current network by a single rewiring of a link , that is , removal of one link and simultaneous addition of a random link between two nodes that are not yet connected .
this actually amounts to two operations on the graph , but has the advantage that the connectivity @xmath42 is kept fixed .
the fitness of a given network is assessed by comparison of the asynchronous dynamics with the synchronous functional attractor .
the initial network state is set according to one randomly chosen state of the synchronous attractor .
the concentration levels are initialized to the same value ( either @xmath43 or @xmath44 ) .
now the system dynamics is explicitly simulated using the stochastic algorithm introduced above ( details given in the appendix ) .
we follow the dynamics for maximally @xmath45 macro time steps ( i.e. rest phases ) .
the fitness score is then obtained by dividing the number of steps correctly following the synchronous behavior by this maximal step number .
the algorithm used in @xcite is an abstraction of the principle that small fluctuations can not on its own drive the system out of its attractor , but only if successively adding up .
thus , in a systematic way , for any possible retardation of signal events it is checked whether it persists in the system for a full progression around the attractor .
if so , the attractor is marginally stable and can in principle lose its synchrony . while this has the advantage of being a deterministic criterion and thus leads to a noiseless fitness function , there are situations in which this criterion is sufficient but not necessary for reliable behavior . by explicitly modeling the time course , as done in this work , only the truly unstable attractors are marked as such .
we do not take into account the transient behavior of the system and define only the limit cycle as the functional attractor . as our reliability definition would be trivially fulfilled in the case of fixed points , we use networks exhibiting a limit cycle from the start . in the evolution process , the fitness of a mutant is compared to the fitness that the mother network scored . a network is only selected , if it scores higher than any other network found before during the evolution . as the dynamics is inherently stochastic , the fitness criterion is noisy , too .
thus , networks which are not more reliable than the mother network might still be selected in the evolution due to variability in the fitness score .
if a network follows the attractor up to a maximum step number , it is said to be `` reliable '' . if during the network process a given number of mutation tries
is exceeded , the evolution process is aborted . the maximal number of mutation tries in the evolution is @xmath46 at each step and @xmath47 during the full course of evolution .
we later discuss the implications of these fixed parameter settings .
we have used the following parameters in the results part .
the delay time @xmath21 is set to unity , the buildup time @xmath18 is @xmath48 .
maximal noise @xmath38 is @xmath49 .
this means , that the impact of any individual perturbation is low and can not itself cause a failure in the fitness test .
only if several perturbations consecutively drive the system away from synchronization , the requirement of an extended static period can be missed . .
in this example , three steps suffice for stabilization .
the structure of each network during the evolution is shown , with the arrows denoting the subsequent step in the evolution . in every step ,
one link is lost ( shown in grey color ) and a new link is added ( denoted by the plus sign ) .
the change of the state space of the network is given in figure [ fevoexampleattractors].,width=302 ] .
for every step , the full attractor landscape is shown .
every dot denotes a state , the subsequent state is connected via a line .
the limit cycle is shown in the center of each attractor basin .
the functional attractor is shown as the upper leftmost attractor in all steps.,width=302 ] in figures [ fevoexamplestructure ] and [ fevoexampleattractors ] we show an example of a typical evolution process for a small network of @xmath50 . during three steps , the network is evolved towards a reliable architecture . the initial network ( upper - left in figure [ fevoexamplestructure ] ) displays three synchronous attractors ( top panel in figure [ fevoexampleattractors ] ) of which the first is chosen as the functional attractor .
the structural changes are depicted in figure [ fevoexamplestructure ] by a grey arrow for the removed link and a plus - sign for the newly added link . as is typical for these evolution processes @xcite , the attractor landscape is affected dramatically during the evolution . in this example ,
only the functional attractor survives the evolution procedure .
we have performed the described network evolution for a variety of different network sizes as well as connectivities . for system sizes of @xmath51 and @xmath52 and connectivities between @xmath53 and @xmath54 the ratio of networks that were stabilized during the evolution is shown in figure [ fevolution ] .
whenever we plot the ratio of stabilized networks , we have calculated the sample errors by a poissonian error estimate , @xmath55 , where @xmath56 is the obtained ratio from @xmath57 sample runs .
one can see that for intermediate connectivities between @xmath58 and @xmath59 , the ratio of stabilized networks is above 80% for all system sizes under investigation .
this means , that starting from any random network , in four out of five cases a simple network evolution is able to find a network that displays the same dynamical attractor , but performs it reliably .
this result matches a very similar dependence on the connectivity that was found for 16 nodes in the infinitesimal scheme in @xcite .
it is interesting to note that for lower connectivities , the ratio of stabilized networks decreases significantly . for all system sizes considered
, there is a sizable decrease of the stabilization ratio for connectivities below two .
this is especially apparent for the large system size @xmath60 .
this is due to the `` essentiality '' of the structure on the dynamics .
changing a link without destroying the dynamical attractor is less likely for lower connectivities . at higher connectivities the larger number of non - essential links in the system aids evolvability towards reliable dynamics via phenotypically neutral mutations .
however , considering large connectivities and large system sizes , the ratio of stabilized networks drops again .
along with the increase of attractor lengths with system size that impairs reproducibility of dynamics .
thus , we find an area of connectivity between @xmath61 and @xmath59 for which the ratio of stabilized networks is similar for all system sizes considered .
the plot in figure [ fevo_steps ] shows the average number of rewiring steps necessary until a stable network realization is found for networks of 32 nodes .
for all connectivities , this number is remarkably low , as the evolution procedure basically implements a biased random walk through structure space .
this is due to the large variation of the fitness score of a single network . despite the rather small evolutionary pressure
, the evolution procedure quickly finds a realization exhibiting reliable dynamics .
interestingly , the number of evolution steps does not monotonically grow with the connectivity , but instead drops for connectivities larger than two .
this again is an indication that networks with higher connectivities are easier to evolve towards reliability .
the ratio of links rewired in the evolution to the total number of links is even monotonically decreasing ( not shown ) .
we want to further investigate the dependence on network size by repeating the evolution procedure with system sizes up to @xmath62 .
this is shown in figure [ fevolutionncomp ] in a log - linear plot of the ratio of stabilized networks vs. system size .
we find that the ability of the process to stabilize a given network decreases with system size .
the line in the figure represents a fit of the function @xmath63 with @xmath64 , @xmath65 thus a relatively slow decay with system size .
one also has to keep in mind that the fixed set of parameters for the number of attempted mutations per evolution step and the total number of attempted mutations during the evolution reduces the success rate for larger networks . for small networks of @xmath66
, 20000 attempted mutations per evolution step suffices for a good estimate of the space of all one - link mutations , but as the number of possible mutations scales with the system size n as @xmath67 , it quickly becomes impossible to check all possibilities .
thus , the results in figure [ fevolutionncomp ] underestimate the probability to find a stable instance .
we have checked the dependence of the results on the selection parameters ( attempted mutations per evolution step @xmath68 and total number of attempted mutations during evolution @xmath69 ) for selected network sizes and connectivities . in figure [ fevolutionncompparams ]
we again show the ratio of stabilized networks against system size , but this time for two different parameter values the original parameter set with @xmath70 ( denoted by ` + ' ) and for an increased value of @xmath71 ( denoted by ` @xmath72 ' ) . for small networks
, the value of this parameter does not significantly affect the results , but for @xmath73 , differences can be clearly seen .
for @xmath60 the ratio rises from @xmath74 at @xmath75 to @xmath76 at @xmath77 .
interestingly , for larger system sizes this effect does not seem to be amplified : for @xmath78 the ratio rises from @xmath79 to @xmath80 .
for @xmath60 we plot the dependence of the ratio of stabilized networks on the parameter @xmath68 in figure [ fevolution50nparams ] .
the largest parameter value used , @xmath77 is about twice the total number of possible rewirings and should thus suffice .
one can see that the decrease in the ratio of successfully evolved networks can be significantly reduced when attempting more mutations per evolution step .
this is due to the fact that an enormous number of mutations is possible of which only a small fraction retains the requested dynamical sequence .
still , one can deduce from these results that it is harder to stabilize large networks than smaller ones : even though there might be a path to a stable network instance , it may not be practically realizable as the chance to find exactly the right mutations may be too small .
however , real world systems display a large amount of modularity that leads to smaller cores of strongly interacting components . we have not taken this into account in our random network approach .
we see this as a model for small networks of key generators as were described in recent boolean models of biological systems @xcite .
the resulting dynamics of the full networks are then influenced by this core without strong feedback .
this allows for rather simple expression patterns of the full network without constraints on the network size .
we have discussed a simple reliability criterion for biological networks and have applied it to network design features that produce reliable dynamics .
we showed that small changes in the network topology can dramatically affect the dynamical behavior of a system and can lead to reliable network dynamics . to investigate how reliability can emerge in real - world systems that have been shaped by evolution
, we studied an evolutionary algorithm that selects networks with a prescribed dynamical behavior if they function more reliably than a given mother network .
we found that a high ratio of random networks can evolve towards instances displaying reliable dynamics . in accordance with other recent work @xcite
, it was shown that the evolution of network structures can lead to reliable dynamics both with a high probability and within short evolutionary time scales .
surprisingly , small connectivities are detrimental to this evolvability .
this is counter - intuitive as sparsely connected networks show rather simple dynamics with short attractor lengths .
however , at the same time they are difficult to evolve because they have a small structural `` buffer '' of links that can be neutrally rewired without changing the dynamics .
this is related to the concepts of `` degeneracy '' and `` distributed robustness '' where additional elements are present in a system that are not strictly necessary for the system s function but have a positive effect on robustness @xcite . here
, these additional elements are links that are not strictly necessary to perform a specific function . thus
, rewiring of these links is possible and allows for a higher probability to find a network with reliable dynamics .
we thus find in our framework that high connectivity , although leading to increasing complexity of the dynamics , can be beneficial for the evolution of networks . for larger system sizes the evolvability towards reliable dynamics decreases .
this is due to the increasing dynamical complexity of such networks ( longer attractor cycles , more non - frozen nodes ) .
our strict criterion requests the reliable reproduction of the exact state sequence for every node , which leads to a more difficult selection process for large system sizes . in summary ,
our results suggest that reliability is an evolvable trait of regulatory networks . in the present simple model , reliability
can be achieved by topological changes alone and without fine - tuning of parameters .
this means that through mutations of the reaction networks , biological systems may have the ability to rapidly acquire the property of reliable functioning in the presence of biochemical stochasticity .
the authors would like to thank maria i. davidich for discussions and helpful comments on the manuscript and fabian zhrer for help with the state - space visualization graph .
this work was supported by deutsche forschungsgemeinschaft grants bo1242/5 - 1 , bo1242/5 - 2 .
the asynchronous algorithm is implemented such that no discretized clock is needed .
only those times will be investigated , when changes in the system happen .
* @xmath81 : time of the last change of buildup / decay behavior * @xmath82 : concentration level at that time * @xmath83 : flag for current behavior - either buildup ( 1 ) or decay ( 0 ) * @xmath84 : current discrete state of node @xmath17 * @xmath85 : discrete state of node @xmath17 that would result from the current states of all nodes : @xmath86 , the system is initialized by setting all values of discrete states @xmath84 equal to the state given by the discrete initial conditions .
the concentration levels are set to the same values ( @xmath43 or @xmath44 ) .
the times of the last behavior changes @xmath81 are set to @xmath37 .
before the simulation is started , for every node @xmath17 it is checked whether the aspired state @xmath85 differs from the current state @xmath84 .
if so , an event is added to the queue q ( sorted by time ) for time @xmath88 , where @xmath36 is a uniformly distributed random number between @xmath37 and @xmath38 .
a simple analytical expression can be given for the times when the concentration levels are crossed ( case 1 ) .
if @xmath89 , the node will not switch its state because the concentration is moving away from the threshold . otherwise , one can calculate the time of the next concentration level to cross the threshold by solving equation ( [ solution ] ) for @xmath90 with @xmath91 : if an event of type 1 happens next , the discrete state of the respective node @xmath17 , @xmath84 , is updated and the effect on other nodes is calculated . for definiteness ,
let us assume this crossing takes place at time @xmath90 .
if this switch causes the aspired state of another node @xmath23 to switch , an event is sorted into the queue @xmath87 at @xmath93 .
when in the queue events for the same node are scheduled to happen at later times , they will be removed .
they are thought to have been `` caught '' by the newly added event . in the second case
, the concentration level of the node at time @xmath90 is calculated according to equation ( [ solution ] ) and saved as @xmath94 with the new time @xmath95 .
the behavior flag @xmath83 is switched to reflect that the node has changed from buildup to decay or vice versa . if the time between any two successive node state changes in the network ( not necessarily of the same node ) is larger than @xmath39 , the node states are recorded and set as a new step to be compared to the synchronous attractor . | the problem of reliability of the dynamics in biological regulatory networks is studied in the framework of a generalized boolean network model with continuous timing and noise . using well - known artificial genetic networks such as the repressilator , we discuss concepts of reliability of rhythmic attractors . in a simple evolution process
we investigate how overall network structure affects the reliability of the dynamics . in the course of the evolution ,
networks are selected for reliable dynamics .
we find that most networks can be easily evolved towards reliable functioning while preserving the original function . |
recently , the voltage control of magnetism without electric current has been extensively studied in order to achieve minimum power consumption and device miniaturization in next - generation of spintronics.@xcite such a control is usually implemented through the magnetoelectric ( me ) coupling in heterostructures which possess coupled magnetic and electric properties .
generally , in me heterostructures voltage can control the magnetism through the interfacial mechanisms such as elastic coupling via strain transfer,@xcite charge modulation,@xcite interface bonding,@xcite and exchange coupling.@xcite for the ferromagnetic / ferroelectric heterostructures , elastic coupling mechanism is most extensively investigated , i.e. a strain generated in a ferroelectric layer by a voltage is transferred to the ferromagnetic layer through the interface and thus can be used to tailor magnetization through the magnetoelastic coupling .
the elastic coupling mechanism can work at the bulk level .
all the other three mechanisms are localized to the region near the interface . among them
, the exchange coupling is localized , but can induce propagation of the magnetization into the bulk .
they offer more freedom to design reasonable and optimum nanoscale devices in which the interface plays a critical role . by using these various mechanisms ,
a voltage induced 180@xmath0 magnetization switching is of great interests .
for example , in order to achieve a high signal - to - noise ratio in magnetic tunnel junction ( mtj ) , a significantly large electric resistance change of mtj is required , which can only be achieved by a 180@xmath0 switching in the free layer of mtj .
trailblazing experiments have demonstrated strain - mediated , charge - mediated , and exchange - coupling - mediated voltage - driven 180@xmath0 switching in ni / batio@xmath3,@xcite bifeo@xmath3-based,@xcite and cofeb / mgo / cofeb@xcite heterostructures , respectively . from the theoretical point , a large number of studies are devoted to the strain - mediated 180@xmath0 switching either by designing the shape of magnets@xcite or by using the 3d precessional switching dynamics.@xcite however , since the substrate of the layered me thin heterostructures is usually clamped , the strain transfer is limited .
the realization of strain - mediated commercial devices still remains challengeable due to some practical issues such as fatigue .
alternatively , interface charge modulation has been deemed as an ideal way for realizing magnetic switching in thin film heterostructures.@xcite but most of the experimental and theoretical investigations show a charge - mediated magnetization switching between two states which are not exactly rotated by 180@xmath0 , i.e. not a 180@xmath0 switching.@xcite meanwhile , most studies deal with the thin film structure with a lateral size of several hundred nanometers to several micrometers or with a thickness of several tens nanometers . in such cases ,
the magnet volume is relatively large so that the temperature effect on the switching dynamics is often ignored,@xcite or shown to be not so strong.@xcite however , in terms of the device miniaturization , if spintronic devices are further scaled down to nanodot shape , i.e. with the size of several tens of nanometers in the lateral direction , the huge reduction of the magnet volume will intensify the temperature effect . for instance , the granular film with @xmath4-ordered fept epitaxially grown on mgo substrate , which has been considered as a promising candidate for memory devices , usually contains fept nanograins with a size of several nanometers to several tens of nanometers.@xcite if such a small - scaled fept / mgo heterostructure is utilized , the consideration of temperature effect is indispensable . in this work , we take epitaxial pt / fept / mgo heterostructures as a model system with a lateral size of several tens of nanometers . combining first - principles calculation and temperature - dependent magnetization dynamics
, we demonstrate the deterministic in - plane and perpendicular 180@xmath0 magnetization switching at room temperature .
it is anticipated that the present study provides valuable insight into the design of voltage control of both in - plane and perpendicular 180@xmath0 switching at room temperature for achieving low - power , high - speed , non - volatile , and highly compact spintronics .
[ s2 ] the material model system for first - principles calculations is shown in fig . [ f1](a ) . since the lattice parameter of bulk mgo is @xmath5 @xmath6 and that of fept is @xmath7 @xmath6 , the lattice mismatch is up to @xmath8 . in experiments ,
mgo with tunable lattice strain can be epitaxially grown at different temperature on different substrates like cu,@xcite , ge,@xcite si,@xcite etc .
so different in - plane epitaxial strains ( @xmath9 ) relative to the equilibrium lattice parameter in mgo at different fept layer number ( n ) are adopted to study their effects .
the material system for micromagnetic dynamics analysis is shown in fig . [ f1](b ) .
the fept nanomagnet is an elliptical cylinder with height @xmath10 , semimajor axis @xmath11 nm , and semiminor axis @xmath12 nm .
for such a small size , a single domain exists@xcite and two angles are used to describe the magnetization state . the model and methodology is presented in the methods section . and @xmath13 .
( a ) electrostatic potential distribution .
( b ) electric field ( @xmath140.9 v / nm ) induced charge density change near pt and fe atoms close to the mgo / fept interface .
( c ) mae as a function of the external electric field .
( d ) atom - resolved mae.,width=317 ] and @xmath1 when the fept layer number is ( a ) @xmath15 , ( b ) @xmath16 , ( c ) @xmath17 , and ( d ) @xmath18.,width=317 ] figure [ f2 ] presents the first - principles results when @xmath13 and @xmath15 . from the electrostatic potential distribution in fig .
[ f2](a ) , we can estimate the external electric field as 0.9 v / nm . under this electric field
, the charge density around pt and fe atoms which are close to the interface is evidently decreased .
the effect of electric field extends to the first two layers of fept next to the interface
. the charge change will affect the d - orbital hybridization and thus the magnetic anisotropy energy ( mae ) , as shown the orbital - resolved mae in fig .
s1 ( in supplementary information , si ) .
the atom - resolved mae analysis in fig .
[ f2](d ) indicates the dominative role of the interfacial pt ( pt1 ) in the perpendicular magnetic anisotropy .
the electric field induced change in mae of pt1 controls the mae change of the system .
orbital - resolved mae analysis in fig .
s1 ( in si ) further confirms the largest mae contribution of the hybridization between @xmath19 and @xmath20 of pt1 .
mae of the system is shown to be linearly dependent on the strength of the electric fields , as presented in fig . [ f2](c ) .
a giant modulation of mae by applying electric fields is observed , with a amplitude as high as 322 fj / vm .
we then calculate @xmath2 ( mae divided by fept volume ) under different @xmath9 and n , as shown in fig .
it can be seen that @xmath2 varies linearly with respect to both @xmath9 and @xmath21 .
@xmath2 is more sensitive to @xmath9 and can be altered from @xmath22 to @xmath23 mj / m@xmath3 when @xmath9 is increased from @xmath24 to @xmath25 .
the strong dependence of @xmath2 on @xmath9 intrigues extensive studies of the strain - mediated magnetization switching .
positive electric field results in an increment in @xmath2 .
the charge - mediated dependence of @xmath2 on @xmath1 provides the feasibility to switch the magnetization .
for example , using the model in fig .
[ f1](b ) , we find that through the charge - mediated @xmath2 change , the electric field can be used to tailor the magnetic hysteresis ( fig . s2 in si ) . for the purely voltage - driven 180@xmath0 switching , we firstly clarify the switching conditions , i.e. how large @xmath9 and @xmath1 make the 180@xmath0 switching possible , through the total energy analysis ( eq . [ eq1 ] in methods ) . in the case of an in - plane 180@xmath0 switching ,
the magnetic easy axis is along the semimajor axis when there is no electric field .
the in - plane 180@xmath0 switching means the magnetization switching from the positive semimajor axis to the negative one , or vice versa .
these two equivalent in - plane easy axes guarantee a non - volatile effect , i.e. when the 180@xmath0 switching is achieved the associated magnetization state will be kept even after the removal of the electric field . the necessary conditions for
the 180@xmath0 switching imply two criteria .
firstly , @xmath2 is small enough to ensure a magnetic easy plane when there is no electric field , i.e. @xmath26 ( eq [ eq1 ] ) in methods ) .
secondly , after applying a positive electric field , @xmath2 is increased to be high enough to achieve a perpendicular easy axis , i.e. @xmath27 . using the criterion @xmath28 , we can estimate the conditions for the possible 180@xmath0 switching , as shown in fig . [ f4](a ) . it can be found that a positive electric field and a small range of @xmath9 potentiate an in - plane 180@xmath0 switching . since more negative @xmath9 already gives rise to larger k at zero electric field ( fig .
[ f3 ] ) , a small increase in k is enough to overcome the in - plane shape anisotropy energy and thus a small positive electric field is sufficient .
this leads to the lower - triangular shape in fig . [
f4](a ) . on the contrary
, the perpendicular 180@xmath0 switching denotes the magnetization switching from the positive perpendicular axis to the negative one , or vice versa .
two criteria required by the necessary conditions for the perpendicular 180@xmath0 switching are opposite to those for the in - plane 180@xmath0 switching . on one hand , without electric field @xmath2 should be large enough to ensure a perpendicular magnetic easy axis .
on the other hand , after a negative electric field is applied , @xmath2 should be decreased to be small enough to ensure an in - plane easy axis .
following the similar method for fig .
[ f4](a ) , the associated necessary conditions are shown in fig . [ f4](b ) .
in contrast to the results in fig .
[ f4](a ) , a negative electric field potentiates a perpendicular 180@xmath0 switching and an upper - triangular shape appears in fig .
[ f4](b ) . ) and pulse width ( @xmath29 ) .
spatial and temporal trajectories of magnetization switching for ( b ) in - plane and ( d ) perpendicular 180@xmath0 switching .
switching time as functions of @xmath30 and @xmath29 for ( c ) in - plane and ( e ) perpendicular 180@xmath0 switching .
the grey region or a switching time of zero in ( c ) and ( e ) means that the 180@xmath0 switching fails .
temperature is set as 0 k.,width=317 ] it should be noted that fig .
[ f4 ] only presents the necessary rather than the sufficient conditions for a 180@xmath0 switching .
if the electric field is always kept until the magnetization reaches the equilibrium state , only a 90@xmath0 switching is achieved . after removing the electric field , in the viewpoint of static analysis
the magnetization possesses the equal possibility to revert back or go forwards . only by using the 3d precessional switching dynamics of magnetization,@xcite a deterministic 180@xmath0 switching can be achieved when the pulse width of the electric field is controlled . in the following ,
we will focus on the switching dynamics in the case of @xmath18 and @xmath31 for an in - plane 180@xmath0 switching and @xmath32 for a perpendicular 180@xmath0 switching .
an electric field pulse with ramp time ( @xmath30 ) and pulse width ( @xmath29 ) shown in fig .
[ f5](a ) is applied to provoke switching . as a first step , we studied the typical switching dynamics at 0 k , as presented in fig .
[ f5](b)-(e ) .
the spatial and temporal magnetization trajectories in fig .
[ f5](b ) and ( d ) indicate an in - plane and perpendicular 180@xmath0 switching , respectively .
it can be seen that the precessional dynamics after the removal of @xmath1 at @xmath29 make the magnetization component towards @xmath33 possible .
the control of @xmath30 and @xmath29 is critical for realizing the 180@xmath0 switching , as shown in fig . [
f5](c ) and ( e ) .
it can be found that in both the in - plane and perpendicular 180@xmath0 switching , a fast switching within @xmath342.5 ns can be attained .
the ramp time plays a role in determining whether the 180@xmath0 switching fails or succeeds , but has little effects on the switching time which is more influenced by the pulse width .
the in - plane 180@xmath0 switching allows a much wider range of electric field pulse width than the perpendicular one .
these features of the switching dynamics are consistent with those from the literatures which also sets the temperature as 0 k. 300 k for an in - plane 180o switching ( @xmath140.6 v / nm , @xmath350.5 ns , @xmath360 ) : ( a ) 180@xmath0 switching fails ; ( b ) 180@xmath0 switching succeeds.,width=264 ] however , if the finite temperature is considered , the switching dynamics can be intrinsically altered .
thereby the thermal fluctuation of torque should be considered , as explained in the methods .
here we do not perform the temperature - dependent first - principles calculations , but focus on the temperature effect in magnetization dynamics . even in the equilibrium states without electric field ,
the magnetization is not exactly aligned along the easy axis .
the finite temperature effect makes the magnetization fluctuate within several degrees around the easy axis ( fig .
s3 in si ) .
figure [ f6 ] presents two individual magnetization trajectories at @xmath37 k , with other conditions the same as those in fig .
[ f5](a ) .
it is clear that fig .
[ f5](a ) without temperature effect shows a deterministic 180@xmath0 switching .
but if @xmath37 k is taken into account , the 180@xmath0 switching can either fail ( fig .
[ f6](a ) ) or succeed ( fig .
[ f6](b ) ) .
it indicates that the temperature effect disturbs the switching behavior and makes the 180@xmath0 switching as probability events .
the deterministic switching demonstrated at 0 k may be not really deterministic at finite temperature , making the previous studies of the 180@xmath0 switching at 0 k to be re - examined .
when the temperature is involved in the switching dynamics , statistical methodology should be applied to get the 180@xmath0 switching probability and the average switching time .
only the switching probability close to @xmath38 guarantees a deterministic switching at @xmath37 k. figures [ f7 ] and [ f8 ] compare the switching behaviours at 0 k with those at 300 k. it can be found from fig . [ f7](a ) and fig .
[ f8](a ) that at 0 k , the 180@xmath0 switching is deterministic and possesses a characteristic switching time .
large electric fields and small pulse width favour a fast switching , with a minimum switching time around 2 ns .
in contrast , at 300 k the switching is not deterministic and a switching probability is obtained , as shown in fig .
[ f7](b ) , [ f7](c ) and fig .
[ f8](b ) . for the in - plane
180@xmath0 switching , fig .
[ f7](b ) shows that the room temperature makes the 180@xmath0 switching only with a maximum probability of @xmath39 and thus a nondeterministic switching .
if we apply a small bias magnetic field @xmath40 mt which is the strength of a typical refrigerator magnet , a switching probability of @xmath41 can be achieved ( fig . [ f7](c ) ) . for the perpendicular 180@xmath0 switching in fig .
[ f8 ] , we observe the similar behaviour .
the only difference is that a @xmath41 switching probability can be achieved without any bias magnetic field , as shown in fig . [
the minimum average switching time in fig .
[ f7](d ) and fig .
[ f8](c ) is close to that in the case of 0 k. it suggests that the room temperature does not dramatically alter the average switching time .
the region with @xmath41 probability is very narrow in fig . [
f7](c ) and fig .
[ f8](b ) , indicating that the magnitude and the pulse width of the electric fields should be carefully designed for a deterministic 180@xmath0 switching at room temperature . in summary
the voltage - driven charge - mediated perpendicular and in - plane 180@xmath0 magnetization switching at 0 k and room temperature is studied by using a multiscale theoretical framework which combines first - principles calculations and temperature - dependent magnetization dynamics . for the epitaxial metal - magnet - insulator ( pt / fept / mgo ) hetero - nanostructure as the model system , it is found from first - principles calculations that the interfacial charges induced by electric fields induce a giant modulation of mae of the nanomagnet .
@xmath2 is revealed to vary linearly with respect to @xmath1 and the epitaxial strain . from the temperature - dependent magnetization dynamics using first - principles results
, it is found that both in - plane and perpendicular 180@xmath0 magnetization switching is possible in the case of suitable epitaxial strain , @xmath1 pulse width , and @xmath1 ramp rate .
but the temperature effect disturbs the switching behavior and makes the 180@xmath0 switching as probability events .
the e magnitude and pulse width should be carefully designed for a deterministic 180@xmath0 switching at room temperature .
statistical analysis indicates that a fast ( around 2 ns ) and deterministic 180@xmath0 switching can be achieved at room temperature .
this work not only demonstrates a charge - mediated way for controlling magnetization by voltage , but also inspires the rational design of miniaturized nanoscale spintronic devices where finite temperature effect plays a critical role .
the magnetic anisotropy energy ( mae ) and magnetization ( @xmath42 ) of the system in fig .
[ f1](a ) are mainly originated from the @xmath4 ordered fept layer .
the supercell is constructed along the ( 001 ) direction , containing @xmath21-layer fept on top of four - layer mgo followed by three - layer pt and a 15-@xmath6-thick vacuum layer .
the fept layer number @xmath21 is chosen to be 5 , 7 , 9 , and 11 .
the electric field is introduced by the dipole layer method,@xcite with the dipole placed in the middle of the vacuum region .
the first - principles calculations were carried out within the density functional theory and the framework of the projector augmented - wave formalism as implemented in the vienna ab initio simulation package ( vasp).@xcite the perdew - burke - ernzerhof ( pbe ) exchange - correlation functional in the generalized gradient approximation ( gga ) was employed .
an energy cutoff of 500 ev and a monkhorst - pack k - mesh @xmath43 were utilized .
the convergence criteria for the structure relaxation at different @xmath9 were set as @xmath44 ev and 2 mev/@xmath6 for the energies and forces , respectively . by using the self - consistent charge density ,
non - self - consistent calculations with spin - orbit coupling were performed to get the total energy as a function of the orientation of the quantization axis .
k was evaluated as the difference of the total energy per unit fept volume when the magnetization was along ( 100 ) and ( 001 ) directions .
positive and negative mae indicates perpendicular and in - plane magnetic anisotropy , respectively . the material system for micromagnetic dynamics analysis is shown in fig .
the fept nanomagnet is an elliptical cylinder with height @xmath10 , semimajor axis @xmath11 nm , and semiminor axis @xmath12 nm .
for such a small size , a single domain exists@xcite and two angles ( @xmath45 ) are used to described the magnetization state .
the total energy of the fept elliptical cylinder can be given as the summation of the demagnetization energy and the magnetocrystalline anisotropy energy , i.e. @xmath46 \end{split } \label{eq1}\ ] ] in which @xmath47 , @xmath48 , and @xmath49 is the demagnetization factor along @xmath50 , @xmath51 , and @xmath52 direction , respectively , and can be calculated as a function of the geometry size.@xcite taking the temperature effect as thermal fluctuations,@xcite the temperature - dependent magnetization dynamics is governed by @xmath53 @xmath54 in which @xmath55 is the gyromagnetic ratio constant , @xmath56 is the damping coefficient,@xcite @xmath57 ps is the time step , and @xmath58 ( @xmath591 , 2 ) is a stochastic process with gaussian distribution , zero mean value , and completely uncorrelated property in time .
the characteristic time @xmath60 is related to volume @xmath61 and temperature @xmath62 as @xmath63 $ ] .
the financial supports from the german federal state of hessen through its excellence programme loewe response and the german science foundation ( individual project xu 121/7 - 1 and the project xu 121/4 - 2 in the forscher gruppe for1509 ) are appreciated .
the authors also greatly acknowledge the access to the lichtenberg high performance computer of tu darmstadt .
m. y. performed the calculations , analyzed the results , and wrote the manuscript .
h. z. and b .- x .
x. analyzed the results and supervised the project .
all authors reviewed the manuscript .
the authors declare no competing interests .
57 hu , j. m. , chen , l. q. @xmath64 nan , c. w. multiferroic heterostructures integrating ferroelectric and magnetic materials .
mater . _ * 28 * , 15 - 39 ( 2016 ) .
tiercelin , n. , dusch , y. , giordano , s. , klimov , a. , preobrazhensky , v. @xmath64 pernod , p. , in nanomagnetic and spintronic devices for energy - efficient memory and computing , eds .
j. atulasimha and s. bandyopadhyay , john wiley @xmath64 sons , chichester , uk2016 , ch . 8 , pp .
221 - 257 .
liu , m. @xmath64 sun , n. x. voltage control of magnetism in multiferroic heterostructures .
r. soc . a _ * 372 * , 20120439 ( 2014 ) .
m . , nan , t. , sun , n. x. @xmath64 chen , l .- q .
multiferroic magnetoelectric nanostructures for novel device applications .
_ mrs bull .
_ * 40 * , 728 - 735 ( 2015 ) .
wang , y. , hu , j. , lin , y. @xmath64 nan , c .- w .
multiferroic magnetoelectric composite nanostructures .
_ npg asia mater . _
* 2 * , 61 - 68 ( 2010 ) .
velev , j. p. , jaswal , s. s. @xmath64 tsymbal , e. y. multi - ferroic and magnetoelectric materials and interfaces .
r. soc . a _ * 369 * , 3069 - 3097 ( 2011 ) . dong , s. , liu , j .-
m . , cheong , s .- w .
@xmath64 ren , z. multiferroic materials and magnetoelectric physics : symmetry , entanglement , excitation , and topology .
phys . _ * 64 * , 519 - 626 ( 2015 ) .
scott , j. f. searching for new ferroelectrics and multiferroics : a user s point of view .
_ npj comput .
* 1 * , 15006 ( 2015 ) . velev , j. p. , burton , j. d. , zhuravlev , m. y. @xmath64 tsymbal , e. y. predictive modelling of ferroelectric tunnel junctions .
_ npj comput .
* 2 * , 16009 ( 2016 ) .
peng , r. c. , hu , j. m. , momeni , k. , wang , j. j. , chen , l. q. @xmath64 nan , c. w. fast 180 degrees magnetization switching in a strain - mediated multiferroic heterostructure driven by a voltage .
_ * 6 * , 27561 ( 2016 ) . yi , m. @xmath64 xu , b. x. a constraint - free phase field model for ferromagnetic domain evolution .
r. soc . a _ * 470 * , 20140517 ( 2014 ) .
yi , m. , xu , b .- x .
@xmath64 shen , z. effects of magnetocrystalline anisotropy and magnetization saturation on the mechanically induced switching in nanomagnets . _ j. appl
. phys . _ * 117 * , 103905 ( 2015 ) .
hu , j. m. , yang , t. , wang , j. , huang , h. , zhang , j. , chen , l. q. @xmath64 nan , c. w. purely electric - field - driven perpendicular magnetization reversal .
_ nano lett . _ * 15 * , 616 - 622 ( 2015 ) .
yi , m. , xu , b .- x .
@xmath64 shen , z. 180@xmath0 magnetization switching in nanocylinders by a mechanical strain .
_ extreme mech .
lett . _ * 3 * , 66 - 71 ( 2015 ) .
wang , j. , hu , j. , wang , h. , jiang , h. , wu , z. , ma , j. , wang , x. , lin , y. @xmath64 nan , c. w. electric - field modulation of magnetic properties of fe films directly grown on bisco@xmath3pbtio@xmath3 ceramics . _ j. appl .
phys . _ * 107 * , 083901 ( 2010 ) .
liu , m. , nan , t. , hu , j .-
m . , zhao , s .- s
. , zhou , z. , wang , c .- y . , jiang , z .- d . , ren , w. , ye , z .- g . ,
chen , l .- q .
@xmath64 sun , n. x. electrically controlled non - volatile switching of magnetism in multiferroic heterostructures via engineered ferroelastic domain states . _ npg asia mater . _
* 8 * , e316 ( 2016 ) .
yi , m. , xu , b .- x .
@xmath64 gross , d. mechanically induced deterministic 180@xmath0 switching in nanomagnets .
mater . _ * 87 * , 40 - 49 ( 2015 ) .
m . , yang , t. n. , chen , l. q. @xmath64 nan , c. w. voltage - driven perpendicular magnetic domain switching in multiferroic nanoislands .
_ j. appl .
phys . _ * 113 * , 194301 ( 2013 ) .
hu , j .- m . @xmath64
nan , c. w. electric - field - induced magnetic easy - axis reorientation in ferromagnetic / ferroelectric layered heterostructures .
b _ * 80 * , ( 2009 ) .
buzzi , m. , chopdekar , r. v. , hockel , j. l. , bur , a. , wu , t. , pilet , n. , warnicke , p. , carman , g. p. , heyderman , l. j. @xmath64 nolting , f. single domain spin manipulation by electric fields in strain coupled artificial multiferroic nanostructures .
_ * 111 * , 027204 ( 2013 ) .
ghidini , m. , maccherozzi , f. , moya , x. , phillips , l. c. , yan , w. , soussi , j. , metallier , n. , vickers , m. e. , steinke , n. j. , mansell , r. , barnes , c. h. , dhesi , s. s. @xmath64 mathur , n. d. perpendicular local magnetization under voltage control in ni films on ferroelectric batio@xmath3 substrates .
* 27 * , 1460 - 1465 ( 2015 ) .
ghidini , m. , pellicelli , r. , prieto , j. l. , moya , x. , soussi , j. , briscoe , j. , dunn , s. @xmath64 mathur , n. d. non - volatile electrically - driven repeatable magnetization reversal with no applied magnetic field .
* 4 * , 1453 ( 2013 ) .
roy , k. , bandyopadhyay , s. @xmath64 atulasimha , j. binary switching in a symmetric potential landscape .
rep . _ * 3 * , 3038 ( 2013 ) .
roy , k. , bandyopadhyay , s. @xmath64 atulasimha , j. switching dynamics of a magnetostrictive single - domain nanomagnet subjected to stress .
b _ * 83 * , ( 2011 ) .
tiercelin , n. , dusch , y. , klimov , a. , giordano , s. , preobrazhensky , v. @xmath64 pernod , p. room temperature magnetoelectric memory cell using stress - mediated magnetoelastic switching in nanostructured multilayers .
lett . _ * 99 * , 192507 ( 2011 ) .
roy , k. , bandyopadhyay , s. @xmath64 atulasimha , j. energy dissipation and switching delay in stress - induced switching of multiferroic nanomagnets in the presence of thermal fluctuations .
_ j. appl .
phys . _ * 112 * , 023914 ( 2012 ) .
brintlinger , t. , lim , s. h. , baloch , k. h. , alexander , p. , qi , y. , barry , j. , melngailis , j. , salamanca - riba , l. , takeuchi , i. @xmath64 cumings , j. in situ observation of reversible nanomagnetic switching induced by electric fields .
_ nano lett . _
* 10 * , 1219 - 1223 ( 2010 ) .
wang , j. j. , hu , j. m. , yang , t. n. , feng , m. , zhang , j. x. , chen , l. q. @xmath64 nan , c. w. effect of strain on voltage - controlled magnetism in bifeo@xmath3-based heterostructures . _
rep . _ * 4 * , 4553 ( 2014 ) .
wang , j. j. , hu , j. m. , ma , j. , zhang , j. x. , chen , l. q. @xmath64 nan , c. w. full 180 degrees magnetization reversal with electric fields .
_ * 4 * , 7507 ( 2014 ) .
weisheit , m. , fahler , s. , marty , a. , souche , y. , poinsignon , c. @xmath64 givord , d. electric field - induced modification of magnetism in thin - film ferromagnets .
_ science _ * 315 * , 349 - 351 ( 2007 ) .
zhu , w. , xiao , d. , liu , y. , gong , s. j. @xmath64 duan , c. g. picosecond electric field pulse induced coherent magnetic switching in mgo / fept / pt(001)-based tunnel junctions : a multiscale study .
_ * 4 * , 4117 ( 2014 ) .
rondinelli , j. m. , stengel , m. @xmath64 spaldin , n. a. carrier - mediated magnetoelectricity in complex oxide heterostructures .
nanotechnol . _
* 3 * , 46 - 50 ( 2008 ) .
duan , c. g. , velev , j. p. , sabirianov , r. f. , zhu , z. , chu , j. , jaswal , s. s. @xmath64 tsymbal , e. y. surface magnetoelectric effect in ferromagnetic metal films .
_ * 101 * , 137201 ( 2008 ) .
maruyama , t. , shiota , y. , nozaki , t. , ohta , k. , toda , n. , mizuguchi , m. , tulapurkar , a. a. , shinjo , t. , shiraishi , m. , mizukami , s. , ando , y. @xmath64 suzuki , y. large voltage - induced magnetic anisotropy change in a few atomic layers of iron .
_ * 4 * , 158 - 161 ( 2009 ) .
niranjan , m. k. , duan , c .- g . ,
jaswal , s. s. @xmath64 tsymbal , e. y. electric field effect on magnetization at the fe / mgo(001 ) interface .
lett . _ * 96 * , 222504 ( 2010 ) .
wang , w. g. , li , m. , hageman , s. @xmath64 chien , c. l. electric - field - assisted switching in magnetic tunnel junctions .
mater . _ * 11 * , 64 - 68 ( 2011 ) .
shiota , y. , nozaki , t. , bonell , f. , murakami , s. , shinjo , t. @xmath64 suzuki , y. induction of coherent magnetization switching in a few atomic layers of feco using voltage pulses .
* 11 * , 39 - 43 ( 2011 ) .
zhou , z. , howe , b. m. , liu , m. , nan , t. , chen , x. , mahalingam , k. , sun , n. x. @xmath64 brown , g. j. interfacial charge - mediated non - volatile magnetoelectric coupling in co@xmath65fe@xmath66/ba@xmath67sr@xmath68tio@xmath3/nb : srtio@xmath3 multiferroic heterostructures .
rep . _ * 5 * , 7740 ( 2015 ) .
yang , s. w. , peng , r. c. , jiang , t. , liu , y. k. , feng , l. , wang , j. j. , chen , l. q. , li , x. g. @xmath64 nan , c. w. non - volatile 180 degrees magnetization reversal by an electric field in multiferroic heterostructures .
* 26 * , 7091 - 7095 ( 2014 ) .
duan , c. g. , jaswal , s. s. @xmath64 tsymbal , e. y. predicted magnetoelectric effect in fe / batio@xmath3 multilayers : ferroelectric control of magnetism .
_ * 97 * , 047201 ( 2006 ) .
fechner , m. , maznichenko , i. v. , ostanin , s. , ernst , a. , henk , j. , bruno , p. @xmath64 mertig , i. magnetic phase transition in two - phase multiferroics predicted from first principles .
b _ * 78 * , ( 2008 ) .
duan , c .- g . , velev , j. p. , sabirianov , r. f. , mei , w. n. , jaswal , s. s. @xmath64 tsymbal , e. y. tailoring magnetic anisotropy at the ferromagnetic / ferroelectric interface .
lett . _ * 92 * , 122905 ( 2008 ) .
yamauchi , k. , sanyal , b. @xmath64 picozzi , s. interface effects at a half - metal / ferroelectric junction .
_ * 91 * , 062506 ( 2007 ) .
duan , c. g. , sabirianov , r. f. , mei , w. n. , jaswal , s. s. @xmath64 tsymbal , e. y. interface effect on ferroelectricity at the nanoscale .
_ nano lett . _
* 6 * , 483 - 487 ( 2006 ) .
fechner , m. , zahn , p. , ostanin , s. , bibes , m. @xmath64 mertig , i. switching magnetization by 180 degrees with an electric field .
lett . _ * 108 * , 197206 ( 2012 ) .
chu , y. h. , martin , l. w. , holcomb , m. b. , gajek , m. , han , s. j. , he , q. , balke , n. , yang , c. h. , lee , d. , hu , w. , zhan , q. , yang , p. l. , fraile - rodriguez , a. , scholl , a. , wang , s. x. @xmath64 ramesh , r. electric - field control of local ferromagnetism using a magnetoelectric multiferroic .
mater . _ * 7 * , 478 - 482 ( 2008 ) .
heron , j. t. , bosse , j. l. , he , q. , gao , y. , trassin , m. , ye , l. , clarkson , j. d. , wang , c. , liu , j. , salahuddin , s. , ralph , d. c. , schlom , d. g. , iguez , j. , huey , b. d. @xmath64 ramesh , r. deterministic switching of ferromagnetism at room temperature using an electric field .
_ nature _ * 516 * , 370 - 373 ( 2014 ) .
wang , j. j. , hu , j. m. , peng , r. c. , gao , y. , shen , y. , chen , l. q. @xmath64 nan , c. w. magnetization reversal by out - of - plane voltage in bifeo@xmath3-based multiferroic heterostructures . _
rep . _ * 5 * , 10459 ( 2015 ) .
kikuchi , n. , okamoto , s. @xmath64 kitakami , o. the critical size between single domain and multidomain in @xmath4-fept particles .
_ j. appl .
phys . _ * 103 * , 07d511 ( 2008 ) .
kim , k. h. , norton , d. p. , christen , d. k. , cantoni , c. , paranthaman , m. @xmath64 aytug , t. epitaxial growth of mgo / tin multilayers on cu . _ vacuum _ * 83 * , 897 - 901 ( 2009 ) .
jeon , k .-
r . , park , c .- y .
@xmath64 shin , s .- c .
epitaxial growth of mgo and cofe / mgo on ge(001 ) substrates by molecular beam epitaxy .
_ cryst . growth des . _ * 10 * , 1346 - 1350 ( 2010 ) .
niu , f. , meier , a. l. @xmath64 wessels , b. w. epitaxial growth and strain relaxation of mgo thin films on si grown by molecular beam epitaxy .
b _ * 24 * , 2586 ( 2006 ) .
neugebauer , j. @xmath64 scheffler , m. adsorbate - substrate and adsorbate - adsorbate interactions of na and k adlayers on al(111 ) .
b _ * 46 * , 16067 - 16080 ( 1992 ) .
kresse , g. @xmath64 joubert , d. from ultrasoft pseudopotentials to the projector augmented - wave method .
b _ * 59 * , 1758 - 1775 ( 1999 ) . beleggia , m. , graef , m. d. , millev , y. t. , goode , d. a. @xmath64 rowlands , g. demagnetization factors for elliptic cylinders .
_ j. phys .
phys . _ * 38 * , 3333 - 3342 ( 2005 ) .
brown , w. f. thermal fluctuations of a single - domain particle .
rev . _ * 130 * , 1677 - 1686 ( 1963 ) .
d . , song , h .- s . , kim , j .- w . , ko , h. s. , sohn , j .- w . ,
park , b .-
@xmath64 shin , s .- c .
gilbert damping and critical real - space trajectory of @xmath4-ordered fept films investigated by magnetic - field - induction and all - optical methods .
express _ * 7 * , 113004 ( 2014 ) . | voltage - driven 180@xmath0 magnetization switching without electric current provides the possibility for revolutionizing the spintronics .
we demonstrated the voltage - driven charge - mediated 180@xmath0 magnetization switching at room temperature by combining first - principles calculations and temperature - dependent magnetization dynamics simulation .
the electric field ( @xmath1 ) induced interface charge is found to allow a giant modulation of the magnetic anisotropy ( @xmath2 ) of the nanomagnet .
particularly @xmath2 is revealed to vary linearly with respect to @xmath1 and the epitaxial strain .
magnetization dynamics simulations using the so - obtained @xmath2 show that both in - plane and perpendicular 180@xmath0 switching can be achieved by @xmath1 pulses .
the temperature effect renders the 180@xmath0 switching as probability events . statistical analysis indicates a fast ( around 2 ns ) and deterministic 180@xmath0 switching achievable at room temperature by controlling the magnitude of e and the pulse width .
the study inspires the rational design of miniaturized nanoscale spintronic devices where finite temperature effect has a great impact . |
dust is the cornerstone of the active galactic nuclei ( agn ) unified scheme that postulates that the diversity of the observed properties of agn are merely a result of the different lines of sight with respect to obscuring material surrounding the active nucleus ( @xcite ; @xcite ; @xcite ) .
the elements necessary for understanding the nature of agn and the variety of their properties are in fact the geometry of the circumnuclear dust and the amount of obscuration .
the latter is also needed in order to understand the physical conditions of the dust enshrouded nuclear region and to estimate the intrinsic uv - to - optical properties of the agn .
obscuring dusty material should re - radiate in the infrared ( ir ) the fraction of accretion luminosity it absorbs , providing thus a direct metric for the study of the medium .
indeed all spectroscopically confirmed agn lying in regions covered by the various spitzer space telescope ( spitzer ) surveys show significant ir emission down to the detection limits .
the conventional wisdom is that dust , distributed in a more or less toroidal form around the nucleus , mainly consists of silicate and graphite grains , leaving unmistakable signatures in the observed seds of agn .
the most important dust features are the t@xmath101500 black - body like rise of the ir emission at @xmath11 , corresponding to the sublimation temperature of the graphite grains and the @xmath129.7 micron feature in emission / absorption attributed to silicates .
while silicates in absorption were already long ago observed in a variety of type 2 objects , iso observations only provided faint indications of their presence in emission in the spectra of type 1 , challenging the unified scheme , as all smooth torus models predicted a feature in emission for the unobscured agn .
clumpy models , on the other hand , provided a natural attenuation of the feature @xcite .
observations of tens of objects carried out with the irs spectrograph onboard spitzer allowed , however , the detection of the silicate feature ( first reported by @xcite and observed by several others since ) predicted by smooth tori models in emission for type 1 agn and quasars ( e.g. @xcite ; @xcite ; @xcite ) .
thus far , there is no consensus on the dust distribution among astronomers ; for an overview of the issues involving the two approaches see @xcite , @xcite and @xcite .
until very recently the number of spectroscopically confirmed type 2 agn was rather limited ; in the last few years however , the advent of large - scale multi - wavelength surveys has led to a revolution in agn studies .
we are now in a position to compile statistically large type 2 agn samples at all redshifts , selected in a multitude of ways . in this work
we try to characterise the observed spectral energy distribution ( sed ) of type 2 agn using the smooth torus models described in @xcite and the methodology described in @xcite , hereafter paper 1 , in which we dealt with the ir properties of quasars .
the aim of the present work is to model a large sample of type 2 agn and , by comparison with the results derived from the study of type 1 quasars in paper 1 , test the unified scheme .
the various type 2 agn samples will be discussed in detail in section [ sec : data ] .
section [ sec : seds ] will briefly describe the various model components as well as the basics of the standard @xmath2 minimisation technique and assumptions .
the results on the individual samples will be presented in section [ sec : fitresults ] while the combined results from all samples , including the quasars ( typical representatives of type 1 agn ) from paper 1 , will be shown in section [ sec : combi ] .
a final discussion on the future of such studies is presented in section [ sec : discuss ] .
in this work we will be analysing a variety of type 2 agn samples , namely a low redshift , low luminosity agn sample , an x - ray and radio selected sample in the cosmos field and a mid - infrared ( mir ) selected sample from the elais fields
. we will be also revisiting the sdss / swire quasar sample studied in paper 1 ( hereafter quasar sample ) .
the low - luminosity low - redshift agn sample , hereafter low-@xmath7 sample , is part of an updated version of an emission line galaxies sample classified as agn .
the selection was made among hundreds of thousands of emission line galaxies from the sdss data release 4 ( dr4 ; @xcite ) , based on the relative strength of the emission lines on a log(/h@xmath13 ) - log(/h@xmath14 ) diagram , after careful subtraction of the stellar continuum , described in detail on @xcite . from the more than 80,000 proposed agn ,
420 reside in the northern swire ( @xcite ; @xcite ) fields elais n1 , elais n2 and lockman . among these objects ,
388 have been detected in at least two out of the four irac bands , with 340 having also mips detections at 24 .
from the 388 ( 340 with an additional 24 detection ) objects , 155 ( 146 ) and 88 ( 81 ) were additionally detected by mips at 70 and 160 , respectively .
the far - infrared ( fir ) measurements characterize or constrain ( with upper limits ) the cold dust emission . due to the nature of the sample
, particular care was taken in the construction of the seds .
we used data from the sdss dr4 , 2mass including the deeper data set 2mass @xmath15 6 taken in the lockman field @xcite , as well as ir data from swire .
for extended objects ( based on the sdss classification ) we used petrosian @xmath1 , @xmath16 , @xmath17 , @xmath18 , @xmath7 magnitudes , 7 `` radius magnitudes for 2mass @xmath19 , and aperture 5 fluxes ( 5.8 '' radius ) , the largest publicly available irac aperture , for the four irac bands , to better capture the total light arising from the resolved host galaxies .
the 5 `` 2mass magnitudes , closer to the 5.8 '' aperture selected for the irac bands , were proven to be inadequate when all the seds were visually inspected as they were clearly falling short of both the sdss and the irac fluxes . for the point sources we used psf sdss magnitudes , default ( psf ) 2mass magnitudes and aperture 2 ( 1.9 " ) irac magnitudes ( for a full description of the swire magnitudes and their adequate for point - like and extended sources see @xcite ) .
the redshift histogram of the 420 objects is shown in fig .
[ fig : zhisto1 ] , with the gray histogram corresponding to the 388 objects that will be studied in detail .
tables [ tab : sdssobj ] and [ tab : swirefluxes ] show the sdss names , coordinates , redshifts , emission line properties , and stellar masses ( taken from the narrow emission line object catalogue ) , and swire fluxes , respectively for the 420 objects .
the objects belonging to the subset ( s ) of 388 entries used for the study are marked in the column `` s '' of table [ tab : sdssobj ] .
the objects marked in column `` agn '' are those that were assigned with an agn component as a result of the fitting procedure . [ cols= " > , < ,
< , > , > , > , > , > , > , > " , ] [ tab : lineratios ] our analysis is only able to identify a torus component in about one - third of the total low-@xmath7 sample ( 75 [ 142 if no cut in the @xmath2 is imposed ] out of 266 [ 388 ] ) . even though evidence of changing behaviour of ( clumpy ) tori @xcite or even complete disappearance of the torus @xcite towards the lower luminosity regimes ( @xmath20 erg / sec ) has been reported , these effects are not what we observe here , as the lowest accretion luminosity estimated for objects with a torus component is of the order of 10@xmath21 erg / sec , seen in the first plot of fig . [
fig : compare_props ] .
we therefore conclude that we either are unable to detect the signature of agn with @xmath22erg / sec or that the criteria used by @xcite to select agn were too generous and many objects with no nuclear activity may have made it into the suggested agn sample .
another issue to consider is the distribution of the resulting covering factor , shown in fig .
[ fig : kauffcf ] . were the computed values for cf correct
, they would translate to a ratio of type2:type1 objects of about 1:1 in the local and low-@xmath7 universe , a much lower value than that observed .
this discrepancy could again be seen as a result of the agn candidate selection and the somewhat sparse samples of the observed seds , the combination of which does not always allow the fitting code to give robust results . allowing for torus models with @xmath23 does not have any effect on the results , as only 18 out of the 388 objects where better fit by a model including such a torus component , with a single one among them assigned a @xmath24 .
the ssps used to reproduce the blue part of the seds provide information about the stellar mass of the host galaxy . to compute the stellar mass we exploit the fact that ssp spectra are given in luminosity units per one solar mass .
we also account for the fact that a certain percentage of stars , depending on the ssp s age , has evolved and is not shining anymore , so the mass value we provide is actually `` luminous mass '' , fig .
[ fig : kauff_mstars ] shows a comparison of the stellar mass computed in this work and that provided in the narrow line agn catalogue @xcite .
the line shows the linear correlation between the two quantities , @xmath25 = 1.59 + ( 1.06@xmath26 ) .
note that the systematic differences in the mass values are the result of the different imf used by @xcite , a kroupa ( 2001 ) imf , which differs from ours both in the slope and in the mass limits .
we now compute the star formation rate ( sfr ) according to @xcite as @xmath27 = 4.5 \times 10^{-44 } \times l_{fir } \label{eqn : sfr}\ ] ] where @xmath28 is the infrared luminosity of the sb component , integrated between 8 and 1000 micron
[ fig : sfrkauff ] shows the distribution of the sfrs for all objects with an sb component and @xmath29 ( black histogram ) , with the grayed region corresponding to the objects without an agn component .
the average sfr of the entire sample without imposing a cut on the @xmath2 values is of 2.8 @xmath30/yr , and drops to 2.2 when only objects with @xmath31 are consider , and to 1.85 when objects with an agn component are excluded .
note that all objects with a torus component are gathered in the part of the histogram with the higher srfs .
we will get back to the discussion on sfrs in section [ sec : sfr ] .
the spectroscopic sample of x - ray emitters selected in the cosmos field presented in @xcite comprises 80 type 2 agn , 47 type 2 agn and red galaxy hybrid and 8 type 2 agn whose classification is not 100% secure .
ninety two out of these 135 objects have a 24 micron counterpart and , as already mentioned in section [ sec : data ] , we will be concentrating on this sub - sample alone .
this was not a constraint applied to the low redshift sample , however , due to the average ( low ) redshift of this sample as well as its nature ( see fig .
[ fig : zhisto2 ] ) .
the seds are dominated by the stellar light all the way to the irac3 band ( 5.8 ) and even though the presence of a torus is somewhat revealed from the lower s@xmath32/s@xmath33 colour ( with respect to objects without a torus component ) , a single data point at 8.0 ( irac4 ) is not enough to constrain the torus properties . for the record , 36 of the remaining objects ( those without a 24 detection )
were indeed assigned a torus component while six others were better represented by ssps alone .
[ fig : tghistochi2 ] shows the minimum @xmath2 distribution for the sub - sample of 92 objects whose seds extend at least up to 24 , ( solid line ) , excluding two objects with minimum @xmath34 that are obviously erroneous and will not be taken into account in what follows . again for consistency , we will adopt a cut - off @xmath2 of 16 , keeping in mind though that including the rest of the fits will not influence the results in any significant way .
examples of sed fits of objects assigned an agn component are shown in figs .
[ fig : cosmosfits ] and [ fig : elaisfits ] , for the cosmos and elais samples , respectively .
the fits for all objects are available as online material .
note that in the upper and lower left plots in fig .
[ fig : cosmosfits ] ( objects cosmos_j100137.69 + 002844.1 and cosmos_j100249.59 + 013520.2 , respectively ) , the observed 24 points lie well above the model ( in both cases consisting only of stars and a torus ) .
such cases indicate the possible existence of an additional sb component , which we did not use for these fits because of the lack of observed 70 and/or 160 data points , as discussed in detail in section [ sec : sedfit ] .
similarly , for the lower right object in fig .
[ fig : elaisfits ] ( elais_j003718 - 421924 ) , the fit fails to reproduce the 160 point , despite the simultaneous existence of a torus and a sb components .
this case shows that either the sb templates used are not always adequate or , most probably , that we are lacking a diffuse component that would account for the cold dust in the host galaxy .
the fitting results show a clear preference for models with dust density decreasing with the distance from the centre , with 95% of the objects finding a better fit with @xmath35 models ( 75% with @xmath36 ) .
a further dependence of the dust spatial distribution on the altitude from the equator was found , with 87% of the objects having a best fit model with @xmath37 .
the resulting tori sizes do not depend on the type of object ( agn or hybrid ) and the estimated average inner tori radii , @xmath38 , are of [email protected] and [email protected] pc , for agn and hybrids , respectively , and with about 60% of the objects better matching torus models with small sizes ( @xmath40=30 ) .
more than half of the objects were found to have a large covering factor ( @xmath41% ) .
a comparative study on the covering factor derived from the various samples will follow shortly . once again , the position of the objects in the irac colour diagram ( fig .
[ fig : colours ] ) and the contribution of the various components in their global seds are in very good agreement .
[ fig : cosmos_colours ] shows the irac colour diagram for the cosmos ( circles ) and elais ( squares ) samples . the majority of objects with a 70 micron detection and an assigned sb template gather in the pah - dominated locus ( upper left part of the diagram ) while the agn contribution to the ir luminosity increases as objects move along the continuum - dominated locus , becoming redder in the @xmath42 axis .
from the 23 objects composing the mir selected type 2 elais agn sample , 21 were found to have a torus component , all of which with @xmath43 , @xmath44=6.0 and @xmath35 , among which 15 found a better fit with @xmath45 templates .
the average estimated sizes of the tori were very similar to those of the cosmos sample .
the reduced @xmath2 distribution for the fits is shown in fig .
[ fig : tghistochi2 ] in dashed lines .
the agn contribution to the ir luminosity for both samples is shown in fig .
[ fig : cosmoselais_agnfrac ] , with the grayed region representing the objects with an additional sb component .
the plain ( dashed ) line and grayed ( black ) region correspond to the cosmos ( elais ) sample .
again , the existence of such a component can be , in general , well constrained only in the presence of data points longward @xmath46and this is valid for only 25% and 50% of cosmos and elais samples , respectively .
we will now attempt to put all the results from the various agn samples together and construct a common description of their dust properties . to this end
, we will first present a summary of the results of paper 1 , that will also be used in this combined study .
the low-@xmath7 sample will be excluded from this analysis , because of the unconfirmed ( stellar vs. starburst ) nature of the objects .
they will be taken into account , however , when discussing the star formation , as the results on this issue seem to be more robust , and only marginally affected by the presence of tori , as suggested from fig .
[ fig : kauff_agnfrac ] .
paper 1 focused on the study of 278 sdss / swire quasars and the properties of dust surrounding them . comparing the results between the `` restricted '' run , where only high @xmath47 models were allowed and the `` full '' run , where optical depths as low as 0.1 were allowed , the hypothesis of the existence of low optical depth tori could not be ruled out . in fact , the majority of objects found a better match of their seds with low @xmath47 models ( @xmath48 . the computed average inner radius was of @xmath122.6 pc with a tendency for higher @xmath40i ( = 100 ) for the full run ( 59% ) with respect to the restricted run ( 46% ) , as reported in paper 1 .
the covering factor depended of course on the choice of the optical depth but had a relatively flat distribution in both cases , taking values from as low as @xmath49 to as high as 0.95 .
as for the dust density , there was no clear preference between @xmath50 and @xmath51 models , in any of the runs , while @xmath52 models were always favoured .
the accretion and ir luminosities were the two better constrained quantities and were almost independent on the choice of @xmath47 .
the subsample of 70 objects with 70 detections enabled an estimation of the contribution of a starburst to the total ir luminosity , indicating contributions of up to 80% .
section [ sec : dataall ] described the observed properties of the combined agn sample . although the sample is not complete , the combined study of the various agn sub - samples composing it , will allow for comparative conclusions .
[ fig : compare_props ] shows the distribution of the various parameters for the different samples . from top to bottom and left to right , and following the colour - coding introduced in the left panel of fig .
[ fig : colours ] we present : the accretion luminosity ; and the ir luminosity attributed to the torus component ( i.e. excluding the stellar emission coming either from the ssps and/or the sb components ) ; the covering factor ; the viewing angle ; the hydrogen column density along the line of site ; and the mass of dust confined inside the torus .
the open histogram corresponds to the `` restricted '' run , where only high optical depth tori models ( @xmath3 ) were allowed ( paper 1 ) .
the combined quasar and type 2 agn samples cover four orders of magnitude in @xmath53 , with the quasar @xmath53 distribution peaking at luminosities of about an order of magnitude brighter than the type 2s ( see fig .
[ fig : compare_props ] - upper left ) .
this observational bias affects , for example , the fraction of the ir luminosity attributed to the torus component ( fig .
[ fig : compare_props ] - upper right ) .
the viewing angle as derived from the fits , shown in the right column , second row , is consistent with the unified scheme postulating that the differences between type 1 and type 2 objects are a line - of - sight effect , with the type 2 objects seen through the obscuring material .
almost all type 2 objects are seen in small viewing angles , i.e. close to the equator and through large amounts of dust , while type 1 objects are seen in large viewing angles especially when only tori with high optical depths are considered ( open green histogram ) .
note that low optical depth tori would result in low obscuration type 2 agn when seen edge - on .
4 in @xcite , e.g. , shows the emission of an agn through a @xmath54 torus for all lines of sights .
while the object is clearly a type 1 when seen face - on , the uv - to - optical nuclear emission can be absorbed by several orders of magnitude when seen through the dust , depending on the wavelength and the line of sight . the majority of type 2 objects are seen through high column densities ( @xmath55 ) , while the few quasars that are seen through the torus , are actually seen through low optical depth and lower than the type 2 column density gas ( fig .
[ fig : compare_props ] , lower left panel ) .
also consistent with the unified scheme is the distribution of the mass of dust ( right column , fourth row in fig .
[ fig : compare_props ] ) .
if the mass of dust can be considered a linear tracer of gas , this implies that the gas reservoirs in the vicinity of type 1 and type 2 objects are comparable .
note that the properties of the radio and x - ray ( cosmos ) and mir - selected ( elais ) agn are remarkably similar with the only notable difference being the slightly higher ir luminosities of the torus components of the elais sample ( upper right plot in fig .
[ fig : compare_props ] ) , simply reflecting the selection effects .
the distribution of the values for the covering factor ( seen in the left column , second row in fig .
[ fig : compare_props ] ) indicates that tori with both high and low covering factors are possible in both types of objects .
the cosmos and elais samples show a stronger tendency for higher values . the estimated type2:type1 ratios based on the cf distribution of the type2 objects is @xmath122 - 2.5:1 ;
the same ratio is found when considering the restricted run of the quasar sample .
the full type 1 run , however , suggests a ratio of 1:1 , with a mean value of the covering factor of @xmath120.47 , very close to the value of 0.4 computed by @xcite for the entire swire quasar population .
this low value , seemingly in conflict with the unified scheme , actually reflects the difference in the accretion luminosity between the type 1 and type 2 samples ( fig .
[ fig : compare_props ] ) , already discussed in section [ sec : compare ] .
from the type 1 quasars with redshifts below @xmath56 we can measure the flux at 5100 , which we can then scale with the accretion luminosity , @xmath53 , as seen in the left panel of fig . [
fig : fracobscagn ] . assuming this relation holds for all redshifts as well as for type 2 objects
, we can translate @xmath53 to @xmath57 .
we can then compute the fraction of obscured agn in bins of @xmath57 , shown in the right panel of fig .
[ fig : fracobscagn ] .
the black circles and open squares make use of the results of the full and restricted runs , respectively .
the full and two dashed lines show the results and uncertainties ( due to bolometric corrections ) obtained by @xcite who conducted a study of 25 high luminosity high redshift ( @xmath58 ) quasars with irs spectroscopy . the observed trend , i.e. the decreasing average torus covering factor with increasing agn luminosity , is in support of the receding torus paradigm , suggested by @xcite . despite the crude conversion of @xmath53 to @xmath57 our results making use of the full run are in very good agreement ( i.e. within the uncertainties ) with those derived by @xcite based on totally independent method and sample , while the results based on the restricted run , even though still consistent ( excluding the brightest two bins ) show a quite different trend .
we can therefore still not conclude on the relative occurence of low @xmath47 tori with respect to high @xmath47 tori but , based on the above , we can still provide some evidence for their existence .
previous studies ( e.g. @xcite ) assume that the ratio between the primary agn radiation ( here measured by @xmath53 ) and the thermal infrared emission attributed to the agn is a direct indicator of the torus covering factor .
our findings confirm this assumption , as shown in fig .
[ fig : lacclircf ] where the accretion - to - ir agn luminosity of each individual object of each sample is plotted against the computed covering factor .
the slope of the correlation is very similar for type 1 and type 2 samples .
type 2 objects , however , have considerably larger @xmath59 ratios than the quasars .
this `` jump '' in the values is due to the different viewing angles and occurs right when the line of sight intercepts the obscuring torus . in order to illustrate this effect
, we show the @xmath59 ratios as a function of the viewing angle for the extreme models of both type 1 and 2 objects ( higher and lower red points ; higher and lower green points in fig .
[ fig : lacclircf ] ) in fig .
[ fig : lacclirvang ] .
the study of objects with at least one data point at wavelengths longward of @xmath46 shows that the ir emission of agn can not be attributed to the torus component alone . in the vast majority of cases , and in order to reproduce the 70 ( and , whenever available , 160 ) points , an additional sb component was necessary .
furthermore , in all those cases the contribution of the agn emission in the infrared is typically smaller than 50% , as seen in figs .
[ fig : kauff_agnfrac ] and [ fig : cosmoselais_agnfrac ] for the type 2 objects , and a bit higher in type 1 quasars ( paper 1 , fig .
[ fig : sfr ] shows the estimated sfrs for the various samples ( with the same colour coding as in fig .
[ fig : colours ] ) computed from eq .
[ eqn : sfr ] for all objects with an sb and an agn components .
the average sfr for the low redshift sample is about 10 @xmath30/yr , while those computed for the cosmos and elais samples are of @xmath1240 @xmath30/yr but with a larger span .
that of type 1 quasars ( shown in green ) , however , has an average sfr of 115 @xmath30/yr , with some cases reaching sfrs of the order of 10@xmath60 @xmath30/yr .
these values are well in agreement with the findings of @xcite , a study based on stacking analysis of a variety of quasar samples .
these values do not necessarily imply that star formation activity is stronger in type 1 quasars _ in general_. more likely , it reflects the observational biases of the samples , both in terms of average redshifts and luminosities ( figs .
[ fig : zhisto2 ] and [ fig : compare_props ] ) .
the starburst - to - agn luminosity ratio for the combined sample shows a slight tendency to decrease with increasing accretion luminosity , as seen in fig .
[ fig : lsblagnlaccz ] .
this is in apparent agreement with previous works on type 1 quasars alone ( again by @xcite ) , and if confirmed for the entire sample , would imply that type 1 and type 2 objects behave in the same way in this respect . on the other hand , we can not assert beyond doubt that there there is no implicit dependency on the redshift .
in fact , examining the quasar sample alone ( open symbols ) we would be tempted to claim an _ increase _ with @xmath53 , but which is likely attributed to the @xmath61 - @xmath7 degeneracy .
this work focuses on the ir properties type-2 agn , the type 1 issues having already been discussed in a previous paper @xcite .
the collection of data used for this paper comes from : 1 ) a low - redshift ( @xmath62 ) agn sample , selected via standard emission - lines ratios criteria ; 2 ) a higher redshift ( @xmath63 ) sample from the cosmos and elais fields , selected with x - ray and mid - infrared criteria .
the observed seds were reproduced by means of a three component model , including emission from stars , hot dust from agn and warm dust from starburst activity , and acceptable fits were obtained for the majority of the objects . despite the fact that the samples we analyzed were not chosen homogeneously , a comparison of the behaviour between the physical properties of the type 1 and type 2 agn , shows an overall agreement .
in paper i we found the covering factor to have a broad trend towards lower values as the accretion luminosity , i.e. the power of the central engine , increased with small , on - average , covering factors . applying
the @xmath53 versus @xmath57 correlation derived for quasars in paper i , we derived the flux at 5100 for the type 2 samples and checked the correlation between the optical luminosity and the covering factor , which in turn , can be converted into a type1-to - type2 fraction .
the results provided by the sed fitting analysis , show a very good agreement with the relation presented in maiolino et al .
( 2007 ) , who find a correlation between the fraction of obscured agn as a function of the optical luminosity ( i.e. at 5100 ) .
another indication of the dusty torus was thought to be ratio between the accretion luminosity and the torus infrared luminosity .
in fact , the latter is just reprocessed radiation which scales linearly with the primary source : what makes the difference is the fraction of `` heating radiation '' which is intercepted by the dust , which again depends on the covering factor .
we find a very similar trend for these quantities , in both type 1 and type 2 objects , also showing how differences in the amount of obscuration are very well explained by _
dust self - absorption _ ,
i.e. thermal dust emission absorbed by dust iteself . for the low-@xmath7 sample , we found no evidence of emission from an agn component for @xmath64% of the objects .
although we can not rule out the absence of an agn from these sources , we can set an upper limit to its luminosity since it is not observed at mid - infrared wavelengths where its emission is the strongest as compared to both that of the stellar and of the starburst - heated dust .
this sample suffers from high contamination from the stellar emission component , even at mid - infrared wavelengths where ( on the contrary ) it s the agn component that usually dominates when present .
the question to ask , therefore , is what can be really constrained with three or four data points .
one of the most reliable quantities should be the optical depth , since low values would make the torus emission stand over the stellar ( sb or stars ) in the mir . in this respect
, the low-@xmath7 sample may be too generous in selecting really active agn .
there is , sometimes , no evidence for a mir excess at all with respect , for example , not only to a `` normal '' starburst emission , but also to a passively evolving galaxy . the absence of any evidence for
a hot dust emission in the mir , which is the place where the stellar component is less important and the agn contribution is increasingly brighter , will remove the agn component from the fits even though this could also be turned into an upper limit . in the cases where there is a mir excess with respect to a pure stellar emission , we explore two possibilities : torus and pah emission . in some cases
a better fit was obtained adding pah -i.e .
starburst component alone- to the model instead of agn which should instead dilute the pah emission @xcite .
comparison between clumpy and smooth tori models ( e.g. @xcite ) indicate that globally the seds produced by the two models are quite similar , but with some details characteristic for one or the other model : the silicate feature observed in absorption in objects seen edge - on is shallower for clumpy models , the average near - ir flux is weaker in smooth models and the clumpy models tend to produce slightly wider seds at certain inclinations .
furthermore , clumpy models can produce very small tori sizes with @xmath65 @xcite , while still producing a broad mir emission .
subsequently , the selection of a smooth torus for the present study might have resulted in overestimated tori sizes without , however , jeopardising the estimates on the properties related to the ir emission or our conclusions on the unification scheme .
sed fitting may not be a sensitive enough method to distinguish between the differences introduced by the two approaches , because of the width of the filters and the scarce sampling of the observed seds , but also because the various characteristics of the torus component can be altered or diluted by , for example , the presence of a starburst component .
notwithstanding these limitations , sed fitting is still the best tool available for extracting the maximum information from large photometric agn samples and is now proving to be a powerful technique in relating the dust properties to the accretion properties as well as the properties of the larger host galaxy . because agn are of order hundred times less numerous than galaxies and also prone to selection biases , such agn studies are only possible with the advent of multi - wavelength surveys that probe large volumes and allow the construction of well - samples seds .
even though we can not address all issues related to dusty tori , our understanding of their properties has been greatly improved over the last decade thanks to the spitzer observatory .
this space infrared telescope allowed , among other things , the construction of seds of hundreds of agn of all types thanks to the irac and mips photometry ( e.g. @xcite ; @xcite ; @xcite ; @xcite ) .
spitzer also opened the doors to detailed and coherent studies of the interplay between agn and starburst activity ( e.g. @xcite and references therein ) , paving the way for herschel .
in fact , herschel with its two cameras / medium resolution spectrographs ( pacs ; @xcite and spire ; @xcite ) and a very high resolution heterodyne spectrometer ( hifi ; @xcite ) will be the first space facility to completely cover the range between 60 and 670 , where the the bulk of energy is emitted in the universe , allowing for a more detailed sampling of the observed fir seds and the cold dust emission .
large parts of the herschel key science will focus on the formation and evolution of galaxies and the studies of star formation .
hermes , the herschel multi - tiered extragalactic survey , is the largest project that will be conducted by herschel , with dedicated @xmath12900 hours that will map over 70 square degrees including most of the swire fields ( where most of the ir photometry of all the objects in this study comes from ) as well as other extragalactic survey fields covered by several missions in all wavelengths ( e.g. cosmos or the groth strip ) , and will address among others the issue of agn and starburst connection ( @xcite ; @xcite ) . additionally with herschel , the astrophysical terahertz large area survey ( atlas ) will cover about 500 square degrees and will observe all sdss quasars with redshift @xmath66 as well as _ all _ the brightest fir sdss quasars ( with luminosities 10 times larger than the mean ) , amounting to @xmath12330 individual detections of quasars in the area covered @xcite ( a factor of @xmath125 more than all the sdss quasars with 70 detections ) .
these studies will allow us to significantly improve our understanding of the agn phenomenon , disentangle the contribution of cold and hot dust emission in their seds and study the concomitant occurence of nuclear activity and star formation .
this work is based on observations made with the _
spitzer space telescope _ , which is operated by the jet propulsion laboratory , california institute of technology under nasa contract 1407 .
support for this work , part of the spitzer space telescope legacy science program , was provided by nasa through an award issued by the jet propulsion laboratory , california institute of technology under nasa contract 1407 .
funding for the creation and distribution of the sdss archive has been provided by the alfred p. sloan foundation , the participating institutions , the national aeronautics and space administration , the national science foundation , the u.s .
department of energy , the japanese monbukagakusho , and the max planck society .
the sdss web site is http://www.sdss.org/. this publication makes use of data products from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by the national aeronautics and space administration and the national science foundation .
adelman - mccarthy j. et al . , 2006 , apjs , 162 , 38 antonucci , r. 1993 , ara&a , 31 , 473 barvainis r. , 1987 , apj , 320 , 537 beichman c.a .
, cutri r. , jarrett t. , stiening r. , skrutskie m. , 2003 . , aj , 125 , 2521 berta s. , fritz j. , franceschini a. , bressan a. , lonsdale c. , 2004 , a&a , 418 , 913 bertelli , g. , bressan , a. , chiosi , c. , fagotto , f. , & nasi , e. 1994 , aaps , 106 , 275 bressan , a. , granato , g. l. , & silva , l. 1998 , aap , 332 , 135 capak p. et al .
, 2007 , apjs , 172 , 99 cardamone c.n .
, urry c.m . , damen m. , van dokkum p. , treister e. , labb i. , virani s.n . , lira p. , gawiser e. , 2008 , apj , 680 , 130 cardelli , j. a. , clayton , g. c. , & mathis , j. s. , 1989 , apj , 345 , 245 de graauw th .
et al . , 2009 , eas publications series , 2009 , 34 , 3 dullemond c.p .
& van bemmel i.m . , 2005 . , a&a , 436 , 47 elitzur m. , 2008 , new astronomy review , 52 , 274 elitzur m. & shlosman i. , 2006 , apj , 648 , 101 engelbracht c.w .
, rieke g.h . ,
gordon k.d .
, smith j .- d.t . ,
werner m.w . , moustakas j. , willmer c.n.a . , vanzi l. , 2008 , apj , 678 , 804 franceschini a. et al . , 2005 ,
aj , 129 , 2074 fritz j. , franceschini a. , hatziminaoglou e. , 2006 , mnras , 366 , 767 granato g.l . , danese l. , 1994 , mnras , 268 , 235 griffin m. et al .
, 2009 , eas publications series , 34 , 33 griffin m.j .
et al . , 2006 , in the proceedings of the conference `` studying galaxy evolution with spitzer and herschel '' , may 2006 , agios nikolaos , crete gruppioni c. et al . , 2008 , apj , 684 , 136 hatziminaoglou e. et al . , 2008 , mnras , 386 , 1252 hatziminaoglou e. et al . , 2007 , aspc , 380 , 367 hatziminaoglou e. et al . , 2005 ,
aj , 129 , 1198 hernan - caballero a. et al . , 2009 , mnras , in press hnig s.f .
& beckert t. , 2007 , mnras , 380 , 1172 jacoby g.h . , hunter d.a .
, christian c.a . , 1984 ,
apjs , 56 , 257 kewley l.j . ,
dopita m.a . ,
sutherland r.s .
, heisler c.a . , trevena j. , 2001 , apj , 556 , 121 kauffmann g. et al . , 2003 , mnras , 346 , 1055 kennicutt r.c . , 1998 , apj , 498 , 541 lacy m. et al . , 2004 , apjs , 154 , 166 lawrence a. , 1991 , mnras , 252 , 586 lonsdale c. et al . , 2003 , pasp , 115 , 897 lonsdale c. et al . , 2004 ,
apjs , 154 , 54 lutz d. , spoon h.w.w . , rigopoulou d. , moorwood a.f.m . , genzel r. , 1998 , apj , 505 , 103l maiolino r. , shemmer o. , imanishi m. , netzer h. , oliva e. , lutz d. , sturm e. , 2007 , a&a , 468 , 979 nenkova m. , sirocky m. m. , nikutta r. , ivezi . , elitzur m. , 2008 , apj , 685 , 160 nenkova m. , ivezic z. , elitzur m. , 2002 , apj , 570 , 9 pier e.a . &
krolik j.h . , 1992 , apj , 401 , 99 poglitsch a. & altieri b. , 2009 , eas publications series , 34 , 43 polletta m. , weedman d. , hoenig s. , lonsdale c.j . , smith h. , houck j. , 2008 , apj , 675 , 960 richards g.t .
et al . , 2006 ,
apjs , 166 , 470 rowan - robinson m. et al . , 2008 , mnras , 386 , 697 sajina a. , lacy m. , scott d. , 2005 , mnras , 621 , 256 sanders d.b .
et al . , 2007 , apjs , 172 , 86 serjeant s. & hatziminaoglou e. , 2009 , mnras , 397 , 265 siebenmorgen r. , haas m. , krgel e. , schulz b. , 2005 , a&a , 436 , l5 surace j. et al . , 2005 , swire data release document 2 , url : http://data.spitzer.caltech.edu/popular/swire/20050603_enhanced/documentation/swire2_doc_083105.pdf tadhunter c. , 2008 , new astronomy review , 52 , 227 trump j.r .
et al . , 2007 , apjs , 172 , 383 urry , c. m. & padovani p. , 1995 ,
pasp , 107 , 803 | this paper is the second part of a work investigating the properties of dusty tori in active galactic nuclei ( agn ) by means of multi - component spectral energy distribution ( sed ) fitting .
it focuses on low luminosity , low redshift ( @xmath0 ) agn selected among emission line galaxies in the overlapping regions between swire and sdss data release 4 as well as x - ray , radio and mid - infrared selected type 2 agn samples from the literature .
the available multi - band photometry covers the spectral range from the @xmath1-band up to 160 .
using a standard @xmath2 minimisation , the observed sed of each object is fit to a set of multi - component models comprising a stellar component , a high optical depth ( @xmath3 ) torus and cold emission from a starburst ( sb ) .
the torus components assigned to the majority of the objects were those of the highest optical depth of our grid of models ( @xmath4 ) .
the contribution of the various components ( stars , torus , sb ) is reflected in the position of the objects on the irac colour diagram , with star- , torus- and starburst - dominated objects occupying specific areas of the diagrams and composite objects lying in between .
the comparison of type 1 ( as derived from paper 1 , @xcite ) and type 2 agn properties is broadly consistent with the unified scheme .
the estimated ratio between type 2 and type 1 objects is about 2 - 2.5:1 .
the agn accretion - to - infrared luminosity ratio is an indicator of the obscuration of the agn since it scales down with the covering factor .
we find evidence supporting the receding torus paradigm , with the estimated fraction of obscured agn , derived from the distribution of the covering factor , decreasing with increasing optical luminosity ( @xmath5 ) over four orders of magnitude .
the average star formation rates are of @xmath6 for the low-@xmath7 sample , @xmath8 for the other type 2 agn and @xmath9 for the quasars ; this result however , might simply reflect observational biases , as the quasars under study were one to two orders of magnitude more luminous than the various type 2 agn .
for the large majority of objects with 70 and/or 160 detections an sb component was needed in order to reproduce the data points , implying that the far - infrared emission in agn arises mostly from star formation ; moreover , the starburst - to - agn luminosity ratio shows a slight trend with increasing luminosity .
galaxies : active quasars : general galaxies : starburst infrared : general |
the compact radio continuum sources comprising w51 ( @xcite ) have long been recognized to constitute some of the most luminous star formation regions in the disk of the galaxy ( @xcite ; @xcite ) .
the high luminosity , the large number of inferred o type stars ( @xcite ) , and the location of these sources within a molecular cloud ( @xcite ) all suggest that the w51 region represents the early formation stages of an ob association .
besides the intrinsic interest in the properties of w51 , this region represents one of the closest analogs in the disk of the milky way to the more vigorous star forming sites found in other galaxies ( e.g. 30 doradus ) .
since these latter regions are quite distant , w51 affords many advantages in investigating the detailed properties of luminous star forming sites and inferring how these regions may originate .
one key to understanding the formation and evolution of any star forming region is establishing the properties of the molecular cloud out of which the stars form . while the molecular gas in the w51 region has been the subject of numerous studies , the interpretation of the results remain controversial .
scoville & solomon ( 1973 ) , primarily on the basis of small strip maps in @xmath1co(10 ) , identified several molecular line components toward w51 and derived a minimum mass of @xmath12 m and a diameter @xmath92030pc for the molecular cloud that they associated with the most intense radio component at @xmath136 cm ( g49.5 - 0.4 ; @xcite ) .
they further suggested that this cloud might be physically related to the several thermal radio continuum sources that make up the w51 hii - region complex ( @xcite ; @xcite ; @xcite ) .
subsequent studies of the molecular gas toward w51 have confirmed the existence of a large molecular cloud ( @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ) , although various models continue to be proposed for the relationship of the multiple spectral features seen in the molecular gas lines and their association with the different hii regions .
the primary difficulty in establishing a definitive model of this region is the unique location of w51 in the galaxy with respect to the sun .
the w51 region has classically been associated with the tangent point of the sagittarius spiral arm , which is oriented such that the line of sight toward w51 intersects the spiral arm over several kiloparsecs of path length ( @xcite ; @xcite ) .
much of the uncertainty surrounding the w51 region stems from establishing whether the numerous radio continuum sources and molecular clouds represent a single , large massive star forming region , or the chance projection of unrelated star forming sites viewed down a spiral arm . to better place the w51 region in context with respect to its location in the galactic plane , figure [ fig1 ]
displays the integrated @xmath1co(10 ) emission in 10km0.2em s@xmath7 velocity bins covering longitudes 4055 from the massachusetts - stony brook @xmath1co survey ( @xcite ) .
the w51 region is distinguished by bright @xmath1co emission extending over a 1@xmath31 area centered on ( @xmath14 ) @xmath9(49.5 , -0.2 ) at velocities @xmath1555km0.2em s@xmath7 .
a `` 3d '' view of the ( @xmath16 ) @xmath1co data cube covering the region surrounding w51 is shown in figure [ fig2 ] .
the @xmath1co isointensity contour surface in this figure clearly illustrates both the relatively large number of smaller molecular clouds with typical internal velocity dispersions of @xmath1735km0.2em s@xmath7 , and the large concentration of @xmath1co emission extending over a @xmath920km0.2em s@xmath7 interval in the w51 region .
much of the @xmath1co emission in this area has centroid velocities that exceed the maximum velocity permitted by pure circular rotation ( i.e. @xmath18 @xmath95457km0.2em s@xmath7 ; @xcite ) .
such velocities have long been noted in 21 cm hi surveys at longitudes near @xmath19 , and have been attributed to large - scale streaming motions of gas in a spiral density wave ( e.g. @xcite ; @xcite ) . in principle
the extent and properties of the molecular clouds located in the w51 region can be established by using the kinematic information in the molecular line data to isolate individual clouds . in practice , previous surveys have had either poor resolution or sparse sampling to make such an attempt feasible .
therefore , we have obtained full beam sampled maps of the w51 region in both @xmath1co(10 ) and @xmath2co(10 ) at subarcminute resolution in order to determine the relationship between the various molecular components .
these maps permit us to disentangle the blends of unrelated clouds along the line of sight and to obtain more accurate mass estimates of the molecular gas .
these data can also be compared with similar maps of more nearby clouds that have recently been obtained by us and others .
the outline of this paper is as follows . in section [ obs ] ,
the observing procedures are described and channels maps of the @xmath1co and @xmath2co emission are presented .
analysis of the different spectral features observed in these maps and a more thorough discussion of the main features associated with the compact radio continuum sources in w51 is given in section [ analysis ] .
in section [ discussion ] , we discuss the current massive star formation events in the region with respect to the various molecular components and comment on the evolution of the w51 giant molecular cloud ( gmc ) .
our conclusions are summarized in section [ summary ] .
a 1.39 x 1.33 region ( 100 x 96 pixels ) toward the w51 region was mapped in @xmath1co(10 ) ( 115.271203 ghz ) and @xmath2co(10 ) ( 110.201370 ghz ) in april 1995 using the quarry receiver array ( @xcite ) on the fcrao 14 m telescope .
the full width at half maximum beam size of the 14 meter antenna at these frequencies is 45 and 47 at 110 and 115 ghz respectively .
the data were taken in position switching mode and calibrated with the standard chopper wheel method of observing an ambient temperature load and sky emission .
the backends for each pixel of the array consisted of an autocorrelator spectrometer set to span the velocity range from @xmath9 0100km0.2em s@xmath7 at 78 khz sampling ( 0.20km0.2em s@xmath7 @ 115 ghz ) and 94 khz resolution ( 0.24km0.2em s@xmath7 ) . during data reduction
the spectra were smoothed to a velocity resolution of 0.5km0.2em s@xmath7 .
previous fcrao measurements indicate that the spillover and scatter efficiency ( @xmath20 ) of the telescope and radome is @xmath90.7 at the observed frequencies .
the observed antenna temperatures corrected by @xmath20 are presented as @xmath21 ( @xcite ) .
a further correction , the source coupling efficiency ( @xmath22 ) , accounts for the coupling of the beam to the source . for a uniform source that fills only the main beam of the 14 m telescope ,
@xmath22 is @xmath90.7 ( i.e. 30% of the power is scattered on angular sizes much greater than the fwhm beam size ) , while for sources with uniform intensity over a diameter of 30 , @xmath22 is 1.0 . in practice ,
the observed structures in the @xmath1co and @xmath2co maps span a range of sizes and shapes , and applying a single coupling efficiency for the entire map is incorrect . for simplicity , we present and analyze the data in the @xmath21 temperature scale .
the typical rms noise in the @xmath1co and @xmath2co maps in 0.5km0.2em s@xmath7 channels is @xmath23@xmath21 @xmath90.7 k and 0.6 k respectively .
images of the integrated @xmath1co and @xmath2co intensity ( @xmath24 in 2km0.2em s@xmath7 wide intervals are presented in figures [ fig3 ] and [ fig4 ] respectively in the velocity range from 40 to 70km0.2em s@xmath7 .
the values printed in the upper left corner of each figure panel denote the centroid velocity of the particular interval .
extended @xmath1co and @xmath2co emission was detected between 0km0.2em s@xmath7 and 25km0.2em s@xmath7 , but these data are not presented here .
this low velocity emission most likely originates from local molecular clouds and is not related to the w51 region of interest here .
little emission was observed between 25 and 35km0.2em s@xmath7 and at velocities in excess of 75km0.2em s@xmath7 ( see [ analysis ] ) .
similar velocity structure is also observed in the 21 cm hi emission lines ( @xcite ) .
the following section analyzes the velocity structure in the molecular line maps and identifies individual molecular clouds .
the @xmath1co and @xmath2co emission toward the w51 region contains a number of discrete velocity components that overlap in projection both spatially and kinematically . to identify and isolate the emission from these velocity components , multiple gaussians convolved with the spectrometer channel widths were fitted to each spectrum in an automated manner .
the free parameters for each gaussian were the amplitude of the spectral line , the mean velocity , and the line width .
the number of gaussians fitted to each spectrum was determined by searching for contiguous channels that contain an integrated intensity with a signal to noise ratio @xmath25 3 .
channels containing a local antenna temperature maxima ( denoted here as channels @xmath26 ) in each such section were then identified .
a local maximum at channel @xmath27 was deemed a `` significant '' peak if the antenna temperature in any channel between c@xmath28 and the neighboring local maximum at channels @xmath29 and @xmath30 decreased by more than 2@xmath31 from the antenna temperature at channel @xmath27 .
the number of significant peaks corresponded to the number of gaussians fitted to that section .
each spectrum was smoothed to a velocity resolution of 1.5km0.2em s@xmath7 prior to identifying the peaks , although the fits were performed on the 0.5km0.2em s@xmath7 resolution data .
spectra with large residuals with respect to the gaussian fits were visually inspected and additional gaussians were added as appropriate .
the @xmath2co data were easily decomposed into gaussians in this manner , but it often became difficult to reliably identify the velocity features in the heavily blended @xmath1co lines .
also , toward the compact hii regions , some of the structure in the @xmath1co and @xmath2co line profiles can be attributed to the absorption of radiation from hot molecular gas by colder foreground material ( see @xcite ) .
away from these compact regions , absorption effects are not as significant , and over most of the cloud , the peaks in the spectral lines should accurately represent the velocity structure along the line of sight .
the results from the gaussian decomposition of the line profiles are synthesized in figure [ fig5 ] .
the upper panel shows histograms of the mean velocities for the @xmath1co ( thick lines ) and @xmath2co ( thin lines ) gaussians , and the lower panel shows the total integrated intensity in the gaussians as a function of the mean velocity . most of the emission is confined to velocity intervals of 025km0.2em s@xmath7 and 3575km0.2em s@xmath7 .
we identify the 025km0.2em s@xmath7 emission as originating from nearby molecular material and the 3575km0.2em s@xmath7 emission with molecular gas in the sagittarius spiral arm .
several velocity components occur repeatedly in both the @xmath1co and @xmath2co gaussian fits as signified by the histogram peaks shown in the top panel in figure [ fig5 ] .
in particular , velocity components at 7 , 1525 , 44 , 49 , 53 , 60 , 63 , and 68km0.2em s@xmath7 are readily apparent . in terms of the @xmath1co and @xmath2co integrated intensity ,
the two major features are the 60km0.2em s@xmath7 and 63km0.2em s@xmath7 components .
a 58km0.2em s@xmath7 component is indicated as well since that is the centroid velocity of the molecular line emission toward the brightest radio continuum source in the w51 region ( @xcite ) .
note , however , that the 58km0.2em s@xmath7 component is not a prominent feature as judged from figure [ fig5 ] .
the following discussion briefly highlights the morphology of the individual velocity components .
the 7km0.2em s@xmath7 velocity component contains weak , narrow lines over nearly the entire mapped region and is undoubtedly a nearby molecular cloud .
the 1525km0.2em s@xmath7 interval appears to contain a few distinct velocity features ( see fig .
[ fig5 ] ) , but it is unclear whether or not these components are physically related .
the 44km0.2em s@xmath7 cloud is elongated parallel to the galactic plane at @xmath32 , although this cloud may form part of a larger structure that extends to lower galactic latitudes .
the molecular gas at these lower latitudes occurs at velocities of @xmath940km0.2em s@xmath7 , which is outside the velocity range assigned to this feature .
the @xmath1co emission from the 49 and 53km0.2em s@xmath7 components is more fragmented than the other features mentioned so far .
these fragments may represent either individual clouds or the remnants of a once larger cloud in the galactic plane .
the 53km0.2em s@xmath7 cloud is distinguished by bright @xmath1co emission near @xmath33 @xmath9@xmath34 that is associated with the compact hii region g49.4 - 0.3 ( see [ dis : hii ] ) .
the 63km0.2em s@xmath7 component extends for nearly a degree in length and is found mainly in the central and eastern part of the mapped region .
this is best observed in the 66km0.2em s@xmath7 panel shown in figure [ fig3 ] , which represents the line wing emission of this velocity component ( as well as emission from the 68km0.2em s@xmath7 cloud discussed below ) .
the 60km0.2em s@xmath7 component consists predominantly of a diffuse patch of emission that extends into a filament to the east , and a second filament to the south .
the 60km0.2em s@xmath7 and 63km0.2em s@xmath7 velocity components , along with the 68km0.2em s@xmath7 cloud discussed below , likely correspond to the `` high velocity stream '' of 21 cm hi emission ( @xcite ) that has been attributed to the streaming motions of gas down the sagittarius spiral arm .
careful inspection of the channel maps indicates that the spatial distribution of the 60km0.2em s@xmath7 and 63km0.2em s@xmath7 components generally do not overlap .
for example , the western edge of the 63km0.2em s@xmath7 component closely matches the eastern edge of the 60km0.2em s@xmath7 emission .
this is best seen in figure [ fig4 ] and the 60km0.2em s@xmath7 and 66km0.2em s@xmath7 velocity panels in figure [ fig3 ] .
further , the extended emission from the eastern portion of the 63km0.2em s@xmath7 component is just above the filamentary extension of the 60km0.2em s@xmath7 component .
such interfaces are unlikely to occur by chance from two unrelated clouds along the line of sight , and suggest that the 60 and 63km0.2em s@xmath7 components represent kinematic structure within a single molecular cloud . velocity differences of this magnitude are commonly observed in nearby molecular clouds ( e.g. @xcite ) .
the 58km0.2em s@xmath7 component is dominated by bright , compact molecular emission and does not contain the diffuse extended emission that characterizes the 60 and 63km0.2em s@xmath7 features .
inspection of the channel maps suggests that this velocity component also reflects the interval kinematic structure within a single cloud encompassing the 60km0.2em s@xmath7 and 63km0.2em s@xmath7 clouds .
for example , the emission from the filament protruding to the southern portion of the mapped region contains primarily a centroid velocity of 58km0.2em s@xmath7 closest to bright compact @xmath1co and @xmath2co emission .
further away from this bright , compact emission region , the velocity of the filament changes to @xmath960km0.2em s@xmath7 .
similar velocity patterns are observed in emission features along the eastern and western edges of the map .
these results suggest that the emission constituting the 58 , 60 , and 63km0.2em s@xmath7 components represent the internal velocity structure within a single molecular cloud .
koo ( 1997 ) reached similar conclusions concerning the atomic hydrogen clouds at these velocities based upon hi absorption observations toward the radio continuum sources .
since the bright @xmath1co emission associated with the 58km0.2em s@xmath7 and 60km0.2em s@xmath7 components are coincident with the brightest radio continuum source in the w51 region ( g49.5 - 0.5 ; see [ dis : hii ] ) , we henceforth refer to the 586063km0.2em s@xmath7 components as the w51 molecular cloud .
the 68km0.2em s@xmath7 component extends for @xmath91 east west across the image and also likely constitutes part of the `` high velocity stream '' identified in hi surveys ( @xcite ) .
interestingly , the 68km0.2em s@xmath7 filament is located at the southern edge of the diffuse emission associated with the w51 molecular cloud at velocities @xmath1563km0.2em s@xmath7 ( see figs . [ fig3 ] and [ fig4 ] ) .
again , such a clear truncation of the w51 molecular cloud at the location of the 68km0.2em s@xmath7 cloud is unlikely to occur from two random clouds along the line of sight , and strongly suggests that these two clouds are physically related objects at a common distance . nonetheless , the elongated appearance of the 68km0.2em s@xmath7 cloud is in stark contrast to the roughly circular shape of the w51 cloud , indicating that these two objects are best treated as individual structures rather than a single molecular cloud .
the physical properties of the molecular clouds in the w51 region can be determined from the gaussian decomposition of the line profiles .
we single out these two clouds since they contain four of the five bright hii regions found in radio continuum surveys ( see [ dis : hii ] ) . in deriving the properties , the distance to the w51 cloud
is assumed to be 7.0 @xmath35 1.5 kpc as determined from proper motion studies of the w51main @xmath36 maser in the g49.5 - 0.4 dense core ( @xcite ) .
the 68km0.2em s@xmath7 cloud was assumed to have the same distance based on its apparent association with the w51 cloud as discussed above .
the other clouds are not included in this analysis since their distances are unknown .
the properties of the w51 and 68km0.2em s@xmath7 clouds are summarized in table [ tbl-1 ] .
the cloud size represents a visual estimate of the extent of the detectable @xmath1co emission along the major ( @xmath37 ) and minor ( @xmath38 ) axis of the cloud .
the cloud line width , @xmath39 , is the full width at half maximum of the sum of the gaussian fits comprising the respective clouds .
two estimates of the cloud mass are provided in table [ tbl-1 ] .
the virial mass , @xmath40 , was calculated using the expression @xmath41 , where @xmath42 is the full width at half maximum line width in kilometers per second and @xmath43 is the cloud radius ( @xmath44 ) in parsecs at the zero intensity level .
while the cloud size is actually measured at a finite @xmath1co intensity level , we find it unlikely that the clouds are appreciably larger at lower intensities , and no correction to the observed cloud size was applied .
note that the above expression for the virial mass is appropriate for a uniform density , spherical cloud .
the equivalent expression for a @xmath45 and @xmath46 density cloud would decrease the constant factor in the virial mass expression to 188 and 125 respectively .
a second mass estimate can in principal be obtained from the @xmath1co and @xmath2co data using the lte analysis ( @xcite ) . however , in blended regions ,
it is often difficult to associate @xmath1co gaussian fits with analogous @xmath2co features . therefore , the h@xmath11 column densities were estimated by applying a constant conversion factor to the @xmath1co integrated intensities ( @xcite ; @xcite ; @xcite ) . to ensure that the galactic conversion factor is indeed valid for the w51 region , the conversion factor was estimated from lines of sight with unblended @xmath1co and @xmath2co lines in the 5671km0.2em s@xmath7 velocity interval that defines the w51 and 68km0.2em s@xmath7 clouds .
the h@xmath11 column densities for these lines of sight were estimated using the procedure outlined by dickman ( 1978 ) and assuming a @xmath2co / h@xmath11 abundance of 1.5 x 10@xmath47 .
a histogram of the ratio of the h@xmath11 column densities to @xmath1co integrated intensities is strongly peaked with a mean value of 1.7 x 10@xmath48 @xmath49 ( kkm0.2em s@xmath7)@xmath7 , and is similar to the conversion factor that has been derived for the galaxy ( between @xmath9 2 - 3 x 10@xmath48 @xmath49 ( kkm0.2em s@xmath7)@xmath7 ; @xcite and references therein ) .
the masses ( @xmath50 ) computed from the @xmath1co integrated intensities were calculated by adopting the conversion factor derived from the w51 data , and include a multiplicative factor of 1.36 to incorporate the mass contribution from heavier elements .
the results summarized in table [ tbl-1 ] indicate that the w51 molecular cloud has a mean diameter of @xmath997 pc and a mass cloud and several other velocity components that we did not associate with the w51 molecular cloud .
] of @xmath910@xmath51 .
the similarity in the two mass estimates indicate the gravitational potential energy is approximately equal to the kinetic energy , and that self
gravity must play a critical role in the evolution of the w51 molecular cloud .
the 68km0.2em s@xmath7 cloud is 136 pc in length but only 22 pc wide over most of the minor axis .
such an obvious departure from spherical symmetry renders the virial mass estimate suspect , and so the mass of the 68km0.2em s@xmath7 cloud was estimated only using the @xmath1co conversion factor .
the mass obtained , @xmath910@xmath52 , is an order of magnitude less than that of the w51 cloud .
a comparison of the w51 cloud properties with the size and mass spectrum of molecular clouds in the galaxy ( e.g. @xcite ; @xcite ) shows that the w51 molecular cloud is one the largest gmcs in the galactic disk . among the @xmath95000 molecular clouds with diameters in excess of @xmath922 pc ( and corresponding masses @xmath910@xmath52
; i.e. gmcs ) , the w51 gmc is in the upper 1% of the clouds by size and the upper 510% by mass . for the 68km0.2em s@xmath7 cloud , although its mass is typical of a relatively low mass gmc , a distinguishing feature is its shape .
the ratio of the major to minor axis of the 68km0.2em s@xmath7 cloud is @xmath96 . in the cloud catalog by solomon et al .
( 1987 ) , 85% of the clouds have an aspect ratio less than 2 , and only one object has an aspect ratio greater than observed for the 68km0.2em s@xmath7 cloud .
clearly an elongated shape over such a large length scale is unusual in the molecular interstellar material and may indicate that the 68km0.2em s@xmath7 cloud is a transient structure originating from a relatively recent dynamical event .
now that the individual molecular clouds in the w51 region have been identified , their relationship to the massive star forming sites can be explored .
three images of the w51 region are shown in figure [ fig6 ] : a map of the integrated @xmath1co(10 ) intensity generated from gaussian fits with mean velocities between 56 and 71km0.2em s@xmath7 ( i.e. the w51 and 68km0.2em s@xmath7 clouds ) , an image of the 60 emission from the iras sky survey atlas , and a @xmath1321 cm radio continuum image ( @xcite ) .
the bulk of the 60 emission is elongated parallel to , but slightly below , the galactic plane , and is coincident with bright radio continuum emission . of the sources labeled in the @xmath1321 cm continuum map ,
w51c has predominantly a non thermal continuum spectrum and is thought to be a supernova remnant ( @xcite ) , while the other sources have thermal spectra and are compact hii regions .
the radio continuum sources g49.4 - 0.3 and g49.5 - 0.4 are classically referred to as w51a ( @xcite ) , with the g49.5 - 0.4 region containing the infrared source w51irs1 ( @xcite ) and the @xmath36 masers w51north , w51south , and w51main ( @xcite ) .
sources g48.9 - 0.3 , g49.1 - 0.4 , and g49.2 - 0.4 are collectively known as w51b ( @xcite ) .
the observed radio continuum fluxes ( @xcite ) and far infrared luminosities ( @xmath9 10@xmath53 ; @xcite ; @xcite ) imply the presence of one or more o stars in each of these regions .
figure [ fig6 ] shows that many of the radio continuum sources have corresponding peaks in the molecular line and far infrared images .
in addition to the spatial coincidences , these radio continuum sources have recombination line velocities ( @xcite ) similar to the velocity components identified from the @xmath1co observations . the g49.5 - 0.4 hii region has a recombination line velocity of 59km0.2em s@xmath7 and is spatially coincident with the strong @xmath1co emission associated with the w51 molecular cloud .
the recombination line velocities toward g48.9 - 0.3 , g49.1 - 0.4 , and g49.2 - 0.4 are 66 , 72 , and 66km0.2em s@xmath7 respectively and are located along the 68km0.2em s@xmath7 molecular cloud .
finally , the g49.4 - 0.3 hii region has a recombination line velocity of 53km0.2em s@xmath7 and is coincident with the bright @xmath1co and @xmath2co emission from the 53km0.2em s@xmath7 molecular cloud .
( note that the molecular gas associated with this hii region is outside the velocity range that defines the w51 gmc . ) hi ( @xcite and h@xmath54co ( @xcite ) spectra toward this source exhibit absorption features at velocities of 63 - 65km0.2em s@xmath7 , indicating that g49.4 - 0.3 must be located behind the w51 gmc .
however , these observations do not indicate whether this source is situated just beyond w51 and hence is physically related to the hii region complex , or if it is a distant , unrelated background massive star forming site .
the molecular line maps presented here provide no compelling reason for ( or against ) such an association . to search for any additional massive star forming sites in the w51 molecular cloud ,
the iras point source catalog was examined for objects that have at least two `` high '' quality detections among the four iras bands and a rising spectral energy distribution toward longer wavelengths .
these criteria were designed to select a reliable sample of objects with far - infrared colors characteristic of embedded star forming regions ( @xcite ) . of the 40 iras point sources within the mapped region that meet these criteria ,
the seven brightest objects at 25 are located along the interface between the w51 and the 68km0.2em s@xmath7 clouds .
visual inspection of iras 60 image in figure [ fig6 ] confirms that the brightest sources are located along this ridge . assuming that the other point sources are located at the distance of the w51 molecular cloud ( although many of them are almost certainly foreground objects ) , point sources found away from this interface have far - infrared luminosities in the four iras bands less than 40,000 @xmath55 and inferred spectral types later than zams b0.5 ( @xcite ) . thus embedded o type stars in the w51 region currently are confined to the southern extreme of the molecular cloud .
the extreme star formation characteristics of the w51 region raises the question as to whether the star formation activity stems from unusual _ global _ properties in the w51 molecular cloud , or from unusual conditions found _ local _ to the massive star forming sites .
these possibilities can be explored by comparing the w51 cloud properties with other star forming regions that have also been extensively mapped in @xmath1co and @xmath2co .
in particular , we shall compare the w51 and 68km0.2em s@xmath7 clouds with the molecular clouds associated with the hii regions sh 140 , sh 155 , sh 235 , sh 247 , sh 252 , and sh 255 which have been mapped with the same receiver and telescope used for the w51 observations ( @xcite ; @xcite ) .
the embedded high mass stellar content in this comparison sample is generally limited to a single early
b / late o type star and is in stark contrast to the cluster of o stars forming in the w51 molecular cloud .
the clouds in the comparison sample have diameters of @xmath92055 pc and masses of @xmath9@xmath56 to @xmath57 .
thus the w51 molecular cloud is @xmath152.5 times larger and @xmath1510 times more massive than these objects .
note that the molecular clouds associated with sh 247 , sh 252 , and sh 255 are distinct regions within the gem ob1 cloud complex , which has a total mass of @xmath58 and a diameter of @xmath9150 pc ( @xcite ) . while the spatial size of the gem ob1 complex is larger than the w51 cloud , the w51 cloud exhibits a continuous structure @xmath9100 pc in size as opposed to the fragmentary appearance of the gem ob1 complex .
the 68km0.2em s@xmath7 cloud is also more massive than the objects in the comparison sample and is significantly more elongated than any of the clouds considered here .
thus the w51 and 68km0.2em s@xmath7 clouds are at the extreme in terms of cloud masses and sizes compared to objects in this sample . with the possible exception of the 68km0.2em s@xmath7 cloud , these clouds are similar though in that they appear to be gravitationally bound .
the size and mass of a cloud are not necessarily the key parameters that control the star formation activity within the cloud .
intuitively , one might expect that the amount of matter above a critical density to be the critical variable for otherwise similar clouds . thus if the massive star formation activity in w51 has resulted from the large scale collapse of the cloud
, one would expect the volume and column density densities to be larger than found in a typical cloud . as an indirect measure of the h@xmath11 column densities , figure [ fig7 ] shows histograms of the observed @xmath2co integrated intensities for each of the clouds in our sample .
the lowest integrated intensity shown for any cloud is 3 kkm0.2em s@xmath7 since that is approximately the highest 3@xmath59 detection limit among the various @xmath2co surveys .
comparison of the @xmath1co and @xmath2co intensities suggests that the @xmath2co emission is optically thin if the @xmath1co and @xmath2co excitation temperatures are equal as assumed in the lte analysis ( see @xcite ) .
therefore , the distribution of @xmath2co integrated intensities should accurately trace the h@xmath11 column density distributions as along as the @xmath2co abundance is roughly constant within a cloud .
the @xmath2co integrated intensities in figure [ fig7 ] can be converted to h@xmath11 column densities for an assumed @xmath2co / h@xmath11 abundance of 1.5 x 10@xmath47 ( @xcite ) with the formula @xmath60 for an excitation temperature of @xmath61 = 10 k , the 3 kkm0.2em s@xmath7 integrated intensity limit imposed for figure [ fig7 ] corresponds to an h@xmath11 column density of 1.5 x 10@xmath62 @xmath49 , or an visual extinction of @xmath91.5@xmath63 ( @xcite ) .
variations in the @xmath2co abundance and excitation conditions will obviously effect the absolute conversion from @xmath2co integrated intensities to h@xmath11 column densities , and the comparisons here are intended to search for large differences ( factors of several or more ) in the typical column densities in these clouds .
figure [ fig7 ] shows that the distributions of @xmath2co integrated intensities peak near the detection limit of 3 kkm0.2em s@xmath7 for each cloud , with a long tail toward the higher integrated intensities .
the tail of these distributions correspond to the high column density regions and are often associated with star forming sites .
the mean @xmath2co integrated intensity among the clouds varies between 4.9 and 9.7 kkm0.2em s@xmath7 , with the w51 cloud containing the third highest mean intensity and the 68km0.2em s@xmath7 cloud the highest .
the high values found for the 68km0.2em s@xmath7 cloud may be a result of an unusual viewing angle , as either this cloud is a sheet of gas observed edge on or a long , narrow filament .
these results imply that most of the mass in each cloud is contained in lines of sight with column densities corresponding to less than a few magnitudes of visual extinction .
thus the w51 gmc is similar to other clouds in that the diffuse envelope contains more mass than the high column density cores .
while the w51 cloud contains a higher column density on average than the other clouds , this can attributed to the fact that it is more than twice as large as some clouds in the sample .
indeed , assuming that the @xmath64 w51 cloud is distributed in a sphere of diameter of 97 pc ( see table [ tbl-1 ] ) , the average h@xmath11 volume density is 40 @xmath65 , comparable to the volume density inferred in nearby molecular clouds ( @xcite ; @xcite ) .
this suggests that the entire w51 molecular is probably not in an advanced stage of collapse , and that the intense star formation activity in w51 likely results from forces acting on a localized region .
in retrospect , this is perhaps not surprising given that the massive star forming regions in w51 are located at the edge of the cloud and not in the center as expected if , for example , the entire cloud was systematically collapsing .
contrary to the global properties , the gas properties local to the w51 massive star forming regions do appear to be unusual compared to the molecular clouds in the solar neighborhood .
submillimeter continuum observations have shown that the core containing the g49.5 - 0.4 hii region contains @xmath9 10@xmath52 of gas with a mean h@xmath11 volume density of @xmath9 5 x 10@xmath66 @xmath65 over a 3 pc radius region ( @xcite ) .
by contrast , submillimeter observations indicate that the most massive cores in nearby molecular clouds typically have masses @xmath923 orders of magnitude less than that of the g49.5 - 0.4 core ( e.g. @xcite ; @xcite ; @xcite ; @xcite ) .
while the g49.5 - 0.4 core does not necessarily contain higher gas densities , it does contain more gas at the densities needed to form stars .
the above discussion suggests that the key to understanding the massive star formation activity in the w51 complex is determining the forces that acted on a localized region within the cloud .
the molecular line maps presented here allow us to speculate on what these forces may be .
as shown in figures [ fig3 ] and [ fig4 ] , the diffuse emission from the w51 gmc truncates at the location of the 68km0.2em s@xmath7 cloud for velocities @xmath15 63km0.2em s@xmath7 .
this morphology was used in [ analysis ] to argue that the w51 and 68km0.2em s@xmath7 clouds are at a common distance since such an interface is unlikely to result from the chance superposition of unrelated molecular clouds .
one way such an interface could form is if the 68km0.2em s@xmath7 cloud has collided into the w51 gmc ( see also @xcite ; @xcite )
. molecular gas does extend below the 68km0.2em s@xmath7 cloud at lower velocities , however . in this picture ,
given the three dimensional structure of the w51 gmc , this material has not crossed the path of the 68km0.2em s@xmath7 cloud .
both h@xmath11co and hi spectra toward the g49.5 - 0.4 hii region ( @xcite ; @xcite ; @xcite ) contain absorption lines at velocities @xmath15 65km0.2em s@xmath7 , and relative to the line of sight , the 68km0.2em s@xmath7 cloud must be located in front of the w51 gmc . since the velocity difference between the two clouds ( at least 5 km0.2em s@xmath7 ) is larger than the sound speed , a shock front will form that will compress the molecular gas and possibly induce star formation ( @xcite ) .
the qualitative model of two colliding clouds accounts for several properties of the w51 region .
first , one would expect that star formation should occur preferentially along the interface region .
the iras image in figure [ fig6 ] shows that this is indeed the case for massive stars , as the northern half of the w51 gmc appears devoid of embedded o type stars despite containing most of the cloud mass .
the cloud collision model would also suggest that the massive star forming regions along the collision interface should have a common age . while the ages of these stars are not known , the lifetime of the mid o type stars found in the w51 region ( @xcite
; see also @xcite ) places an upper limit to the stellar ages of @xmath95 myr ( @xcite )
. the actual ages may be considerably less since these o stars are still in the compact hii region phase , which has a lifetime of less than 1 myr ( @xcite ; @xcite ) .
further support for a cloud cloud collision model comes from considering the projected distance ( 73 pc ) between the two furthest separated star forming regions along the w51/68km0.2em s@xmath7 cloud interface .
the sound travel time across this distance is nearly two orders of magnitude larger than the o star lifetime .
therefore , these star forming regions must have been created by either a single event operating along the entire southern edge of the cloud or from up to 4 separate , but nearly simultaneous , events .
given the scarcity of embedded o stars in the galaxy , the cloud collision model would provide a natural explanation for simultaneous star formation along the ridge .
ultimately , the suggested collision between the w51 and 68km0.2em s@xmath7 clouds may be related to a spiral density wave .
the anomalous gas velocities in the w51 region have long been attributed to streaming motions in the sagittarius spiral arm ( @xcite ; @xcite ) . the associated spiral density wave may have indirectly led to the massive star formation activity in the w51 region by enhancing the number density of clouds and increasing the probability of a cloud - cloud collision .
further , a spiral wave shock , if present ( see @xcite ) , will compress and flatten any clouds ( @xcite ) .
indeed , such a shock could account for the highly elongated shape of the 68km0.2em s@xmath7 cloud .
the w51 gmc , however , remains roughly circular in shape , and globally its evolution is likely still dominated by self gravity . finally , we briefly consider the implications of these results for other star forming regions . while the w51 star forming region exceeds all nearby embedded star forming sites in terms of the number of o stars , bolometric luminosity , and dense core mass , w51 itself is dwarfed by some star forming regions in nearby galaxies .
most notably , the 30 doradus region in the large magellanic cloud contains an order magnitude more o stars than w51 ( @xcite ) . at larger distances ,
many interacting galaxies contain even yet more vigorous massive star forming regions that may be dense enough to represent young globular clusters ( @xcite ) .
most of the molecular gas appears to have been dispersed in these clusters already , and in any event , the distances to these systems precludes any detailed studies of the natal clouds .
the most significant piece of information afforded by the w51 molecular maps is that despite containing one of the most massive dense cores known in our galaxy , most of the mass in the w51 molecular cloud is not currently forming massive stars .
thus it is easy to imagine that the increase in the number of cloud cloud collisions that presumably results in interacting galaxies may lead to a larger number of star forming regions throughout a single cloud or more intense star formation activity within a small region .
in fact , the mass within the w51 molecular cloud is comparable to that in globular clusters , and it may not require the conglomeration of many giant molecular clouds to form these stellar systems , but the large scale collapse of a single gmc .
we have mapped a 1.39 x 1.33 region toward the w51 hii region complex at 45 - 47 resolution and 50 sampling in the j=10 transitions of @xmath1co and @xmath2co . from these data
we have identified the major molecular clouds and have associated these clouds with the massive embedded star forming sites in the w51 region .
we find that : \(1 ) the two most prominent clouds in the w51 region are the 58 - 6063km0.2em s@xmath7 cloud ( defined as the w51 gmc ) and the 68km0.2em s@xmath7 cloud .
the w51 gmc is associated with the brightest @xmath136 cm continuum source in w51 ( g49.5 - 0.4 ) , and the 68km0.2em s@xmath7 cloud contains the hii regions g48.9 - 0.3 ,
g49.1 - 0.4 , and g49.2 - 0.4 . published absorption line spectra ( @xcite ; @xcite ) indicate that the fifth bright hii region in the area , g49.4 - 0.3 , must be located behind the w51 gmc , but it remains unclear whether or not it is physically associated with the other star forming regions .
\(2 ) the mass of the w51 gmc and the 68km0.2em s@xmath7 clouds are @xmath9 1.2@xmath310@xmath51 and @xmath67 respectively .
the w51 molecular cloud is roughly circular in shape with a mean diameter of @xmath997 pc and appears to be gravitationally bound .
compared to the @xmath95000 gmcs in the galactic disk , w51 is among the top 1% by size and the top 510% in terms of cloud mass .
the 68km0.2em s@xmath7 cloud is an elongated filament of @xmath9 136pc@xmath322pc in size
. the 6:1 aspect ratio of the major and minor axis in the 68km0.2em s@xmath7 are rare in the galaxy over such a large size scale ( @xcite ) , and suggests that this molecular cloud may represent a transient feature .
\(3 ) the properties of the w51 and 68km0.2em s@xmath7 clouds are compared with nearby clouds that have been studied in a similar manner but contain lower levels of massive star formation activity . while the w51 cloud is larger and more massive than nearby clouds , the mean h@xmath11 column density is not unusual given the large size , and the mean h@xmath11 volume density is comparable . the w51 gmc is similar to other clouds in that most of the molecular mass
is contained in a diffuse molecular envelop that is not forming massive stars .
the 68km0.2em s@xmath7 cloud contains the largest mean column density among the clouds studied here , but this may be a result of an unusual viewing angle for this elongated cloud .
we suggest that much of the star formation activity in the w51 region has not resulted from global collapse of the w51 cloud , but from forces acting on localized regions within the cloud .
\(4 ) we speculate that much of the massive star formation activity in w51 has resulted from a collision between the w51 and 68km0.2em s@xmath7 molecular clouds .
this conjecture can explain the string of embedded o stars that are spread out for 70 pc along the interface between the w5168km0.2em s@xmath7 clouds , and why massive star formation is currently confined to the southern ridge of the w51 gmc .
we would like to thank mark heyer for completing the @xmath2co map of w51 and bon chul koo for making available his @xmath1321 cm continuum image of w51 .
jmc acknowledges support from the james clerk maxwell telescope fellowship .
dbs was supported in part by nasa grant nagw-3938 .
the five college radio astronomy observatory is operated with support from nsf grant 9420159 .
arnal , e. m. , & goss , w. m. 1985 , , 145 , 369 bachiller , r. & cernicharo , j. 1986 , , 166 , 283 bally , j. , stark , a. a. , wilson , r. w. , & langer , w. d. 1987 , , 312 , l45 bieging , j. 1975 , in lecture notes in physics 42 , hii regions and related topics , ed . t. l. wilson and d. downes ( berlin : springer ) , 443 blitz , l. 1991 , in the physics of star formation , eds . c. j. lada & n. kylafis , ( dordrecht : kluwer ) , 1 bohlin , r. c. , savage , b. d. , & drake , j. f. 1978 , , 224 , 132 brand , j. , & blitz , l. 1993 , , 275 , 67 burton , w. b. 1970 , , 2 , 291 carpenter , j. m. , snell , r. l. , & schloerb , f. p. 1995
, , 445 , 246 churchwell , e. 1990 , , 2 , 79 comern , f. & torra , j. 1996 , , 314 , 776 dame , t. m. , ungerechts , h. , cohen , r. s. , de geus , e. j. , grenier , i. a. , may , j. , murphy , d. c. , nyman , l .- , & thaddeus , p. 1987
, , 322 , 706 dickman , r. l. 1975 , , 202 , 50 dickman , r. l. 1978 , , 37 , 407 elmegreen , b. g. , & elmegreen , d. m. 1978 , , 220 , 1051 erickson , n. r. , goldsmith , p. f. , novak , g. , grosslein , r. m. , viscuso , p. j. , erickson , r. b. , & predmore , c. r. 1992 , ieee trans . on microwave theory and techniques , 40 , 1 genzel , r. , & downes , d. 1977 , , 30 , 145 genzel , r. , downes , d. , schneps , m. h. , reid , m. j. , moran , j. m. , kogan , l. r. , kostenko , v. i. , matveyenko , l. i. , & r@xmath68nn@xmath69ng , b. 1981 , , 247 , 1039 georgelin , t. m. , & georgelin , y. p. 1976
, , 49 , 57 goss , w. m. , & shaver , p. a. 1970 , australian j. phys , ap .
, 14 , 1 harper , d. a. , & low , f. j. 1971 , , 165 , l9 harvey , p. m. , joy , m. , lester , d. f. , & wilking , b. a. 1986 , , 300 , 737 heyer , m. h. , carpenter , j. m. , & ladd , e. f. 1996 , , 463 , 630 hoffman , w. f. , frederick , c. l. , & emery , r. j. 1971 , , 170 , l89 jaffe , d. t. , davidson , j. a. , dragovan , m. , & hildebrand , r. h. 1984 , , 284 , 637 koo , b. 1997 , , 108 , 489 koo , b. , & moon , d. 1997 , , 475 , 194 kundu , n. r. , & velusamy , t. 1967 , ann . d@xmath70astrophys .
30 , 59 kutner , m. l. , & ulich , b. k. 1981 , , 250 , 341 lubow , s. h. , balbus , s. a. , & cowie , l. l. 1986 , , 309 , 496 mehringer , d. m. 1994 , , 91 , 713 meynet , g. , maeder , a. , schaller , g. , schaerer , d. , & charbonnel , c. 1994 , , 103 , 97 mezger , p. g. , sievers , a. w. , haslam , c. g. t. , kreysa , e. , lemke , r. , mauersberger , r. , & wilson , t. l. 1992 , , 256 mufson , s. l. , & liszt , h. s. 1979 , , 232 , 451 nakamura , t. , kodaira , s. , ishi , k. , inatani , j. , & ohishi , m. 1984 , , 36 , 123 ohishi , m. , nakamura , t. , kodaira , s. , ishi , k. , & inatani , j. 1984 , , 36 , 505 oldham , p. g. , griffin , m. j. , richardson , k. j. , & sandell , g. 1994 , , 284 , 559 pankonin , v. , payne , h. e. , & terzian , y. 1979 , , 75 , 365 rengarajan , t. n. , cheung , l. h. , fazio , g. g. , shivanandan , k. , mcbreen , b. 1984 , , 286 , 573 sanders , d. b. , clemens , d. p. , scoville , n. z. , & solomon , p. m. 1986 , , 60 , 1 sanders , d. b. , scoville , n. z. , & solomon , p. m. 1985 , , 289 , 373 schloerb , f. p. , snell , r. l. , & schwartz , p. r. 1987 , , 319 , 426 scoville , n. z. & solomon , p. m. 1973 , , 180 , 31 scoville , n. z. , yun , m. s. , clemens , d. p. , sanders , d. b. , & waller , w. h. 1987 , , 63 , 821 shane , w. w. & bieger - smith , g. p. 1966
, , 18 , 263 sievers , a. w. , mezger , p. g. , gordon , m. a. , kreysa , e. , haslam , c. g. t. , & lemke , r. 1991 , , 251 , 231 solomon , p. m. , rivolo , a. r. , barrett , j. , & yahil , a. 1987 , , 319 , 730 strong , a. w. , bloemen , j. b. g. m. , dame , t. m. , grenier , i. a. , hermsen , w. , lebrun , f. , nyman , l. . , pollock , a. m. t. , & thaddeus , p. 1988
, , 207 , 1 subrahmanyan , r. & goss , w. m. 1995 , , 275 , 755 vacca , w. d. , robert , c. , leitherer , c. , & conti , p. s. 1995 , , 444 , 647 vacca , w. d. , garmany , c. d. , & shull , j. m. 1996 , , 460 , 914 walker , h. j. , cohen , m. , volk , k. , wainscoat , r. j. , & schwartz , d. e. 1989 , , 98 , 2163 westerhout , g. 1958 , , 14 , 215 whitmore , b. c. , schweizer , f. , leitherer , c. , borne , k. , & robert , c. 1993 , , 106 , 1354 wilson , t. l. , mezger , p. g. , gardner , f. f. , & milne , d. k. 1970 , , 5 , 99 woodward , p. r. 1976 , , 207 , 484 wynn - williams , c. g. , becklin , e. e. , & neugebauer , g. 1974 , , 187 , 473 | we present 45 - 47 angular resolution maps at 50@xmath0 sampling of the @xmath1co and @xmath2co j=1 - 0 emission toward a 1.39 @xmath3 1.33 region in the w51 hii region complex .
these data permit the spatial and kinematic separation of several spectral features observed along the line of sight to w51 , and establish the presence of a massive ( @xmath4 ) , large ( @xmath5 ) giant molecular cloud ( gmc ) , defined as the w51 gmc , centered at ( @xmath6km0.2em s@xmath7 ) . a second massive ( @xmath8 ) ,
elongated ( 136 pc @xmath3 22 pc ) molecular cloud is found at velocities of @xmath968km0.2em s@xmath7 along the southern edge of the w51 gmc .
of the five radio continuum sources that classically define the w51 region , the brightest source at @xmath10 cm ( g49.5 - 0.4 ) is spatially and kinematically coincident with the w51 gmc and three ( g48.9 - 0.3 , g49.1 - 0.4 , and g49.2 - 0.4 ) are associated with the 68km0.2em s@xmath7 cloud . published absorption line spectra indicate that the fifth prominent continuum source ( g49.4 - 0.3 ) is located behind the w51 molecular cloud .
the w51 gmc is among the upper 1% of clouds in the galactic disk by size and the upper 510% by mass .
while the w51 gmc is larger and more massive than any nearby molecular cloud , the average h@xmath11 column density is not unusual given its size and the mean h@xmath11 volume density is comparable to that in nearby clouds . the w51 gmc is also similar to other clouds in that most of the molecular mass is contained in a diffuse envelope that is not currently forming massive stars .
we speculate that much of the massive star formation activity in this region has resulted from a collision between the 68km0.2em s@xmath7 cloud and the w51 gmc .
10000 10000 |
molecular hydrogen is the most abundant molecular species in the universe , but its spectrum is not so well suited for direct observations , because the strong dipole - allowed transitions in the lyman ( @xmath7 ) and werner ( @xmath8 ) systems fall in the range of the vacuum ultraviolet ( vuv ) , hence at wavelengths for which the earth s atmosphere is opaque . the first observation of molecular hydrogen in space was reported in 1970 @xcite using a rocket borne spectrometer observing from high altitudes , therewith evading atmospheric absorption of the vuv radiation .
subsequently , absorption lines in the vuv - spectrum of molecular hydrogen were observed with the _ copernicus _ satellite telescope @xcite .
later the vuv - absorption spectrum of h@xmath5 was observed with the _ far ultraviolet spectroscopic explorer _
@xcite and the _ hubble space telescope _
@xcite , all satellite - based instruments .
spectra of molecular hydrogen in highly redshifted objects , e.g. in the line of sight of quasars , can be observed with ground - based telescopes provided the redshift of the absorbing cloud is at @xmath9 .
early observations from earth with tentative assignment of molecular hydrogen were performed by @xcite .
surveys of damped - lyman-@xmath10 systems in search for pronounced h@xmath5 absorption features were conducted by @xcite .
the molecular hydrogen spectra in highly redshifted objects were analyzed and compared with classical @xcite and laser - based laboratory investigations of the lyman bands @xcite , initially resulting in some indications for a possible variation of the proton - electron mass ratio ( @xmath11 ) at cosmological redshifts @xcite .
further detailed and accurate studies of molecular hydrogen at high redshift resulted in constraints on a varying proton - electron mass ratio @xcite now yielding an overall limit of @xmath12 @xcite .
fundamental physics theories predict variation of fundamental constants on cosmological temporal and spatial scales , but the couplings between light scalar fields causing those variations may also generate dependencies of fundamental constants on local gravitational fields @xcite .
the latter phenomena were explored recently in search of a possible dependence of the fine structure constant @xmath10 @xcite and the proton - electron mass ratio @xmath13 @xcite in the strong gravitational fields prevailing in the photosphere of white dwarf stars .
the search for a @xmath13-dependence on a gravitational potential as strong as 10@xmath14 times that at the earth s surface was based on h@xmath5 lyman band absorptions in white dwarfs gd133 and gd29 - 38 initially discovered by @xcite , and reanalyzed by @xcite . for an analysis of the h@xmath5 absorption spectrum in the hot environment of white dwarf photospheres , at typical temperatures of 12,000 - 14,000 k ,
accurate molecular physics data are required on excitations from highly excited ro - vibrational states in the lyman band system .
it is the purpose of the present paper to collect and re - evaluate the information on laboratory wavelengths @xmath2 for the relevant lyman bands , and perform a calculation on sensitivity coefficients @xmath1 to @xmath13-variation for the relevant lines .
together with existing information on line oscillator strengths for the transitions @xmath3 and natural lifetime parameters @xmath4 for the upper states @xcite a constraining value for @xmath15 can be derived in the presence of the strong gravitational fields near the surface of white dwarfs . in section [ moldat ] an evaluation of the wavelengths @xmath2
will be performed , which will be listed alongside with values for @xmath3 and @xmath4 . in section [ ksec ]
a calculation of sensitivity coefficients @xmath1 will be presented , following different methods , one based on the semi - empirical analysis of mass dependencies of level energies , and one based on an _ ab initio _ calculation .
the combined molecular physics information on h@xmath5 will be applied in section [ wd ] for the analysis of white dwarf spectra of gd29 - 38 and gd133 , observed with the cosmic origins spectrograph on the hubble space telescope ( cos - hst ) .
parts of the absorption spectra of the photospheres of white dwarfs gd29 - 38 ( wd2327 + 049 ) and gd133 ( wd1116 + 026 ) obtained with cos - hst are plotted in figs .
[ wd-29 - 38 ] and [ wd-133 ] .
the observed spectra @xcite extend over the range 1144 - 1444 , with @xmath16@xmath17 lyman band lines unambiguously assignable in the window 12981444 , and tentative indications of weak @xmath18@xmath17 werner band transitions in the range 11441290 .
the spectra reflect the high temperatures in the white dwarf photospheres , by the multiple vibrationally and rotationally excited levels of the ground electronic state .
the most intense lines belong to the @xmath19 bands of the lyman system .
lines in the lyman bands are indicated by the ( blue ) vertical sticks .
the regions dominated by ci lines ( from the white dwarf photosphere ) and cii lines ( from the interstellar medium ) are left out of the model calculations . , width=321 ] at the high temperatures in white dwarf photospheres populations of vibrational levels up to @xmath20 must be considered .
based on the calculations of franck - condon factors @xcite it can be deduced what bands in the lyman system produce most of the absorption intensity . whereas for excitation from @xmath21 a sequence of vibrational bands contributes to the absorption , in case of excitation from @xmath22 the franck - condon factors yield most of the strength for excitation to @xmath23 , while in case of excitation from @xmath24 excitation to a single @xmath25 level outweighs all other contributions by over an order of magnitude .
this pattern is reflected in the calculations of individual line oscillator strengths @xmath3 for the lyman bands as performed by @xcite .
similarly , high rotational angular momentum states are populated , peaking at @xmath26 for temperatures of 12,000 k , while transitions originating from states up to @xmath27 are found @xcite to contribute to the absorption in the spectra shown in figs .
[ wd-29 - 38 ] and [ wd-133 ] . in order to model the spectra at the highest accuracy possible the wavelengths @xmath2 of the relevant lyman absorption lines have to be included in a model .
the most accurate wavelengths follow from listed level energies for excited @xmath28 and ground @xmath29 quantum states .
starting from level energies is the superior approach over using listed experimental wavelengths , since those level energy values represent averages over several well - calibrated spectral lines .
this approach was followed in the two - step determination of @xmath28 level energies @xcite , first measuring anchor line energies of some intermediate states in h@xmath5 @xcite .
a subsequent study of transitions between excited states , e.g. the @xmath30 system , provides the high accuracy on @xmath31 level energies from the large data base of lines measured in a broad wavelength interval , including the infrared .
these combined procedures resulted in level energies for @xmath32 up to @xmath33 , for @xmath34 up to @xmath26 , and for @xmath35 up to @xmath36 at accuracies generally at 0.001 @xmath37 or better @xcite .
these precision experiments do not extend to very high rotational quantum numbers , needed for the analysis of the white dwarf spectra .
for those higher rotational states the most accurate values are those of @xcite , where level energies are determined from analysis of thousands of lines observed in lyman band emissions obtained from a low - pressure discharge @xcite .
the values of @xmath38 level energies were calculated up to @xmath39 in an _ ab initio _ model in which the input potential curves were slightly adapted to represent the experimental data at an accuracy of 0.1 @xmath37 .
these calculations were further tested in a vuv - laser - based experiment performed in a plasma in which high ro - vibrational levels in the @xmath40 system were probed , yielding agreement within the estimated uncertainties of a few 0.1 @xmath37 @xcite . the comparison of the level energies from @xcite and @xcite for low @xmath41 quantum numbers yield good agreement , but with a noticeable systematic offset that is still within the estimated uncertainty of the latter . based on this comparison ,
an energy correction of between 0.04 - 0.06 @xmath37 was applied for each vibrational series in the dataset of @xcite .
the combined dataset comprising the low @xmath41 level values from @xcite and the ( corrected ) high @xmath41 level energies from @xcite was employed in the calculation of the transition wavelengths for the analysis .
similar procedures were also applied to the level energies of the @xmath42 state where the low @xmath41 level energies were taken from @xcite . for the high @xmath41 levels , energy corrections of between 0.04 - 0.13
@xmath37 was applied for the vibrational series of the @xmath43 states , and 0.06 - 0.11 @xmath37 for the @xmath44 states from @xcite . for the ground state
one may rely on the full - fledged _ ab initio _ calculations of all 302 bound @xmath29 quantum levels in the ground state of h@xmath5 .
those values have been calculated , including born - oppenheimer energies , adiabatic and non - adiabatic corrections as well as relativistic and quantum - electrodynamic contributions by @xcite .
their accuracy at 0.001 @xmath37 has been tested in a number of precision experimental studies on the ground state level energies @xcite well verifying the claimed accuracy .
wavelengths for lines in the @xmath45 lyman bands are calculated from these data sets on level energies for excited and ground states .
values for the wavelengths and their accuracies , varying between 0.001 @xmath37 and 0.1 @xmath37 ( on a frequency / energy scale ) , are listed for vibrational levels @xmath46 and @xmath47 in the supplementary material , lyman lines , about a third is used in subsequent analysis of white dwarf absorption spectra .
similar lists for werner transitions are also provided . ] while a subset of the most intense lines is listed in table [ lyman ] .
the corresponding wavelengths for werner transitions are also listed in supplementary material with a few of the most intense lines listed in table [ werner ] .
the tables contain the molecular physics knowledge of the lyman lines to be observed in the hot environments of white dwarf photospheres ; wavelengths @xmath2 , line oscillator strengths @xmath3 , natural line broadening coefficients @xmath4 @xcite , and the sensitivity coefficients @xmath1 to be discussed in the next section .
-coefficients of some intense lines in the @xmath48branch of h@xmath5 lyman bands .
the ( green ) squares represent the lines in the @xmath49 band typically observed in the cold environment of high redshift objects toward quasars ( although only for @xmath50 ) , while ( red ) triangles and ( blue ) circles represent @xmath51 bands ( most intense for @xmath52 ) are typically observed in the hot photospheres of white dwarfs .
@xmath1-coefficients for most intense werner q - branch lines at @xmath53 k , belonging the @xmath54 bands are also shown . ,
width=302 ] [ cols="^,^,^,>,<,<,<,^,^",options="header " , ] the use of h@xmath5 lyman band lines for constraining @xmath55 may be mimicked by offsets in the wavelength calibration of the astronomical spectra , in particular in cases where a correlation exists between wavelength and sensitivity coefficients @xmath1 .
this possibility was discussed in the context of possible cosmological variation of @xmath13 , or limitations thereof , as probed through h@xmath5 absorption in the colder environments of high redshift galaxies , where typical excitation temperatures of 50 k prevail @xcite . under such conditions only a few rotational states , @xmath50 ,
are populated for a single @xmath56 vibrational level , and the @xmath1-coefficients for the lyman bands strongly correlate with wavelength @xcite . for those studies
the simultaneous use of lyman and werner band absorptions lifts the correlation somewhat , as shown in fig . 4 in @xcite .
for the present case of the lyman absorption bands at 12000 k the situation is much different as exemplified in fig .
[ k - coeffs - wd ] showing the calculated @xmath57-coefficients plotted as a function of wavelength , while the size of the data points reflects the ( relative ) line intensities of the specific lines in the spectrum .
this figure demonstrates that the correlation between @xmath57 and @xmath58 is lifted at each wavelength in the spectrum various lines with differing @xmath1 contribute to the absorption spectrum .
this feature makes the derivation of constraints on @xmath55 more robust .
-coefficients plotted vs. wavelength with relative line intensities as they appear in the spectra of the photospheres of white dwarfs marked as the sizes of the points .
the line intensities refer to the product of line oscillator strengths @xmath59 and the fractional population @xmath60 at a temperature of 12,000 k. , width=321 ]
in this study the molecular physics information on the @xmath61 lyman bands and @xmath62 werner bands of the h@xmath5 molecule is collected and re - evaluated resulting in a listing of wavelengths @xmath2 , line oscillator strengths @xmath3 , and radiative damping factors @xmath4 , while sensitivity coefficients @xmath1 for a possible change of the proton - electron mass ratio @xmath13 are calculated .
this database is implemented for the analysis of h@xmath5 absorption spectra in the photospheres of gd133 and gd29 - 38 white dwarfs stars yielding a constraint on a possible dependence of the proton - electron mass ratio of @xmath63 in a gravitational potential of @xmath64 .
the study focuses on the lyman bands for @xmath65 and @xmath66 , corresponding to the strongest h@xmath5 absorption features , observed with the cosmic origins spectrograph aboard the hubble space telescope .
future spectroscopic studies on similar hot photosphere white dwarfs may cover the shorter wavelength region where @xmath62 werner lines contribute . for that purpose calculations on the molecular physics data on the werner bands are tabulated in the present study . for improving the analysis methods extended data
would be required along the following lines : ( i ) the laboratory wavelengths @xmath2 for lyman and werner lines involving rotational quantum states @xmath67 are only accurate to 0.1 @xmath37 ; laser - based studies could improve those values by two orders of magnitude , and it was shown by @xcite that laser - based studies on very high @xmath41-states are feasible .
( ii ) if white dwarf photospheres with enhanced column densities can be detected , spectroscopic features of the hd molecule should become observable . for a detailed analysis of such environments the molecular physics properties of the lyman and werner bands of hd
should then be investigated , along the lines as pursued by @xcite .
the authors acknowledge support from the netherlands foundation for fundamental research of matter ( fom ) .
we have used data from the cycle 18 program 12290 ( pi m. jura ) observations with the hubble space telescope .
abgrall , h. launay , f. , roueff , e. and roncin , j .- y . j. chem .
phys 87 , 2036 ( 1987 ) .
abgrall , h. roueff , e. , launay , f. , roncin , j .- y . , and subtil , j .- l
ser . 101 , 273 ( 1993 ) .
abgrall , h. roueff , e. , launay , f. , roncin , j .- y . , and subtil , j .- l
ser . 101 , 323 ( 1993 ) .
abgrall , h. , roueff , e. launay , f. , roncin , j .- y . , and
subtil , j .- l . j. mol . spectr .
157 , 512 ( 1993 ) .
abgrall , h. , roueff , e. , launay , f. and roncin , j .- y .
j. phys 72 , 856 ( 1994 ) .
abgrall , h. , roueff , e. , and drira , i. , astron .
141 , 297 ( 2000 ) .
moos , h. w. _ et al .
_ , fuse science team , fuse instrument team and fuse science operation team , astroph . j. lett . 538 , l1 ( 2000 ) .
morton , d. c. and dinerstein , h. l. , astroph .
j. 204 , 1 ( 1976 ) .
niu , m. l. , salumbides , e. j. , dickenson , g. d. , eikema , k. s. e. , and ubachs , w. , j. mol .
300 , 44 ( 2014 ) . | recently an accurate analysis of absorption spectra of molecular hydrogen , observed with the cosmic origins spectrograph aboard the hubble space telescope , in the photosphere of white dwarf stars gd133 and gd29 - 38 was published in a letter [ phys .
rev .
lett .
113 , 123002 ( 2014 ) ] , yielding a constraint on a possible dependence of the proton - electron mass ratio on a gravitational field of strength 10,000 times that at the earth s surface . in the present paper further details of that study
are presented , in particular a re - evaluation of the spectrum of the @xmath0 lyman bands relevant for the prevailing temperatures ( 12,000 - 14,000 k ) of the photospheres .
an emphasis is on the calculation of so - called @xmath1-coefficients , that represent the sensitivity of each individual line to a possible change in the proton - electron mass ratio .
such calculations were performed by semi - empirical methods and by _ ab initio _ methods providing accurate and consistent values .
a full listing is provided for the molecular physics data on the lyman bands ( wavelengths @xmath2 , line oscillator strengths @xmath3 , radiative damping rates @xmath4 , and sensitivity coefficients @xmath1 ) as required for the analyses of h@xmath5-spectra in hot dwarf stars .
a similar listing of the molecular physics parameters for the @xmath6 werner bands is provided for future use in the analysis of white dwarf spectra .
[ firstpage ] white dwarfs : molecules : h@xmath5 white dwarfs : photospheres stars : individual : gd133 , gd29 - 38 |
one of the main goals of string theory is the inclusion of the standard model ( sm ) of particle physics in an ultraviolet complete and consistent theory of quantum gravity .
the hope is a unified theory of all fundamental interactions : gravity as well as strong and electroweak interactions within the @xmath7 sm .
recent support for the validity of the particle physics standard model is the 2012 discovery of the `` so - called '' higgs boson .
how does this fit into known string theory ?
ideally one would have hoped to derive the standard model from string theory itself , but up to now such a program has not ( yet ) been successful .
it does not seem that the sm appears as a prediction of string theory . in view of
that we have to ask the question whether the sm can be embedded in string theory .
if this is possible we could then scan the successful models and check specific properties that might originate from the nature of the underlying string theory .
known superstring theories are formulated in @xmath0 space time dimensions ( or @xmath8 for m theory ) while the sm describes physics in @xmath2 .
the connection between @xmath0 and @xmath2 requires the compactification of six spatial dimensions .
the rather few superstring theories in @xmath0 give rise to a plethora of theories in @xmath2 with a wide spectrum of properties .
the search for the sm and thus the field of so - called `` string phenomenology '' boils down to a question of exploring this compactification process in detail . but
how should we proceed ? as the top - down approach is not successful we should therefore analyse in detail the properties of the sm and then use a bottom - up approach to identify those regions of the `` string landscape '' where the sm is most likely to reside .
this will provide us with a set of `` rules '' for @xmath2 model constructions of string theory towards the constructions of models that resemble the sm .
the application of these rules will lead us to `` fertile patches '' of the string landscape with many explicit candidate models .
given these models we can then try to identify those properties of the models that make them successful .
they teach us some lessons towards selecting the string theory in @xmath0 as well as details of the process of compactification . in the present paper
we shall describe this approach to `` string phenomenology '' . in section [ sec : fivegoldenrules ]
we shall start with `` five golden rules '' as they have been formulated some time ago @xcite .
these rules have been derived in a bottom - up approach exploiting the particular properties of quark- and lepton representations in the sm .
they lead to some kind of ( grand ) unified picture favouring @xmath9 and @xmath10 symmetries in the ultraviolet .
however , these rules are not to be understood as strict rules for string model building
. you might violate them and still obtain some reasonable models .
but , as we shall see , life is more easy if one follows these rules
. in section [ sec : minilandscape ] we shall start explicit model building along these lines .
we will select one of those string theories that allow for an easy incorporation of the rules within explicit solvable compactifications .
this leads us to orbifold compactifications of the heterotic @xmath11 string theory @xcite as an example
. we shall consider this example in detail and comment on generalizations and alternatives later .
the search for realistic models starts with the analysis of the so - called @xmath12-ii orbifold @xcite .
we define the search strategy in detail and present the results known as the `` minilandscape '' @xcite , a fertile patch of the string landscape for realistic model building .
we analyse the several hundred models of the minilandscape towards specific properties , as e.g. the location of fields in extra - dimensional space .
the emerging picture leads to a notion of `` local grand unification '' , where some of the more problematic aspects of grand unification ( gut ) can be avoided .
we identify common properties of the successful models and formulate `` lessons '' from the minilandscape that should join the `` rules '' for realistic model building .
section [ sec : orbifoldlandscape ] will be devoted to the construction of new , explicit mssm - like models using all @xmath13 and certain @xmath14 orbifold geometries resulting in approximately 12000 orbifold models .
then , in section [ sec : rulesorbifoldlandscape ] we shall see how the lessons of the minilandscape will be tested in this more general `` orbifoldlandscape '' . in section [ sec : generallandscape ]
we shall discuss alternatives to orbifold compactifications , as well as model building outside the heterotic @xmath11 string .
the aim is a unified picture of rules and lessons for successful string model building . section [ sec : nvconclusions ] will be devoted to conclusions and outlook .
let us start with a review of the `` five golden rules for superstring phenomenology '' , which can be seen as phenomenologically motivated guidelines to successful string model building @xcite .
the rules can be summarized as follows : we need 1 .
spinors of @xmath10 for sm matter 2 .
incomplete gut multiplets for the higgs pair 3
. repetition of families from geometrical properties of the compactification space 4 .
@xmath15 supersymmetry 5 .
@xmath16-parity and other discrete symmetries let us explain the motivation for these rules in some detail in the following .
it is a remarkable fact that the spinor @xmath18 of @xmath10 is the unique irreducible representation that can incorporate exactly one complete generation of quarks and leptons , including the right - handed neutrino .
thereby , it can explain the absence of gauge - anomalies in the standard model for each generation separately .
furthermore , it offers a simple explanation for the observed ratios of the electric charges of all elementary particles .
in addition , there is a strong theoretical motivation for grand unified theories like @xmath10 from gauge coupling unification at the gut scale @xmath19 gev .
hence , the first golden rule for superstring phenomenology suggests to construct string models in such a way that at least some generations of quarks and leptons reside at a location in compact space , where they are subject to a larger gauge group , like @xmath10 .
hence , these generations come as complete representations of that larger group , e.g. as @xmath18 of @xmath10 . the heterotic string offers this possibility through the natural presence of the exceptional lie group @xmath20 , which includes an @xmath10 subgroup and its spinor representation . furthermore , using orbifold compactification the four - dimensional standard model gauge group can be enhanced to a local gut , i.e to a gut group like @xmath10 which is realized locally at an orbifold singularity in extra dimensions .
in addition , there are matter fields ( originating from the so - called twisted sectors of the orbifold ) localised at these special points in extra dimensions and hence they appear as complete multiplets of the local gut group , for example as @xmath18-plets of @xmath10 .
on the other hand , the spinor of @xmath10 is absent in ( perturbative ) type ii string theories , which can be seen as a drawback of these theories .
often this drawback manifests itself in an unwanted suppression of the top quark yukawa coupling . on the other hand , f - theory ( and m - theory )
can cure this through the non - perturbative construction of exceptional lie groups like e.g. @xmath21 .
when two seven - branes with @xmath10 gauge group intersect in the extra dimensions , a local gut can appear at the intersection .
there , the gauge group can be enhanced to a local @xmath21 and a spinor of @xmath10 can appear as matter representation . beside complete spinor representations of @xmath10 for quarks and leptons , the ( supersymmetric extension of the ) standard model needs split , i.e. incomplete @xmath10 multiplets for the gauge bosons and the higgs(pair ) .
their unwanted components inside a full gut multiplet would induce rapid proton decay and hence need to be ultra - heavy . in the case of the higgs doublet , this problem
is called the doublet
triplet splitting problem , because for the smallest gut @xmath9 a higgs field would reside in a five - dimensional representation of @xmath9 , which includes beside the higgs doublet an unwanted higgs triplet of @xmath22 .
this problem might determine the localisations of the higgs pair and of the gauge bosons in the compactification space : they need to reside at a place in extra dimensions where they feel the breaking of the higher - dimensional gut to the 4d sm gauge group .
hence , incomplete gut multiplets , e.g. for the higgs , can appear .
this is the content of the second golden rule . in this way local guts exhibit grand unified gauge symmetries only at some special `` local '' surroundings in extra dimensions , while in 4d the gut group seems to be broken down to the standard model gauge group .
this allows us to profit from some of the nice properties of guts ( like complete representations for matter as described in the first golden rule ) , while avoiding the problematic properties ( like doublet triplet splitting ) . in the case of the heterotic string on orbifolds the so - called untwisted sector ( i.e. the 10d bulk )
can naturally provide such split @xmath10 multiplets for the gauge bosons and the higgs .
in particular , when the orbifold twist acts as a @xmath23 in one of the three complex extra dimensions , one can obtain an untwisted higgs pair that is vector - like with respect to the full ( i.e. observable and hidden ) gauge group . combined with an ( approximate ) @xmath16-symmetry
this can yield a solution to the @xmath24-problem of the mssm .
furthermore , as all charged bulk fields originate from the 10d @xmath11 vector multiplet this scenario naturally yields gauge higgs unification . finally , an untwisted higgs pair in the framework of heterotic orbifolds can relate the top quark yukawa coupling to the gauge coupling and hence give a nice explanation for the large difference between the masses of the third generation compared to the first and second one . in order to achieve this , the top quark needs to originate either from the bulk ( as it is often the case in the minilandscape @xcite of @xmath12-ii orbifolds ) or from an appropriate fixed torus , i.e. a complex codimension one singularity in the extra dimensions .
the triple repetition of quarks and leptons as three generations with the same gauge interactions but different masses is a curiosity within the standard model and asks for a deeper understanding .
one approach from a bottom - up perspective is to engineer a so - called flavour symmetry : one introduces a ( non - abelian ) symmetry group , discrete or gauge , and unifies the three generations of quarks and leptons in , for example , a single three - dimensional representation of that flavour group .
however , as the yukawa interactions violate the flavour symmetry , it must be broken spontaneously by the vacuum expectation value of some standard model singlet , the so - called flavon .
this might explain the mass ratios and mixing patterns of quarks and leptons .
the third golden rule for superstring phenomenology asks for the origin of such a flavour symmetry .
the rule suggests to choose the compactification space such that some of its geometrical properties lead to a repetition of families and hence yields a discrete flavour symmetry . in this case
, the repetition of the family structure comes from topological properties of the compact manifold . within the framework of type ii string theories
, the number of families can be related to intersection numbers of d - branes in extra dimensions , while for the heterotic string it can be due to a degeneracy between orbifold singularities . in the latter case , one can easily obtain non - abelian flavour groups which originate from the discrete symmetry transformations that interchange the degenerate orbifold singularities , combined with a stringy selection rule that is related to the orbifold space group @xcite . in any case the number of families will be given by geometrical and topological properties of the compact six - dimensional manifold .
superstring theories are naturally equipped with @xmath26 or 2 supersymmetry in 10d . however ,
generically all supersymmetries are broken by the compactification to 4d .
the fourth golden rule suggests to choose a `` non - generic '' compactification space such that @xmath26 survives in 4d .
examples for such special spaces are calabi
yaus , orbifolds and orientifolds .
motivation for this is a solution of the so - called `` hierarchy problem '' between the weak scale ( a tev ) and the string ( planck ) scale .
supersymmetry can stabilize this large hierarchy .
since such a supersymmetry appears naturally in string theory , we assume that @xmath26 supersymmetry will survive down to the tev - scale . apart from the gauge symmetries of string theory , we need more symmetries to describe particle physics phenomena of the supersymmetric standard model
. these could provide the desired textures of yukawa couplings , explain the absence of flavour changing neutral currents , help to avoid too fast proton decay , provide a stable particle for cold dark matter and solve the so - called @xmath24-problem . we know
that ( continuous ) global symmetries might not be compatible with gravitational interactions . hence , local discrete symmetries might play this role in string theory .
one of these symmetries is the well - known matter parity of the minimal supersymmetric extension of the standard model ( mssm ) : it forbids proton decay via dim .
4 operators and leads to a stable neutral wimp candidate .
other discrete gauge symmetries are required to explain the flavour structure of quark / lepton masses and mixings .
as we have seen in our review in section [ sec : fivegoldenrules ] , the five golden rules @xcite naturally ask for exceptional lie groups . @xmath10 , although it is not an exceptional group , fits very well in the chain of exceptional groups @xmath28 .
therefore , the @xmath11 heterotic string is the prime candidate and we choose it as our starting point .
alternatives to obtain @xmath20 in string theory are m- and f - theory , where such gauge groups can appear in non - perturbative constructions .
the implementation of the rules in string theory started with the consideration of orbifold compactifications of the @xmath11 heterotic string .
this lead to the so - called `` heterotic brane world '' @xcite where toy examples have been constructed in the framework of the @xmath29 orbifold . there ,
the explicit `` geographical '' properties of fields in extra dimensions have been presented and the local guts at the orbifold fixed points were analysed , see e.g. fig .
[ fig : gaugegrouptopo ] . a first systematic attempt at realistic model constructions @xcite was based on the @xmath12-ii orbifold @xcite of the @xmath11 heterotic string .
this orbifold considers a six - torus defined by the six - dimensional lattice of @xmath30 modded out by two twists , each acting in four of the six extra dimensions : @xmath31 of order 2 ( @xmath32 ) and @xmath33 of order 3 ( @xmath34 ) , see fig .
[ fig : thetaomegasector ] . in ref .
@xcite the embedding of the twists into the @xmath11 gauge group was chosen in such a way that at an intermediate step there are local @xmath10 guts with localised @xmath18-plets for quarks and leptons .
this choice can be motivated by rule i , as discussed in the previous section .
further breakdown of the gauge group to @xmath7 is induced by two orbifold wilson lines @xcite . in this
set - up , a scan for realistic models was performed using the following strategy : * choose appropriate wilson lines ( and identify inequivalent models ) * sm gauge group @xmath35 times a hidden sector * hypercharge @xmath36 is non - anomalous and in @xmath9 gut normalisation * ( net ) number of three generations of quarks and leptons * at least one higgs pair * exotics are vector - like w.r.t .
the sm gauge group and can be decoupled using the above criteria , the computer assisted search led to a total of some 200 and 300 mssm - like models in refs .
@xcite and @xcite , respectively .
the models typically have additional vector - like exotics as well as unbroken @xmath37 gauge symmetries , one of which is anomalous .
this anomaly induces an fayet iliopoulos term ( fi - term ) , hence a breakdown of the additional @xmath37 s and thus allows for a decoupling of the vector - like exotics .
explicit examples are given in ref .
@xcite as benchmark models .
all fields of the models can be attributed to certain sectors with specific geometrical properties . in the present case
there is an untwisted sector with fields in 10d ( bulk ) , as well as twisted sectors where fields are localised at certain points ( or two - tori ) in the six - dimensional compactified space . the @xmath38 twisted sector ( fig .
[ fig : thetaomegasector ] ) has fixed points and thus yields fields localised at these points in extra dimensions that can only propagate in our four - dimensional space time .
the @xmath31 and @xmath33 twisted sectors , in contrast , have fixed two - tori in extra dimensions .
fields in these sectors are confined to six space time dimensions .
many properties of the models depend on these `` geographic '' properties of the fields in extra dimensions .
for example , yukawa couplings between matter and higgs fields and in particular their coupling strengths are determined by the `` overlap '' of the fields in extra dimensions .
given this large sample of realistic models , we can now analyse their properties and look for similarities and regularities . which geometrical and geographical properties in extra dimensions are important for realistic models ? by construction
, all the models have observable sector gauge group @xmath7 and possibly some hidden sector gauge group relevant for supersymmetry breakdown .
there is a net number of three generations of quarks and leptons and at least one pair of higgs doublets @xmath39 and @xmath40 .
the higgs triplets are removed and the doublet
triplet splitting problem is solved .
a first question concerns a possible `` @xmath24-term '' : @xmath41 and we shall start our analysis with the higgs
system , following the discussion of ref .
@xcite .
the higgs
system is vector - like and a @xmath24-term @xmath41 is potentially allowed . as this is a term in the superpotential we would like to understand why @xmath24 is small compared to the gut - scale : this is the so - called @xmath24-problem . to avoid this problem
one could invoke a symmetry that forbids the term .
however , we know that @xmath24 has to be non - zero . hence , the symmetry has to be broken and this might reintroduce the @xmath24-problem again . in string theory the problem is often amplified since typically we find several ( say @xmath42 ) higgs doublet pairs . in the procedure to remove the vector - like exotics ( as described above ) we have to make @xmath43 pairs heavy while keeping one light .
in fact , in many cases the small @xmath24-parameter is the result of a specific fine - tuning in such a way to remove all doublet pairs except for one .
we do not consider this as a satisfactory solution .
fortunately , the models of the minilandscape are generically not of this kind .
many minilandscape models provide one higgs pair that resists all attempts to remove it . this is related to a discrete r - symmetry @xcite that can protect the @xmath24-parameter in the following way : in some cases the discrete r - symmetry is enlarged to an approximate @xmath44 @xcite .
symmetries can explain vanishing vacuum energy in susy vacua .
] therefore , a @xmath24-parameter is generated at a higher order @xmath45 in the superpotential @xmath46 , where the approximate @xmath44 is broken to its exact discrete subgroup .
this yields a suppression @xmath47 , where @xmath48 is set by the fi parameter .
the crucial observation for this mechanism to work is the localisation of the higgs pair @xmath39 and @xmath40 in agreement with our second golden rule : both reside in the 10d bulk originating from gauge fields in extra dimensions .
furthermore , the higgs pair is vector - like with respect to all symmetries , gauge and discrete .
this is related to the @xmath23 orbifold action in one of the two - tori .
hence , each term in the superpotential @xmath49 also couples to the higgs pair , i.e. @xmath50 .
as susy breakdown requires a non - vanishing vev of the superpotential the @xmath24-term is related to the gravitino mass , i.e. @xmath51 .
this is a reminiscent of a field theoretical mechanism first discussed in ref .
@xcite . among all quarks and leptons
the top - quark is very special : its large mass requires a large top - quark yukawa coupling .
many minilandscape models address this naturally via the localisation of the top - quark in extra dimensions : both @xmath52 and @xmath53 reside in the 10d bulk , along with the higgs pair .
hence , we have gauge - yukawa unification and the trilinear yukawa coupling of the top is given by the gauge coupling .
typically the top - quark is the only matter field with trilinear yukawa coupling .
the location of the other fields of the third family is strongly model - dependent , but in general they are distributed over various sectors : the third family could be called a `` patchwork family '' .
the first two families are found to be located at fixed points in extra dimensions ( fig .
[ fig : thetaomegasector ] ) . as such
they live at points of enhanced symmetries , both gauge and discrete .
the discrete symmetry is the reason for their suppressed yukawa couplings . in the @xmath12-ii example shown in the figure
two families live at adjacent fixed points in the third extra - dimensional torus : one family is located at @xmath54 , the other at @xmath55 and @xmath56 ( see fig .
[ fig : thetaomegasector ] ) . technically , this is a consequence of a vanishing wilson line in the @xmath57 direction .
this leads to a @xmath58 flavour symmetry @xcite .
the two localised families form a doublet , while the third family transforms in a one - dimensional representation of @xmath58 .
this set - up forbids sizeable flavour changing neutral currents and thus relieves the so - called `` flavour problem '' .
furthermore , the geometric reason for small yukawa couplings of the first and second family is their minimal overlap with the bulk higgs fields .
this leads to yukawa couplings of higher order and a hierarchical generation of masses based on the froggatt
nielsen mechanism @xcite , where the fi - term provides a small parameter @xmath59 that controls the pattern of masses .
in addition , the first two families live at points of enhanced gauge symmetries and therefore build complete representations of the local grand unified gauge group , e.g. as @xmath18-plets of @xmath10 .
hence , they enjoy the successful properties of `` local grand unification '' outlined in the first golden rule . the question of supersymmetry breakdown is a complicated process and we shall try to extract some general lessons that are rather model - independent . specifically we would consider gaugino condensation in the hidden sector @xcite realized explicitly in the minilandscape @xcite , see also section [ sec : orbifoldlandscaperule4 ] . a reasonable value for the gravitino mass can be obtained if the dilaton is fixed at a realistic value @xmath60 .
thus , the discussion needs the study of moduli stabilization , which , fortunately , we do not have to analyse here .
in fact we can rely on some specific pattern of supersymmetry breaking which seems to be common in various string theories , first observed in the framework of type iib theory @xcite and later confirmed in the heterotic case @xcite : so - called `` mirage mediation '' .
its source is a suppression of the tree level contribution in modulus mediation ( in particular for gaugino masses and a - parameters ) .
the suppression factor is given by the logarithm of the `` hierarchy '' @xmath61 , which numerically is of the order @xmath62 .
non - leading terms suppressed by loop factors can now compete with the tree - level contribution . in its simplest form
the loop corrections are given by the corresponding @xmath63-functions , leading to `` anomaly mediation '' if the tree level contribution is absent . without going into detail ,
let us just summarise the main properties of mirage mediation : * gaugino masses and a - parameters are suppressed compared to the gravitino mass by the factor @xmath61 * we obtain a compressed pattern of gaugino masses ( as the @xmath64 @xmath63-function is negative while those of @xmath65 and @xmath66 are positive ) * soft scalar masses @xmath67 are more model - dependent . in general we would expect them to be as large as @xmath68 @xcite
. the models of the minilandscape inherit this generic picture .
but they also teach us something new on the soft scalar masses , which results in lesson 4 .
the scalars reside in various localisations in the extra dimensions that feel susy in different ways : first , the untwisted sector is obtained from simple torus compactification of the 10d theory leading to extended @xmath69 supersymmetry in @xmath2 .
hence , soft terms of bulk fields are protected ( at least at tree level ) and broken by loop corrections when they communicate to sectors with less susy .
next , scalars localised on fixed tori feel a remnant @xmath70 susy and might be protected as well .
finally , fields localised at fixed points feel only @xmath26 susy and are not further protected @xcite .
therefore , we expect soft terms @xmath71 for the localised first two families , while other ( bulk ) scalar fields , in particular the higgs bosons and the stop , feel a protection from extended susy .
consequently , their soft masses are suppressed compared to @xmath68 ( by a loop factor of order @xmath72 ) .
this constitutes lesson 4 of the minilandscape .
the 10d heterotic string compactified on a six - dimensional toroidal orbifolds provides an easy and calculable framework for string phenomenology @xcite .
a toroidal orbifold is constructed by a six - dimensional torus divided out by some of its discrete isometries , the so - called point group . for simplicity
we assume this discrete symmetry to be abelian .
combined with the condition on @xmath26 supersymmetry in 4d one is left with certain @xmath13 and @xmath14 groups , in total 17 different choices . for each choice , there are in general several inequivalent possibilities , e.g. related to the underlying six - torus .
recently , these possibilities have been classified using methods from crystallography , resulting in 138 inequivalent orbifold geometries with abelian point group @xcite .
the orbifolder @xcite is a powerful computer program to analyse these abelian orbifold compactifications of the heterotic string .
the program includes a routine to automatically generate a huge set of consistent ( i.e. modular invariant and hence anomaly - free ) orbifold models and to identify those that are phenomenologically interesting , e.g. that are mssm - like .
a crucial step in this routine is the identification of inequivalent orbifold models in order to avoid an overcounting : even though the string theory input parameters of two models ( i.e. so - called shifts and orbifold wilson lines ) might look different , the models can be equivalent and share , for example , the same massless spectrum and couplings .
the current version ( 1.2 ) of the orbifolder uses simply the massless spectrum in terms of the representations under the full non - abelian gauge group in order to identify inequivalent models .
however , there are typically five to ten 1 factors and the corresponding charges are neglected for this comparison of spectra , because they are highly dependent on the choice of 1 basis . as pointed out by groot nibbelink and loukas @xcite one can easily improve this by using in addition to the non - abelian representations also the @xmath36 hypercharge as it is uniquely defined for a given mssm model .
we included this criterion into the orbifolder .
however , it turns out that using this refined comparison method the number of inequivalent mssm - like orbifold models increases only by 3% . using the improved version of the orbifolder we performed a scan in the landscape of all @xmath13 and certain @xmath14 heterotic orbifold geometries for mssm - like models ,
where our basic requirements for a model to be mssm - like are : * sm gauge group @xmath73 times a hidden sector * hypercharge @xmath36 is non - anomalous and in @xmath9 gut normalisation * ( net ) number of three generations of quarks and leptons * at least one higgs pair * all exotics must be vector - like with respect to the sm gauge group we identified approximately 12000 mssm - like orbifold models that suit the above criteria .
given the large number of promising models we call them the `` orbifoldlandscape '' .
a summary of the results can be found in the appendix in tabs . [
tab : mssmsummary ] and [ tab : mssmsummary2 ] .
furthermore , the orbifolder input files needed to load these models into the program can be found at @xcite .
the scan did not reveal any mssm - like models from orbifold geometries with @xmath74 , @xmath75 and @xmath76-ii point group .
this is most likely related to the condition of @xmath9 gut normalisation for hypercharge .
note that this search for mssm - like orbifold models is by far not complete .
for example , we only used the standard @xmath14 orbifold geometries ( i.e. those with label ( 1 - 1 ) following the nomenclature of ref .
in addition , our search was performed in a huge , but still finite parameter set of shifts and wilson lines .
finally , the routine to identify inequivalent orbifold models can surely be improved further .
hence , presumably only a small fraction of the full heterotic orbifold landscape has been analysed here .
let us compare our findings to the literature .
the @xmath12-ii ( 1 - 1 ) orbifold has been studied intensively in the past , see e.g. @xcite .
also the minilandscape @xcite was performed using this orbifold geometry , see section [ sec : reviewminilandscape ] . in the first paper @xcite local @xmath10 and @xmath21 guts were used as a search strategy and thus one was restricted to four out of 61 possible shifts , resulting in 223 mssm - like models . in the second paper @xcite this restriction was lifted , resulting in almost 300 mssm - like models .
they are all included in our set of 348 mssm - like models from @xmath12-ii ( 1 - 1 ) , see tab .
[ tab : mssmsummary ] in the appendix .
similar to @xmath12-ii , the @xmath77 orbifold geometry has been conjectured to be very promising for mssm model building @xcite . here , we can confirm this conjecture : we found 3632 mssm - like models from @xmath77 ( 1 - 1 ) the largest set of models in our scan . also from a geometrical point of view
, the @xmath77 orbifold is very rich : there are in total 41 different orbifold geometries with @xmath77 point group , i.e. based on different six - tori and roto
translations @xcite .
we considered only the standard choice here , labelled ( 1 - 1 ) .
hence , one can expect a huge landscape of mssm - like models to be discovered from @xmath77 .
recently , groot nibbelink and loukas performed a model scan in all @xmath78-i and @xmath78-ii geometries @xcite .
they also used a local gut search strategy ( based on @xmath9 and @xmath10 local guts ) and hence started with 120 and 108 inequivalent shifts for @xmath78-i and @xmath78-ii , respectively .
their scan resulted in 753 mssm - like models . without imposing the local gut strategy our search revealed in total 1713 mssm - like models from @xmath78 ,
see tab .
[ tab : mssmsummary ] .
further orbifold mssm - like models have been constructed using the @xmath79-i orbifold geometry @xcite .
this orbifold seems also to be very promising as we identified 750 mssm - like models in this case , see tab .
[ tab : mssmsummary ] .
finally , we confirm the analysis of ref .
@xcite for the orbifold geometries @xmath12-i and @xmath13 with @xmath80 and standard lattice ( 1 - 1 ) . in the next section
we will apply the strategies described by the `` five golden rules of superstring phenomenology '' to our orbifoldlandscape and search for common properties of our 12000 mssm - like orbifold models .
thereby , we will see how many mssm - like models would have been found following the `` five golden rules '' strictly and how many would have been lost .
hence , we will estimate the prosperity of the `` five golden rules '' .
in the following we focus on the golden rules i - iv . a detailed analysis of rule
v is very model - dependent and will thus not be discussed here . as discussed in sec .
[ sec : firstgoldenrules ] at least some generations of quarks and leptons might originate from spinors of @xmath10 sitting at points in extra dimensions with local @xmath10 gut .
hence , we perform a statistic on the number of such localisations in our 12000 mssm - like orbifold models .
the results are summarised in tab .
[ tab : mssmsummary2 ] and displayed in fig .
[ fig : so10spinors ] .
it turns out that 25% of our models have at least one local @xmath10 gut .
furthermore , we find that some orbifolds seem to forbid local @xmath10 guts with @xmath18-plets ( for example @xmath12-i @xcite ) . on the other hand , the mssm - like models from @xmath12-ii and @xmath78-i ( 1 - 1 ) and ( 2 - 1 ) prefer zero or two localised @xmath18-plets of @xmath10 .
three local @xmath18-plets are very uncommon , they mostly appear in @xmath77 . note that the number of local guts can be greater than three even though the model has a ( net ) number of three generations of quarks and leptons .
obviously , an anti - generation of quarks and leptons is needed in such a case .
the maximal number we found in our scan is four local @xmath10 guts with @xmath18-plets for matter in the cases of @xmath29 and @xmath77 orbifold geometries .
in addition , we analyse our 12000 models for local @xmath9 guts with local matter in @xmath81-plets .
the results are summarised in table [ tab : mssmsummary2 ] and displayed in fig .
[ fig : localsu5guts ] .
we find this case to be very common : almost 40% of our mssm - like models have at least one local @xmath81-plet of a local @xmath9 gut .
next , we also look for local @xmath21 guts with @xmath82-plets . we find only a few cases , most of them appear in @xmath14 orbifold geometries , see table [ tab : mssmsummary2 ] . finally , we scan our models for localised sm generations ( i.e. localised left handed quark doublets ) transforming in a complete multiplet of any local gut group that unifies the sm gauge group . again
, our results are listed in table [ tab : mssmsummary2 ] and visualised in fig .
[ fig : localanyguts ] .
we find most of our models , i.e. 70% , have at least one local gut with a localised sm generation . in summary ,
the first golden rule , which demands for local guts in extra dimensions in order to obtain complete gut multiplets for matter , is very successful : most of our 12000 mssm - like models share this property automatically , it was not imposed by hand in our search .
since the higgs doublets reside in incomplete gut multiplets , they might be localised at some region of the orbifold where the higher - dimensional gut is broken to the 4d standard model gauge group .
this scenario yields a natural solution to the doublet
triplet splitting problem .
the untwisted sector ( i.e. bulk ) would be a prime candidate for such a localisation , but there can be further possibilities .
the numbers of such gut breaking localisations are summarised in table [ tab : mssmsummary ] and displayed in fig .
[ fig : higgs ] .
we see that gut breaking localisations are very common among our mssm - like models .
only a very few models do not contain any gut breaking localisations that yield incomplete gut multiplets for at least one higgs . on the other hand ,
there are 4223 cases with one gut breaking localisation in most cases ( 4097 out of 4223 ) this is the bulk .
in addition , there are many models that have more than one possibility for naturally split higgs multiplets , but in almost all cases the bulk is among them .
note that most of our mssm - like models have additional exotic higgs - like pairs , mostly two to six additional ones .
in contrast to the mssm higgs pair they often originate from complete multiplets of some local gut . on the other hand , we identified 1011 mssm - like models with exactly one higgs pair .
cases with exactly one higgs pair , originating from the bulk might be especially interesting . in summary ,
the second golden rule , which explains incomplete gut multiplets for the higgs using gut breaking localisations in extra dimensions , is very successful as in the case of the first golden rule , most of our 12000 mssm - like models follow this rule automatically .
the standard model contains three generations of quarks and leptons with a peculiar pattern of masses and mixings .
this might be related to a ( discrete ) flavour symmetry . or @xmath22 is also possible .
some of the models in our orbifoldlandscape realise this possibility , but we do not analyse these cases in detail here . ] from the orbifold perspective discrete flavour symmetries naturally arise from the symmetries of the orbifold geometry @xcite . however , certain background fields ( i.e. orbifold wilson lines @xcite ) can break these symmetries .
the maximal number of orbifold wilson lines is six corresponding to the six directions of the compactified space .
the orbifold rotation , however , in general identifies some of those directions .
hence , the corresponding wilson lines have to be equal . for example
, the @xmath74 orbifold allows for maximally three independent wilson lines . in general , one can say that the more wilson lines vanish the larger is the discrete flavour symmetry . on the other hand ,
non - vanishing wilson lines are generically needed in order to obtain the standard model gauge group and to reduce the number of generations to three .
hence , it is interesting to perform a statistic on the number of vanishing wilson lines for our 12000 mssm - like orbifold models , see tab . [
tab : mssmsummary ] in the appendix and fig .
[ fig : generations ] .
there are orbifold geometries , like @xmath83 , @xmath12-i and @xmath79-i , apparently demanding for all possible orbifold wilson lines to be non - trivial in order to yield the mssm , see tab . [
tab : mssmsummary ] .
these mssm - like models are expected to have no discrete , non - abelian flavour symmetries .
on the other hand , there are several orbifold geometries that seem to require at least one vanishing wilson line in order to reproduce the mssm with its three generations , for example @xmath12-ii , @xmath29 , @xmath77 , @xmath84 and @xmath85 .
in general , the case of vanishing wilson lines is very common : we see that in 75% of our mssm - like orbifold models at least one allowed orbifold wilson line is zero . in these cases
non - abelian flavour symmetries are expected .
for example , most of the mssm - like models from @xmath12-ii ( 1 - 1 ) have a @xmath86 flavour symmetry with the first two generations transforming as a doublet and the third one as a singlet @xcite . in summary , the third golden rule , which explains the origin of three generations of quarks and leptons by geometrical properties of the compactification space , is generically satisfied for our 12000 mssm - like orbifold models .
by construction , i.e. by choosing the appropriate orbifold geometries , our 12000 mssm - like orbifold models preserve @xmath26 supersymmetry in four dimensions .
this is expected to be broken by non - perturbative effects , i.e. by hidden sector gaugino condensation @xcite . here , we follow the discussion of @xcite where low energy supersymmetry breaking in the minilandscape of @xmath12-ii orbifolds was analysed . see also @xcite for a related discussion . in detail , our mssm - like models typically possess a non - abelian hidden sector gauge group with little or no charged matter representations .
the corresponding gauge coupling depends via the one - loop @xmath63-functions on the energy scale .
if the coupling becomes strong at some ( intermediate ) energy scale @xmath87 the respective gauginos condensate and supersymmetry is broken spontaneously by a non - vanishing dilaton @xmath88-term . assuming that susy breaking is communicated to the observable sector via gravity the scale of soft susy breaking is given by the gravitino mass , i.e.@xmath89 where @xmath90 denotes the planck mass and the scale of gaugino condensation @xmath87 is given by @xmath91 for every mssm - like orbifold model we compute the @xmath63-function of the largest hidden sector gauge group under the assumption that any non - trivial hidden matter representation of this gauge group can be decoupled in a supersymmetric way .
furthermore , we assume dilaton stabilization at a realistic value @xmath60 .
hence , we obtain the scale @xmath87 of gaugino condensation .
our results are displayed in fig .
[ fig : gauginocondenstaion ] . for an intermediate scale @xmath92
one obtains a gravitino mass in the @xmath93 range , which is of phenomenological interest .
the models in the orbifoldlandscape seem to prefer low energy susy breaking . this result
is strongly related to the heterotic orbifold construction : the @xmath11 gauge group in 10d is broken by orbifold shift and wilson lines , which are highly constrained by string theory ( i.e. modular invariance ) . therefore , both @xmath20 factors get broken and not only the observable one .
it turns out that the unbroken gauge group from the hidden @xmath20 has roughly the correct size to yield gaugino condensation at an intermediate scale and hence low energy susy breaking .
note that our analysis is just a rough estimate as various effects have been neglected , for example the decoupling of hidden matter , the identification of the gaugino condensation and ( string ) threshold corrections .
these effects can in principle affect the scale of susy breaking even by 2 - 3 orders of magnitude .
with these considerations we have only scratched the surface of the parameter space of potentially realistic models .
in addition , we have used `` five golden rules '' as a prejudice for model selection and it has to be seen whether this is really justified . for general model building in the framework of ( perturbative )
string theory we have the following theories at our disposal : * type i string with gauge group @xmath94 * heterotic @xmath94 * heterotic @xmath11 * type iia and iib orientifolds * intersecting branes with gauge group @xmath95 as we explained in detail , our rule i points towards exceptional groups and hence towards the @xmath11 heterotic string . on the other hand ,
type ii orientifolds typically provide gauge groups of type @xmath96 or @xmath97 and products thereof .
although we have @xmath98 gauge groups in these schemes , matter fields do not come as spinors of @xmath98 , but originate from adjoint representations . in the intersecting brane models based on @xmath95 gauge groups matter
transforms in bifundamental representations of @xmath99 ( originating from the adjoint of @xmath100 ) .
while this works nicely for the standard model representations , it appears to be difficult to describe a grand unified picture with e.g. gauge group @xmath9 . trying to obtain a gut yields a gauge group at least as large as @xmath101 and one has problems with a perturbative top - quark yukawa coupling .
one possible way out is the construction of string models without the prejudice for guts , see e.g. @xcite .
a comprehensive review on these intersecting brane model constructions can be found in the book of ibez and uranga @xcite or other reviews @xcite .
these models have a very appealing geometric interpretation , see e.g.@xcite : fields are located on branes of various dimensions .
thus , physical properties of the models can be inferred from the localisation of the brane fields in the extra dimensions and by the overlap of their wave functions , similar to the heterotic minilandscape .
this nice geometrical set - up leads to attempts to construct so - called `` local models '' . here
, one assumes that all particle physics properties of the model are specified by some local properties at some specific point or sub - space of the compactified dimensions and that the `` bulk '' properties can be decoupled .
however , the embedding of the local model into an ultraviolet complete and consistent string model is an assumption and its validity remains an open question .
further schemes include `` non - perturbative '' string constructions : * m - theory in @xmath8 * heterotic m - theory @xmath11 * f - theory these non - perturbative constructions are conjectured theories that generalize string theories or known supergravity field theories in higher dimensions .
the low energy limit of m - theory is 11-dimensional supergravity .
heterotic m - theory is based on a @xmath8 theory bounded by two @xmath0 branes with gauge group @xmath20 on each boundary and f - theory is a generalization of type iib theory , where certain symmetries can be understood geometrically .
this non - perturbative construction allows for singularities in extra dimensions that lead to non - trivial gauge groups according to the so - called a - d - e classification .
groups of the a - type ( @xmath102 ) and d - type ( @xmath98 ) can also be obtained in the perturbative constructions with d - branes and orientifold branes , while exceptional gauge groups can only appear through the presence of e - type singularities .
this allows for spinors of @xmath10 and can produce a non - trivial top - quark yukawa coupling within an @xmath9 grand unified theory . in that sense ,
f - theory can be understood as an attempt to incorporate rule i within type iib theory .
unfortunately , it is difficult to control the full non - perturbative theory and the search for realistic models is often based on local model building .
many questions are still open but there is enough room for optimism that promising models can be embedded in a consistent ultraviolet completion .
a general problem of string phenomenology is the difficulty to perform the explicit calculations needed to check the validity of the model .
this is certainly true for the non - perturbative models , where we have ( at best ) some effective supergravity description . but also in the perturbative constructions we have to face this problem .
we have to use simplified compactification schemes to be able to do the necessary calculations we need a certain level of `` berechenbarkeit '' . in our discussion
we used the flat orbifold compactification that allows the use of conformal field theory methods . in principle , this enables us to do all the necessary calculations to check the models in detail . in the @xmath12-ii minilandscape this has been elaborated to a large extend . for the more general orbifold landscape , this still has to be done .
other constructions with full conformal field theory control are the free fermionic constructions @xcite and the `` tensoring '' of conformal field theory building blocks : so - called gepner models @xcite .
they share `` berechenbarkeit '' with the flat orbifold models , but the geometric structure of compactified space is less transparent .
we have to hope that these simplified compactifications ( or approximations ) lead us to realistic models . in the generic situation
one needs smooth manifolds , e.g. calabi - yau spaces , and some specific models have been constructed @xcite . however , these more generic compactifications require more sophisticated methods for computations that are only partially available , for example in order to determine yukawa couplings . more recently
a simplification based on the embedding of line bundles has allowed the constructions of many models @xcite . still the calculational options are limited .
it would be interesting to get a better geometric understanding of the compact manifold . at the moment
the `` determination '' of couplings is based on a supergravity approximation using @xmath37 symmetries .
these symmetries are exact in this approximation at the `` stability wall '' but are expected to be broken to discrete symmetries in the full theory .
this is in concord with rule v asking for the origin of discrete symmetries .
furthermore , this question has recently been analysed intensively within the various string constructions @xcite .
we have seen that there is still a long way to go in the search for realistic particle physics models from string theory .
there are many possible roads but we are limited by our calculational techniques .
thus , in the near future we are still forced to make choices . here
, we have chosen to follow `` five golden rules '' outlined in section [ sec : fivegoldenrules ] , which are mainly motivated by the quest for a unified picture of particle physics interactions .
this strategy seems to require an underlying structure provided by exceptional groups pointing towards the @xmath11 heterotic string and f - theory .
even given these rules , there are stumbling blocks because of the complexity of the compact manifolds .
we can not resolve these problems in full generality : we have to use simplified compactification schemes or approximations .
we have to hope that nature has chosen a theory that is somewhat close to these simplified schemes .
of course , any method to go beyond this simplified assumptions should be seriously considered . however , there is some hope that this assumption might be justified : the orbifold models studied in this work have enhanced ( discrete ) symmetries that could be the origin of symmetries of the standard model , especially with respect to the flavour structure and symmetries relevant for proton stability as well as the absence of other rare processes .
generically , these symmetries are slightly broken as we go away from the orbifold point .
this gives rise to some hierarchical structures , for example for the ratio of quark masses in the spirit of froggatt and nielsen @xcite .
the analysis of the minilandscape can be seen as an attempt to study these questions in detail .
based on the availability of conformal field theory techniques we can go pretty far in the analysis of explicit models .
a detailed analysis of the `` orbifoldlandscape '' has not been performed yet , but should be possible along the same lines .
in section [ sec : orbifoldlandscape ] we started this enterprise of model building by constructing 12000 mssm - like models . in a next step ,
the detailed properties of promising models have to be worked out .
especially the framework of the @xmath77 @xcite should provide new insight into the properties of realistic models and might teach us further key properties shared by successful models . one key property that we have learned is the geography of fields in the extra dimensions .
the localisation of matter fields and the gauge group profiles in extra dimensions are essential for the properties of the low energy model .
this is the first message of the heterotic orbifold construction and shared by the `` braneworld '' constructions in type ii string theory and f - theory .
further lessons are : * the higgs pair is a bulk field .
this allows for a convincing solution of the @xmath24-problem using a ( discrete ) r - symmetry and yields doublet
triplet splitting .
* a sizeable value of the top - quark yukawa coupling requires a sufficient overlap with the higgs fields in extra dimensions .
thus , the top - quark should extend to the bulk as well . * the matter fields of the first and second generation should be localised in a region of the extra - dimensional space where they are subject to an enhanced gauge symmetry , like @xmath10 .
this local gut forces them to appear as complete representations , e.g. as spinors of @xmath10 .
furthermore , the geometrical structure can manifest itself in a discrete flavour symmetry . *
the quest for low energy supersymmetry is the guiding principle in string model building .
still , it has to be seen whether this is realised in nature . at the moment
no sign of supersymmetry has been found at the lhc , although the value of the higgs mass is consistent with susy .
the analysis of the models of the minilandscape and the location of the fields suggests a certain structure where even some remnants of extended supersymmetry ( for fields in the bulk ) seem to be at work .
this picture of `` heterotic supersymmetry '' @xcite can hopefully be tested experimentally in the not too far future .
this work was partially supported by the sfb
transregio tr33 `` the dark universe '' ( deutsche forschungsgemeinschaft ) and the dfg cluster of excellence `` origin and structure of the universe '' ( www.universe-cluster.de ) .
.statistics on mssm - like models ( using the search criteria listed in sec . [ sec : searchstrategy ] ) obtained from a random scan in all @xmath13 and certain @xmath14 heterotic orbifold geometries .
the first column labels the geometry following the nomenclature from @xcite .
the second column gives the number of inequivalent mssm - like models found in our scan .
next , we give the maximal number of independent wilson lines ( wls ) possible for the respective orbifold geometry and in the fourth column we count the number of mssm - like models with a certain number ( i.e. 0,1,2,3,4 ) of vanishing wilson lines , see sec . [
sec : orbifoldlandscaperule3 ] . in the fifth column
we count the number of locations with broken local gut such that higgs - doublets without triplets appear , see sec .
[ sec : orbifoldlandscaperule2 ] .
finally , in the last column we give the number of models without @xmath103 , i.e. without fi term .
[ cols="^,^ , > , > , > , > , > , > , > , > , > , > , > , > , > , > , > " , ] h. p. nilles , ` arxiv : hep - th/0410160 ` [ hep - th ] , hep - th/0410160 .
l. j. dixon , j. a. harvey , c. vafa , and e. witten , nucl.phys .
* b261 * ( 1985 ) , 678 . l. j. dixon , j. a. harvey , c. vafa , and e. witten , nucl.phys . *
b274 * ( 1986 ) , 285 .
t. kobayashi , s. raby , and r .- j .
zhang , nucl.phys .
* b704 * ( 2005 ) , 3 , ` arxiv : hep - ph/0409098 ` [ hep - ph ] .
w. buchmller , k. hamaguchi , o. lebedev , and m. ratz , nucl.phys . *
b712 * ( 2005 ) , 139 , ` arxiv : hep - ph/0412318 ` [ hep - ph ] .
w. buchmller , k. hamaguchi , o. lebedev , and m. ratz , phys.rev.lett .
* 96 * ( 2006 ) , 121602 , ` arxiv : hep - ph/0511035 ` [ hep - ph ] .
w. buchmller , k. hamaguchi , o. lebedev , and m. ratz , nucl.phys . *
b785 * ( 2007 ) , 149 , ` arxiv : hep - th/0606187 ` [ hep - th ] .
o. lebedev , h. p. nilles , s. raby , s. ramos - snchez , m. ratz , et al .
, phys.lett . *
b645 * ( 2007 ) , 88 , ` arxiv : hep - th/0611095 ` [ hep - th ] .
o. lebedev , h. p. nilles , s. raby , s. ramos - snchez , m. ratz , et al .
, phys.rev . *
d77 * ( 2008 ) , 046013 , ` arxiv:0708.2691 ` [ hep - th ] .
m. ratz , ` arxiv:0711.1582 ` [ hep - ph ] , arxiv:0711.1582 [ hep - ph ] . o. lebedev , h. p. nilles , s. ramos - snchez , m. ratz , and p. k. vaudrevange , phys.lett . *
b668 * ( 2008 ) , 331 , ` arxiv:0807.4384 ` [ hep - th ] . t. kobayashi , h. p. nilles , f. plger , s. raby , and m. ratz , nucl.phys . * b768 * ( 2007 ) , 135 , ` arxiv : hep - ph/0611020 ` [ hep - ph ] . h. p. nilles , s. ramos - snchez , m. ratz , and p. k. vaudrevange , eur.phys.j . * c59 * ( 2009 ) , 249 , ` arxiv:0806.3905 ` [ hep - th ] . s. frste , h. p. nilles , p. k. vaudrevange , and a. wingerter , phys.rev . *
d70 * ( 2004 ) , 106008 , ` arxiv : hep - th/0406208 ` [ hep - th ] .
l. e. ibez , h. p. nilles , and f. quevedo , phys.lett .
* b187 * ( 1987 ) , 25 . h. p. nilles , eur.phys.j .
* c74 * ( 2014 ) .
r. kappl , h. p. nilles , s. ramos - snchez , m. ratz , k. schmidt - hoberg , et al .
, phys.rev.lett .
* 102 * ( 2009 ) , 121602 , ` arxiv:0812.2120 ` [ hep - th ] .
f. brmmer , r. kappl , m. ratz , and k. schmidt - hoberg , jhep * 1004 * ( 2010 ) , 006 , ` arxiv:1003.0084 ` [ hep - th ] . j. casas and c. muoz , phys.lett . * b306 * ( 1993 ) , 288 , ` arxiv : hep - ph/9302227 ` [ hep - ph ] . h. p. nilles , m. ratz , and p. k. vaudrevange , fortsch.phys .
* 61 * ( 2013 ) , 493 , ` arxiv:1204.2206 ` [ hep - ph ] . c. froggatt and h. b. nielsen , nucl.phys .
* b147 * ( 1979 ) , 277 .
h. p. nilles , phys.lett . * b115 * ( 1982 ) , 193 . s. ferrara , l. girardello , and h. p. nilles , phys.lett . * b125 * ( 1983 ) , 457 .
j. derendinger , l. e. ibez , and h. p. nilles , phys.lett .
* b155 * ( 1985 ) , 65 .
m. dine , r. rohm , n. seiberg , and e. witten , phys.lett . * b156 * ( 1985 ) , 55 . o. lebedev , h .-
nilles , s. raby , s. ramos - snchez , m. ratz , et al . , phys.rev.lett . * 98 * ( 2007 ) , 181602 , ` arxiv : hep - th/0611203 ` [ hep - th ] .
k. choi , a. falkowski , h. p. nilles , m. olechowski , and s. pokorski , jhep * 0411 * ( 2004 ) , 076 , ` arxiv : hep - th/0411066 ` [ hep - th ] .
k. choi , a. falkowski , h. p. nilles , and m. olechowski , nucl.phys .
* b718 * ( 2005 ) , 113 , ` arxiv : hep - th/0503216 ` [ hep - th ] .
o. lebedev , h. p. nilles , and m. ratz , phys.lett . *
b636 * ( 2006 ) , 126 , ` arxiv : hep - th/0603047 ` [ hep - th ] .
v. nilles and j. plank , ccr * 42 * , 736 .
o. lebedev , h. p. nilles , and m. ratz , ` arxiv : hep - ph/0511320 ` [ hep - ph ] , 211 , hep - ph/0511320 .
o. loaiza - brito , j. martin , h. p. nilles , and m. ratz , aip conf.proc .
* 805 * ( 2006 ) , 198 , ` arxiv : hep - th/0509158 ` [ hep - th ] .
o. lebedev , v. lwen , y. mambrini , h. p. nilles , and m. ratz , jhep * 0702 * ( 2007 ) , 063 , ` arxiv : hep - ph/0612035 ` [ hep - ph ] .
v. lwen and h. p. nilles , phys.rev .
* d77 * ( 2008 ) , 106007 , ` arxiv:0802.1137 ` [ hep - ph ] . v. lwen and h. p. nilles , nucl.phys . * b827 * ( 2010 ) , 337 , ` arxiv:0907.4983 ` [ hep - ph ] , arxiv:0907.4983 [ hep - ph ] .
s. krippendorf , h. p. nilles , m. ratz , and m. w. winkler , phys.lett . *
b712 * ( 2012 ) , 87 , ` arxiv:1201.4857 ` [ hep - ph ] .
m. badziak , s. krippendorf , h. p. nilles , and m. w. winkler , jhep * 1303 * ( 2013 ) , 094 , ` arxiv:1212.0854 ` [ hep - ph ] . m. fischer , m. ratz , j. torrado , and p. k. vaudrevange , jhep * 1301 * ( 2013 ) , 084 , ` arxiv:1209.3906 ` [ hep - th ] . h. p. nilles , s. ramos - snchez , p. k. vaudrevange , and a. wingerter , comput.phys.commun .
* 183 * ( 2012 ) , 1363 , ` arxiv:1110.5229 ` [ hep - th ] .
s. groot nibbelink and o. loukas , jhep * 1312 * ( 2013 ) , 044 , ` arxiv:1308.5145 ` [ hep - th ] . h. p. nilles and p.
k. vaudrevange , _ the orbifoldlandscape : input files for the orbifolder _ , 2014 , `
http://www.th.physik.uni-bonn.de/ nilles / orbifoldlandscape/ ` .
d. k. m. pena , h. p. nilles , and p .- k .
oehlmann , jhep * 1212 * ( 2012 ) , 024 , ` arxiv:1209.6041 ` [ hep - th ]
. j. e. kim and b. kyae , nucl.phys . * b770 * ( 2007 ) , 47 , ` arxiv : hep - th/0608086 ` [ hep - th ] .
j. e. kim , j .- h .
kim , and b. kyae , jhep * 0706 * ( 2007 ) , 034 , ` arxiv : hep - ph/0702278 ` [ hep - ph ] .
s. ramos - snchez , fortsch.phys .
* 10 * ( 2009 ) , 907 , ` arxiv:0812.3560 ` [ hep - th ] , ph.d.thesis ( advisor : h.p .
nilles ) . k. r. dienes , phys.rev . * d73 * ( 2006 ) , 106010 , ` arxiv : hep - th/0602286 ` [ hep - th ] . k. r. dienes , m. lennek , d. senechal , and v. wasnik , phys.rev . * d75 * ( 2007 ) , 126005 , ` arxiv:0704.1320 ` [ hep - th ] . b. gato - rivera and a. schellekens , ` arxiv:1401.1782 ` [ hep - ph ] , arxiv:1401.1782 [ hep - ph ] . l. e. ibez and a. m. uranga , cambridge , uk : univ . pr .
( 2012 ) 673 p. r. blumenhagen , b. krs , d. lst , and s. stieberger , phys.rept .
* 445 * ( 2007 ) , 1 , ` arxiv : hep - th/0610327 ` [ hep - th ] .
g. honecker , m. ripka , and w. staessens , nucl.phys .
* b868 * ( 2013 ) , 156 , ` arxiv:1209.3010 ` [ hep - th ] .
a. e. faraggi , phys.lett . *
b278 * ( 1992 ) , 131 .
t. dijkstra , l. huiszoon , and a. schellekens , nucl.phys .
* b710 * ( 2005 ) , 3 , ` arxiv : hep - th/0411129 ` [ hep - th ] .
r. donagi , b. a. ovrut , t. pantev , and d. waldram , adv.theor.math.phys .
* 5 * ( 2002 ) , 93 , ` arxiv : hep - th/9912208 ` [ hep - th ] .
v. braun , y .- h .
he , b. a. ovrut , and t. pantev , phys.lett . * b618 * ( 2005 ) , 252 , ` arxiv : hep - th/0501070 ` [ hep - th ] . l. b. anderson , j. gray , a. lukas , and e. palti , phys.rev . * d84 * ( 2011 ) , 106005 , ` arxiv:1106.4804 ` [ hep - th ] . l. b. anderson , j. gray , a. lukas , and e. palti , jhep * 1206 * ( 2012 ) , 113 , ` arxiv:1202.1757 ` [ hep - th ] .
m. berasaluce - gonzalez , l. e. ibez , p. soler , and a. m. uranga , jhep * 1112 * ( 2011 ) , 113 , ` arxiv:1106.4169 ` [ hep - th ] .
l. ibez , a. schellekens , and a. uranga , nucl.phys .
* b865 * ( 2012 ) , 509 , ` arxiv:1205.5364 ` [ hep - th ] . m. berasaluce - gonzalez , p. camara , f. marchesano , d. regalado , and a. uranga , jhep * 1209 * ( 2012 ) , 059 , ` arxiv:1206.2383 ` [ hep - th ] .
anastasopoulos , m. cveti , r. richter , and p. k. vaudrevange , jhep * 1303 * ( 2013 ) , 011 , ` arxiv:1211.1017 ` [ hep - th ] . g. honecker and w. staessens , jhep * 1310 * ( 2013 ) , 146 , ` arxiv:1303.4415 ` [ hep - th ] . | string theoretical ideas might be relevant for particle physics model building .
ideally one would hope to find a unified theory of all fundamental interactions .
there are only few consistent string theories in @xmath0 or @xmath1 space - time dimensions , but a huge landscape in @xmath2 .
we have to explore this landscape to identify models that describe the known phenomena of particle physics .
properties of compactified six spatial dimensions are crucial in that respect .
we postulate some useful rules to investigate this landscape and construct realistic models .
we identify common properties of the successful models and formulate lessons for further model building .
+ * geography of fields in extra dimensions : + string theory lessons for particle physics * * hans peter nilles@xmath3 and patrick k.s .
vaudrevange@xmath4 * + @xmath5 bethe center for theoretical physics + and + physikalisches
institut der universitt bonn , + nussallee 12 , 53115 bonn , germany + @xmath6 excellence cluster universe , technische universitt mnchen , + boltzmannstr .
2 , d-85748 , garching , germany + and + arnold sommerfeld center for theoretical physics , lmu + theresienstrae 37 , 80333 mnchen , germany |
random tensor models ( see @xcite for a recent review ) generalize in dimension three ( and higher ) the celebrated matrix models ( see , for example , @xcite for a review ) . indeed , in the same way matrix models are related to combinatorial maps @xcite , tensor models in dimension three are related to tensor graphs or 3d maps . a certain case of random tensor models , the so - called colored tensor models , have been intensively studied in the recent years ( see @xcite for a review ) .
the graphs associated to these models are regular edge - colored graphs .
an important result is the asymptotic expansion in the limit @xmath1 ( @xmath0 being the size of the tensor ) , an expansion which was obtained in @xcite .
the role played by the genus in the 2d case is played here by a distinct integer called the _ degree_. the dominant graphs of this asymptotic expansion are the so - called melonic graphs , which correspond to particular triangulations of the three - dimensional sphere @xmath2 .
let us also mention here that a universality result generalizing matrix universality was obtained in the tensor case in @xcite .
a particularly interesting combinatorial approach for the study of these colored graphs was proposed recently by guru and schaeffer in @xcite , where they analyze in detail the structure of colored graphs of fixed degree and perform exact and asymptotic enumeration .
this analysis relies on the reduction of colored graphs to some terminal forms , called _
schemes_. an important result proven in @xcite is that the number of schemes of a given degree is finite ( while the number of graphs of a given degree is infinite ) .
nevertheless , a certain drawback of colored tensor models is that a large number of tensor graphs is discarded by the very definition of the model .
thus , a different type of model was initially proposed in @xcite , the 3d multi - orientable ( mo ) tensor model .
this model is related to tensor graphs which correspond to 3d maps with a particular eulerian orientation .
the set of mo tensor graphs contains as a strict subset the set of colored tensor graphs ( in 3d ) .
the asymptotic expansion in the limit @xmath1 for the mo tensor model was studied in @xcite , where it was shown that the same class of tensor graphs , the melonic ones , are the dominant graphs in this limit . the sub - dominant term of this expansion was then studied in detail in @xcite . in this paper
we implement a guru - schaeffer analysis for the mo random tensor model .
we investigate in detail the general term of the asymptotic expansion in the limit @xmath1 . as in the colored case ,
this is done by defining appropriate terminal forms , the schemes .
nevertheless , our analysis is somehow more involved from a combinatorial point of view , since , as already mentioned above , a larger class of 3d maps has to be taken into consideration .
also an important difference with respect to the colored model , which only allows for integer degrees , is that the mo model allows for both half - odd - integer and integer degrees .
this leads to the the fact that the dominant schemes are different from the ones identified in @xcite for the colored model ( interestingly , in both cases , dominant schemes are naturally associated to rooted binary trees ) .
let us also mention that the analysis of this paper may further allow for the implementation of the the so - called _
double scaling limit _ for the mo tensor model .
this is a particularly important mechanism for matrix models ( see again @xcite ) , making it possible to take , in a correlated way , the double limit @xmath1 and @xmath3 where @xmath4 is a variable counting the number of vertices of the graph and @xmath5 is some critical point of the generating function of schemes of a given degree .
in this section we recall the main definitions related to mo tensor graphs .
a _ map _ ( also called a fat - graph ) is a graph ( possibly with loops and multiple edges , possibly disconnected ) such that at each vertex @xmath6 the cyclic order of the @xmath7 incident half - edges ( @xmath7 being the degree of @xmath6 ) in clockwise ( cw ) order around @xmath6 is specified .
a _ corner _ of a map is defined as a sector between two consecutive half - edges around a vertex , so that a vertex of degree @xmath7 has @xmath7 incident corners .
a _ 4-regular map _ is a map where all vertices have degree @xmath8 .
a _ multi - orientable tensor graph _
, shortly called mo - graph hereafter , is a 4-regular map where each half - edge carries a sign , @xmath9 or @xmath10 , such that each edge has its two half - edges of opposite signs , and the two half - edges at each corner also have opposite signs .
in addition , for convenience , the half - edges at each vertex are turned into 3 parallel strands , see figure [ fig : exemple_mo_graph ] for an example . the strand in the middle is called _ internal _ , the two other ones are called _
external_. an external strand is called _ left _ if it is on the left side of a positive half - edge or on the right side of a negative half - edge ; an external strand is called _ right _ if it is on the right side of a positive half - edge or on the left side of a negative half - edge .
a _ face _ of an mo - graph is a closed walk formed by a closed ( cyclic ) sequence of strands .
external faces ( faces formed by external strands ) are the classical faces of the 4-regular map , while _ internal faces _ ( faces formed by internal strands ) , also called _ straight faces _ thereafter , are not faces of the 4-regular map . note also that external faces are either made completely of left external strands , or are made completely of right external strands ; accordingly external faces are called either left or right .
we finally define a _ rooted mo - graph _ as a connected mo - graph with a marked edge , which is convenient ( as in the combinatorial study of maps ) to avoid symmetry issues ( indeed , as for maps , mo graphs are _
unlabelled _ ) .
the _ degree _ of an mo - graph @xmath11 is the quantity @xmath12 defined by @xmath13 where @xmath14 are respectively the numbers of connected components , vertices , and faces ( including internal faces ) of @xmath11 . for @xmath15 and @xmath16 two mo - graphs , denote by @xmath17 the mo - graph made of disconnected copies of @xmath15 and @xmath16 .
note that the quantities @xmath14 for @xmath11 result from adding the respective quantities from @xmath15 and @xmath16 .
hence we have : [ claim : disconnected ] for two mo - graphs @xmath15 and @xmath16 , the degree of @xmath17 is the sum of the degrees of @xmath15 and @xmath16 . in particular , the degree of an mo - graph is the sum of the degrees of its connected components .
that @xmath12 follows from the following observation @xcite : @xmath11 gives rise to three 4-regular maps @xmath18 ( resp .
@xmath19 , @xmath20 ) , called the _ jackets _ of @xmath11 , which are obtained from @xmath11 by deleting straight faces ( resp . deleting right faces , deleting left faces ) , see figure [ fig : jacketex ] for an example . note that @xmath18 is an orientable map , while @xmath19 and @xmath20 are typically only locally orientable ( in a usual ribbon representation , edges have twists ) .
let @xmath21 , @xmath22 and @xmath23 be the respective genera of @xmath18 , @xmath19 , @xmath20 ( since @xmath18 is orientable , @xmath24 , while @xmath22 and @xmath23 are in @xmath25 ) .
let @xmath26 , @xmath27 , @xmath28 be the numbers of left faces , right faces , and straight faces of @xmath11 . denoting by @xmath29 the number of edges of @xmath11 , the euler relation gives @xmath30 , @xmath31 , @xmath32 . since @xmath33 and @xmath34 ( because the map is 4-regular ) we conclude that @xmath35 we call @xmath36 the _ ( canonical ) genus _ of @xmath11 .
an mo - graph @xmath11 is called _ planar _ if @xmath37 .
.,width=377 ] seeing edges as directed from the positive to the negative half - edge , mo - graphs may be seen as 4-regular maps where the edges are directed such that each vertex has two ingoing and two outgoing half - edges , and the two ingoing ( resp .
two outgoing ) half - edges are opposite , see figure [ fig : exemple_mo_graph_oriented ] for an example .
we call _ admissible orientations _ such orientations of 4-regular maps .
this point of view of mo - graphs as oriented 4-regular maps is quite convenient and will be adopted most of the time in the rest of the article ( except when we need to locally follow the strand structure ) .
note that the left faces are directed counterclockwise ( ccw ) , and the right faces are directed cw ; and the straight faces are closed walks that alternate in direction at each vertex they pass by .
, represented as an oriented 4-regular map.,width=302 ] since straight faces alternate at each vertex they pass by , straight faces have even length .
note also that admissible orientations are completely _ characterized _ by the property that the faces of the 4-regular map ( not including the straight faces ) are either cw or ccw ; this is also equivalent to coloring the faces of the map in two possible colors , say red or blue , such that at each edge there is a blue face on one side and a red face on the other side ( this is also directly visible given the definition with strands , where right faces are red and left faces are blue , see figure [ fig : exemple_mo_graph ] ) .
in other words , a 4-regular map can be endowed with an admissible orientation iff the dual of every of its connected components is a bipartite quadrangulation . if this is the case , and if there is a marked directed edge in each connected component , then there is a unique admissible orientation that fits with the prescribed edge directions ( existence follows from the above discussion , and there is clearly a unique way to propagate the directions starting from the marked directed edges ) .
recall that , in the context of maps , a _ rooted map _ means a connected map with a marked directed edge .
the above paragraph yields the following statement , which we think is worth mentioning even if it will not be needed to perform the combinatorial study ( scheme extraction and associated analysis ) of mo - graphs : [ rk : bij_4regular ] rooted mo - graphs having genus @xmath38 and @xmath39 vertices may be identified with rooted 4-regular maps having genus @xmath38 and @xmath39 vertices , and with the property that the dual rooted quadrangulation is bipartite ( these are themselves well - known to be in bijection with rooted connected maps having genus @xmath38 and @xmath39 edges ) . in genus
@xmath40 ( planar case ) , all rooted 4-regular maps have this property .
we conclude this subsection with a simple lemma relating the degree and the genus to the number of straight faces : [ lem : straight_faces ] let @xmath11 be a connected mo - graph .
let @xmath41 be the number of vertices , @xmath28 the number of straight faces , and for each @xmath42 , @xmath43 the number of straight faces of length @xmath44 in @xmath11 .
then the quantity @xmath45 satisfies @xmath46 where @xmath47 and @xmath38 are respectively the degree and the genus of @xmath11 .
hence @xmath48 .
since the number @xmath29 of edges is twice the number of vertices , we have @xmath49 , so that @xmath50 , hence @xmath51 . let @xmath26 , @xmath27 be the numbers of left faces and right faces of @xmath11 .
we have seen in section [ sec : degree ] that @xmath52 , where @xmath53 and @xmath54 substracting the sum of the two last equalities from the first one , we obtain : @xmath55 so that @xmath56 .
[ rk : doodles ] for @xmath37 , according to remark [ rk : bij_4regular ] , rooted mo - graphs correspond to rooted 4-regular planar maps .
seeing such a map as the planar projection of an entangled link that lives in the 3d space ( vertices of the map correspond to crossings , where the under / over information is omitted ) , @xmath28 is classically interpreted as the number of _ knot - components_. lemma [ lem : straight_faces ] gives @xmath57 . since @xmath58 and @xmath59 , the extremal cases are : ( 1 ) @xmath60 , in which case @xmath61 ; ( 2 ) @xmath62 , in which case @xmath63 ( the mo - graph of figure [ fig : exemple_mo_graph_oriented ] , having @xmath64 , @xmath65 , and thus @xmath66 , is intermediate ) . while the case @xmath60 is combinatorially well - understood ( as will be recalled in section [ sec : extract_melon_free_core ] ) , the case @xmath62 is notoriously difficult : in that case , 4-regular planar maps are projected diagrams of one - component links ( i.e. , knot diagrams ) , see @xcite and references therein .
we will go back to this remark in section [ sec : gen_funct ] ( on generating functions ) , where we will establish , for each fixed @xmath12 , the asymptotic enumeration of rooted 4-regular planar maps with @xmath67 vertices and with @xmath68 knot - components , as @xmath69 goes to infinity . as recalled here from @xcite ,
mo - graphs form a superfamily of the well studied _ regular colored graphs _ of dimension @xmath71 . for @xmath72
a _ regular colored graph of dimension @xmath73 _ is defined as a @xmath74-regular bipartite graph @xmath11 ( vertices are either black or white ) where the edges have a color in @xmath75 , such that at each vertex the @xmath76 incident edges have different colors .
a _ rooted _ colored graph is a connected colored graph with a marked edge of color @xmath40 . for @xmath77 , a _
face _ of type @xmath78 in a colored graph of dimension @xmath73 is a cycle made of edges that alternate colors in @xmath79 ; let @xmath80 be the number of faces of type @xmath78 in @xmath11 .
let @xmath81 be the number of connected components , @xmath82 the number of vertices and @xmath83 the total number of faces of @xmath11 , i.e. , @xmath84 .
the _ degree _ of @xmath11 is the integer @xmath47 given by @xmath85 dimensions .
right : the induced rooted mo - graph.,width=453 ] note that a colored graph has a canonical realization as a map , where the edge colors in cw ( resp .
ccw ) order around black ( resp .
white ) vertices are @xmath86 .
let @xmath11 be a rooted colored graph of dimension @xmath71 that is canonically embedded .
orienting the edges of even color from the black to the white extremity and the edges of odd color from the white to the black extremity , we obtain a rooted mo - graph @xmath87 , see figure [ fig : colored ] for an example .
clearly this gives an injective mapping , since there is a unique way ( when possible ) to propagate the edge colors starting from the root - edge .
hence , rooted colored graphs of dimension @xmath71 form a subfamily of rooted mo - graphs .
in addition @xmath88 so that @xmath87 and @xmath11 have the same total number of faces .
hence the degree formula for colored graphs is consistent with the degree formula for mo - graphs .
from now on , it is convenient to consider that a `` fake '' vertex of degree @xmath89 , called the _ root - vertex _ , is inserted in the middle of the root - edge of any rooted mo - graph ( so that the root - edge is turned into two edges ) . by convention also , it is convenient to introduce the following mo - graph : the _ cycle - graph _ is defined as an oriented self - loop carrying no vertex .
the cycle - graph is connected , has @xmath90 , @xmath91 ( one face in each type ) , hence has degree @xmath40 . in its
rooted version , the ( rooted ) _ cycle - graph _ is made of an oriented loop incident to the root - vertex . in a ( possibly rooted ) mo - graph @xmath11 , a _ melon _ is a triple edge @xmath92 such that none of the @xmath71 edges is the root - edge ( if @xmath11 is rooted ) , @xmath93 form a left face of length @xmath89 , @xmath94 form a right face of length @xmath89 , and @xmath95 form a straight face of length @xmath89 , see figure [ fig : melon ] . define the _ removal _ of a melon as the operation below ( where possibly @xmath96 , and possibly @xmath97 or @xmath6 might be the root if the mo - graph is rooted ) : the reverse operation ( where @xmath98 is allowed to be a loop , and is allowed to be incident to the root - vertex if the mo - graph is rooted ) is called the _ insertion _ of a melon at an edge . a rooted mo - graph @xmath11 is called _ melonic _ if it can be reduced to the rooted cycle - graph by successive removals of melons , see figure [ fig : melonic ] for an example .
an important remark is that , for a rooted melonic mo - graph , _ any _ greedy sequence of melon removals terminates at the cycle - graph .
it is known @xcite that rooted mo - graphs of degree @xmath40 are exactly the rooted melonic graphs . for @xmath11 a ( possibly rooted )
mo - graph , @xmath98 an edge of @xmath11 ( possibly a loop , possibly incident to the root - vertex ) , and @xmath99 a rooted mo - graph , define the operation of _ substituting @xmath98 by @xmath99 in @xmath11 _ as in the following generic drawing : we have the following bijective statement , which is the counterpart for mo - graphs of ( * ? ? ?
4 ) : [ prop : melon_free ] each rooted mo - graph @xmath11 is uniquely obtained as a rooted melon - free mo - graph @xmath100 called the _ melon - free core _ of @xmath11 where each edge @xmath98 is substituted by a rooted melonic mo - graph ( possibly the rooted cycle - graph , in which case @xmath98 is unchanged ) .
in addition , the degree of @xmath11 equals the degree of @xmath100 .
let @xmath100 be the rooted melon - free graph obtained from @xmath11 after performing a maximal greedy sequence of melon removals .
conversely , @xmath11 is obtained from @xmath100 where a sequence of melon insertions is performed .
hence @xmath11 is equal to @xmath100 where each edge @xmath98 is substituted by a rooted melonic graph @xmath101 .
this gives the existence of a melon - free core .
uniqueness of the melon - free core is given by the observation that any other maximal greedy sequence of melon removals starting from @xmath11 has to progressively shell the melonic components @xmath101 , hence terminates at @xmath100 .
finally , it is clear that @xmath11 and @xmath100 have the same degree , since a melon insertion preserves the degree ( it preserves the number of connected components , increases the number of vertices by @xmath89 , and increases the number of faces by @xmath71 ) . , two faces of length @xmath102 , and have degree @xmath103.,width=226 ] in this section we go a step further and define the process of extracting a _ scheme _ from a rooted melon - free mo - graph .
2-dipole _ , shortly a dipole , in a ( possibly rooted ) mo - graph @xmath11 as a face of length @xmath89 incident to two distinct vertices , and not passing by the root if @xmath11 is rooted . accordingly , one has three distinct types of dipoles : l , r , or s , see figure [ fig : dipoles ] ( left part ) . as the figure shows , each dipole has two exterior half - edges on one side and two exterior half - edges on the other side .
notice also that a double edge does not necessarily delimit a dipole , as shown in figure [ fig : dipoles ] ( right part ) .
[ rk : after_def_dipole ] a face of length @xmath89 is actually always incident to two distinct vertices , except in the mo - graphs that are made of one vertex and two loops .
these two graphs , shown in figure [ fig : infinity ] , are called the clockwise and the counterclockwise infinity graph , respectively .
they have no dipole but have two faces of length @xmath89 . in an mo - graph @xmath11 , define a _
chain _ as a sequence of dipoles @xmath104 ( not passing by the root if @xmath11 is rooted ) such that for each @xmath105 , @xmath106 and @xmath107 are connected by two edges involving two half - edges on the same side of @xmath106 and two half - edges on the same side of @xmath107 , see figure [ fig : chain_dipoles ] .
a chain is called _ unbroken _ if all the @xmath108 dipoles are of the same type .
a _ broken chain _ is a chain which is not unbroken .
proper chain _ is a chain of at least two dipoles .
a proper chain is called _ maximal _ if it can not be extended into a larger proper chain . by very similar arguments as in lemma 8 of
@xcite one obtains the following result : dipoles .
right : a sequence of 3 dipoles that do not form a chain ( the 3rd element , a dipole of type @xmath109 , is connected to the 2nd element by two half - edges on opposite sides).,width=453 ] [ claim : disjoint ] in a rooted mo - graph , any two maximal proper chains are vertex - disjoint .
let @xmath11 be a rooted melon - free mo - graph .
the _ scheme _ of @xmath11 is the graph obtained by simultaneously replacing any maximal proper chain of @xmath11 by a so - called _ chain - vertex _ , as shown in figure [ fig : chains ] , see also figure [ fig : example_scheme ] for an example . ) in an mo - graph with chain - vertices.,width=170 ] since a scheme is not an mo - graph ( due to the presence of chain - vertices ) , we need to extend the definition and main properties seen so far in order to allow for chain - vertices . in the following , an _ mo - graph with chain - vertices
_ is an oriented 4-regular map with two types of vertices : standard vertices and chain - vertices ( each chain - vertex being labelled by either @xmath110 ) , so as to satisfy the local rules of figure [ fig : rules_mo_graph_chain ] .
such a graph is possibly rooted , i.e. , has a fake - vertex of degree @xmath89 in the middle of some edge .
note that the class of ( rooted ) mo - graph with chain - vertices is _ larger _ than the class of schemes ( we will see below which rooted mo - graphs with chain - vertices are schemes ) . in order to compute the number of faces of an mo - graph @xmath87 with chain - vertices , we need to specify locally the _ strand structure _ at every type of chain - vertex ; as for mo - graphs , we imagine here that each edge is turned into @xmath71 parallel strands and we have to specify at each chain - vertex how the incident strands go through the vertex .
the specification is given by figure [ fig : chain_strands ] ; in the case of an unbroken chain - vertex the two strands that go through the chain - vertex are called the _ crossing strands _ at that chain - vertex .
note that this specification gives the natural strand structure to expect whenever a chain - vertex @xmath111 is to be consistently substituted by a chain @xmath81 ( for instance a chain - vertex of type @xmath112 is to be susbtituted by an unbroken chain of dipoles of type @xmath112 ) , that is , the strand structure of @xmath111 reflects how the strands arriving at @xmath81 are routed ( some bouncing back , some going through @xmath81 ) .
then , as for classical mo - graphs , a _
face _ is a closed walk formed from strands .
the _ degree _ @xmath47 of @xmath87 is defined as @xmath113 where @xmath81 , @xmath41 , @xmath83 are as usual the numbers of connected components , standard vertices , and faces , and where @xmath114 and @xmath115 stand respectively for the numbers of unbroken chain - vertices and broken chain - vertices .
an edge @xmath98 of @xmath87 is said to be _ adjacent _ to a chain - vertex if the two half - edges of @xmath98 are the two half - edges on the same side of a chain - vertex of @xmath87 .
then @xmath87 is said to be _ melon - free _ if it has no melon nor an edge adjacent to a chain - vertex .
it is easy to see that an mo - graph is melon - free iff its scheme is melon - free . in a melon - free (
possibly rooted ) mo - graph @xmath87 with chain - vertices , define a _ chain _ as a sequence @xmath104 of elements that are either dipoles ( not passing by the root if @xmath87 is rooted ) or chain - vertices , such that for each @xmath105 , @xmath106 and @xmath107 are connected by two edges involving two half - edges on the same side of @xmath106 and two half - edges on the same side of @xmath107 , see figure [ fig : chain_dipoles_chain_vert ] . a _
proper chain _ is a chain of at least two elements .
now define a _
reduced scheme _ as a rooted melon - free mo - graph with chain - vertices and with no proper chain . by construction ,
the scheme of a rooted melon - free mo - graph ( with no chain - vertices ) is a reduced scheme .
claim [ claim : disjoint ] then easily yields the following bijective statement : every rooted melon - free mo - graph is uniquely obtained as a reduced scheme where each chain - vertex is consistently substituted by a chain of at least two dipoles ( consistent means that if the chain - vertex is of type @xmath112 , then the substituted chain is an unbroken chain of l - dipoles , etc ) .
the following result ensures that the degree definition for mo - graphs with chain - vertices is consistent with the replacement of chains by chain - vertices : [ lem : substitute ] let @xmath11 be an mo - graph with chain - vertices . and let @xmath99 be an mo - graph with chain - vertices obtained from @xmath11 by consistently substituting a chain - vertex by a chain of dipoles .
then the degrees of @xmath11 and @xmath99 are the same .
let @xmath39 be the chain - vertex where the substitution takes place , and let @xmath116 be the number of dipoles in the substituted chain .
denote by @xmath117 the parameters ( degree , numbers of connected components , standard vertices , faces , unbroken chain - vertices , broken chain - vertices ) for @xmath11 , and by @xmath118 the parameters for @xmath99 .
clearly @xmath119 , @xmath120 .
if @xmath39 is unbroken , then @xmath121 and @xmath122 .
say @xmath39 is of type @xmath112 ( the arguments for types @xmath123 are similar ) .
then the strand structure remains the same , except for @xmath116 new faces of type @xmath112 ( of degree @xmath89 ) , @xmath124 new faces of type @xmath125 ( of degree @xmath8 ) and @xmath124 new faces of type @xmath109 ( of degree @xmath8 ) inside the substituted chain .
hence @xmath126 .
consequently , @xmath127 .
if @xmath39 is broken , then @xmath128 and @xmath129 .
the strand structure remains the same , except for @xmath124 new faces of each type @xmath130 inside the substituted chain ( an easy case inspection ensures that substituting a chain of two dipoles of different types always bring @xmath71 new faces , one of each type , and any additional dipole brings @xmath71 new faces , one of each type ) .
hence @xmath131 .
consequently , @xmath127 .
a rooted melon - free mo - graph has the same degree as its scheme .
the main result in this section is the following : [ prop : finite ] for each @xmath12 , the set of reduced schemes of degree @xmath47 is finite . similarly as in @xcite , this result will be obtained from two successive lemmas , proved respectively in section [ sec : proof_lem_bound1 ] and section [ sec : proof_lem_bound2 ] : [ lem : bound1 ] for each reduced scheme of degree @xmath47 , the sum @xmath132 of the numbers of dipoles and chain - vertices satisfies @xmath133 .
[ lem : bound2 ] for @xmath134 and @xmath12 , there is a constant @xmath135 ( depending only on @xmath116 and @xmath47 ) such that any connected unrooted mo - graph ( without chain - vertices ) of degree @xmath47 with at most @xmath116 dipoles has at most @xmath135 vertices .
let us show how lemmas [ lem : bound1 ] and [ lem : bound2 ] imply proposition [ prop : finite ] .
let @xmath109 be a reduced scheme of degree @xmath47 , and let @xmath136 be the rooted mo - graph with no chain - vertices obtained by substituting each chain - vertex of @xmath109 by a chain of dipoles of length @xmath89 or @xmath71 as shown in figure [ fig : substitute_canonical ] .
as already seen in lemma [ lem : substitute ] , such substitutions preserve the degree , and the mapping @xmath137 is injective . in addition , the number of dipoles of @xmath136 is at most three times the total number of dipoles and chain - vertices of @xmath109 , hence @xmath136 has at most @xmath138 dipoles according to lemma [ lem : bound1 ] .
unrooting @xmath136 might increase the number of dipoles by up to @xmath89 ( recall that dipoles are not counted in rooted mo - graphs if they pass by the root , and at most @xmath89 dipoles can pass by the root , the case of @xmath89 additional dipoles happening when the root - edge belongs to a melon ) . hence
, writing @xmath139 , lemma [ lem : bound2 ] ensures that @xmath136 has at most @xmath135 vertices . since there is an injective mapping from reduced schemes of degree @xmath47 to rooted mo - graphs of size bounded by the fixed quantity @xmath135
, we conclude that the number of reduced schemes of degree @xmath47 is finite .
we finally state the following `` loop - removal '' lemma , which will be useful at some points in the following : [ lem : remove_loop ] let @xmath11 be an mo - graph of degree @xmath12 , let @xmath140 be a loop of @xmath11 , and let @xmath99 be the mo - graph obtained from @xmath11 by erasing the loop and its incident vertex , as shown in figure [ fig : remove_loop ] .
then @xmath99 has degree @xmath141 .
hence @xmath11 has at most @xmath142 loops .
clearly @xmath99 has the same number of connected components as @xmath11 , has one vertex less , and has one face less ( which has length @xmath102 ) .
we analyze here how the degree of an mo - graph with chain - vertices ( not necessarily a reduced scheme , possibly with melons ) evolves when removing a chain - vertex .
we then have a similar analysis for the removal of a dipole .
let @xmath11 be an mo - graph with chain - vertices , and let @xmath39 be a chain - vertex of @xmath11 .
the _ removal _ of @xmath39 consists of the following operations ( see the left part of figure [ fig : deletion ] ) : 1 . delete @xmath39 from @xmath11 ; 2 . on each side of @xmath39 , connect together the two detached legs ( without creating a new vertex ) .
denote by @xmath99 the resulting graph .
the chain - vertex @xmath39 is said to be _ non - separating _
if @xmath99 has the same number of connected components as @xmath11 , and _ separating _ otherwise ( in which case @xmath99 has one more connected component ) . in the separating case
, it might be that one of the two resulting connected components is the cycle - graph ( this happens if two half - edges on one side of @xmath39 belong to the same edge ) .
let @xmath143 be the parameters for @xmath11 ( degree , numbers of connected components , standard vertices , faces , unbroken chain - vertices , broken chain - vertices ) , and let @xmath144 be the parameters for @xmath99 .
assume first that @xmath39 is a broken chain - vertex .
we clearly have @xmath145 , @xmath146 ( the strand structure is the same before and after removal ) , @xmath128 , @xmath129 and @xmath119 ( resp .
@xmath147 ) if @xmath39 is non - separating ( resp . separating ) .
hence , if @xmath39 is separating , then @xmath148 , so that @xmath127 ; and if @xmath39 is non - separating , then @xmath149 , so that @xmath150 .
assume now that @xmath39 is an unbroken chain - vertex .
we clearly have @xmath145 , @xmath121 , @xmath122 , and @xmath119 ( resp .
@xmath147 ) if @xmath39 is non - separating ( resp . separating ) .
if the two crossing strands at @xmath39 belong to a same face , then that face splits into two faces when removing @xmath39 , hence @xmath151 ; as opposed to that , if the two crossing strands belong to different faces , then these two faces are merged when removing @xmath39 , hence @xmath152 . moreover , the two crossing strands are easily seen to always be in the same face if @xmath39 is separating .
we conclude that , if @xmath39 is separating , then @xmath153 , so that @xmath127 ; and if @xmath39 is non - separating then @xmath154 , where @xmath155 , hence either @xmath156 or @xmath150 .
we now consider the operation of removing a dipole @xmath7 from @xmath11 , removal which consists of the following operations ( the 3 cases are shown in the right - part of figure [ fig : deletion ] ) : 1 .
delete the two vertices and the two edges of @xmath7 from @xmath11 ; 2 . on each side of @xmath7 ,
connect together the two detached legs ( without creating a new vertex ) .
let @xmath99 be the resulting graph .
again , @xmath7 is called _ non - separating _ if @xmath99 has as many connected components as @xmath11 , and _ separating _ otherwise .
observe that the operation of removing @xmath7 is the same as substituting @xmath7 by an unbroken chain - vertex @xmath39 ( of the same type as @xmath7 , so that the substitution does not change the degree according to lemma [ lem : substitute ] ) and then removing @xmath39 . from the discussion on the degree variation when removing @xmath39
, we conclude that , if @xmath7 is separating , then the degree is the same in @xmath99 as in @xmath11 , and if @xmath7 is non - separating then the degree of @xmath99 equals the degree of @xmath11 minus a quantity that is either @xmath102 or @xmath71 .
this analysis thus leads to the following statement : [ lem : removals ] the degree is unchanged when removing a separating chain - vertex or a separating dipole ( hence , by claim [ claim : disconnected ] , the degree is distributed among the resulting components ) .
the degree decreases by @xmath71 when removing a non - separating broken chain - vertex .
the degree decreases by @xmath102 or @xmath71 when removing a non - separating unbroken chain - vertex or a non - separating dipole .
[ coro : removals ] an mo - graph with chain - vertices and with degree @xmath157 has no non - separating dipole nor non - separating chain - vertex .
we show lemma [ lem : bound1 ] , i.e. , that for each reduced scheme @xmath11 of degree @xmath47 , the sum @xmath132 of the numbers of dipoles and chain - vertices in @xmath11 satisfies @xmath133 . the total number of ( uncolored ) dipoles.,width=377 ] .
then , iteratively , one removes at each step a non - separating dipole or a non - separating chain - vertex ( at each step the non - separating dipole or chain - vertex to be removed next is surrounded ) .
let @xmath99 be the mo - graph with chain - vertices thus obtained ( where colored edges are drawn bolder ) .
as the last two drawings show , the removal of uncolored dipoles and chain - vertices ( which are all separating ) of @xmath99 yields a tree of components ( a tree edge is labelled @xmath158 if it comes from a chain - vertex of type @xmath159 and is labelled @xmath160 if it comes from an uncolored dipole of type @xmath159).,width=491 ] as a first step , starting from @xmath11 , as long as there is at least one non - separating dipole or one non - separating chain - vertex , delete it ( as in figure [ fig : deletion ] ) and color the two resulting edges .
let @xmath99 be the resulting graph with chain - vertices , and let @xmath161 be the number of removals from @xmath11 to @xmath99 . according to lemma [ lem : removals ] ,
each non - separating removal decreases the degree by at least @xmath102 , hence @xmath162 , @xmath99 has degree at most @xmath163 , and @xmath99 has at most @xmath164 blue edges . during the sequence of steps from @xmath11 to @xmath99 ,
denote by @xmath0 the current sum of the numbers of chain - vertices and uncolored dipoles ( dipoles with no colored edge ) .
it is easy to see that each removal decreases @xmath0 by at most @xmath71 ( the worst case is shown in figure [ fig : remove_three ] ) .
hence , if we denote by @xmath165 the value of @xmath0 for @xmath99 , we have @xmath166 .
call a _ connection _ a chain - vertex or uncolored dipole of @xmath99 . since all dipoles and chain - vertices of @xmath99 are separating , @xmath99 can be seen as a tree @xmath167 of `` components '' ( see the last row of figure [ fig : from_scheme_to_tree ] for an example ) , where the edges of @xmath167 correspond to the @xmath165 connections of @xmath99 , and the nodes of @xmath167 correspond to the connected components of @xmath99 left after removing the connections ( note that some of these connected components might be cycle - graphs , including the component carrying the root ) . for a component @xmath168 of @xmath167 ,
the _ adjacency order _ of @xmath168 is the number of adjacent components in @xmath167 .
an edge @xmath98 of @xmath168 resulting from a ( removed ) connection is called a _ marked edge_. as shown in figure [ fig : merger ] , every marked edge @xmath98 is involved in @xmath169 connections ( hence the adjacency order of a component is at least its number of marked edges ) , and results from the merger of @xmath170 edges of @xmath99 ; @xmath98 is called _ colored _ if at least one of these @xmath170 edges is colored .
a component @xmath168 is called _ colored _ if if it bears either a colored edge of @xmath99 left untouched by removals , or a colored marked edge ; and it is called _ uncolored _ otherwise . thus , there are at most @xmath164 colored components in @xmath99 . of a component @xmath168 of @xmath99 such that @xmath98 is involved in @xmath169 connections , @xmath98 results from the merger of @xmath170 edges of @xmath99 ( @xmath171 in the example , @xmath98 results from the merger of @xmath172).,width=453 ] similarly as in @xcite , an important remark is the following : [ claim : deg_three ] an uncolored component @xmath168 of @xmath167 of degree @xmath40 and not containing the root has adjacency order at least @xmath71 .
the first unrooted mo - graphs of degree @xmath40 are the cycle - graph ( no vertex ) and the so - called `` quadruple - edge graph '' ( the mo - graph whose underlying map is of genus @xmath40 with two vertices connected by @xmath8 edges ) . and any other mo - graph of degree @xmath40 can be obtained from the quadruple - edge graph by successive melon insertions .
this easily implies that , if @xmath168 is not the cycle - graph and has adjacency order smaller than @xmath71 ( hence has less than @xmath71 marked edges ) , then @xmath168 has a dipole not passing by any marked edge , a contradiction . if @xmath168 is the cycle - graph and has adjacency order @xmath102 , then it clearly yields a melon in @xmath11 , a contradiction . finally , if @xmath168 is the cycle - graph and has adjacency order @xmath89 , then the two edges of @xmath167 at @xmath168 each of which either corresponds to a separating dipole or to a chain - vertex of @xmath11 together form a proper chain ( of two elements ) of @xmath11 , a contradiction . denote by @xmath173 the number of uncolored components in @xmath99 of degree @xmath40 and not containing the root , @xmath174 the number of other components of degree @xmath40 ( colored or containing the root ) , and @xmath175 the number of components of positive degree .
the degrees of the components add up to the degree of @xmath99 ( recall that , for separating removals , the degree is distributed among the components ) , and each component of positive degree has degree at least @xmath103 , hence @xmath176 .
the components counted by @xmath173 have at least @xmath71 neighbours in @xmath167 , hence @xmath177 .
since @xmath167 has at most @xmath164 colored components , we have @xmath178 ( the @xmath179 accounting for the component containing the root ) . and
since @xmath167 is a tree , we have @xmath180 hence , if we eliminate @xmath173 using @xmath177 , we obtain @xmath181 so that @xmath182 . since @xmath183 , we conclude that @xmath133 .
we prove here lemma [ lem : bound2 ] , i.e. , that for any @xmath12 and @xmath134 , there exists some constant @xmath135 such that any connected unrooted mo - graph ( with no chain - vertices ) of degree @xmath47 with at most @xmath116 dipoles has at most @xmath135 vertices , so that there are finitely many mo - graphs of degree @xmath47 with @xmath116 dipoles .
the bound @xmath135 we obtain is linear in @xmath116 and @xmath47 , with quite large multiplicative constants ( we have not pushed to improve the bound , also to avoid making the proof too complicated ) .
the proof is a bit more technical than the analogous statement for colored graphs in @xcite , because in the mo model , only straight faces have even length , while left and right faces are allowed to have odd lengths .
precisely , our proof strategy is to show in a series of claims that some properties ( e.g. being incident to a dipole ) are satisfied only by a limited ( bounded by a quantity depending only on @xmath116 and @xmath47 ) number of vertices .
hence if @xmath11 could be arbitrarily large , it would have vertices not safisfying these properties .
as we will see , this would yield the contradiction that @xmath11 has to be reduced to a certain mo - graph with @xmath184 vertices .
in all the proof , @xmath11 denotes a connected unrooted mo graph with no chain - vertices , of degree @xmath47 and with at most @xmath116 dipoles ; @xmath185 are respectively the numbers of vertices , edges , and faces of @xmath11 . moreover , for @xmath42
, @xmath186 is the number of straight faces of length @xmath44 . according to remark [ rk : after_def_dipole ] ,
the only mo - graphs where the set of dipoles is not equal to the set of faces of length @xmath89 are the two infinity graphs ( both having one vertex ) .
hence we can assume that @xmath11 is not equal to these two graphs .
thus @xmath187 , which gives : [ claim : pf1 ] there are at most @xmath188 vertices of @xmath11 that are incident to a dipole . also , lemma [ lem : straight_faces ] directly implies @xmath189 since @xmath190 for @xmath191 , the total length @xmath192 of straight faces of length larger than @xmath8 is bounded by @xmath193 , so that we obtain : [ claim : pf2 ] there are at most @xmath193 vertices of @xmath11 incident to a straight face of length larger than @xmath8 .
a cycle is called _ self - intersecting _ if it passes twice by a same vertex .
that does not create a loop . on the right
we see that it is not possible to orient the edges so as to satisfy the local condition at the vertices @xmath97 and @xmath6 if the local condition is satisfied at @xmath194.,width=377 ] [ claim : pf3 ] there are at most @xmath142 self - intersecting straight faces of length @xmath8 in @xmath11 .
hence there are at most @xmath195 vertices incident to such faces .
as we see in figure [ fig : self_intersect ] , a self - intersecting straight face of length @xmath8 must contain a loop , which yields the first statement according to lemma [ lem : remove_loop ] ( there are at most @xmath142 loops ) .
the second statement just follows from the fact that these cycles ( due to the self - intersection ) have at most @xmath71 vertices .
given a non - self - intersecting cycle @xmath168 of @xmath11 , a _
chordal path _ for @xmath168 is a path that starts from @xmath168 , ends at @xmath168 ( possibly at the same vertex ) , and is outside of @xmath168 inbetween . in the context of maps
, we call a chordal path _ faulty _ if it starts on one side of @xmath168 and ends on the other side of @xmath168
. a cycle @xmath168 is called _ faulty _ if it has a faulty chordal path , and , for @xmath134 , is called _ @xmath116-faulty _ if it has a chordal path of length at most @xmath116 . for @xmath168 a faulty cycle ,
we denote by @xmath196 the union of @xmath168 and of a ( canonically chosen ) shortest possible faulty chordal path of @xmath168 ; @xmath196 is called the _ faulty extension _ of @xmath168 .
. there are at most @xmath198 4-faulty cycles of length @xmath8 in @xmath11 .
hence there are at most @xmath199 vertices incident to such cycles . for @xmath168 a faulty cycle
, @xmath196 is clearly a topological minor of @xmath11 of genus @xmath102 .
hence , if we have @xmath200 vertex - disjoint faulty extensions of faulty cycles , these yield a topological minor of genus @xmath201 .
since the genus of @xmath11 is bounded by @xmath142 , we conclude that there can not be more than @xmath142 vertex - disjoint faulty extensions of faulty cycles .
let @xmath202 be the set of 4-faulty cycles of length @xmath8 of @xmath11 , and let @xmath203 ( where we distinguish the cycle from the faulty path in every faulty extension , so that @xmath204 ) .
any @xmath205 has at most @xmath206 vertices , and an easy calculation ensures that any given vertex @xmath6 of @xmath11 belongs to at most @xmath197 elements of @xmath207 ( this can probably be improved , but we just aim at a certain fixed bound ) .
hence @xmath196 can intersect at most @xmath208 other elements of @xmath207 .
we easily conclude that @xmath209 ( otherwise one could construct incrementally a subset of @xmath210 vertex - disjoint elements of @xmath207 ) .
let @xmath211 be the set of vertices of @xmath11 that are either incident to a dipole , or incident to a straight face of length larger than @xmath8 , or incident to a self - intersecting straight face of length @xmath8 , or incident to a 4-faulty cycle of length @xmath8 .
the four claims above give @xmath212 , width=302 ] and let @xmath213 be the set of vertices in @xmath211 or adjacent to a vertex in @xmath211 . since @xmath11 is @xmath8-regular , @xmath214 .
we are now going to show that if @xmath11 has a vertex @xmath6 not in @xmath213 , then @xmath11 is actually reduced to having only @xmath184 vertices . since @xmath215 , it is incident to two non - intersecting straight faces @xmath216 and @xmath217 , both of length @xmath8 . since @xmath216 is not faulty , it has to meet @xmath217 at a vertex different from @xmath6 ( otherwise @xmath217 would be a faulty chordal path for @xmath216 ) .
let @xmath218 be an intersection of @xmath216 and @xmath217 different from @xmath6 .
for @xmath219 , let @xmath106 be the distance of @xmath218 from @xmath6 on @xmath220 ( note that @xmath221 ) . if @xmath222 then either the two corresponding edges @xmath223 of @xmath216 and @xmath217 form an external dipole , or otherwise @xmath224 is a faulty chordal edge for @xmath216 .
both cases are excluded since @xmath225 , hence @xmath226 .
we write the cycle @xmath216 as @xmath227 and the cycle @xmath217 as @xmath228 , see figure [ fig : cycles ] .
let @xmath229 be the straight face ( non - intersecting and of length @xmath8 , since @xmath230 ) passing by @xmath231 and different from @xmath216 .
since @xmath230 , @xmath229 is not faulty , hence it intersects the 4-cycle @xmath232 at a vertex different from @xmath231 ; this other intersection must be at @xmath233 because @xmath6 and @xmath218 have their @xmath8 incident edges already in @xmath234 .
similarly @xmath229 intersects the 4-cycle @xmath235 at a vertex different from @xmath231 , and this other intersection must be @xmath236 ; and @xmath229 intersects the 4-cycle @xmath216 at a vertex different from @xmath231 , and this other intersection must be @xmath237 , see figure [ fig : cycles ] .
hence @xmath238 forms a connected induced subgraph that saturates all its vertices ( each of the @xmath184 vertices of @xmath99 is incident to @xmath8 edges of @xmath99 ) . hence @xmath239 .
in other words , if @xmath11 has a vertex outside of @xmath213 , then @xmath11 has @xmath184 vertices .
we can now easily conclude the proof .
define @xmath240 , and assume @xmath11 has more than @xmath135 vertices .
since @xmath214 , @xmath11 has a vertex not in @xmath213 , which implies that @xmath11 is reduced to a graph with @xmath184 vertices , giving a contradiction .
in this section we recover the results of @xcite about the structure of mo - graphs of degree @xmath103 , using ingredients from the preceding sections . [
prop : deg_one_half ] the only unrooted connected melon - free mo - graphs of degree @xmath103 are the cw and the ccw infinity graph .
assume there is a connected melon - free mo - graph @xmath11 of degree @xmath103 different from the cw or the ccw infinity graph .
according to remark [ rk : after_def_dipole ] the set of dipoles of @xmath11 equals the set of its faces of length @xmath89 .
note that in degree @xmath103 , the genus @xmath38 has to be @xmath40 ( since @xmath38 is a nonnegative integer bounded by @xmath47 ) .
hence lemma [ lem : straight_faces ] gives @xmath241 , so that @xmath242 , i.e. , @xmath11 has a dipole @xmath7 ( of type @xmath109 ) . since @xmath243
, @xmath7 has to be separating according to corollary [ coro : removals ] .
let @xmath15 and @xmath16 be the connected components resulting from the removal of @xmath7 . since @xmath7 is separating , the degrees of @xmath15 and @xmath16 add up to @xmath103 , hence one has degree @xmath40 and the other has degree @xmath103 . by convention
let @xmath15 be the one of degree @xmath40 , considered as rooted at the edge resulting from the deletion of @xmath7 .
there are two cases : ( i ) if @xmath15 is the cycle - graph , then @xmath7 is part of a melon of @xmath11 , giving a contradiction ; ( ii ) if @xmath15 is not the cycle - graph , then it has a melon ( since rooted mo - graphs of degree @xmath40 are melonic ) , giving again a contradiction .
this yields the following bijective result , where we recover ( * ? ? ?
3.1 ) : every unrooted connected mo - graph of degree @xmath103 is uniquely obtained from the cw or the ccw infinity graph where each of the two edges is substituted by a rooted melonic graph .
all the arguments in the proof of proposition [ prop : melon_free ] ( extraction of a unique melon - free core ) can be directly recycled to give the following statement : `` for @xmath244 , every unrooted connected mo - graph of degree @xmath47 is uniquely obtained as an unrooted connected melon - free mo - graph of degree @xmath47 where each edge is substituted by a rooted melonic graph '' , which together with proposition [ prop : deg_one_half] yields the result
. another direct consequence of proposition [ prop : deg_one_half ] is the following statement that will prove useful in the next section : [ coro : dipole12 ] every rooted connected mo - graph of degree @xmath103 different from the ( rooted ) cw or ccw infinity graph has a dipole . according to proposition [ prop : deg_one_half ] ,
the underlying unrooted mo - graph ( obtained by erasing the fake - vertex of degree @xmath89 ) has a melon , hence there is a dipole not including the root edge ; by definition this dipole is also a dipole of the rooted mo - graph ( recall that dipoles of rooted mo - graphs are required not to pass by the root ) .
a reduced scheme @xmath109 of degree @xmath12 is called _ dominant _ if it maximizes ( over reduced schemes of degree @xmath47 ) the number @xmath245 of broken chain - vertices ( as we will see in section [ sec : gen_funct ] , @xmath245 determines the singularity order of the generating function of rooted mo - graphs of degree @xmath47 and reduced scheme @xmath109 , the larger @xmath245 the larger the singularity order ) . in this section
we show that , in each degree @xmath47 , the maximal number of broken chain - vertices is @xmath246 , and we precisely determine what are the dominant schemes .
let @xmath109 be a reduced scheme of degree @xmath47 , and let @xmath245 be the number of broken chain - vertices of @xmath109 .
we now determine a bound for @xmath245 in terms of @xmath47 .
the discussion is very similar to the one of section [ sec : proof_lem_bound1 ] , with the difference that only the broken chain - vertices are removed .
first , as long as there is a non - separating broken chain - vertex , we remove it ( as shown in figure [ fig : deletion ] ) and color the two resulting edges .
let @xmath161 be the number of such removals until all broken chain - vertices are separating ; at this stage , denote by @xmath247 the resulting mo - graph with chain - vertices , which has degree @xmath248 according to lemma [ lem : removals ] , and has at most @xmath164 colored edges .
then @xmath247 can be seen as a tree @xmath167 of components that are obtained by removing all separating broken chain - vertices .
it is also convenient to consider the tree @xmath249 of components that are obtained by removing all separating chain - vertices and all separating uncolored dipoles dipoles with no colored edge of @xmath247 ( in @xmath167 and @xmath249 some components of degree @xmath40 might be cycle - graphs ) .
note that @xmath249 is a refinement of @xmath167 , i.e. , each component @xmath168 of @xmath167 `` occupies '' a subtree @xmath250 of @xmath249 . for a component @xmath168 of @xmath167 ( resp . of @xmath249 ) , the _ adjacency order _ of @xmath168 is defined as the number of adjacent components in @xmath167 ( resp . in @xmath249 ) .
an edge @xmath98 of @xmath168 resulting from a separating removal for @xmath167 ( resp .
@xmath249 ) is called a _ marked edge_. such an edge @xmath98 results from the merger of edges of @xmath247 . then @xmath98 is called _ colored _ if at least one of these edges is a colored edge of @xmath247 . and a component @xmath168 of @xmath167 ( resp . of @xmath249 )
is called _ uncolored _ if it bears no colored edge ( marked or unmarked ) . since @xmath247 has at most @xmath164 colored edges , @xmath167 ( resp .
@xmath249 ) has at most @xmath164 colored components .
consider an uncolored component @xmath168 of @xmath167 of degree @xmath40 and not containing the root .
note that @xmath168 has no non - separating dipoles nor non - separating chain - vertices , by corollary [ coro : removals ] .
in addition , by the arguments of claim [ claim : deg_three ] , any component of @xmath250 must have adjacency order at least @xmath71 , otherwise it would yield a melon or a proper chain in @xmath109 .
this clearly implies that @xmath168 can not have adjacency order smaller than @xmath71 , and if it has adjacency order @xmath71 , then @xmath250 consists of a unique component .
let @xmath173 be the number of components of @xmath167 of degree @xmath40 , uncolored , and not containing the root .
let @xmath174 be the number of other components of degree @xmath40 .
let @xmath175 be the number of components of positive degree .
moreover , let @xmath108 be the number of edges of @xmath167 , which is also the number of broken chain - vertices of @xmath247 , so that @xmath251 .
the degrees of the components add up to the degree @xmath248 of @xmath99 ( recall that , for separating removals , the degree is distributed among the components ) , and each component of positive degree has degree at least @xmath103 , hence @xmath252 the components counted by @xmath173 have at least @xmath71 neighbours in @xmath167 , hence @xmath253 the graph @xmath247 has at most @xmath164 blue edges , hence @xmath254 where the @xmath179 accounts for the component containing the root .
since @xmath167 is a tree , we have @xmath255 hence , eliminating @xmath173 using @xmath256 , we obtain @xmath257 so that @xmath258 .
hence @xmath259 .
if a scheme @xmath109 is such that @xmath245 reaches the upper bound @xmath246 , then it implies that all the above inequalities are tight , hence @xmath260 ( all broken chain - vertices are separating ) , @xmath261 ( the component containing the root has degree @xmath40 ) , @xmath262 ( all the components of positive degree and the component containing the root are leaves of @xmath167 , and the other components of degree @xmath40 have @xmath71 neighbours in @xmath167 ) , and @xmath263 ( all positive degree components have degree @xmath103 ) .
we take a closer look at the components of degree @xmath40 .
let @xmath168 be such a component not containing the root ; note that @xmath168 has no non - separating dipole nor non - separating chain - vertex according to corollary [ coro : removals ] .
since @xmath168 has adjacency order @xmath71 and is uncolored ( because @xmath260 ) , as already mentioned @xmath250 must have a single component , hence @xmath168 has no chain - vertex and any dipole of @xmath168 must pass by a marked edge of @xmath168 . recall that an unrooted mo - graph of degree @xmath40 is either the cycle - graph , or can be obtained from the `` quadruple - edge '' mo - graph ( two vertices connected by @xmath8 edges ) by successive insertions of melons .
note that , as soon as at least one melon is inserted in the quadruple - edge mo - graph , there are two vertex - disjoint melons .
hence , an mo - graph of degree @xmath40 with at most @xmath71 marked edges and strictly more than @xmath89 vertices has a dipole that avoids the marked edges .
it follows that @xmath168 has to be either the cycle - graph or the quadruple - edge graph .
now , if @xmath168 is the component containing the root , since it has only one adjacent component @xmath264 , it has to be the ( rooted ) cycle - graph .
indeed , the case of a rooted quadruple - edge mo - graph is excluded ; in that case , among the 4 edges of @xmath168 , one carries the root , one is marked , and the two other ones form a dipole that , together with the chain - vertex connecting @xmath168 to @xmath264 , would form a proper chain of @xmath109 , a contradiction .
we now take a closer look at components of positive degrees ( of degree @xmath103 ) .
let @xmath168 be such a component ( which has a unique marked edge since it is at a non - root leaf of @xmath167 ) .
since @xmath168 has degree @xmath103 it can not have non - separating chain - vertices or non - separating unbroken chain - vertices , according to corollary [ coro : removals ] . and by the arguments of claim [ claim : deg_three ] , components of degree @xmath40 of @xmath250
must have adjacency order at least @xmath71 . since the degrees of components in @xmath250 add up to @xmath103
, this easily implies that the unique possibility is @xmath250 having a single node .
hence @xmath168 has no dipole avoiding the marked edge nor chain - vertices , so that @xmath168 must be the cw or the ccw infinity graph , according to corollary [ coro : dipole12 ] . with @xmath265 leaves .
( b ) one of the @xmath266 dominant schemes of degree @xmath89 arising from @xmath267.,width=302 ] [ prop : structure_dominant ] for @xmath244 , the dominant schemes of degree @xmath47 arise from rooted binary trees ( see figure [ fig : dominant ] for an example ) with @xmath268 leaves , @xmath269 inner nodes , and @xmath246 edges , where the root - leaf is occupied by the rooted cycle - graph , the @xmath142 other leaves are occupied by ( cw or ccw ) infinity graphs , the @xmath269 inner nodes are occupied either by the cycle - graph or by the quadruple - edge graph , and the @xmath246 edges are occupied by separating broken chain - vertices .
each rooted binary tree with @xmath268 leaves yields @xmath270 dominant schemes . in the analysis above we have seen that @xmath271 and that any scheme with @xmath272 must be of the stated form . conversely any scheme of the stated form is a valid scheme ( no melon nor proper chain ) of degree @xmath47 with @xmath272 chain - vertices .
hence these schemes are exactly the dominant schemes at degree @xmath47 .
each rooted binary tree with @xmath268 leaves gives rise to @xmath273 dominant schemes .
indeed , at each inner node , one has to decide whether it is occupied by the cycle - graph or by the quadruple - edge graph , and in the second case , one has to choose among the three possible configurations for the free edge ( the unique edge not involved in any connection ) since the straight face passing by the free edge can go either toward the left child , or the right child , or the parent .
finally , one has to decide at each non - root leaf if it is occupied by the cw or the ccw infinity graph . as we have recalled in section [ sec :
regular_subfamily ] , regular colored graphs in dimension @xmath71 naturally form a subfamily of mo - graphs . however , since half - integer degrees are not possible in the model of regular colored graphs , the dominant schemes differ ; as shown in @xcite , in degree @xmath274 , the dominant schemes are associated to rooted binary trees with @xmath275 leaves ( and @xmath276 inner nodes ) , where the root - leaf is occupied by a root - melon , while the @xmath47 non - root leaves are occupied by the unique scheme of degree @xmath102 .
let @xmath12 , and let @xmath277 be the ( finite ) set of reduced schemes of degree @xmath47 . for each @xmath278 ,
let @xmath279 be the generating function of rooted melon - free mo - graphs of reduced scheme @xmath109 , where @xmath97 marks half the number of non - root vertices ( i.e. , for @xmath280 , a weight @xmath281 is given to rooted mo graphs with @xmath44 non - root vertices ) .
let @xmath108 be half the number of non - root standard vertices of @xmath109 , @xmath245 the number of broken chain - vertices , @xmath282 the number of unbroken chain - vertices of type @xmath112 or @xmath125 , @xmath283 the number of even straight chain - vertices , and @xmath284 the number of odd straight chain - vertices .
the generating functions for unbroken chains of type l ( resp .
r ) is clearly @xmath285 , the one for even straight chains is @xmath286 , the one for odd straight chains is @xmath287 , and the one for broken chains is @xmath288 .
therefore @xmath289 denoting by @xmath81 the total number of chain - vertices and by @xmath290 the total number of straight chain - vertices , this simplifies as @xmath291 it is well - known that the generating function of rooted melonic graphs is given by @xmath292 in addition , a rooted melon - free mo - graph with @xmath44 non - root vertices has @xmath293 edges ( recall that the root - edge is split into two edges ) where one can insert a rooted melonic graph .
therefore , defining @xmath294 , the generating function @xmath295 of rooted mo - graphs of reduced scheme @xmath109 is given by @xmath296 and the generating function @xmath297 of rooted mo - graphs of degree @xmath47 is simply given by @xmath298 as discussed in @xcite , @xmath299 has its main singularity at @xmath300 , @xmath301 , and @xmath302 .
therefore , the dominant terms are those for which @xmath245 is maximized . as shown in section [ sec : dominant ] , these schemes are naturally associated to rooted binary trees with @xmath269 inner nodes , as stated in proposition [ prop : structure_dominant ] . for a fixed rooted binary tree @xmath303 with @xmath269 inner nodes , the total contribution of schemes arising from @xmath303 to the generating function of rooted melon - free mo - graphs of degree @xmath47 is @xmath304 indeed , such schemes have @xmath272 , @xmath305 , @xmath306 , each of the @xmath142 non - root leaves is occupied either by the cw or ccw infinity graph , and each of the @xmath269 inner nodes is either occupied by the cycle - graph or the quadruple - edge graph that has @xmath71 possible configurations .
hence the total contribution to the generating function of rooted mo - graphs is @xmath307 this is to be multiplied by the catalan number @xmath308 of rooted binary trees with @xmath269 inner nodes .
note that , by design , the dominant schemes contribute only to graphs with @xmath67 non - root vertices such that @xmath69 , which is consistent here with the prefactor @xmath309 .
graphs with @xmath67 non - root vertices such that @xmath310 have schemes with strictly less than @xmath246 broken chain - vertices , hence have a singular behaviour of type @xmath311 , with @xmath312 .
after elementary calculations and applying transfer theorems of analytic combinatorics @xcite , we obtain : [ prop : asympt ] for @xmath47 and @xmath313 in @xmath25 , let @xmath314 be the number of rooted mo - graphs with @xmath67 vertices and degree @xmath47 .
then , @xmath47 being fixed , for @xmath315 ( and @xmath316 denoting the euler gamma function ) , @xmath317 and @xmath318 . according to proposition [ prop : asympt ] , for @xmath274 @xmath314
is asymptotically @xmath319 , where the multiplicative constant involves @xmath308 . as a comparison ,
as shown in @xcite , for @xmath274 the number @xmath320 of rooted colored graphs ( in dimension @xmath71 ) of degree @xmath47 with @xmath67 vertices is asymptotically @xmath321 , where the multiplicative constant involves @xmath322 .
we also deduce from these estimates that for fixed @xmath274 , the probability that a random rooted mo - graph of degree @xmath47 with @xmath67 vertices ( where @xmath323 ) is a regular colored graph is @xmath324 .
now , going back to remark [ rk : doodles ] , we can also easily obtain the asymptotic enumeration under the planarity constraint , based on the following : each rooted melon - free mo - graph whose reduced scheme is dominant is planar . as shown in section [ sec : dominant ] , a rooted melon - free mo - graph @xmath11 of degree @xmath325 has a total of @xmath142 loops ( one at each non - root leaf of the associated binary tree , see figure [ fig : dominant ] ) .
according to lemma [ lem : remove_loop ] , erasing all these @xmath142 loops yields a ( rooted ) mo - graph @xmath99 of degree @xmath40 , hence a planar mo - graph .
clearly the @xmath142 erased loops can be inserted back without breaking planarity , from which we conclude that @xmath11 is planar .
.,width=302 ] given any dominant scheme such as shown in figure [ fig : dominant ] ( naturally drawn following the shape of a rooted binary tree ) , one can easily figure out how to retract / extend some parts of the drawing so as to obtain a planar redrawing , see figure [ fig : planar_redraw ] .
[ coro : almost_sure ] for each fixed @xmath12 and for @xmath69 , the probability that a rooted mo - graph of degree @xmath47 with @xmath67 vertices is planar tends to @xmath102 as @xmath326 .
it just follows from the observation that the edge - substitution by melonic components clearly preserves planarity , so that all rooted mo - graphs arising from a dominating reduced scheme are planar .
this concludes the proof since these rooted mo - graphs are the ones that dominate the asymptotic expansion . according to remark [ rk : doodles ] , in the planar case , rooted mo - graphs correspond ( bijectively ) to rooted 4-regular maps , the straight faces of the mo - graph identify to the knot - components of the map , and
the number of straight faces is equal to @xmath327 , with @xmath41 the number of vertices and @xmath47 the degree . under this rephrasing , proposition [ prop : asympt ] and corollary [ coro : almost_sure ]
yield the following : for @xmath328 and @xmath329 , let @xmath330 be the number of rooted 4-regular planar maps with @xmath67 vertices and @xmath116 knot - components .
then @xmath331 for @xmath332 .
in addition , for each fixed @xmath12 , and for @xmath69 , @xmath333 has the same asymptotic estimate as @xmath314 , given by .
in this article we have shown that , similarly as in the colored model @xcite , the combinatorial study of mo - graphs can be done by extraction of so - called _ schemes _ , such that there are _ finitely many _ schemes in each fixed degree @xmath12 .
we have also identified the dominant schemes in each degree , whose shapes are naturally associated to rooted binary trees . having determined the dominant schemes we have obtained an explicit asymptotic estimate for the number of ( rooted ) mo - graphs of fixed degree as the number of vertices tends to infinity . as already mentioned in the introduction , a first perspective for future
work is the implementation of the double scaling limit for the mo tensor model . for the sake of completeness ,
let us mention that the double scaling mechanism was implemented , for a different type of colored model , in @xcite , using quantum field theoretical - inspired methods ( the so - called intermediate field method ) . from a probabilistic point a view , a different perspective for future work is to investigate whether the new dominant schemes we have exhibited in this paper correspond or not to phases different from the one of branched polymers ( it was recently proved that the melon graphs correspond , from this point of view , to branched polymers @xcite ) .
another appealing perspective is to extend the mo model ( currently only developed in dimension @xmath70 ) and results of this paper to higher dimensions .
finally , one can address the issue of proving counting theorems for 3d maps using tensor integral techniques , generalizing the fact that counting theorems for maps can be obtained using matrix integral techniques .
the authors acknowledge rzvan guru and gilles schaeffer for very stimulating and instructive discussions .
adrian tanasa is partially supported by the grants anr jcjc combphysmat2tens " and pn 09 37 01 02 .
ric fusy is partially supported by the anr grant `` cartaplus '' 12-js02 - 001 - 01 , and the anr grant `` egos '' 12-js02 - 002 - 01 .
the authors also acknowledge the erwin schrdinger international institute for the valuable work environment provided to them during the combinatorics , geometry and physics " programme .
v. rivasseau , `` the tensor track , iii , '' fortsch .
* 62 * ( 2014 ) 81 .
g. schaeffer and p. zinn - justin , _ on the asymptotic number of plane curves and alternating knots _ , experimental mathematics , 13(4 ) ( 2004 ) 483494 .
adrian tanasa + _ lipn , institut galile , cnrs umr 7030 , universit paris 13 , sorbonne paris cit , _ + _ 99 av .
clement , 93430 villetaneuse , france , eu
_ + _ horia hulubei national institute for physics and nuclear engineering , p.o.b .
mg-6 , 077125 magurele , romania , eu _ | three - dimensional random tensor models are a natural generalization of the celebrated matrix models .
the associated tensor graphs , or 3d maps , can be classified with respect to a particular integer or half - integer , the degree of the respective graph . in this paper
we analyze the general term of the asymptotic expansion in @xmath0 , the size of the tensor , of a particular random tensor model , the multi - orientable tensor model .
we perform their enumeration and we establish which are the dominant configurations of a given degree . |
the concepts of dependence and independence are ubiquitous in many scientific disciplines such as experimental physics , social choice theory , computer science , and cryptography . dependence logic @xmath3@xcite and
its so - called team semantics have given rise to a new logical framework in which various notions of dependence and independence can be formalized and studied .
dependence logic extends first - order logic by dependence atoms @xmath4 expressing that the value of the variable @xmath5 is functionally dependent on the values of @xmath6 .
the formulas of dependence logic are evaluated over _ teams _ ,
i.e. , sets of assignments , and not over single assignments as in first - order logic . in @xcite a modal variant of dependence logic @xmath0was introduced . in the modal framework teams are sets of worlds , and a dependence atom @xmath7 holds in a team @xmath8 if there is a boolean function that determines the value of the propositional variable @xmath9 from those of @xmath10 in all worlds in @xmath8 .
one of the fundamental properties of @xmath0(and of dependence logic ) is that its formulas satisfy the so - called downwards closure property : if @xmath11 , and @xmath12 , then @xmath13 .
still , the modal framework is very different from the first - order one , e.g. , dependence atoms between propositional variables can be eliminated with the help of the classical disjunction @xmath14 @xcite . on the other hand , it was recently shown that eliminating dependence atoms using disjunction causes an exponential blow - up in the formula size , that is , any formula of @xmath15 logically equivalent to the atom in is bound to have length exponential in @xmath16 @xcite . the central complexity theoretic questions regarding @xmath0have been solved in @xcite .
extended modal dependence logic , @xmath17 , was introduced in @xcite .
this extension is defined simply by allowing @xmath18formulas to appear inside dependence atoms , instead of only propositions .
@xmath17can be seen as the first step towards combining dependencies with temporal reasoning .
@xmath17is strictly more expressive than @xmath0but its formulas still have the downwards closure property . in fact , @xmath17has recently been shown to be equivalent to the logic @xmath15 @xcite . in the first - order case ,
several interesting variants of the dependence atoms have been introduced and studied .
the focus has been on independence atoms @xmath19 and inclusion atoms @xmath20 which were introduced in @xcite and @xcite , respectively .
the intuitive meaning of the independence atom is that the variables @xmath21 and @xmath22 are independent of each other for any fixed value of @xmath23 , whereas the inclusion atom declares that all values of the tuple @xmath24 appear also as values of @xmath25 . in @xcite a modal variant , @xmath26 , of independence logic
was introduced .
the logic @xmath26contains @xmath0as a proper sublogic , in particular , its formulas do not in general have the downwards closure property . in @xcite
it was also noted that all @xmath26formulas are invariant under bisimulation when this notion is lifted from single worlds to a relation between sets of words in a natural way . at the same time ( independently ) in @xcite it was shown that @xmath17and @xmath15 can express exactly those properties of kripke structures and teams that are downwards closed and invariant under @xmath27-bisimulation for some @xmath28 . a famous theorem by johan van benthem @xcite states that modal logic is exactly the fragment of first - order logic that is invariant under ( full ) bisimulation .
in this paper we study the analogues of this theorem in the context of team semantics .
our main result shows that an analogue of the van benthem theorem for team semantics can be obtained by replacing @xmath18by _ modal team logic _ ( @xmath1 ) .
@xmath1was introduced in @xcite and extends @xmath18(and @xmath0 ) by classical negation @xmath29 .
more precisely , we show that for any team property @xmath30 the following are equivalent : a. [ enum : main theorem : mtl expressible ] there is an @xmath1-formula which expresses @xmath30 , b. [ enum : main theorem : fo expressible and bisimulation invariant ] there is a first - order formula which expresses @xmath30 and @xmath30 is bisimulation - invariant , c. [ enum : main theorem : bounded bisimulation invariant ] @xmath30 is invariant under @xmath27-bisimulation for some @xmath27 , d. [ enum : main theorem : bisimulation invariant and local ] @xmath30 is bisimulation - invariant and local .
we also study whether all bisimulation invariant properties can be captured by natural variants of @xmath17 .
we consider extended modal independence and extended modal inclusion logic ( @xmath31and @xmath32 , respectively ) , which are obtained from @xmath17by replacing the dependence atom with the independence ( resp .
inclusion ) atom .
we show that both of these logics fail to capture all bisimulation invariant properties , and therefore in particular are strictly weaker than @xmath1 . on the other hand , we show that @xmath33 ( @xmath32extended with classical disjunction ) is in fact as expressive as @xmath34 , but the analogously defined @xmath35 is strictly weaker .
finally , we show that the extension @xmath36 of @xmath18by all first - order definable generalized dependence atoms ( see @xcite ) gives rise to a logic that is as well equivalent to @xmath1 .
a _ kripke model _ is a tuple @xmath37 where @xmath38 is a nonempty set of worlds , @xmath39 , and @xmath40 , where @xmath30 is the set of propositional variables .
a _ team _ of a model @xmath41 as above is simply a set @xmath42 .
the central basic concept underlying vnnen s modal dependence logic and all its variants is that modal formulas are evaluated not in a world but in a team .
this is made precise in the following definitions .
we first recall the usual syntax of modal logic @xmath18 : @xmath43 where @xmath44 is a propositional variable .
note that we consider only formulas in negation normal form , i.e. , negation appears only in front of atoms . as will become clear from the definition of team semantics of @xmath18 , that we present next , @xmath44 and @xmath45 are not dual formulas , consequently _ tertium non datur _ does not hold in the sense that it is possible that @xmath46 and @xmath47 ( however , we still have that @xmath48 for all models @xmath41 and teams @xmath8 ) .
it is worth noting that in @xcite , the connective @xmath49 is allowed to appear freely in @xmath0formulas ( with semantics generalizing the atomic negation case of definition [ def : semanticsml ] below , note that classical negation as allowed in @xmath1is not allowed in @xmath0 ) .
the well - known dualities from classical modal logic are also true for @xmath0formulas hence any @xmath18-formula ( even @xmath0 ) can be rewritten in such a way that @xmath49 only appears in front of propositional variables .
[ def : semanticsml ] let @xmath37 be a kripke model , let @xmath42 be a team , and let @xmath50 be an @xmath18-formula .
we define when @xmath11 holds inductively : if @xmath51 , then @xmath11 iff @xmath52 , if @xmath53 , then @xmath11 iff @xmath54 , if @xmath55 for some formulas @xmath56 and @xmath57 , then @xmath11 iff @xmath58 with @xmath59 and @xmath60 , if @xmath61 for some formulas @xmath56 and @xmath57 , then @xmath11 iff @xmath62 and @xmath63 , if @xmath64 for some formula @xmath56 , then @xmath11 iff there is some team @xmath65 of @xmath41 such that @xmath66 and for each @xmath67 , there is some @xmath68 with @xmath69 , and for each @xmath68 , there is some @xmath67 with @xmath69 . if @xmath70 for some formula @xmath56 , then @xmath11 iff @xmath66 , where @xmath65 is the set @xmath71 .
analogously to the first - order setting , @xmath72-formulas satisfy the following _ flatness _ property @xcite . here , the notation @xmath73 in item [ enum : flatness individual worlds ] refers to the standard semantics of modal logic ( without teams ) .
[ proposition : flatness ] let @xmath41 be a kripke model and @xmath8 a team of @xmath41 .
let @xmath50 be an @xmath72-formula .
then the following are equivalent : 1 .
@xmath11 , 2 .
@xmath74 for each @xmath67 , 3 .
[ enum : flatness individual worlds ] @xmath73 for each @xmath67 . _ modal team logic
_ extends @xmath18by a second type of negation , denoted by @xmath29 , and interpreted just as classical negation .
the syntax is formally given as follows : @xmath75 where @xmath44 is a propositional variable .
the semantics of @xmath1is defined by extending def .
[ def : semanticsml ] by the following clause : * if @xmath76 for some formula @xmath56 , then @xmath11 iff @xmath77 .
we note that usually ( see @xcite ) , @xmath1also contains dependence atoms ; however since these atoms can be expressed in @xmath1we omit them in the syntax ( see proposition [ defatoms ] below ) .
the classical disjunction @xmath14 ( in some other context also referred to as _ intuitionistic disjunction _ ) is also readily expressed in @xmath1 : @xmath78 is logically equivalent to @xmath79 .
for an @xmath18formula @xmath50 , we let @xmath80 denote the formula that is obtained by transforming @xmath81 to negation normal form . now by proposition [ proposition : flatness ] it follows that @xmath82 hence @xmath83 if and only if there is some @xmath67 with @xmath84 . we therefore often write @xmath85 instead of @xmath86 .
note that @xmath87 is not a global operator stating existence of a world anywhere in the model , but @xmath87 is evaluated in the current team .
it is easy to see ( and follows from proposition [ proposition : team bisimulation implies equivalence ] ) that a global `` exists '' operator can not be expressed in @xmath1 .
the next proposition shows that dependence atoms can be easily expressed in @xmath1 .
[ defatoms ] the dependence atom can be expressed in @xmath1by a formula that has length polynomial in @xmath16 .
note first that , analogously to the first - order case @xcite , is logically equivalent with @xmath88 where @xmath89 is the so - called intuitionistic implication with the following semantics : @xmath90 the connective @xmath91 has a short logically equivalent definition in @xmath1(see @xcite ) : @xmath92 is equivalent to @xmath93 , where @xmath94 is the dual of @xmath95 , i.e. , @xmath96 , and @xmath97 is a shorthand for the formula @xmath98 .
finally , @xmath99 can be written as @xmath100 , hence the claim follows .
the intuitionistic implication used in the proof above has been studied in the modal team semantics context in @xcite .
we now introduce the central concept of bisimulation @xcite .
intuitively , two _ pointed models _
( i.e. , pairs of models and worlds from the model ) @xmath101 and @xmath102 are bisimilar , if they are indistinguishable from the point of view of modal logic .
the notion of @xmath27-bisimilarity introduced below corresponds to indistinguishability by formulas with modal depth up to @xmath27 : for a formula @xmath50 in any of the logics considered in this paper , the _ modal depth _ of @xmath50 , denoted with @xmath103 , is the maximal nesting degree of modal operators ( i.e. , @xmath104 and @xmath105 ) in @xmath50 .
let @xmath106 and @xmath107 be kripke models .
we define inductively what it means for states @xmath108 and @xmath109 to be @xmath27-bisimilar , for some @xmath110 , written as @xmath111 .
* @xmath112 holds if for each propositional variable @xmath44 , we have that @xmath113 if and only if @xmath114 .
* @xmath115 holds if the following three conditions are satisfied : 1 .
@xmath112 , 2 . for each successor @xmath116 of @xmath117 in @xmath118
, there is a successor @xmath119 of @xmath120 in @xmath121 such that @xmath122 ( _ forward _ condition ) , 3 . for each successor @xmath119 of @xmath120 in @xmath121
, there is a successor @xmath116 of @xmath117 in @xmath118 such that @xmath122 ( _ backward _ condition ) .
full bisimilarity is defined similarly : pointed models @xmath101 and @xmath102 are _ bisimilar _ if there is a relation @xmath123 such that @xmath124 , and for all @xmath124 , we have that @xmath117 and @xmath120 satisfy the same propositional variables , and for each successor @xmath116 of @xmath117 in @xmath118 , there is a successor @xmath119 of @xmath120 in @xmath121 with @xmath125 ( forward condition ) , and analogously for each successor @xmath119 of @xmath120 in @xmath121 , there is a successor @xmath116 of @xmath117 in @xmath118 with @xmath125 ( back condition ) . in this case
we simply say that @xmath126 and @xmath127 are _ bisimilar_. it is easy to see that bisimilarity implies @xmath27-bisimilarity for each @xmath27 .
let @xmath106 and @xmath107 be kripke models , and let @xmath108 , @xmath109 . then @xmath101 and @xmath102 are @xmath27-equivalent for some @xmath110 , written @xmath128 if for each modal formula @xmath50 with @xmath129 , we have that @xmath130 if and only if @xmath131 .
again , we simply write @xmath132 if the models @xmath118 and @xmath121 are clear from the context . as mentioned above , @xmath27-bisimilarity and @xmath27-equivalence coincide .
the following result is standard ( see , e.g. , @xcite ) : [ prop : bisimulation and equivalence ] let @xmath106 and @xmath107 be kripke models , and let @xmath108 , @xmath109 .
then @xmath111 if and only if @xmath128 . for @xmath1and more generally logics with team semantics ,
the above notion of bisimulation can be lifted to teams .
the following definition is a natural adaptation of @xmath27-bisimilarity to the team setting : let @xmath106 and @xmath107 be kripke models , let @xmath133 and @xmath134 be teams of @xmath118 and @xmath121 . then @xmath135 and @xmath136 are @xmath27-bisimilar , written as @xmath137 if the following holds : * for each @xmath138 , there is some @xmath139 such that @xmath111 , * for each @xmath139 , there is some @xmath138 such that @xmath111 .
full bisimilarity on the team level is defined analogously . in this case
we again say that @xmath135 and @xmath136 are bisimilar , and write @xmath140 , if there is a relation @xmath123 satisfying the forward and backward conditions as above , and which additionally satisfies that for each @xmath138 , there is some @xmath139 with @xmath124 , and for each @xmath139 , there is some @xmath138 with @xmath124 .
this notion of team - bisimilarity was first introduced in @xcite and @xcite . the following result
is easily proved by induction on the formula length : propositionpropositionteambisimulationimpliesequivalence[proposition : team bisimulation implies equivalence ] let @xmath118 and @xmath121 be kripke models , let @xmath133 and @xmath134 be teams of @xmath118 and of @xmath121
. then 1 . if @xmath141 , then for each formula @xmath142 with @xmath129 , we have that @xmath143 if and only if @xmath144 .
2 . if @xmath145 , then for each formula @xmath142 , we have that @xmath143 if and only if @xmath144 .
it suffices to show the first claim , since @xmath145 implies @xmath141 for every @xmath110 .
we show the first claim by induction over the construction of @xmath50 . in this proof
, we omit the models in the notation @xmath146 ( for the entire proof , the model on the left - hand side of @xmath146 is always @xmath118 , and the model on the right - hand side is always @xmath121 ) .
clearly , it suffices to show that if @xmath143 , then @xmath144 .
hence assume that @xmath143 * assume that @xmath51 for a variable @xmath44 ( the case @xmath53 is analogous ) .
since @xmath143 , we know that @xmath113 for each @xmath138 .
let @xmath139 .
since @xmath147 , there is some @xmath138 with @xmath148 . by the above , we know that @xmath113 , and by definition of @xmath146 , it follows that @xmath114 . hence @xmath144 as required . *
assume that @xmath61 .
since @xmath149 , we know that @xmath150 and @xmath151 . by induction , it follows that @xmath152 and @xmath153 .
therefore , @xmath144 as required .
* assume that @xmath55 .
then @xmath154 with @xmath155 and @xmath156 .
+ let @xmath157 and @xmath158 .
it can easily be verified that @xmath159 and @xmath160 and @xmath161 . by induction
, it follows that @xmath162 and @xmath163 , and hence @xmath144 as required . *
assume that @xmath76
. then @xmath164 .
assume indirectly that @xmath152 .
due to induction , it then follows that @xmath150 , a contradiction . *
assume that @xmath70 , let @xmath165 , and @xmath166 . from @xmath143
, we know that @xmath155 , and we need to show that @xmath162 . due to induction , and since @xmath167 , it suffices to show that @xmath168 . +
hence let @xmath169 .
then @xmath116 is the @xmath170-successor of some @xmath138 .
since @xmath171 , there is some @xmath139 with @xmath148 . in particular , there is some @xmath119 such that @xmath119 is an @xmath172-successor of @xmath120 , and @xmath173 .
the other direction is symmetric . *
assume that @xmath64 .
since @xmath143 , there is a team @xmath174 of @xmath118 such that for each @xmath138 , there is some @xmath169 and @xmath175 .
we define the team @xmath176 as follows : + @xmath177 + @xmath178 by definition , @xmath176 only contains @xmath172-successors of worlds in @xmath134 , and for each world @xmath119 in @xmath176 , there is some @xmath169 with @xmath173 . to complete the proof
, it therefore remains to show that : 1 . for each @xmath139
, there is some @xmath172-successor @xmath119 of @xmath120 with @xmath179 , 2 . for each @xmath169
, there is some @xmath180 with @xmath173 .
+ we now prove these claims : 1 .
let @xmath139 .
then there is some @xmath138 with @xmath148 . by choice of @xmath174,there
is some @xmath169 with @xmath175 . since @xmath148
, there is an @xmath172-successor @xmath119 of @xmath176 with @xmath181 . hence by choice of @xmath176 , we know that @xmath180 , and @xmath119 is an @xmath172-successor of @xmath120 .
2 . let @xmath169 .
then @xmath116 is the @xmath170-successor of some @xmath138 .
since @xmath147 , there is some @xmath139 with @xmath148 .
by choice of @xmath174 , there is an @xmath170-successor @xmath116 of @xmath117 with @xmath169 .
since @xmath148 , there is some @xmath172-successor @xmath119 of @xmath120 with @xmath181 .
hence , by choice of @xmath176 , we have @xmath180 as required .
the expressive power of classical modal logic ( i.e. , without team semantics ) can be characterized by bisimulations . in particular , for every pointed model @xmath182 , there is a modal formula of modal depth @xmath27 that exactly characterizes @xmath182 up to @xmath27-bisimulation . in the following ,
we restrict ourselves to a finite set of propositional variables .
in this section , we study the expressive power of @xmath1 . as usual , we measure the expressive power of a logic by the set of properties expressible in it . a _ team property _ is a class of pairs @xmath183 where @xmath41 is a kripke model and @xmath184 a team of @xmath41 . for an @xmath1-formula @xmath50
, we say that @xmath50 _ expresses _ the property @xmath185 .
note that most variants of modal dependence logic have the _ empty team property _ , i.e. , for all @xmath186 and all kripke structures @xmath41 , we have @xmath187 , which obviously does not hold for @xmath34
. however , it immediately follows from the bisimulation invariance of @xmath1that for every @xmath1formula @xmath50 one of the two possibilities hold : * for all kripke structures @xmath41 , @xmath187 . * for all kripke structures @xmath41 , @xmath188 . for this reason
we exclude the empty team in the statement of our results below , but we note that by the remarks above all results cover also the empty team .
let @xmath30 be a team property .
then @xmath30 is _ invariant under @xmath27-bisimulation _ if for each pair of kripke models @xmath118 and @xmath121 and teams @xmath133 and @xmath134 with @xmath141 and @xmath189 , it follows that @xmath190 .
we introduce some ( standard ) notation . in a model @xmath41 ,
the _ distance _ between two worlds @xmath117 and @xmath120 of @xmath41 is the length of a shortest path from @xmath117 to @xmath120 ( the distance is infinite if there is no such path ) . for a world @xmath191 of a model @xmath41 and a natural number @xmath192 ,
the _ @xmath192-neighborhood of @xmath191 in @xmath41 _ , denoted @xmath193 , is the set of all worlds @xmath194 of @xmath41 such that the distance from @xmath191 to @xmath194 is at most @xmath192 .
for a team @xmath8 , with @xmath195 we denote the set @xmath196 .
we often identify @xmath195 and the model obtained from @xmath41 by restriction to the worlds in @xmath195 . a team property @xmath30 is _
@xmath192-local _ for some @xmath197 if for all models @xmath41 and teams @xmath8 , we have @xmath198 we say that @xmath30 is _ local _ , if @xmath30 is @xmath192-local for some @xmath197 .
since our main result establishes a connection between team properties definable in mtl and team properties definable in first - order logic , we also define what it means for a team property to be expressed by a first - order formula . for a finite set of propositional variables @xmath199
, we define @xmath200 as the first - order signature containing a binary relational symbol @xmath201 ( for the edges in our model ) , a unary relational symbol @xmath8 ( for representing a team ) , and , for each variable @xmath202 , a unary relational symbol @xmath203 ( representing the worlds in which @xmath204 is true ) .
kripke models @xmath41 with teams @xmath8 ( where we only consider variables in @xmath199 ) directly correspond to @xmath200 structures : a model @xmath37 and a team @xmath8 uniquely determines the @xmath200-structure @xmath205 with universe @xmath38 and the obvious interpretations of the symbols in @xmath200 .
we therefore say that a first - order formula @xmath50 over the signature @xmath200 expresses a team property @xmath30 , if for all models @xmath41 with a team @xmath8 , we have that @xmath206 if and only if @xmath207 .
we can now state the main result of this paper : [ theorem : mtl expressiveness characterization ] let @xmath30 be a team property .
then the following are equivalent : a. [ enum : main theorem : mtl expressible ] there is an @xmath1-formula which expresses @xmath30 , b. [ enum : main theorem : fo expressible and bisimulation invariant ] there is a first - order formula which expresses @xmath30 and @xmath30 is bisimulation - invariant , c. [ enum : main theorem : bounded bisimulation invariant ] @xmath30 is invariant under @xmath27-bisimulation for some @xmath27 , d. [ enum : main theorem : bisimulation invariant and local ] @xmath30 is bisimulation - invariant and local .
this result characterizes the expressive power of @xmath1 in several ways .
the equivalence of points [ enum : main theorem : mtl expressible ] and [ enum : main theorem : fo expressible and bisimulation invariant ] is a natural analog to the classic van benthem theorem which states that standard modal logic directly corresponds to the bisimulation - invariant fragment of first - order logic .
it is easy to see that characterizations corresponding to items [ enum : main theorem : bounded bisimulation invariant ] and [ enum : main theorem : bisimulation invariant and local ] also hold in the classical setting .
our result therefore shows that @xmath1 plays the same role for team - based modal logics as @xmath18 does for standard modal logic .
the connection between our result and van benthem s theorem @xcite is also worth discussing . essentially , van benthem s theorem is the same result as ours , where `` @xmath1 '' is replaced by `` @xmath18 '' and properties of pointed models ( i.e. , singleton teams ) are considered . in @xmath18 ,
classical negation is of course freely available ; however the property of a team being a singleton is clearly not invariant under bisimulation but the property of a team having only one element _ up to bisimulation _ is .
it therefore follows that each property of singleton teams that is invariant under bisimulation and that can be expressed in @xmath1 can already be expressed in @xmath18 .
the remainder of section [ sect : mtl expressiveness ] is devoted to the proof of theorem [ theorem : mtl expressiveness characterization ] .
the proof relies on various formulas that characterize pointed models , teams of pointed models , or team properties up to @xmath27-bisimulation , for some @xmath110 . in table
[ fig : formulas overview ] , we summarize the notation used in the following and explain the intuitive meaning of these formulas . we start with a natural characterization of the semantics of splitjunction @xmath95 for @xmath18-formulas .
propositionlemmabigsplitjunctioncharacterization[lemma : big splitjunction characterization ] let @xmath208 be a non - empty finite set of @xmath18-formulas , let @xmath41 be a model and @xmath8 a team .
then @xmath209 if and only if for each world @xmath67 , there is a formula @xmath210 with @xmath74 . let @xmath41 be a model and @xmath8 a team
. for @xmath211 , the claim follows from proposition [ proposition : flatness ] .
therefore assume that the lemma holds for @xmath212 , and consider a set @xmath213
. then @xmath214 .
first assume that @xmath215
. then @xmath216 . therefore , @xmath58 with * @xmath217 , * @xmath218 . due to the induction assumption , we know that for every world @xmath219 , there is a formula @xmath220 with @xmath73 . due to proposition [ proposition : flatness ] , we know that each world @xmath221 satisfies the formula @xmath222 , therefore each world @xmath67 satisfies some formula @xmath210 as claimed . for the converse , assume that for each @xmath67 , there is some formula @xmath210 with @xmath73 .
let @xmath223 , and let @xmath224 . by the prerequisites we know that @xmath58 .
due to induction , we have that @xmath217 , and , due to proposition [ proposition : flatness ] , we know that @xmath218 .
it therefore follows that @xmath225 as claimed .
.formulas and sets of formulas used in the proof of theorem [ theorem : mtl expressiveness characterization ] [ cols="<,<,<",options="header " , ] + the following result is standard : [ theorem : hintikka formulas ] * ( * ? ? ?
* theorem 32)*- for each pointed kripke model @xmath182 and each natural number @xmath27 , there is a _ hintikka formula _
( or characteristic formula ) @xmath226 with @xmath227 such that for each pointed model @xmath228 , the following are equivalent : 1 .
@xmath229 , 2 .
@xmath230 .
clearly , we can choose the hintikka formulas such that @xmath231 is uniquely determined by the bisimilarity type of @xmath182 .
this implies that for @xmath27-bisimilar pointed models @xmath101 and @xmath102 , the formulas @xmath232 and @xmath233 are identical .
it it clear that theorem [ theorem : hintikka formulas ] does not hold for an infinite set of propositional symbols , since a finite formula can only specify the values of finitely many variables .
we now define the set of all hintikka formulas that will appear in our later constructions .
informally , @xmath234 is the set of all hintikka formulas characterizing models up to @xmath27-bisimilarity : [ definition : hintikkaformulaslevel ] for @xmath28 , the set @xmath234 is defined as @xmath235 an important observation is that @xmath234 is a finite set : this follows since above , we chose the representatives @xmath231 to be identical for @xmath27-bisimilar pointed models , and since there are only finitely many pointed models up to @xmath27-bisimulation .
since @xmath234 is finite , we can in the following freely use disjunctions over arbitrary subsets of @xmath234 and still obtain a finite formula .
we will make extensive use of this fact in the remainder of section [ sect : mtl expressiveness ] , often without reference .
our next definition is used to characterize a team , again up to @xmath27-bisimulation . since teams are sets of worlds , we use sets of formulas to characterize teams in the natural way , by choosing , for each world in the team , one formula that characterizes it . [ definition : hintikkasetofpointedmodelsinteam ] for a model @xmath41 and a team @xmath8 ,
let @xmath236 since @xmath237 , it follows that @xmath238 is finite as well .
in fact , it is easy to see that @xmath239 is exactly the number of @xmath27-bisimilarity types in @xmath8 , i.e. , the size of a maximal subset of @xmath8 containing only worlds such that the resulting pointed models are pairwise non-@xmath27-bisimilar .
we now combine the formulas from @xmath238 to be able to characterize @xmath41 and @xmath8 ( up to @xmath27-bisimulation ) by a single formula : [ definition : hintikkamodelwithteam ] for a model @xmath41 with a team @xmath184 , let @xmath240 intuitively , the formula @xmath241 expresses that in a model @xmath242 and @xmath65 with @xmath243 , for each world @xmath67 there must be some @xmath68 such that @xmath230 , and conversely , for each @xmath68 , there must be some @xmath67 with @xmath230 , which then implies that @xmath183 and @xmath244 are indeed @xmath27-bisimilar . from the above
, it follows that @xmath241 is a finite @xmath1-formula .
therefore , with the above intuition , it follows that @xmath241 expresses @xmath27-bisimilarity with @xmath183 .
propositionlemmateambisimulationcharacterizationwithhintikkaformulas [ proposition : team bisimulation characterization with hintikka formulas ] let @xmath245 be kripke models with teams nonempty @xmath246 .
then the following are equivalent : * @xmath247 * @xmath248 .
first assume that @xmath247 . to see that @xmath249 , let @xmath250 . by definition
, @xmath50 is an @xmath18-formula with @xmath129 and there is some world @xmath139 such that @xmath131 . due to the bisimulation condition , we know that there is a world @xmath138 with @xmath251 . from proposition [ prop : bisimulation and equivalence ] , it follows that @xmath130 .
since @xmath138 , it follows that @xmath252 .
since this is true for each @xmath250 , it follows that @xmath249 .
it remains to show that @xmath253 .
due to lemma [ lemma : big splitjunction characterization ] , it suffices to show that for each world @xmath138 , there is some formula @xmath250 with @xmath130 .
therefore , let @xmath138 . due to the bisimulation condition , we know that there is some @xmath139 with @xmath111 .
due to the definition of @xmath254 , we know that for the formula @xmath255 we have that @xmath250 and @xmath131 . since @xmath251 , it follows that @xmath130 , and hence for each @xmath138 there is a formula @xmath50 as required . for the converse , assume that @xmath248 .
in particular , it follows that @xmath256 .
due to lemma [ lemma : big splitjunction characterization ] , it follows that for each @xmath138 , there is some formula @xmath250 with @xmath130 . due to the definition of @xmath254 , we know that @xmath257 , and that there is some @xmath139 with @xmath131 . from theorem [ theorem : hintikka formulas ] , it therefore follows that for each @xmath138 , there is some @xmath139 with @xmath251 .
let @xmath120 be a world in @xmath134 , and let @xmath255 .
it then follows that @xmath250 .
since @xmath249 , it follows from the definition of the operator @xmath87 that there is some @xmath138 with @xmath130 . due to the choice of @xmath50 , theorem [ theorem : hintikka formulas ] implies that @xmath251 as required . in this section
, we prove our main result , theorem [ theorem : mtl expressiveness characterization ] . the direction [ enum : main theorem : mtl expressible ] @xmath91 [ enum : main theorem : bounded bisimulation invariant ] follows immediately from proposition [ proposition : team bisimulation implies equivalence ] .
for the converse , assume that @xmath30 is invariant under @xmath27-bisimulation .
without loss of generality assume @xmath259 .
we claim that the formula @xmath260 expresses @xmath30 .
first note that @xmath261 can be written as the disjunction of only finitely many formulas : each @xmath241 is uniquely defined by a subset of the finite set @xmath234 , therefore there are only finitely many formulas of the form @xmath241 .
we now show that for each model @xmath41 and team @xmath8 , we have that @xmath206 if and only if @xmath262 . first assume that @xmath206 .
then the fact that @xmath263 for each model @xmath41 , each world @xmath191 and each number @xmath27 and proposition [ proposition : team bisimulation characterization with hintikka formulas ] imply that @xmath264 .
therefore , @xmath262 . for the converse , assume that @xmath262 .
then there is some @xmath265 with @xmath266 . due to proposition [ proposition : team bisimulation characterization with hintikka formulas ] , it follows that @xmath267 .
since @xmath30 is invariant under @xmath27-bisimulation , it follows that @xmath206 as required .
it suffices to show that @xmath30 can be expressed in first - order logic .
this follows using essentially the standard translation from modal into first - order logic .
since classical disjunction is of course available in first - order logic , the proof of the implication [ enum : main theorem : bounded bisimulation invariant ] @xmath91 [ enum : main theorem : mtl expressible ] shows that it suffices to express each @xmath241 ( expressing team - bisimilarity to @xmath268 ) in first - order logic .
each of the hintikka formulas @xmath231 ( expressing bisimilarity to the pointed model @xmath269 ) is a standard modal formula , therefore an application of the standard translation gives a first - order formula @xmath270 with a free variable @xmath204 such that for all models @xmath242 and worlds @xmath194 , we have that @xmath229 if and only if @xmath271 .
we now show how to express @xmath241 ( expressing team - bisimilarity to @xmath268 ) in first - order logic .
recall that @xmath241 is defined as @xmath272 .
therefore , a first - order representation of @xmath241 is given as @xmath273 where @xmath274 is the standard translation of @xmath50 into first - order logic as mentioned above .
this concludes the proof .
let @xmath50 be the first - order formula expressing @xmath30 . since @xmath50 is first - order , we know that @xmath50 is hanf - local for a brief discussion of hanf - locality ] .
let @xmath192 be the hanf - locality rank of @xmath50 .
we show that @xmath50 is @xmath275-local
. therefore , let @xmath41 be a model with team @xmath8 .
we show that @xmath207 if and only if @xmath276 .
since @xmath50 is bisimulation - invariant , it suffices to construct models @xmath118 and @xmath121 containing @xmath8 such that * @xmath277 and @xmath183 are team - bisimilar , * @xmath278 and @xmath279 are team - bisimilar , * @xmath280 if and only if @xmath281 .
we first define @xmath282 as the model obtained from @xmath41 by disconnecting @xmath283 from the remainder of the model , i.e. , by removing all edges between @xmath283 and @xmath284 . since @xmath282 is also obtained from @xmath283 by adding the remainder of the model @xmath41 without connecting the added worlds to @xmath283 , it is obvious that @xmath285 .
we now define the models @xmath118 and @xmath121 such that @xmath286 and @xmath287 ( and hence @xmath288 ) as follows : * @xmath118 and @xmath121 are obtained from @xmath41 and @xmath282 by adding the exact same components : for each @xmath289 ( note that @xmath41 and @xmath282 have the exact same set of worlds ) , countably infinitely many copies of @xmath290 and of @xmath291 are added to both @xmath118 and @xmath121 . * for @xmath292 , and @xmath293 , with @xmath294 , we denote the @xmath16-th copy of @xmath291 in @xmath295 , the _ center _ of @xmath294 is the copy of @xmath191 in @xmath294 . * for @xmath292 , and @xmath293 , with @xmath296 , we denote the @xmath16-th copy of @xmath290 in @xmath295 , the _ center _ of @xmath296 is the copy of @xmath191 in @xmath296 . in the above ,
when we `` copy '' a part of a ( kripke ) model , this includes copying the values of the involved propositional variables in these worlds ( this is reflected in the resulting first - order models in the obvious way ) .
however , we stress that the team @xmath8 is treated differently : the set @xmath8 is not enlarged with the copy operation , i.e. , a copy of a world in @xmath8 is itself not an element of @xmath8 . since @xmath118 and @xmath121
are obtained from @xmath41 and @xmath282 by adding new components that are not connected to the original models , it clearly follows that @xmath183 and @xmath277 are team - bisimilar , and @xmath297 and @xmath278 are team - bisimilar .
note that each @xmath191 in the @xmath41-part of @xmath118 is the center of a @xmath275-environment isomorphic to @xmath298 , and each @xmath191 in the @xmath282-part of @xmath121 is the center of a @xmath275-environment isomorphis to @xmath299 . since the models @xmath41 ( @xmath282 ) contain one copy of each @xmath290 ( @xmath291 ) , both @xmath118 and @xmath121 contain countably infinitely many copies of each @xmath290 and each @xmath291 .
let @xmath300 be the subset of @xmath118 containing only the points from the @xmath41-part of @xmath118 , plus the center of each @xmath301 , and the center of each @xmath299 .
similarly , let @xmath302 be the subset of @xmath121 containing only the points from the @xmath282-part of @xmath121 plus the centers of the added components .
since @xmath118 and @xmath121 contain the same number of copies of each relevant neighborhood , there is a bijection @xmath303 such that for each @xmath304 , the @xmath275-neighborhoods of @xmath191 and @xmath305 are isomorphic .
now @xmath306 can be modified such that for each @xmath289 which has distance at most @xmath192 to a world in @xmath8 , the value @xmath305 is the corresponding world in the @xmath282-part of @xmath121 .
the thus - modified @xmath306 now satisfies that for each @xmath304 , the @xmath192-neighborhoods of @xmath191 and @xmath305 are isomorphic .
we can easily extend @xmath306 to worlds in @xmath307 and @xmath308 that are not the center of their respective components by mapping such a world @xmath191 in @xmath307 to the copy of @xmath191 in @xmath309 , and analogously for @xmath308 .
therefore , we have constructed a bijection @xmath310 such that for each @xmath311 , the @xmath192-neighborhood of @xmath191 in @xmath118 is isomorphic to the @xmath192-neighborhood of @xmath191 in @xmath121 .
since @xmath50 is hanf - local with rank @xmath192 , this implies that @xmath280 if and only if @xmath281 , as required .
the proof of this implication uses ideas from otto s proof of van benthem s classical theorem presented in @xcite .
however our proof is based on the hanf - locality of first - order expressible properties , whereas otto s proof uses ehrenfeucht - frass games , as a consequence , our construction requires an infinite number of copies of each model due to cardinality reasons .
assume that @xmath30 is invariant under bisimulation , and @xmath30 is @xmath27-local for some @xmath110 .
we show that @xmath30 is invariant under @xmath27-bisimulation .
hence let @xmath137 .
since @xmath30 is invariant under bisimulation , we can without loss of generality assume that @xmath118 and @xmath121 are directed forests , that @xmath118 contains only worlds connected to worlds in @xmath133 , and analogously for @xmath121 and @xmath134 .
since @xmath30 is also @xmath27-local , we can also assume that @xmath118 contains no world with a distance of more than @xmath27 to @xmath133 , and analogously for @xmath121 and @xmath134 .
from these assumptions , it immediately follows that @xmath140 , and , since @xmath30 is invariant under bisimulation , this implies that @xmath189 if and only if @xmath136 in @xmath30 , as required .
research on variants of ( modal ) dependence logic has concentrated on logics defined in terms of independence and inclusion atoms .
analogously to @xmath0 , these logics are invariant under bisimulation but are strictly less expressive than @xmath1@xcite . on the other hand , extended modal dependence logic , @xmath17 , uses dependence atoms but allows them to be applied to @xmath18-formulas instead of just proposition symbols @xcite .
this variant is also known to be a proper sub - logic of @xmath1being able to express all downwards - closed properties that are invariant under @xmath27-bisimulation for some @xmath28 , and equivalent to @xmath15 @xcite . in this section
we systematically study the expressive power of variants of @xmath17replacing dependence atoms by independence and inclusion atoms . depending on whether we also allow classical disjunction or not , this gives four logics , namely @xmath312 ( extended modal independence logic ) , @xmath35 ( @xmath31extended with classical disjunction ) , @xmath313 ( extended modal inclusion logic ) and @xmath314 ( @xmath32extended with classical disjunction ) .
we study the expressiveness of these logics , and show that while @xmath314 is as expressive as @xmath1 , for each of the other three logics there is an @xmath1-expressible property that can not be expressed in the logic . in the last section , we also study the extension of @xmath18by first - order definable generalised dependence atoms , and show that the resulting logic even without the addition of classical disjunction is equivalent to @xmath1 .
we first consider _ extended modal independence logic _
( @xmath31 ) .
syntactically , @xmath31extends @xmath18by the following : if @xmath315 , @xmath316 , and @xmath317 are finite sets of @xmath18-formulas , then @xmath318 is an @xmath31-formula .
the semantics of this _ extended independence atom _ are defined by lifting the definition for propositional variables given in @xcite to @xmath18-formulas as follows . for a formula @xmath50 and a world @xmath191
, we write @xmath319 for the function defined as @xmath320 if @xmath74 , and @xmath321 otherwise ( the model @xmath41 will always be clear from the context ) . for a set of formulas @xmath322 and
worlds @xmath323 , we write @xmath324 if @xmath325 for each @xmath326 .
@xmath327 the extension of @xmath31by classical disjunction @xmath14 is denoted by @xmath35 .
we will next show that @xmath35 is a proper sub - logic of @xmath1 .
the following lemma will be used in the proof .
[ emptyr ] let @xmath37 be a kripke model such that @xmath328 and @xmath42 a team . then for all @xmath329 it holds that if @xmath330 , then @xmath331 for all @xmath67 . a straight - forward induction on the construction of @xmath50 using the facts that a singleton team trivially satisfies all independence atoms , and the empty team satisfies all formulas of @xmath35 .
theoremtheoremnotexpressibleinemil[theorem : not expressible in emil ] @xmath332 .
the first inclusion follows from the fact that dependence atoms can be expressed by independence atoms .
the inclusion is strict since @xmath17is downwards - closed and @xmath31is not . for the last inclusion ,
note that every property expressible in @xmath35 is invariant under bisimulation , hence it follows that @xmath1can express every @xmath35-expressible property due to theorem [ theorem : mtl expressiveness characterization ] .
for the strictness , we show that there is a team property that is invariant under @xmath333-bisimulation and which can not be expressed in @xmath35 .
let @xmath30 be the property @xmath334 i.e. , the class of @xmath183 such that @xmath8 contains at least one world in which @xmath44 is satisfied .
consider the model @xmath41 with worlds @xmath117 and @xmath120 , where @xmath44 is true in @xmath117 and false in @xmath120 , and the accessability relation @xmath328 .
let @xmath335 , and let @xmath336 . obviously , @xmath337 and @xmath338 .
by lemma [ emptyr ] for all @xmath35-formulas @xmath50 , if @xmath339 , then @xmath340 .
this shows that there is no @xmath35-formula expressing @xmath30 .
analogously to @xmath31 , we now define _ extended modal inclusion logic _ , @xmath32 .
@xmath32extends the syntax of @xmath18with the following rule : if @xmath341 and @xmath342 are @xmath18-formulas , then @xmath343 is an @xmath32-formula .
the semantics of this _ inclusion atom _ are lifted from the first - order setting @xcite to the extended modal case : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ @xmath344 if for every world @xmath67 there is a world @xmath345 such that @xmath346 for each @xmath347 .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the extension of @xmath32by classical disjunction @xmath14 is denoted by @xmath314 .
analogously to first - order inclusion logic @xcite , the truth of @xmath32-formulas is preserved under unions of teams .
hence we get the following result .
theorememincllessthanmtl @xmath32is strictly less expressive than @xmath1 .
we show that there is a @xmath333-bisimulation invariant property that can not be expressed with @xmath32 .
for this , let @xmath30 be the property @xmath348 clearly ( and also due to theorem [ theorem : mtl expressiveness characterization ] ) , @xmath30 is invariant under @xmath333-bisimulation .
now , let @xmath41 be a model with worlds @xmath117 and @xmath120 such that in @xmath349 , the variable @xmath350 is true and @xmath351 is false .
let @xmath352 , and @xmath336 .
then , by construction , @xmath337 and @xmath353 , but @xmath354 .
now assume that @xmath50 is an @xmath32-formula that expresses @xmath30 .
then , in particular @xmath339 , @xmath340 and @xmath355 .
however , it easily follows that @xmath32is _ union_-closed , i.e. if @xmath339 and @xmath340 , then also @xmath356 ( see , e.g. , @xcite , the property trivially transfers to the modal setting ) .
therefore , we have a contradiction
. next we want to show that @xmath357is as powerful as @xmath1 .
[ theorem : eminclc expressiveness ] let @xmath30 be a team property
. then the following are equivalent : 1 .
[ enum : theorem : eminclc expressiveness : bisimulation invariance ] @xmath30 is invariant under @xmath27-bisimulation .
[ enum : theorem : eminclc expressiveness : eminclc expressiveness ] there is an @xmath357-formula @xmath50 with @xmath358 that characterizes @xmath30 . the direction from [ enum : theorem : eminclc expressiveness : eminclc expressiveness ] to [ enum : theorem : eminclc expressiveness : bisimulation invariance ] follows by a straight - forward extension of the proof of proposition [ proposition : team bisimulation implies equivalence ] .
for the converse , assume that @xmath30 is invariant under @xmath27-bisimulation .
from the proof of theorem [ theorem : mtl expressiveness characterization ] , we know that it suffices to construct an @xmath357-formula @xmath50 that is equivalent to the @xmath1-formula @xmath359 . since the @xmath14-operator is available in @xmath314 , it suffices to show how to express the formula @xmath241 for each model @xmath41 and team @xmath8 as an @xmath357-formula .
recall that @xmath240 the second conjunct already is an @xmath357-formula , hence it suffices to show how @xmath360 can be expressed for an @xmath18-formula @xmath50 .
as discussed earlier , @xmath361 for an @xmath18-formula @xmath50 if and only if there is a world @xmath67 with @xmath74 .
hence from the semantics of the inclusion atom , it is clear that @xmath360 is equivalent to @xmath362 .
this concludes the proof . in this section
we show that @xmath1 , and the bisimulation invariant properties , can be captured as the extension of @xmath18by all generalized dependence atoms definable in first - order logic without identity .
the notion of a generalized dependence atom in the modal context was introduced in @xcite . a closely related notion
was introduced and studied in the first - order context in @xcite .
the semantics of a generalized dependence atom @xmath363 is determined essentially by a property of teams . in the following we are interested in generalized dependence atoms definable by first - order formulae , defined as
follows : suppose that @xmath363 is an atom of width @xmath16 , that is , an atom that applies to @xmath16 propositional variables ( for example the atom in ) .
we say that @xmath363 is @xmath364-definable if there exists a @xmath364-sentence @xmath365 over signature @xmath366 such that for all kripke models @xmath367 and teams @xmath8 , @xmath368 where @xmath369 is the first - order structure with universe @xmath8 and relations @xmath370 for @xmath371 , where for all @xmath67 , @xmath372 .
in our `` extended '' setting the arguments to a generalized dependence atom @xmath373 can be arbitrary @xmath18-formulas instead of propositional variables .
hence the relation @xmath374 is now interpreted by the worlds of @xmath8 in which @xmath375 is satisfied .
we denote by @xmath376the extension of @xmath18by all generalized dependence atoms @xmath363 that are @xmath364-definable without identity .
[ thm : mtl = ml^fo ] @xmath376is equally expressive as @xmath1 . in the proof of theorem 6.8 in
@xcite it is showed that @xmath376is invariant under bisimulation in the case where generalised atoms may be applied only to propositional variables .
the proof easily extends to the setting where arbitrary @xmath18-formulas may appear as arguments to a generalised dependence atom .
therefore , @xmath376is not more expressive than @xmath1 . for the converse ,
let @xmath30 be a property that can be expressed in @xmath1 . from theorem [ theorem : mtl expressiveness characterization ]
, it follows that @xmath30 is invariant under @xmath27-bisimulation , and from the proof of theorem [ theorem : mtl expressiveness characterization ] , we know that it suffices to express the formula @xmath359 in @xmath376 .
we can do this with the following first - order definable atom ( by suitably choosing the parameters @xmath377 ) : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ @xmath378 if and only if there is some @xmath379 such that each @xmath67 satisfies some @xmath380 , and for each @xmath381 , there is some @xmath67 that satisfies @xmath382 . _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the atom @xmath363 can now be fo - defined by replacing the exists / for all quantifiers on the indices with disjunctions / conjunctions : @xmath383 then , the atom @xmath363 applied to the formulas in @xmath241 for all @xmath206 gives a formula expressing @xmath30 .
our results show that , with respect to expressive power , modal team logic is a natural upper bound for all the logics studied so far in the area of modal team semantics .
overall , an interesting picture of the characterization of the expressiveness of modal logics in terms of bisimulation emerges : let us say that `` invariant under bounded bisimulation '' means invariant under @xmath27-bisimulation for some finite @xmath27 .
then we have the following hierarchy of logics : * due to van benthem s theorem @xcite , @xmath18can exactly express all properties of pointed models that are fo - definable and invariant under bisimulation . * due to @xcite , @xmath18with team semantics and extended with classical disjunction @xmath14 can exactly express all properties of teams that are invariant under bounded bisimulation and additionally downwards - closed . * our result shows that @xmath18with team semantics and extended with classical negation @xmath29 can exactly express all properties of teams that are invariant under bounded bisimulation .
a number of open questions in the realm of modal logics with team semantics remain : 1 . in the proof of theorem [ thm : mtl = ml^fo ] , for each @xmath27
, there is only a finite width of the @xmath363-operator above needed to express all properties that are invariant under @xmath27-bisimulation .
however , the theorem leaves open the question whether there is a `` natural '' atom @xmath363 or an atom with `` restricted width '' that gives the entire power of @xmath1 .
can we axiomatize @xmath1 ?
axiomatizability of sublogics of @xmath1has been studied , e.g. , in @xcite and @xcite .
while we mentioned a number of complexity results on modal dependence logic and some of its extensions , this issue remains unsettled for full @xmath1 .
in particular , what is the complexity of satisfiability and validity of @xmath1 ? 10 samson abramsky and jouko a. vnnen . from if to bi . , 167(2):207230 , 2009 .
patrick blackburn , maarten de rijke , and yde venema .
, volume 53 of _ cambridge tracts in theoretical computer scie_. cambridge university press , cambridge , 2001 .
j. ebbing and p. lohmann .
complexity of model checking for modal dependence logic . in _
theory and practice of computer science ( sofsem ) _ , number 7147 in lecture notes in computer science , pages 226237 , berlin heidelberg , 2012 .
springer verlag .
johannes ebbing , lauri hella , arne meier , julian - steffen mller , jonni virtema , and heribert vollmer .
extended modal dependence logic . in leonid libkin , ulrich kohlenbach , and
ruy j. g. b. de queiroz , editors , _ wollic _ , volume 8071 of _ lecture notes in computer science _ , pages 126137 .
springer , 2013 .
pietro galliani . inclusion and
exclusion dependencies in team semantics - on some logics of imperfect information .
, 163(1):6884 , 2012 .
pietro galliani and lauri hella . inclusion logic and fixed point logic . in simona ronchi della rocca , editor , _ csl _ , volume 23 of _ lipics _ ,
pages 281295 .
schloss dagstuhl - leibniz - zentrum fuer informatik , 2013 .
goranko and martin otto . ,
chapter model theory of modal logic , pages 255325 .
elsevier , 2006 .
rajeev gor , barteld p. kooi , and agi kurucz , editors . .
college publications , 2014 .
erich grdel and jouko vnnen . dependence and independence .
, 101(2):399410 , 2013 . william hanf . .
in _ symposium in the theory of models _ , pages 132145 , 1965 .
lauri hella , kerkko luosto , katsuhiko sano , and jonni virtema . the expressive power of modal dependence logic . in gor
@xcite , pages 294312 .
juha kontinen , julian - steffen mller , henning schnoor , and heribert vollmer .
modal independence logic . in gor
@xcite , pages 353372 .
antti kuusisto .
a double team semantics for generalized quantifiers .
, 24(2):149191 , 2015 .
leonid libkin . .
springer , 2004 .
peter lohmann and heribert vollmer .
complexity results for modal dependence logic . , 101(2):343366 , 2013 .
julian - steffen mller .
complexity of model checking in modal team logic .
, 2012 .
julian - steffen mller . .
phd thesis , leibniz universitt hannover , 2014 .
m. otto .
elementary proof of the van benthem - rosen characterisation theorem .
technical report 2342 , darmstadt university of technology , 2004 .
david michael ritchie park .
concurrency and automata on infinite sequences . in peter
deussen , editor , _ theoretical computer science , 5th gi - conference , karlsruhe , germany , march 23 - 25 , 1981 , proceedings _ ,
volume 104 of _ lecture notes in computer science _ ,
pages 167183 .
springer , 1981 .
katsuhiko sano and jonni virtema . axiomatizing propositional dependence logics .
, abs/1410.5038 , 2014 .
merlijn sevenster .
model - theoretic and computational properties of modal dependence logic . , 19(6):11571173 , 2009 .
jouko vnnen .
. cambridge university press , 2007 .
jouko vnnen .
modal dependence logic .
, 5:237254 , 2009 .
johan van benthem . .
phd thesis , universiteit van amsterdam , instituut voor logica en grondslagenonderzoek van de exacte wetenschappen , 1977 .
johan van benthem . .
bibliopolis , 1985 .
fan yang . .
phd thesis , university of helsinki , 2014 .
in the proof of the implication proof of implication [ enum : main theorem : fo expressible and bisimulation invariant ] @xmath91 [ enum : main theorem : bisimulation invariant and local ] of theorem [ theorem : mtl expressiveness characterization ] , we used hanf - locality of first - order formualas . in the following ,
we briefly introduce the relevant definitions and state the main result .
our presentation is based on ( * ? ? ?
* section 4 ) . since in this paper , we only consider properties of kripke models , our first - oder models are defined over a relational signature that includes ( at most ) a single binary relation @xmath384 representing the kripke relation of our models , a unary relation @xmath8 representing a team in a kripke model , and for each propositional variable , a unary relation representing the worlds in which the variable is true . in the following , we assume a fixed finite set of variables . for a world @xmath289 , and a natural number @xmath385 , the _ radius @xmath385 ball _ around @xmath191 is the set of worlds @xmath194 such that there is an ( undirected ) path in @xmath41 between @xmath191 and @xmath194 with length at most @xmath385 .
the @xmath385-neighborhood of @xmath191 in @xmath41 , denoted with @xmath386 , is the model @xmath41 restricted to the radius @xmath385 ball around @xmath191 , with an additional constant interpreted as @xmath191 .
a property @xmath30 of kripke models is _ hanf - local _ if there is some @xmath391 such that for every kripke models @xmath118 and @xmath121 with @xmath392 , we have that @xmath118 has property @xmath30 if and only if @xmath121 has property @xmath30 . the smallest number @xmath192 for which this is true is the hanf - locality rank of @xmath30 . | the famous van benthem theorem states that modal logic corresponds exactly to the fragment of first - order logic that is invariant under bisimulation . in this article
we prove an exact analogue of this theorem in the framework of modal dependence logic @xmath0and team semantics .
we show that modal team logic @xmath1 , extending @xmath0by classical negation , captures exactly the @xmath2-definable bisimulation invariant properties of kripke structures and teams .
we also compare the expressive power of @xmath1to most of the variants and extensions of @xmath0recently studied in the area . |
a key unsolved problem in general relativity is to compute the gravitational radiation produced by a small object spiralling into a much larger black hole .
this problem is of direct observational relevance .
inspirals of compact objects into intermediate mass black holes ( @xmath0 may be observed by ligo and other ground based interferometers @xcite ; recent observations suggest the existence of black holes in this mass range @xcite .
in addition , a key source for the space - based gravitational wave detector lisa is the final epoch of inspiral of a stellar - mass compact object into a massive ( @xmath1 ) black hole at the center of a galaxy .
@xcite have estimated that lisa should see over a thousand such inspiral events during its multi - year mission lifetime , based on monte carlo simulations of the dynamics of stellar cusps by freitag @xcite .
observations of these signals will have several major scientific payoffs @xcite : * from the observed waveform , one can measure the mass and spin of the central black hole with fractional accuracies of order @xmath2 @xcite .
the spin can provide useful information about the growth history ( mergers versus accretion ) of the black hole @xcite .
* likewise , one obtains a census of the inspiralling objects masses with precision @xmath3 , teaching us about the stellar mass function and mass segregation in the central parsec of galactic nuclei .
* the measured event rate will teach us about dynamics in the central parsec of galaxies . *
the gravitational waves will be wonderful tools for probing the nature of black holes in galactic nuclei , allowing us to observationally map , for the first time , the spacetime geometry of a black hole , and providing a high precision test of general relativity in the strong field regime @xcite . to be in the lisa waveband , the mass @xmath4 of the black hole must be in the range @xmath5 @xcite .
the ratio between the mass @xmath6 of the compact object and @xmath4 is therefore in the range @xmath7 .
these systems spend the last year or so of their lives in the very relativistic regime close to the black hole horizon , where the post - newtonian approximation has completely broken down , and emit @xmath8 cycles of waveform @xcite . realizing
the above science goals will require accurate theoretical models ( templates ) of the gravitational waves .
this is because the method of matched filtering will be used both to detect the signals buried in the detector noise , and to measure the parameters characterizing detected signals .
the accuracy requirement is roughly that the template should gain or lose no more than @xmath9 cycle of phase compared to the true waveform over the @xmath10 cycles of inspiral .
these sources must therefore be modeled with a fractional accuracy @xmath11 .
the past several years have seen a significant research effort in the relativity community aimed at providing these accurate templates . to date
, there have been several approaches to generating waveforms .
the foundation for all the approaches is the fact that , since @xmath12 , the field of the compact object can be treated as a linear perturbation to the large black hole s gravitational field . on short timescales ,
the compact object moves on a geodesic of the kerr geometry , characterized by its conserved energy @xmath13 , @xmath14-component of angular momentum @xmath15 , and carter constant @xmath16 . over longer timescales ,
radiation reaction causes the parameters @xmath13 , @xmath15 and @xmath16 to evolve and the orbit to shrink .
the various approaches are : 1 . _
use of post - newtonian methods : _ fairly crude waveforms can be obtained using post - newtonian methods @xcite .
these have been used to approximately scope out lisa s ability to detect inspiral events @xcite and to measure the waveform s parameters @xcite . however , since the orbital speeds are a substantial fraction of the speed of light , these waveforms are insufficiently accurate for the eventual detection and data analysis of real signals .
use of conservation laws : _ in this approach @xcite one uses the teukolsky - sasaki - nakamura ( tsn ) formalism @xcite to compute the fluxes of energy @xmath13 and angular momentum @xmath15 to infinity and down the black hole horizon generated by a a compact object on a geodesic orbit . imposing global conservation of energy and angular momentum ,
one infers the rates of change of the orbital energy and angular momentum . for certain special classes of orbits ( circular and equatorial orbits )
this provides enough information that one can also infer the rate of change of the carter constant @xmath16 , and thus the inspiralling trajectory .
direct computation of the self - force : _ in this more fundamental approach one computes the self - force or radiation - reaction force arising from the interaction of the compact object with its own gravitational field .
a formal expression for this force in a general vacuum spacetime in terms of the retarded green s function was computed several years ago @xcite .
translating this expression into a practical computational scheme for kerr black holes is very difficult and is still in progress .
roughly 100 papers devoted to this problem have appeared in the last few years ; see , for example poisson @xcite for an overview and references .
time - domain numerical simulations : _ another technique is to numerically integrate the teukolsky equation as a 2 + 1 pde in the time domain @xcite , and to model the compact object as a finite - sized source .
this approach faces considerable challenges : ( i ) there is a separation of timescales the orbital period is much shorter than the radiation reaction timescale .
( ii ) there is a separation of lengthscales the compact object is much smaller than the black hole .
( iii ) the self - field of the small object must be computed with extremely high accuracy , as the piece of the self - field responsible for the self - force is a tiny fraction ( @xmath17 ) of the divergent self - field .
this approach may eventually be competitive with ( ii ) and ( iii ) but is currently somewhat far from being competitive [ flux accuracies are @xmath18 as compared to @xmath19 for ( ii ) ] .
a key feature of these systems is that they evolve adiabatically : the radiation reaction timescale @xmath20 is much longer than the orbital timescale @xmath21 , by a factor of the inverse @xmath22 of the mass ratio .
this has implications for the nature of the signal .
the self - acceleration @xmath23 of the compact object can be expanded in powers of the mass ratio as @xmath24\;. \label{eq : selfaccel}\ ] ] here @xmath25 and @xmath26 are the dissipative and conservative pieces of the leading - order self - acceleration computed in refs .
similarly , @xmath27 and @xmath28 are the corresponding pieces of the first correction to the self - acceleration , which has not yet been computed ( although see ref.@xcite for work in this direction )
. the effect of the dissipative pieces of the self force will accumulate secularly , while the effect of the conservative pieces will not .
hence the effect of the dissipative pieces on the phase of the orbit will be larger than that of the conservative pieces by a factor of the number of cycles of inspiral , @xmath29 .
consider now , for example , the azimuthal phase @xmath30 of the orbit .
this can be expanded in powers of the mass ratio using a two - timescale expansion as @xcite @xmath31\ ; , \label{eq : emri_phase}\ ] ] where the leading - order phase @xmath32 is determined by @xmath25 , and the higher order correction @xmath33 to the phase is determined by @xmath26 and @xmath27 . here @xmath34 is a `` slow '' time variable which satisfies @xmath35 .
we shall call leading - order waveforms containing only terms corresponding to @xmath32 _
adiabatic waveforms_. from eq.([eq : emri_phase ] ) these waveforms are accurate to @xmath9 cycle over the inspiral ( more precisely , the phase error is independent of @xmath36 in the limit @xmath37 ) . to compute adiabatic waveforms
we need only keep the dissipative piece of the self - force , and we can discard any conservative pieces . the conservation - law method discussed above yields waveforms that are accurate to this leading order , but is limited in that it can not be used for generic orbits .
adiabatic waveforms will likely be sufficiently accurate for detecting inspiral events with lisa @xcite .
the correction @xmath38 to the phase of the waveform can be estimated using the post - newtonian approximation , and amounts to @xmath39 cycle over the inspiral for typical parameter values . while the post - newtonian approximation is not strictly valid in the highly relativistic regime near the horizon of interest here , it suffices to give some indication of the accuracy of adiabatic waveforms .
this computation is described in [ sec : accuracy ] .
recently yasushi mino derived a key result that paves the way for computations of adiabatic waveforms for generic inspirals in kerr @xcite .
mino showed that the time average of the self - force formula of @xcite for bound orbits in kerr yields the same result as the gradient of one half the difference between the retarded and advanced metric perturbations .
this `` half retarded minus half advanced '' prescription is the standard result for electromagnetic radiation reaction in flat spacetime due to dirac @xcite .
this prescription was also posited , without proof , for scalar , electromagnetic and gravitational radiation reaction in the kerr spacetime by galtsov @xcite .
however , prior to mino s analysis , the prescription was not generally thought to be applicable or relevant in kerr
. as one might expect , knowledge of the infinite - time - averaged self - force is sufficient to compute the leading - order , adiabatic waveforms @xcite , although there are subtleties related to the choice of gauge @xcite .
the sufficiency can also be established using a two - timescale expansion @xcite . in this paper
we apply mino s result to compute an explicit expression for the time - averaged time derivative of the carter constant .
we specialize to the case of a particle endowed with a scalar charge , coupled to a scalar field .
this computation is useful as a warm up exercise for the more complicated case of a particle emitting gravitational waves @xcite .
we start in sec .
[ sec : scalar ] by describing the model of a point particle coupled to a scalar field , and review how the self - force causes both an acceleration of the particle s motion and also an evolution of the renormalized rest mass of the particle @xcite . in sec .
[ sec : geosesics ] we review the properties of generic bound geodesic orbits in the kerr spacetime
. a crucial result we use later is that the @xmath40 and @xmath41 motions are periodic when expressed as functions of a particular time parameter we call mino time @xcite , and that the @xmath42 and @xmath43 motions consist of linearly growing terms plus terms that are periodic with the period of the @xmath40-motion , plus terms that are periodic with the period of the @xmath41-motion @xcite . section [ sec : minoproof ] reviews mino s derivation of the half - retarded - minus - half - advanced prescription for radiation reaction in the adiabatic limit @xcite . in sec .
[ sec : modes ] we discuss a convenient basis of modes for solutions of the scalar wave equation in kerr , namely the `` in '' , `` out '' , `` up '' and `` down '' modes used by chrzanowski @xcite and galtsov @xcite .
we also review several key properties of these modes including symmetry relations and relations between the various reflection and transmission coefficients that appear in the definitions of the modes .
section [ sec : retarded ] reviews the standard derivation of the mode expansion of the retarded green s function , and sec .
[ sec : radiative ] reviews the derivation of the mode expansion of the radiative green s function given by galtsov @xcite , correcting several typos in galtsov which are detailed in [ sec : typos ] .
next , in sec .
[ sec : harmonic ] we turn to the source term in the wave equation for the scalar field .
we review the derivation of drasco and hughes of the harmonic decomposition of this source term in terms of a discrete sum over frequencies in which harmonics of three different fundamental frequencies occur @xcite .
we derive expressions for the mode coefficients in the expansion of the retarded field near future null infinity and near the future event horizon .
these coefficients are expressed as integrals over a torus in phase space , which is filled ergodically by the geodesic motion .
section [ sec : edot ] combines the results of the preceding sections to derive expressions for the time - averaged rates of change of two of the conserved quantities of geodesic motion , namely the energy @xmath13 and the angular momentum @xmath15 .
the expressions are derived from the radiative self force , following galtsov @xcite .
galtsov also shows that identical expressions are obtained by using the fluxes of energy and angular momentum to infinity and down the black hole horizon .
[ this result has recently been independently derived for circular , equatorial orbits in ref . .
] we also show that the time - averaged rate of change of the renormalized rest mass of the particle vanishes .
in section [ sec : kdot ] we derive an expression for the time - averaged rate of change of the carter constant , using the radiative self - force and the mode expansion of the radiative green s function .
this expression [ eq .
( [ eq : dkdt ] ) below ] is the main new result in this paper .
it involves two new amplitudes that are computed in terms of integrals over the torus in phase space , just as for the amplitudes appearing in the energy and angular momentum fluxes . finally , in sec .
[ sec : circular ] we show that our result correctly predicts the known result that circular orbits remain circular while evolving under the influence of radiation reaction .
this prediction serves as a check of our result . as apparent from the above summary
, about 25 percent of this paper is new material , and the remaining 75 percent is review .
the intent is to give a complete and self - contained treatment of scalar radiation reaction in the kerr spacetime , in a single unified notation , starting with the kerr metric , and ending with formulae for the evolution of all three constants of the motion that are sufficiently explicit to be used immediately in a numerical code .
we consider a point particle of scalar charge @xmath44 coupled to a scalar field @xmath45 .
we denote by @xmath46 the bare rest mass of the particle , to be distinguished from a renormalized rest mass @xmath47 which will occur below .
the particle moves in a spacetime with metric @xmath48 which is fixed ; we neglect the gravitational waves generated by the particle .
the worldline of the particle is @xmath49 , where @xmath50 is proper time .
the action is taken to be s = - d^4 x ( ) ^2 - d\{_0 - q } .
[ eq : action ] varying this action with respect to the worldline yields the equation of motion @xmath51 where @xmath52 is the 4-velocity and @xmath53 is the 4-acceleration .
following poisson @xcite , we define the renormalized mass @xmath47 by and add the electromagnetic coupling term @xmath54 to the action ( ) , the equation of motion ( [ eq : acc ] ) becomes @xmath55 , where @xmath56 is the faraday tensor .
this shows that the renormalized mass ( [ eq : mudef ] ) is the mass that would be measured by coupling to other fields . ]
( ) = _ 0 - q .
[ eq : mudef ] the equation of motion ( [ eq : acc00 ] ) can then be written as @xmath57 it is also useful to rewrite the definition ( [ eq : mudef ] ) of the renormalized mass in terms of a differential equation for its evolution with time : ( ) = - q u^_. [ eq : dmudt0 ] the phenomenon of evolution of renormalized rest - mass is discussed further in refs .
@xcite . equations ( [ eq : acc ] ) and ( [ eq : dmudt0 ] ) can also be combined to give the expression f^= u^ _ ( u^ ) = q ^. [ eq : tsf ] for the total self force .
the components of @xmath58 parallel to and perpendicular to the four velocity yield the quantity @xmath59 and @xmath47 times the self - acceleration , respectively .
varying the action ( [ eq : action ] ) with respect to the field @xmath45 gives the equation of motion ( x ) = t(x ) , [ eq : wave ] where the scalar source is ( x ) = - q _
^(4)[x , z ( ) ] .
[ eq : sourcedef ] here ^(4)(x , x ) = ^(4)(x - x)/ [ eq : deltafndef ] is the generalized dirac delta function , where @xmath60 .
we will assume that there is no incoming scalar radiation , so that the physical solution of the wave equation ( [ eq : wave ] ) is the retarded solution _
ret(x ) = d^4 x g_ret(x , x ) t(x ) = - q dg_ret[x , z ( ) ] . [ eq : phiret ] here @xmath61 is the retarded green s function for the scalar wave equation ( [ eq : wave ] ) . the retarded field ( [ eq : phiret ] ) must of course be regularized before being inserted into the expression ( [ eq : acc ] ) for the self - acceleration and into the expression ( [ eq : dmudt0 ] ) for @xmath59 , or else divergent results will be obtained .
the appropriate regularization prescription has been derived by quinn @xcite .
the regularized self - acceleration at the point @xmath62 is @xcite a^ ( ) = ( g^ + u^u^ ) , [ eq : selfa1 ] and the regularized expression for @xmath59 is ( ) = - r + q^2 _ 0 _ -^- d u^_g_ret[z(),z( ) ] .
[ eq : dmudt1 ] as discussed in the introduction , we specialize in this paper to bound motion about a kerr black hole of mass @xmath4 with @xmath63 . in this case
the terms in eqs.([eq : selfa1 ] ) and ( [ eq : dmudt1 ] ) involving the ricci tensor @xmath64 vanish .
also , as long as the orbit is not very close to the innermost stable orbit , the evolution of the orbit is adiabatic : the orbital evolution timescale @xmath20 is much longer than the orbital timescale @xmath21 .
this adiabaticity allows a significant simplification of the formulae ( [ eq : selfa1 ] ) and ( [ eq : dmudt1 ] ) for the self - acceleration and for the rate of change of mass @xmath59 , as shown by mino @xcite .
namely , these formulae reduce to the simple expressions ( [ eq : acc ] ) and ( [ eq : dmudt0 ] ) , but with @xmath45 replaced by the radiative field @xmath65 here @xmath66 is the advanced solution of eq.([eq : wave ] ) . in sec . [
sec : minoproof ] below we review mino s argument , specialized to the scalar case .
to summarize , the starting point for our analysis is the self - acceleration expression a^= ( g^ + u^u^ ) _ _ rad .
[ eq : selfacc0 ] together with the expression for the evolution of rest mass ( ) = - q u^ _ _ rad .
[ eq : dmudt2 ] equations ( [ eq : selfacc0 ] ) and ( [ eq : dmudt2 ] ) are equivalent to the equation f^= u^ _ ( u^ ) = q ^_rad . [ eq : tsf1 ] for the total self - force .
this section summarizes the notation and results of drasco & hughes @xcite , and also some of the results of mino @xcite , for generic bound geodesic orbits in kerr . in boyer - lindquist coordinates @xmath67 , the kerr metric is @xmath68 here @xmath69 and @xmath4 , @xmath70 are the black hole mass and spin parameter . throughout the rest of this paper we use units in which @xmath71 , for simplicity . the square root of the determinant of the metric is = , [ eq :
detg ] and the wave operator is given by @xmath72 \phi_{,tt } - \frac{4 a r}{\delta } \phi_{,t\phi } + \left ( \frac{1}{\sin^2 \theta } - \frac{a^2}{\delta } \right ) \phi_{,\phi\phi } \nonumber \\ \lo + ( \delta \phi_{,r})_{,r } + \frac{1}{\sin \theta } ( \sin \theta \phi_{,\theta})_{,\theta}. \label{eq : waveoperator}\end{aligned}\ ] ] the differential operator that appears on the right hand side of this equation is the teukolsky differential operator @xcite for spin @xmath73 .
we will use later the kinnersley null tetrad @xmath74 , @xmath75 , @xmath76 , @xmath77 , which is given by = _
t + _ r + _ , [ eq : vecldef ] = _ t - _ r + _ , [ eq : vecndef ] and = ( i a _ t + _ + _ ) .
the corresponding one - forms are = - dt + a ^2 d+ dr , = - dt + d- dr , and = ( - i a dt + d+ i ^2 d ) .
the basis vectors obey the orthonormality relations @xmath78 and @xmath79 while all other inner products vanish .
the metric can be written in terms of the basis one - forms as g _
( n _ ) + 2 m _ ( m_)^*. [ eq : gabformula ] we define the conserved energy per unit rest mass @xmath47 e = - u , [ eq : edef ] the conserved @xmath14-component of angular momentum divided by @xmath80 l_z = u , [ eq : lzdef ] and carter constant divided by @xmath81 q = u_^2 - a^2 ^2 e^2 + ^2 l_z^2 + a^2 ^2 .
[ eq : qdef ] the geodesic equations can then be written in the form @xcite
@xmath82 ^ 2- \delta\left[r^2 + ( l_z - a e)^2 + q\right ] \equiv v_r(r)\ ; , \label{eq : rdot}\\ % \fl \left(\frac{d\theta}{d\lambda}\right)^2 = q - \cot^2\theta l_z^2 -a^2\cos^2\theta(1 - e^2 ) \equiv v_\theta(\theta)\ ; , \label{eq : thetadot}\\ % \fl \frac{d\phi}{d\lambda } = \csc^2\theta l_z + ae\left(\frac{r^2+a^2}{\delta } - 1\right ) - \frac{a^2l_z}{\delta } \equiv v_\phi(r,\theta)\ ; , \label{eq : phidot}\\ % \fl \frac{dt}{d\lambda } = e\left[\frac{(r^2+a^2)^2}{\delta } - a^2\sin^2\theta\right ] + al_z\left(1 - \frac{r^2+a^2}{\delta}\right ) \equiv
v_t(r,\theta)\;. \label{eq : tdot}\end{aligned}\ ] ] here @xmath83 is the mino time parameter @xcite , related to proper time @xmath50 by d= d. [ eq : minotime ] also these equations define the potentials @xmath84 , @xmath85 , @xmath86 and @xmath87 . , @xmath88 , @xmath89 and @xmath90 in ref . .
we no not use this notation here since it would clash with the functions @xmath91 , @xmath92 and @xmath45 defined in eq .
( ) below . ] sometimes it will be convenient to use instead of the carter constant @xmath16 the quantity k = q + ( l_z - a e)^2 . for convenience we will also call this quantity
the `` carter constant '' . in the schwarzschild limit @xmath16 and @xmath93
are given by @xmath94 and @xmath95 .
the quantity @xmath93 can be written as k = k^ u_u _ , [ eq : kdef ] where @xmath96 is the killing tensor k^ = 2 m^ ( |m^ ) - a^2 ^2 g^. using the identity ( [ eq : gabformula ] ) this can also be written as k^ = 2 l^ ( n^ ) + r^2 g^. [ eq : kalphabeta ] using the formulae ( [ eq : vecldef ] ) and ( [ eq : vecndef ] ) for the null vectors @xmath74 and @xmath75 together with the definitions ( [ eq : edef ] ) and ( [ eq : lzdef ] ) of @xmath13 and @xmath15 , we obtain from eq .
( [ eq : kalphabeta ] ) the following formula for @xmath93 : k = ( ^2 e - a l_z)^2 - u_r^2 - r^2 .
[ eq : kformula1 ] solving this for @xmath97 gives u_r = ; [ eq : urformula ] this formula will be useful later . following mino @xcite
, we parameterize any geodesic by seven parameters : e , l_z , q,_r0,_0,t_0,_0 . here
@xmath98 and @xmath99 are the values of @xmath100 and @xmath101 at @xmath102 .
the quantity @xmath103 is the value of @xmath83 nearest to @xmath104 for which @xmath105 , where @xmath106 is the minimum value of @xmath40 attained on the geodesic .
similarly @xmath107 is the value of @xmath83 nearest to @xmath104 for which @xmath108 , where @xmath109 is the minimum value of @xmath41 attained on the geodesic .
this parameterization is degenerate because of the freedom to reparametrize the geodesic via @xmath110 .
we discuss this degeneracy further in sec.[sec : degeneracy ] . frequently in this paper we will focus on the _ fiducial geodesic _ associated with the constants @xmath13 , @xmath15 and @xmath16 , namely the geodesic with _
r0 = _ 0 = t_0 = _ 0 = 0 .
it follows from the geodesic equations ( [ eq : rdot ] ) and ( [ eq : thetadot ] ) that the functions @xmath111 and @xmath112 are periodic .
we denote the periods by @xmath113 and @xmath114 , respectively , so r(+ _ r ) = r ( ) , ( + _ ) = ( ) .
[ eq : periodic ] using the initial condition @xmath115 we can write the solution @xmath111 to eq .
( [ eq : rdot ] ) explicitly as r ( ) = r(- _ r0 ) , [ eq : rmotion ] where the function @xmath116 is defined by _ r_min^r ( ) = .
[ eq : hatrdef ] similarly we can write the solution @xmath112 to eq .
( [ eq : thetadot ] ) explicitly as ( ) = ( - _ 0 ) , [ eq : thetamotion ] where the function @xmath117 is defined by _
_ min^ ( ) = .
[ eq : hatthetadef ] the functions @xmath116 and @xmath117 are just the @xmath40 and @xmath41 motions for the fiducial geodesic
. next , the function @xmath87 that appears on the right hand side of eq .
( [ eq : tdot ] ) is a sum of a function of @xmath40 and a function of @xmath41 : v_t(r , ) = v_tr(r ) + v_t ( ) , where @xmath118 and @xmath119 . therefore using @xmath120 we obtain t ( ) = t_0 + _ 0^d \ { v_tr[r( ) ] + v_t[( ) ] } .
[ eq : tmotion0 ] next we define the averaged value @xmath121 of @xmath122 to be v_tr = _ 0^_r dv_tr[r ( ) ] = _ 0^_r dv_tr[r ( ) ] . here
the second equality follows from the representation ( [ eq : rmotion ] ) of the @xmath40 motion together with the periodicity condition ( [ eq : periodic ] ) .
similarly we define v_t = _ 0^ _ dv_t [ ( ) ] = _ 0^ _ dv_t [ ( ) ] . inserting these definitions into eq .
( [ eq : tmotion0 ] ) allows us to write @xmath100 as a sum of a linear term and terms that are periodic @xcite : @xmath123 here we have defined the constant = v_tr + v_t , [ eq : gammadef ] and the functions t_r ( ) = _ 0^d \ { v_tr[r( ) ] - v_tr } , [ eq : deltatrdef ] t _ ( ) = _ 0^d \ { v_t[( ) ] - v_t } .
[ eq : deltatthetadef ] the key property of these functions is that they are periodic : t_r(+ _ r ) = t_r ( ) , t_(+ _ ) = t _ ( ) ; this follows from the definitions ( [ eq : deltatrdef ] ) and ( [ eq : deltatthetadef ] ) together with the periodicity condition ( [ eq : periodic ] ) . we can exhibit the dependence of these functions on the parameters @xmath103 and @xmath107 by substituting the formulae ( [ eq : rmotion ] ) and ( [ eq : thetamotion ] ) for @xmath111 and @xmath112 into eqs .
( [ eq : deltatrdef ] ) and ( [ eq : deltatthetadef ] ) .
the result is t_r ( ) = t_r(- _ r0 ) - t_r(-_r0 ) , [ eq : deltatrformula ] t _ ( ) = t_(- _ 0 ) - t_(-_0 ) , [ eq : deltatthetaformula ] where the functions @xmath124 and @xmath125 are defined by _ r ( ) = _ 0^d \ { v_tr[r( ) ] - v_tr } , [ eq : hattrdef ] _
( ) = _ 0^d \ { v_t[( ) ] - v_t } .
[ eq : hattthetadef ] the motion in @xmath43 can be analyzed in exactly the same way as the motion in @xmath42 .
first , the function @xmath86 that appears on the right hand side of eq .
( [ eq : phidot ] ) is a sum of a function of @xmath40 and a function of @xmath41 : v_(r , ) = v_r(r ) + v _ ( ) , where @xmath126 and @xmath127 . therefore using @xmath128 we obtain ( )
0^d \ { v_r[r( ) ] + v_[( ) ] } .
[ eq : phimotion0 ] next we define the averaged value @xmath129 of @xmath130 to be v_r = _ 0^_r dv_r[r ( ) ] = _ 0^_r dv_r[r ( ) ] . here
the second equality follows from the representation ( [ eq : rmotion ] ) of the @xmath40 motion together with the periodicity condition ( [ eq : periodic ] ) .
similarly we define v _ = _ 0^ _ dv _ [ ( ) ] = _ 0^ _ dv _ [ ( ) ] . inserting these into eq .
( [ eq : phimotion0 ] ) and using @xmath131 allows us to write @xmath101 as a sum of a linear term and terms that are periodic @xcite : @xmath132 here we have defined the constant _ = v_r + v _ , [ eq : upsilonphidef ] and the functions _ r ( ) = _ 0^d \ { v_r[r( ) ] - v_r } , [ eq : deltaphirdef ] _ ( ) = _ 0^d \ { v_[( ) ] - v _ } .
[ eq : deltaphithetadef ] the key property of these functions is that they are periodic : _ r(+ _
( ) , _ ( + _ ) = _ ( ) ; this follows from the definitions ( [ eq : deltaphirdef ] ) and ( [ eq : deltaphithetadef ] ) together with the periodicity condition ( [ eq : periodic ] ) . we can exhibit the dependence of these functions on the parameters @xmath103 and @xmath107 by substituting the formulae ( [ eq : rmotion ] ) and ( [ eq : thetamotion ] ) for @xmath111 and @xmath112 into eqs .
( [ eq : deltaphirdef ] ) and ( [ eq : deltaphithetadef ] ) . the result is _ r ( ) = _ r(- _ r0 ) - _ r(-_r0 ) , [ eq : deltaphirformula ] _ ( ) = _ ( - _ 0 ) - _ ( -_0 ) , [ eq : deltaphithetaformula ] where the functions @xmath133 and @xmath134 are defined by _ r ( ) = _ 0^d \ { v_r[r( ) ] - v_r } , [ eq : hatphirdef ] _ ( ) = _ 0^d \ { v_[( ) ] - v _ } .
[ eq : hatphithetadef ] not all of the parameters @xmath13 , @xmath15 , @xmath16 , @xmath103 , @xmath107 , @xmath98 and @xmath99 that characterize the geodesic are independent .
this is because of the freedom to change the dependent variable @xmath83 via @xmath135 . under this change of variable
the parameters @xmath103 and @xmath107 transform as _
r0 _ r0 = _ r0 + , [ eq : t1 ] _ 0 _ 0 = _ 0 + .
[ eq : t2 ] we can compute how the parameters @xmath98 and @xmath99 transform as follows . combining eqs .
( [ eq : tmotion1 ] ) , ( [ eq : deltatrformula ] ) and ( [ eq : deltatthetaformula ] ) gives the following formula for the @xmath42 motion : t = t_0 + + t_r(- _
r0 ) - t_r(-_r0 ) + t_(- _ 0 ) - t_(-_0 ) .
rewriting the right hand side in terms of @xmath136 , @xmath137 and @xmath138 yields @xmath139 comparing the first and second lines here allows us to read off the value of @xmath140 : _ 0 = t_0 - + t_r(-_r0 - ) - t_r(-_r0 ) + t_(-_0 - ) - t_(-_0 ) .
[ eq : t3 ] similarly we obtain _ 0 _ 0 = _ 0 - _ + _ r(-_r0 - ) - _ r(-_r0 ) + _ ( -_0 - ) - _ ( -_0 ) .
[ eq : t4 ] in sec .
[ sec : dep ] below we will explicitly show that all of the amplitudes and waveforms we compute are invariant under these transformations .
we will often encounter functions @xmath141 of mino time @xmath83 that are periodic with period @xmath114 . for such functions
the average over mino time is given by f__= _
0^ _ f _ ( ) d .
[ eq : thetaaverage ] in order to compute such averages it will be convenient to change the variable of integration from @xmath83 to a new variable @xmath142 .
this variable is a generalization of the newtonian true anomaly and is defined as follows @xcite .
the potential @xmath85 defined by eq .
( [ eq : thetadot ] ) can be written in terms of the variable z ^2 as v_(z ) = .
[ eq : thetaformula ] here @xmath143 and @xmath144 and @xmath145 are the two zeros of @xmath146 .
they are ordered such that @xmath147 .
the variable @xmath142 is defined by ^2 ( ) = z_- ^2 , together with the requirement that @xmath142 increases monotonically as @xmath83 increases . from the definition ( [ eq : hatthetadef ] ) of @xmath117 together with the formula ( [ eq : thetaformula ] ) for @xmath146 we get = . therefore the average ( [ eq : thetaaverage ] ) can be written as f__= _ 0 ^ 2 d. [ eq : thetaaverage1 ] a similar analysis applies to averages of functions @xmath148 that are periodic with period @xmath113 @xcite .
the average over @xmath83 of such a function is f_r _ = _
0^_r f_r ( ) d .
[ eq : raverage ] now the potential @xmath84 in eq .
( [ eq : rdot ] ) can be written as @xcite v_r(r ) = ( 1-e^2)(r_1-r)(r - r_2)(r - r_3)(r - r_4 ) , where @xmath149 , @xmath150 , @xmath151 and @xmath152 are the four roots of the quartic @xmath153 , ordered such that @xmath154 . for stable orbits
the motion takes place in @xmath155 .
the orbital eccentricity @xmath156 and semi - latus rectum @xmath157 are defined by r_1 = and r_2 = .
we also define the parameters @xmath158 and @xmath159 by r_3 = , r_4 = . then following ref .
@xcite we define the parameter @xmath160 by ( ) = .
[ eq : psidef ] it follows from these definitions that @xcite = p ( ) , where ( ) = ^1/2 ^1/2 .
the average ( [ eq : raverage ] ) over @xmath83 is therefore f_r _ = _
0 ^ 2 d = .
[ eq : raverage1 ] note that the viewpoint on the new variables @xmath142,@xmath160 adopted here is slightly different to that in @xcite : here they are defined in terms of the fiducial motions @xmath116 and @xmath117 instead of the actual motions @xmath111 and @xmath112 .
the new viewpoint facilitates the computation of amplitudes and fluxes for arbitrary geodesics ; if one is working with the fiducial geodesic the distinction is unimportant .
finally suppose we have a function @xmath161 of two parameters @xmath162 and @xmath163 which is biperiodic : @xmath164 the average value of this function is f _ = _ 0^_r d_r _ 0^ _ d_f(_r , _ ) . by combining the results ( [ eq : thetaaverage1 ] ) and ( [ eq : raverage1 ] ) we can write this average as a double integral over the new variables @xmath142 and @xmath160 : f _ = _ 0 ^ 2 d_0 ^ 2 d. [ eq : averageidentity ] this formula will be used in later sections .
in this section we review the proof by mino @xcite that for computing radiation reaction in the adiabatic limit in kerr , one can use the `` half retarded minus half advanced '' prescription .
we specialize to the scalar case .
we start by defining some notation for self - forces .
suppose that we have a particle with scalar charge @xmath44 at a point @xmath165 in a spacetime @xmath4 with metric @xmath48 .
suppose that the 4-velocity of the particle at @xmath165 is @xmath52 , and that we are given a solution @xmath45 ( not necessarily the retarded solution ) of the wave equation ( [ eq : wave ] ) for the scalar field , for which the source is a delta function on the geodesic determined by @xmath165 and @xmath52 .
the self - force on the particle is then some functional of @xmath165 , @xmath52 , @xmath48 and @xmath45 , which we write as f^. here we suppress the trivial dependence on @xmath44 .
note that this functional does not depend on a choice of time orientation for the manifold , and also it is invariant under @xmath166 .
next , we define the retarded self - force as f^_ret= f^ , [ eq : selfret ] where @xmath167 is the retarded solution to the wave equation ( [ eq : wave ] ) using the time orientation that is determined by demanding that @xmath52 be future directed . similarly we define the advanced self force by f^_adv= f^ , where @xmath66 is the advanced solution .
it follows from these definitions that f^_ret= f^_adv .
[ eq : flip ] the derivation of the half - retarded - minus - half - advanced prescription in the adiabatic limit rests on two properties of the self - force .
the first property is the fact that the self - force can be computed by subtracting from the divergent field @xmath45 a locally constructed singular field @xmath168 that depends only on the spacetime geometry in the vicinity of the point @xmath165 and on the four velocity at @xmath165 .
this property is implicit in the work of quinn @xcite and quinn and wald @xcite , and was proved explicitly by detweiler and whiting @xcite .
we can write this property as f^= q ^ ( - _ sing [ p , u^,g _ ] ) .
[ eq : property0 ] the second property is a covariance property .
suppose that @xmath160 is a diffeomorphism from the manifold @xmath4 to another manifold @xmath169 that takes @xmath165 to @xmath170 .
we denote by @xmath171 the natural mapping of tensors over the tangent space at @xmath165 to tensors over the tangent space at @xmath170 .
we also denote by @xmath171 the associated natural mapping of tensor fields on @xmath4 to tensor fields on @xmath169
. then the covariance property is f^= ^ * f^. this expresses the fact that the self - force does not depend on quantities such as a choice of coordinates .
it follows from the definition ( [ eq : selfret ] ) that the retarded self - force @xmath172 satisfies a similar covariance relation : f^_ret[(p ) , ^ * u^ , ^ * g _ ] = ^ * f^_ret[p , u^ , g _ ] .
[ eq : covariance ] next we review a property of bound geodesics in kerr upon which the proof depends .
this is the fact that for generic bound geodesics there exists isometries @xmath160 of the kerr spacetime of the form t 2 t_1 - t , 2 _ 1 - , [ eq : isometry ] where @xmath34 and @xmath173 are constants , which come arbitrarily close to mapping the geodesic onto itself . to see this , note that if a geodesic is mapped onto itself by the mapping ( [ eq : isometry ] ) , then the point @xmath174 $ ] is mapped onto @xmath175 $ ] , which must equal @xmath176 $ ] where @xmath177 . if we specialize to @xmath178 then we see that @xmath179 .
we denote this point by @xmath180 .
next , since the geodesic is determined by @xmath181 and by the four - velocity @xmath52 at @xmath182 , the geodesic is mapped onto itself by @xmath160 if and only if @xmath183 at @xmath181
. this will be true if and only if the @xmath40 and @xmath41 components of @xmath52 vanish .
thus , a geodesic will be invariant under a map of the form ( [ eq : isometry ] ) if and only if it contains a point @xmath181 which is simultaneously a turning point of the @xmath40 and @xmath41 motions .
consider now a generic bound geodesic .
such a geodesic can be characterized by the parameters @xmath13 , @xmath15 , @xmath16 , @xmath103 , @xmath107 , @xmath98 , @xmath99 in the notation of sec .
[ sec : geodesicparameters ] .
the turning points of the @xmath40 motion occur at @xmath184 , where @xmath185 is an integer .
similarly the turning points of the @xmath41 motion occur at @xmath186 , where @xmath187 is an integer .
since generically @xmath113 and @xmath114 will be incommensurate , for any @xmath188 we can find values of the integers @xmath187 and @xmath185 so that |_r , n - _ , m| . thus we can find a point on the geodesic that is arbitrarily to close to being a turning point for both the @xmath40 and @xmath41 motions , and hence there is an isometry of the form ( [ eq : isometry ] ) that comes arbitrarily close to mapping the geodesic onto itself .
below we will simply assume that the isometry does map the geodesic onto itself ; the error associated with this assumption can be made arbitrarily small in the adiabatic limit .
the self - acceleration @xmath53 at a point @xmath165 on a geodesic has three independent components , since @xmath189 .
these three components are determined by the time derivatives of the three conserved quantities @xmath190 in order to compute the evolution of the orbit in the adiabatic limit , it is sufficient to know the time averaged quantities _
t , _ t , _ t , [ eq : timeaverages ] where _ t _ t 1 2 t _
-t^t dt de dt(t ) .
this was shown in ref .
@xcite ( although see refs .
@xcite for caveats related to gauge issues ) , and is also discussed in detail using a two - timescale expansion in ref.@xcite .
we now combine the various results discussed above to derive the half retarded minus half advanced prescription .
we parameterize the geodesic as @xmath191 , and denote by f^_ret(t ) = f_ret^the retarded self force at @xmath192 .
the key idea is to use the covariance property ( [ eq : covariance ] ) , with @xmath160 taken to be the isometry associated with the geodesic .
if we take @xmath165 to be @xmath192 , then the left hand side of ( [ eq : covariance ] ) evaluates to @xmath193 & = & f_{\rm adv}^\alpha [ z^\alpha(2 t_1 -t ) , u^\alpha(2 t_1 - t ) , g_{\alpha\beta } ] \nn \\ & = & f_{\rm adv}^\alpha(2 t_1 - t).\end{aligned}\ ] ] here we have used the property ( [ eq : flip ] ) and the fact that @xmath194 .
this gives f_adv^(2
t_1 -t ) = ^ * f_ret^(t ) .
[ eq : flip1 ] now let @xmath195 denote one of the three conserved quantities @xmath13 , @xmath15 or @xmath93 .
the time derivative of @xmath195 depends linearly on the self force , so we can write ( t ) = _ ( t ) f^_ret(t ) [ eq : dcaledt000 ] for some @xmath196 . below we will compute @xmath196 explicitly for the three different conserved quantities and derive the key property that ^ * _ ( t ) = - _ ( 2 t_1 -t ) .
[ eq : flip2 ] now acting on eq .
( [ eq : dcaledt000 ] ) with @xmath171 and using the relations ( [ eq : flip1 ] ) and ( [ eq : flip2 ] ) yields ( _ f^_ret)(t ) = - ( _ f^ _ adv)(2t_1 -t ) .
taking a time average gives _
t = - _ t , and hence _
t = _ t = 1 2 _ t. finally we use the explicit formula ( [ eq : property0 ] ) for the self force .
the contributions from the locally constructed singular field @xmath197 cancel , and we obtain _
t = q 2 _ t. this can be written as _
t = _ t where f_rad^= q 2 ^ ( _ ret - _ adv ) is the self - force one obtains from the half retarded minus half advanced prescription , cf .
( [ eq : selfacc0 ] ) and ( [ eq : dmudt2 ] ) above .
what we have shown is that although the quantity @xmath198 differs from the true self - force @xmath199 , the difference averages to zero when one takes a time average over the entire geodesic .
it remains to derive the formula ( [ eq : flip2 ] ) . for energy and angular momentum , we can write the conserved quantity as the inner product of a killing vector @xmath200 with the 4-velocity of the particle : = _ u^. here @xmath201 for @xmath202 , and @xmath203 for @xmath204 . taking a derivative with respect to proper time @xmath50
gives = u^ _ ( _ u^ ) = ( _ _ ) u^u^+ _
( u^_u^ ) = _ a^ , where @xmath53 is the 4-acceleration .
here we have used the fact that @xmath205 is a killing vector so @xmath206 .
next we use the geodesic equations ( [ eq : tdot ] ) and ( [ eq : minotime ] ) to translate from a proper time derivative to coordinate time derivative , and we use the relation between self - force and self - acceleration given by eqs .
( [ eq : acc ] ) and ( [ eq : tsf ] ) . using the definition ( [ eq : dcaledt000 ] ) of @xmath196 now gives _
= ( ^_+ u_u^ ) _ .
the formula ( [ eq : flip2 ] ) now follows from @xmath207 and @xmath208 .
turn now to the carter constant .
taking a time derivative of the expression ( [ eq : kdef ] ) for @xmath93 gives @xmath209 here we have used the killing tensor equation @xmath210 .
proceeding as before now gives the expression _
= ( ^_+ u_u^ ) k _
u^. the formula ( [ eq : flip2 ] ) now follows from the relations @xmath207 and @xmath211 .
separation of variables for the scalar wave equation in kerr was first carried out by dieter brill and others in 1972 @xcite , and subsequently generalized to higher spins by teukolsky @xcite .
we substitute the ansatz ( t , r , , ) = r(r ) ( ) e^i m e^-i t [ eq : ansatz ] into the homogeneous version of eq .
( [ eq : wave ] ) , and make use of the expression ( [ eq : waveoperator ] ) for the wave operator .
the result is the two equations ( _ , ) _ , - + a^2 ^2 ^2= - , [ eq : thetaeqn ] and ( r_,r)_,r + ( ^2 ^4 - 4 r a m + m^2 a^2 ) r = ( + ^2 a^2 ) r. [ eq : reqn ] here @xmath83 is the separation constant we label the successive eigenvalues @xmath83 of eq .
( [ eq : thetaeqn ] ) by an integer @xmath212 , with @xmath213 , and we denote these eigenvalues by @xmath214 . in the schwarzschild limit @xmath215
we have @xmath216 .
since the differential equation ( [ eq : thetaeqn ] ) does not depend on the sign of @xmath217 or the sign of @xmath187 , we have _ -l - m = _ lm , [ eq : lambdaidentity ] a property which will be used later .
we denote by @xmath218 the corresponding solutions to eq .
( [ eq : thetaeqn ] ) .
the functions s_lm ( , ) e^i m _ lm ( ) [ eq : somegalmdef ] are the spheroidal harmonics .
these harmonics are orthogonal on the sphere , and we can choose them to be orthonormal like the spherical harmonics : d^2 s_lm(,)^ * s _ lm ( , ) = _ l l _
m m. [ eq : orthonormal ] following galtsov @xcite we also choose the phases of the spheroidal harmonics to satisfy ( p s_lm ) ( , ) s_lm(- , + ) = ( -1)^l s_lm ( , ) [ eq : zparity ] and s_-l -m(,)^ * = ( -1)^m s_lm ( , ) .
[ eq : zparity1 ] here @xmath219 is the parity operator . the radial equation ( [ eq : reqn ] )
can now be simplified by defining u(r ) = r(r ) , [ eq : udef ] and by using the tortoise coordinate @xmath220 defined by dr^ * = ( ^2/ ) dr . [ eq : rstardef0 ] an explicit expression for @xmath220 is r^ * = r + - , [ eq : rstardef ] where @xmath221 are the two roots of @xmath222 . [ recall that we are using units in which the mass @xmath4 of the black hole is unity . ]
the resulting simplified radial equation is ( r^ * ) + v_lm(r^ * ) u(r^ * ) = 0 , [ eq : diffsimple ] where the potential is @xmath223 .
\label{eq : potential}\end{aligned}\ ] ] from eq .
( [ eq : lambdaidentity ] ) it follows that this potential satisfies v_-l - m(r ) = v_lm(r ) , [ eq : videntity ] an identity which will be used later . in this section
we follow the treatment of chrzanowski @xcite as modified slightly by galtsov @xcite .
we follow galtsov s notation except that we omit his subscripts @xmath224 denoting spin , since we have specialized to @xmath73 .
we start by analyzing the asymptotic behavior of the potential ( [ eq : potential ] ) as @xmath225 and as @xmath226 .
we consider first the limit @xmath225 , the past and future event horizons . in this limit @xmath227 and @xmath228 .
we find @xmath229 , where p_m = - m _ + , [ eq : kdef ] and _ + = is the angular velocity of rotation of the horizon .
therefore the solutions of the radial equation near the horizons are of the form u(r ) e^i p_m r^*. in the limit of @xmath226 ( past and future null infinity ) , @xmath230 , so the solutions are of the form u(r )
e^i r^*. we now define , following galtsov , the following solution u_lm^in(r^ * ) = _ lm \ { ll _
lm |p_m|^-1/2 e^- i p_m r^ * , & + .
[ eq : uindef ] this equation serves to define the mode as well as the complex transmission and reflection coefficients @xmath231 and @xmath232 . the coefficient @xmath233 is a normalization constant whose value is arbitrary ; we will discuss a convenient choice of @xmath233 later .
we will often denote the set of subscripts @xmath234 collectively by @xmath235 : thus @xmath236 , @xmath237 , and sometimes we will omit the subscript and use simply @xmath238 , @xmath50 .
the `` in '' mode ( [ eq : uindef ] ) is a mixture of outgoing and ingoing components at past and future null infinity , since the mode function is multiplied by @xmath239 . at the past and future event horizons , the mode is purely ingoing when the sign of @xmath240 is the same as the sign of @xmath217 .
however , from the definition ( [ eq : kdef ] ) of @xmath240 we see that the sign of @xmath241 can be negative ; this occurs for superradiant modes .
thus , at the future event horizon the `` in '' modes can be either ingoing or outgoing .
the crucial feature of the `` in '' modes is that they vanish on the past event horizon .
this feature will be used later in constructing the various green s functions .
a more precise statement of the result is that a solution @xmath45 of the wave equation which is a linear combination of `` in '' modes with coefficients @xmath242 , such that the coefficients depend smoothly on @xmath217 ( a reasonable requirement ) , must vanish at the past event horizon . to see this , note from eqs .
( [ eq : ansatz ] ) and ( [ eq : udef ] ) that the solution can be written as ( t , r , , ) = _ -^d_l=0^_m =- l^l e^-i t c_lm s_lm ( , ) .
we now insert the asymptotic form ( [ eq : uindef ] ) of the mode function near the horizon , and we use the definition ( [ eq : kdef ] ) of @xmath240 .
this gives @xmath243 now all of the quantities that depend on @xmath217 in the integrand are smooth functions of @xmath217 .
since fourier transforms of smooth functions go to zero at infinity , it follows that the function @xmath244 defined by eq .
( [ eq : gdef ] ) satisfies @xmath245 as @xmath246 , where @xmath247 .
thus , @xmath45 will vanish as @xmath248 , on the past event horizon .
another important property of the `` in '' modes is the identity u^in_-l - m(r^ * ) = u^in_lm(r^*)^*. [ eq : uinidentity ] to derive this , note that the potential @xmath249 is real and also satisfies the identity ( [ eq : videntity ] ) .
it follows that the function @xmath250 satisfies the same differential equation ( [ eq : diffsimple ] ) as @xmath251 , and has the same asymptotic behavior as @xmath226 if we choose the normalization constants to satisfy _
-l - m = _ lm^*. [ eq : alphacondt ] the result ( [ eq : uinidentity ] ) now follows .
in particular the identity ( [ eq : uinidentity ] ) implies that _ -l - m = _ lm^ * and _ -l - m = _ lm^*. [ eq : tauidentity ] we next define the `` up '' modes u_lm^up(r^ * ) = _ lm \ { ll |p_m|^-1/2 , & + .
[ eq : uupdef ] as before this equation defines both the mode as well as the complex coefficients @xmath252 and @xmath253 . the coefficient @xmath254 is a normalization constant whose value is arbitrary ; we will discuss a convenient choice of @xmath254 later .
the `` up '' modes are a mixture of ingoing and outgoing components at the past and future event horizons . at future null infinity ,
the mode is purely outgoing . by using an argument similar to that given above
, one can show that these modes vanish at past null infinity .
also we can argue as before that u_-l - m^up(r^ * ) = u_lm^up(r^*)^ * , [ eq : uupidentity ] _
-l - m = _ lm^ * , and _
-l - m = _ lm^ * , as long as we choose the normalization constants to satisfy _
-l - m = _ lm^*. [ eq : betacondt ] since the `` in '' modes vanish at the past event horizon , and the `` up '' modes vanish at past null infinity , the two sets of modes are orthogonal to one another and together form a complete basis of modes .
next we note that @xmath255 is a solution of the differential equation ( [ eq : diffsimple ] ) since the potential ( [ eq : potential ] ) is real .
we denote this quantity by u^out_lm = ( u^in_lm)^*. [ eq : uoutdef ] similarly we define u^down_lm = ( u^up_lm)^*. [ eq : udowndef ] these `` down '' and `` out '' modes together form a basis , as the `` out '' modes vanish on the future event horizon and the `` down '' modes vanish at future null infinity . the properties of all of these modes are summarized in fig . [
fig : modes ] .
the four sets of constants @xmath232 , @xmath231 , @xmath252 and @xmath253 are not all independent
. one can derive relations between these coefficients by using the fact that the wronskian w(u_1,u_2 ) = u_1 - u_2 .
[ eq : wronskiandef ] is conserved for any two solutions @xmath256 and @xmath257 of the homogeneous equation ( [ eq : diffsimple ] ) . by evaluating @xmath258 at @xmath259 and at @xmath260 using the asymptotic relations ( [ eq : uindef ] ) and ( [ eq : uupdef ] ) and equating the results ,
we obtain _ lm _ lm=1 . [ eq : widentity1 ] a similar calculation with the modes @xmath261 and @xmath262 yields _ lm^ * _ lm = - _ lm^*. [ eq : widentity2 ] finally , using the modes @xmath263 and @xmath262 yields |
_ lm|^2 + | _ lm|^2 = 1 .
[ eq : widentity3 ] the `` down '' and `` out '' modes can be expressed as linear combinations of the the `` in '' and `` up '' modes , since the latter form a basis of modes . using the asymptotic forms ( [ eq : uindef ] ) and ( [ eq : uupdef ] ) of the modes at @xmath226 together with the definition ( [ eq : udowndef ] ) allows us to identity the coefficients for @xmath264 , giving u^down_lm(r ) = u^in_lm(r ) - _
lm u^up _ lm(r ) .
[ eq : udownformula ] a similar computation using the definition ( [ eq : uoutdef ] ) and using the asymptotic forms at @xmath225 gives u^out_lm(r ) = u^up_lm(r ) - u^in_lm(r ) .
this can be simplified using the relations ( [ eq : widentity1 ] ) and ( [ eq : widentity2 ] ) to give u^out_lm(r )
= | _ lm|^2 u^up_lm(r ) + _ lm^ * u^in_lm(r ) .
[ eq : uoutformula ] finally , following galtsov @xcite , we define the complete mode functions ^in_lm(t , r , , ) = e^-i t s_lm ( , ) u^in_lm(r^ * ) , [ eq : piindef ] ^up_lm(t , r , , ) = e^-i t s_lm ( , ) u^up_lm(r^ * ) , [ eq : piupdef ] ^out_lm(t , r , , ) = e^-i t s_lm ( , ) u^out_lm(r^ * ) , [ eq : pioutdef ] and ^down_lm(t , r , , ) = e^-i t s_lm ( , ) u^down_lm(r^ * ) . [ eq : pidowndef ] galtsov actually includes a factor of @xmath265 in these definitions , where @xmath219 is the parity operator defined by ( pf)(t , r , , ) = f(t , r,-,+ ) for any function @xmath266 and @xmath267 is an additional mode index .
however when acting on the scalar modes considered in this paper , we have from eq.([eq : zparity ] ) that 1 + pp = 1 + p ( -1)^l = 2 _ p,(-1)^l .
thus the @xmath157 index is redundant in the scalar case ; we have simply dropped it and replaced the factor of @xmath268 by @xmath269 .
the retarded green s function @xmath61 is defined such that if @xmath45 obeys the scalar wave equation with source @xmath270 = t , then the retarded solution is _
ret(x ) = d^4 x g_ret(x , x ) t(x ) .
[ eq : gretdef ] the expression for the retarded green s function in terms of the modes defined in the last section is however eq . ( 2.25 ) of galtsov says @xmath271 ; thus we disagree with galtsov by a factor of @xmath272 .
for this reason we do not assume any values for the normalization constants @xmath233 and @xmath254 in our computations , but leave them unevaluated .
] @xmath273 .
\label{eq : gretformula0}\end{aligned}\ ] ] here @xmath274 is the step function , defined to be @xmath275 for @xmath276 and @xmath277 otherwise .
note that the expression ( [ eq : gretformula0 ] ) is independent of the values chosen for the normalization constants @xmath233 and @xmath254 , since the factor of @xmath278 cancels a factor of @xmath279 present in the definition ( [ eq : uindef ] ) of the `` in '' modes , and similarly for @xmath280 and the `` up '' modes .
the expression ( [ eq : gretformula0 ] ) can be expanded into a more explicit form by using the definitions ( [ eq : piindef ] ) ( [ eq : pidowndef ] ) of the complete mode functions @xmath281 in terms of the radial mode functions @xmath282 , together with the definitions ( [ eq : uoutdef ] ) and ( [ eq : udowndef ] ) of the `` out '' and `` down '' modes .
this gives @xmath283 .
\label{eq : gretformula1}\end{aligned}\ ] ] in this section we review the standard derivation of the formula ( [ eq : gretformula1 ] ) . the key idea behind
the derivation is the following .
suppose that the source @xmath284 is non - zero only in the finite range of values of @xmath40 r_min r r_max .
then , the retarded solution @xmath285 will be a solution of the homogeneous equation in the regions @xmath286 and @xmath287 .
now , the retarded solution is determined uniquely by the condition that it vanish on the past event horizon and on past null infinity .
this property will be guaranteed if : 1 . when we expand @xmath167 in the region @xmath286 on the basis of solutions @xmath288 and @xmath289 of the homogeneous equation , only the `` in '' modes contribute .
then , since the `` in '' modes vanish on the past event horizon , @xmath167 must also vanish on the past event horizon .
2 . when we expand @xmath167 in the region @xmath290 on the basis of solutions @xmath288 and @xmath289
, only the `` up '' modes contribute .
then , since the `` up '' modes vanish on past null infinity , @xmath167 must also vanish on past null infinity .
we start by defining the fourier transformed quantities ( , r , , ) = _
-^dt e^i t t(t , r , , ) and ( , r , , ) = _ -^dt e^i t
( t , r , , ) .
we make the following ansatz for the green s function : g_ret(t , r,,;t,r,, ) = _ -^ e^-i ( t - t ) g_ret(r,,;r,, ; ) .
[ eq : ansatz0 ] inserting these definitions into the defining relation ( [ eq : gretdef ] ) and using the formula ( [ eq : detg ] ) for @xmath291 gives ( , r , , ) = _
0^dr d^2 ( r, ) g_ret(r,,;r,, ; ) ( , r,, ) .
[ eq : step1 ] next , we decompose the quantities @xmath292 and @xmath293 on the basis of spheroidal harmonics : ( , r , , ) = _ lm s_lm ( , ) r_lm(r ) [ eq : philmdef ] and ( r , ) ( , r , , ) = r^2 _ lm s_lm ( , ) _
lm(r ) [ eq : tlmdef ] we include the factor of @xmath294 on the left hand side of eq.([eq : tlmdef ] ) because of the appearance of a factor of @xmath294 in the integrand of eq .
( [ eq : step1 ] ) .
the factor of @xmath295 on the right hand side of eq .
( [ eq : tlmdef ] ) is so that the coefficients @xmath296 reduce to the conventional spherical harmonic coefficients in the schwarzschild limit @xmath297 . from the orthogonality relation ( [ eq : orthonormal ] )
the inverse transformations are r_lm(r ) = d^2 s_lm(,)^ * ( , r , , ) and r^2 _ lm(r ) = d^2 s_lm(,)^ * ( r , ) ( , r , , ) .
[ eq : inverse2 ] next , we insert the decompositions ( [ eq : philmdef ] ) and ( [ eq : tlmdef ] ) of @xmath292 and @xmath298 into the fourier transform of the differential equation ( [ eq : wave ] ) using eq .
( [ eq : waveoperator ] ) .
this gives + v_lm(r^ * ) u_lm(r^ * ) = s_lm , [ eq : ulmeqn ] where @xmath299 , the potential @xmath249 is given by eq .
( [ eq : potential ] ) , and the source term is s_lm = _ lm . [ eq : slmdef ] we denote by @xmath300 the relevant green s function for the differential equation ( [ eq : ulmeqn ] ) : u_lm(r^ * ) = _ -^dr^ * g_lm(r^*,r^ * ) s_lm(r^ * ) .
[ eq : gomegalmdef ] we will derive an explicit formula for @xmath301 shortly . first , we note that the fourier - transformed retarded green s function @xmath302 can be expressed in terms of @xmath303 via the formula _
ret(r,,;r,, ; ) = _ lm s_lm ( , ) s_lm(,)^ * .
[ eq : ansatz1 ] to see this , insert the ansatz ( [ eq : ansatz1 ] ) into the relation ( [ eq : step1 ] ) and simplify using the definition ( [ eq : inverse2 ] ) of @xmath296 .
this gives ( , r , , ) = _ lm _
0^dr s_lm ( , ) g_lm(r^*,r^ * ) _ lm(r ) . comparing this with the definition ( [ eq : philmdef ] ) of @xmath304 and simplifying using the relations ( [ eq : udef ] ) , ( [ eq : rstardef0 ] ) and ( [ eq : slmdef ] )
finally yields the formula ( [ eq : gomegalmdef ] ) .
hence the ansatz ( [ eq : ansatz1 ] ) is correct .
finally we turn to the derivation of the formula for the green s function @xmath303 for the differential equation ( [ eq : ulmeqn ] ) . from the discussion at the start of this section , the relevant boundary conditions to impose are that g_lm(r^*,r^ * ) u^in_lm(r^ * ) , r^ * -[eq : b1 ] and g_lm(r^*,r^ * ) u^up_lm(r^ * ) , r^ * .
[ eq : b2 ] consider now the expression @xmath305 , \label{eq : ansatz2}\end{aligned}\ ] ] where @xmath306 is the conserved wronskian ( [ eq : wronskiandef ] ) .
this expression satisfies the boundary conditions ( [ eq : b1 ] ) and ( [ eq : b2 ] ) .
also one can directly verify that it satisfies the differential equation ( [ eq : ulmeqn ] ) with the source replaced by @xmath307 , using the fact that the `` in '' and `` up '' modes satisfy the homogeneous version of the differential equation .
this establishes the formula ( [ eq : ansatz2 ] ) .
next , we compute the wronskian @xmath308 using the asymptotic expressions ( [ eq : uindef ] ) and ( [ eq : uupdef ] ) for the mode functions for @xmath226 .
this gives w(u^in_lm , u^up_lm ) = 2 i _ lm _ lm . inserting this into eq .
( [ eq : ansatz2 ] ) and then into eqs.([eq : ansatz1 ] ) and ( [ eq : ansatz0 ] ) finally yields the formula ( [ eq : gretformula1 ] ) .
using the retarded green s function @xmath61 discussed in the last section we can construct the retarded solution @xmath285 of the wave equation ( [ eq : wave ] ) .
similarly , we can construct the advanced solution @xmath309 using the advanced green s function @xmath310 ; this is the unique solution which vanishes on the future event horizon and at future null infinity .
one half the retarded solution minus one half the advanced solution gives the radiative solution : _
rad(x ) = . clearly the radiative solution is given in terms of a radiative green s function _
rad(x ) = d^4 x g_rad(x , x ) t(x ) , where g_rad(x , x ) = .
[ eq : graddef ] the expression for the radiative green s function in terms of the modes defined in sec .
[ sec : modes ] is @xmath311 .
\label{eq : gradformula0}\end{aligned}\ ] ] note that this expression actually independent of the values chosen for the normalization constants @xmath233 and @xmath254 , since the factor of @xmath312 cancels factors of @xmath279 present in the definition ( [ eq : uindef ] ) and ( [ eq : uoutdef ] ) of the `` out '' modes , and similarly for @xmath280 and the `` down '' modes . the expression ( [ eq : gradformula0 ] )
can be expanded into a more explicit form by using the definitions ( [ eq : piindef ] ) ( [ eq : pidowndef ] ) of the complete mode functions @xmath281 in terms of the radial mode functions @xmath282 .
this gives @xmath313 .
\label{eq : gradformula1}\end{aligned}\ ] ] in this section we review the derivation of the formula ( [ eq : gradformula1 ] ) given in galtsov @xcite .
we start by deriving the identity g_adv(x , x ) = g_ret(x,x ) .
[ eq : identitya ] for any two functions @xmath314 and @xmath315 we have the identity ^a ( _ a - _ a ) = - . integrating this identity over a spacetime region @xmath316 and using stokes theorem gives _
v ( ^a - ^a ) d^3 _ a = _ v d^4 x ( - ) .
[ eq : identity0 ] if we choose @xmath316 to be the region of spacetime outside the black hole , then the boundary @xmath317 of @xmath316 consists of the past and future event horizons , and also past and future null infinity .
we now fix two points @xmath318 and @xmath319 in @xmath316 , and we choose ( x ) = g_ret(x , x_1 ) and ( x ) = g_adv(x , x_2 )
. then @xmath320 and @xmath321 , where @xmath322 is the covariant delta function ( [ eq : deltafndef ] ) .
the right hand side of eq.([eq : identity0 ] ) therefore evaluates to ( x_2 ) - ( x_1 ) = g_ret(x_2,x_1 ) - g_adv(x_1,x_2 ) .
[ eq : rhsid ] consider now the left hand side of eq .
( [ eq : identity0 ] ) .
since @xmath45 vanishes on the past event horizon and at past null infinity , and @xmath323 vanishes on the future event horizon and at future null infinity , there is no portion of @xmath317 where both @xmath45 and @xmath323 are nonzero .
therefore the left hand side vanishes , so the expression ( [ eq : rhsid ] ) must vanish , proving the identity ( [ eq : identitya ] ) .
we now insert the identity ( [ eq : identitya ] ) into the definition ( [ eq : graddef ] ) of @xmath324 , and use the explicit expansion ( [ eq : gretformula1 ] ) of @xmath325 .
we also specialize to _ lm = _ lm = 1 ; [ eq : normalizationchoice ] this causes no loss in generality since the final result ( [ eq : gradformula1 ] ) is independent of the values of @xmath233 and @xmath326 .
this gives @xmath327 \nn \\ \mbox { } & & - \frac{1}{8
\pi i } \int_{-\infty}^\infty d\omega e^{i \omega ( t - t ' ) } \sum_{lm } s_{\omega lm}(\theta',\phi ' ) s_{\omega lm}(\theta,\phi)^ * \frac{\omega}{|\omega| } \frac{1}{\varpi \varpi ' } \nn \\ \mbox { } & & \times \left [ u^{\rm up}_{\omega lm}(r ' ) u^{\rm in}_{\omega lm}(r ) \theta(r'-r ) + u^{\rm in}_{\omega lm}(r ' ) u^{\rm up}_{\omega lm}(r ) \theta(r - r ' ) \right ] , \label{eq : gradformula3}\end{aligned}\ ] ] where @xmath328 and @xmath329 . in the second sum
we relabel @xmath330 and @xmath331 , and we use the formula ( [ eq : zparity1 ] ) .
this gives @xmath332 \theta(r - r ' ) \nn \\ \mbox { } & & + \left [ u^{\rm in}_{\omega lm}(r ) u^{\rm up}_{\omega lm}(r ' ) + u^{\rm in}_{-\omega l - m}(r ) u^{\rm up}_{-\omega l - m}(r ' ) \right ] \theta(r'-r ) \bigg\}. \label{eq : gradformula4}\end{aligned}\ ] ] next , since our choice ( [ eq : normalizationchoice ] ) of the normalization constants satisfies the conditions ( [ eq : alphacondt ] ) and ( [ eq : betacondt ] ) , we can use the identities ( [ eq : uinidentity ] ) and ( [ eq : uupidentity ] ) .
this yields @xmath333 \theta(r - r ' ) \nn \\ \mbox { } & & + \left [ u^{\rm in}_{\omega lm}(r ) u^{\rm up}_{\omega lm}(r ' ) + u^{\rm in}_{\omega lm}(r)^ * u^{\rm up}_{\omega lm}(r')^ * \right ] \theta(r'-r ) \bigg\}. \label{eq : gradformula5}\end{aligned}\ ] ] consider now the coefficient of @xmath334 inside the curly brackets in eq .
( [ eq : gradformula5 ] ) .
we denote this quantity by @xmath335 : h(r , r ) = u^up(r ) u^in(r ) + u^up(r)^ * u^in(r)^ * . [ eq : hdef ] here and in the next few paragraphs
we omit for simplicity the subscripts @xmath234 .
next we use the formula ( [ eq : uoutformula ] ) for the `` out '' modes in terms of the `` in '' and `` up '' modes , together with our choice ( [ eq : normalizationchoice ] ) of normalization constants , to obtain u^up(r ) = .
[ eq : uupformula1 ] here p_m / |p_m| is a variable which is @xmath275 for normal modes and @xmath336 for superradiant modes . inserting this into eq .
( [ eq : hdef ] ) and expanding yields @xmath337 .
\label{eq : step3}\end{aligned}\ ] ] next , by using eq .
( [ eq : uupformula1 ] ) to evaluate the quantity @xmath338 we obtain the identity @xmath339 we now use this identity to eliminate the last two terms inside the square brackets in eq .
( [ eq : step3 ] ) , and simplify using the identity ( [ eq : widentity3 ] ) .
this gives h(r , r ) = u^in(r)^ * u^in(r ) + ||^2 u^up(r)^ * u^up(r ) .
[ eq : hformula ] now the right hand side of eq .
( [ eq : hformula ] ) is explicitly invariant under the combined transformations of interchanging @xmath40 and @xmath340 and taking the complex conjugate .
however , from the definition ( [ eq : hdef ] ) of @xmath335 , the left hand side is invariant under complex conjugation .
it follows that both sides of eq.([eq : hformula ] ) are real and also symmetric under interchange of @xmath40 and @xmath340 : h(r , r ) = h(r,r ) .
[ eq : hsym ] next , the quantity inside the curly brackets in the expression ( [ eq : gradformula5 ] ) for @xmath324 is h(r , r ) ( r - r ) + h(r,r ) ( r-r ) . using the symmetry property ( [ eq : hsym ] ) together with @xmath341 ,
this can be written simply as @xmath335 .
therefore we can replace the expression in curly brackets in ( [ eq : gradformula5 ] ) with the expression ( [ eq : hformula ] ) for @xmath335 .
this gives @xmath342 .
\label{eq : gradformula6}\end{aligned}\ ] ] finally , rewriting this in terms of the `` down '' and `` out '' modes using the definitions ( [ eq : uoutdef ] ) and ( [ eq : udowndef ] ) , and reinserting the appropriate factors of the normalization constants yields the formula ( [ eq : gradformula1 ] ) .
using the expression ( [ eq : gretformula0 ] ) for the retarded green s function together with the integral expression ( [ eq : gretdef ] ) , we can compute the retarded field @xmath285 generated by the source @xmath284 .
for the case we are interested in , @xmath284 will be nonzero only in a finite range of values of @xmath40 of the form r_min r r_max . for @xmath290
only the first term in the square brackets in eq .
( [ eq : gretformula0 ] ) will contribute , and the function @xmath334 will always be @xmath343 .
this gives _
ret(x ) = _ -^d_l=0^_m =- l^l z^out_lm ^ up_lm(x ) , r r_max , where the coefficients @xmath344 are given by the inner product of the `` out '' modes with the source : z^out_lm ( ^out_lm , t ) . [ eq : zoutdef ] here following galtsov @xcite we have defined the inner product of any functions @xmath314 and @xmath315 on spacetime by ( , ) d^4 x ( x)^ * ( x ) .
[ eq : innerproductdef ] similarly for @xmath286 we obtain _
ret(x ) = _ -^d_l=0^_m =- l^l z^down_lm ^ in_lm(x ) , r r_min , where z^down_lm ( ^down_lm , t ) . [ eq : zdowndef ] finally , the expression ( [ eq : gradformula0 ] ) for the radiative green s function together with the definitions ( [ eq : zoutdef ] ) and ( [ eq : zdowndef ] )
give the following expression for the radiative field @xmath345 .
\label{eq : phiradformula0}\end{aligned}\ ] ] all of these expressions depend on the amplitudes @xmath346 and @xmath347 . for the case considered here where the source @xmath284 is a point particle on a bound geodesic orbit , the amplitudes @xmath348 and @xmath347 can be expressed as discrete sums over delta functions @xcite : z_lm^out , down = _ k=-^_n=-^ z^out , down_lmkn ( - _ mkn ) .
[ eq : harmonicdecomposition ] here _ mkn = m _ + k _ + n _
r , [ eq : omegamkndef ] _
= , _ = , _ r = , _ = , _ r = , [ eq : upsilondefs ] and @xmath349 and @xmath350 are defined by eqs .
( [ eq : gammadef ] ) and ( [ eq : upsilonphidef ] ) . the formula for the coefficients @xmath351 and
@xmath352 is @xmath353}{\varpi[r(\lambda_r ) ] } \theta_{\omega_{mkn}lm}[\theta(\lambda_\theta)]^ * e^{- i m \delta \phi_r(\lambda_r ) } e^{-i m \delta \phi_\theta(\lambda_\theta ) } \nn \\
\lo \times e^{i \omega_{mkn } \delta t_r(\lambda_r ) } e^{i \omega_{mkn } \delta t_\theta(\lambda_\theta ) } u^{\rm out , down}_{\omega_{mkn}lm}[r(\lambda_r)]^*. \label{eq : zlmknformula}\end{aligned}\ ] ] here the functions @xmath354 , @xmath355 , @xmath356 and @xmath357 are defined in eqs.([eq : deltatrdef ] ) , ( [ eq : deltatthetadef ] ) , ( [ eq : deltaphirdef ] ) , and ( [ eq : deltaphithetadef ] ) .
also the function @xmath358 is defined by the decomposition of the spheroidal harmonic @xmath359 : s_lm ( , ) = _ lm ( ) e^i m , [ eq : sdecompose ] cf .
( [ eq : somegalmdef ] ) above . for evaluating the double integral ( [ eq : zlmknformula ] )
it is convenient to use the identity ( [ eq : averageidentity ] ) that expresses the integral in terms of the variables @xmath142 and @xmath160 . in this section
we review the derivation of the harmonic decomposition ( [ eq : harmonicdecomposition ] ) given by drasco and hughes @xcite , adapted to the scalar case , and generalized from fiducial geodesics to arbitrary bound geodesics .
we consider only the `` out '' amplitude @xmath360 ; the `` down '' case is exactly analogous .
we start by inserting the definition ( [ eq : sourcedef ] ) of the source @xmath284 into the definition of @xmath361 given by eqs.([eq : zoutdef ] ) and ( [ eq : innerproductdef ] ) .
this gives an expression consisting of an integral along the geodesic of the mode function : z_lm^out = - q _
-^d ^ out_lm [ t(),r(),(),()]^*. [ eq : zlmout1 ] inserting the expression ( [ eq : pioutdef ] ) for the mode function @xmath362 in terms of the radial mode function @xmath363 , and using the decomposition ( [ eq : sdecompose ] ) of the spheroidal harmonic @xmath364 now gives z_lm^out = - 2 q _ -^d _ lm[()]^ * e^-i m ( ) e^i t ( ) u_lm^out[r()]^*. we next change the variable of integration from proper time @xmath50 to mino time @xmath83 using eq .
( [ eq : minotime ] ) , and we use the expressions ( [ eq : tmotion00 ] ) and ( [ eq : phimotion00 ] ) for the functions @xmath100 and @xmath101 .
this gives @xmath365 \theta_{\omega lm}[\theta(\lambda)]^ * \nn \\
\lo \times e^{-i m \delta
\phi_r(\lambda ) } e^{-i m \delta \phi_\theta(\lambda ) } e^{i \omega \delta t_r(\lambda ) } e^{i \omega \delta t_\theta ( \lambda ) } \frac{1}{\varpi[r(\lambda ) ] } u_{\omega lm}^{\rm out}[r(\lambda)]^*. \label{eq : zoutformula1}\end{aligned}\ ] ] we now define the function of two variables @xmath366 \theta_{\omega lm}[\theta(\lambda_\theta)]^ * e^{-i m \delta \phi_r(\lambda_r ) } \nn \\ \mbox { } & \times & e^{-i
m \delta \phi_\theta(\lambda_\theta ) } e^{i \omega \delta t_r(\lambda_r ) } e^{i \omega \delta t_\theta ( \lambda_\theta ) } \frac{1}{\varpi[r(\lambda_r ) ] } u_{\omega lm}^{\rm out}[r(\lambda_r)]^*. \label{eq : jomegalmdef}\end{aligned}\ ] ] this function has two key properties .
first , when evaluated at @xmath367 , it coincides with the integrand of eq .
( [ eq : zoutformula1 ] ) , up to the first exponential factor : z_lm^out = _ -^d j_lm ( , ) e^i ( - m _ ) .
[ eq : zoutformula2 ] second , the function @xmath368 is biperiodic : @xmath369 this follows from the fact that the functions @xmath111 , @xmath370 and @xmath371 are all periodic with period @xmath113 , and the functions @xmath112 , @xmath372 and @xmath373 are all periodic with period @xmath114 , cf .
[ sec : geosesics ] above .
standard properties of biperiodic functions now imply that @xmath374 can be written in terms of a double fourier series : j_lm(_r , _ ) = _
k=-^_n=-^j_lmkn e^-i ( k _ _ + n _ r _ r ) , [ eq : doublefourierseries ] where @xmath375 and @xmath376 are given by eq.([eq : upsilondefs ] ) , and where the coefficients @xmath377 are given by j_lmkn = _ 0^_r d_r _ 0^ _ d _ e^i ( k _ _ +
n _ r _ r ) j_lm(_r , _ ) .
[ eq : jomegalmkndef ] we now insert the fourier series ( [ eq : doublefourierseries ] ) for @xmath368 , evaluated at @xmath367 , into the expression ( [ eq : zoutformula2 ] ) for @xmath344 .
this gives @xmath378 where we have used the definition ( [ eq : omegamkndef ] ) of @xmath379 . now comparing with eq.([eq : harmonicdecomposition ] ) we can read off that z^out_lmkn = j__mkn lmkn .
[ eq : zlmknans ] combining this with the definitions ( [ eq : jomegalmdef ] ) and ( [ eq : jomegalmkndef ] ) evaluated at @xmath380 finally gives the formula ( [ eq : zlmknformula ] ) .
note that it follows from the harmonic decomposition ( [ eq : harmonicdecomposition ] ) that for geodesic sources , the continuous frequency @xmath217 and the discrete indices @xmath212 , @xmath187 are replaced with the four discrete indices @xmath381,@xmath185,@xmath212 , and @xmath187 . in this context the operation - , m - m , l l associated with the symmetries ( [ eq : lambdaidentity ] ) , ( [ eq : zparity1 ] ) , ( [ eq : videntity ] ) , ( [ eq : uinidentity ] ) ( [ eq : tauidentity ] ) and ( [ eq : uupidentity ] ) ( [ eq : betacondt ] ) is replaced by the operation k -k , n -n , m -m , l l. the amplitude @xmath382 is a function of the parameters characterizing the geodesic , namely @xmath13 , @xmath15 , @xmath16 , @xmath98 , @xmath99 , @xmath103 , and @xmath107 , cf .
sec.[sec : geodesicparameters ] above .
we write this dependence as z^out_lmkn = z^ out_lmkn(e , l_z , q , t_0,_0,_r0,_0 ) .
we now specialize to the fiducial geodesic associated with the constants @xmath13 , @xmath15 and @xmath16 , i.e. the geodesic for which @xmath383 . for this case we can simplify the formula ( [ eq : zlmknformula ] ) by setting @xmath98 and @xmath99 to zero , by replacing the motions @xmath111 and @xmath112 with the fiducial motions @xmath116 and @xmath117 defined by eqs .
( [ eq : hatrdef ] ) and ( [ eq : hatthetadef ] ) , and by replacing the functions @xmath354 , @xmath355 , @xmath356 and @xmath357 with the functions @xmath384 , @xmath385 , @xmath133 , and @xmath134 .
defined by eqs .
( [ eq : hattrdef ] ) , ( [ eq : hattthetadef ] ) , ( [ eq : hatphirdef ] ) , and ( [ eq : hatphithetadef ] ) .
this yields @xmath386}{\varpi[{\hat r}(\lambda_r ) ] } \theta_{\omega_{mkn}lm}[{\hat \theta}(\lambda_\theta)]^ * e^{- i m { \hat \phi}_r(\lambda_r ) } \nn \\ \lo \times e^{-i m { \hat \phi}_\theta(\lambda_\theta ) } e^{i \omega_{mkn } { \hat t}_r(\lambda_r ) } e^{i \omega_{mkn } { \hat t}_\theta(\lambda_\theta ) } u^{\rm out}_{\omega_{mkn}lm}[{\hat r}(\lambda_r)]^*. \label{eq : zlmknformulafiducial}\end{aligned}\ ] ] for more general geodesics , the amplitude @xmath387 depends on the parameters @xmath98 , @xmath99 , @xmath103 and @xmath107 only through an overall phase : z^out_lmkn(e , l_z , q , t_0,_0,_r0,_0 ) = e^i _ lmkn(t_0,_0,_r0,_0 ) z^out_lmkn(e , l_z , q,0,0,0,0 ) , [ eq : changeparameter ] where @xmath388 \nn \\ \mbox { } & & -\omega_{mkn } \left [ { \hat t}_r(-\lambda_{r0 } ) + { \hat t}_\theta(-\lambda_{\theta0 } ) - t_0 \right ] .
\label{eq : chilmnkdef}\end{aligned}\ ] ] this formula can be derived by substituting the expressions ( [ eq : deltatrformula ] ) , ( [ eq : deltatthetaformula ] ) , ( [ eq : deltaphirformula ] ) and ( [ eq : deltaphithetaformula ] ) for the functions @xmath354 , @xmath355 , @xmath356 and @xmath357 into eq.([eq : zlmknformula ] ) , making the changes of variables in the integral _ r _ r = _ r - _ r0 , _ _ = _ - _ 0 , and comparing with eq .
( [ eq : zlmknformulafiducial ] ) .
finally we note that the phase ( [ eq : chilmnkdef ] ) and amplitude ( [ eq : changeparameter ] ) are invariant under the transformations ( [ eq : t1 ] ) , ( [ eq : t2 ] ) , ( [ eq : t3 ] ) , and ( [ eq : t4 ] ) that correspond to the re - parameterization @xmath110 .
this invariance serves as a consistency check of the formulae , since we expect the invariance on physical grounds .
let @xmath195 be one of the three conserved quantities of geodesic motion , @xmath13 , @xmath15 or @xmath16 . for the purpose of evolving the orbit we would like to compute the quantity _
t , [ eq : desired ] that is , the average with respect to the boyer - lindquist time coordinate @xmath42 of the derivative of @xmath195 with respect to @xmath42 . however , the quantity that is most naturally computed is the derivative with respect to proper time @xmath50 , and the type of average that is most easily computed is the average with respect to mino time @xmath83 . in this section
we therefore rewrite the quantity ( [ eq : desired ] ) in terms of a mino - time average of @xmath389 . in the adiabatic limit
, we can choose a time interval @xmath390 which is long compared to the orbital timescales but short compared to the radiation reaction time is the geometric mean of the orbital time and the radiation reaction time ; this is the time it takes for the phase difference between the geodesic orbit and the true orbit to become of order unity . ] .
then , to a good approximation we have _
t = , where @xmath391 is the change in @xmath195 over this interval .
now let @xmath392 be the change in mino time over the interval . from eq .
( [ eq : tmotion1 ] ) we have t = + oscillatory terms . now
the oscillatory terms will be bounded as @xmath390 is taken larger and larger , and therefore in the adiabatic limit they will give a negligible fractional correction to @xmath390 .
hence we get @xmath393 where the @xmath83 subscript on the angular brackets means an average with respect to @xmath83 .
note that using the definition ( [ eq : gammadef ] ) of @xmath349 we can rewrite this formula as @xmath394 finally we can use eq .
( [ eq : minotime ] ) to rewrite the mino - time derivative in eq .
( [ eq : dcaledt00 ] ) in terms of a proper time derivative .
this gives the final formula which we will use : @xmath395 we now specialize to the cases of energy and angular momentum , @xmath396 or @xmath204 .
for these cases the result for the time - averaged time derivative is @xmath397 here p_mkn p_m_mkn = _ mkn - m _ + , cf . eq . ( [ eq : kdef ] ) above
also the symbol @xmath235 is a shorthand for the set of indices @xmath398 ; we define _
= _ lmkn _ _ mknlm , [ eq : alphalambdadef ] and similarly for @xmath399 and @xmath237 .
the sum over @xmath235 is defined as _ =
_ k=-^_n=-^_l=0^_m =- l^l .
the expression in curly brackets in eq .
( [ eq : dcaledt2 ] ) means either a factor of @xmath379 ( for energy ) or a factor of @xmath187 ( for angular momentum ) .
the first term in the square brackets corresponds to the flux through future null infinity , and the second term corresponds the flux through the horizon .
note that for energy the future null infinity term is always negative ; this makes sense since the expression ( [ eq : dcaledt2 ] ) is the rate of change of orbital energy
. the flux expression ( [ eq : dcaledt2 ] ) has been derived from the radiation reaction force .
as shown by galtsov , the same result is obtained by computing the flux directly using the stress - energy tensor of the scalar field @xcite .
note that the final answer ( [ eq : dcaledt2 ] ) is independent of our choices for the normalization constants @xmath400 and @xmath399 , since the amplitudes @xmath401 and @xmath402 contain factors of @xmath400 and @xmath399 that compensate for the factors that appear explicitly in eq .
( [ eq : dcaledt2 ] ) . for energy and angular momentum , we can write the conserved quantity as the inner product of a killing vector @xmath200 with the 4-velocity of the particle : = _ u^. here @xmath201 for @xmath202 , and @xmath203 for @xmath204 . taking a time derivative gives = u^ _ ( _ u^ )
= ( _ _ ) u^u^+ _ ( u^_u^ ) = _ a^ , where @xmath53 is the 4-acceleration . here
we have used the fact that @xmath205 is a killing vector so @xmath206 .
next , we use the formula ( [ eq : selfacc0 ] ) for the self - acceleration .
this gives @xmath403 \nn \\ \mbox { } & = & \frac{q}{\mu } \xi^\alpha \nabla_\alpha \phi_{\rm rad } + \frac{q}{\mu } { \cal e } \frac{d \phi_{\rm rad}}{d \tau}. \label{eq : dcaledtau0}\end{aligned}\ ] ] now the second term here is a total time derivative to leading order in @xmath44 , since @xmath195 is conserved to zeroth order in @xmath44 .
hence the change in @xmath195 over an interval from @xmath404 to @xmath405 associated with the second term will be .
this is a term which will oscillate but will not grow secularly with time ( since @xmath406 does not grow secularly with time ) . hence in the adiabatic limit
this term will be smaller than the change due to the first term in eq.([eq : dcaledtau0 ] ) by the ratio of the orbital timescale to the inspiral timescale , which is negligible in the approximation we are using . dropping the second term in eq.([eq : dcaledtau0 ] ) and substituting into eq .
( [ eq : used ] ) we get _ t = _ .
[ eq : dcaledtau1 ] next , we can obtain an explicit expression for the radiative field @xmath406 by substituting the harmonic decomposition ( [ eq : harmonicdecomposition ] ) into eq .
( [ eq : phiradformula0 ] ) .
this gives @xmath407 .
\label{eq : phiradformula0a}\end{aligned}\ ] ] here as above @xmath235 denotes the set of indices @xmath398 , and we have defined ^out , down_lmkn = ^out , down__mknlm .
we now substitute this expression into eq.([eq : dcaledtau1 ] ) , and use the fact that the operator @xmath408 gives a factor of @xmath409 ( for the energy ) or @xmath410 ( for the angular momentum ) .
we get @xmath411 , \label{eq : dcaledt1}\end{aligned}\ ] ] where the expression in curly brackets means either a factor of @xmath379 or a factor of @xmath187 .
we now discuss the evaluation of the quantity _ which appears in eq .
( [ eq : dcaledt1 ] ) .
this is the time average of the `` out '' mode function evaluated on the geodesic , and can be easily evaluated using the harmonic decomposition of sec.[sec : harmonic ] . by comparing the integrands of eqs.([eq : zlmout1 ] ) and
( [ eq : zoutformula2 ] ) we find that _ _ mknlm^out[z^ ( ) ] = - j__mknlm(,)^ * e^-i ( _ mkn - _ m ) , where @xmath412 $ ] is the geodesic . now using the fourier series expansion ( [ eq : doublefourierseries ] ) of @xmath368 gives _
_ mknlm^out[z^ ( ) ] = - _ k,n j__mknlmkn^*e^-i ( _ mkn - _ mkn ) . from this expression
it is clear that averaging over @xmath83 kills all the terms in the sum except @xmath413 , @xmath414 , and we obtain _
= - j__mknlmkn^ * = - ( z^out_lmkn)^*. [ eq : zoutuseful ] for the last equation we have used eq .
( [ eq : zlmknans ] ) . finally substituting this result , together with a similar equation for the `` down '' modes , in to eq .
( [ eq : dcaledt1 ] ) we obtain the final result ( [ eq : dcaledt2 ] ) equation ( [ eq : dmudt2 ] ) for the time derivative of the renormalized rest mass @xmath415 can be immediately integrated with respect to @xmath50 between two times @xmath404 and @xmath405 .
this yields ( _ 2 ) - ( _ 1 ) = q _
rad[z(_2 ) ] - q _
rad[z(_1 ) ] , where @xmath49 is the geodesic .
now since the zeroth - order orbit is bound , the right hand side of this equation does not contain any secularly growing terms .
instead it consists only of oscillatory terms .
therefore , over long timescales and in the adiabatic limit ( the regime in which we are working in this paper ) , the rest mass is conserved : = 0 .
we now turn to the corresponding computation for the carter constant @xmath93 defined by eq .
( [ eq : kdef ] ) .
the result is @xmath416 .
\label{eq : dkdt}\end{aligned}\ ] ] this expression has the same structure as the expression ( [ eq : dcaledt2 ] ) for the time derivatives of the energy and angular momentum , except that the squared amplitudes @xmath417 and @xmath418 have been replaced with products of the amplitudes @xmath401 and @xmath402 with two new amplitudes @xmath419 and @xmath420 .
these new amplitudes are defined by the equation @xmath421^ * \nn \\ \mbox { } \times e^{- i m \delta \phi_r(\lambda_r ) } e^{-i m \delta \phi_\theta(\lambda_\theta ) } e^{i \omega_{mkn } \delta t_r(\lambda_r ) } e^{i \omega_{mkn } \delta t_\theta(\lambda_\theta ) } \nn \\ \mbox { } \times \left\ { g_{mkn}[\lambda_r,\lambda_\theta ] r^{\rm out , down}_{\omega_{mkn}lm}[r(\lambda_r)]^ * + g[\lambda_r,\lambda_\theta ] \frac{d r^{\rm out , down}_{\omega_{mkn}lm}}{dr}[r(\lambda_r)]^ * \right\ } , \nonumber \\ \label{eq : tildezlmknformula}\end{aligned}\ ] ] where r^out , down__mknlm(r ) = , g_mkn(_r , _ ) - ( ^2 e - a l_z ) ( ^2 _ mkn- a m ) + 2 i r u_r , [ eq : gmnkdef ] g(_r , _ ) = - i u_r .
[ eq : gdef1 ] on the right hand sides of eqs .
( [ eq : gmnkdef ] ) and ( [ eq : gdef1 ] ) , it is understood that @xmath422 , @xmath423 and @xmath424 are evaluated at @xmath425 and @xmath426 .
also it is understood that @xmath97 is given as a function of @xmath40 by eq.([eq : urformula ] ) , and hence as a function of @xmath162 by using @xmath427 and by resolving the sign ambiguity in eq .
( [ eq : urformula ] ) using eq .
( [ eq : psidef ] ) .
several features of the formulae ( [ eq : dkdt ] ) and ( [ eq : tildezlmknformula ] ) are worth noting : * the formula ( [ eq : tildezlmknformula ] ) for the new amplitudes @xmath428 and @xmath429
. has a very similar structure to the formula ( [ eq : zlmknformula ] ) defining the amplitudes @xmath382 and @xmath430 .
in particular , the derivation of the formula ( [ eq : changeparameter ] ) for the dependence of the amplitudes on the parameters @xmath98 , @xmath99 , @xmath103 and @xmath107 of the geodesic ( via an overall phase ) carries through as before .
this means that the time derivative of the carter constant is independent of these parameters , as expected , since the phase from the untilded amplitudes cancels the phase from the tilded amplitudes in eq.([eq : dkdt ] ) .
* although we have no general definition for `` flux of carter '' , presumably the first term in eq .
( [ eq : dkdt ] ) corresponds to something like the flux of carter to future null infinity , and the second term to the flux of carter down the black hole horizon .
* as was the case for the original amplitudes , to evaluate the double integral ( [ eq : tildezlmknformula ] ) it is convenient to use the identity ( [ eq : averageidentity ] ) to express the integral in terms of the variables @xmath142 and @xmath160 instead of @xmath163 and @xmath162 .
taking a time derivative of the expression ( [ eq : kdef ] ) for @xmath93 gives @xmath209 here we have used the killing tensor equation @xmath210 .
now substituting the expression ( [ eq : selfacc0 ] ) for the self - acceleration gives @xmath431 \nn \\ \mbox { } & = & \frac{2 q}{\mu } k \frac{d \phi_{\rm rad}}{d \tau } + \frac{2 q}{\mu } k^{\alpha\beta } u_{\alpha } \nabla_\beta \phi_{\rm rad}. \label{eq : dkdtau0}\end{aligned}\ ] ] now the first term here is a total time derivative to leading order in @xmath44 , since @xmath93 is conserved to zeroth order in @xmath44 .
hence this term can be neglected for the reason explained in sec.[sec : derivation1 ] .
dropping the first term and substituting into eq .
( [ eq : used ] ) we get _ t = _ .
[ eq : dkdtau1 ] next , we use the explicit expression ( [ eq : phiradformula0a ] ) for the radiation field @xmath406 .
this gives @xmath432 .
\label{eq : dkdt1}\end{aligned}\ ] ] we now define the amplitudes ^out_lmkn = _ [ eq : tildezoutdef ] and ^down_lmkn = _ .
[ eq : tildezdowndef ] substituting these definitions into eq .
( [ eq : dkdt1 ] ) gives the formula ( [ eq : dkdt ] ) .
therefore it remains only to derive the formula ( [ eq : tildezlmknformula ] ) for the amplitudes @xmath433 and @xmath434 .
we will derive the formula for @xmath435 ; the derivation of the formula for @xmath434 is similar .
we start by simplifying the differential operator @xmath436 which appears in the definitions ( [ eq : tildezoutdef ] ) and ( [ eq : tildezdowndef ] ) . using the expression ( [ eq : kalphabeta ] ) for the killing tensor
we can write this as k^ u _ _ = ( l^u _ ) n^ _ + ( n^u _ ) l^_+ r^2 . using the definitions ( [ eq : vecldef ] ) and ( [ eq : vecndef ] ) of @xmath74 and @xmath437 ,
the definitions ( [ eq : edef ] ) and ( [ eq : lzdef ] ) of @xmath13 and @xmath15 , and the fact that @xmath438 and @xmath439 reduce to @xmath440 and @xmath441 when acting on @xmath442 gives @xmath443 consider now the contribution of the first term in eq.([eq : k3 ] ) to the expression ( [ eq : tildezoutdef ] ) for @xmath428 . using the relation ( [ eq : minotime ] ) between @xmath444 and @xmath445 we can write this as @xmath446 we can integrate this by parts with respect to @xmath83 ; the boundary term which is generated can be neglected for the reason explained in sec .
[ sec : derivation1 ] .
this gives @xmath447 where we have used @xmath448 from eqs.([eq : minotime ] ) and ( [ eq : kerrmetric ] ) . combining this with the result obtained from substituting the second and third terms in eq .
( [ eq : k3 ] ) into eq .
( [ eq : tildezoutdef ] ) gives ^out_lmkn = _ , [ eq : tildezoutformula1 ] where the functions @xmath449 and @xmath450 are [ cf .
( [ eq : gmnkdef ] ) and ( [ eq : gdef1 ] ) above ] g_mkn(_r , _ ) - ( ^2 e -a l_z ) ( ^2 _ mkn - a m ) + 2 i r u_r , [ eq : gmnkdef2 ] g(_r , _ ) = - i u_r .
[ eq : gdef2 ] as explained above , it is understood that the functions @xmath422 , @xmath423 and @xmath424 on the right hand sides of eqs.([eq : gmnkdef2 ] ) and ( [ eq : gdef2 ] ) are evaluated at @xmath425 and @xmath426 .
also it is understood that @xmath97 is given as a function of @xmath40 by eq.([eq : urformula ] ) , and hence as a function of @xmath162 by using @xmath427 and by resolving the sign ambiguity in eq .
( [ eq : urformula ] ) using eq .
( [ eq : psidef ] ) .
the average over @xmath83 in eq .
( [ eq : tildezoutformula1 ] ) can be evaluated using the same techniques as in secs .
[ sec : harmonic ] and [ sec : edot ] above . using the definition ( [ eq : pioutdef ] ) of the `` out '' mode function and the definition ( [ eq : omegamkndef ] ) of @xmath379 , the quantity inside the angular brackets in eq .
( [ eq : tildezoutformula1 ] ) can be written as _
mkn ( , ) , where @xmath451^ * \nn \\ \mbox { } \times e^{- i m \delta \phi_r(\lambda_r ) } e^{-i m \delta \phi_\theta(\lambda_\theta ) } e^{i \omega_{mkn } \delta t_r(\lambda_r ) } e^{i \omega_{mkn } \delta t_\theta(\lambda_\theta ) } \nn \\
\mbox { } \times \left\ { g_{mkn}[\lambda_r,\lambda_\theta ] r^{\rm out , down}_{\omega_{mkn}lm}[r(\lambda_r)]^ * + g[\lambda_r,\lambda_\theta ] \frac{d r^{\rm out , down}_{\omega_{mkn}lm}}{dr}[r(\lambda_r)]^ * \right\}. \nonumber \\ \label{eq : tildejformula}\end{aligned}\ ] ] this function is biperiodic : @xmath452 hence it can be expanded as a double fourier series , and the average over @xmath83 of @xmath453 is _ = _ 0^_r d_r _ 0^ _ d__mkn(_r , _ ) .
[ eq : jav ] substituting eq .
( [ eq : tildejformula ] ) into eq .
( [ eq : jav ] ) and back into eq .
( [ eq : tildezoutformula1 ] ) now gives the formula ( [ eq : tildezlmknformula ] ) .
in this section we study the prediction of the formula ( [ eq : dkdt ] ) for the time derivative of the carter constant , for the special case of circular orbits ( orbits of constant boyer - lindquist radius @xmath40 ) .
it is known that circular orbits remain circular while evolving under the influence of radiation reaction in the adiabatic regime @xcite . also for circular orbits
it is possible to relate the carter constant @xmath93 to the energy @xmath13 and angular momentum @xmath15 of the orbit : specializing eq .
( [ eq : kformula1 ] ) to zero radial velocity @xmath454 yields @xmath455 in order for this relation to be preserved under slow evolution of @xmath93 , @xmath13 and @xmath15 , we must have _
[ eq : circtocirccheck ] we now verify that our expression ( [ eq : dkdt ] ) for @xmath456 satisfies the condition ( [ eq : circtocirccheck ] ) . for circular orbits , the following simplifications apply to the harmonic decomposition of the geodesic source derived in sec.[sec :
harmonic ] .
first , the functions @xmath457 , @xmath124 , @xmath371 and @xmath458 defined by eqs .
( [ eq : deltatrformula ] ) , ( [ eq : hattrdef ] ) , ( [ eq : deltaphirformula ] ) and ( [ eq : hatphirdef ] ) vanish identically .
second , the biperiodic function @xmath459 defined in eq.([eq : jomegalmdef ] ) is therefore independent of @xmath162 .
it follows that the double fourier series ( [ eq : doublefourierseries ] ) is replaced by the single fourier series j_lm ( _ ) = _ k=-^j_lmk e^-i k _ _ , [ eq : singlefourierseries ] where the coefficients @xmath460 are given by j_lmk = _ 0^ _ d _ e^i k _ _ j_lm ( _ ) .
[ eq : jomegalmkdef ] the effect of this change on the subsequent formulae for fluxes and amplitudes in secs .
[ sec : edot ] and [ sec : kdot ] can be summarized as : ( i ) remove the sums over @xmath185 ; ( ii ) remove the indices @xmath185 from the amplitudes @xmath461 ; ( iii ) remove the averaging operation @xmath462 from the formulae for the amplitudes ; and ( iv ) evaluate the integrands in the formulae for the amplitudes at @xmath463 .
also the formula ( [ eq : omegamkndef ] ) for the frequency @xmath379 is replaced by _
mk = m _ + k _ . with these simplifications ,
the formula ( [ eq : dcaledt2 ] ) for the time derivatives of energy and angular momentum becomes @xmath464 where @xmath202 or @xmath15 , and the expression in curly brackets means either @xmath465 ( for @xmath13 ) or @xmath187 ( for @xmath15 ) .
similarly the expression ( [ eq : dkdt ] ) for the time derivative of the carter constant becomes @xmath466 .
\label{eq : dkdtcirc}\end{aligned}\ ] ] comparing eqs .
( [ eq : dcaledt2circ ] ) and ( [ eq : dkdtcirc ] ) , we see that the condition ( [ eq : circtocirccheck ] ) will be satisfied as long as the amplitudes @xmath467 satisfy _ lmk^out , down = ( ^2_mk - m a ) z_lmk^out , down .
[ eq : amplitudeidentity ] therefore it suffices to derive the identity ( [ eq : amplitudeidentity ] ) .
we will derive this identity for the `` out '' modes ; the `` down '' derivation is identical
. the particular form of the expressions for the amplitudes that are most useful here are eqs .
( [ eq : zoutuseful ] ) and ( [ eq : tildezoutformula1 ] ) : z^out_lmk = - _ , [ eq : zoutuseful1 ] and ^out_lmk = _ .
[ eq : tildezoutformula2 ] here we have defined @xmath468 .
it follows from eq .
( [ eq : gdef2 ] ) together with @xmath469 that the function @xmath470 vanishes , so the formula for @xmath471 simplifies to ^out_lmk = _ .
[ eq : tildezoutformula3 ] we now substitute into eq .
( [ eq : tildezoutformula3 ] ) the expression ( [ eq : gmnkdef2 ] ) for @xmath472 . the second term in eq .
( [ eq : gmnkdef2 ] ) vanishes since @xmath469 , and of all the factors in the first term , only @xmath294 depends on @xmath83 ; the remaining factors are constants and can be pulled out of the average over @xmath83 .
this yields _ lmk^out = - ( ^2 _ mk - a m ) _ . comparing this with the formula ( [ eq : zoutuseful1 ] ) for @xmath473 now
yields the identity ( [ eq : amplitudeidentity ] ) .
this research was supported in part by nsf grants phy-0140209 and phy-0244424 and by nasa grant nagw-12906 .
we thank the anonymous referees for several helpful comments .
in this appendix we estimate the accuracy of the adiabatic waveforms , expanding on the brief treatment given in ref .
specifically we estimate the magnitude of the correction to the waveform s phase that corresponds to the term @xmath474 in eq.([eq : emri_phase ] ) .
we can roughly estimate this term by using post - newtonian expressions for the waveform in which terms corresponding to @xmath474 are readily identified . while the post - newtonian approximation is not strictly valid in the highly relativistic regime near the horizon of interest here , it suffices to give some indication of the accuracy of the adiabatic waveforms . in this appendix
we specialize for simplicity to circular , equatorial orbits .
we also specialize to gravitational radiation reaction , unlike the body of the paper which dealt with the scalar case .
our conclusion is that adiabatic waveforms will likely be sufficiently accurate for signal detection .
the procedure we use is as follows .
we focus attention on the phase @xmath475 of the fourier transform of the dominant , @xmath476 piece of the gravitational waveform . here
@xmath266 is the gravitational wave frequency . now
a change to the phase function of the form @xmath477 is not observable ; the constant @xmath390 corresponds to a change in the time of arrival of the signal .
therefore we focus on the observable quantity @xmath478 , and we write this as ( f ) = g(f , m_1,m_2 ) .
[ eq : gadef ] here @xmath479 is a function , @xmath480 is the mass of the black hole , and @xmath481 is the mass of the inspiralling compact object , the total mass , and @xmath482 , the reduced mass . throughout the body of the paper we referred to the mass of the inspiralling object as @xmath47 and the mass of the black hole as @xmath4 ; this is valid to leading order in @xmath36 . in this appendix
however we need to be more accurate , and hence we revert to using @xmath480 and @xmath481 for the two masses . ] .
the specific form of @xmath479 obtained from post - newtonian theory is discussed below .
we next expand the function @xmath479 as a power series in @xmath481 , at fixed @xmath266 and @xmath480 .
the leading order term scales as @xmath483 , and we obtain g(f , m_1,m_2 ) = + g_1(f , m_1 ) + g_2(f , m_1 ) m_2 + [ eq : gexpand ] the first term in this expression corresponds to the term @xmath484 in eq .
( [ eq : emri_phase ] ) ; it gives the leading - order , adiabatic waveform .
the error incurred from using adiabatic waveforms is therefore g(f , m_1,m_2 ) = g(f , m_1,m_2 ) - = g_1(f , m_1 ) + g_2(f , m_1 ) m_2 + we want to estimate the effects of this phase error .
it is useful to split the phase error into two terms : g(f , m_1,m_2 ) = g_1(f , m_1,m_2 ) + g_2(f , m_1,m_2 ) , [ eq : deltagsplit ] where g_1(f , m_1,m_2 ) = g(f , m_1,m_2 ) - [ eq : deltag1def ] and g_2(f , m_1,m_2 ) = - .
the phase error @xmath485 corresponds to the error that would be caused by using an incorrect value of the mass of the black hole .
the effect of this error on the data analysis would be to cause a systematic error in the inferred best fit value of the black hole mass .
however , the fractional error would be of order @xmath12 .
this error will not have any effect on the ability of adiabatic waveforms to detect signals when used as search templates .
therefore , we will neglect this error , and focus on the remaining error term @xmath486 .
we now compute explicitly the phase error @xmath487 that corresponds to the term @xmath486 in eq .
( [ eq : deltagsplit ] ) .
we use the post-3.5-newtonian expression for @xmath475 , which can be computed from the orbital energy @xmath488 and gravitational wave luminosity @xmath489 via the equation @xcite = - 2 . for the energy @xmath488
we use the expression given in eq .
( 50 ) of ref.@xcite , and for the luminosity @xmath490 we use eq.(12.9 ) of ref .
@xcite . from ref . and
@xmath491 from refs . .
] this gives ( f ) = 2 f t_c + _ c + ( f ) , [ eq : psi0 ] where @xmath492 and @xmath493 are constants , @xmath494 , and @xmath495 here @xmath496 is euler s constant , @xmath70 is the dimensionless spin parameter of the black hole , and @xmath497 .
we have also added to eq .
( [ eq : fullphase ] ) the spin - dependent terms up to post-2-newtonian order , taken from eq .
( 1 ) of ref.@xcite , specialized to a non - spinning particle on a circular , equatorial orbit .
we now insert the expression ( [ eq : psi0 ] ) for @xmath475 into eqs.([eq : gadef ] ) and ( [ eq : deltag1def ] ) , and integrate twice with respect to @xmath266 to obtain @xmath487 .
the result is @xmath498 , \label{eq : deltapsians}\end{aligned}\ ] ] where @xmath499 .
now as discussed above , a change to the phase function of the form @xmath477 is not observable .
this freedom allows us to set @xmath500 where @xmath501 is a fixed frequency , which amounts to replacing our expression @xmath487 for the phase error with _
1(f ) - _ 1(f_0 ) - _ 1(f_0)(f - f_0 ) .
we now evaluate the phase error @xmath502 for some typical sources .
first , we take @xmath503 , @xmath504 , @xmath505 . for this case the last year of inspiral extends from @xmath506 hz to @xmath507 hz @xcite ; the maximum value of @xmath508 is @xmath509 cycles if we take @xmath510 hz . as a second example with a higher mass ratio , we take @xmath511 , @xmath512 , @xmath513 . the last year of inspiral for this case extends from @xmath514 hz to @xmath515 hz @xcite ; the maximum value of @xmath508 is @xmath516 cycles if we take @xmath517 hz ( see fig .
[ fig : phase ] ) larger . ] . for signals from intermediate mass black holes that may be detectable by ligo , the phase errors are yet smaller .
these phase errors are sufficiently large that adiabatic waveforms can not be used as data - analysis templates ; the template will not quite stay in phase with the signal over the entire @xmath8 cycles of inspiral .
however , the requirements on detection templates are much less stringent : phase coherence is only needed for @xmath518 weeks , rather than a year @xcite . in addition , in the matched filtering search process phase error will tend to be compensated for by small systematic errors in the best - fit mass parameters . because the adiabatic waveforms are _ almost _ good enough for data - analysis templates ( phase errors @xmath519 cycle in the worst cases )
, adiabatic waveforms will likely be accurate enough for detection templates both for ground - based and space - based detectors .
this appendix lists some of the typos in the paper by galtsov @xcite : * the discussion of the range of summation of the indices @xmath212 , @xmath187 given after eq .
( 2.5 ) is incorrect
. it should be summation @xmath212,@xmath187 with @xmath520 and @xmath521 rather than @xmath522 and @xmath523 .
* in the definition of the variable @xmath524 given just before eq .
( 2.7 ) , the factor @xmath525 should be @xmath526 . * in the first of the four equations in eqs . ( 2.9 ) , the term @xmath527 should be @xmath528 . * the right hand side of eq . ( 2.25 ) should be divided by @xmath529 , as already discussed in sec .
[ sec : retardedformula ] above . * in eq .
( 2.30 ) , the quantity @xmath530 should be replaced by its complex conjugate @xmath531 . * in the first line of eq .
( 3.2 ) , the quantity @xmath532 should be replaced by its complex conjugate @xmath533 . * in the second of eqs .
( 3.3 ) , the right hand side should read @xmath534 rather than @xmath535
i.e. the closing bracket is in the wrong place .
* in the second term on the first line of eq .
( 3.6 ) , the first factor of @xmath536 should be omitted . * in eq .
( 3.9 ) , the argument of the last factor should be @xmath537 rather than @xmath538 .
* galtsov s equation ( 4.13 ) for the rate of change of energy or angular momentum of the orbit is correct only for sources which are smooth functions of frequency .
however , for bound geodesics , the inner products @xmath539 are sums of delta functions in frequency , so it does not make sense to square these inner products as galtsov does .
the correct version of this equation is our eq .
( [ eq : dcaledt2 ] ) above .
10 lee samuel finn and kip s. thorne . gravitational waves from a compact star in a circular , inspiral orbit , in the equatorial plane of a massive , spinning black hole , as observed by lisa . ,
d62:124021 , 2000 , gr - qc/0007074 .
kostas glampedakis and daniel kennefick .
zoom and whirl : eccentric equatorial orbits around spinning black holes and their evolution under gravitational radiation reaction .
, d66:044002 , 2002 , gr - qc/0203086 .
luc blanchet , thibault damour , and gilles esposito - farese .
dimensional regularization of the third post - newtonian dynamics of point particles in harmonic coordinates . ,
d69:124007 , 2004 , gr - qc/0311052 .
luc blanchet , thibault damour , gilles esposito - farese , and bala r. iyer
. gravitational radiation from inspiralling compact binaries completed at the third post - newtonian order .
, 93:091101 , 2004 , gr - qc/0406012 . | a key source for lisa will be the inspiral of compact objects into massive black holes .
recently mino has shown that in the adiabatic limit , gravitational waveforms for these sources can be computed by using for the radiation reaction force the gradient of one half the difference between the retarded and advanced metric perturbations . using post - newtonian expansions
, we argue that the resulting waveforms should be sufficiently accurate for signal detection with lisa .
data - analysis templates will require higher accuracy , going beyond adiabaticity ; this remains a significant challenge .
we describe an explicit computational procedure for obtaining waveforms based on mino s result , for the case of a point particle coupled to a scalar field .
we derive an explicit expression for the time - averaged time derivative of the carter constant , and verify that the expression correctly predicts that circular orbits remain circular while evolving under the influence of radiation reaction .
the derivation uses detailed properties of mode expansions , green s functions and bound geodesic orbits in the kerr spacetime , which we review in detail .
this paper is about three quarters review and one quarter new material .
the intent is to give a complete and self - contained treatment of scalar radiation reaction in the kerr spacetime , in a single unified notation , starting with the kerr metric , and ending with formulae for the time evolution of all three constants of the motion that are sufficiently explicit to be used immediately in a numerical code . |
in the recent years the quantum impurity problem out - of - equilibrium has attracted great interest .
the most relevant realizations are the quantum dots connected to at least two metallic leads@xcite and short metallic wires containing magnetic impurities@xcite . in the impurity problem
the exact methods play distinguished roles especially the bethe ansatz and conformal invariance .
the generalization of these methods to out - of - equilibrium situations are the most challenging new directions .
mehta and andrei are aiming to solve the kondo problem on a dot with two contact attached . first a simple resonant level without spin was studied to test the new generalization of the bethe ansatz method@xcite .
their elegant suggestion is very provocative . in order to test this kind of new methods
we perform a detailed study of that problem using different weak coupling perturbative methods combined with nrg . as the final goal
we calculate the current flowing through the impurity when a finite voltage is applied on the contacts .
the most challenging claim of mehta and andrei is that the current is a non - monotonic function of the strength of the coulomb coupling between the electron on the dot and conduction electrons in the two leads . in order to make the comparison more explicite
we generalize the time - ordered scattering formalism for non - equilibrium in the next leading logarithmic order . in this way the current is calculated as a function of the applied voltage and the coulomb coupling strength .
increasing the coulomb coupling strength we find also a non - monotonic feature but the order of increasing and decreasing regions is the opposite to the finding of mehta and andrei@xcite .
the model to be treated is the following : a single impurity orbital is coupled to two reservoirs of fermi gas via hopping but the two reservoirs have different chemical potentials @xmath0 and @xmath1 on left and right of the impurity in a one dimensional model .
@xmath2 is determined by the applied voltage @xmath3 ( @xmath4 is the electronic charge ) .
the coulomb interaction acts between the electron on the impurity level and the conduction electrons at the impurity position .
thus the hamiltonian has the form @xmath5 with @xmath6 where @xmath7 and @xmath8 , @xmath9 is the fermi velocity , @xmath10 is the creation operator of the spinless fermion in lead @xmath11 , while @xmath12 is the energy of the local level and @xmath13 is the creation operator for the electron on that site .
the interaction term is @xmath14 where @xmath15 is the coulomb coupling which in a physical case @xmath16 , @xmath17 , and @xmath18 is the length of the chain .
the existence of the substraction of @xmath19 is not essential , they can be omitted and than @xmath12 is shifted as @xmath20 and a local potential @xmath21 is acting on the electrons , but the latter one can be taken into account by changing the electron density of states in the leads at the position of the impurity .
the hybridization between the lead electrons and the localized electron is described by @xmath22 where @xmath23 is the hybridization matrix element . in case of equilibrium
it is useful to generalize the model to @xmath24 reservoirs instead of @xmath18 , @xmath25 , and then @xmath26 runs through @xmath27 and @xmath28 . then the hybridization term in @xmath29
is chosen in a specific form @xmath30 indicating that only the electrons with @xmath31 are hybridizing while the others are taking part only in the coulomb screening .
namely , only those electrons are hybridizing which have the symmetry of the localized orbital ( @xmath32-like ) . as a result of the screening the electron gas
is polarized depending on the occupation of the localized state and that polarizations lead to orthogonality catastrophe@xcite . the model with @xmath33
is known as a resonant level model and has been studied in great detail@xcite and the one with @xmath34 has been introduced to study finite range interaction in 3d@xcite . the goal of the present paper is to provide weak coupling results for @xmath35 .
but before doing that the @xmath36 equilibrium case is studied in the weak coupling limit by diagram technique and then to extend the results for stronger couplings wilson s numerical renormalization group ( nrg)@xcite and anderson - yuval coulomb gas method@xcite in order to check the validity of weak coupling results concerning a specific behavior .
namely at some stage of the calculation in the exponent of the renormalized quantities a combination @xmath37 appears . for @xmath16 that is changing sign at @xmath38 and
that leads in changing the increasing to decreasing behavior but that crossover is well beyond the validity of the perturbation theory at least for @xmath39 . in order to judge the real situation ,
an nrg study will be performed including the weak ( @xmath40 ) as well as strong coupling regions ( @xmath41 ) to get an insight whether the crossover indicated above is expected or it is likely artifact of the weak coupling theory .
we also map the problem to a one - dimensional coulomb model closely following the work of anderson and yuval , where the screening can even be in the strong coupling limit .
all of these methods are suggesting a coherent picture of the crossover and they agree very well especially for @xmath42 .
the study of such a crossover is especially relevant as in the work of mehta and andrei@xcite such a crossover is suggested in the current flowing in the non - equilibrium case @xmath35 at @xmath43 . if we could find the crossover already in equilibrium then it is obvious to expect that in the non - equilibrium situation . the paper is organized in the following way : in section [ sec : weak_coup ] we provide the analytical perturbative method up to next to leading logarithmic order introducing extra channels for screening , where the non - monotonic competion of the vertex and self - energy correction is already demonstrated in equilibrium . in section [ sec : nrg ] the equilibrium calculation is extended to strong coupling by using wilson s numerical renormalization group technique and the result is compared to that of the analytical calculation . in section [ app : karcsi ]
the anderson - yuval method is presented . in section [ sec : out_of_equ ] the time dependent scattering method is applied for non - equilibrium closely following the generalized version of anderson s poor man s scaling in the next to leading order and the current is calculated . in section [ sec : summary ]
the results are summarized . in appendix a some cancellation due to ward indetities are discussed .
the resonant level model is given by eqs .
( [ hs ] ) , ( [ h_0 ] ) and ( [ h_2 ] ) . that does not contain non - commuting terms , thus kondo behavior
is not expected in the weak coupling limit . in the strong coupling limit that model , however , can be mapped to an anisotropic kondo model@xcite but that mapping is not considered here .
the model shows strong similarities to the x - ray absorption@xcite , as the strength of the interaction ( invariant charge ) between conduction electrons and the electron on the impurity level is scale invariant .
the system shows scaling in terms of reduction of the conduction electron band width , @xmath44 . in case @xmath33 the scaling equations
were derived by schlottmann@xcite and those can be easily extended for arbitrary @xmath24 . ) and crosses ( @xmath3 ) . in case of conduction electrons
the channel indices are also indicated . ]
there are two different kinds of vertex corrections depicted in fig.[fig : vertex_corr ] a , b , c , where the solid lines stand for conduction electrons , the dotted line for electron on the impurity level and the interactions are indicated by dots ( @xmath15 ) and crosses ( @xmath3 ) . in case of conduction electrons the channel indices are also indicated .
the hartree - fock energy shift can be incorporated by @xmath12 .
the self - energy of the electron on the impurity is depicted in fig.[fig : vertex_corr ] d. in the calculation of the self - energy counter terms are introduced to eliminate the constant terms to keep @xmath45 unrenormalized . closely following the earlier works @xcite , the invariant charge for the coulomb interaction takes the form @xmath46 where @xmath47 is the vertex function and @xmath48 can be determined perturbatively starting with 1 where @xmath49 is the renormalized one - electron green s function .
the functions @xmath50 and @xmath51 are @xmath52 where @xmath53 is the conduction electron density of states , for spinless electrons in one of the channels @xmath54 .
as the coulomb interaction is independent of @xmath55 , thus the factor @xmath24 occurs . as the consequence of the ward identity relating the vertex correction and the self - energy depicted in fig.[fig : vertex_corr]c and d cancel out in @xmath56@xcite @xmath57 thus @xmath58 .
the renormalization group gives @xmath59 for arbitrary @xmath55 .
the hybridization contains vertex correction for @xmath60 ( fig.[fig : v_corr ] ) @xmath61 but it does not contain linear contribution in @xmath62 in the order of @xmath63 . the relevant invariant charge @xmath64 is @xmath65 as the interaction is connected by only one impurity line .
thus the terms linear in @xmath62 are @xmath66 the result of the renormalization equation is @xmath67 the second term in the exponent appears as reduction of @xmath64 describing the coulomb screening in the @xmath24 channels . for @xmath68
the @xmath64 interaction is always decreasing as @xmath69 is reduced but for @xmath16 it depends on the strength of @xmath15 .
for large enough @xmath15 the screening dominates thus @xmath70 that behavior will be further discussed in sec.[sec : summary ] . the scaling regions , however , are not unlimited as the impurity level has its own width , @xmath71 .
there are two contributions to the level width @xmath71 .
the hybridization broadens the impurity level just in case of the anderson model and that is in the second order in @xmath3 @xmath72 where also the effect of renormalization is taken into account .
there is also a korringa like broadening due to the creation of electron - hole pairs thus @xmath73 , where @xmath69 comes from the phase space .
that is important only for large @xmath69 where everything is smooth thus the broadening is not effective .
the broadening due to the hybridization cuts off the renormalization procedure at energy @xmath74 .
combining eqs.([eq : v_scaling ] ) and ( [ eq : gamma ] ) and inserting the condition given above provides the final @xmath75 value as @xmath76 for different channel numbers .
the intermediate maximum can diverge only for @xmath77 , in all the other cases the increases are also rather moderate . ] as shown in fig.[fig : v_ren ] , for @xmath68 @xmath78 renormalizes downwards and for even more negative @xmath15 down to zero . for @xmath16 and @xmath79
first @xmath75 increases with increasing @xmath15 but for @xmath80 it starts to decrease and tends to zero again .
the intermediate maximum appears at @xmath81 . for @xmath39
that maximum is , however , already outside the weak coupling limit , where the calculation can not be trusted .
the question still remains unanswered whether the non - monotonic behavior for @xmath16 can be traced in strong coupling calculation or not .
the conclusions for the crossover might be trusted only for large @xmath82 , which does not have physical relevance .
in order to determine the region of validity of the weak coupling approach in equilibrium , we have performed numerical renormalization group@xcite ( nrg ) analysis for the @xmath39 and @xmath42 cases . in wilson s nrg technique after
the logarithmic discretization of the conduction band one maps the original hamiltonian of an impurity problem to a semi - infinite chain with the impurity at the end of the chain . as a consequence of the logarithmic discretization
the hopping amplitude along the chain decreases exponentially as @xmath83 where @xmath84 is a discretization parameter ( we have used @xmath85 throughout the calculations ) while @xmath55 is the site index .
the separation of energy scales provided by the exponentially vanishing hopping amplitude allows us to diagonalize the hamiltonian iteratively to approximate the ground state and the excitation spectrum of the full chain .
since we know the eigenenergies and eigenvectors of the hamiltonian , we can calculate dynamical quantities such as the density of states using the lehman representation of the spectral function@xcite .
first let us focus on the physically relevant case of @xmath16 and @xmath39 . to compare the numerical data with the weak coupling results
, we have calculated the impurity density of states for different values of the interaction strength @xmath15 .
the results are shown in fig .
[ fig : sp_func ] .
the numerical data validates the weak coupling results for @xmath86 . in our nrg calculation
we considered a flat band with constant density of states @xmath87 , where @xmath44 stands for the half bandwidth . in the lower panel of fig .
[ fig : sp_func ] the renormalized value of the hybridization , @xmath75 is shown as a function of the interaction strength @xmath15 . in nrg calculation , we have defined @xmath75 from the finite size spectrum directly .
the finite size spectrum as a function of iteration number crosses over from the initial fixed point to the strong coupling one characterized by single particle phase shifts @xmath88 at around @xmath89 .
@xmath89 is determined by the renormalized hybridization , @xmath90 .
we take @xmath91 when the energy of the first excited state exceeds @xmath92 of its fixed point value . and @xmath93 as obtained by perturbative rg and wilson s nrg .
the lower panel shows the renormalized value of the hybridization , @xmath75 as a function of the interaction strength @xmath15 .
the numerical data supports the weak coupling results for @xmath94 . ] to answer the question whether an intermediate maximum appears outside the weak coupling limit or not , we have performed calculation with very large values of the interaction strength up to @xmath95 .
the results are shown in fig .
[ fig : v_ren_n2n4 ] .
our conclusion is that even for @xmath39 such a non - monotonic behavior is found but the position of the maximum as well as the shape of the curve for large @xmath15 differs essentially from those obtained by weak coupling calculation .
it still remains a question whether for case of many channels the weak coupling calculation is reliable or not . to treat many channels with nrg
is very challenging but to see the tendency with increasing channel number , we performed the numerical analysis of the case @xmath42 .
the results are plotted in fig .
[ fig : v_ren_n2n4 ] as well .
our data suggests that already for @xmath42 the position of the turning point as well as the decay of the curve at large @xmath15 is reproduced by the weak coupling calculation with a much better accuracy than in case of @xmath39 . 0.3 cm as a function of the interaction strength @xmath15 for @xmath39 and @xmath42 . ,
title="fig : " ] 0.3 cm
in most of the physical cases the coulomb interaction @xmath15 dominates over the hopping term @xmath96 . to overcome that difficulty , anderson and yuval@xcite introduced a path integral method for the kondo problem where the interaction @xmath15 is described in terms of phase shifts while the hopping is treated as perturbation .
the similarity between the kondo and the present problem can be exploited in the following way : the complex time axis is divided into intervals and as it is shown in fig . [
fig : time_intvals ] where the solid lines represents the time interval when the impurity level is occupied and the light ones stand for unoccupied level .
the conduction electrons can join the time line at the end points of the intervals where hopping @xmath96 takes place while they can touch the time line at any other points due to the coulomb interaction .
those are indicated by dashed lines which are labeled according to the channel indices @xmath97 . thus the incoming and outgoing conduction electron lines should be connected all possible ways and finally a summation over all possible configuration of channel labellings @xmath97 must be carried out .
0.5 cm takes place while they can touch the time line at any points due to the coulomb interaction .
those are indicated by dashed lines which are labeled according to the channel indices @xmath97 .
, title="fig : " ] 0.5 cm some of the connections are indicated in fig .
[ fig : time_intvals ] by dotted lines .
the final result for the partition function can be given in analytical form as in refs.@xcite where the kondo or the two level system problems were treated .
the partition function has the form of a one dimensional coulomb gas with appropriately defined vector charges @xmath98 ( @xmath99 ) .
the scaling equations are derived by eliminating short time intervals at the short time cutoff @xmath100 with its initial value taken as the inverse bandwidth @xmath101 .
the interaction @xmath96 must be also made dimensionless by a factor @xmath102 .
the phase shifts for electrons in case of filled and empty impurity levels are : @xmath103 and @xmath104 , respectively .
only their difference will appear in the scaling : @xmath105 .
the phase shifts are limited : @xmath106 , @xmath107 .
the friedel sum rule requires @xmath108 which expresses that the one electron difference between the filled and empty sites must be screened by charge oscillations formed by the conduction electrons .
the hybridization can be associated with the fugacity @xmath109 the interaction can be represented by charges at the interaction points as @xmath110 where the index @xmath111 labels hopping in and out of the impurity level assisted by an electron annihilation or creation in channel 0 .
the lengthy derivation of the scaling equations closely follows refs . ) and the final result is @xmath112 or expressed in terms of phase @xmath113 ( phase shifts ) they are @xmath114 here the scaling is carried out by increasing @xmath100 to reduce the electronic bandwidth .
the term @xmath115 in the first line of eq.([eq : sca_eqs_yz ] ) originates in the explicite factor @xmath116 in the definition of @xmath117 ( [ eq : fugacity ] ) and disappears from the corresponding scaling equation of @xmath3 .
the system of equations eqs .
( [ eq : sca_eqs_yz ] ) must be solved for initial values @xmath118 but @xmath119 can be arbitrary .
the fugacity can either increase or decrease exponentially depending on the quantity @xmath120 .
similar expressions were obtained in ref . by matching the perturbative results with expression in terms of phase shifts .
the regions for attractive interaction and large enough repulsive one will be treated separately . in the first case @xmath121 ( @xmath68,@xmath122 ) .
the solution is @xmath123 thus @xmath117 is decreasing , therefore @xmath119 ( @xmath124 ) are slowly varying thus the @xmath100 dependence can be ignored .
thus @xmath125 the situation is different for repulsive interaction ( @xmath126 ) .
there are two regions in that case . in the first one @xmath127 and @xmath117 is increasing .
then eq .
( [ eq : vscal_yuval ] ) is valid as far as @xmath128 is satisfied
. there is , however , a crossover where @xmath129 to the second region where @xmath117 decreases again and the screening dominates .
the larger @xmath24 the stronger the decrease is .
the crossover between the increasing and decreasing regions is at @xmath130 as @xmath119-s are very slightly renormalized thus the unrenormalized values can be used and @xmath131 is independent of @xmath26 ( @xmath132 ) .
thus the crossover is at @xmath133 and then for @xmath134 comparing with the results of nrg in case @xmath39 the turning point is at @xmath135 thus the agreement is not complete but at least it could be argued that it is inside the accuracy of the scaling equation .
the weak coupling scaling result is very poor as it was expected ( see fig.[fig : sp_func ] ) . for @xmath42
all the methods give very similar results .
the general solution of the scaling equations can be searched in the form @xmath136 . then @xmath137 the scaling trajectories are @xmath138 during the renormalization @xmath117 is fast increasing and the scaling is stopped where it reaches unity
. meanwhile @xmath139 decreases slowly and its renormalized value can be extracted from eq.([eq : trajectory ] ) .
the result in leading order in @xmath117 reads as @xmath140 outside that region the long time approximation for the conduction electron green s function can not be applied .
considering theoretical methods two ways can be followed : the keldysh green s function method or the calculation of the scattering amplitude by time - ordered perturbation theory . here
the second method will be followed , where the initial conduction electron states can be arbitrary non - equilibrium states and for the intermediate and final states the actual non - equilibrium distributions are taken into account .
that method has been earlier applied in the leading logarithmic approximation@xcite , which is a generalization of anderson s poor man scaling@xcite . here
the extension of that method is presented to next leading order . for equilibrium first the kondo model was treated by that way @xcite .
the basic idea is to calculate the development of the initial @xmath141 state to the final @xmath142 one , but in second order the renormalization of the norms of those states must be corrected also .
thus the scattering matrix element to be considered is @xmath143 t { [ \langle f| % 1+[1/(\omega - h_0)]t \sum\limits_{n=0}^{\infty}\left(\frac{1}{\omega - h_0}h_{\rm int}\right)^n \langle i| %
1+[1/(\omega - h_0)]t \sum\limits_{n=0}^{\infty}\left(\frac{1}{\omega - h_0}h_{\rm int}\right)^n \label{eq : state_norm}\ ] ] where @xmath69 is the initial energy of state @xmath141 and the hamiltonian is split as @xmath144 . in the present problem
the correction of the normalization occurs for the impurity electron states while in the kondo model for the spin states .
those normalizations appeared in the previous treatment as @xmath145 in eq.([eq : v_inv ] ) .
the scaling equation can be obtained after changing the cutoff @xmath146 by adjusting the coupling constant to keep @xmath147 invariant for appropriate @xmath141 and @xmath142 . in the following
we use the original left and right states ( @xmath148 ) with @xmath149 as a start . in order to derive the scaling for coupling @xmath15
the initial and final states should be @xmath150 where @xmath151 is the non - equilibrium state at applied voltage @xmath152 , for which disregarding @xmath3 the state is the noninteracting ground state .
considering the occupation @xmath153 of the @xmath154-level the occupation probability in the steady state must be determined in presence of @xmath152 . that value will be @xmath155 for @xmath45 but in the general case
@xmath156 it can depend on @xmath152 .
the diagrams of the numerator up to @xmath157 order are shown in fig.[fig : ucorr ] ( i)-(vi ) where the diagrams should be decorated by the direction of the lines in all possible ways . the diagram of the self - energy is shown in fig.[fig : ucorr](vii ) .
order contributing to the numerator of eq.([eq : state_norm ] ) .
the diagrams should be decorated by the direction of the lines in all possible ways .
the diagram of the self - energy is shown in panel ( vii ) . ] in logarithmic approximation only the diagrams linear in @xmath158 are contributing to the scaling equations thus the relevant vertex corrections are ( ii),(v ) and ( vi ) while ( iii ) and ( iv ) are not as these provide @xmath159 .
the type ( ii ) diagrams with the parallel and antiparallel lines cancel each other .
as the logarithmic terms in diagrams ( v),(vi ) and ( vii ) come from closed electron loops which are independent of the actual value of @xmath0 and @xmath1 thus to the logarithmic term the left and right contacts contribute separately which should be independent of the applied voltage .
that simplification is not sustained in higher order contributions where the left and right lines simultaneously occur . the self - energy correction in ( vi ) contributes by adding that to either of the incoming and outgoing @xmath154-lines .
one of those corrections is cancelled by the denominator in eq.([eq : state_norm ] ) .
as it is well known from the spinless fermionic case e.g. the x - ray absorption problem @xcite the remaining diagram is cancelled by ( v ) .
thus single logarithmic term does not remain .
that is similar to the equilibrium case , see eq.([eq : uinv_noscaling ] ) thus the invariant coupling @xmath160 .
( for the details see appendix a. ) in the following the renormalization of the hybridization depicted in fig.[fig : vertex_corr]b and [ fig : v_corr ] is crucial where e.g. @xmath161 and @xmath162 .
keeping terms up to @xmath163 .
after taking the denominator in eq.([eq : state_norm ] ) into account the final form of @xmath147 is @xmath164 where the first correction is due to the vertex depicted in fig.[fig : v_corr ] , while the second one arises from the self - energy on the leg of diagram reduced by the denominator by a factor 1/2 .
this result agrees with eq.([eq : v_corr2 ] ) ( @xmath39 ) at @xmath165 . here
taking the special case @xmath45 the voltage @xmath152 serves as a low energy cutoff .
as it has been mentioned earlier considering the @xmath154-level there is a steady state occupation @xmath153 .
that value is determined from the ballance of the in and outflow of the conduction electrons . to determine it for @xmath156 two other quantities
must be known , namely the changes in the level position and the spectral function of the @xmath154-level , @xmath166 and @xmath167 due to the applied voltage .
that calculation can be carried out numerically in a self - consistent way .
the probability of scattering of an electron coming from the left ( l ) or right ( r ) into the @xmath154-level is denoted by @xmath168 while the opposite process by @xmath169 .
these quantities are @xmath170 and @xmath171 where @xmath172 are determined from renormalization group equations with the appropriate infrared cutoffs and @xmath173 is the fermi distribution function for the leads in presence of the voltage .
those will be taken at zero temperature @xmath174 .
the steady state is determined by @xmath175 this equation combined with eqs.([eq : w+])([eq : w- ] ) gives @xmath176d\varepsilon } { \int\varrho_d(\varepsilon)\left[v_{l}^2(\varepsilon)+ v_{r}^2(\varepsilon)\right]d\varepsilon}\;. \label{eq : nd}\ ] ] if electron - hole symmetry holds thus @xmath45 , then @xmath177 .
the next step is to determine the self - energy of the @xmath154-electron .
the @xmath154-electron propagator is @xmath178 which can be simply developed because the occupied @xmath154-level determines the time flow .
the self - energy corrections appear also in the normalization .
the effect of hybridization is just to give an extra broadening of the @xmath154-level to be considered later . without hybridization
the self - energy is @xmath179 which can be evaluated as @xmath180 and @xmath181 in the equilibrium calculation the term proportional to @xmath182 is contributing to the function @xmath51 in eqs.([eq : gamma_d ] ) and the last constant term @xmath183 is eliminated by the applied counter term to keep @xmath12 unrenormalized while @xmath184 is a korringa type of relaxation . to be noted that the voltage does not occur as the energy goes directly into the electron - hole creation of the same electrode . as we mentioned at the end of sec.[sec : weak_coup ] , that broadening is less important at small energies .
the hybridization of the @xmath154-level gives the essential part of the broadening just like in the anderson impurity model @xmath185 where the voltage dependent hybridization strength must be used .
the the @xmath154-electron spectral function is @xmath186 with the help of these quantities we are ready to calculate the current through th impurity : @xmath187 combining this equation with the expression of @xmath188 given in eqs.([eq : w+])([eq : w- ] ) the current takes the form @xmath189 \varrho_d(\varepsilon)d\varepsilon\nonumber\\ -2\pi\varrho_0\int\left[f(\varepsilon - ev/2)v_r^2(\varepsilon)-f(\varepsilon+ev/2 ) v_l^2(\varepsilon)\right ] \varrho_d(\varepsilon)d\varepsilon\;. \label{eq : curr2}\ ] ] the numerical calculation goes as follows : for a fixed value of @xmath152 we discretize the energy interval @xmath190 $ ] and calculate the renormalized hybridization @xmath191 and the impurity self - energy .
the latter is evaluated in such a way that the renormalized @xmath154-level position zero @xmath192 . by performing a sum over @xmath193 we cab
calculate at the level occupation @xmath194 and the current @xmath195 for the given value of the voltage .
the result is shown in fig.[fig : current ] for @xmath196 . 0.5 cm , and @xmath197 .
for this range of interaction strength the weak coupling method was reliable in equilibrium . ,
title="fig : " ] 0.5 cm 0.5 cm , and @xmath198 .
note that the weak coupling result is reliable for @xmath94 only .
, title="fig : " ] 0.5 cm in that regime of @xmath15 the weak coupling rg gave good results in equilibrium ( presented in sec .
[ sec : nrg ] ) therefore one expects reliable results in out - of - equilibrium as well . as it is shown in fig.[fig : current ] , the current increases with increasing interaction strength .
the results can be interpreted as follows : with increasing @xmath15 the impurity spectral function gets broadened since the hybridization is enhanced . in the linear regime ( @xmath199 ) the @xmath15-dependence drops out since the increase of coupling to the leads @xmath200 is cancelled by the decrease of the height of the @xmath154-level spectral function which scale as @xmath201 . for larger values of the voltage @xmath202 the @xmath154-level
is experienced not only at the peak of the spectral function ( @xmath203 ) thus the current is not linear in @xmath152 any more . for @xmath204
the complete @xmath154-level contributes to the current thus it saturates and the asymptotic value is proportional to @xmath200 .
the resonant level model studied has very different behavior for attractive and repulsive interaction .
this difference can be understood using the site representation for conduction electrons in the strong @xmath15 limit by the following argument : \(i ) _ in case of attractive interaction _ the particle on the @xmath154-level attracts electrons to pile up around the impurity occupying the next site thus the electron on the @xmath154-level is blocked for hopping to the conduction band .
\(ii ) _ in case of repulsive interaction _ the site next to the occupied @xmath154-level is empty thus that electron can easily jump to the conduction band .
all the methods predict that increasing the strength of an attractive interaction the hopping rate @xmath3 is reduced and for strong enough coupling it even goes to zero ( see fig.[fig : v_ren_n2n4 ] for @xmath205 ) .
the effect of orthogonality catastrophe reducing the hopping is less relevant because that have been already reduced by the vertex correction . in the repulsive case for large @xmath15 the orthogonality catastrophe ( self - energy correction ) is reducing essentially the hopping rate .
thus the effective hopping @xmath206 is first enhanced by the effect described above and can reach a maximum which is followed by a reduction due to the orthogonality catastrophe .
the position of the maximum can be pushed to lower energies by increasing the number of the screening channels @xmath24 thus the perturbative result becomes more and more reliable . in case of @xmath39 the latter method leads to a pronounced maximum but the nrg indicates only slowly varying bump . in the anderson - yuval approach the maximum is even pushed to infinity .
thus the existence of the maximum is supported only by the nrg . considering the time ordered scattering formalism the results are in accordance with the expectation of the weak coupling result for @xmath39 .
the increasing @xmath15 results in increasing current as first @xmath3 is increased but for larger @xmath15 due to the orthogonality catastrophe the current is essentially reduced ( see fig.[fig : current_decrease ] ) . as the nrg does not give a sharp crossover for the hopping rate @xmath206 thus the corresponding effect in the current must be less pronounced in reality .
of course , for @xmath207 the crossover must clearly exist .
unfortunately , in the work of mehta and andrei the crossover is in the range of reduced current , which is just the opposite what is expected on the ground of the physical picture established and results obtained by different methods for the hopping rate .
very detailed further studies of the bethe ansatz method are needed to understand and resolve the origin of the presented discrepancies .
the authors acknowledge the fruitful discussions with n. andrei , p. mehta , j. von delft , j. kroha , p. wlfle , p. schmitteckert and g. zarnd .
this work was supported by projects otka d048665 , t048782 , ts049881 and t046303 .
a.z . is grateful to the humboldt foundation to support his stay in munich .
l.b . acknowledges the support of the jnos bolyai scholarship .
contributing to the numerator of eq.([eq : state_norm ] ) .
a ward identity ensures the cancellation of these diagrams for small frequencies meaning that no logarithmic term survives in second order thus the invariant coupling @xmath160 . ] the diagrams of the numerator of eq.([eq : state_norm ] ) up to @xmath157 order are shown in fig.[fig : ucorr ] ( i)-(vi ) and the diagram of the self - energy is shown in fig.[fig : ucorr](vii ) .
as noted earlier , the logarithmic terms come from diagrams ( v),(vi ) and ( vii ) .
note that the self - energy correction in ( vi ) ( see also fig.[fig : app]b ) contributes by adding that to either of the incoming or outgoing @xmath154-lines .
one of those corrections is cancelled by the diagram shown in fig.[fig : ucorr](vii ) ) in the denominator in eq.([eq : state_norm ] ) . therefore the two relevant diagrams are those depicted in fig.[fig : app ] .
the contribution of fig.[fig : app]b can be written as on the other hand , the contribution of fig.[fig : app]a for small frequencies can be evaluated and the @xmath154-electron lines indicated by arrows occur twice just like the denominator squared in eq.([eq : appdiag2 ] ) .
that ward identity ensures the cancellation of diagram shown in fig.[fig : ucorr](v ) by one of the leg one in fig.[fig : ucorr](vi ) .
this means that no logarithmic term survives in second order thus the invariant coupling @xmath160 .
99 see e.g. l.i .
glazman , m. pustilnik , in `` nanophysics : coherence and transport , '' eds .
bouchiat et al .
( elsevier , 2005 ) , p. 427 .
see e.g. n.o .
birge and f. pierre , in `` fundamental problems of mesoscopic physics , interactions and decoherence '' edited by i.v .
lerner b.l .
altshuler and y. gefen ( kluwer academic , dortrecht 2004 ) p. 3 .
p. mehta and n. andrei , phys .
lett . * 96 * , 216802 ( 2006 ) .
anderson , phys .
lett . * 18 * , 1049 ( 1967 ) .
see e.g. p.b .
vigman and a.m. finkelstein , zh .
. fiz . * 75 * , 204 ( 1978 ) ( sov .
jetp * 48 * , 102 ( 1978 ) ) .
p. schlottmann , phys .
b * 25 * , 4815 ( 1982 ) .
t. giamarchi , c.m .
varma , a.e .
ruckenstein and p. nozires , phys .
rev . lett . * 70 * , 3967 ( 1993 ) .
wilson , rev .
* 47 * , 773 ( 1975 ) ; t. costi , in _ density matrix renormalization _ , edited by i. peschel _
( springer 1999 ) .
g. yuval and p.w .
anderson , phys .
b * 1 * , 1522 ( 1970 ) .
b. roulet , f. gavoret , p. nozires , phys . rev . * 178 * , 1072 ( 1969 ) .
p. nozires and c.t .
de dominicis , phys . rev . * 178 * , 1097 ( 1969 ) . m. fowler and a. zawadowski , solid state commun . *
9 * , 471 ( 1970 ) and a.a .
abrikosov and a.a .
migdal , j. low temp . phys . * 3 * , 319 ( 1970 ) .
j. slyom , j. phys f : met .
* 4 * , 2269 ( 1974 ) .
k. vladr , a. zawadowski , and g.t .
zimnyi , phys .
b * 37 * 2001 ( 1988);_ibid _ * 37 * 2015 ( 1988 ) .
k. vladr , phys .
b * 44 * 1019 ( 1991 ) .
a. zawadowski in _ electronic correlations : from meso- to nano - physics _ , edited by t. martin , g. montambaux and j. trn thanh vn ( edp sciences , paris , 2001 ) p389 .
a. rosch , j. paaske , j. kroha , and p. wlfle , phys .
90 * , 076804 ( 2003 ) ; j. paaske , a. rosch , and p. wlfle , phys .
b * 69 * , 155330 ( 2004 ) .
o. jsghy , a. jakovc , and a. zawadowski , phys .
b * 72 * , 205119 ( 2005 )
anderson , j. phys .
c * 3 * , 2436 ( 1970 ) . j. slyom and a. zawadowski , j. phys .
* 4 * , 80 ( 1974 ) . | the spinless resonant level model is studied when it is coupled by hopping to one of the arbitrary number of conduction electron channels .
the coulomb interaction acts between the electron on the impurity and in the different channels . in case of repulsive or attractive interaction the conduction electrons
are pushed away or attracted to ease or hinder the hopping by creating unoccupied or occupied states , respectively . in the screening of the hopping orthogonality catastrophe
plays an important role . in equilibrium
the weak and strong coupling limit the renormalizations are treated by perturbative , numerical and anderson - yuval coulomb gas methods . in case of two leads the current due to applied voltage is treated in the weak coupling limit .
the presented detailed study should help to test other methods suggested for non - equilibrium transport . |
thin covers from another material are often applied in engineering to improve the functional properties of the surfaces of components of machines and structures . on the other hand ,
thin covers with certain mechanical properties are used for modeling of real microstructure of the surfaces , adhesion and glue bondings @xcite .
the classical methods for solution of contact problems for bodies with thin covers are grounded on integral equations and are reviewed in work @xcite .
nowadays , one of the most effective numerical methods for such contact problems are methods , based on variational formulations and finite element approximations .
efficient approach for solution of multibody contact problems is the use of domain decomposition methods ( ddms ) .
many ddms for contact problems without covers are obtained on discrete level @xcite . among ddms , proposed on continuous level for contact problems without covers
are methods presented in @xcite .
domain decomposition methods for solution of problem of ideal contact between two bodies , connected through nonlinear winkler layer are proposed in @xcite .
these methods are based on saddle - point formulation and conjugate gradient methods . in current contribution
we consider the problem of unilateral contact between bodies with nonlinear winkler covers .
we give variational formulations of this problem in the form of nonlinear variational inequality on convex set and variational equation in the whole space , and present theorems about existence and uniqueness of their solution .
furthermore , we propose on continuous level a class of parallel domain decomposition methods for solving the nonlinear variational equation , which corresponds to original contact problem . in each iteration of these methods we have to solve in a parallel way linear variational equations in separate bodies , which are equivalent in a weak sense to linear elasticity problems with robin boundary conditions on possible contact areas .
these ddms are based on abstract nonstationary iterative methods for variational equations in banach spaces .
they are the generalization of domain decomposition methods , proposed by us earlier in @xcite for unilateral contact problems without covers .
some particular cases of proposed ddms can be viewed as a modification of semismooth newton method @xcite .
the numerical analysis of obtained ddms is made for plane contact problems using finite element approximations .
consider a unilateral contact of @xmath0 elastic bodies @xmath1 with piecewise smooth boundaries @xmath2 , @xmath3 ( fig .
suppose that across each contact surface there is a nonlinear winkler layer .
denote @xmath4 .
a stress - strain state in point @xmath5 of each solid @xmath6 is described by the displacement vector @xmath7 , the tensor of strains @xmath8 and the tensor of stresses @xmath9 .
these quantities satisfy the following relations : @xmath10 @xmath11 @xmath12 where @xmath13 are the components of volume forces vector @xmath14 , and @xmath15 are symmetric elasticity constants , which are bounded in the following sense : @xmath16 introduce on boundary @xmath2 a local orthonormal coordinate system @xmath17 , where @xmath18 is an outer unit normal .
then the vectors of displacements and stresses on @xmath19 can be written in the following way : @xmath20 suppose , that the boundary @xmath21 consists of three disjoint parts : + @xmath22 , @xmath23 , @xmath24 , @xmath25 . on the part @xmath26 homogenous dirichlet boundary conditions are prescribed , and on the part @xmath27 we consider neumann boundary conditions : @xmath28 the part @xmath29 , @xmath30 is the possible contact area of body @xmath31 with the other bodies . here
@xmath32 is the possible unilateral contact area of body @xmath33 with body @xmath34 , and @xmath35 is the set of the indices of all bodies in contact with body @xmath33 .
we assume that the surfaces @xmath36 and @xmath37 are sufficiently close ( @xmath38 ) , and @xmath39 , @xmath40 , @xmath41 , where @xmath42 is the projection of point @xmath43 on @xmath32 .
let @xmath44 be a distance between bodies @xmath33 and @xmath34 before the deformation .
we suppose that possible contact areas @xmath45 and @xmath46 , @xmath47 , @xmath48 have nonlinear winkler covers .
total compression @xmath49 of these covers is related with normal contact stress as follows : @xmath50 , @xmath51 , @xmath52 , where @xmath53 is given nonlinear continuous function , which satisfy the next conditions : @xmath54 @xmath55 on possible contact zones @xmath56 , @xmath57 , @xmath3 we consider the following unilateral contact conditions through nonlinear winkler layers : @xmath58 @xmath59 @xmath60 @xmath61\sigma _
{ \alpha n } ( { \bf x})=0,\,{\bf x}'=p({\bf x}),\,{\bf x}\in s_{\alpha \beta } .\ ] ]
for each body @xmath33 consider sobolev space @xmath62^{3 } $ ] and the closed subspace @xmath63 .
all values of the elements from these spaces on the parts of boundary @xmath2 should be understood as traces .
the trace of element @xmath64 on the part @xmath65 should belong to space @xmath66^{3}$ ] , and the trace of element from @xmath67 on the part @xmath68 should belong to @xmath69^{3}$ ] .
define hilbert space @xmath70 with scalar product + @xmath71 and norm @xmath72 , @xmath73 .
moreover , introduce following spaces @xmath74 and @xmath75 , where @xmath76 . in space
@xmath77 consider the closed convex set of all displacements , which satisfy nonpenentration contact conditions : @xmath78 where @xmath79 , @xmath80 .
let us introduce bilinear form @xmath81 , such that @xmath82 represents the total elastic deformation energy of the bodies , linear form @xmath83 , which is equal to external forces work , and nonquadratic functional @xmath84 , which represents the total deformation energy of nonlinear winkler layers : @xmath85 @xmath86 @xmath87 ds\,,\,\,\,\,{\bf u},{\bf v } \in v_{0}\,,\,\,\,\,{\bf w } \in w\,,\ ] ] where @xmath88^{3}$ ] , @xmath89^{3}$ ] , @xmath90 .
we have shown , that if condition ( [ prokopyshyni_contrib__3 _ ] ) holds then bilinear form @xmath91 is symmetric , continuous and coercive , and nonquadratic functional @xmath92 is gateaux differentiable : @xmath93 * theorem 1 . * _ suppose that conditions ( [ prokopyshyni_contrib__3 _ ] ) , ( [ prokopyshyni_contrib__6 _ ] ) , ( [ prokopyshyni_contrib__7 _ ] ) hold
. then problem ( [ prokopyshyni_contrib__1_])([prokopyshyni_contrib__2 _ ] ) , ( [ prokopyshyni_contrib__4 _ ] ) , ( [ prokopyshyni_contrib__8_])([prokopyshyni_contrib__11 _ ] ) has an alternative weak formulation as the following minimization problem : _
@xmath94 _ moreover , there exists a unique solution of problem ( [ prokopyshyni_contrib__17 _ ] ) , and this problem is equivalent to the following nonlinear variational inequality on set @xmath95 : _
@xmath96 except this variational formulation , we also have proposed another weak formulation of original contact problem in the form of nonlinear variational equation .
let us introduce the following nonquadratic functional in space @xmath97 : @xmath98 ds\,,\,\,\,\,{\bf u } \in v_{0}\,,\ ] ] where @xmath99 is nonlinear function .
functional @xmath100 is nonnegative and gateaux differentiable in @xmath97 : @xmath101\,ds\,.\ ] ] we have shown that if conditions ( [ prokopyshyni_contrib__6 _ ] ) and ( [ prokopyshyni_contrib__7 _ ] ) hold , then gateaux differential @xmath102 satisfies the following properties : @xmath103 @xmath104 @xmath105 these properties helped us to prove the next theorem
. * theorem 2 . * _ suppose that conditions ( [ prokopyshyni_contrib__3 _ ] ) , ( [ prokopyshyni_contrib__6 _ ] ) and ( [ prokopyshyni_contrib__7 _ ] ) hold
. then the contact problem ( [ prokopyshyni_contrib__1_])([prokopyshyni_contrib__2 _ ] ) , ( [ prokopyshyni_contrib__4 _ ] ) , ( [ prokopyshyni_contrib__8_])([prokopyshyni_contrib__11 _ ] ) is equivalent to problem ( [ prokopyshyni_contrib__1_])([prokopyshyni_contrib__2 _ ] ) , ( [ prokopyshyni_contrib__4 _ ] ) , ( [ prokopyshyni_contrib__8 _ ] ) with the following nonlinear boundary value conditions on the possible contact areas : _ @xmath106 _ and it is equivalent in weak sense to the next nonquadratic minimization problem : _
@xmath107 _ moreover , problem ( [ prokopyshyni_contrib__26 _ ] ) has a unique solution and is equivalent to the next nonlinear variational equation in space @xmath97 : _
in reflexive banach space @xmath109 consider an abstract nonlinear variational equation @xmath110 where @xmath111 is a functional , which is linear in @xmath112 , but nonlinear in @xmath113 , and @xmath114 is linear continuous form . for numerical solution of ( [ prokopyshyni_contrib__29 _ ] ) consider the next nonstationary iterative method @xcite : @xmath115 , \,\,\,\
, k=0,1 , ... \,\ , , \ ] ] where @xmath116 are some given bilinear forms , @xmath117 are iterative parameters , and @xmath118 is the _ k_-th approximation to the exact solution of problem ( [ prokopyshyni_contrib__29 _ ] ) .
* * theorem 3.**@xcite _ suppose that functional @xmath119 satisfies the following properties : _
@xmath120 @xmath121 @xmath122 _ then nonlinear variational equation ( [ prokopyshyni_contrib__29 _ ] ) has a unique solution @xmath123 .
in addition , suppose that bilinear forms _ @xmath124 , @xmath125 _ are symmetric , continuous with constant _
@xmath126 , _ coercive with constant _
@xmath127 , _ and the next conditions hold : _ @xmath128 @xmath129\ , \right\},\ ] ] _ where @xmath130 is obtained by iterative method ( [ prokopyshyni_contrib__30 _ ] ) .
now let us apply nonstationary iterative method ( [ prokopyshyni_contrib__30 _ ] ) for solving nonlinear variational equation ( [ prokopyshyni_contrib__28 _ ] ) , which corresponds to original contact problem .
this equation can be written in form ( [ prokopyshyni_contrib__29 _ ] ) , where @xmath131 , @xmath132 , @xmath133 , @xmath134 , and iterative method ( [ prokopyshyni_contrib__30 _ ] ) applied to solve ( [ prokopyshyni_contrib__28 _ ] ) rewrites as follows : @xmath135,\,k=0,1, .... \ ] ] note , that in general case iterative method ( [ prokopyshyni_contrib__37 _ ] ) does not lead to domain decomposition .
let us propose such variants of this method , which involve the domain decomposition .
at first , let us take bilinear forms @xmath124 in method ( [ prokopyshyni_contrib__37 _ ] ) as follows : @xmath136 @xmath137\left[v_{\alpha n } + v_{\beta
n } \right]ds,\ ] ] @xmath138^{-}=\{\,0,\ , d_{\alpha \beta } -u_{\alpha n}^{k } -u_{\beta n}^{k } \ge 0 \,\}\vee \{\,1,\ , \textrm{else}\,\}.\ ] ] here @xmath139 and @xmath140 are the second subdifferentials of functionals @xmath141 and @xmath142 in point @xmath143 . in the case when @xmath144 , @xmath145 , iterative method ( [ prokopyshyni_contrib__37 _ ] ) with bilinear forms ( [ prokopyshyni_contrib__38 _ ] ) corresponds to semismooth newton method for variational equation ( [ prokopyshyni_contrib__28 _ ] ) .
however , this method does not lead to domain decomposition .
now , let us take bilinear forms @xmath124 in the following way : @xmath146 @xmath147 where @xmath148 are characteristic functions of some given subsets @xmath149 of possible contact areas .
let us show , that such choice of bilinear forms @xmath124 involves the domain decomposition . introduce a notation @xmath150 .
then iterative method ( [ prokopyshyni_contrib__37 _ ] ) with bilinear forms ( [ prokopyshyni_contrib__40 _ ] ) can be written in such way : @xmath151 @xmath152 since the common quantities of the subdomains are known from the previous iteration , variational equation ( [ prokopyshyni_contrib__41 _ ] ) splits into @xmath0 separate equations in subdomains @xmath153 , and iterative method ( [ prokopyshyni_contrib__41_])([prokopyshyni_contrib__42 _ ] ) can be written in the following equivalent form : @xmath154 @xmath155 @xmath156 @xmath157 in each iteration @xmath158 of method ( [ prokopyshyni_contrib__43_])([prokopyshyni_contrib__44 _ ] ) , we have to solve @xmath0 linear variational equations ( [ prokopyshyni_contrib__43 _ ] ) in parallel , which correspond to linear elasticity problems in separate bodies @xmath31 with robin boundary conditions on possible contact areas .
therefore , this method refers to parallel robin robin type domain decomposition schemes . by taking different characteristic functions
@xmath159 , we can obtain different particular cases of domain decomposition method ( [ prokopyshyni_contrib__43_])([prokopyshyni_contrib__44 _ ] ) .
thus , taking @xmath160 @xmath161 , @xmath162 , @xmath163 , we get parallel neumann
neumann domain decomposition scheme .
other borderline case is when @xmath164 @xmath165 , @xmath162 , @xmath163 .
moreover , we can choose characteristic functions @xmath159 by formula ( [ prokopyshyni_contrib__39 _ ] ) , i.e. @xmath166 .
numerical experiments , provided by us , have shown , that such ddm has higher convergence rate than other particular domain decomposition schemes .
numerical investigations of proposed ddms have been made for plane problem of unilateral contact between two isotropic bodies @xmath167 and @xmath168 , one of which has a groove ( fig .
the bodies are uniformly loaded by normal stress with intencity @xmath169 .
each body has length @xmath170 and height @xmath171 .
the young s moduli and poisson s ratios of the bodies are the same : @xmath172 , @xmath173 .
the distance between bodies is @xmath174^{+ } \right\}^{3/2}$ ] , @xmath175 , where @xmath176 , @xmath177 , @xmath178 , @xmath179,\ , \ , x_{2 } = h\right\}$ ] . across possible contact area
@xmath180 there is a nonlinear winkler layer .
the relationship between normal contact stresses and displacements of this layer are described by the following power function : + @xmath181 , @xmath182 , where parameters @xmath183 and @xmath184 are taken from the intervals @xmath185 @xmath186\,$ ] , @xmath187 $ ] . for such choice of these parameters the nonlinear winkler layer models a roughness of the possible contact surface @xcite .
this problem has been solved by ddm ( [ prokopyshyni_contrib__43_])([prokopyshyni_contrib__44 _ ] ) with stationary iterative parameters @xmath188 , @xmath189 and characteristic functions @xmath190 , taken by formula ( [ prokopyshyni_contrib__39 _ ] ) , i.e. @xmath191 , @xmath189 . for solving linear variational problems ( [ prokopyshyni_contrib__43 _ ] ) in each iteration @xmath158
we have used finite element method with 8192 linear triangular elements for each body .
we have used the following initial guesses for displacements @xmath192 , @xmath193 and the next stopping criterion : @xmath194 @xmath195 , @xmath196 , where @xmath197^{2}}$ ] is discrete norm , @xmath198 are finite element nodes on the possible contact area , and @xmath199 is relative accuracy . at fig .
2a the relative error @xmath200 of displacement @xmath201 on different iterations @xmath158 , obtained for @xmath202 , @xmath203 , is represented for different values of parameter @xmath204 .
curves 19 correspond to @xmath205 , 0.02 , 0.05 , 0.6 , 0.8 ( 0.3 ) , 0.9 , 0.95 , 0.98 , 0.99 .
for these values of parameter @xmath204 , ddm ( [ prokopyshyni_contrib__43_])([prokopyshyni_contrib__44 _ ] ) reaches the accuracy @xmath206 in 193 , 124 , 65 , 7 , 13 , 27 , 55 , 134 iterations respectively .
thus , we conclude , that the best convergence rate reaches if @xmath207 .
the convergence rate is good if @xmath208 $ ] .
however , it becomes slow when @xmath204 is close to 0 or to 1 . for @xmath209
the method is still convergent , but the convergence becomes nonmonotone . for @xmath210
the method is not anymore convergent . at fig .
2b the normal contact stress @xmath211 , obtained by ddm ( [ prokopyshyni_contrib__43_])([prokopyshyni_contrib__44 _ ] ) for @xmath212 and different values of parameters @xmath184 is represented .
curves 14 correspond to numerical solution for @xmath213 , 0.6 , 0.8 , 1 .
dashed curve represents the analytical solution , obtained in @xcite for contact of two halfspaces without nonlinear layer . here
we conclude , that for small values of @xmath184 ( @xmath214 ) the influence of nonlinear layer on the contact behavior is not so large and the numerical solutions are close to the solution without layer .
however , for larger values of @xmath184 ( @xmath215 ) the influence of nonlinear layer becomes more significant and can not be neglected .
the positive feature of proposed domain decomposition methods are the simplicity of their algorithms .
these methods have only one iteration loop , which deals with domain decomposition , nonlinearity of winkler layers and contact conditions .
dostl , z. , kozubek , t. , vondrk , v. , brzobohat , t. , markopoulos , a. : scalable tfeti algorithm for the solution of multibody contact problems of elasticity . int .
methods engrg . *
41 * , 675696 ( 2010 ) dyyak , i.i . ,
prokopyshyn , i.i . : domain decomposition schemes for frictionless multibody contact problems of elasticity . in : g.k .
numerical mathematics and advanced applications 2009 , pp .
springer ( 2010 ) dyyak , i.i . ,
prokopyshyn , i.i . ,
prokopyshyn , i.a . : penalty robin
robin domain decomposition methods for unilateral multibody contact problems of elasticity : convergence results ( 2012 ) .
http://arxiv.org/pdf/1208.6478.pdf prokopyshyn , i.i . ,
dyyak , i.i . ,
martynyak , r.m . ,
prokopyshyn , i.a . : penalty robin
robin domain decomposition schemes for contact problems of nonlinear elastic bodies ( 2012 ) . | in this paper we propose on continuous level a class of domain decomposition methods of robin robin type to solve the problems of unilateral contact between elastic bodies with nonlinear winkler covers .
these methods are based on abstract nonstationary iterative algorithms for nonlinear variational equations in reflexive banach spaces .
we also provide numerical investigations of obtained methods using finite element approximations .
* key words : * unilateral contact , nonlinear winkler layers , nonlinear variational inequalities , nonlinear variational equations , iterative methods , domain decomposition * msc2010 : * 65n55 , 74s05 |
the largest source of information today is the world wide web .
the estimated number of documents nears 10 billion .
similarly , the number of documents changing on a daily basis is also enormous .
the ever - increasing growth of the web presents a considerable challenge in finding novel information on the web .
in addition , properties of the web , like scale - free small world ( sfsw ) structure @xcite may create additional challenges .
for example the direct consequence of the scale - free small world property is that there are numerous urls or sets of interlinked urls , which have a large number of incoming links .
intelligent web crawlers can be easily trapped at the neighborhood of such junctions as it has been shown previously @xcite .
we have developed a novel artificial life ( a - life ) method with intelligent individuals , crawlers , to detect new information on a news web site .
we define a - life as a population of individuals having both static structural properties , and structural properties which may undergo continuous changes , i.e. , adaptation .
our algorithms are based on methods developed for different areas of artificial intelligence , such as evolutionary computing , artificial neural networks and reinforcement learning .
all efforts were made to keep the applied algorithms as simple as possible subject to the constraints of the internet search .
evolutionary computing deals with properties that may be modified during the creation of new individuals , called multiplication. descendants may exhibit variations of population , and differ in performance from the others .
individuals may also terminate .
multiplication and selection is subject to the fitness of individuals , where fitness is typically defined by the modeler .
for a recent review on evolutionary computing , see @xcite . for reviews on related evolutionary theories and the dynamics of self - modifying systems
see @xcite and @xcite , respectively .
similar concepts have been studied in other evolutionary systems where organisms compete for space and resources and cooperate through direct interaction ( see , e.g. , @xcite and references therein . )
selection , however , is a very slow process and individual adaptation may be necessary in environments subject to quick changes . the typical form of adaptive learning is the connectionist architecture , such as artificial neural networks .
multilayer perceptrons ( mlps ) , which are universal function approximators have been used widely in diverse applications .
evolutionary selection of adapting mlps has been in the focus of extensive research @xcite . in a typical reinforcement learning ( rl )
problem the learning process @xcite is motivated by the expected value of long - term cumulated profit .
a well - known example of reinforcement learning is the td - gammon program of tesauro @xcite .
the author applied mlp function approximators for value estimation .
reinforcement learning has also been used in concurrent multi - robot learning , where robots had to learn to forage together via direct interaction @xcite .
evolutionary learning has been used within the framework of reinforcement learning to improve decision making , i.e. , the state - action mapping called policy @xcite . in this paper
we present a selection based algorithm and compare it to the well - known reinforcement learning algorithm in terms of their efficiency and behavior . in our problem ,
fitness is not determined by us , but fitness is implicit .
fitness is jointly determined by the ever changing external world and by the competing individuals together .
selection and multiplication of individuals are based on their fitness value .
communication and competition among our crawlers are indirect .
only the first submitter of a document may receive positive reinforcement .
our work is different from other studies using combinations of genetic , evolutionary , function approximation , and reinforcement learning algorithms , in that i ) it does not require explicit fitness function , ii ) we do not have control over the environment , iii ) collaborating individuals use value estimation under ` evolutionary pressure ' , and iv ) individuals work without direct interaction with each other .
we performed realistic simulations based on data collected during an 18 days long crawl on the web .
we have found that our selection based weblog update algorithm performs better in scale - free small world environment than the rl algorithm , eventhough the reinforcement learning algorithm has been shown to be efficient in finding relevant information @xcite .
we explain our results based on the different behaviors of the algorithms .
that is , the weblog update algorithm finds the good relevant document sources and remains at these regions until better places are found by chance .
individuals using this selection algorithm are able to quickly collect the new relevant documents from the already known places because they monitor these places continuously .
the reinforcement learning algorithm explores new territories for relevant documents and if it finds a good place then it collects the existing relevant documents from there .
the continuous exploration of rl causes that it finds relevant documents slower than the weblog update algorithm .
also , crawlers using weblog update algorithm submit more different documents than crawlers using the rl algorithm .
therefore there are more relevant new information among documents submitted by former than latter crawlers .
the paper is organized as follows . in section [
s : related ] we review recent works in the field of web crawling .
then we describe our algorithms and the forager architecture in section [ s : architecture ] .
after that in section [ s : experiments ] we present our experiment on the web and the conducted simulations with the results . in section [ s : discussion ]
we discuss our results on the found different behaviors of the selection and reinforcement learning algorithms .
section [ s : conclusion ] concludes our paper .
our work concerns a realistic web environment and search algorithms over this environment . we compare selective / evolutionary and reinforcement learning methods .
it seems to us that such studies should be conducted in ever changing , buzzling , wabbling environments , which justifies our choice of the environment .
we shall review several of the known search tools including those @xcite that our work is based upon .
readers familiar with search tools utilized on the web may wish to skip this section .
there are three main problems that have been studied in the context of crawlers .
rungsawang et al . @xcite and references therein and menczer @xcite studied the topic specific crawlers .
risvik et al . @xcite and references therein address research issues related to the exponential growth of the web .
cho and gracia - molina @xcite , menczer @xcite and edwards et .
al @xcite and references therein studies the problem of different refresh rates of urls ( possibly as high as hourly or as low as yearly ) .
rungsawang and angkawattanawit @xcite provide an introduction to and a broad overview of topic specific crawlers ( see citations in the paper ) .
they propose to learn starting urls , topic keywords and url ordering through consecutive crawling attempts .
they show that the learning of starting urls and the use of consecutive crawling attempts can increase the efficiency of the crawlers .
the used heuristic is similar to the weblog algorithm @xcite , which also finds good starting urls and periodically restarts the crawling from the newly learned ones .
the main limitation of this work is that it is incapable of addressing the freshness ( i.e. , modification ) of already visited web pages .
menczer @xcite describes some disadvantages of current web search engines on the dynamic web , e.g. , the low ratio of fresh or relevant documents .
he proposes to complement the search engines with intelligent crawlers , or web mining agents to overcome those disadvantages .
search engines take static snapshots of the web with relatively large time intervals between two snapshots .
intelligent web mining agents are different : they can find online the required recent information and may evolve intelligent behavior by exploiting the web linkage and textual information .
he introduces the infospider architecture that uses genetic algorithm and reinforcement learning , also describes the myspider implementation of it .
menczer discusses the difficulties of evaluating online query driven crawler agents .
the main problem is that the whole set of relevant documents for any given query are unknown , only a subset of the relevant documents may be known . to solve this problem he introduces two new metrics that estimate the real recall and precision based on an available subset of the relevant documents . with these metrics search engine and online crawler performances
can be compared .
starting the myspider agent from the 100 top pages of altavista the agent s precision is better than altavista s precision even during the first few steps of the agent .
the fact that the myspider agent finds relevant pages in the first few steps may make it deployable on users computers .
some problems may arise from this kind of agent usage .
first of all there are security issues , like which files or information sources are allowed to read and write for the agent .
the run time of the agents should be controlled carefully because there can be many users ( google answered more than 100 million searches per day in january - february 2001 ) using these agents , thus creating huge traffic overhead on the internet .
our weblog algorithm uses local selection for finding good starting urls for searches , thus not depending on any search engines .
dependence on a search engine can be a suffer limitation of most existing search agents , like myspiders .
note however , that it is an easy matter to combine the present algorithm with urls offered by search engines .
also our algorithm should not run on individual users s computers .
rather it should run for different topics near to the source of the documents in the given topic
e.g. , may run at the actual site where relevant information is stored .
risvik and michelsen @xcite mention that because of the exponential growth of the web there is an ever increasing need for more intelligent , ( topic-)specific algorithms for crawling , like focused crawling and document classification . with these algorithms crawlers and search engines
can operate more efficiently in a topically limited document space .
the authors also state that in such vertical regions the dynamics of the web pages is more homogenous .
they overview different dimensions of web dynamics and show the arising problems in a search engine model .
they show that the problem of rapid growth of web and frequent document updates creates new challenges for developing more and more efficient web search engines .
the authors define a reference search engine model having three main components : ( 1 ) crawler , ( 2 ) indexer , ( 3 ) searcher .
the main part of the paper focuses on the problems that crawlers need to overcome on the dynamic web . as a possible solution the authors propose a heterogenous crawling architecture .
they also present an extensible indexer and searcher architecture .
the crawling architecture has a central distributor that knows which crawler has to crawl which part of the web .
special crawlers with low storage and high processing capacity are dedicated to web regions where content changes rapidly ( like news sites ) .
these crawlers maintain up - to - date information on these rapidly changing web pages .
the main limitation of their crawling architecture is that they must divide the web to be crawled into distinct portions manually before the crawling starts .
a weblog like distributed algorithm as suggested here my be used in that architecture to overcome this limitation .
cho and garcia - molina @xcite define mathematically the freshness and age of documents of search engines .
they propose the poisson process as a model for page refreshment .
the authors also propose various refresh policies and study their effectiveness both theoretically and on real data .
they present the optimal refresh policies for their freshness and age metrics under the poisson page refresh model .
the authors show that these policies are superior to others on real data , too .
they collected about 720000 documents from 270 sites . although they show that in their database more than 20 percent of the documents are changed each day , they disclosed these documents from their studies .
their crawler visited the documents once each day for 5 months , thus can not measure the exact change rate of those documents . while in our work we definitely concentrate on these frequently changing documents .
the proposed refresh policies require good estimation of the refresh rate for each document .
the estimation influences the revisit frequency while the revisit frequency influences the estimation .
our algorithm does not need explicit frequency estimations .
the more valuable urls ( e.g. , more frequently changing ) will be visited more often and if a crawler does not find valuable information around an url being in it s weblog then that url finally will fall out from the weblog of the crawler .
however frequency estimations and refresh policies can be easily integrated into the weblog algorithm selecting the starting url from the weblog according to the refresh policy and weighting each url in the weblog according to their change frequency estimations .
menczer @xcite also introduces a recency metric which is 1 if all of the documents are recent ( i.e. , not changed after the last download ) and goes to 0 as downloaded documents are getting more and more obsolete . trivially immediately after a few minutes run of an online crawler the value of this metric will be 1 , while the value for the search engine will be lower .
edwards et al .
@xcite present a mathematical crawler model in which the number of obsolete pages can be minimized with a nonlinear equation system .
they solved the nonlinear equations with different parameter settings on realistic model data .
their model uses different buckets for documents having different change rates therefore does not need any theoretical model about the change rate of pages .
the main limitations of this work are the following : * by solving the nonlinear equations the content of web pages can not be taken into consideration .
the model can not be extended easily to ( topic-)specific crawlers , which would be highly advantageous on the exponentially growing web @xcite , @xcite , @xcite . * the rapidly changing documents ( like on news sites ) are not considered to be in any bucket , therefore increasingly important parts of the web are disclosed from the searches .
however the main conclusion of the paper is that there may exist some efficient strategy for incremental crawlers for reducing the number of obsolete pages without the need for any theoretical model about the change rate of pages .
there are two different kinds of agents : the foragers and the reinforcing agent ( ra ) .
the fleet of foragers crawl the web and send the urls of the selected documents to the reinforcing agent .
the ra determines which forager should work for the ra and how long a forager should work .
the ra sends reinforcements to the foragers based on the received urls .
we employ a fleet of foragers to study the competition among individual foragers .
the fleet of foragers allows to distribute the load of the searching task among different computers .
a forager has simple , limited capabilities , like limited number of starting urls and a simple , content based url ordering .
the foragers compete with each other for finding the most relevant documents . in this way
they efficiently and quickly collect new relevant documents without direct interaction . at first
the basic algorithms are presented .
after that the reinforcing agent and the foragers are detailed .
a forager periodically restarts from a url randomly selected from the list of starting urls .
the sequence of visited urls between two restarts forms a path .
the starting url list is formed from the @xmath0 first urls of the weblog . in the weblog
there are @xmath1 urls with their associated weblog values in descending order .
the weblog value of a url estimates the expected sum of rewards during a path after visiting that url .
the weblog update algorithm modifies the weblog before a new path is started ( algorithm [ t : weblog_pseudo ] ) .
the weblog value of a url already in the weblog is modified toward the sum of rewards in the remaining part of the path after that url .
a new url has the value of actual sum of rewards in the remaining part of the path .
if a url has a high weblog value it means that around that url there are many relevant documents .
therefore it may worth it to start a search from that url . ' '' '' ' '' '' [ t : weblog_pseudo]*weblog update*. @xmath2 was set to 0.3 ' '' '' xxx = xx = xx = xx = xx = xx = xx = xx = xx = ` input ` + @xmath3 the steps of the given path + @xmath4 the sum of rewards for each step in the given path + ` output ` + starting url list + ` method ` + @xmath5 cumulated sum of @xmath6 in reverse order + @xmath7 not having value in @xmath8 + @xmath9 having value in @xmath8 + ` for each ` @xmath10 + @xmath11 + ` endfor ` + ` for each ` @xmath12 + @xmath13 + @xmath14 + ` endfor ` + @xmath15 descending order of values in @xmath8 + @xmath15 truncate @xmath8 after the @xmath16 + element + starting url list @xmath17 first @xmath18 elements of @xmath8 ' '' '' ' '' '' without the weblog algorithm the weblog and thus the starting url list remains the same throughout the searches .
the weblog algorithm is a very simple version of evolutionary algorithms . here
, evolution may occur at two different levels : the list of urls of the forager is evolving by the reordering of the weblog . also , a forager may multiply , and its weblog , or part of it may spread through inheritance . this way , the weblog algorithm incorporates most basic features of evolutionary algorithms .
this simple form shall be satisfactory to demonstrate our statements .
a forager can modify its url ordering based on the received reinforcements of the sent urls .
the ( immediate ) profit is the difference of received rewards and penalties at any given step .
immediate profit is a myopic characterization of a step to a url .
foragers have an adaptive continuous value estimator and follow the _ policy _ that maximizes the expected long term cumulated profit ( ltp ) instead of the immediate profit .
such estimators can be easily realized in neural systems @xcite .
policy and profit estimation are interlinked concepts : profit estimation determines the policy , whereas policy influences choices and , in turn , the expected ltp .
( for a review , see @xcite . ) here , choices are based on the greedy ltp policy : the forager visits the url , which belongs to the _ frontier _ ( the list of linked but not yet visited urls , see later ) and has the highest estimated ltp . in the particular simulation
each forager has a @xmath19 dimensional probabilistic term - frequency inverse document - frequency ( prtfidf ) text classifier @xcite , generated on a previously downloaded portion of the geocities database .
fifty clusters were created by boley s clustering algorithm @xcite from the downloaded documents .
the prtfidf classifiers were trained on these clusters plus an additional one , the @xmath20 , representing general texts from the internet .
the prtfidf outputs were non - linearly mapped to the interval [ -1,+1 ] by a hyperbolic - tangent function . the classifier was applied to reduce the texts to a small dimensional representation .
the output vector of the classifier for the page of url @xmath21 is @xmath22 .
( the @xmath20 output was dismissed . )
this output vector is stored for each url ( algorithm [ t : pageinfo_urlordering_pseudo ] ) . ' '' '' ' '' '' [ t : pageinfo_urlordering_pseudo]*page information storage * ' '' '' xxx = xx = xx = xx = xx = xx = xx = xx = xx = ` input ` + @xmath23 urls of pages to be stored + ` output ` + @xmath24 the classifier output vectors for pages of @xmath25 + ` method ` + ` for each ` @xmath26 + @xmath27 text of page of @xmath28 + @xmath29 classifier output vector for @xmath30 + ` endfor ` ' '' '' ' '' '' a linear function approximator is used for ltp estimation .
it encompasses @xmath31 parameters , the _ weight vector _ @xmath32 .
the ltp of document of url @xmath21 is estimated as the scalar product of @xmath33 and @xmath34 : @xmath35 . during url ordering the url with highest ltp estimation
is selected .
the url ordering algorithm is shown in algorithm [ t : urlordering_pseudo ] . ' '' '' ' '' '' [ t : urlordering_pseudo]*url ordering * ' '' '' xxx = xx = xx = xx = xx = xx = xx = xx = xx = ` input ` + @xmath36 the set of available urls + @xmath24 the stored vector representation of the urls + ` output ` + @xmath37 url with maximum ltp value + ` method ` + ` for each ` @xmath38 + @xmath39 + ` endfor ` + @xmath37 url with maximal ltp @xmath40 ' '' '' ' '' '' the weight vector of each forager is tuned by temporal difference learning @xcite .
let us denote the current url by @xmath41 , the next url to be visited by @xmath42 , the output of the classifier for @xmath43 by @xmath44 and the estimated ltp of a url @xmath43 by @xmath45 .
assume that leaving @xmath41 to @xmath42 the immediate profit is @xmath46 .
our estimation is perfect if @xmath47 .
future profits are typically discounted in such estimations as @xmath48 , where @xmath49 .
the error of value estimation is @xmath50 we used throughout the simulations @xmath51 . for each step @xmath52 the weights of the value function were tuned to decrease the error of value estimation based on the received immediate profit @xmath46 .
the @xmath53 estimation error was used to correct the parameters .
the @xmath54 component of the weight vector , @xmath55 , was corrected by @xmath56 with @xmath57 and @xmath58 .
these modified weights in a stationary environment would improve value estimation ( see , e.g , @xcite and references therein ) .
the url ordering update is given in algorithm [ t : urlordering_update_pseudo ] . ' '' '' ' '' '' [ t : urlordering_update_pseudo]*url ordering update * ' '' '' xxx = xx = xx = xx = xx = xx = xx = xx = xx = ` input ` + @xmath59 the step for which the reinforcement is received + @xmath60 the previous step before @xmath42 + @xmath61 reinforcement for visiting @xmath42 + ` output ` + @xmath62 the updated weight vector + ` method ` + @xmath63 + @xmath64 ' '' '' ' '' '' without the update algorithm the weight vector remains the same throughout the search .
a document or page is possibly relevant for a forager if it is not older than 24 hours and the forager has not marked it previously .
algorithm [ t : relevant_pseudo ] shows the procedure of selecting such documents .
the selected documents are sent to the ra for further evaluation . ' '' '' ' '' '' [ t : relevant_pseudo]*document relevancy at a forager * ' '' '' xxx = xx = xx = xx = xx = xx = xx = xx = xx = ` input ` + @xmath65 the pages to be examined + ` output ` + @xmath66 the selected pages + ` method ` + @xmath67 previously selected relevant pages + @xmath66 all pages from @xmath68 which are + not older than 24 hours and + not contained in @xmath69 + @xmath67 add @xmath70 to @xmath69 ' '' '' ' '' '' during multiplication the weblog is randomly divided into two equal sized parts ( one for the original and one for the new forager ) .
the parameters of the url ordering algorithm ( the weight vector of the value estimation ) are either copied or new random parameters are generated . if the forager has a url ordering update algorithm then the parameters are copied .
if the forager does not have any url ordering update algorithm then new random parameters are generated , as shown in algorithm [ t : multiplication_pseudo ] . ' '' '' ' '' '' [ t : multiplication_pseudo]*multiplication * ' '' ''
xxx = xx = xx = xx = xx = xx = xx = xx = xx = ` input ` + @xmath8 + weight vector of url ordering + ` output ` + @xmath71 + @xmath72 + ` method ` + @xmath73 randomly selected + urls and values from @xmath8 + @xmath15 delete @xmath71 from @xmath8 + ` if ` forager has url ordering update algorithm + @xmath74 copy the weight vector of url ordering + ` else ` + @xmath74 generate a new random weight vector + ` endif ` ' '' '' ' '' '' a reinforcing agent controls the `` life '' of foragers .
it can start , stop , multiply or delete foragers .
ra receives the urls of documents selected by the foragers , and responds with reinforcements for the received urls .
the response is @xmath75 ( a.u . ) for a relevant document and @xmath76 ( a.u . ) for a not relevant document .
a document is relevant if it is not yet seen by the reinforcing agent and it is not older than 24 hours .
the reinforcing agent maintains the score of each forager working for it .
initially each forager has @xmath77 score .
when a forager sends a url to the ra , the forager s score is decreased by @xmath78 .
after each relevant page sent by the forager , the forager s score is increased by @xmath79 ( algorithm [ t : manageurl_pseudo ] ) . ' '' '' ' '' '' [ t : manageurl_pseudo]*manage received url * ' '' '' xxx = xx = xx = xx = xx = xx = xx = xxxxx = ` input ` + @xmath80 received url from forager + ` output ` + reinforcement to forager + updated forager score + ` method ` + @xmath81 relevant pages seen by the ra + @xmath27 get page of @xmath28 + decrease forager s score with @xmath82 + ` if ` @xmath83 or page date is older than 24 hours + send @xmath84 to forager + ` else ` + @xmath81 add @xmath30 to @xmath85 + send @xmath86 to forager + increase forager s score with @xmath87 + ` endif ` ' '' '' ' '' '' when the forager s score reaches @xmath88 and the number of foragers is smaller than @xmath89 then the forager is multiplied .
that is a new forager is created with the same algorithms as the original one has , but with slightly different parameters .
when the forager s score goes below @xmath90 and the number of foragers is larger than @xmath91 then the forager is deleted ( algorithm [ t : manageforager_pseudo ] ) .
note that a forager can be multiplied or deleted immediately after it has been stopped by the ra and before the next forager is activated . ' '' '' ' '' '' [ t : manageforager_pseudo ] * : manage forager * ' '' '' xxx = xx = xx = xx = xx = xx = xx = xxxxx = ` input ` + @xmath92 the forager to be multiplied or deleted + ` output ` + possibly modified list of foragers + ` method ` + ` if ` ( @xmath93 s score @xmath94 @xmath95 ` and ` + number of foragers @xmath96 @xmath97 ) + @xmath98 call @xmath93 s + * multiplication , alg .
[ t : multiplication_pseudo ] * + @xmath93 may modify it s own weblog + @xmath99 create a new forager with the received + @xmath8 and @xmath100 + set the two foragers score to @xmath101 + ` else if ` ( @xmath93 s score @xmath102 @xmath103 ` and ` + number of foragers @xmath104 @xmath105 ) + delete @xmath93 + ` endif ` ' '' '' ' '' '' foragers on the same computer are working in time slices one after each other .
each forager works for some amount of time determined by the ra .
then the ra stops that forager and starts the next one selected by the ra .
the pseudo - code of the reinforcing agent is given in algorithm [ t : reinforcing_pseudo ] . ' '' '' ' '' '' [ t : reinforcing_pseudo ] * : reinforcing agent * ' '' '' xxx = xx = xx = xx = xx = xx = xx = xxxxx = ` input ` + seed urls + ` output ` + @xmath81 found relevant documents + ` method ` + @xmath81 empty set /*set of all observed relevant pages + initialize @xmath105 foragers with the seed urls + set one of them to be the next + ` repeat ` + start next forager + receive possibly relevant url + call * manage received url , alg .
[ t : manageurl_pseudo ] * with url + stop forager if its time period is over + call * manage forager , alg . [ t : manageforager_pseudo ] * with this forager + choose next forager + ` until ` time is over ' '' '' ' '' '' a forager is initialized with parameters defining the url ordering , and either with a weblog or with a seed of urls ( algorithm [ t : initforager_pseudo ] ) . after its initialization a forager crawls in search paths , that is after a given number of steps the search restarts and the steps between two restarts form a path . during each path
the forager takes @xmath106 number of steps , i.e. , selects the next url to be visited with a url ordering algorithm . at the beginning of a path a url
is selected randomly from the starting url list .
this list is formed from the 10 first urls of the weblog .
the weblog contains the possibly good starting urls with their associated weblog values in descending order .
the weblog algorithm modifies the weblog and so thus the starting url list before a new path is started .
when a forager is restarted by the ra , after the ra has stopped it , the forager continues from the internal state in which it was stopped .
the pseudo code of step selection is given in algorithm [ t : stepselection_pseudo ] . ' '' '' ' '' '' [ t : initforager_pseudo]*initialization of the forager * ' '' '' xxx = xx = xx = xx = xx = xx = xx = xx = xx = ` input ` + weblog or seed urls + url ordering parameters + ` output ` + initialized forager + ` method ` + set path step number to @xmath107 /*start new path + set the weblog + either with the input weblog + or put the seed urls into the weblog with 0 weblog value + set the url ordering parameters in url ordering algorithm ' '' '' ' '' '' ' '' '' ' '' '' [ t : stepselection_pseudo]*url selection * ' '' '' xxx = xx = xx = xx = xx = xx = xx = xx = xx = ` input ` + @xmath36 set of urls available in this step + @xmath108 set of visited urls in this path + ` output ` + @xmath109 selected url to be visited next + ` method ` + ` if ` path step number @xmath110 + @xmath109 selected url by * url ordering , alg .
[ t : urlordering_pseudo ] * + increase path step number + ` else ` + call the * weblog update , alg .
[ t : weblog_pseudo ] * to update the weblog + @xmath109 select a random url from the starting url list + set path step number to 1 + @xmath36 empty set + @xmath108 empty set + ` endif ` ' '' '' ' '' '' the url ordering algorithm selects a url to be the next step from the frontier url set .
the selected url is removed from the frontier and added to the visited url set to avoid loops .
after downloading the pages , only those urls ( linked from the visited url ) are added to the frontier which are not in the visited set . in each step
the forager downloads the page of the selected url and all of the pages linked from the page of selected url .
it sends the urls of the possibly relevant pages to the reinforcing agent .
the forager receives reinforcements on any previously sent but not yet reinforced urls and calls the url ordering update algorithm with the received reinforcements .
the pseudo code of a forager is shown in algorithm [ t : forager_pseudo ] . ' '' '' ' '' '' [ t : forager_pseudo]*forager * ' '' '' xxx = xx = xx = xx = xx = xx = xx = xx = xx = ` input ` + @xmath36 set of urls available in the next step + @xmath108 set of visited urls in the current path + ` output ` + sent documents to the ra + modified @xmath111 and @xmath112 + modified @xmath8 and url ordering weight vector + ` method ` + ` repeat ` + @xmath109 call * url selection , alg .
[ t : stepselection_pseudo ] * + @xmath36 remove @xmath113 from @xmath111 + @xmath108 add @xmath113 to @xmath112 + @xmath27 download the page of @xmath113 + @xmath114 links of @xmath30 + @xmath115 @xmath116 which are not @xmath112 + @xmath36 add @xmath117 to @xmath111 + download pages of @xmath116 + call * page information storage , alg . [ t : pageinfo_urlordering_pseudo ] * with @xmath117 + @xmath66 call * document relevancy , alg .
[ t : relevant_pseudo ] * for + all pages + send @xmath70 to reinforcing agent + receive reinforcements for sent but not yet reinforced pages + call * url ordering update , alg .
[ t : urlordering_update_pseudo ] * with + the received reinforcements + ` until ` time is over ' '' '' ' '' ''
we conducted an 18 day long experiment on the web to gather realistic data .
we used the gathered data in simulations to compare the weblog update ( section [ sss : weblog ] ) and reinforcement learning algorithms ( section [ sss : rl ] ) . in web experiment
we used a fleet of foragers using combination of reinforcement learning and weblog update algorithms to eliminate any biases on the gathered data .
first we describe the experiment on the web then the simulations .
we analyze our results at the end of this section .
we ran the experiment on the web on a single personal computer with celeron 1000 mhz processor and 512 mb ram .
we implemented the forager architecture ( described in section [ s : architecture ] ) in java programming language . in this experiment a fixed number of foragers were competing with each other to collect news at the cnn web site .
the foragers were running in equal time intervals in a predefined order .
each forager had a 3 minute time interval and after that interval the forager was allowed to finish the step started before the end of the time interval .
we deployed 8 foragers using the weblog update and the reinforcement learning based url ordering update algorithms ( 8 wlrl foragers ) .
we also deployed 8 other foragers using the weblog update algorithm but without reinforcement learning ( 8 wl foragers ) .
the predefined order of foragers was the following : 8 wlrl foragers were followed by the 8 wl foragers .
we investigated the link structure of the gathered web pages . as it is shown in fig .
[ f : sf ] the links have a power - law distribution ( @xmath118 ) with @xmath119 for outgoing links and @xmath120 for incoming links .
that is the link structure has the scale - free property .
the clustering coefficient @xcite of the link structure is 0.02 and the diameter of the graph is 7.2893 .
we applied two different random permutations to the origin and to the endpoint of the links , keeping the edge distribution unchanged but randomly rewiring the links .
the new graph has 0.003 clustering coefficient and 8.2163 diameter .
that is the clustering coefficient is smaller than the original value by an order of magnitude , but the diameter is almost the same .
therefore we can conclude that the links of gathered pages form small world structure . ) . vertical axis : relative frequency of number of edges at different urls ( @xmath121 ) .
dots and dark line correspond to outgoing links , crosses and gray line correspond to incoming links .
, width=240 ] the data storage for simulation is a centralized component .
the pages are stored with 2 indices ( and time stamps ) .
one index is the url index , the other is the page index .
multiple pages can have the same url index if they were downloaded from the same url .
the page index uniquely identifies a page content and the url from where the page was download . at each page download of any foragers we stored the followings ( with a time stamp containing the time of page download ) :
1 . if the page is relevant according to the ra then store `` relevant '' 2 .
if the page is from a new url then store the new url with a new url index and the page s state vector with a new page index 3 . if the content of the page is changed since the last download then store the page s state vector with a new page index but keep the url index 4 . in both previous cases store the links of the page as links to page indices of the linked pages 1 .
if a linked page is from a new url then store the new url with a new url index and the linked page s state vector with a new page index 2 .
if the content of the linked page is changed since the last check then store the page s state vector with a new page index but same url index for the simulations we implemented the forager architecture in matlab .
the foragers were simulated as if they were running on one computer as described in the previous section . during simulations we used the web pages that we gathered previously to generate a realistic environment
( note that the links of pages point to local pages ( not to pages on the web ) since a link was stored as a link to a local page index ) : * simulated documents had the same state vector representation for url ordering as the real pages had * simulated relevant documents were the same as the relevant documents on the web * pages and links appeared at the same ( relative ) time when they were found in the web experiment - using the new url indices and their time stamps * pages and links are refreshed or changed at the same relative time as the changes were detected in the web experiment using the new page indices for existing url indices and their time stamps * simulated time of a page download was the average download time of a real page during the web experiment .
we conducted simulations with two different kinds of foragers .
the first case is when foragers used only the weblog update algorithm without url ordering update ( wl foragers ) .
the second case is when foragers used only the reinforcement learning based url ordering update algorithm without the weblog update algorithm ( rl foragers ) .
each wl forager had a different weight vector for url value estimation during multiplication the new forager got a new random weight vector .
rl foragers had the same weblog with the first 10 urls of the gathered pages that is the starting url of the web experiment and the first 9 visited urls during that experiment . in both cases initially there were 2 foragers and they were allowed to multiply until reaching the population of 16 foragers .
the simulation for each type of foragers were repeated 3 times with different initial weight vectors for each forager .
the variance of the results show that there is only a small difference between simulations using the same kind of foragers , even if the foragers were started with different random weight vectors in each simulation .
table [ t : params ] shows the investigated parameters during simulations . [ cols= " < ,
< " , ] from table [ t : data ] we can conclude the followings : * rl and wl foragers have similar download efficiency , i.e. , the efficiencies from the point of view of the news site are about the same .
* wl foragers have higher sent efficiencies than rl foragers , i.e. , the efficiency from the point of view of the ra is higher .
this shows that wl foragers divide the search area better among each other than rl foragers .
sent efficiency would be 1 if none of two foragers have sent the same document to the ra .
* rl foragers have higher relative found url value than wl foragers .
rl foragers explore more than wl foragers and rl found more urls than wl foragers did per downloaded page .
* wl foragers find faster the new relevant documents in the already found clusters .
that is freshness is higher and age is lower than in the case of rl foragers .
[ f : efficiency ] shows other aspects of the different behaviors of rl and wl foragers .
download efficiency of rl foragers has more , higher , and sharper peaks than the download efficiency of wl foragers has .
that is wl foragers are more balanced in finding new relevant documents than rl foragers .
the reason is that while the wl foragers remain in the found good clusters , the rl foragers continuously explore the new promising territories .
the sharp peaks in the efficiency show that rl foragers _ find and recognize _ new good territories and then _ quickly collect _ the current relevant documents from there .
the foragers can recognize these places by receiving more rewards from the ra if they send urls from these places .
the predefined order did not influence the working of foragers during the web experiment . from fig .
[ f : efficiency ] it can be seen that foragers during the 3 independent experiments did not have very different efficiencies . on fig .
[ f : freshness ] we show that the foragers in each run had a very similar behavior in terms of age and freshness , that is the values remains close to each other throughout the experiments .
also the results for individual runs were close to the average values in table [ t : data ] ( see the standard deviations ) . in each individual run
the foragers were started with different weight vectors , but they reached similar efficiencies and behavior .
this means that the initial conditions of the foragers did not influence the later behavior of them during the simulations .
furthermore foragers could not change their environment drastically ( in terms of the found relevant documents ) during a single 3 minute run time because of the short run time intervals and the fast change of environment large number of new pages and often updated pages in the new site . during
the web experiment foragers were running in 8 wlrl , 8 wl , 8 wlrl , 8 wl , temporal order . because of the fact that initial conditions does not influence the long term performance of foragers and the fact that the foragers can not change their environment fully we can start to examine them after the first run of wlrl foragers .
then we got the other extreme order of foragers , that is the 8 wl , 8 wlrl , 8 wl , 8 wlrl , temporal ordering . for the overall efficiency and behavior of foragers it did not really matter
if wlrl or wl foragers run first and one could use mixed order in which after a wlrl forager a wl forager runs and after a wl forager a wlrl forager comes . however , for higher bandwidths and for faster computers , random ordering may be needed for such comparisons .
our first conjecture is that selection is efficient on scale - free small world structures .
lrincz and kkai @xcite and rennie et al .
@xcite showed that rl is efficient in the task of finding relevant information on the web .
here we have shown experimentally that the weblog update algorithm , selection among starting urls , is at least as efficient as the rl algorithm .
the weblog update algorithm finds as many relevant documents as rl does if they download the same amount of pages .
wl foragers in their fleet select more different urls to send to the ra than rl foragers do in their fleet , therefore there are more relevant documents among those selected by wl foragers then among those selected by rl foragers . also the freshness and age of found relevant documents are better for wl foragers than for rl foragers . for the weblog update algorithm ,
the selection among starting urls has no fine tuning mechanism . throughout its life
a forager searches for the same kind of documents goes into the same ` direction ' in the state space of document states determined by its fixed weight vector .
the only adaptation allowed for a wl forager is to select starting urls from the already seen urls .
the wl forager can not modify its ( ` directional ' ) preferences according goes newly found relevant document supply , where relevant documents are abundant .
but a wl forager finds good relevant document sources in its own direction and forces its search to stay at those places . by chance the forager can find better sources in its own direction if the search path from a starting url is long enough . on fig .
[ f : efficiency ] it is shown that the download efficiency of the foragers does not decrease with the multiplication of the foragers .
therefore the new foragers must found new and good relevant document sources quickly after their appearances .
the reinforcement learning based url ordering update algorithm is capable to fine tune the search of a forager by adapting the forager s weight vector .
this feature has been shown to be crucial to adapt crawling in novel environments @xcite .
an rl forager goes into the direction ( in the state space of document states ) where the estimated long term cumulated profit is the highest .
because the local environment of the foragers may changes rapidly during crawling , it seems desirable that foragers can quickly adapt to the found new relevant documents .
relevant documents may appear lonely , not creating a good relevant document source , or do not appear at the right url by a mistake .
this noise of the web can derail the rl foragers from good regions .
the forager may `` turn '' into less valuable directions , because of the fast adaptation capabilities of rl foragers .
our second conjecture is that selection fits sfsw better than rl .
we have shown in our experiments that selection and rl have different behaviors .
selection selects good information sources , which are worth to revisit , and stays at those sources as long as better sources are not found by chance .
rl explores new territories , and adapts to those .
this adaptation can be a disadvantage when compared with the more rigid selection algorithm , which sticks to good places until ` provably ' better places are discovered .
therefore wl foragers , which can not be derailed and stay in their found ` niches ' can find new relevant documents faster in such already known terrains than rl foragers can .
that is , freshness is higher and age is lower for relevant documents found by wl foragers than for relevant documents found by rl foragers .
also , by finding good sources and staying there , wl foragers divide the search task better than rl foragers do , this is the reason for the higher sent efficiency of wl foragers than of rl foragers .
we have rewired the network as it was described in section [ ss : real ] . this way a scale - free ( sf ) but not so small world was created . intriguingly , in this sf structure , rl foragers performed better than wl ones .
clearly , further work is needed to compare the behavior of the selective and the reinforcement learning algorithms in other then sfsw environments .
such findings should be of relevance in the deployment of machine learning methods in different problem domains . from the practical point of view
, we note that it is an easy matter to combine the present algorithm with urls offered by search engines .
also , the values reported by the crawlers about certain environments , e.g. , the environment of the url offered by search engines represent the neighborhood of that url and can serve adaptive filtering .
this procedure is , indeed , promising to guide individual searches as it has been shown elsewhere @xcite .
we presented and compared our selection algorithm to the well - known reinforcement learning algorithm .
our comparison was based on finding new relevant documents on the web , that is in a dynamic scale - free small world environment .
we have found that the weblog update selection algorithm performs better in this environment than the reinforcement learning algorithm , eventhough the reinforcement learning algorithm has been shown to be efficient in finding relevant information @xcite .
we explain our results based on the different behaviors of the algorithms .
that is the weblog update algorithm finds the good relevant document sources and remains at these regions until better places are found by chance .
individuals using this selection algorithm are able to quickly collect the new relevant documents from the already known places because they monitor these places continuously . the reinforcement learning algorithm explores new territories for relevant documents and if it finds a good place then it collects the existing relevant documents from there .
the continuous exploration and the fine tuning property of rl causes that rl finds relevant documents slower than the weblog update algorithm . in our future work
we will study the combination of the weblog update and the rl algorithms .
this combination uses the wl foragers ability to stay at good regions with the rl foragers fine tuning capability . in this way
foragers will be able to go to new sources with the rl algorithm and monitor the already found good regions with the weblog update algorithm .
we will also study the foragers in a simulated environment which is not a small world .
the clusters of small world environment makes it easier for wl foragers to stay at good regions .
the small diameter due to the long distance links of small world environment makes it easier for rl foragers to explore different regions .
this work will measure the extent at which the different foragers rely on the small world property of their environment .
this material is based upon work supported by the european office of aerospace research and development , air force office of scientific research , air force research laboratory , under contract no .
fa8655 - 03 - 1 - 3036 .
this work is also supported by the national science foundation under grants no .
int-0304904 and no .
any opinions , findings and conclusions or recommendations expressed in this material are those of the author(s ) and do not necessarily reflect the views of the european office of aerospace research and development , air force office of scientific research , air force research laboratory .
j. edwards , k. mccurley , and j. tomlin , _ an adaptive model for optimizing performance of an incremental web crawler _ , proceedings of the tenth international conference on world wide web , 2001 , pp . 106113 .
b. gbor , zs .
palotai , and a. lrincz , _ value estimation based computer - assisted data mining for surfing the internet _ , int .
joint conf . on neural networks ( piscataway , nj 08855 - 1331 ) , ieee operations center , 26 - 29 july , budapest , hungary 2004 , pp .
1035 . , ieee catalog number : 04ch37541c , ijcnn2004 cd rom conference proceedings .
thorsten joachims , _ a probabilistic analysis of the rocchio algorithm with tfidf for text categorization _ , proceedings of icml-97 , 14th international conference on machine learning ( nashville , us ) ( douglas h. fisher , ed . ) , morgan kaufmann publishers , san francisco , us , 1997 , pp . 143151 .
k. tuyls , d. heytens , a. nowe , and b. manderick , _ extended replicator dynamics as a key to reinforcement learning in multi - agent systems _ , ecml 2003 , lnai 2837 ( n. lavrac et al . , ed . ) , springer - verlag , berlin , 2003 , pp . | in this paper we compare the performance characteristics of our selection based learning algorithm for web crawlers with the characteristics of the reinforcement learning algorithm .
the task of the crawlers is to find new information on the web .
the selection algorithm , called weblog update , modifies the starting url lists of our crawlers based on the found urls containing new information .
the reinforcement learning algorithm modifies the url orderings of the crawlers based on the received reinforcements for submitted documents .
we performed simulations based on data collected from the web .
the collected portion of the web is typical and exhibits scale - free small world ( sfsw ) structure .
we have found that on this sfsw , the weblog update algorithm performs better than the reinforcement learning algorithm .
it finds the new information faster than the reinforcement learning algorithm and has better new information / all submitted documents ratio .
we believe that the advantages of the selection algorithm over reinforcement learning algorithm is due to the small world property of the web . |
antiferromagnets in an external magnetic field show a variety of phase diagrams depending on the interaction terms present in the spin hamiltonian@xcite .
the spin interaction may be isotropic , anisotropic with an easy axis and/or single - ion anisotropy terms , where the anisotropy is in the direction of the external magnetic field .
the phase diagram of such models exhibit a multicritical point , where several transition lines meet . at a _
bicritical _ point three phases - an antiferromagnetic phase , a spin flop phase and the paramagnetic phase - are in coexistence .
the phase transition lines to the paramagnetic phase are second order transition lines , whereas the transition line between the spin flop and the antiferromagnetic phase is of first order . at the _ tetracritical _ point four phases - an antiferromagnetic phase , a spin flop phase , an intermediate or mixed phase and the paramagnetic phase - are in coexistence .
all transition lines are of second order in this case .
a field theoretic description of these models starts with a static functional for an @xmath6-component field @xmath7 of @xmath8 symmetry ( @xmath9 ) leading to different multicritical behavior connected with the stable fixed point ( fp ) found in the renormalization group treatment@xcite .
bicritical behavior has been connected with the stability of the well known _
isotropic heisenberg _ fixed point of @xmath10 symmetry , whereas tetracriticality has been connected with a fixed point of @xmath0 symmetry , which might be either the so called _ biconical _ fp or the _ decoupling _ fp . in the last fp the parallel and the perpendicular components of the order parameter ( op ) are asymptotically decoupled .
the important questions which theory should give an answer to is , ( i ) which of these fps is the stable one in a three- or two - dimensional system , and ( ii ) what are the differences in the critical behavior at the multicritical point ?
these questions have been risen and considered in one loop order@xcite , where the heisenberg fp turns out to be the stable one in @xmath11 for the case @xmath4 and @xmath5 , but this picture is changed in higher loop order . in a five loop order @xmath12 expansion it has been found that the biconical fp is the stable one @xcite .
it also has been found that the differences between the exponents at the different multicritical points are much smaller than in the one loop order calculation .
physical examples where such multicritical behavior has been found are the anisotropic antiferromagnets@xcite ( with the magnetic field in the hard direction ) like@xcite gdalo@xmath13 and@xcite mnf@xmath14 , as well as@xcite mncl@xmath144d@xmath14o or@xcite mn@xmath14as@xmath15 ( a = si or ge ) .
other examples with a single ion anisotropy might be layered cuprate antiferromagnets like ( ca , la)@xmath16cu@xmath17o@xmath18 . besides the examples with @xmath4 and @xmath5 one might consider other cases : @xmath4 and @xmath19 when additional anisotropies are present as in@xcite nicl@xmath14 or high-@xmath20 superconductors representing a system with @xmath21 ( corresponding to the superconductor op ) and @xmath22 ( corresponding to the antiferromagnetic op ) .
quite recently the possible types of phase diagrams in the magnetic field - temperature plane of @xmath11 of uniaxially anisotropic antiferromagnets have been studied by monte carlo simulations@xcite . for @xmath4 and @xmath5
a phase diagram with a bicritical point has been found in agreement with earlier simulations @xcite , but contrary to the results of renormalization group theory in higher loop orders@xcite .
a general picture is obtained when one considers a generalized model with an @xmath6 component order parameter ( op ) , which splits into @xmath1 parallel op components and @xmath2 perpendicular op components and quartic interaction terms of @xmath0 symmetry .
both parallel and perpendicular op components become critical at the multicritical point . in the @xmath1-@xmath2-space regions of different types of multicriticality
exist touching each other at stability border lines ( phase border lines ) where the fixed points change their stability ( such a picture of the different stability regions might be called a phase diagram ) .
in addition to the stability of a fixed point we want to mention that one has to consider also the attraction regions of a fixed point to answer the question wether one can reach the stable fixed point . in order to discuss the attraction regions one has to consider the flow of the couplings from the nonuniversal initial ( background ) values . we therefore reconsider the critical behavior of systems with @xmath0 symmetry .
being interested in criticality of three dimensional systems , we will work within the minimal subtraction scheme and evaluate the results at fixed dimension @xmath11.@xcite for the universal properties ( as asymptotic critical exponents and marginal dimensions ) it turns out that already the two loop calculations refined by resummation are in good quantitative agreement with previous resummed higher order @xmath23-expansion results.@xcite however , contrary to previous calculations , the technique we use gives a possibility to analyze non - universal effective critical behavior which is manifested in a broader temperature interval near the ( multi)critical point .
such calculations are out - of - reach the @xmath23-expansion and will be performed below on the base of analysis of the renormalization group flow . the paper is organized as follows : starting from the static functional ( sec .
[ func ] ) we introduce the renormalization in sec .
[ renorm ] and calculate the field theoretic functions in sec . [ zeta ] .
the perturbative expansions being asymptotic , we apply in sec .
[ fixp ] the resummation technique to restore their convergence and to extract numerical values of the fixed points of the renormalization group transformations and their stability .
we discuss the stability border lines between the fixed points and show that they are shifted considerably compared to the one loop calculation . as a result for the isotropic antiferromagnet in a magnetic field represented by the point ( 1,2 ) in the @xmath1-@xmath2-space
the _ biconical fixed point is stable predicting tetracritical behavior if the fixed point is reached from the background_. the very neighborhood of the stability border lines is characterized by very small transient exponents .
moreover , looking at the attraction regions a surface in the space of the fourth order couplings exists above which _ no finite fixed point _ can be reached .
this indicates the possibility of a scenario mentioned already earlier@xcite where the multicritical point is a _
triple point _ and the second order lines separating the paramagnetic phase from the ordered phases contain a tricritical point . in section [ expon ]
the critical exponents are defined and their asymptotic values are calculated for the physically interesting case @xmath4 , @xmath5 .
the flow equations and effective exponents are discussed in sec .
[ flow ] leading to our final conclusions and outlook in sec .
[ out ] . in the appendices we discuss the perturbative expansion for
the vertex functions ( appendix a ) and explain the resummation procedure exploited in our calculations ( appendix b ) .
the critical behavior of an isotropic system ( @xmath24 symmetry ) with short range interaction is determined by the static functional @xmath25 which is known as the ginzburg - landau - wilson(glw)-functional .
the order parameter @xmath26 is assumed to be a @xmath6-component real vector .
the symbol @xmath27 denotes the scalar product between vectors .
@xmath28 is proportional to the temperature distance to the critical point and @xmath29 is the fourth order coupling in which perturbation expansion is usually performed .
systems represented by such a static functional have been extensively studied in the last decades with different renormalization procedures , and the corresponding critical exponents and amplitude ratios are known up to high loop orders ( see e.g.@xcite ) . in order to describe bicritical behavior the @xmath6-dimensional space of the order parameter components will be divided into two subspaces with dimensions @xmath2 and @xmath1 with the property @xmath30 in the following .
correspondingly the order parameter separates into @xmath31 where @xmath32 is the @xmath2-dimensional order parameter of the @xmath2-subspace , and @xmath33 is the @xmath1-dimensional order parameter of the @xmath1-subspace .
performing the separation in the glw - functional ( [ hglw ] ) one obtains @xmath34 this functional contains three fourth order couplings @xmath35 and instead of one parameter @xmath28 as in ( [ hglw ] ) , in ( [ hbicrit ] ) two different parameters @xmath36 and @xmath37 appear referring to different temperature distances .
the decomposition in parallel and perpendicular op components allows to describe the critical behavior at the meeting point of two critical lines : ( i ) the line where @xmath36 becomes zero and the @xmath2-dimensional components @xmath32 are the op , and ( ii ) the line where @xmath37 becomes zero and the @xmath1-dimensional components @xmath38 are the op . at the meeting point
both quadratic terms become zero and both components of @xmath39 have to be taken into account . the critical behavior of this multicritical point , has been described already in one loop order@xcite and three different types of multicritical behavior have been found ; ( i ) one described by the well known isotropic @xmath6 component heisenberg fixed point where all fourth order couplings are equal , ( ii ) one described by a decoupling fixed point , which consists of a combination of two @xmath2 and @xmath1 component isotropic heisenberg fixed points of two decoupled systems and ( iii ) a new type of fixed point called the biconical fixed point . which of these fixed point is the stable one depends on the number of components @xmath2 and @xmath1 and the dimension @xmath3 of space .
the scaling properties depend on the symmetry of stable fixed point when the @xmath24 symmetry in the op space is broken to @xmath0 symmetry , then also the spatial correlations are different for the two op subspaces .
the procedure used to obtain the vertex functions appropriate for the renormalization is described in more details in the appendix [ vertex ] . from the general structure of the two point vertex functions presented
therein follows that the order parameter functions @xmath32 and @xmath33 in the subspaces may be renormalized by the scalar renormalization factors @xmath40 the above relations and the definitions ( [ correlperp ] ) and ( [ correlpara ] ) imply that the correlation lengths , @xmath41 and @xmath42 , do not renormalize .
they constitute together with the wave vector modulus @xmath43 the independent lengths of the system .
the fourth order couplings may also be renormalized by scalar renormalization factors @xmath44 with the geometrical factor @xmath45 analogously to dohm@xcite .
in ( [ ad ] ) @xmath46 denotes the euler gamma - function and @xmath47 is the surface of the @xmath3-dimensional unit sphere .
we want to remark that it would be possible to introduce a @xmath48-matrix for the renormalization of the three fourth order couplings @xmath49 as performed in @xcite .
the resulting @xmath50-functions are the same anyway .
thus it remains a matter of taste if one uses a scalar or a matrix renormalization for the three fourth order couplings .
the situation changes if one considers @xmath51- insertions .
the vertex function @xmath52 splits up into four functions @xmath53 , @xmath54 , @xmath55 and @xmath56 ( for the notation see appendix [ vertex ] ) .
a consistent renormalization of the @xmath51- insertions is only possible by introducing @xmath57 with the renormalization matrix @xmath58 the zeroth order of the perturbation expansion only appears in the diagonal elements @xmath59 , while the off - diagonal elements @xmath60 start with the one loop order . with ( [ phiren ] ) and ( [ phi2ren ] ) an arbitrary vertex function ( [ vertexfunc3 ] ) renormalizes as @xmath61 the indices @xmath62 , @xmath63 and @xmath64 are running over @xmath65 and @xmath66 .
the renormalization introduced in the above relations removes the poles from all vertex functions except the functions @xmath67 which have to be considered separately .
thus equation ( [ vertexren ] ) is valid for all vertex functions except @xmath67 .
the multiplicative renormalization in ( [ vertexren ] ) does not remove all singularities at @xmath68 in these function . in order to remove the remaining poles
an additional additive renormalization is necessary .
the multiplicative renormalization leads to functions @xmath69 a finite function can be obtained by subtracting the singular part @xmath70_s$ ] from ( [ gamma02r ] ) .
thus we may introduce @xmath71_s\right)\ ] ] as renormalized functions .
the additive renormalizations are then defined by @xmath72_s\ ] ] the three functions @xmath73 , @xmath74 and @xmath75 in ( [ au ] ) represent the extension of the function @xmath76 in the isotropic case @xcite and may be written as the symmetric matrix @xmath77 within statics it may also be convenient to work with the temperature dependent vertex functions without introducing the correlation length , as described in the appendix at stage ( [ vertexfunc2 ] ) .
this allows to avoid functions with @xmath51-insertions except for the specific heat . in this case a renormalization for the temperature distances @xmath78 and @xmath79 has to be introduced as performed in @xcite .
in order to obtain renormalization factors which do not contain ratios of the temperature distances a matrix renormalization @xmath80 with the matrices @xmath81 has to be used .
the renormalization factors in the matrix above are obtained by collecting the @xmath23-poles proportional to @xmath82 and @xmath83 in the two vertex functions @xmath84 and @xmath85 .
the matrices in ( [ zrzphi ] ) are related to the renormalization matrix @xmath86 for the @xmath51-insertions by @xmath87 which represents the matrix counterpart of the scalar relation in the isotropic case .
the superscript @xmath88 denotes the transposed matrix .
from the scalar renormalization factors @xmath90 ( @xmath91 ) the @xmath89-functions @xmath92 are derived .
the @xmath93-derivatives , also in the following definitions , always are taken at fixed unrenormalized parameters . in two loop order we obtain the @xmath89-functions @xmath94 from the matrix @xmath86 of the renormalization of the @xmath51-insertions it is convenient to introduce the @xmath89-matrix @xmath95_{ij}(\{u\})=-\sum_l \left(\kappa\frac{d}{d\kappa}[\mbox{\boldmath$z$}_{\phi^2}]_{il}\right ) [ \mbox{\boldmath$z$}_{\phi^2}^{-1}]_{lj}\ ] ] a matrix @xmath96 following from @xmath97 , which has been defined in ( [ zrzphi ] ) , also can be introduced analogous to ( [ zetaphiq ] ) by replacing @xmath86 with @xmath97 . in two loop order
the components of the matrix @xmath98 are @xmath99_{11}&= & \frac{n_\perp+2}{6}u_\perp \left(1-\frac{5}{12}u_\perp\right)-\frac{n_\|}{72}u_\times^2 \\ \label{zetaps12 } [ \mbox{\boldmath$\zeta$}_{\phi^2}]_{12}&= & \frac{n_\perp}{6}u_\times \left(1-\frac{u_\times}{3}\right ) \\
\label{zetaps21 } [ \mbox{\boldmath$\zeta$}_{\phi^2}]_{21}&= & \frac{n_\|}{6}u_\times \left(1-\frac{u_\times}{3}\right ) \\
\label{zetaps22 } [ \mbox{\boldmath$\zeta$}_{\phi^2}]_{22 } & = & \frac{n_\|+2}{6}u_\| \left(1-\frac{5}{12}u_\|\right)-\frac{n_\perp}{72}u_\times^2\end{aligned}\ ] ] the non diagonal elements ( [ zetaps12 ] ) and ( [ zetaps21 ] ) are proportional to @xmath100 and differ only in a prefactor @xmath2 and @xmath1 respectively .
thus we may write @xmath101_{12}=n_\perp u_\times c(\{u\})= \frac{n_\perp}{n_\|}[\mbox{\boldmath$\zeta$}_{\phi^2}]_{21}\ ] ] with a function @xmath102 in two loop order . relation ( [ zrzphi2rel ] ) between @xmath86 and @xmath97 implies @xmath103 where the diagonal matrix @xmath104 has been introduced . from the additive renormalization ( [ amatrix ] ) the function @xmath105_{ij}(\{u\})=\kappa^\varepsilon \sum_{n , p}[\mbox{\boldmath$z$}_{\phi^2}]_{in}[\mbox{\boldmath$z$}_{\phi^2}]_{jp } \nonumber \\ \times\sum_{l
, m}\kappa\frac{d}{d\kappa}\kappa^{-\varepsilon } [ \mbox{\boldmath$z$}_{\phi^2}^{-1}]_{nl}[\mbox{\boldmath$z$}_{\phi^2}^{-1}]_{pm } [ \mbox{\boldmath$a$}]_{lm}(\{u\})\end{aligned}\ ] ] can be introduced which is the extension of the scalar function @xmath106 in the isotropic case ( see ref.@xcite for definitions ) . calculating ( [ bphi2 ] ) in two loop order
we obtain @xmath107 the @xmath50-functions of the four couplings are defined as @xmath108 with @xmath109 . in two loop order
the explicit expressions of the @xmath50-functions are @xmath110 the flow equations of the fourth order couplings @xmath111 are @xmath112 where @xmath113 is the flow parameter .
.fixed points and stability exponents of the @xmath8 model .
[ tab1 ] [ cols="<,<,<,<,<,<,<",options="header " , ] therefore we choose another way to estimate the numerical values of the exponents making use of the fixed @xmath3 expansions .
there , we proceed as follows .
first , based on the two - loop expressions ( [ zetaps11])-([zetaps22 ] ) for the matrix @xmath114_{ij}(\{u\})$ ] ( [ zetaphiq ] ) we find its eigenvalues as expansions in renormalized couplings . in turn , the exponents @xmath115 , @xmath116 are expressed in terms of the corresponding eigenvalues as series in renormalized couplings as well .
corresponding series are found also for the magnetic susceptibility exponents .
the series are then resummed ( as described in the appendix [ resummation ] ) and evaluated at the fps .
numerical results of the exponents @xmath115 , @xmath116 , @xmath117 , and @xmath118 obtained by such evaluation are given in table [ tab3 ] in the ( stable ) biconical fp @xmath119 and , for comparison , in the ( unstable ) heisenberg fp @xmath120 .
the expressions for the exponents @xmath121 and @xmath122 are too short to be resummed .
therefore their numerical values are found from the familiar scaling relations @xmath123 , @xmath124 using resummed values of @xmath125s and @xmath126s .
for the decoupling fp one has to modify the scaling relations according to the statements above and has to use @xmath127 , @xmath124 . note that within the accuracy of calculations results for the exponents that correspond to parallel and perpendicular fields do not differ ( the difference shows up within the fourth digit ) .
the overall agrement with the 5 loop @xmath23-expansion is very good especially for the stable biconical fp .
the resummation of the @xmath50-function has the big advantage to find the fixed point values and asymptotic exponents but in addition the flow of the couplings from their background values to their fp values .
the flow is important for the crossover behavior but also for the discussion of the asymptotic multicritical behavior of physical representatives of such systems with different the background values of the couplings .
their location in the attraction region of a fp defines the critical behavior .. fig .
[ flowres ] shows the flow lines in the space of the coupling parameters for different initial conditions calculated from the resummed @xmath50-functions .
mean field theory states a criterion@xcite which has to be fulfilled by the fourth order couplings for the existence of a tetracritical point ; it reads @xmath128 the flow equations show that @xmath129 is not an invariant surface of the flow .
however it contains several separatrices with the corresponding fixed points : the decoupling fps with @xmath130 ( @xmath131 , @xmath132 , @xmath133 , @xmath134 ) and the heisenberg fp @xmath135 .
therefore in the region shown in fig .
1 the mean field condition @xmath136 represents quite well the surface ( called mfs ) separating the attraction region of the biconical fp @xmath119 and the region of run away flows . thus one can draw the following conclusions .
a system with initial conditions @xmath137 lies outside the attraction region of the stable fb and its runaway flow indicates that a first order transition is expected . one would conclude that the multicritical point is a triple point . systems with initial conditions where roughly @xmath136 would flow to the heisenberg fp @xmath135 indicating that the multicritical point is a bicritical point . then finally if the initial conditions are such that @xmath138 the biconical fp @xmath119 is reached and the multicritical point is a tetracritical point . in this way the three scenarios sketched in ref .
@xcite are realized .
the important and open point is to connect the initial conditions of the field theoretic flow equations to the interaction parameters in the appropriate spin hamiltonian containing the anisotropic interactions defining the antiferromagnetic system .
having available the solutions of the flow equations one can define the effective exponents evaluating the field theoretic @xmath89-functions at the values of the couplings given by the flow according to the definitions of the exponents in eqs .
( [ etaid ] ) and ( [ nu ] ) .
we substitute the couplings obtained from the resummed flow equations into the resummed @xmath89-functions appearing in the expressions for exponents . in this way the effective exponents become functions of the flow parameter @xmath113 , e.g. @xmath139 the effective crossover exponent @xmath140 follows from eq .
( [ crossphi ] ) as @xmath141 results for the exponents are shown in the figs .
[ gammaeff ] , [ nueff ] and [ phieff ] .
the two exponents of the parallel and perpendicular op susceptibilities are almost equal even in the nonasymptotic region but might be quite different from the asymptotic value .
especially due to the slow transient present the values for the flows 1 - 3 are smaller than the expected asymptotic values .
the same holds for the effective exponent of the correlation length @xmath142 .
the flow 6 which does not reach a finite fixed point lies within the expected values .
the formal decrease of the effective exponents to unphysical negative values are due to the flow to infinite values of the couplings .
we have shown that the two loop order perturbation theory together with appropriate resummation techniques is sufficient to calculate the multicritical behavior appearing in systems with @xmath0 symmetry at fixed dimension .
the advantage of such a procedure lies in the accessibility of the corresponding flow equations , which allow a discussion of attraction regions and effective ( crossover ) critical behavior .
we confirmed the shift of the one loop stability borderlines , with the consequence that the multicritical behavior for the case @xmath4 and @xmath5 is characterized by the stable biconical fixed point and not by to heisenberg fixed point .
the discussion of the attraction region of this fixed point leads to the possibility of different phase diagrams depending on the nonuniversal initial parameters entering the flow equations . in a next step
the results will be used to reconsider dynamical critical behavior of the pure relaxational dynamics@xcite of these systems . for the bicritical dynamics of an antiferromagnet in an external magnetic field an extension of the above mentioned model
is necessary since besides the op conserved densities ( cd ) have to be taken into account . in statics
these couplings only appear up to second order terms and can be integrated out . in dynamics
they lead to a coupling of the two dynamic equations .
a complete description of the dynamical multicritical behavior for has not been given in two loop order so far .
finally mode coupling terms have to be taken into account up to two loop order .
acknowledgement : we thank w. selke for useful discussions .
this work was supported by the fonds zur frderung der wissenschaftlichen forschung under project no .
with the static functional ( [ hbicrit ] ) the vertex functions @xmath143 can be calculated in a definite loop order by collecting all one - particle - irreducible contributions . @xmath144 and @xmath145 are the total numbers of @xmath146- and @xmath51- insertions and the indices @xmath147 and @xmath63 indicate if the corresponding insertion is of type @xmath65 or @xmath66 . @xmath148 and @xmath149 act as placeholder for the two temperature distances and three fourth order couplings . between the lower critical dimension @xmath150 and the upper critical dimension @xmath151
the vertex functions ( [ vertexfunc ] ) also contain singularities at @xmath11 .
they have their origin in a non analytical shift of the critical temperature as function of the four point couplings ( for more details see the third refeerence in ref.@xcite ) .
n order to remove them , two parameters @xmath152 and @xmath153 are introduced .
they are determined by @xmath154 this defines to functions @xmath155 and @xmath156 . introducing @xmath157 and rewriting the expressions for ( [ vertexfunc ] ) leads to vertex functions @xmath158 for dimensions larger than @xmath159 all perturbation contributions now have only singularities at least at @xmath68 .
further it may be convenient to introduce the correlation length instead of the temperature distance .
in the present case two correlation lengths @xmath160 have to be introduced . inserting the reversed equations @xmath161 and @xmath162 into ( [ vertexfunc2 ] ) leads to functions @xmath163 introducing the correlation lengths in the vertex functions
is a resummation procedure which removes the expanded contributions of the correlation length from the vertex functions .
this leads to expressions which are simplified considerably .
moreover this is true for the calculations in dynamic models ( see for instance part ii ) . within statics
the two point functions reveal then the general structure @xmath164 in two loop order the @xmath43-independent function @xmath165 is @xmath166 where we have introduced the short notation @xmath167 the @xmath43-dependent function @xmath168 reads in two loop order @xmath169 in order to obtain @xmath170 and @xmath171 one only has to interchange @xmath65 and @xmath66 . the two loop integral @xmath172 is defined as @xmath173 in the limit @xmath174 we have @xmath175 applying the limit to @xmath176 ( see ( [ falphai ] ) and ( [ galphai ] ) ) one obtains @xmath177 and the two point vertex functions reduce in the asymptotic region to @xmath178 because the poles do not depend on @xmath43 , the two functions @xmath179 and @xmath176 contain the same pole terms .
thus the two point functions each with may be renormalized by scalar renormalization factors which remove the poles from the functions @xmath179 .
in this appendix we describe a procedure we use to resum divergent expansions for the two - loop rg functions . starting from the rg function that has a form of truncated polynomial in renormalized couplings : @xmath180 one first represents it in a form of a resolvent series@xcite in variable @xmath181 :
@xmath182 the expansion coefficients @xmath183 in ( [ ap2 ] ) explicitly depend on couplings and coefficient @xmath184 ( [ ap1 ] ) .
obviously , for @xmath185 the function ( [ ap1 ] ) reproduces the initial rg function ( [ ap2 ] ) : @xmath186 .
then , the function ( [ ap2 ] ) is resummed as a single variable function and further evaluated at @xmath185 to recover ( [ ap1 ] ) . to perform the resummation we use the pad - borel technique.@xcite namely , assuming factorial growth of the expansion coefficients
@xmath183 we define the borel transform@xcite of ( [ ap2 ] ) by : @xmath187 where @xmath46 is euler gamma - function .
analytical continuation of function ( [ ap3 ] ) is achieved by representing it in a form of a pad approximant.@xcite in our case , working within a two - loop approximation we use the diagonal [ 1/1 ] pad approximant : @xmath188(t).\ ] ] finally , the resummed function is obtained via an inverse borel transform : @xmath189(t ) e^{-t}.\ ] ] the procedure described above was used to analyze the rg flows and exponents .
note however , that the inverse borel transform ( [ ap5 ] ) is well defined , when no poles in the denominator of pad approximant ( [ ap4 ] ) appear .
otherwise one may estimate its principal value .
the poles do not appear for a sign - alternating series ( as the series for the @xmath50-functions are ) . to deal with sign alternating series during an evaluation of critical exponents
, we have resummed the functions @xmath190 .
99 k .- s .
liu and m. e. fisher , j. low temp .
* 10 * , 655 ( 1972 ) .
d. nelson , j. m. kosterlitz , and m. e. fisher , phys .
lett . * 33 * , 813 ( 1974 ) .
a. aharony and a. d. bruce , phys .
lett . * 33 * , 427 ( 1974 ) .
j. m. kosterlitz , d. nelson , and m. e. fisher , phys .
b * 13 * , 412 ( 1976 ) .
p. calabrese , a. pelissetto , and e. vicari , phys .
b * 67 * , 054505 ( 2003 ) . for an older review
see : y. shapira , in _ multicritical phenomena _ , ( plenum , new york and london , 1983 ) page 35 .
h. rohrer , phys .
lett . * 34 * , 1638 ( 1975 ) .
h. rohrer , and ch .
gerber , phys .
lett . * 38 * , 909 ( 1977 ) .
a. r. king and h. rohrer , phys . rev .
* 19 * , 5864 ( 1979 ) .
r. a. butera et al . , phys .
b * 24 * , 1244 ( 1981 ) .
k. ohgushi and y. ueda , phys .
lett . * 95 * , 217202 ( 2005 ) . c. c. beccera , n. f. oliveira , jr .
a. paduan - filho , w. figueiredo , and m. v. p. souza , phys .
b * 38 * , 6887 ( 1988 )
. j. a. j. basten , e. frikkee , and w. j. m. de jonge , phys .
b * 22 * , 1429 ( 1980 ) .
m. holtschneider , w. selke , and r.leidl , phys . rev . * 72 * , 064443 ( 2005 ) .
w. selke , m. holtschneider , r.leidl , s.wessel , and g.bannasch , _ computer simulation studies in condensed matter physics xxi _ , eds .
d. p. landau , s.p .
lewis , and h. b. schttler ( springer - verlag , heidelberg , berlin , 2008 ) .
d. p. landau and k. binder , phys . rev . * 72 * , 2328 ( 1978 ) . v. dohm , z. phys .
b * 60 * , 61 ( 1985 ) ; r. schloms and v. dohm , europhys .
lett . , * 3 * , 413 ( 1987 ) ; r. schloms and v. dohm , nucl .
b * 328 * , 639 ( 1989 ) .
e. domany , d. mukamel , and m. e. fischer , phys.rev .
b * 15 * , 5432 ( 1977 ) .
s. murakami and n. nagaosa , j. phys .
. jpn . * 69 * , 2395 ( 2000 ) .
j. zinn - justin , _ quantum field theory and critical phenomena ( international series of monographs on physics , 92 ) _
( oxford univ press , 1996 ) ; h. kleinert , v. schulte - frohlinde _ critical properties of @xmath191-theories _ ( singapore : world scientific , 2001 ) .
v. dohm , _ report of the kernforschungsanlage jlich nr . 1578 _ ( 1979 ) .
r. folk and g. moser , j. phys .
* 39 * , r207 ( 2006 ) .
this is certainly true for the non - degenerate @xmath50-functions .
a counterexample is given by the random site ising model .
there , a degeneracy of the one - loop @xmath50-functions leads to appearance of additional fps on the two - loop level , the last are proportional to the @xmath192 . for a review
see e.g. r. folk , yu .
holovatch , and t. yavorskii .
physics - uspiekhi * 46 * , 169 ( 2003 ) [ uspiekhi fizichieskikh nauk * 173 * , 175 ( 2003 ) ] . m. dudka , yu .
holovatch , and t. yavorskii , j. phys .
a , * 37 * , 10727 ( 2004 ) ; b. delamotte , yu .
holovatch , d. ivaneyko , d. mouhanna , and m. tissier , j. stat .
p03014 ( 2008 ) .
h. kleinert , j. neu , v. schulte - frohlinde , k. g. chetyrkin , and s. a. larin , phys .
b * 272 * , 39 ( 1991 ) ; erratum : phys .
b * 319 * , 545 ( 1993 ) .
a. aharony and s. fishman , phys .
* 37 * , 1587 ( 1976 ) ; a. aharony , phys .
lett . * 88 * , 059703 ( 2002 ) .
j. keyley and d.j .
wallace , j. phys .
a * 6 * , 1667 ( 1973 ) .
d. j. amit and y. y. goldschmidt , ann . physics * 114 * , 356 ( 1978 ) .
r. guida and j. zinn - justin , j. phys .
a * 31 * , ( 1998 ) 8103 .
v. dohm and h .- k .
janssen , phys .
* 17 * , 946 ( 1977 ) p. j. s. watson , j. phys . a * 7 * , l167 ( 1974 ) .
g. a. baker , b. g. nickel , m. s. green and d. i. meiron , phys .
rev . lett .
, * 36 * , 1351 ( 1976 ) ; g. a. baker , b. g. nickel and d. i. meiron , phys . rev
. b , * 17 * , 1365 ( 1978 ) .
g. h. hardy , _ divergent series _
( oxford , 1948 ) .
g. a. baker jn and p. graves - morris , _ pad approximants _
( addison - wesley : reading , ma , 1981 ) . | we calculate the static critical behavior of systems of @xmath0 symmetry by renormalization group method within the minimal subtraction scheme in two loop order .
summation methods lead to fixed points describing multicritical behavior .
their stability boarder lines in the space of order parameter components @xmath1 and @xmath2 and spatial dimension @xmath3 are calculated .
the essential features obtained already in two loop order for the interesting case of an antiferromagnet in a magnetic field ( @xmath4 , @xmath5 ) are the stability of the biconical fixed point and the neighborhood of the stability border lines to the other fixed points leading to very small transient exponents .
we are also able to calculate the flow of static couplings , which allows to consider the attraction region .
depending on the nonuniversal background parameters the existence of different multicritical behavior ( bicritical or tetracritical ) is possible including a triple point . |
the theory of discrete analytic functions has been developed in recent years based on the theory of circle packings or circle patterns , which was initiated by thurston s idea of using circle packings as an approximation of the riemann mapping@xcite .
so far many important properties have been established for discrete analytic functions , such as the discrete maximum principle and schwarz s lemma@xcite , the discrete uniformization theorem@xcite , and so forth . for a comprehensive introduction to the theory of discrete analytic functions
, we refer to @xcite .
it is known that certain circle patterns with fixed regular combinatorics admit rich structure .
for example , it has been pointed out that the circle patterns with square grid combinatorics introduced by schramm@xcite and the hexagonal circle patterns@xcite are related to integrable systems .
some explicit examples of discrete analogues of analytic functions have been presented which are associated with schramm s patterns : @xmath1 , @xmath2 , airy function@xcite , @xmath3 , @xmath4@xcite . also , discrete analogues of @xmath3 and @xmath4 associated with hexagonal circle patterns are discussed in @xcite . among those examples
, it is remarkable that the discrete analogue of the power function @xmath3 associated with the circle patterns of schramm type has a close relationship with the sixth painlev equation ( p@xmath5)@xcite .
it is desirable to construct a representation formula for the discrete power function in terms of the painlev transcendents as was mentioned in @xcite .
the discrete power function can be formulated as a solution to a system of difference equations on the square lattice @xmath6 with a certain initial condition . a correspondence between the dependent variable of this system and the painlev transcendents can be found in @xcite , but the formula seems somewhat indirect .
agafonov has constructed a formula for the radii of circles of the associated circle pattern at some special points on @xmath7 in terms of the gauss hypergeometric function@xcite . in this paper
, we aim to establish an explicit representation formula of the discrete power function itself in terms of the hypergeometric @xmath0 function of p@xmath5 which is valid on @xmath8 and for @xmath9 .
based on this formula , we generalize the domain of the discrete power function to a discrete analogue of the riemann surface . on the other hand ,
the fact that the discrete power function is related to p@xmath5 has been used to establish the immersion property@xcite and embeddedness@xcite of the discrete power function with real exponent .
although we can not expect such properties and thus the correspondence to a certain circle pattern for general complex exponent , we have found a special case of @xmath10 where the discrete power function is an immersion .
another purpose of this paper is to prove the immersion property of this case .
this paper is organized as follows . in section 2
, we give a brief review of the definition of the discrete power function and its relation to p@xmath5 . the explicit formula for the discrete power function
is given in section 3 .
we discuss the extension of the domain of the discrete power function in section 4 . in section 5 , we show that the discrete power function for @xmath11 is an immersion .
section 6 is devoted to concluding remarks .
for maps , a discrete analogue of conformality has been proposed by bobenko and pinkall in the framework of discrete differential geometry@xcite .
[ d - conformal ] a map @xmath12 is called discrete conformal if the cross - ratio with respect to every elementary quadrilateral is equal to @xmath13 : @xmath14 the condition ( [ cr ] ) is a discrete analogue of the cauchy - riemann relation . actually , a smooth map @xmath15 is conformal if and only if it satisfies @xmath16 for all @xmath17 .
however , using definition [ d - conformal ] alone , one can not exclude maps whose behaviour is far from that of usual holomorphic maps .
because of this , an additional condition for a discrete conformal map has been considered@xcite .
a discrete conformal map @xmath18 is called embedded if inner parts of different elementary quadrilaterals @xmath19 do not intersect .
an example of an embedded map is presented in figure [ embedded ] .
this condition seems to require that @xmath20 is a univalent function in the continuous limit , and is too strict to capture a wide class of discrete holomorphic functions .
in fact , a relaxed requirement has been considered as follows@xcite .
a discrete conformal map @xmath18 is called immersed , or an immersion , if inner parts of adjacent elementary quadrilaterals @xmath19 are disjoint .
see figure [ immersed ] for an example of an immersed map .
let us give the definition of the discrete power function proposed by bobenko@xcite .
[ def : dpower1 ] let @xmath21 be a discrete conformal map .
if @xmath18 is the solution to the difference equation @xmath22 with the initial conditions @xmath23 for @xmath24 , then we call @xmath25 a discrete power function . the difference equation ( [ eq - f ] )
is a discrete analogue of the differential equation @xmath26 for the power function @xmath27 , which means that the parameter @xmath28 corresponds to the exponent of the discrete power function .
it is easy to get the explicit formula of the discrete power function for @xmath29 ( or @xmath30 ) . when @xmath29 , ( [ eq - f ] ) is reduced to a three - term recurrence relation . solving it with the initial condition @xmath31
, we have @xmath32 \displaystyle\prod_{k=1}^l\dfrac{2k+\gamma}{2k-\gamma}&(n=2l+1 ) , \end{array } \right.\ ] ] for @xmath33 . when @xmath34 ( or @xmath35 ) , agafonov has shown that the discrete power function can be expressed in terms of the hypergeometric function@xcite .
one of the aims of this paper is to give an explicit formula for the discrete power function @xmath18 for arbitrary @xmath36 . [ cols="^,^ " , ]
agafonov and bobenko have shown that the discrete power function for real @xmath28 is an immersion and thus defines a circle pattern of schramm type .
we have generalized the discrete power function to complex @xmath28 .
it is natural to ask whether there are other cases where the discrete power function is associated with circle patterns .
we have the following result : [ thm : immersion ] the mapping @xmath37 satisfying ( [ cr ] ) and ( [ eq - f ] ) with the initial condition @xmath38 is an immersion when @xmath39 . in this section ,
we give the proof of theorem [ thm : immersion ] along with the discussion in @xcite .
we also use the explicit formulae given in previous sections . setting @xmath40 we associate the discrete power function with circle patterns of schramm type .
the proof of theorem [ thm : immersion ] is then reduced to properties of the radii of those circles . [
lem : equidistance ] a discrete power function @xmath18 defined by ( [ cr ] ) and ( [ eq - f ] ) with initial condition @xmath41 for arbitrary @xmath42 has the equidistant property @xmath43 for any @xmath44 .
moreover , if and only if @xmath39 , we have latexmath:[\[\label{equi_f:2 } @xmath44 . *
* proof.**by using the formulae in theorem [ formula_f ] ( or proposition [ formula_v ] ) , we have @xmath46 which proves ( [ equi_f:1 ] ) .
we also have @xmath47 putting @xmath48 , we obtain @xmath49 which implies @xmath50 . using the first equation of ( [ equi_f:1 ] ) , we see that the first equation of ( [ equi_f:2 ] ) follows .
the second equation of ( [ equi_f:2 ] ) can be proved in a similar manner .
suppose that ( [ equi_f:2 ] ) holds , then from ( [ equi_f:3 ] ) we have @xmath51 which leads us to @xmath39.@xmath52 [ prop : kite ] let @xmath18 satisfy ( [ cr ] ) and ( [ eq - f ] ) in @xmath53 with initial condition ( [ initial_f ] ) . then all the elementary quadrilaterals @xmath19 are of the kite form , namely , all edges at each vertex @xmath18 with @xmath54 are of the same length , latexmath:[\[|f_{n+1,m}-f_{n , m}|=|f_{n , m+1}-f_{n , m}|= the neighbouring edges at the vertex @xmath18 with @xmath56 are equal to @xmath57 . *
* proof.**for three complex numbers @xmath58 ( @xmath59 ) , we introduce a notation @xmath60 = \arg \frac{z_1-z_2}{z_3-z_2}.\ ] ] we first consider the quadrilateral @xmath61 . notice that @xmath62 which implies that @xmath63=\frac{\pi}{2}$ ] .
then it follows from ( [ cr ] ) that @xmath64=\frac{\pi}{2}$ ] , @xmath65 and @xmath66 .
we next consider the quadrilateral @xmath67 where , from lemma [ lem : equidistance ] , we have @xmath68 .
we see from ( [ cr ] ) that @xmath69=[f_{1,0},f_{2,0},f_{2,1}]=\frac{\pi}{2}$ ] and @xmath70 . from lemma [ lem : equidistance ] and
@xmath71=\frac{\pi}{2}$ ] , we see that @xmath72=\frac{\pi}{2}$ ] .
then a similar argument can be applied to the quadrilateral @xmath73 and so forth . in this manner ,
proposition [ prop : kite ] is proved inductively.@xmath52 from proposition [ prop : kite ] , it follows that the circumscribed circles of the quadrilaterals @xmath74 with @xmath54 form a circle pattern of schramm type @xcite , namely , the circles of neighbouring quadrilaterals intersect orthogonally and the circles of half - neighbouring quadrilaterals with a common vertex are tangent ( see figure [ fig : f_config_init ] ) .
conversely , for a given circle pattern of schramm type , it is possible to construct a discrete conformal mapping @xmath18 as follows .
let @xmath75 be a circle pattern of schramm type on the complex plane .
define @xmath76 in the following manner : * if @xmath54 , then @xmath18 is the center of @xmath77 . * if @xmath56 , then @xmath78 . by construction
, it follows that all elementary quadrilaterals @xmath19 are of the kite form whose angles between the edges with different lengths are @xmath79 .
therefore , ( [ cr ] ) is satisfied automatically . in what follows ,
the function @xmath18 , defined by ( a ) and ( b ) , is called a discrete conformal map corresponding to the circle pattern @xmath80 .
we now use the the radii of corresponding circle patterns to characterize the necessary and sufficient condition that the discrete power function is an immersion .
[ thm : immersion_r ] let @xmath18 satisfying ( [ cr ] ) and ( [ eq - f ] ) with initial condition ( [ initial_f ] ) be an immersion
. then @xmath81 defined by @xmath82 satisfies @xmath83 and @xmath84r_{n , m+1}(r_{n+1,m}+r_{n-1,m } ) + nr_{n+1,m}(r_{n , m+1}^2-r_{n+1,m}r_{n-1,m } ) } { [ ( m+1)r_{n+1,m}-\delta r_{n , m+1}](r_{n+1,m}+r_{n-1,m } ) -n(r_{n , m+1}^2-r_{n+1,m}r_{n-1,m } ) } , \label{eq_2}\ ] ] for @xmath85 .
conversely , let @xmath86 satisfy ( [ eq_1 ] ) and ( [ eq_2 ] )
. then @xmath81 defines an immersed circle pattern of schramm type .
the corresponding discrete conformal map @xmath18 is an immersion and satisfies ( [ eq - f ] ) .
* proof . *
the proof of theorem [ thm : immersion_r ] occupies the remainder of this subsection .
suppose that the discrete power function @xmath18 is immersed . for @xmath56 , we may parametrize the edges around the vertex @xmath18 as @xmath87 where @xmath88 ( @xmath89 ) are the radii of the corresponding circles , since all the angles around @xmath18 are @xmath79 . the constraint ( [ eq - f ] )
reads @xmath90 [ lem : kite ] for @xmath56 we have : @xmath91 * proof . *
the kite form of the quadrilateral @xmath19 implies @xmath92}$ ] ( see figure [ fig : f_config_fr ] ) . the first equation of ( [ eqn : kite1 ] ) follows by noticing that @xmath93=\frac{r_2}{r_1}$ ] .
the second equation is derived by a similar consideration on the quadrilateral @xmath94.@xmath52 setting @xmath95 and substituting ( [ eqn : edges_param1])([eqn : kite1 ] ) into ( [ eq - f ] ) at the point @xmath96 , one arrives at @xmath97 the real part of ( [ eq_r ] ) gives @xmath98 which coincides with ( [ eq_1 ] ) .
now we parametrize the edges around the vertex @xmath99 with @xmath100 as @xmath101 & f_{n , m+1}=f_{n+1,m+1}-r_2e^{i\beta'},\quad f_{n+1,m}=f_{n+1,m+1}-ir_1e^{i\beta'}. \end{split}\ ] ] from the first equation of ( [ eqn : kite1 ] ) and noticing that all angles around the vertex @xmath99 are @xmath79 , we have the following relation between @xmath102 and @xmath103 : @xmath104 one can express @xmath105 in two ways as ( see figure [ fig : f_config_fr ] ) @xmath106 and @xmath107 the compatibility implies @xmath108 then , from the imaginary part of ( [ eq_r ] ) , one obtains @xmath109 or solving ( [ eqn : nmr ] ) with respect to @xmath110 @xmath111r_{n+1,m}(r_{n , m+1}+r_{n , m-1 } ) + mr_{n , m+1}(r_{n+1,m}^2-r_{n , m+1}r_{n , m-1 } ) } { [ ( n+1)r_{n , m+1}+\delta r_{n+1,m}](r_{n , m+1}+r_{n , m-1 } ) -m(r_{n+1,m}^2-r_{n , m+1}r_{n , m-1})}. \label{(2,1)}\ ] ] we may rewrite ( [ eq - f ] ) at @xmath112 in terms of @xmath113 ( @xmath89 ) @xmath114 , @xmath115 as @xmath116 & = i\dfrac{r_1-ir_2}{r_1+ir_2}\left [ 2(n+1)\dfrac{r_2r_1}{r_2+r_1}+2i(m+1)\dfrac{r_1r_2}{r_1+r_2}\right ] .
\end{split}\ ] ] eliminating @xmath28 from ( [ eq_r ] ) and ( [ eqn : nmr2 ] ) , we get @xmath117 we then eliminate @xmath114 using ( [ eqn : nmr ] ) to obtain @xmath118r_2(r_1+r_3)+nr_1(r_2 ^ 2-r_1r_3 ) } { [ ( m+1)r_1-\delta r_2](r_1+r_3)-n(r_2 ^ 2-r_1r_3)},\ ] ] which coincides with ( [ eq_2 ] ) .
this proves the first part of theorem [ thm : immersion_r ] . to prove the second part
, we use the following lemma .
let @xmath119 satisfy ( [ eq_1 ] ) and ( [ eq_2 ] ) .
then , @xmath120 satisfies ( [ ( 2,1 ) ] ) and @xmath121(r_{n , m+1}+r_{n , m-1})\\[2 mm ] -m(r_{n+1,m}^2-r_{n , m+1}r_{n , m-1})(r_{n , m-1}+r_{n+2,m-1})=0 . \end{array } \label{(2,-1)}\ ] ] * * proof.**substituting ( [ eq_1 ] ) at @xmath122 and at @xmath112 into ( [ eq_2 ] ) to eliminate @xmath123 and @xmath124 , we get ( [ ( 2,1 ) ] ) . substituting ( [ eq_2 ] ) at @xmath125 into ( [ ( 2,1 ) ] ) , we get ( [ ( 2,-1 ) ] ) under the condition @xmath126 which can be verified from the compatibility with ( [ eq_1 ] ) and ( [ eq_2 ] ) .
@xmath52 eliminating @xmath127 from ( [ ( 2,1 ) ] ) and ( [ ( 2,-1 ) ] ) , we get @xmath128 in @xcite , it was proven that , given @xmath81 satisfying ( [ 5pts ] ) , the circle pattern with radii of the circles @xmath81 is immersed .
thus , the corresponding discrete conformal map @xmath18 is an immersion .
let us finally show that the discrete conformal map @xmath18 satisfies ( [ eq - f ] ) .
putting @xmath29 in ( [ eq_1 ] ) , we have @xmath129 .
this means that @xmath130 for any @xmath131 . by using the ambiguity of translation and scaling of the circle pattern
, one can set @xmath31 without loss of generality , and set @xmath132 putting @xmath133 in ( [ ( 2,1 ) ] ) we have @xmath134 a geometric consideration leads us the following lemma : [ angles ] we have @xmath135 * proof . *
.,title="fig:",scaledwidth=95.0% ] +
( i ) @xmath136 .,title="fig:",scaledwidth=95.0% ] + ( ii ) @xmath137 first , we consider the case of @xmath138 , see figure [ fig : delta_theta1 ] ( i ) . from @xmath139
we obtain @xmath140 which yields @xmath141=0.\ ] ] since @xmath142 , we have @xmath143 and @xmath144 thus we get ( [ theta ] ) from ( [ delta_r ] ) .
when @xmath145 , the configuration of points is shown in figure [ fig : delta_theta1 ] ( ii ) .
the equality @xmath139 implies @xmath146 which gives the same result as the case of @xmath138 .
let us investigate the case of @xmath147 , see figure [ fig : delta_theta2 ] ( i ) .
since @xmath148 we have @xmath149 which leads us to @xmath150=0.\ ] ] .,title="fig:",scaledwidth=75.0% ] + ( i ) @xmath151 .,title="fig:",scaledwidth=70.0% ] + ( ii ) @xmath152 then we have @xmath153 , and thus ( [ tan_theta ] ) .
figure [ fig : delta_theta2 ] ( ii ) illustrates the case of @xmath154 .
we see from @xmath155 that @xmath156 which also leads us to ( [ eq_h ] ) , and thus ( [ tan_theta ] ) .
therefore we have proved lemma [ angles].@xmath52 from ( [ f_and_theta ] ) , ( [ theta ] ) and the initial condition @xmath31 , we see by induction that the points @xmath157 satisfy @xmath158 similarly , we see that @xmath159 satisfy @xmath160 thus it is possible to determine @xmath18 in @xmath161 by using ( [ cr ] ) . since ( [ cr ] ) is compatible with ( [ eq - f ] ) , @xmath18 satisfies ( [ eq - f ] ) simultaneously .
this proves the second part of theorem [ thm : immersion_r].@xmath52 theorem [ thm : immersion_r ] claims that if @xmath81 satisfying ( [ eq_1 ] ) and ( [ eq_2 ] ) is positive , then the corresponding @xmath18 is an immersion . in order to establish theorem [ thm : immersion ]
, we have to prove the positivity of @xmath81 determined by ( [ eq_1 ] ) and ( [ eq_2 ] ) with the initial condition @xmath162 , @xmath163 .
first , we show that positivity of @xmath164 for @xmath85 is reduced to that of @xmath165 for @xmath33 .
[ positivity ] let the solution @xmath81 of ( [ eq_1 ] ) for @xmath85 and ( [ eq_2 ] ) for @xmath166 with initial data @xmath167 be positive for @xmath168 , namely , @xmath169 for any @xmath170 . then @xmath81 is positive everywhere in @xmath171 and satisfies ( [ eq_2 ] ) for @xmath85 . * * proof.**equation ( [ eq_2 ] ) for @xmath172 with initial data ( [ init_r ] ) determines @xmath173
. we use ( [ eq_1 ] ) and ( [ eq_2 ] ) for @xmath174 inductively to get @xmath165 and @xmath175 . as was mentioned before , we see that @xmath176 and @xmath177 for all @xmath178 by putting @xmath30 and @xmath29 , respectively , in ( [ eq_1 ] ) . with these data
one can determine @xmath81 in @xmath171 by using ( [ eq_1 ] ) .
when @xmath179 , we use ( [ eq_1 ] ) in the form of @xmath180 for positive @xmath181 and @xmath182 , we get @xmath183 .
when @xmath184 , one can show in a similar way that @xmath185 for given positive @xmath186 and @xmath123 by using ( [ eq_1 ] ) .
one can show by induction that we have ( [ ( 2,-1 ) ] ) for @xmath187 , and we get ( [ eq_2 ] ) for @xmath188 .
similarly , one can show by induction that we have ( [ ( 2,-1 ) ] ) at @xmath189 for @xmath190 , and ( [ eq_2 ] ) for @xmath189 .
thus we obtain ( [ eq_2 ] ) for @xmath179 .
one can show in a similar way that we have ( [ eq_2 ] ) for @xmath191 by using ( [ ( 2,-1 ) ] ) at @xmath192 as an auxiliary relation .
@xmath52 due to proposition [ positivity ] , the discrete function @xmath193 with @xmath39 is an immersion if and only if @xmath169 for all @xmath170 .
we next reduce the positivity to the existence of unitary solution to a certain system of difference equations .
[ prop : xy ] the map @xmath194 satisfying ( [ cr ] ) and ( [ eq - f ] ) with the initial condition @xmath195 is an immersion if and only if the solution @xmath196 to the system of equations @xmath197 \dfrac{\gamma}{2 } = \dfrac{n+2}{1-x_{n+1}^{-1}y_{n+1}^{-1 } } -\dfrac{n+1}{1+x_n^{-1}y_{n+1}},\quad x_0y_0=\dfrac{\gamma}{\gamma-2 } , \end{array } \label{alpha&varphi}\ ] ] with @xmath198 is of the form @xmath199 , where @xmath200 . * * proof.**let @xmath18 be an immersion .
define @xmath200 through @xmath201 using proposition [ prop : kite ] , one obtains @xmath202 f_{n , n+2}-f_{n , n+1}=r_{n , n+1}e^{i(2\varphi_n-\pi/2 + 2\aa_n+\beta_n ) } , & f_{n , n+1}-f_{n , n}=r_{n , n+1}e^{i(\beta_n+\pi/2 ) } , \end{array}\ ] ] where @xmath203 . figure [ fig : f_configuration ] is a schematic diagram illustrating the configuration of the relevant quadrilaterals . .
] now the constraint ( [ eq - f ] ) for @xmath204 is equivalent to @xmath205 putting these expressions into the equality @xmath206 together with @xmath207 and @xmath208 , one obtains @xmath209 + 2\cos\aa_n\left(\dfrac{n}{1+e^{2i(\varphi_n-\aa_{n-1 } ) } } + \dfrac{n+1}{e^{-2i\aa_n}-e^{2i\varphi_n}}\right)=-\gamma e^{i\aa_n}. \end{array } \label{alpha&varphi'}\ ] ] on the other hand , equation ( [ eq_2 ] ) for @xmath210 is reduced to @xmath211 by a similar geometric consideration to the proof of lemma [ angles ] , this implies @xmath212 which can be transformed to @xmath213 by using ( [ alpha&varphi ] ) , we see that ( [ alpha&varphi ] ) yields the first equation of ( [ alpha&varphi ] ) with @xmath214 and @xmath215 .
the second equations of ( [ alpha&varphi ] ) come from ( [ alpha&varphi ] ) .
this proves the necessity part .
now let us suppose that there is a solution @xmath216 of ( [ alpha&varphi ] ) with @xmath200 .
this solution together with ( [ alpha_varphi ] ) and ( [ cr ] ) determines a sequence of orthogonal circles with their centers on @xmath217 , and thus the points @xmath218 .
now ( [ cr ] ) determines @xmath18 on @xmath161 . since @xmath200 ,
the inner parts of the quadrilaterals @xmath219 and of the quadrilaterals @xmath220 are disjoint , which means that we have positive solution @xmath165 and @xmath175 of ( [ eq_1 ] ) and ( [ eq_2 ] ) .
given @xmath175 and @xmath165 , ( [ eq_1 ] ) determines @xmath81 for all @xmath85 . due to proposition [ positivity ] ,
@xmath81 is positive , and satisfies ( [ eq_1 ] ) and ( [ eq_2 ] ) . theorem [ thm : immersion_r ] implies that the discrete conformal map @xmath221 corresponding to the circle pattern @xmath80 determined by @xmath81 is an immersion and satisfies ( [ eq - f ] ) . since @xmath222 and @xmath223 , equation ( [ cr ] ) implies @xmath224 .
this proves proposition [ prop : xy ] .
@xmath52 note that although ( [ alpha&varphi ] ) is a system of equations , a solution @xmath196 of ( [ alpha&varphi ] ) is determined by its initial value @xmath225 . the system of equation ( [ alpha&varphi ] ) can be written in the following recurrent form : @xmath226},\ ] ] or @xmath227 @xmath228 and @xmath229 .
it is easy to see that @xmath230 when @xmath231 and that @xmath232 when @xmath233 , which implies that this system possesses unitary solutions .
moreover , we have the following theorem as for the arguments of the unitary solutions : there exists a unitary solution @xmath196 to the system of equation ( [ alpha&varphi ] ) with @xmath234 , where @xmath235\}.\ ] ] * * proof**we first investigate the properties of the function @xmath236 and @xmath237 restricted to the torus @xmath238 . * property 1 . *
the function @xmath236 is continuous on @xmath239 .
the function @xmath237 is continuous on @xmath240 for any @xmath170 .
( continuity on the boundary of @xmath240 is understood to be one - sided . )
the points of discontinuity must satisfy @xmath241 the first identity holds only for @xmath242 .
the second never holds for unitary @xmath243 . * property 2 . * for @xmath244 , we have @xmath245 . for @xmath246 , we have @xmath247 , where @xmath248\}$ ] and @xmath249 .
property 2 is verified as follows : using the transformation @xmath250 where @xmath251and @xmath215 , we see that the first equation of ( [ alpha&varphi ] ) takes the form @xmath252 it is obvious that @xmath253 $ ] when @xmath254\times[0,+\infty]\bs\{(0,0),(+\infty,+\infty)\}$ ] . the second equation of ( [ alpha&varphi ] )
can be expressed as @xmath255 by using the variables @xmath256 , we have @xmath257 where @xmath258u-(2n+3)v+\delta(1+uv ) } { ( n+2)-(n+1)v^2+(2n+3)uv+\delta v(1+uv)}.\ ] ] [ lem : alpha_n+1 ] it holds that @xmath259 $ ] for @xmath260 $ ] .
* * proof.**let us investigate the function @xmath261 for @xmath262 $ ] .
it is easy to see that @xmath263 on @xmath264 ^ 2 $ ] except for the points satisfying @xmath265 .
consider the values of @xmath261 on the boundary of @xmath264 ^ 2 $ ] .
it is easy to see that @xmath266 we find that @xmath267 except for the point @xmath268 , and that @xmath269 is monotone increasing .
note that @xmath270 when @xmath271 , we see that @xmath272 except for the point @xmath273 and that @xmath274 except for the point @xmath275 .
when @xmath276 , we see that @xmath272 and that @xmath274 except for the points @xmath277 .
the singular points of @xmath261 can be expressed by @xmath278 , where @xmath279 , if @xmath280 .
note that @xmath254 ^ 2 $ ] .
then we find that the singular points lie in @xmath281,u\in[0,(n+1)/\dl]$ ] ( when @xmath271 ) or in @xmath282,u\in[0,+\infty]$ ] ( when @xmath276 ) , and that @xmath283 is monotone increasing ( see figure [ fig : fuv ] ) .
therefore we see that @xmath284 $ ] when @xmath260 $ ] .
@xmath52 . ]
the final equation in ( [ alpha&varphi ] ) leads us to @xmath285 it is easy to see that @xmath286 $ ] when @xmath287 .
therefore property 2 is established .
now let us introduce @xmath288 s_{\rm iv}(k):=\{y_0\in a_{\rm i}\,| \,x_k\in a_{\rm iv},\,x_l\in a_{\rm i}\,(l < k)\ } , \end{array}\ ] ] where @xmath196 is the solution of ( [ alpha&varphi ] ) . from property 1
, it follows that @xmath289 and @xmath290 are open sets in the induced topology of @xmath291 .
denote @xmath292 which are also open .
these sets are nonempty since @xmath293 and @xmath294 are nonempty ( see the proof of lemma [ lem : alpha_n+1 ] ) .
note that @xmath295 or @xmath296 can be empty .
finally , introduce @xmath297 it is obvious that @xmath298 and @xmath299 are mutually disjoint .
property 2 implies @xmath300 since the connected set @xmath291 can not be covered by two open disjoint subsets @xmath301 and @xmath299 , we see that @xmath302 .
so there exists @xmath225 such that the solution @xmath303 for any @xmath170 .
suppose that @xmath304 hold at a certain @xmath305 .
then we get @xmath306 and @xmath307 from ( [ eqn : xy ] ) , or ( [ ev : v ] ) and ( [ ev : alpha ] ) . similarly ,
if @xmath308 hold at a certain @xmath305 , we get @xmath309 and then @xmath310 .
it means that in both cases @xmath311 .
suppose that @xmath312 hold at a certain @xmath305 .
then we get @xmath313 and then @xmath310 from ( [ eqn : xy ] ) , or ( [ ev : v ] ) and ( [ ev : alpha ] ) . similarly , if @xmath314 hold at a certain @xmath305 , we get @xmath315 and then @xmath307 .
it also means that in both cases @xmath316 .
thus , it follows that @xmath317 and @xmath318 for the solution @xmath303 for any @xmath170 .
@xmath52 we have shown that @xmath319 is not empty . in order to establish theorem [ thm : immersion ] ,
let us finally show the uniqueness of the initial condition which gives rise to the solution @xmath320 , namely , the circle pattern .
indeed , the initial condition is nothing but that for the discrete power function .
take a solution @xmath196 such that @xmath321 and consider the corresponding circle pattern .
let @xmath322 be radii of circles with centers at @xmath323 , i.e. , @xmath324 .
we have the following lemma .
an explicit formula for @xmath322 is given by @xmath325-i\zeta \left[\left(\frac{r - k+1}{2}\right)_k+\left(\frac{r - k}{2}\right)_k\right ] } { ( -1)^k \left[\left(\frac{r - k+1}{2}\right)_k+\left(\frac{r - k}{2}\right)_k\right]-i\zeta \left[\left(\frac{r - k+1}{2}\right)_k-\left(\frac{r - k}{2}\right)_k\right]},\ ] ] where @xmath326 .
* * proof**the radii @xmath322 are defined by @xmath327 ( see ( [ f - v ] ) ) . from proposition
[ formula_v ] , we have @xmath328 where @xmath329 is given by ( see ( [ gauss ] ) ) @xmath330 and @xmath331 it is easy to see that @xmath332 under @xmath10 . due to the formulae@xcite @xmath333 \begin{split } & f(a , b , a - b+2;-1)=2^{-a}\pi^{1/2}(b-1)^{-1}\gamma(a - b+2)\\ & \hskip40mm\times\left[\dfrac{1}{\gamma(\hf a)\gamma(\hf a - b+\frac{3}{2 } ) } -\dfrac{1}{\gamma(\hf a+\hf)\gamma(\hf a - b+1)}\right ] , \end{split}\end{aligned}\ ] ] and the contiguity relation @xmath334f(a , b , c;t)\\ & \hskip40 mm + c(c-1)(t-1)f(a , b , c-1;t)=0 , \end{split}\ ] ] we get @xmath335-i\zeta \left[\left(\frac{r - k+1}{2}\right)_k+\left(\frac{r - k}{2}\right)_k\right ] } { ( -1)^k \left[\left(\frac{r - k+1}{2}\right)_k+\left(\frac{r - k}{2}\right)_k\right]-i\zeta \left[\left(\frac{r - k+1}{2}\right)_k-\left(\frac{r - k}{2}\right)_k\right]}.\ ] ] then we can easily verify that @xmath336 satisfy the recurrence relation @xmath337 where @xmath338 . on the other hand , ( [ ( 2,1 ) ] ) for @xmath29 is reduced to @xmath339 then noticing @xmath340 , we see that @xmath341 . @xmath52 [ unique ] the set @xmath319 consists of only one element @xmath342 . *
* proof**by using the formulae @xmath343 and the asymptotic formula as @xmath344 @xmath345 we get @xmath346 for large @xmath347 . ] the relation ( [ theta ] ) implies @xmath348 .
figure [ fig : f_asymptotic ] illustrates the configuration around @xmath349 for sufficiently large @xmath347 .
note that the three points @xmath349 , @xmath350 , @xmath351 are asymptotically collinear .
we see from ( [ eqn : lim_r ] ) that @xmath322 should behave as @xmath352 irrespective of the parity of @xmath347 , since the four points @xmath353 form a quadrilateral .
then we find that @xmath354 .
this completes the proof of proposition [ unique ] .
@xmath52 therefore @xmath18 satisfying ( [ cr ] ) and ( [ eq - f ] ) is an immersion if and only if the initial condition is given by ( [ initial_f_immsesion ] ) .
this completes the proof of theorem [ thm : immersion ] .
the discrete logarithmic function and cases where @xmath355 were excluded from the considerations in the previous sections . from the viewpoint of the theory of hypergeometric functions , these cases lead to integer differences in the characteristic exponents .
thus we need a different treatment for the precise description of these cases .
however , they may be obtained by some limiting procedures in principle .
in fact , agafonov has examined the case where @xmath356 and @xmath357 by using a limiting procedure@xcite , the former is the discrete power function @xmath358 and latter is the discrete logarithmic function . in general , one may obtain a description of these cases by introducing the functions @xmath359 and @xmath360 as @xmath361 \displaystyle\lim_{r\to - j } \dfrac{(-r+1)_j}{(r+1)_j}f_{n , m } , & \mbox{for $ \gamma=-2j\in 2\mathbb{z}_{<0}$ } \end{array } \right.\ ] ] and @xmath362 respectively . the function @xmath359 might coincide with the counterpart defined in section 6 of @xcite . moreover , it has been shown that the discrete power function and logarithmic function associated with hexagonal patterns are also described by some discrete painlev equations@xcite .
it may be an interesting problem to construct the explicit formula for them .
the authors would like to express our sincere thanks to professor masaaki yoshida for valuable suggestions and discussions .
this work was partially supported by jsps grant - in - aid for scientific research no .
21740126 , 21656027 and 23340037 , and by the global coe program education and research hub for mathematics - for - industry from the ministry of education , culture , sports , science and technology , japan . addtoresetequationsection
as a preparation , we give a brief review of the bcklund transformations and some of the bilinear equations for the @xmath0 functions@xcite .
it is well - known that p@xmath5 ( [ p6 ] ) is equivalent to the hamilton system @xmath363 whose hamiltonian is given by @xmath364f_2 + \alpha_2(\alpha_1+\alpha_2)f_0.\ ] ] here @xmath365 and @xmath366 are defined by @xmath367 and @xmath368 with @xmath369 .
the bcklund transformations of p@xmath5 are described by @xmath370 @xmath371 @xmath372 s_6 : & \alpha_0\leftrightarrow\alpha_3,\quad\alpha_1\leftrightarrow\alpha_4,\quad f_2\mapsto -\dfrac{f_4(f_4f_2+\alpha_2)}{t},\quad f_4\mapsto\dfrac{t}{f_4},\\[4 mm ] s_7 : & \alpha_0\leftrightarrow\alpha_4,\quad\alpha_1\leftrightarrow\alpha_3,\quad f_2\mapsto\dfrac{f_3(f_3f_2+\alpha_2)}{t-1},\quad f_4\mapsto\dfrac{f_0}{f_3 } , \end{array}\ ] ] where @xmath373 is the cartan matrix of type @xmath374 .
then the group of birational transformations @xmath375 generate the extended affine weyl group @xmath376 .
in fact , these generators satisfy the fundamental relations @xmath377 and @xmath378 s_5s_6=s_6s_5,\quad s_5s_7=s_7s_5,\quad s_6s_7=s_7s_6 . \end{array}\ ] ] we add a correction term to the hamiltonian @xmath379 as follows , @xmath380 + \dfrac{1}{4 }
\left[(\alpha_1+\alpha_4)^2+(\alpha_3+\alpha_4)^2 + 4\alpha_2\alpha_4\right].\ ] ] this modification gives a simpler behavior of the hamiltonian with respect to the bcklund transformations . from the corrected hamiltonian , we introduce a family of hamiltonians @xmath381 as @xmath382 next , we also introduce @xmath0 functions @xmath383 by @xmath384 . imposing the condition that the action of the @xmath385 s on the @xmath0 functions also commute with the derivation @xmath386 , one can lift the bcklund transformations to the @xmath0 functions .
the action of @xmath376 is given by @xmath387 and @xmath388^{\frac{1}{4}}\tau_1,\quad \tau_1\mapsto [ t(t-1)]^{-\frac{1}{4}}\tau_0,\\[1 mm ] & \tau_3\mapsto t^{-\frac{1}{4}}(t-1)^{\frac{1}{4}}\tau_4,\quad \tau_4\mapsto t^{\frac{1}{4}}(t-1)^{-\frac{1}{4}}\tau_3,\quad \tau_2\mapsto [ t(t-1)]^{-\frac{1}{2}}f_0\tau_2 , \end{array}\ ] ] @xmath389 @xmath390 & \tau_1\mapsto(-1)^{\frac{3}{4}}(t-1)^{-\frac{1}{4}}\tau_3,\quad \tau_3\mapsto(-1)^{-\frac{3}{4}}(t-1)^{\frac{1}{4}}\tau_1,\\[1 mm ] & \tau_2\mapsto -i(t-1)^{-\frac{1}{2}}f_3\tau_2 .
\end{array}\ ] ] we note that some of the fundamental relations are modified @xmath391 and @xmath392 s_5s_7\tau_{\{0,1,2,3,4\}}=\{i ,- i,-1,i ,- i\}s_7s_5\tau_{\{0,1,2,3,4\}},\\[1 mm ] s_6s_7\tau_{\{0,1,2,3,4\}}=\{-i ,- i,-1,i , i\}s_7s_6\tau_{\{0,1,2,3,4\}}. \end{array}\ ] ] let us introduce the translation operators @xmath393 whose action on the parameters @xmath394 is given by @xmath395 \widehat{t}_{40}(\vec{\alpha})=\vec{\alpha}+(-1,0,0,0,1),\\[1 mm ] \widehat{t}_{34}(\vec{\alpha})=\vec{\alpha}+(0,0,0,1,-1),\\[1 mm ] t_{14}(\vec{\alpha})=\vec{\alpha}+(0,1,-1,0,1 ) .
\end{array}\ ] ] we denote @xmath396 . by using this notation
, we have @xmath397^{\frac{1}{4}}\tau_1,\\[1 mm ] \tau_{0,-1,-1,0}=(-1)^{-\frac{3}{4}}t^{\frac{1}{4}}\tau_3 , & \tau_{0,-1,0,0}=(-1)^{-\frac{3}{4}}(t-1)^{\frac{1}{4}}\tau_4,\\[1 mm ] \tau_{-1,-2,-1,1}=(-1)^{-\frac{1}{4}}s_0(\tau_0 ) , & \tau_{0,-1,0,1}=(-1)^{-\frac{3}{4}}[t(t-1)]^{\frac{1}{4}}s_1(\tau_1),\\[1 mm ] \tau_{-1,-1,0,1}=-it^{\frac{1}{4}}s_3(\tau_3 ) , & \tau_{-1,-1,-1,1}=(t-1)^{\frac{1}{4}}s_4(\tau_4 ) , \end{array } \label{lattice - tau}\ ] ] for instance .
when the parameters @xmath398 take the values @xmath399 the function @xmath400 relates to the hypergeometric @xmath0 function @xmath401 introduced in proposition [ hg_p6 ] by@xcite @xmath402 where we denote @xmath403 and @xmath404 , and the constants @xmath405 are determined by the recurrence relations @xmath406 \omega_{k , l+1,m , i}\omega_{k , l-1,m , i } = -i\hat{b}(\hat{c}-\hat{b})\omega_{k , l , m , i}^2,\\[1 mm ] \omega_{k , l , m+1,i}\omega_{k , l , m-1,i } = ( \hat{c}-\hat{a})(\hat{c}-\hat{b})\omega_{k , l , m , i}^2 \end{array } \quad ( i=0,1)\ ] ] and @xmath407 with initial conditions @xmath408 and @xmath409 from the above formulation , one can obtain the bilinear equations for the @xmath0 functions .
for example , let us express the bcklund transformations @xmath410 in terms of the @xmath0 functions @xmath411 .
we have by using ( [ act_on_tau ] ) @xmath412 \alpha_2t^{-\frac{1}{2}}\tau_0\tau_4 -s_1(\tau_1)s_2s_3(\tau_3)+s_3(\tau_3)s_2s_1(\tau_1)=0,\\[1 mm ] \alpha_2t^{-\frac{1}{2}}\tau_0\tau_3 -s_1(\tau_1)s_2s_4(\tau_4)+s_4(\tau_4)s_2s_1(\tau_1)= 0 . \end{array } \label{bi_bt}\ ] ] applying the affine weyl group @xmath413 on these equations , we obtain @xmath414 ( \alpha_0+\alpha_2+\alpha_4)\,t^{\frac{1}{2}}\tau_1\tau_3 -\tau_4s_4s_2s_0(\tau_0)+\tau_0s_0s_2s_4(\tau_4)=0 , \end{array } \label{bi_bt:1}\ ] ] @xmath415 ( \alpha_0+\alpha_1+\alpha_2)\,t^{\frac{1}{2}}s_1(\tau_1)\tau_3 -s_4(\tau_4)s_1s_2s_0(\tau_0)+\tau_0s_0s_1s_2s_4(\tau_4)=0 , \end{array } \label{bi_bt:2}\ ] ] @xmath416 ( \alpha_2+\alpha_3+\alpha_4)\,t^{-\frac{1}{2}}s_4(\tau_4)\tau_0 -s_1(\tau_1)s_4s_2s_3(\tau_3)+\tau_3s_3s_4s_2s_1(\tau_1)=0 , \end{array } \label{bi_bt:3}\ ] ] and @xmath417 ( \alpha_1+\alpha_4+\alpha_2)t^{-\frac{1}{2}}\tau_0\tau_3 -\tau_1s_1s_2s_4(\tau_4)+\tau_4s_4s_2s_1(\tau_1)=0 . \end{array
} \label{bi_bt:4}\ ] ] for instance , the first equation in ( [ bi_bt:1 ] ) can be obtained by applying @xmath418 on the first one in ( [ bi_bt ] ) .
we also get the second equation in ( [ bi_bt:1 ] ) by applying @xmath419 on the second one in ( [ bi_bt ] ) .
other equations can be derived in a similar manner . by applying the translation @xmath420 to the bilinear relations ( [ bi_bt:1 ] ) and noticing ( [ lattice - tau ] )
, we get @xmath421 ( \alpha_0+\alpha_2+\alpha_4-m)\tau_{k-1,l-1,m-1,n'}\tau_{k , l-1,m-1,n'}\\ \hskip20 mm + \tau_{k , l-1,m , n'}\tau_{k-1,l-1,m-2,n'}-\tau_{k , l , m , n'}\tau_{k-1,l-2,m-2,n'}=0 , \end{array}\ ] ] and then ( [ bi_n ] ) for the hypergeometric @xmath0 functions .
similarly , we obtain for the hypergeometric @xmath0 functions ( [ bi_m ] ) , ( [ bi_m ] ) and ( [ bi_bt ] ) from ( [ bi_bt:2 ] ) , ( [ bi_bt:3 ] ) and ( [ bi_bt:4 ] ) , respectively . the constraints @xmath422 yield @xmath423 \tau_0s_1s_2s_0(\tau_0 ) = \tau_4s_1s_2s_4(\tau_4)-ts_1(\tau_1)s_2s_1(\tau_1),\\[1 mm ] \tau_3s_4s_2s_3(\tau_3 ) = s_4(\tau_4)s_2s_4(\tau_4)-\tau_1s_4s_2s_1(\tau_1 ) , \end{array}\ ] ] and @xmath424 t\tau_3s_3(\tau_3)-\tau_0s_0(\tau_0)=(t-1)\tau_4s_4(\tau_4 ) , \end{array}\ ] ] from which we obtain ( [ bi_constraint1 ] ) and ( [ bi_constraint2 ] ) , respectively . due to ( [ def_h ] ) we have the relation @xmath425.\ ] ] then we get the bilinear relations @xmath426
d\,\tau_0\cdot\tau_3 = t^{\frac{1}{2}}\tau_4s_4s_2s_1(\tau_1 ) + \dfrac{1}{2}(1-\alpha_3+\alpha_4)\tau_0\tau_3,\\[3 mm ] d\,\tau_0\cdot\tau_3 = t^{\frac{1}{2}}\tau_1s_1s_2s_4(\tau_4)+\dfrac{1}{2}(\alpha_0-\alpha_1)\tau_0\tau_3 , \end{array } \label{bi_d}\ ] ] where @xmath427 denotes hirota s differential operator defined by @xmath428 . by applying the translation @xmath420 to the first bilinear relation of ( [ bi_d ] ) ,
one gets @xmath429\ , \tau_{k , l , m , n'}\cdot\tau_{k , l-1,m-1,n ' } = -t^{\frac{1}{2}}(t-1)^{-\frac{1}{2 } } \tau_{k-1,l-1,m-1,n'+1}\tau_{k+1,l , m , n'-1 } , \end{array}\ ] ] which is reduced to the first relation of ( [ bi_d ] ) .
the second and third relations of ( [ bi_d ] ) also yield their counterparts in ( [ bi_d ] ) .
00 handbook of mathematical functions with formulas , graphs , and mathematical tables , edited by m. abramowitz and i. a. stegun .
dover publications , inc . ,
new york , 1992 .
agafonov , s. i. `` imbedded circle patterns with the combinatorics of the square grid and discrete painlev equations . '' _ discrete comput .
_ 29 , no . 2 ( 2003 ) : 305319 .
agafonov , s. i. `` discrete riccati equation , hypergeometric functions and circle patterns of schramm type . '' _ glasg .
j. _ 47 , no .
a ( 2005 ) : 116 .
agafonov , s. i. , and bobenko , a. i. `` discrete @xmath430 and painlev equations . '' _ internat .
res . notices _
2000 , no .
4 : 165193 .
agafonov , s. i. , and bobenko , a. i. `` hexagonal circle patterns with constant intersection angles and discrete painlev and riccati equations . ''
_ j. math .
phys . _ 44 , no . 8
( 2003 ) : 34553469 .
beardon , a. f. , and stephenson , k. `` the uniformization theorem for circle packing . '' _ indiana univ .
39 , no . 4 ( 1990 ) : 13831425 .
bobenko , a. i. `` discrete conformal maps and surfaces '' in symmetries and integrability of difference equations ( canterbury , 1996 ) , 97108 , london math .
lecture note ser .
255 ( cambridge univ . press , cambridge , 1999 ) .
bobenko , a. i. , and hoffmann , t. `` hexagonal circle patterns and integrable systems : patterns with constant angles . '' _ duke math .
j. _ 116 , no . 3 ( 2003 )
: 525566 .
bobenko , a. i. , hoffmann , t. , and suris , y. b. `` hexagonal circle patterns and integrable systems : patterns with the multi - ratio property and lax equations on the regular triangular lattice . '' _ int . math .
res . not .
_ 2002 , no .
3 : 111164 .
bobenko , a. i. , and pinkall , u. `` discrete isothermic surfaces . '' _ j. reine angew .
_ 475 ( 1996 ) : 187208 .
bobenko , a. i. , and pinkall , u. `` discretization of surfaces and integrable systems . ''
discrete integrable geometry and physics ( vienna , 1996 ) , 358 , oxford lecture ser .
appl . , 16 , oxford univ . press , new york , 1999 .
hay , m. , kajiwara , k. , and masuda , t. `` bilinearization and special solutions to the discrete schwarzian kdv equation . ''
_ j. math - for - ind .
_ 3 , ( 2011 ) : 5362 .
masuda , t. `` classical transcendental solutions of the painlev equations and their degeneration . '' _ tohoku math .
j. _ 56 , no . 4 ( 2004 )
: 467490 .
nijhoff , f. w. , ramani , a. , grammaticos , b. , and ohta , y. `` on discrete painlev equations associated with the lattice kdv systems and the painlev vi equation . '' _ stud .
_ 106 , no . 3 ( 2001 ) : 261314 .
rodin , b. `` schwarz s lemma for circle packings . ''
_ 89 , no . 2 ( 1987 ) : 271289 .
schramm , o. `` circle patterns with the combinatorics of the square grid . '' _ duke math .
86 , no . 2 ( 1997 )
: 347389 .
stephenson , k. _ introduction to circle packing _ , new york : cambridge university press , 2005 .
thurston , w. p. `` the finite riemann mapping theorem . '' invited address , international symposium in celebration of the proof of the bieberbach conjecture ( purdue university , 1985 ) . | we present an explicit formula for the discrete power function introduced by bobenko , which is expressed in terms of the hypergeometric @xmath0 functions for the sixth painlev equation .
the original definition of the discrete power function imposes strict conditions on the domain and the value of the exponent .
however , we show that one can extend the value of the exponent to arbitrary complex numbers except even integers and the domain to a discrete analogue of the riemann surface .
moreover , we show that the discrete power function is an immersion when the real part of the exponent is equal to one . |
analyses of galaxies in the sloan digital sky survey ( sdss ) have demonstrated two distinct bimodalities in galaxy properties : a bimodality between recent - burst dominated and more continuous star formation histories ( sfhs ) as a function of stellar mass @xmath5 , divided at @xmath6 m@xmath4 @xcite , and a bimodality between blue late - type and red early - type galaxy sequences as a function of optical color , divided at @xmath7 @xcite .
recently , @xcite have partially unified these observations , demonstrating a color transition within each of the two galaxy sequences at @xmath8 m@xmath4 , as well as an increase in the relative number density of red sequence galaxies above @xmath125@xmath9 m@xmath4 .
they also argue that the number density of the red sequence is consistent with a major - merger origin .
however , the cause of the color and sfh transitions at @xmath123@xmath9 m@xmath4 remains to be explained .
several physical processes that influence sfhs may imprint a transition mass on the galaxy population .
supernova - driven gas blow - away will preferentially affect halos with small escape velocities @xcite , although simulations suggest that the baryonic mass threshold for blow - away may be closer to @xmath10 m@xmath4 than to @xmath11 m@xmath4 @xcite .
cold - mode gas accretion may dominate in low - mass halos whose gas fails to shock to the virial temperature @xcite ; here analytic estimates give a threshold mass of a few times @xmath12 m@xmath4 including dark matter , so a link to the observed transition at @xmath13 23 @xmath9 m@xmath4 is plausible .
finally , observations suggest that inefficient star formation may be typical of disk - dominated galaxies with @xmath14 , possibly reflecting the relative importance of supernova feedback as opposed to other turbulence drivers in supporting the interstellar medium against gravitational instability @xcite .
all of these processes involve gas its expulsion , accretion , or rate of consumption . thus examining
how the gas properties of galaxies vary with color and stellar mass may offer vital clues to the origin of the transition mass and the color shifts within the red and blue sequences .
unfortunately , tracing the dominant neutral phase of the interstellar medium requires hi 21-cm line observations , which are challenging even at the modest redshifts probed by the sdss . to make full use of the statistical power of the sdss
, an alternate strategy is required .
building on earlier optical work ( e.g. , * ? ? ?
* ) , @xcite has shown a remarkably tight correlation between hi mass - to-@xmath15-band luminosity ratios and @xmath16 colors .
going one step further , the present work describes a method for estimating atomic - gas - to - stellar mass ratios using @xmath0 colors from the sdss and two micron all sky survey ( 2mass ) databases .
this `` photometric gas fraction '' technique is calibrated using hi data from the recently expanded hyperleda hi catalog . when the technique is applied to a sample of @xmath135,000 sdss-2mass galaxies at @xmath17 , the transition mass of 23@xmath9 m@xmath4
is observed to correspond to a shift in gas richness found separately in both galaxy color sequences .
this result implies that any explanation of the transition mass via gas physics must directly or indirectly affect both early- and late - type galaxies .
optical , near - ir , and hi data were obtained from the sdss second data release ( dr2 , * ? ? ? * ) , the 2mass all - sky extended source catalog ( xsc , * ? ? ?
* ) , and the hyperleda homogenized hi catalog @xcite .
merged catalogs were constructed containing all @xmath17 , @xmath18 , @xmath19 galaxies with positions matched to within 6@xmath20 and with reliable redshifts and magnitudes based on data flags and cataloged errors ( magnitude errors @xmath21 in @xmath22 , @xmath23 in hi , and @xmath24 in @xmath25 ) .
the 2mass magnitude limit was set fainter than the completeness limit to improve statistics on dwarf and low surface brightness galaxies .
as the 2mass xsc has uneven depth , it probes significantly fainter than the completeness limit in some areas of the sky .
because of their marginal detectability , galaxies with hi - derived gas - to - stellar mass ratios greater than two were targeted for individual inspection , and eight were rejected as having unreliable 2mass or sdss pipeline reductions .
these rejections exacerbate the shortage of ir - faint galaxies .
the final samples are : sdss - hyperleda ( 575 galaxies ) , sdss-2mass - hyperleda ( 346 galaxies ) , and sdss-2mass ( 35,166 galaxies ) .
an additional requirement for the sdss-2mass sample was that the local group motion - corrected redshift be greater than 1000 .
all optical and ir magnitudes used here are fitted magnitudes , i.e. sdss model magnitudes and 2mass extrapolated total magnitudes .
the sdss magnitudes are corrected for galactic extinction using the dr2 tabulated values and k - corrected to redshift zero using * kcorrect v3.2 * @xcite , while the 2mass @xmath22-band magnitudes are k - corrected using @xmath26 @xcite .
distances are computed in the concordance cosmology @xmath27 , @xmath28 , @xmath29 .
[ fg : both]a shows the basic correlation between @xmath30-band and 21-cm apparent magnitudes @xmath31 and @xmath32 for the sdss - hyperleda sample .
its existence is not surprising : @xmath30-band light is a tracer of young massive stars , and the birth rate of young stars is known to depend on the available gas reservoir ( as in the global correlation between disk - averaged star formation rate and gas surface density , * ? ? ?
the presence of young massive stars may also enhance hi detection ( e.g. , * ? ? ? * ) .
the absolute magnitude correlation obtained by distance - correcting @xmath31 and @xmath32 is of course far stronger than the correlation in fig . [ fg : both]a , but at the cost of non - independent axes . in any case , what is relevant for predicting @xmath32 from @xmath31 is not correlation strength but scatter .
most of the 0.92 mag scatter in the @xmath31@xmath32 relation is not explained by the errors .
this scatter likely represents variations in @xmath30-band extinction , molecular - to - atomic gas ratios , and the physical conditions required to convert a gas reservoir into young stars .
even without calibrating these factors , the @xmath31@xmath32 relation is sufficiently tight for the present application . figs .
[ fg : both]b and [ fg : both]c plot atomic - gas - to - stellar mass ratios ( g / s ) against @xmath33 and @xmath0 colors for the sdss-2mass - hyperleda sample .
gas masses are derived from hi fluxes with a helium correction factor of 1.4 , and stellar masses are derived from @xmath22-band fluxes using stellar mass - to - light ( m / l ) ratios estimated from @xmath34 colors as in @xcite .
the resulting correlations are distance - independent and extremely strong , with spearman rank correlation coefficients of 0.75 and 0.69 for @xmath0 and @xmath33 respectively .
note that the calibration sample spans the color@xmath5 relation ( fig .
[ fg : colormass]a ) , and the color
g / s relation is tighter than the color@xmath5 relation for these galaxies by @xmath125% in both @xmath0 and @xmath33 . the strength of the @xmath0 color
g / s relation derives both from the underlying @xmath31@xmath32 relation and from the close correspondence between @xmath22-band light and stellar mass .
the latter correspondence is assumed within this work and may not apply to all starbursting systems @xcite ; however , @xcite find that spectroscopically determined m / l ratios generally agree well with color - based m / l ratios , even in the low - mass regime where starbursts are common . the large dynamic range of the @xmath0 color
g / s relation makes the relation forgiving of errors and thus well - suited to low - precision estimation of photometric gas fractions .
the 0.37 dex scatter in the relation provides a basis for error estimation .
furthermore , galaxies of low and high mass define broadly similar @xmath0 color
g / s relations in their regime of overlap ( triangles and dots in fig .
[ fg : both]c ) .
it should be borne in mind that the generality of the photometric gas fraction technique as currently formulated relies on the fact that heavily dust - enshrouded star formation , as in luminous infrared galaxies , is rare in the low-@xmath35 universe @xcite . in dusty systems one might find high g / s linked to _ red _
@xmath0 colors .
also , the calibration given here could significantly underestimate actual gas - to - stellar mass ratios if stellar m / l ratios are much lower than assumed and/or molecular gas corrections are large ( a controversial topic ; see * ? ? ? * ; * ? ? ?
* ) . the m / l ratios used here are roughly consistent with maximum disk assumptions for spiral galaxies @xcite . within the @xmath36 magnitude set , the best alternative to the @xmath0 color
g / s relation is the @xmath33 color g / s relation .
its larger scatter may in part reflect the fact that @xmath22-band magnitude errors will move points along the @xmath0 color g / s relation , but away from the @xmath33 color g / s relation . however , the effect from cataloged errors is quite small ( arrows in figs .
[ fg : both]b and [ fg : both]c ) , so the greater @xmath33 scatter seems to be mostly physical . fig .
[ fg : colormass]a plots @xmath33 color vs. @xmath37 for the @xmath135,000-galaxy sdss-2mass sample , with points color - coded to indicate photometric gas fractions g / s@xmath38 ( computed from @xmath0 colors and the fit in fig .
[ fg : both]c ) .
the well - known red and blue sequences ( referring to galaxy color , not color - coding ) are roughly separated by the dashed line ( an approximation to the separator of * ? ? ?
* ) . within each sequence ,
the color - coding reveals a shift in gas fractions near a threshold mass of @xmath39 13 @xmath40 @xmath11 m@xmath4 .
massive red - sequence galaxies are extremely gas - poor ( g / s@xmath38 as low as 1:100 , red / yellow color - coded points ) , whereas for red - sequence galaxies below @xmath41 , intermediate gas fractions ( g / s@xmath38 @xmath1 1:10 , green points ) are the norm .
likewise , massive blue - sequence galaxies have intermediate gas fractions ( g / s@xmath38 @xmath1 1:10 , green points ) , but blue - sequence galaxies below @xmath41 are typically gas rich ( g / s@xmath38 @xmath1 1:1 , blue points ) .
these results are shown in binned form in fig .
[ fg : colormass]b , where the contours show the conditional probability distribution of @xmath33 on @xmath37 . in this plot ,
a vertical slice through the contours at a given mass gives the one - dimensional probability distribution for @xmath33 colors at that mass .
the conditional probability distribution is formed by weighting each galaxy by 1/v@xmath42 , where v@xmath42 is the maximum volume within which it could have been detected , then normalizing the counts in each color@xmath40mass bin by the total counts in each mass column .
bins with fewer than four galaxies are not considered .
this algorithm is most robustly applied to truly magnitude - limited samples , but the results shown here look qualitatively similar to those of @xcite for a magnitude - limited sdss sample .
each galaxy is assigned the smallest of three v@xmath42 estimates , based on ( i ) the magnitude limit in @xmath43 , ( ii ) the magnitude limit in @xmath22 , and ( iii ) the distance limit imposed by the @xmath17 selection requirement . the color - coded symbols in fig .
[ fg : colormass]b illustrate gas fraction and morphology trends using 1/v@xmath42-weighted bin averages .
ovals and s - shapes identify early and late type morphologies , respectively , based on the concentration index @xmath44 ( defined as @xmath45 , where @xmath46 and @xmath47 are the 90% and 50% petrosian radii ) .
a value of @xmath48 is commonly used to divide early and late types when true morphological type information is lacking @xcite .
[ fg : colormass]b indicates transitional values ( @xmath49 ) with both type symbols . some of the bluest low - mass galaxy bins show transitional types , which will not be examined here .
otherwise , the blue sequence forms a broad , low @xmath44 strip , with a shift in gas richness at @xmath41 @xmath113 @xmath40 10@xmath50 m@xmath4 . at a similar mass
, the red sequence shows a coordinated shift in gas richness and @xmath44 .
intriguingly , the increased gas content of the red sequence below @xmath41 causes the red and blue sequences to fuse ( nearly ) in a plot of g / s@xmath38 vs. @xmath5 ( fig .
[ fg : colormass]c ) .
the double - peaked , slightly s - shaped contours of the g / s@xmath38 vs. @xmath5 distribution thus look qualitatively similar to the bimodal distribution of sfh vs.@xmath5 reported by @xcite .
this similarity suggests that the bimodality in sfhs may be intimately related to changes in g / s .
the contours in fig .
[ fg : colormass]c show a broad shift near @xmath110@xmath50 m@xmath4 , while the individual trends in the red and blue sequences change slope over the range @xmath113 @xmath40 10@xmath50 m@xmath4 ( though the latter changes are not independent of the method of separating the two sequences ) .
kauffmann et al .
find a single - sequence bimodal structure in @xmath44 vs. @xmath5 as well , perhaps because of the decreasing @xmath44 of the red sequence below @xmath41 .
despite low @xmath44 values and high gas contents , the low - mass red sequence does fit into the general merger scenario for the origin of early - type galaxies , in a way that may illuminate the kauffmann et al .
the low - mass red sequence maps closely to an abundant population of faint , moderately gas - rich s0 and s0/a galaxies observed in the nearby field galaxy survey , a survey designed to represent the natural distribution of galaxy types over a wide range of luminosities ( * ? ? ?
* ; * ? ? ?
* see fig . 3 of the latter ) .
based on the frequency of gas - stellar counterrotation in this population and the scarcity of dwarf intermediate - type spiral galaxies , @xcite argue that @xmath5150% of low - luminosity s0 s may form via late - type dwarf mergers .
such gas - rich mergers would naturally produce remnants with modest @xmath44 , explaining why low - luminosity early types are predominantly s0 rather than e galaxies .
moreover , low - luminosity s0 s quite often have blue , starbursting centers , despite red outer disks @xcite , and such color gradients correlate with evidence of _ recent _ interactions ( in all morphological types , * ? ? ?
the sfh measures adopted by @xcite , which are based on 3@xmath20-aperture spectroscopy , may emphasize the starbursting centers of this class of galaxies ( and conversely , the quiescent bulges of many high - mass late - type systems ) , reinforcing the single - sequence structure of the resulting sfh vs. @xmath5 plots .
a connection between star formation histories and gas fractions is in some sense obvious : gas must be consumed to form stars , so old red stellar populations will tend to be associated with diminished gas supplies .
however , galaxies also accrete and expel gas , so this simple view misses much of the story .
a complete picture must explain why gas fractions depend on mass , and in particular why there is such a close coincidence between the transition to gas richness ( atomic - gas - to - stellar mass ratio @xmath1 1:1 in the blue sequence ) and the shift to recent - burst dominated sfhs below @xmath1
23 @xmath40 10@xmath50 m@xmath4 .
critical transition masses are predicted by scenarios involving starburst - driven gas blow - away , inefficient star formation below a gravitational instability threshold , and/or cold - mode gas accretion @xcite .
the abundance of gas in low - mass ( 10@xmath5210@xmath50 m@xmath4 ) galaxies seen in this paper is hard to explain if global gas blow - away is a dominant process in this mass regime
. however , localized gas blowout or strong feedback could inhibit efficient widespread star formation in low - mass disk galaxies .
such scenarios would not explicitly account for the gas in low - mass red - sequence galaxies , but formation via mergers of low - mass late - type systems , as discussed above , could allow low - mass red - sequence galaxies to acquire their modest gas excesses from gas - rich progenitors and thereby inherit the progenitors threshold mass for increased gas fractions . alternatively , it is possible that the transition in star formation modes at @xmath123 @xmath40 10@xmath50 m@xmath4 is not a cause , but an _ effect _ of changing gas fractions , as in cold - mode accretion scenarios .
if so , low - mass galaxies may form stars reasonably efficiently and still appear gas rich .
the excess gas in low - mass red - sequence galaxies could in this case represent post - merger cold - mode accretion , possibly as part of a process of disk regrowth . in conclusion , this paper has demonstrated a shift in galaxy gas mass fractions at @xmath41 @xmath1 13 @xmath40 10@xmath50 m@xmath4 , likely related to the shift in sfhs observed near the same stellar mass by @xcite . the link may be causal in either direction , depending on the relative importance of supernova blowout , feedback , and cold - mode accretion processes in determining @xmath41 . to establish this result , a technique has been introduced to estimate photometric gas fractions based on the correlation between @xmath0 colors and atomic - gas - to - stellar mass ratios .
this correlation is interesting in its own right ( see also * ? ? ?
* ) and will be further examined and applied in future work .
i thank d. mar for being a helpful sounding board and s. faber for sparking my interest in galaxy bimodalities .
n. martimbeau and j. huchra generously shared a catalog merging code for me to modify .
i am also indebted to s. jester , c. gerardy , d. mar , and p. hoeflich for assistance with database and software issues . k. gebhardt and r. kennicutt provided useful comments on the original manuscript .
d. mcintosh , l. matthews , a. baker , and the referee helped to improve the clarity of the final version .
this research has used the hyperleda hi catalog ( vizier catalog vii/238 ) and data from 2mass , a joint project of the u. of massachusetts and ipac / caltech , funded by nasa and the nsf .
it has also used the sdss ( see acknowledgement at http://www.sdss.org/collaboration/credits.html ) . | galaxies over four decades in stellar mass are shown to obey a strong correlation between @xmath0 colors and atomic - gas - to - stellar mass ratios ( g / s ) , using stellar mass - to - light ratios derived from optical colors .
the correlation holds for g / s ranging from nearly 10:1 to 1:100 for a sample obtained by merging the sdss dr2 , 2mass , and hyperleda hi catalogs .
this result implies that @xmath0 colors can be calibrated to provide `` photometric gas fractions '' for statistical applications . here
this technique is applied to a sample of @xmath135,000 sdss-2mass galaxies to examine the relationship of gas fractions to observed bimodalities in galaxy properties as a function of color and stellar mass . the recently identified transition in galaxy properties at stellar masses
@xmath2@xmath3 m@xmath4 corresponds to a shift in gas richness , dividing low - mass late - type galaxies with g / s @xmath1 1:1 from high - mass galaxies with intermediate - to - low g / s .
early - type galaxies below the transition mass also show elevated g / s , consistent with formation scenarios involving mergers of low - mass gas - rich systems and/or cold - mode gas accretion . |
leptoquarks ( lq ) are hypothetical particles that are predicted by many extensions @xcite of the standard model of particle physics ( sm ) , such as grand unification theory , technicolor and composite models .
they carry both baryon and lepton numbers and thus couple to a lepton and a quark .
they carry fractional electric charge , they are color triples , and can have either zero or one unit of spin ( i.e. can be either scalar or vector particles ) . existing experimental limits on lepton number violation , flavor changing neutral current , proton decay , and other rare processes favor three generations of leptoquarks , with no inter - generational mixing . the production and decay of lqs
are characterized by : the mass of the lq particle ( m@xmath0 ) , its decay branching ratio into a lepton and a quark ( usually denoted as @xmath1 ) and the yukawa coupling at the lq - lepton - quark vertex ( @xmath2 ) . at hadron colliders ,
leptoquarks are mainly produced in pairs , via gluon - gluon fusion and quark anti - quark annihilation .
the dominant pair production mechanisms do not depend strongly on @xmath2 and the single production of leptoquarks does not become significant ( and it does not in any case invalidate the search results presented here ) in the range of lq masses probed with the 2010 lhc data .
the final state event signatures from the decay of pair produced lqs can be classified as : dilepton and jets ( both lq and anti - lq decay into a charged lepton and a quark ) ; lepton , missing transverse energy and a jet ( one lq decays into a charged lepton and a quark , while the other decays into a neutrino and a quark ) ; and missing transverse energy and jets ( both lq and anti - lq decay into neutrinos and quarks ) .
the three signatures have branching ratios of @xmath3 , 2@xmath4 , and @xmath5 , respectively .
the charged leptons can be either electrons , muons , or tau leptons , corresponding to the three generation of lqs .
only electrons and muons are considered here .
searches for first and second generation pair - produced scalar lqs were performed and published on 34 - 36 pb@xmath6 of 7 tev proton - proton collision data recorded by the cms detector @xcite during the 2010 lhc run .
first generation results include searches both in the dilepton and jets final state ( @xmath7 ) , and in the lepton , missing transverse energy and jets final state ( @xmath8 ) .
second generation result consist of a search in the dilepton and jets final state ( @xmath9 ) .
the @xmath7 and @xmath8 results were combined to attain the best possible exclusion reach in all of the parameter space of the first generation pair produced scalar lq search , i.e. ( @xmath1 and @xmath10 ) .
the analysis in all of the channels aims at identifying the existence of new heavy particles by establishing an excess of events characteristic of the decay of heavy objects .
the analysis starts with using either a single or a double lepton trigger path , which is robust and very efficient .
an event signature preselection isolates events with high transverse momentum final state objects ( two or more isolated leptons and two or more jets , or one isolated lepton , two or more jets , and significant missing transverse energy indicative of the presence of a neutrino ) .
event variables are identified to effectively separate a possible lq signal from standard model backgrounds ( these are the @xmath11 , @xmath12 , and @xmath13 variables described below ) and lower thresholds are placed on these variables at preselection level .
backgrounds from standard model sources are estimated and first compared with data at the preselection level .
major sources of background are @xmath14+jets and @xmath15jets processes and @xmath16 ( with single top production , diboson processes , and qcd multijet processes being smaller contributions ) .
major backgrounds are either directly determined from data control samples or determined with monte carlo samples ( therefore using kinematic shape information from the mc ) but normalized to data in selected control regions .
a cut - based variable approach is used for the final selection , where the selection is optimized for different lq mass hypotheses by minimizing the expected upper limit on the lq cross section in the absence of an observed signal using a bayesian approach @xcite .
the three variables used in the optimization are : the scalar sum of the final objects transverse momenta , @xmath11 ; the invariant mass of the dilepton pair , @xmath17 , for the @xmath7 and @xmath9 channels ; and the transverse mass , @xmath13 , of the lepton and neutrino in the @xmath8 analysis .
after final selection , the data are well described by the standard model background predictions . in the absence of an observed signal , an upper limit on the lq cross section
is set using a bayesian method @xcite with a flat signal prior .
a lognormal prior is used to integrate over the nuisance parameters . using poisson statistics , a 95% confidence level ( c.l . )
upper limit is obtained on the lq pair - production cross - section times branching ratio as a function of lq mass . comparing with the nlo predictions @xcite for the scalar lq pair production cross section a lower limit on @xmath10
is determined for @xmath18 and @xmath19 ( electron only , for the latter ) . in the electron channel
, the @xmath7 and @xmath8 channels are combined to further maximize the exclusion in @xmath1 and @xmath10 , specially for the case of @xmath19 , where the @xmath7 adds to the sensitivity of the @xmath8 channel .
the @xmath7 analysis requires the presence of a single or double electromagnetic trigger ( with an efficiency close to 100@xmath20 ) , two or more isolated electrons with @xmath21 gev and @xmath22 , and two or more jets with @xmath21 gev and @xmath23 .
in addition , at pre - selection level , @xmath24 is required together with a minimum threshold of @xmath25 gev and @xmath26 gev .
data and standard model background predictions agree well at the level of pre - selection , as shown in fig .
[ eeprelpte ] , and [ eeprelst ] .
this channel uses an integrated luminosity of 33.2 pb@xmath6 .
similarly , the @xmath9 analysis starts from the requirement of a single muon trigger ( with an efficiency of 99@xmath20 ) , two or more isolated muons with @xmath21 gev and @xmath27 ( one of which must be within @xmath28 ) , and two or more jets with @xmath21 gev and @xmath23 .
the two muons are required to be separated in @xmath29 by at least 0.3 .
minimum thresholds of @xmath30 gev and @xmath31 gev complete the preselection requirements .
good agreement is observed between data and standard model background predictions at this level ( fig.s [ mmprelptm ] and [ mmprelst ] ) .
this channel uses an integrated luminosity of 34.0 pb@xmath6 .
the major backgrounds from standard model processes to the dilepton and jets channels come from @xmath32 jets and @xmath16 production . the @xmath33 contribution is determined from mc , but it is rescaled to the data in the z peak region at preselection .
the invariant mass of the dilepton pair is shown in fig .
[ eepreldimass ] and in fig .
[ mmpreldimass ] for the @xmath7 and @xmath9 channel , respectively .
the normalization factors are @xmath34 for @xmath7 and @xmath35 for @xmath9 .
the systematic uncertainty on the @xmath32 jets background prediction in both channels comes from the statistically dominated uncertainty in the normalization factors and from a shape uncertainty obtained by comparing the yields of @xmath32 jets mc samples generated with different renormalization and factorization scales and matching thresholds .
the @xmath16 normalization uncertainty is based on a cms measurement of the @xmath16 cross - section @xcite .
smaller @xmath15 jets and diboson plus jets backgrounds are determined entirely from mc and are found to be negligible .
the background contributions from qcd multijet processes are determined from data control regions and are also found to be negligible . at final selection , a dilepton invariant mass cut ( @xmath36125 gev and @xmath37115 gev ) is placed to suppress most of the @xmath33 jets background and a @xmath11 threshold is optimized for each @xmath10 hypothesis .
@xmath11 is effective at removing most of the @xmath16 background and any residual backgrounds surviving the preselection criteria and the cut on the dilepton invariant mass .
the number of data , signal mc , and background events for each of the final selection criteria optimized for different @xmath10 are listed in table [ tab : eejjsummary ] and table [ tab : mumujjsummary ] for the @xmath7 and @xmath9 channels respectively , together with the optimized @xmath11 thresholds . in absence of an excess above standard model backgrounds expectation , an upper limit on the lq cross section
is set using a bayesian method similar to the one used in the @xmath7 analysis .
the @xmath8 channel has maximum sensitivity for @xmath19 , and also offers sensitivity for lower values of @xmath1 , where the @xmath7 channel quickly runs out of sensitivity due to the quadratic behavior of the branching fraction of the lq pair into the above channel .
thus , combining it with the @xmath7 channel allows one to extend the sensitivity of the search in the intermediate @xmath1 range compared to either of the two analysis considered separately .
the two channels are combined using the same bayesian approach used to set the individual limits . assuming a flat prior for the signal cross section , the expected signal yield in the @xmath7 [ @xmath8 ] channel is found by multiplying the cross section by the signal efficiency , integrated luminosity , and the branching fraction @xmath3 [ @xmath39 .
since the major sources of systematic uncertainty between the two channels are correlated , a combined likelihood is constructed from the individual likelihoods for the two channels . the 95% c.l .
mass limit as a function of @xmath1 is shown in fig .
[ eecombbeta ] , along with the individual limits from the two channels . with the assumption
that @xmath40 , first generation scalar leptoquarks with masses less than 340 gev are excluded at 95% c.l .
@xcite .
searches for pair production of first and second generation scalar leptoquarks have been performed at cms in the final states with : two charged leptons and two jets , one charged lepton ( electron ) , missing transverse energy and two jets , with the full 2010 statistics . in the absence of a signal we exclude first generation lq with masses below 384 gev ( @xmath41 ) , second generation lq with masses below 394 gev ( @xmath41 ) , and first generation lq with masses below 340 gev ( @xmath19 ) . the @xmath18 results were the most stringent limits at the time of publication .
the @xmath19 and @xmath18 results have been combined and have been submitted for publication .
the analysis of the 2011 data is ongoing . | this article describes the search for pair production of scalar leptoquarks performed by the cms collaboration using the data from the 2010 proton - proton run of the large hadron collider at a center of mass energy of 7 tev .
the following final state signatures from the decay of a pair of scalar leptoquarks are described in here : the di - lepton plus jets final state ( where the leptons are either both electrons or both muons ) and the lepton plus jets and transverse missing energy final state ( where the lepton is an electron ) . |
in the sections which follow , we shall be concerned predominantly with the theory of short @xmath0-ray bursts . if the concepts there proposed are indeed relevant to an understanding of the nature of these sources , then their existence becomes inextricably linked to the _ metabolic pathways _ through which gravity , spin , and energy can combine to form collimated , ultra relativistic outflows .
these threads are few and fragile , as we are still wrestling with understanding non - relativistic processes , most notably those associated with the electromagnetic field and gas dynamics .
if we are to improve our picture - making we must make more and stronger ties to physical theory . but in reconstructing the creature , we must be guided by our eyes and their extensions . in this introductory chapter
we have therefore attempted to briefly summarize the observed properties of these ultra - energetic phenomena .
there are five sections : [ sec : his ] gives a brief account of their history from birth to present - age ; [ sec : metabolics ] is devoted to their metabolism in other words , to their gross energetics , spectra and time variability ; [ sec : aft ] describes the attributes of the afterglow signals , which , as fading beacons , mark the location of the fiery and brief @xmath0-ray event . these afterglows in turn enable the measurement of redshift distances , the identification of host galaxies at cosmological distances , and provide evidence that many short @xmath0-ray bursts are associated with old stellar populations and possibly with no bright supernova .
these threads will be woven in [
sec : field ] .
finally , [ sec : stage ] gives a compendium of the observational _
facts_. the manifestations of sgrb activity are extremely diverse .
sgrbs are observed throughout the whole electromagnetic spectrum , from ghz radio waves to 10 mev @xmath0-rays , but until recently , they were known predominantly as bursts of @xmath0-rays , largely devoid of any observable traces at any other wavelengths . before 2005 ,
most of what we knew about sgrbs was based on observations from the burst and transient source experiment ( batse ) on board the _ compton gamma ray observatory _ ,
whose results have been summarized by fishman & meegan @xcite .
batse , which measured about 3000 events , detected approximately one burst on a typical day . while they are on , they outshine every other source in the @xmath0-ray sky , including the sun .
although each is unique , the bursts fall into one of two rough categories .
bursts that last less than two seconds are classified as short , and those that last longer
the majority as long @xcite .
the two categories differ spectroscopically , with short bursts having relatively more high - energy @xmath0-rays than long bursts do .
figure [ fig : hard ] shows the hardness ratio as a function of the duration of the emission .
it is a measure of the slope of the spectrum , where larger values mean that the flux at high energies dominates .
arguably the most important result from batse concerned the spatial distribution of bursts .
both long and short events occur isotropically that is , evenly over the entire sky with no dipole and quadrupole components , suggesting a cosmological distribution .
this finding cast doubt on the prevailing wisdom , which held that bursts came from sources within the milky way .
the uniform distribution instead led most astronomers to conclude that the instruments were picking up some kind of cosmological event .
unfortunately , @xmath0-rays alone did not provide enough information to settle the question for sure .
the detection of radiation from bursts at other wavelengths would turn out to be essential .
visible light , for example , could reveal the galaxies in which the bursts took place , allowing their distances to be measured .
attempts were made to detect these burst counterparts , but they proved fruitless .
observations of burst counterparts @xcite were restricted to the class of long duration bursts until , in 2005 , the _ swift _ spacecraft succeeded in obtaining high - resolution x - ray images @xcite of the fading afterglow of grb 050509b so named because it occurred on may 9 , 2005 .
this detection , followed by a number of others at an approximate rate of 10 per year , led to accurate positions , which allowed the detection and follow - up of the afterglows at optical and longer wavelengths @xcite .
this paved the way for the measurement of redshift distances , the identification of candidate host galaxies , and the confirmation that they were at cosmological distances @xcite . _
swift _ is equipped with @xmath0-ray , x - ray and optical detectors for on - board follow - up , and capable of relaying to the ground arc - second quality burst coordinates within less than a minute from the burst trigger , allowing even mid - size ground - based telescopes to obtain prompt spectra and redshifts .
sgrbs are brief flashes of radiation at soft and hard @xmath0-ray energies that display a wide variety of time histories .
they were first detected at soft @xmath0-ray energies with wide field - of - view instruments , with peak soft @xmath0-ray fluxes reaching hundreds of photons @xmath1 s@xmath2 in rare cases .
the batse instrument was sensitive in the 50 - 300 kev band , and provided the most extensive data base of sgrb observations during the prompt phase . of various sgrbs .
the sgrb spectrum is non thermal , the number of photons varying typically as @xmath3 , where @xmath4 at low energies changes to @xmath5 to 3 above a photon energy @xmath6 1 mev.,width=336 ] sgrbs typically show a very hard spectrum in the soft to hard @xmath0-ray regime .
the photon index breaks from @xmath7 at energies @xmath8 kev , to a @xmath9 to @xmath10 spectrum at @xmath11 several hundred kev @xcite .
consequently , the peak photon energies , @xmath12 , of the time - averaged @xmath13 spectra of batse sgrbs are typically found in the 500 kev - several mev range @xcite .
the general trend is that the spectrum softens , and @xmath12 decreases , with time .
more precise statements must , however , wait for larger area detectors . in figure
[ fig : spec ] representative spectra are plotted , in the conventional coordinates @xmath14 and @xmath13 , the energy radiated per logarithmic frequency interval .
some obvious points should be emphasized .
we measure directly only the specific luminosity @xmath15 ( the energy radiated in the direction of the earth per second per steradian per logarithmic frequency interval by a source at luminosity distance @xmath16 ) , and its dimensionless distance - independent ratios between two frequencies ( batse triggers , for example , are based on the count rate between 50 kev and 300kev )
. the apparent bolometric luminosity @xmath17 may be quite different from the _ true _ bolometric luminosity @xmath18 if the source is not isotropic .
the grb spectra shown in figure [ fig : spec ] are those of grb 910609 , grb 970704 , and grb 931101 which were observed by batse .
the time integrated spectrum on those detectors ranges from 25 kev to 10mev @xcite .
a _ typical _ sgrb if there is such a thing lasts for a fraction of a second .
observed durations vary , however , by three orders of magnitude , from several milliseconds @xcite to a few seconds @xcite .
the shortest batse burst had a duration of 5ms with a 0.2ms structure @xcite .
similarly to long bursts , sgrbs have complicated and irregular time profiles which vary drastically from one burst to another @xcite .
they range from smooth , fast rise and quasi - exponential decay , through curves with several peaks , to variable curves with many peaks .
various profiles , selected from the batse catalog , are shown in figure [ fig : lc ] .
-rays are detected , above background , from the same direction.,width=432 ] the duration of a sgrb is defined by the time during which the middle 50% ( @xmath19 ) or 90% ( @xmath20 ) of the counts above background are measured .
a bimodal duration distribution is measured , irrespective of whether the @xmath19 or @xmath20 durations are considered @xcite .
about two - thirds of batse grbs are long - duration grbs with @xmath21 s , with the remainder comprising the sgrbs .
the integral size distribution of batse sgrbs in terms of peak flux @xmath22 is very flat below @xmath23 ph @xmath1 s@xmath2 , and becomes steeper than the expected power law with index @xmath24 of a euclidean distribution of sources , at @xmath25 ph @xmath1 s@xmath2 ( see figure 12 in @xcite ) .
this follows from a cosmological origin of grb sources , with the decline in the number of faint bursts due to cosmic expansion .
follow - up x - ray observations with _ swift _ and _ hete - ii _ have permitted redshift determinations that firmly establish the distance scale to the sources of sgrbs @xcite .
the redshifts of about half a dozen sgrbs are now known ( late 2006 ) , with the median @xmath26 and the largest _ spectroscopically _ inferred redshift at @xmath27 .
it should be noted that the inference of redshift distances relies at present on the statistical connection to a putative host galaxy and spectroscopy of the host , and that none are based on the absorption - line systems seen in the spectra of the afterglows .
the corresponding distances imply apparent isotropic @xmath0-ray energy releases @xmath28-@xmath29 ergs , where @xmath30 is the solid angle into which the @xmath0-rays are beamed ( figure [ fig : energ ] ) .
for a solar - mass object , this implies that an unusually large fraction of the energy is converted into @xmath0-ray photon energy .
this spread in the inferred luminosities obtained under the assumption of isotropic emission may be reduced if most grb outflows are jet - like .
a beamed jet would alleviate the energy requirements , and some observational evidence does suggest the presence of a jet ( see [ sec : aft ] ) . -ray energy as a function of redshift .
the energy is calculated , assuming isotropic emission , for 6 sgrbs with estimated redshifts . for comparison ,
the isotropic @xmath0-ray energies for 22 long grbs are also plotted using the compilation of @xcite.,width=480 ] among the first sgrbs localized by _ swift _ was grb 050509b .
the satellite slewed rapidly in the direction of the burst , enabling its narrow - field x - ray telescope to pinpoint to within 10 arc seconds a source of faint x - ray emission less than a minute after the trigger . the source decayed with a power - law behaviour , @xmath31 , with @xmath32 @xcite .
despite intense follow - up in the optical and radio , no counterparts were discovered @xcite . however , the lack of a detectable afterglow at other wavelengths is not surprising considering , at face value , the existing grb afterglow theory @xcite .
it took about two months before the observation of a second short burst , grb 050709 , pinpointed by _ hete - ii _ @xcite .
other detections have followed , at an approximate rate of 10 per year , and permitted the observation and follow - up of afterglows at optical and longer wavelengths .
grb 050709 was the first sgrb from which an optical counterpart was observed @xcite , and grb 050724 was the first sgrb for which a radio afterglow was measured @xcite .
grb 051221a is probably the best - sampled sgrb to date @xcite .
of the few bursts localized by the _ swift _
xrt , only four were bright enough to permit detailed study : grb 050724 , grb 051221a , grb 051210 , and grb 051227 @xcite .
various x - ray light curves selected from the _ swift _ catalog , are shown in figure [ fig : flares ] .
_ chandra _ follow - up observations obtained for grb 050724 and grb 051221a provide the most constraining observations for sgrb jets to date @xcite .
the lack of an observed early downturn in the x - ray light curves has been interpreted as supporting a mild degree of anisotropy in sgrbs @xcite .
anisotropy in the burst outflow and emission affects the light curve at the time when the inverse of the bulk lorentz factor equals the opening angle of the outflow . if the critical lorentz factor is less than 3 or so ( i.e. the opening angle exceeds 20@xmath33 ) such a transition might be masked by the change from ultra relativistic to mildly relativistic flow
it would then be generically difficult to limit the late - time afterglow opening angle in this way if it exceeds 20@xmath33 .
since some afterglow light curves are unbroken power laws for over 20 days ( e.g. , grb 050724 ) , if the energy input were indeed just a simple impulsive shell the opening angle of the late - time afterglow at long wavelengths would probably be greater than 25@xmath33 , i.e. @xmath34 @xcite . however , even this still means that the energy estimates from the afterglow assuming isotropy could be @xmath35 times too high .
the beaming angle for the @xmath0-ray emission,@xmath30 , could be smaller , and is much harder to constrain directly .
an observation that attracted much attention was the discovery of long duration ( @xmath36 s ) x - ray flares following several sgrbs @xcite after a delay of @xmath37 s ( see , e.g. , grb 050724 in figure [ fig : flares ] ) .
there is also independent support that x - ray emission on these time scales is detected when lightcurves of many bursts are stacked @xcite .
these observations may indicate that some sources display continued activity ( at a variable level ) over a period of minutes @xcite .
additional structure in the light curves has also emerged from a continued analysis of some of these objects down to the faintest x - ray flux levels . in some bursts ( e.g. , grbs 050724 and 051221a )
, the x - ray light curve exhibits evidence for a late " energy refreshment to the blast wave , on time scales comparable to the afterglow time scale .
a reason for the late rise would be , say , if slower moving ejecta catches up with the main shock front injecting a substantial amount of energy and momentum @xcite . however , there are other mechanisms that can produce rising afterglow fluxes @xcite . starting with the detection of grb 050509b ,
a growing body of evidence has suggested that sgrbs are associated with an older and lower - redshift galactic population than long grbs and , in a few cases , with large ( @xmath38 10 kpc ) projected offsets from the centers of their putative host galaxies ( table [ tab : hosts ] ; @xcite ) .
lccccc grb & @xmath39 & host & in a galaxy & sfr & metallicity + & & classification & cluster & ( @xmath40 yr@xmath2)&(@xmath41 ) + 050509b & 0.225 & e & yes & @xmath420.1 & @xmath61 + 050709 & 0.160 & irr / late - type dwarf & no & @xmath43 & 0.25 + 050724 & 0.258 & early ( e+s0 ) & no & @xmath44 & 0.2 + 050813 & 0.722?(1.8 ) & e ? & likely & @xmath420.2 & @xmath61 + 051221a & 0.5459 & late - type dwarf ? & ? & @xmath61.5 & @xmath61 + 060502b & 0.287 & e ? & ? & @xmath420.4 & @xmath60.2 + the discovery of grb 050509b and its fading x - ray afterglow @xcite led to the first redshift and host galaxy association @xcite for a sgrb .
a chance association with such a galaxy was deemed unlikely even under conservative assumptions and stood in stark contrast with the lines - of - sight of long grbs , for which no association with an early - type host was ever made . further _
swift _ and _ hete - ii _ detections of sgrbs have continued to support this hypothesis , though sgrbs are not ubiquitously found at large offsets and associated with early - type galaxies . grb 050724 @xcite , grb 050813 @xcite , and grb 060502b like 050509b , were found to be in close association with old , red galaxies .
grb 050724 had optical and radio afterglow emission that pinpointed its location to be within its red host , making the association completely unambiguous , though the association of grb 050813 and grb 060502b with any single host remains tentative @xcite .
the absence of observable h@xmath45 and [ o ii ] emission constrains the unobscured star formation rates in these galaxies to @xmath46 yr@xmath2 , and the lack of balmer absorption lines implies that the last significant star forming event occurred @xmath47 gyr ago .
based on positions of the afterglows , two bursts ( 050509b and 050813 ) are very likely associated with clusters of galaxies @xcite . not all hosts lack active star formation : grb 050709 @xcite and grb 051221a @xcite both had optical afterglows and were associated with galaxies showing evidence for current , albeit low , star formation .
the host of grb 051221a , moreover , exhibits evidence for an evolved stellar population . despite the availability of both x - ray and optical afterglow locations ,
no nearby host has successfully been identified for either grb 060121 or grb 060313 .
these observations indicate that these sgrbs occurred during the past @xmath48 gyr of the universe ( @xmath49 ) in galaxies with diverse physical characteristics .
in contrast to what is found for sgrbs , all of the confirmed long grb host galaxies are actively forming stars with integrated , unobscured sfrs @xmath50 yr@xmath2 @xcite .
these host galaxies have small stellar masses and bluer colors than present - day spiral galaxies ( suggesting a low metallicity ; @xcite ) .
the ages implied for the long - soft grbs are estimated to be @xmath42 0.2 gyr @xcite , which is significantly younger than the minimum ages derived for the early - type galaxies found to be associated with sgrbs .
the cluster environments of at least two sgrbs contrast strikingly with the observation that no well - localized long - soft grb has yet been associated with a cluster @xcite . on the whole ,
the hosts of sgrbs , and by extension the progenitors , are not drawn from the same parent population of long grbs .
sgrbs appear to be more diffusely positioned around galaxies , and their associated hosts contain a generally older population of stars .
it has been noted @xcite that the giant flare ( gf ) observed from the putative galactic magnetar source sgr1806 - 20 in december 2004 @xcite could have looked like a classical sgrb had it occurred much farther away , thus making the tell - tale periodic signal characteristic of the neutron star rotation in the fading emission undetectable .
the two previously recorded gfs of this type , one each from sgr 0520 - 66 on 5 march 1979 @xcite and sgr 1900 + 14 on 27 august 1998 @xcite , would have been detectable by existing instruments only out to @xmath6 8 mpc , and it was therefore not previously thought that they could be the source of sgrbs .
the main spike of the 27 december event would have resembled a short , hard grb if it had occurred within @xmath6 40 mpc , a distance scale encompassing the virgo cluster @xcite .
however , the paucity of observed giant flares in our own galaxy has so far precluded observationally based determinations of either their luminosity function or their rate . in the magnetar model ,
sgrs are isolated neutron stars with teragauss exterior magnetic fields and even stronger fields within @xcite , making them the most magnetized objects in the universe .
the large - scale reorganization of the magnetic field is thought to produce the observed gfs .
the formation rate of magnetars is expected to track that of stars , which is @xmath51 in our galaxy @xcite .
this suggests that batse would have triggered on such events as sgrbs at a rate of @xmath52 , compared with the all sky batse rate of about 150 @xmath53 . here
@xmath54 is the average rate of gfs in the galaxy similar to the 27 december event .
the observed isotropic distribution of short batse grbs on the sky and the lack of excess events from the direction of the virgo cluster suggests that only a small fraction , @xmath55 5% , of these events can be sgr gfs within 40 mpc , implying that @xmath56 on average for a galaxy like our own @xcite .
before _ swift _ detected sgrbs , searches for nearby galaxies within narrow inter planetary network ( ipn ) error boxes revealed already that only up to @xmath57 15% of them could be accounted for by magnetars capable of producing gfs @xcite .
finally , a search for sgrbs with spectral characteristics similar to that of gfs ( thermal spectra with @xmath58 kev ) in the batse catalog has concluded that a small fraction ( up to a few percent ) of them could have originated from magnetars in nearby galaxies @xcite .
magnetars are thought to be formed during core - collapse events inside massive stars , and because of their relatively short lifetimes as observable sources , @xmath59 yr , would naturally be located in predominantly star - forming galaxies , while essentially none should be seen in ellipticals .
one possible distinction of these from the classic sgrb population may well come from radio observations , because their radio afterglows should not be detectable beyond @xmath6 1 mpc @xcite .
the fraction of sgr events among what are now classified as short grbs may not be dominant , but it should be detectable and can be tested with future _ swift _ observations .
current observational limits @xcite indicate that any supernova - like event accompanying sgrbs would have to be over 50 times fainter than normal type ia sne or type ic hypernovae , 5 times fainter than the faintest known ia or ic sne , and fainter than the faintest known type ii sne .
these limits strongly constrain progenitor models for sgrbs .
the limits derived for grb 050709 ( filled triangles ) and grb 050509b ( empty triangles ) are plotted in figure [ fig : snlim ] along with two sn light curves as they would appear at z = 0.225 .
the type ic sne plotted are the very energetic type ic sn 1998bw associated with the long grb 980425 @xcite and the faint , fast - rise type ic sn 1994i @xcite .
the absence of a sn severely rules out models predicting a normal sn ia associated with sgrbs .
likewise , current observations disfavor a progenitor uniquely associated with massive stellar collapse . observations of long grbs at @xmath60 are consistent with all having sn bumps @xcite .
the situation is less clear at @xmath61 where all but two long grbs have had sn features : grb 980425 @xcite ; grb 031203 @xcite ; grb 030329 @xcite ; grb 011121 @xcite ; but no sn was detected in grbs 060505 and 060614 down to a flux limit at least hundreds of times fainter than sn 1998bw @xcite .
the distribution of time delays between progenitor formation and explosion is not yet well understood .
it can be constrained using , for example , the sgrb rate as a function of redshift @xcite .
if sgrb explosions lag star formation by some considerable amount of time , then the intrinsic redshift distribution of sgrbs would be significantly skewed to lower @xmath39 when compared to the universal star formation rate ( sfr )
. a cursory comparison of the redshift distribution of sgrbs with the sfr reveals what appears to be a significant time delay of a few gyr .
it should be noted that there are inherent biases in the discovery of a grb at a given redshift which are often difficult to quantify , such as complex trigger efficiencies and non - detections .
an alternative and perhaps less restrictive approach may be to use the rates of sgrbs in different types of galaxies @xcite .
on average , early - type galaxies have their stars formed earlier than late - type galaxies , and this difference , together with the time delay between progenitor formation and sgrb outburst , inevitably leads to different burst rates in the two types of galaxies .
for instance , the morphological types for the bursts in table [ tab : hosts ] reflect a higher incidence of early - type galaxies than type ia supernovae and suggest progenitor lifetimes significantly exceeding 6 gyr @xcite .
one difficulty with this idea is that star formation processes in these two types of galaxies might not be identical .
for example , elliptical galaxies can form by the merging of two gas - rich galaxies .
many globular clusters can form in the merging process , which could enhance , for example , the fraction of binary progenitors and also change the lifetime distribution .
obviously , these estimates are only sketchy and should be taken as an order of magnitude estimate at present . however , it should improve as more hosts are detected . on the other hand
, a large progenitor lifetime would help explain the apparent high incidence of galaxy cluster membership .
this can be naturally explained by the fact that galaxies in over dense regions form earlier in hierarchical cosmologies and thus make a substantial contribution to the local stellar mass inventory @xcite .
detailed observations of the astrophysics of individual grb host galaxies may thus be essential before stringent constraints on the lifetime of sgrb progenitors can be placed . if confirmed with further host observations , this tendency of sgrb progenitors to be relatively old can help differentiate between various ways of forming a sgrb . + this concludes our compendium of the _ facts_. for ease of reference in the chapters that follow , they have been assembled here with a minimum of speculative interpretation .
sgrbs have been observed assiduously throughout the electromagnetic spectrum only recently and although we now know much about their collective and individual properties , we are still long way from being sure how they operate . these sgrb sources involve energies that can exceed @xmath62 ergs , the mass equivalent of 1/10,000 of a sun . compared with the size of the sun , the seat of this activity is extraordinarily compact , as indicated by rapid variability of the radiation flux on time scales as short as milliseconds .
it is unlikely that mass can be converted into energy with better than a few ( up to ten ) percent efficiency ; therefore , the more powerful sgrb sources must `` process '' upwards of @xmath63 through a region which is not much larger than the size of a neutron star ( ns ) or a stellar mass black hole ( bh ) .
no other entity can convert mass to energy with such a high efficiency , or within such a small volume .
well - known arguments connected with opacity , variability time scales and so forth require highly relativistic and variable outflow @xcite .
best - guess numbers are lorentz factors @xmath64 in the range @xmath65 to @xmath66 , allowing rapidly - variable emission to occur at radii in the range @xmath67 to @xmath68 cm @xcite .
because the emitting region must be several powers of ten larger than the compact object that acts as trigger , there is a further physical requirement : the original energy outflowing in a wind would , after expansion , be transformed into bulk kinetic energy .
this energy can not be efficiently radiated as @xmath0-rays unless it is re - randomized .
the emitted energy is an observable diagnostic of the microphysical processes of particle acceleration and cooling occurring within the bulk flow @xcite .
it is beyond the scope of this article to describe the properties of these outflows and the physical processes occurring within , and we shall confine our attention to properties directly relevant to their origins ( an excellent account of the structure and energetics of sgrb outflows is given in @xcite . ) .
this review is therefore not complete and in its emphasis reflects the biases of the authors .
based on current models , the simplest hypothesis that the afterglow is due to a relativistic expanding blast wave seems to agree with the present data @xcite .
the complex time - structure of some bursts suggests that the central engine may remain active for up to 100 seconds @xcite .
however , at much later times all memory of the initial time - structure would be lost : essentially all that matters is how much energy and momentum has been injected , its distribution in angle and velocity .
however we can at present only infer the energy per solid angle ; there are reasons to suspect that the afterglow is not too narrowly beamed ; on the other hand the constraints on the angle - integrated @xmath0-ray energy are not strong @xcite .
as regards the trigger , there remain a number of key questions .
what is the identity of their progenitors ?
what is the nature of the triggering mechanism , the transport of the energy and the time scales involved ?
does it involve a black hole orbited by a dense torus ? and , if so , can we decide between the various alternative ways of forming it ?
the presence of sgrbs in old stellar populations helps rule out a source uniquely associated with recent star formation , while the lack of an accompanying supernova is strong evidence against a core - collapse origin .
there is now a stronger motivation to develop models in fuller detail .
this article outlines some of these issues .
in this section , we present a partial summary of some general ways in which gravity , angular momentum and the electromagnetic field can couple to power ultra relativistic outflows , along with a review of the most popular current models for the central source .
there are four sections : [ sec : general ] gives a brief account of the arguments in favor of a gravitationally fueled origin ; [ sec : bestiary ] describes the attributes of the most widely favored and conventional progenitors ; the various modes of energy extraction from such systems are then discussed in
[ sec : pathways ] ; finally , [ sec : obs ] gives a compendium of the types of observation that might help discriminate among the various progenitor models .
shortly after the discovery of quasars in 1963 , it was suggested that accretion of gas onto a compact massive body was responsible for their enormous energy output @xcite , which can be hundreds of times larger than that of entire galaxies .
such a body is essentially a very efficient converter of gravitational binding energy into radiation .
the deeper the gas can fall into the potential well before the radiation is converted , the more efficient the process , hence the appealing nature of compact objects . for black holes approximately
@xmath69 erg g@xmath2 ( where @xmath70 is the radius of the marginally stable orbit ) can be released , and even more if the hole is endowed with a large angular momentum .
this efficiency is over a hundred times that traditionally associated with thermonuclear reactions ( hydrogen burning releases @xmath71 erg g@xmath2 ) .
since the 1960s , we have also learned about stellar mass black holes , where @xmath72 m@xmath73 , in galactic binary systems and ultra - luminous x - ray sources which , with less confidence , we also associate with black holes , primarily on energetic grounds . in these objects ,
accretion ( and the accompanying radiation ) is usually thought to be limited by the self regulatory balance between newtonian gravity and radiation pressure .
a fiducial luminosity is the _ eddington _ limit associated with quasi - spherical accretion , at which radiation pressure balances gravity .
if thomson scattering provides the main opacity and the relevant material is fully ionized hydrogen , then this luminosity is @xmath74 @xmath75 being the thomson cross - section .
this may be converted to a mass accretion rate if one considers that the accretion luminosity is @xmath76 , giving @xmath77 where @xmath78 is the radius of the compact object .
now clearly this applies strictly only in a quasi - spherical configuration and a steady state .
relaxing these assumptions allows for greater luminosities in transient events , or for configurations in which the energy release is somehow collimated .
however they will not be greater than the expression given above by more than a factor of a few .
the photon luminosity , for the duration of a typical short burst ( a few seconds at most ) , is thousands of times larger than that of any active galactic nucleus ( thought to involve supermassive black holes ) , and is 12 orders of magnitude above the limit ( [ eq : edd ] ) .
the total energy , however , is not very far off from that of other phenomena encountered in astrophysics , and is in fact reminiscent of that released in the core of a supernova .
the eddington photon limit ( [ eq : edd ] ) is circumvented if the main cooling agent is emission of neutrinos rather than electromagnetic waves .
the associated interaction cross section is then many orders of magnitude smaller , and the allowed accretion rates and luminosities are correspondingly higher .
for example , using the cross section for neutrino pair production , the eddington limit can be rewritten as @xmath79 with an associated accretion rate , assuming unit efficiency for conversion of mass into neutrino energy , @xmath80 the time it would take an object to radiate away its entire rest mass energy in this way is a mass - independent _ eddington time _ given by @xmath81 while the time scale over which an accretion - driven source would double its mass is @xmath82 .
the dynamical time scales near black holes are modest multiples of @xmath83 , where @xmath84 is the characteristic size of the collapsed object @xmath85 and are therefore much shorter than @xmath86 . a fiducial eddington density , characteristic near the horizon when the hole accretes at the eddington rate , is @xmath87 it should be noted that the typical thomson optical depth under these conditions is @xmath88 and so , as expected , photons are incapable of escaping and constitute part of the fluid . for completeness
, we can also define an eddington temperature , as the black body temperature if a luminosity @xmath89 emerges from a sphere of radius @xmath90 , @xmath91 or @xmath92 and an eddington magnetic field strength @xmath93 finally , for comparison , we define @xmath94 as the temperature the accreted material would reach if its gravitational potential energy were turned entirely into thermal energy .
it is given by @xmath95 some authors use the related concept of the virial temperature , @xmath96 , for a system in mechanical and thermal equilibrium . in general
, the radiation temperature is expected to be @xmath97 . in deriving the above estimates we have assumed that the radiating material can be characterized by a single temperature .
this may not apply , for example , when a hot corona deforms the neutrino spectrum away from that of a cooler thermal emitter @xcite .
although we have considered here the specific case of neutrino pair creation , the estimates vary little when one considers for example coherent scattering of neutrinos by nuclei and/or free nucleons ( except for the energy scaling ) .
similar overall fiducial numbers also hold for neutron stars , except that the simple mass scalings obtained here are lost .
it is thus clear from the above estimates that when mass accretes onto a ( stellar - mass ) black hole or neutron star under these conditions , the densities and temperatures are so large ( @xmath98 g @xmath99 , @xmath100 k ) that : ( i ) photons are completely trapped ; and ( ii ) neutrinos , being copiously emitted , are the main source of cooling since they can mostly escape .
this regime , which requires correspondingly large accretion rates , is termed hypercritical accretion , and was considered for sn1987a shortly after the explosion @xcite .
such high accretion rates are never reached for black holes in xrbs or agn , where the luminosity remains well below the photon eddington rate ( [ eq : edd ] ) .
they can , however , be achieved in the process of forming neutron stars and stellar mass black holes during the collapse of massive stellar cores .
note that at sufficiently high accretion rates , the density reaches the threshold for optical thickness even to neutrinos , which can not then simply stream out ( this occurs at @xmath101 g @xmath99 ) .
the current view is that sgrbs arise in a very small fraction ( approximately @xmath102 ) of stars which undergo a catastrophic energy release event toward the end of their evolution .
one conventional possibility is the coalescence of binary neutron stars @xcite .
double neutron star binaries , such as the famous psr1913 + 16 @xcite , will eventually coalesce due to angular momentum and energy losses to gravitational radiation . the resulting system could be top heavy and unable to survive as a single neutron star
. however , a black hole would be unable to swallow the large amount of angular momentum present .
the expected outcome would then be a spinning hole , orbited by a torus of ns debris .
other types of progenitor have been suggested - e.g. a ns - bh merger @xcite , where the neutron star is tidally disrupted before being swallowed by the hole ; the merger of a white dwarf ( wd ) with a black hole @xcite ; the coalescence of binary wds @xcite ; or accretion induced collapse ( aic ) of a ns @xcite , where the collapsing neutron star has too much angular momentum to collapse quietly into a black hole . in an alternative class of models , it is supposed that the compact objects are contained within a globular cluster , and that the binary system will evolve mainly through hardening of the binary through three
body interactions @xcite or physical star - star collisions @xcite rather than by pure gravitational wave emission .
finally , violent reconfigurations of the magnetic field on magnetars @xcite may offer another possibility .
table [ tab : rates ] provides a summary of the various rate estimates for some of these possible sgrb progenitors . aside from the rate of sne events ,
the rate of sgrbs and plausible progenitors in table [ tab : rates ] are highly imprecise .
the main sources of uncertainty are related to supernova kicks , the mass ratio distribution in binaries , mass limits for black hole formation , stellar radii and common envelope evolution @xcite .
lc progenitor & rate ( @xmath103 ) + ns ns & 1 - 800 + bh
ns & 0.11000 + bh wd & 0.01 - 100 + ns aic & 0.1 - 100 + wd - wd & 3000 + sn ib / c & 60000 + sn ia & 150000 + sgrbs & 10@xmath104 + the rate of gravitational mergers among bh - bh and bh - ns binaries depends on the orbital periods after spiral in , and hence on the ill understood details of the process .
a binary with total mass @xmath105 , mass ratio @xmath106 , and period @xmath107 ( in days ) in a nearly circular orbit will merge in @xmath108 thus for a coalescence to occur within a hubble time , spiral in must have reduced the orbital period to @xmath55 1 day . since a black hole can have a mass exceeding that of its companion helium core , spiral in need not occur , as opposed to the case of a lower mass object like a neutron star ( except in very non conservative mass transfer ) . if the companion has a strong wind , spiral in might never begin . if it does , envelope ejection might be less efficient than that by a neutron star , since black hole accretion could have a very low efficiency . if spiral in does occur , however , the subsequent evolution may be simpler than in the double neutron star case .
helium stars less massive than 3@xmath40 expand their envelopes dramatically during core carbon burning @xcite , requiring a second spiral - in to make a ns - ns binary which could merge in a hubble time .
the more massive he cores which could leave black hole remnants do not expand much , so a second spiral - in is not needed .
thus bh - ns binaries form at rates comparable to the ns - ns rate , but the fraction which merge depends on the miasma of mass transfer and spiral - in which determine the final period distribution .
gravitational waveforms of any merging systems will allow the masses of the merging bodies to be accurately determined @xcite , shedding light on the underlying physics , as well as on the minimum black - hole mass determined by post - collapse infall .
the much more certain ns - ns mergers will also allow the mass distribution of neutron stars to be determined .
how might such a progenitor generate a relativistic outflow or a sudden release of electromagnetic energy ?
we now address various possible routes .
it has become increasingly apparent in the last few years that most plausible sgrb progenitors suggested so far ( e.g. ns - ns or ns - bh , wd - bh mergers or physical collisions , and ns aic ) are expected to lead to a black hole plus debris torus system ( figure [ fig : prog ] ) . in this case there is no external agent feeding the accretion disk , and thus the event is over roughly on an accretion time scale ( which would be on the order of one second ) .
a possible exception includes the formation from accretion @xcite or from a wd - wd @xcite or ns - ns merger @xcite of a rapidly rotating ( in some cases very massive ) neutron star with an ultrahigh magnetic field ( figure [ fig : prog ] ) .
if there is an ordered field @xmath109 , and a characteristic angular velocity @xmath110 , for a spinning compact source of radius @xmath78 , then the magnetic dipole moment is @xmath111 .
general arguments suggest @xcite that the non thermal magnetic - dipole - like luminosity will be @xmath112 , and simple scaling from these familiar results of pulsar theory require fields of order @xmath113 g to carry away the rotational or gravitational energy ( which is @xmath114 erg ) in a time scale of seconds @xcite .
one of the most perceptive theoretical discoveries that was made about black holes was that , when they spin , a fraction of their mass can be ascribed to rotational energy and is , in principle , extractable @xcite .
this is most convincingly demonstrated by observing that there exist orbits of test particles with negative total energy , ( including their rest mass ) , within the ergosphere .
if an infalling plasma cloud is attached to magnetic field lines anchored at a large distance ( e.g. , in an accreting torus ) , and the field drags the cloud backwards relative to the rotation of the hole placing it on an orbit with negative energy , the work that has been performed can be thought of as energy that has been effectively extracted from the spin of the black hole @xcite .
the binding energy of the orbiting debris , and the spin energy of the bh are thus the two main reservoirs for the case of a black hole central engine : up to 42% of the rest mass energy of the torus , and 29% of the rest mass energy of the black hole itself can be extracted for a maximal black hole spin .
sgrb activity can be powered by the black hole only as long as there is interaction with the surrounding gas , which will probably need to be centrifugally supported ( although even for relatively low angular momentum a considerable amount of energy can still be extracted @xcite ) . the angular momentum is quite generally a crucial parameter , in many ways determining the geometry of the accretion flow .
even a little rotation can make a big difference , breaking the spherical symmetry and producing accretion _ disks _ instead of _ radial _ inflow , as envisaged originally by bondi @xcite .
if the gas has no angular momentum and the magnetic field is dynamically unimportant , there will be essentially radial inflow .
spherical accretion onto black holes is relatively inefficient despite the deep potential well , because the gas is compressed , but not shocked , and thus can not easily convert gravitational to thermal energy .
the flow pattern changes dramatically if the inflowing gas has a small amount of angular momentum .
the quasi - spherical approximation breaks down when the gas reaches a radius @xmath115 , where @xmath116 is the angular momentum per unit mass , and if injection occurs more or less isotropically at large radii , a familiar accretion disk will form ( as occurs in x ray binaries and white dwarf binary systems ) .
the matter will instead dissipate its motion perpendicular to the plane of symmetry and form a differentially rotating disk , the rotational velocity at each point being approximately keplerian , and then gradually spiral inwards as viscosity transports its angular momentum outwards .
if the emission process is very efficient , the disk is dynamically cold and geometrically thin , in the sense that locally , @xmath117 and the pressure scale height @xmath118 . if gas passing through a thin disk reaches a radius within which the internal pressure builds up either because it is unable to cool in an inflow time or because the radiation pressure force is competitive with gravity the disk will become geometrically thick , with @xmath119 . in a thick disk or torus
the pressure provides substantial support in the radial as well as the vertical direction , and the angular momentum distribution ( now as a function of height as well as radius ) may be far from keplerian . both types of configurations have been studied extensively and the agreement in some cases between observation and theory is extremely good , assuming a fundamentally empirical recipe for angular momentum transport @xcite . as mentioned above in [ sec : general ] , in principle flow onto a compact object can liberate gravitational potential energy at a rate approaching a few tenths of @xmath120 , where @xmath121 is the mass inflow rate . even for such high efficiencies the mass requirements of the more luminous sgrb sources are rather high , with @xmath122 where @xmath123 is the overall efficiency .
the inner regions of disks with mass fluxes in this range are generally able to cool by neutrinos on time scales shorter than the inflow time .
if @xmath124 , then the bulk of the neutrino radiation comes from a region only a few gravitational radii in size , and the physical conditions can be scaled in terms of the _ eddington quantities _ defined in [ sec : general ] .
the remaining relevant parameter , related to the angular momentum , is @xmath125 , where @xmath126 is the free fall velocity .
the inward drift speed @xmath127 would be of order @xmath128 for supersonic radial accretion . when angular momentum is important , this ratio depends on the mechanism for its transport through the disk , which is related to the effective shear viscosity . for a thin disk ,
the factor @xmath129 is of order @xmath130 , where @xmath131 is the scale height at radius @xmath132 and @xmath45 is the phenomenological viscosity parameter .
suppose that a given accretion rate yields a luminosity @xmath133 with an efficiency @xmath134 .
then the characteristic density , at a distance @xmath132 from the hole , with account of the effects of rotation , is @xmath135 and the maximum magnetic field , corresponding to equipartition with the bulk kinetic energy , would be @xmath136 any neutrinos emerging directly from the central _ core _ would have energies of a few mev .
note that , as mentioned above , @xmath137 is far below the _ virial _ temperature @xmath138 .
the flow pattern when accretion occurs would be then determined by the value of the parameters @xmath139 , which determine the importance of radiation pressure and gravity , and the ratio @xmath140 , which fixes the temperature if a stationary flow pattern is set up .
the preceding general discussion of neutrino cooled accretion flows thus provides a basis for general models .
so far we have discussed the power output that might be generated by the accretion process , but we have made no attempt to describe in detail the flow of the accreting gas . a hint that its dynamics may not be straightforward is provided by the existence of the eddington limit for neutrinos , clearly illustrating that for the high accretion rates expected in sgrbs , forces other than gravity may be important .
in addition , in many cases ( and probably in most ) the accreting matter possesses considerable angular momentum per unit mass which it has to lose somehow in order to be accreted at all .
the reader is referred to [ sec : disks ] for fuller details .
we now concentrate briefly on the origins of the jets which provide the power for sgrb sources .
pairs or photons ( i ) .
the neutrinos that are emitted from the inner regions of the debris deposit part of their energy in the outer parts of the disk ( ii ) , driving a strong baryonic outflow .
this wind may be responsible for collimating the jet .
_ panel b : _ strong magnetic fields anchored in the dense matter can convert the binding and/or spin energy into a poynting outflow .
a dynamo process of some kind is widely believed to be able to operate in accretion disks , and simple physical considerations suggest that fields generated in this way would have a canonical length - scale of the order of the disk thickness ( i ) .
open field lines can connect the disk outflow and may drive a hydromagnetic wind ( ii ) .
the above mechanism can tap the binding energy of the debris torus , while a rapidly rotating hole could contain an even larger energy reservoir , extractable in principle through mhd coupling to the exterior ( iii;iv ) .
b _ adapted from @xcite.,width=480 ] two ingredients are necessary for the production of jets : first , there must be a source of material with sufficient free energy to escape the gravitational field of the compact object ; second , there must be a way of imparting some directionality to the escaping flow .
our eventual aim must be to understand the overall flow pattern around a central compact object , involving accretion , rotation , and directional outflow but we are still far from achieving this .
most current works who have discussed outflow and collimation have simply invoked some central supply of energy and material .
a self - consistent model incorporating outflow and inflow must explain why some fraction of the matter can acquire a disproportionate share of energy ( i.e. , a high enthalpy ) .
a brief summary of the various metabolic pathways is presented in figure [ fig : diag ] .
the neutrino luminosity emitted when disk material accretes via viscous ( or magnetic ) torques on a time scale @xmath141 s is roughly @xmath142 for a canonical radiation efficiency of 0.1 .
one fairly direct solution is to reconvert some of this energy via collisions outside the disk into electron - positron pairs or photons @xcite .
if this occurs in a region of low baryon density ( e.g. , along the rotation axis , away from the equatorial plane of the disk ) a relativistic pair - dominated wind can be produced . an obvious requirement for this mechanism to be efficient
is that the neutrinos escape ( free streaming , or diffusing out if the density is high enough ) in a time scale shorter than that for advection into the black hole .
the efficiency for conversion into pairs ( scaling with the square of the neutrino density ) is too low if the neutrino production is too gradual , so this can become a delicate balancing act .
typical estimates suggest a lower bound of @xmath143 when the entire surface area emits close to a single temperature black body
. the efficiency may be significantly larger if dissipation takes place in a corona like environment @xcite .
one attractive energy extraction mechanism that could circumvent the above restriction in efficiency is a relativistic magneto hydrodynamic ( mhd ) wind @xcite .
such a wind carries both bulk kinetic energy and ordered poynting flux , and it is possible that gamma - ray production occurs mainly at large distances from the source @xcite . a rapidly rotating neutron star ( or accretion disk ) releases energy via magnetic torques at a rate @xmath144 where @xmath145 is the spin period , and @xmath109 is the strength of the poloidal field at a radius @xmath132 .
the last stable orbit for a schwarzschild hole lies at a coordinate distance @xmath146 km , to be compared with @xmath147 km for an extremal kerr hole .
thus the massive neutron disk surrounding a schwarzschild black hole of approximately @xmath148 should emit a spin - down luminosity comparable to that of a millisecond neutron star .
a similar mhd outflow would result if angular momentum were extracted from a central kerr hole via electromagnetic torques @xcite .
the field required to produce @xmath149 is enormous , and may be provided by a helical dynamo operating in hot , convective nuclear matter with a millisecond period @xcite .
a dipole field of the order of @xmath113 g appears weak compared to the strongest field that can in principle be generated by differential rotation ( @xmath150^{-1}\;{\rm g}$ ] ) , or by convection ( @xmath151 ) , although how this may come about in detail is not resolved .
note , however , that it only takes a residual torus ( or even a cold disk ) of @xmath152to confine a field of @xmath113 g. orbiting debris with such large magnetic seed fields and turbulent fluid motions will give rise to a plethora of electromagnetic activity ( figure [ fig : diag ] ) .
the topic lies beyond the scope of this paper and we refer the reader to the excellent review by blandford @xcite .
a potential death - trap for such relativistic outflows is the amount of entrained baryonic mass from the surrounding medium .
for instance , a poynting flux of @xmath153 erg could not accelerate an outflow to @xmath154 if it had to drag more than @xmath155 of baryons with it .
a related complication renders the production of relativistic jets even more challenging , because the high neutrino fluxes are capable of ablating baryonic material from the surface of the disk at a rate @xcite @xmath156 thus a rest mass flux @xmath157 limits the bulk lorentz factor of the wind to @xmath158 assuming that the external poloidal field strength is limited by the vigor of the convective motions , the spin - down luminosity scales with neutrino flux as @xmath159 , where @xmath160 is the convective velocity .
the ablation rate given in equation ( [ eq : ablation ] ) then indicates that the limiting bulk lorentz factor @xmath161 of the wind decreases as @xmath162 .
thus the burst luminosity emitted by a magnetized neutrino cooled disk may be self - limiting .
mass loss could , however , be suppressed if the relativistic wind were somehow collimated into a jet .
this suggests that centrifugally driven mass loss will be heaviest in the outer parts of the disk , and that a detectable burst may be emitted only within a relatively small solid angle centered on the rotation axis . in the preceding sections , we have endeavored to outline some of the basic physical processes that are believed to be of most relevance to interpreting sgrb sources , not so much because we believe that there is strong evidence in favor of them , but instead because it provides a framework in which to discuss the observations and to demonstrate that , as yet , sgrbs pose no threat to conventional physics .
of course , we are conscious that most observations tell us less about the primary source than the about secondary reprocessing of this power in the circumburst and galactic environment from which we can learn about the central engine only by a chain of uncertain inferences . unless sgrbs are eventually found to be accompanied by telltale emission features like the supernovae of long - duration grbs , the only definitive understanding of the progenitors will come from possible associations to direct gravitational or neutrino signals .
spectral investigations will nonetheless be important as probes of the basic energetics and microphysical parameters of relativistic shocks @xcite .
we are more sanguine about the possibility of deriving accurate particle densities from afterglow observations . however , there is the possibility that occurrence in an unusually low density environment could cause the external shock to occur at much larger radii and over a much longer time scale than in usual afterglows @xcite
. the x - ray afterglow intensity could be then below the threshold for triggering
. this may be the case for sgrbs arising from compact binaries that are ejected from the host galaxy into an external environment that is much less dense that the ism assumed for usual models .
another possibility for an unusually low density environment , made up only of very high energy but extremely low density electrons , is if the grb goes off inside a pulsar cavity inflated by one of the neutron stars in the precursor binary @xcite .
such bubbles can be as large as fractions of a parsec or more , giving rise to a deceleration shock months after the sgrb with a consequently much lower brightness that could avoid triggering and detection .
the difference between the low - density and high - density environment cases could be tested if future observations of afterglows reveal a correlation with the degree of galaxy clustering or with individual galaxies .
the association of sgrbs with both star - forming galaxies and with ellipticals dominated by old stellar populations , to some researchers , suggested an analogy to type ia supernovae , as it indicated a class of progenitors with a wide distribution of delay times between formation and explosion .
similarly , just as core - collapse supernovae are discovered almost exclusively in late - time star - forming galaxies , so too are long grbs .
indeed , a detailed census of the types of host galaxies , burst locations and redshifts should help decide between the various alternatives suggested in [ sec : bestiary ] .
this is because if the progenitor lifetime is long and the systemic kick is small , then the bursts should correspond spatially to the oldest populations in a given host galaxy . for early - type galaxies ,
the distribution would presumably follow the light of the host @xcite .
in contrast , a neutron star binary could take millions of years to spiral together , and could by then , especially if given a substantial kick velocity on formation , have moved many kiloparsecs from its point of origin .
the burst offsets would then presumably be larger for smaller mass hosts .
as new redshifts , offsets , and host galaxies of sgrbs are gathered , the theories of the progenitors will undoubtedly be honed .
sgrb activity manifests itself on many different scales . the simplest hypothesis , however , is that the central _ prime mover _ or _ trigger _ is qualitatively similar in all events .
the primary energy may be reprocessed in different ways , depending on the details of the circumburst ( and galactic ) environment .
a brief summary of the various evolutionary pathways that may be involved in producing a sgrb is given in
[ sec : bestiary ] .
the important message is that most current alternatives involve either an accreting stellar mass black hole ( neutron star ) or a precursor stage whose inevitable end point is such an object .
furthermore , as we discussed in [ sec : general ] , a black hole ( neutron star ) embedded in infalling matter offers a more efficient power source than any other conceivable progenitor ; it is our firm prejudice that sgrbs are energized by mechanisms that involve either neutron stars or black holes . for this reason , in this section , we turn out attention to the formation of the _ trigger _ : the actual astrophysical system that is capable of , and presumably powers the observed burst .
it is beyond the scope of this article to describe completely all evolutionary pathways , and we shall confine our attention to binary mergers and physical collisions . the study of compact binaries , particularly those involving pulsars , has evolved dramatically over the past thirty years , and carries a great deal of historical baggage . consequently , most theoretical work has been directed towards describing these binary encounters .
our discussion will reflect this bias .
there are eight sections :
[ sec : mass ] gives a summary of observed compact object masses ;
[ sec : couples ] gives an overall picture of the merger scenario as we best understand it now ; then in [ sec : early ] we describe early considerations of these events ; this is followed in
[ sec : dynstab ] by a summary of dynamical stability considerations in such systems ; the merger itself is considered in
[ sec : merger ] ; the effects of general relativity and the fate of the remnant core are addressed in
[ sec : greffects ] and [ sec : corefate ] respectively ; finally , the outcome of compact object collisions is studied in [ sec : colls ] .
the mass of the compact object dictates a characteristic length and luminosity scale for sgrb activity , so we must consider what we know ( and what we do not ) from observations of systems containing compact objects in a more leisurely state of affairs . measuring stellar masses , although a considerable observational challenge , is not nearly as difficult as measuring their radii ( which is why the equation of state at nuclear densities remains one of the primary open questions in compact object and nuclear physics studies ) .
when they occur in binaries , accurate mass determinations are possible ( figure [ masses ] ) , and for cases where there are two compact objects in tight orbits , the constraints can be truly spectacular ( psrb1913 + 16 @xcite , psrb1534 + 12 @xcite , and psrj0737 - 3039a , b @xcite ) .
neutron star masses are tightly clustered around the chandrasekhar mass , at 1.4@xmath163 .
those that deviate farthest from this value towards high masses are found in low masss x ray binaries @xcite , where the slow but steady mass transfer from a companion has produced top
heavy objects ( the most dramatic example being j0751 + 1807 at 2.1m@xmath164 @xcite ) . at such an extreme ,
it is possible that conversion to strange ( quark ) matter @xcite may have occurred ( see jaikumar et al . for a review @xcite ) .
incredibly , it is possible that observations of superbursts ( unusually long type i x ray bursts ) may be able to discern if this is actually the case @xcite . and @xmath165 ( where the scatter is much larger ) .
the data are taken from @xcite for neutron stars , and @xcite for black holes.,width=432 ] regarding black holes , the scatter is considerably greater , probably reflecting the fact that a threshold mass is necessary to create one in a core collapse event , but no obvious hard limit exists at high masses , and a degree of variation is expected @xcite .
the constraints are also less stringent in most cases , due to a combination of effects , but nevertheless a boundary can be drawn at roughly 3m@xmath164 , separating the most massive neutron star from the least massive black hole .
this is also close to the maximum mass allowed for a neutron star in hydrostatic equilibrium in general relativity , based only on the condition that the equation of state must preserve causality @xcite .
one would thus expect that encounters between objects of this type would have mass ratios of about unity if two neutron stars are involved , and 0.2 - 0.3 if one of each interacts .
although white dwarfs are not usually considered as direct progenitors of grbs , one should keep an open mind about their possible role , particularly in the light of unexpected discoveries made by _
swift _ concerning late time variability and the inferences about the parent population of sgrbs .
they may be responsible for the violent birth of neutron stars through accretion induced collapse or the merger of two dwarfs in a binary @xcite , and it is possible that this remnant could power a grb @xcite .
the mass of a white dwarf is somewhat lower that that of a neutron star , but they are larger , by a factor of at least @xmath166 , leading to a natural ( dynamical ) time scale that is greater than for neutron stars and black holes by a factor @xmath167 . as a stellar binary revolves ,
gravitational waves carry away its energy and angular momentum .
the loss rates depend on the system mass , semi major axis , @xmath168 , and eccentricity , and rapidly increase as orbital decay progresses . in the stage where @xmath169 , where @xmath78 is the typical stellar radius , the evolution can be computed to high accuracy in the weak field limit , using post
newtonian expansions for point masses @xcite .
the characteristic decay time for circular orbits of period @xmath145 is given by @xmath170 which , when applied to the binary pulsars psr1913 + 16 @xcite and psrj0737 - 3039 @xcite , gives @xmath171 myr and @xmath172 myr , respectively .
these waveforms , calculated for the last few minutes of the binary system s life , will be required as templates for detection with interferometric gravitational wave detectors such as ligo @xcite and virgo @xcite . for more common binaries with much longer periods
, the decay rate will be negligible and the emission may be observed with the space based mission lisa . for systems reaching small separations , the decay rate will increase catastrophically and a collision will ensue , followed by relaxation to a final , nearly steady state in which , in many cases , only a black hole will remain .
this is the ring down phase ,
in which the emission of gravitational waves and the configuration of the fossil black hole can be studied through perturbative methods on a kerr background ( see @xcite for a review ) .
a composite gravitational radiation signal is shown in figure [ gravwaves ] as an example .
the decay at early times ( @xmath173 ) was computed using the weak - field limit approximation for point masses , while the merger waveform itself ( @xmath174 ) is the result of a three - dimensional numerical simulation for the disruption of a neutron star by a point mass .
the initially slow decay accelerates ( while the amplitude and frequency of the signal increase ) and comes to an abrupt end when the separation is of the order of the stellar radius , giving a complex signal which depends on the details of the neutron star structure . in principle ,
observation of such radiation and its frequency power spectrum could help constrain the mass radius relationship for neutron stars @xcite . and
adiabatic index @xmath175 ( left column @xcite ) and @xmath176 ( right column @xcite ) . in this example , the observer is directly above the orbital plane of the binary along the rotation axis .
the rise in amplitude and frequency shows the characteristic accelerated phase of decay , followed by a rapid transition to a ring down after the violent tidal disruption of the star at @xmath177 . in the bottom panels , details of the merger and tidal disruption waveform
are shown , from whose spectrum one could in principle extract valuable information regarding the equation of state at neutron star densities .
note that the neutron star with the more compressible equation of state is fully and rapidly disrupted , resulting in the total disappearance of the gravitational wave signal , whereas the one with a stiff equation of state leaves behind a low - mass orbiting remnant , which produces persistent emission at late times ( see also figure [ profiles]).,width=480 ] the middle ground , during which a violent redistribution of angular momentum and energy takes place , will produce a burst of gravitational , neutrino , and electromagnetic energy , perhaps in the form of a classical grb @xcite .
there is no approximate , analytical solution to be found in this regime , and multidimensional numerical simulations are clearly required to address the behavior of the system .
the process is highly dynamical , with thermodynamics and emission processes mattering little , and gravitational dynamics ruling the evolution ( with the possible exception of systems containing white dwarfs ) .
even before high resolution numerical simulations were carried out in essentially newtonian gravity , it was assumed that the dynamical merger would indeed result in the formation of an accretion disk with enough mass and internal energy to account for the energy budget of a typical grb , either through the tidal disruption of the neutron star in the former , or post merger collapse of most of the central core in the latter .
compact binaries were thus seriously suggested as possible progenitors for grbs at cosmological distances , with possible maximum power in excess of 10@xmath178 erg s@xmath2 @xcite .
the characteristics of the problem , and the qualitatively different nature of the questions that can be of interest ( or addressed ) has dictated a range of approaches when searching for a solution .
on the one hand , there is the question of generation of gravitational waves , clearly deserving of substantial effort in its own right , and tackled as such by several researchers . on the other , the generation of an electromagnetic signal ( as a grb , or another observable transient ) requires the consideration of a different kind of input physics , mainly in the domain of thermodynamics , the equation of state under a wide array of conditions , nuclear physics , weak interactions and magnetohydrodynamics .
efforts in the field have generally proeeded along these largely orthogonal axes , with little overlap between them until quite recently , and it is only now that numerical simulations with any realistic microphysics are beginning to employ the full machinery of general relativity .
it is thus only natural that results midway along this path have not always been in agreement , and indeed are not in some aspects even now .
the recent and spectacular advances in the successfull evolution of black hole binaries in numerical relativity give us today a glimpse of what may be possible to study in greater detail in the coming years ( see the recent overview by j. centrella @xcite ) . to our knowledge , the merger of compact objects ( i.e. , neutron stars and/or black holes ) was first considered by james lattimer and david schramm @xcite .
they initially studied the dynamics and disruption of a neutron star by a black hole in a bound system , and the discovery of the hulse taylor pulsar in 1974
@xcite soon made it clear that such encounters were an inevitable , if rare in terms of the lifetime of a typical galaxy , consequence of binary stellar evolution pathways .
the computations were carried out in full general relativity , with certain simplifications to make them tractable and their interest originally laid in the possibility of such events ejecting neutron rich material to the interstellar medium , which might subsequently undergo r
process nucleosynthesis and contribute to the observed abundances of heavy elements @xcite .
grbs had only recently been made known to the astrophysical community @xcite , and the likelihood of them being of galactic origin made lattimer & schramm consider such events to be unrelated . in their words : `` although the frequency of these events is too small to be important as far as currently observed anti
neutrino , @xmath0ray , or optical bursts are concerned , enough neutron star material may be ejected to be of nucleosynthetic importance.''@xcite .
over thirty years after their pioneering work , coalescing compact binaries stand today as one of the leading runners in models for sgrbs .
it is noteworthy that only now are detailed computer simulations of such collisions beginning to fully consider the effects of general relativity in the dynamical merger phase of tidal disruption ( john a. wheeler s `` tube of toothpaste '' mechanism @xcite ) .
the stiffness of the nuclear equation of state , or equivalently , the compressibility , is a parameter derived from microphysics that has a predominant role in the final outcome of a merger .
it is most easily expressed by writing the pressure
density relation as @xmath179 where @xmath64 is the adiabatic index .
figure [ profiles ] shows the effect of its variation on the structure of a star by plotting the density profiles of static , spherically symmetric configurations in hydrostatic equilibrium . for @xmath180 ,
the result is a nearly constant density ( a stiff equation of state ) , while for @xmath181 ( lower values will not produce a stable solution in newtonian gravity ) the star is highly centrally condensed ( a soft equation of state ) .
it is immediately obvious that when placed in the vicinity of a massive companion , the star made of more compressible material will feel tidal effects to a much lesser degree .
the issue of how matter responds to pressure at extremely high densities is of course far from being solved .
for example , while the stiffness of ordinary matter may be quite high at @xmath182 , with @xmath183 , the presence of exotic condensates at supra
nuclear densities could soften the equation of state considerably @xcite , leading to potentially different behavior . ) .
the varying degree of central condensation makes the stiffer stars ( with higher values of @xmath64 ) more susceptible to tidal effects when placed in a binary system.,width=384 ] the equilibrium configurations for an incompressible fluid , and their stability , were considered by chandrasekhar @xcite , both for single and binary stars , in combinations that could represent double neutron stars or neutron stars in orbit around point masses .
this work was extended later by lai and collaborators @xcite to compressible configurations by the use of polytropic relations as equations of state . in a series of papers
, they showed that purely newtonian hydrodynamic effects can de stabilize close binaries if the equation of state is sufficiently stiff ( @xmath184 ) .
this can occur before contact , when the separation is typically @xmath185 stellar radii .
the effective binary potential becomes so steep that a dynamical ( i.e. , on an orbital time scale ) plunge occurs , forcing the two stars to come together .
this is independent of any consideration of general relativity , and clearly shows just how powerful the departures from point
mass behavior can be when the combination of a finite stellar radius and compressibility are considered .
for illustrative purposes , we show in figure [ jvsr ] the equilibrium curves of total orbital angular momentum for a newtonian neutron star point mass binary for two compressibilities , as well as the point mass result ( which is simply @xmath186 ) .
this has important implications for the initiation of mass transfer from one star to the other , the occurrence of tidal disruption and the final configuration of the system once it comes into near contact ( see e.g. , chapter 4 in @xcite ) . in a newtonian approximation .
the neutron star is modeled as a non spinning tri axial roche
riemann ellipsoid ( see @xcite ) of varying adiabatic index ( curves for @xmath187 are shown ) .
tidal effects increase the steepness of the effective interaction potential and produce a minimum in @xmath188 , indicating the onset of a dynamical instability .
the effect of the equation of state is clear , with the less compressible neutron star encountering instability , which will induce a dynamical plunge , at a greater separation than its more compressible counterpart . for reference ,
the equilibrium curve for point masses is also shown . as expected , it is a monotonically increasing function of radius and exhibits none of the turning points characteristic of the appearance of an unstable branch.,width=384 ] the compressibility also impacts directly upon the mass radius relation of a star . for polytropes , we have @xmath189 so the star will expand upon mass loss ( or contract ) for @xmath190 ( or @xmath191 ) .
thus it can further overflow its roche lobe once mass transfer begins , in which case the process itself is unstable leads to complete disruption , or retreat from the roche surface and , in principle , shut off mass transfer in the absence of external driving ( which will not occur , because gravitational radiation emission will remove orbital angular momentum continuously ) . in either case , the details of the dynamics will depend on the global adiabatic index at high densities .
the numerical studies of coalescing binaries began in the 1980s and 1990s , with computations of the gravitational wave emission and determinations of the stability and dynamics of the system at separations comparable with the stellar radius under various simplifying approximations @xcite .
the thermodynamical evolution of the fluid during merger , with the potential for electromagnetic energy release and grbs was also considered by various groups @xcite .
these studies used newtonian or post - newtonian gravity with additional terms included in the equations of motion to mimic gravitational radiation reaction ( which is the ultimate agent driving the binary evolution ) , and in some cases approximations to the equations of general relativity . , going from stiff to soft ( left to right ) .
logarithmic density contours of density ( spaced every quarter decade , with the lowest one at @xmath192}=11 $ ] ) , are shown in the orbital plane , with additional shading for @xmath193 .
the distance scale is indicated in each panel , and a black disk of radius @xmath194 represents the black hole .
all three cases initially had identical mass ratios , @xmath195 @xcite.,width=480 ] figure [ bhnsmergers ] shows the dynamical evolution of a merging neutron star
black hole binary , during which the neutron star ( modeled as a polytrope ) is tidally disrupted and an accretion disk promptly forms @xcite .
the general conclusion of this body of work ( which includes a mock radiation reaction force to account for gravitational wave emission ) is that when the orbital separation is of the order of a few stellar radii , the star becomes greatly deformed due to tidal effects , and the system becomes dynamically unstable . for stellar components with
relatively stiff equations of state ( @xmath196 ) the instability is due to a strong steepening of the effective potential because of tidal effects ( figure [ profiles ] ) . in the opposite limit , the mass radius relationship for the star is such that mass loss leads to an expansion and roche lobe overflow , further accelerating the process and leading to a runaway in which disruption occurs .
all of the features anticipated in the work of lai and collaborators concerning the stability of the binary are apparent , and are seen as well in the numerical simulations of merging binaries performed by rasio and collaborators @xcite .
the main additional result in terms of the dynamics is that configurations that are stable in principle up to roche lobe overflow are de stabilized by the mass transfer process itself , and fully merge as well within a few orbital periods .
it is also quite clear that in the presence of gravitational radiation reaction ( even in its crudest approximation ) the system enters a dynamical infall at small separations that no amount of ( even conservative ) mass transfer can revert .
qualitatively , the result is largely independent of the mass ratio , the assumed initial condition in terms of neutron star spin ( both tidally locked and irrotational configurations were explored in @xcite ) and the assumed compressibility .
quantitatively , the details are different , and are primarily reflected in the final disk mass and the gravitational wave signal ( see , e.g. , @xcite ) .
after a few initial orbital periods , a black hole of @xmath197 , surrounded by a thick and hot debris disk ( @xmath198 ) approximately @xmath199 cm across is all that remains of the initial couple ( figure [ assembly ] ) .
} = $ ] 8 ; 12 ) . in the first panel ,
the tidally deformed star is still visible to the left of the black hole . by 4
ms , the disk is growing around the black hole , and a slice of the tidal tail ( see figure [ bhnsmergers ] ) is visible to the left of the disk . by 17 ms the disk is practically fully formed , and is draining onto the black hole on a much longer time scale , determined by angular momentum transport effects .
this particular calculation used @xmath176 , and the box shown covers @xmath200 km @xcite.,width=480 ] for double nss the situation is essentially the same in terms of the orbital dynamics prior to merger .
synchronization during spiral - in is impossible because the viscosity in neutron star matter is too small @xcite , and so they merge with spin frequencies close to zero , compared with the orbital frequency ( such configurations are termed _ irrotational _ ) .
contact occurs at subsonic velocities and a shear layer then develops at the interface . modeling
this numerically is extremely difficult , because the layer is unstable at all wavelengths and vortices develop all along it @xcite , possibly amplifying the magnetic field to extremely large values very quickly @xcite .
the mass ratio in double ns binaries is quite close to unity , so a fairly symmetrical remnant would be a natural and expected outcome . however , in general even small departures from unity can have important consequences in the inner structure of the final object : the lighter star is disrupted and spread over the surface of its massive companion , which can remain largely undisturbed @xcite .
the final configuration now consists of a supra massive neutron star ( i.e. , one with more mass than a cold , non rotating configuration could support ) surrounded again by a thick shock heated envelope and hot torus similar to that formed in bh - ns mergers .
the center of the remnant is rapidly , and in many cases , differentially rotating , which can have a profound impact on its subsequent evolution ( see below , [ sec : corefate ] ) . by definition , in either case
the merger entails the formation of an object rotating at the break up limit .
material dynamically stripped from a star is thus violently ejected by tidal torques through the outer lagrange point , removing energy and angular momentum and forming a large tail . for double neutron star binaries
there are two such structures , and obviously only one in a bh - ns pair . the mass ( typically up to a few tenths of a solar mass ) , energy and composition of these structures can be more easily studied by lagrangian particle based methods , as they are several thousand kilometers across by the time the merger is over , and have thus moved off the numerical domain in grid based codes . by the inherent asymmetry in bh - ns binaries ( in form and mass ratio ) ,
the black hole can receive a substantial kick ( up to 10@xmath201 km s@xmath2 @xcite ) during the process , in addition to that arising from gravitational wave emission .
some of the fluid ( as much as a few hundredths of a solar mass ) in these flows is often gravitationally unbound , and could , as originally envisaged by lattimer & schramm , undergo r
process nucleosynthesis @xcite .
the rest will eventually return to the vicinity of the compact object , with possible interesting consequences for sgrb late time emission ( see [ sec : prolonged ] ) .
it is important to note that the debris disks are formed in only a few dynamical time scales ( a few milliseconds ) , which is practically instantaneous considering their later , more leisurely evolution ( which will be detailed below ) .
the possibility that a fraction of the neutron star in a bh - ns binary would survive the initial mass transfer episode and produce cycles of accretion was explored at long time scales by davies et al .
this was motivated by the fact that numerical simulations employing relatively stiff equations of state at high binary mass ratios @xcite showed that this might actually occur . however , this is most likely not the case , because : ( i ) the required equation of state was unrealistically stiff ; and ( ii ) gravity was essentially computed in a newtonian formalism .
more recent calculations with pseudo
newtonian potentials @xcite and the use of general relativity @xcite consistently fail to reproduce this behavior for a range of compressibilities in the equation of state . instead , the neutron star is promptly and fully disrupted soon after the onset of mass transfer . performing calculations with full general relativity
is clearly necessary , but is not an easy task .
more importantly , it is not even clear how to estimate its effects with simple analytical considerations .
computations of the location of the _ innermost stable circular orbit _ ( isco ) in black hole systems indicate that tidal disruption may be avoided completely , with the star plunging directly beyond the horizon , essentially being accreted whole in a matter of a millisecond @xcite .
this would preclude the formation of a grb lasting 10 to 100 times longer .
but we stress that , just as for stability considerations in the newtonian case , even if the location of the isco can be a useful guide in some cases , it can not accurately describe the dynamical behavior of the system once mass transfer begins . in pseudo newtonian numerical simulations @xcite and post - newtonian orbital evolution estimates @xcite , even when a near radial plunge is observed , the star is frequently distorted enough by tidal forces that long tidal tails and disk like structures can form .
the outcome is particularly sensitive to the mass ratio @xmath202 and initial general relativity calculations showed that the spin of the black hole is also very important , with rotating bhs favoring the creation of disks @xcite .
dynamical calculations of bh - ns systems in pseudo newtonian potentials that mimic general relativity effects typically show that for mass ratios @xmath203 it is possible to form a disk , although of lower mass than previously thought , @xmath204@xcite .
a long one armed tidal tail is formed as in the newtonian calculations , with some material being dynamically ejected from the system . recently , faber et al .
@xcite have presented their first results for the dynamical merger phase of a bh - ns binary in general relativity for low mass ratios ( @xmath205 ) .
while the the equation of state is simple ( a polytropic relation with @xmath206 was assumed ) , the final outcome is clearly dependent on the complex dynamics of mass transfer .
even though at one point in the evolution most of the stellar material lies within the analytically computed isco , tidal torques transfer enough angular momentum to a large fraction of the fluid , producing an accretion disk with @xmath207 by the end of the simulation at @xmath208 ms .
the black hole horizon is a point of no return , but the innermost stable circular orbit is most definitely _ not_. extending our own work , and further motivated by the first accurate localizations of sgrb afterglows ( starting with grb050509b , @xcite ) , we computed the binary evolution of bh - ns systems in a pseudo newtonian potential in an attempt to better estimate the circumstances under which a disk may form , and its specific configuration if it does @xcite .
the most important parameters are the disk mass and size , as these set the global energy , density , and time scales for its evolution .
the calculations employed the same formalism and code as before @xcite , substituting the pseudo - newtonian potential of paczyski & wiita @xcite , and using irrotational binaries as initial conditions ( i.e. , ones in which the neutron star spin is negligible ) .
we explored variations in both the mass ratio ( @xmath209 ) and neutron star compressibility ( @xmath210 ) , and found that full tidal disruption takes place as the star approaches the black hole , with a torus forming around it .
it contains , however , substantially less mass than in the newtonian case , because a larger fraction is directly accreted by the black hole .
additionally , a large portion of the remaining material is on highly eccentric orbits in the long tidal tails , and , while it will return to the vicinity of the black hole , it does not constitute immediately part of the accretion disk .
overall the torus mass is at least one order of magnitude lower than previously estimated , and somewhat colder , since it has not been shock heated to the same degree by self
interaction of the accretion stream ( see figure [ bhnsmergers ] ) .
rosswog @xcite has performed high resolution calculations of this type as well , but with the use of a realistic equation of state , and finds similar evolution and outcomes . in the case of merging neutron star pairs , general relativistic calculations
@xcite have progressed even further , going beyond the adoption of a simple polytropic pressure density relation @xcite .
the outcome is predictably complicated , and depends sensitively on the total mass of the system and the initial mass ratio .
asymmetric binaries tend to produce more massive disks than those with identical components .
as mentioned earlier for the case of newtonian calculations , a small deviation from a mass ratio of unity is significant , since it can lead to the nearly complete tidal disruption of the lighter component , while the more massive star is largely unaffected .
shibata & taniguchi @xcite find that the lower the mass ratio , the more massive the resulting accretion disk , reaching @xmath211 for @xmath212 .
both psr1913 + 16 and psrj0737 - 3039 have @xmath213 @xcite , so this mass ratio may be unrealistically low , but when @xmath214 a considerable disk containing @xmath215 still forms .
this can easily account for the @xmath216 erg required for a typical sgrb , simply based on its gravitational binding energy .
we give in table [ diskmass ] a summary of estimates for the disk properties for different progenitors , based on the calculations of various groups described above .
the characteristic masses , maximum densities and temperatures are @xmath217 , @xmath218 g @xmath99 and @xmath219k respectively , and provide a starting point for the more detailed disk evolution calculations we will describe below in [ sec : disks ] .
@lllll prog .
& @xmath220 & gravity & & eq .
of state & ref .
+ & & method & & + bh / ns&0.1 - 0.3&n , sph & polytropes & @xcite + bh / ns&0.03 - 0.04&pw , sph & polytropes & @xcite + bh / ns&0.26 - 0.67&n , grid & ls@xcite&@xcite + bh / ns&0.001-.01&pw , sph & shen@xcite&@xcite + bh / ns&0.001 - 0.01&gr , sph & polytropes&@xcite + ns / ns&0.2 - 0.5&n & sph , polytropes&@xcite + ns / ns&0.4&n , sph & polytropes&@xcite + ns / ns&0.01 - 0.25&n , grid & ls@xcite&@xcite + ns / ns&0.25 - 0.55&n , sph & ls@xcite , shen@xcite&@xcite + ns / ns&0.05 - 0.26&gr , sph & shen@xcite&@xcite + ns / ns&0.0001 - 0.01&gr , grid & apr@xcite&@xcite + in the case of merging neutron stars , the ultimate fate of the central object is still unresolved , and depends on the maximum mass that a hot , differentially rotating configuration can support . it has been known for some time that the maximum allowed mass can be increased by the effects of rotation ( see , e.g. , @xcite ) , and in particular , _ differential _ rotation @xcite .
in addition , the post
merger core is certainly not cold , as shock heating ( at least for a mass ratio of unity ) can raise the internal energy substantially .
shibata & taniguchi @xcite find that this threshold for collapse can be estimated as @xmath221 , where @xmath222 is the corresponding value for a non rotating and spherical , cold configuration .
based on the recent observation of psrj0751 + 1807 @xcite , @xmath223 , so a total mass greater than @xmath224 is required for prompt collapse to a black hole . if @xmath225 , various mechanisms ( e.g. , emission of gravitational waves , redistribution of angular momentum , magnetic fields ) could act to dissipate and/or transport energy and angular momentum , possibly inducing collapse after a delay which could range from seconds to weeks ( this has also been suggested as a two step process which would be responsible for a sn / grb association in the case of long grbs @xcite ) .
differential rotation , besides providing additional support against collapse , could amplify seed magnetic fields in the core to large values @xcite and turn the it into a powerful magnetar . spin down
torques , magnetic field winding and flaring alone could then in principle power a grb , independently of the presence of a torus like accretion structure @xcite .
note that this would also be one ingredient of the expected outcome of accretion induced collapse of a wd , or a double wd merger
. the surrounding environment would be much different , clearly , and its effect on the generation of a relativistic fireball would certainly be important ( and possibly catastrophic @xcite ) .
the magneto rotational instability ( mri ) may also operate in such an environment , transferring angular momentum to the debris disk fast enough to keep it from being swallowed by the core when ( and if ) it collapses @xcite .
an alternative possibility is that the core may deform into a bar - like structure , provided the equation of state is sufficiently stiff @xcite .
gravitational torques in the presence of this asymmetry may then transfer angular momentum to the orbiting debris and increase the mass of the disk . in either of these last two cases
one would then have a system similar to that occurring in a bh - ns merger .
in general , the details remain highly uncertain , but powering a grb is still a possible evolutionary pathway ( see figure 21 in @xcite ) . the collision of unbound compact objects , rather than their merger in binaries , has not received much attention in the context of grbs .
janka & ruffert @xcite computed the interaction of two identical neutron stars on parabolic orbits , and found a post - collision environment so contaminated with baryons that they concluded it was not viable as a grb central engine .
in addition , they estimated the event rate within galaxies , and found it much too low to be of interest .
we have reconsidered this problem @xcite , and find that if interactions in globular clusters ( gc ) are taken into account , the rates can have an important effect on the production of sgrbs , if the outcome of the collision is a favorable one .
exactly how frequent a collision takes place depends on the velocity dispersion and the stellar density in a given gc , as well as the number density of gcs in a given galaxy @xcite . a similar scenario , the tidal disruption of stars by massive black holes ( where the mass ratio is typically @xmath226 ) was considered as a mechanism for feeding agn , thus accounting for their luminosity @xcite .
simple initial estimates and more detailed analytical calculations @xcite were subsequently confirmed in their fundamental aspects through numerical simulations by various groups @xcite . the relevant quantity fixing the strength of the interaction is @xmath227 where @xmath228 and @xmath229 are the stellar and black hole mass , @xmath230 is the pericenter distance and @xmath231 is the stellar radius . when @xmath232 the tidal field is sufficiently strong to disrupt the star ( the energy for this is ultimately extracted from the orbital motion ) . for the scenario considered by rees @xcite ,
the star moves essentially in a fixed background metric since @xmath233 ( a condition fully exploited in the numerical calculations later carried out ) , and the typical pericenter distance is up to one hundred times the stellar radius .
thus the star does not directly impact its companion , and the interaction is entirely gravitational .
it was found that approximately half the mass of the star is ejected from the system ( at greater than escape velocity ) and the other half remains bound . using the keplerian relation @xmath234 for the orbiting material , where @xmath235 is the specific energy , and the fact that the differential mass distribution with energy , @xmath236 , is practically constant , the rate at which mass returns to the black hole is given by @xmath237 .
much can be learned from these calculations , and applied to the case of colliding compact objects of comparable mass .
much is different as well , primarily because the mass ratio is of order @xmath238 , rather than 10@xmath239 .
this means that the pericenter distance is now comparable to the stellar radius .
the initial encounter may thus resemble more a direct impact , with immediate exchange of mass , and a corresponding alteration to the mass ratio on a dynamical time scale . a second effect , as for mergers
, is related to the compressibility of the material , and how the star responds in radius once mass transfer begins . and 1.4m@xmath164 ( left ) , a neutron star and a black hole with masses 1.4m@xmath164 and 4.5m@xmath164 ( middle ) , and a white dwarf and a massive neutron star with masses 0.5m@xmath164 and 2.5m@xmath164 ( right ) .
all neutron stars are modeled with adiabatic indices @xmath206 , and the white dwarf with index @xmath176 , appropriate for a low mass configuration .
the scale is indicated in each panel , with logarithmic contours of density shown in the orbital plane ( spaced every quarter decade , with the lowest one at @xmath192}=10 ; 10 ; 0 $ ] for each column respectively ) .
the two tidal tails and the accretion disk are clearly visible in the bh - ns collision .
note the large difference in spatial and temporal scales for the case involving the wd .
the estimated delay for the wd core to return to the vicinity of the bh is 1000 s. all encounters have a parameter @xmath240 , as defined in equation ( [ tidaleta]).,width=441 ] we have now computed the outcome of parabolic encounters between black holes , neutron stars and white dwarfs , in various combinations , to determine the structure of the remnant and investigate its viability as a possible grb central engine .
the obvious difference between systems containing a wd and those that do not is that the time scales are typically longer in the former by a factor @xmath241 .
it is therefore natural to expect accretion and possible flares at hundreds , or even thousands of seconds , rather than tens of seconds after the main interaction .
whether these manifest themselves in @xmath0-rays at all is another matter , and one that can not be easily tackled numerically yet .
the reason is that the wd radius is typically a thousand times larger than that of its companion , and hence the dynamical range that needs to be properly resolved is correspondingly increased .
we find that collisions occur in two stages ( see figure [ collisions ] ) . during the first periastron passage
, an outward tidal stream is stripped from the low mass member and ejected at high velocity .
this act binds the surviving core , which rapidly returns for a second passage and is entirely disrupted , forming a second tidal tail , as well as an accretion disk around the massive companion .
the duration of the evolution up to this point is comparable to the final disruption in a binary ( about 10 ms ) .
the end result is the formation of a system similar to that occurring in a binary merger , in that it contains a massive body ( with the same uncertainties in delay and pathway to collapse to a black hole as before ) surrounded by a debris disk of a few tenths to a few hundredths of a solar mass .
we also considered the case of a white dwarf colliding with a massive neutron star ( or low mass black hole ) .
we were unable at this point to follow the entire evolution of the system , because of the longer time scales involved , but a large tidal tail is also formed at first encounter , with the surviving core being bound as a result . already , however , a disk is apparent , being a thousand times larger than in the ns / bh encounter .
a detailed dynamical study of this type of collision will be presented elsewhere @xcite .
+ despite a variety of possible initial configurations , our best guess for the ultimate prime mover thus consists of a similar astrophysical object : a stellar mass black hole surrounded by a hot debris torus ; moreover the overall energetics of these various progenitors differ by at most a few orders of magnitude , the spread reflecting the differing spin energy in the hole and the different masses left behind in the orbiting debris . on the hypothesis that the central engine involves hyperaccreting black holes
, we would obviously expect the hole mass , the rate at which the gas is supplied to the hole , and the angular momentum of the hole to be essential parameters . in this case
there is no external agent feeding the accretion disk ( excluding tidal tail interactions , which we will consider in detail in [ sec : prolonged ] ) , and thus the event is over roughly on an accretion time scale .
any attempt to explain their properties thus requires that we now consider its evolution and associated energy release .
the way in which gas flows onto an accreting object depends largely on the conditions where it is injected . in this section we address the general dynamical evolution of neutrino cooled accretion flows under the conditions expected to occur in sgrb central engines .
there are five sections :
[ sec : history ] is devoted to general considerations and a review of previous work addressing neutrino cooled flows ; [ sec : thermo ] gives a summary of the microphysical and thermodynamical conditions in the fluid and considers the relevant equation of state and cooling processes in the disk ; detailed calculations in a pseudo - general relativistic potential are presented in [ sec : evolution ] ; the generation of disk driven winds and their accompanying outflows is discussed in [ sec : winds ] ; finally , stability considerations are addressed in
[ sec : stability ] . despite all of the complications inherent in the study of neutrino cooled flows ( and accretion flows in general ) , two main ingredients , easier to qualify , are crucial for understanding the life and death of the accreting source .
the first is simply the total mass available in the system at the outset .
this fixes the global energy scales , both thermal and gravitational , and allows for an approximate determination of the available energy , with @xmath242 note that we have estimated the thermal and gravitational energies to be of the same order .
this is different than for classical thin disks , because the hypercritical accretion flows we are to consider are born in highly dynamical situations where there is an enormous store of internal energy ( this is true for binary mergers as well as for collapsar type scenarios ) .
the fundamental reason is that in either case , internal energy was the primary source of support prior to the catastrophic gravitationally induced collapse or tidal disruption .
the second is the nature of the mechanism feeding the inner disk with mass and energy as a function of time . in the standard picture following core collapse , normally envisaged for long grbs @xcite , the inflowing mass supplies an accretion rate @xmath243 for a time @xmath244 s and
is thereafter fed by injection of infalling matter at a rate that drops off as @xmath245 . for the case of merging compact objects , it is usually assumed that no external agent feeds the previously assembled disk , and thus that the event is over roughly on an accretion time scale .
this picture can be altered by the presence of tidal tails formed with material stripped from the stars during the initial merger phase and ejected on eccentric orbits .
we will address this later .
regardless of the tidal tails , the physical conditions and instantaneous structure within the flow are similar in both scenarios , and conclusions drawn from one case may be used to infer the situation in the other .
the equivalent eddington luminosity for neutrinos fixes a rough upper bound for the power output ( considering coherent scattering off free nucleons as a source of opacity ) of @xmath246 erg s@xmath2 . combined with the available energy in the disk
, one finds a characteristic lifetime at this power level of @xmath247 ms .
two important effects alter this estimate : ( i ) the power is in fact much smaller than @xmath89 , by at least two orders of magnitude , and ( ii ) accretion onto the black hole drains the disk of mass and energy on a viscous time scale , @xmath248 , which can be shorter if the efficiency of angular momentum transport is high .
speculation about the structure and evolution of a massive accretion disk surrounding a new - born black hole began as soon as they were suggested to be the prime movers of grb sources @xcite . despite the large theoretical uncertainties
, it was clear from the start that indeed dense , rapidly evolving structures with huge accretion rates ( on the order of one solar mass per second ) were to be expected , with natural time scales ( dynamical and viscous ) that could in principle account for grbs .
whether or not the resulting signal turns into a burst or not is still a matter of debate , but as stan woosley put it @xcite `` if the signature of a @xmath249 black hole accreting stellar masses of material in a minute is not a gamma ray burst , what is it ? '' .
the physical regime at the expected densities and temperatures places these systems alongside core
collapse supernovae in the bestiary of astrophysics , and possibly makes them even more energetic .
the detailed structure of neutrino cooled accretion flows ( ndafs ) has been investigated by a number of groups over the past few years . initially only steady - state , azimuthally symmetric and vertically integrated solutions were considered , with increasing level of detail mainly in what concerns the thermodynamics and the equation of state @xcite , but also in the effects of general relativity @xcite .
these studies indicated that the instantaneous power output for plausible accretion rates , if maintained , could account for the overall energetics of grbs .
neutrino cooling , when initiated at the proper temperature , would be an efficient mechanism for the removal of internal energy , and the disk would thin rather abruptly .
the 1d solutions of popham et al .
@xcite were matched in the inner regions quite nicely with the initial numerical results for collapsar simulations @xcite .
narayan et al .
@xcite considered among other things , how the radius of matter injection would affect the flow , and concluded that it needed to be rather small in order for neutrino cooling to be efficient .
otherwise , a large fraction of the accreting mass at large radii would , being unable to get rid of its internal energy , simply be blown away in a large scale outflow .
kohri & mineshige @xcite then considered microphysical effects in greater detail , both in terms of the equation of state and the cooling processes .
finally , the effects of general relativity on the disk structure have been studied by chen & beloborodov @xcite for stationary flows around schwarzschild black holes .
more recently , time
dependent calculations in one @xcite , two @xcite assuming azimuthal symmetry and three dimensions @xcite have tackled the question of time dependence and instabilities .
the trade off between relaxing the assumption of @xmath250symmetry or not is reflected in the time interval than can be reliably modeled at high resolution : @xmath251 ms in 3d vs. @xmath252 s in 2d .
the former allows for a full exploration of azimuthal modes and instabilities , while the latter may provide reliable physical estimates on time scales comparable to the duration of sgrbs .
both 2d and 3d simulations have relied on initial conditions taken from 3d binary merger calculations @xcite , and as such are a natural extension of these models .
it has become clear from these studies that the neutrino energy release is rather rapid .
as long as the accretion ( or viscous ) time scale @xmath253 is longer than the cooling time @xmath254 , the fluid will be able to radiate most of its internal energy .
the power is thus maintained at a fairly constant level for up to 100 ms , then drops rapidly .
however , the density in the disk remains fairly high for essentially @xmath255 , which can be as long as a few seconds .
it is thus in principle able to anchor strong magnetic fields capable of driving mhd flows .
another important result is that for high enough accretion rates ( or equivalently , densities ) , the inner most regions of the disk can become opaque to neutrinos .
this has important consequences for the cooling time scale , since the internal energy can not escape immediately , but must diffuse out .
it is also important for the composition , since a negative radial lepton gradient , akin to that occurring above proto neutron stars following core
collapse @xcite , can be established . despite the stabilizing influence of differential rotation @xcite
, we have found that the disk can become convectively unstable , which can have important consequences for the generation of magnetic fields and nucleosynthetic products in a possible outflow .
general relativity , as already pointed out , not only plays a crucial role in the merger dynamics of compact binaries , but may also be important in determining the evolution and energy output from the accretion disks thus formed . the disk is small enough that a substantial amount of material lies in the vicinity of the marginally stable orbit , where the potential departs significantly from its newtonian form .
the black hole itself may be rapidly rotating , depending on the previous binary history , and thus produce inertial frame dragging . finally , the emitted neutrinos are subject to light bending effects , which can alter the efficiency for annihilation , and thus the corresponding spectrum
. dynamical simulations of post merger disk evolution have been performed without @xcite and with @xcite account of these differences , through the use of a pseudo potential of the form proposed by paczyski & wiita @xcite and artemova et al .
@xcite for rotating black holes .
these effects have been taken into account in the computation of the energy deposition rates in the vicinity of the disk @xcite , taking realistic configurations derived from compact mergers as input conditions in recent computations @xcite .
these disks are typically compact , with the bulk of the mass residing within @xmath199 cm of the black hole ( which contains @xmath256 ) .
they are dense ( @xmath257 ) and hot ( @xmath258 ) .
the temperature is in fact high enough that the nuclei are practically fully photodisintegrated in the inner regions .
neutronization due to the high densities can cause the electron fraction to drop substantially below @xmath259 , and capture of @xmath260 pairs by free neutrons and protons provides the bulk of the cooling for @xmath261 g @xmath99 .
the gas is composed essentially of @xmath45 particles and free nucleons in nuclear statistical equilibrium ( nse ) , @xmath260 pairs ( the relative importance of positrons is quite sensitive to the degeneracy parameter @xmath262 , where @xmath263 is the electron chemical potential ) , trapped photons and neutrinos of all species ( @xmath260 captures and annihilation produce only electron neutrinos , but nucleon nucleon bremsstrahlung and plasmon decays can produce @xmath264 and @xmath265 neutrinos as well ) . studying the thermal evolution and
the corresponding energy release thus requires a detailed equation of state and consideration of weak interaction rates and emission processes .
in addition , the background gravitational field is intense enough that relativistic effects can come into play , at least in the inner regions of the flow .
finally , the whole situation can hardly be considered to be in a steady state , since it originates , in most cases , either from the collapse of the fe - core in a massive star , or from the merger or collision of two compact objects .
we have thus considered an equation of state in which the total pressure is given by @xmath266 where @xmath267 is the pressure associated with neutrinos , @xmath268 , @xmath269 and @xmath270 is the mass fraction of photodisintegrated ( and ideal ) nuclei . as an approximate ( but quite accurate ) solution to the equations of nuclear statistical equilibrium between free nucleons and helium ( we do not consider the creation of iron like nuclei ) we use @xmath271 the degeneracy parameter of fermions can vary over a large range in the disk , so it is necessary to use an expression ( due to @xcite ) that allows for it ( but under the condition of relativity , which translates to @xmath272 g @xmath99 ) , namely @xmath273,\ ] ] and @xmath274\ ] ] for the electron fraction .
this formula reduces to the appropriate limits when the temperature is low ( @xmath275 , implying @xmath276 for a cold fermi gas ) and when it is high ( @xmath277 , giving @xmath278 for relativistic @xmath260 pairs ) .
note that this automatically takes into account the presence of pairs in the limit of an ultra relativistic gas , making it unnecessary to alter the factor @xmath279 appearing in the term for radiation pressure .
there is an additional effect which alters the equilibrium composition of the fluid , related to the optical depth .
we approximate this by following beloborodov @xcite and computing the electron fraction as @xmath280 in the optically thin and mildly degenerate case ( @xmath281c^{2}$ ] ) , and @xmath282 in the optically thick case ( this follows from setting the neutrino chemical potential equal to zero in the reaction @xmath283 and using maxwell boltzmann statistics for the non degenerate nucleons ) .
we use an interpolated fit weighted with factors involving @xmath284 to smoothly transition from one regime to the other .
we have included neutrino energy losses from @xmath260 pair captures onto free nucleons by the use of tables @xcite and @xmath260 pair annihilation from fitting functions @xcite .
additionally plasmon decays @xcite and nucleon - nucleon bremsstrahlung @xcite losses are considered , although their contribution to the total luminosity is negligible .
the resulting cooling rates are accurate over a wide range of temperature and density , and are vary consistently with the changing composition and conditions within the flow ( particularly for the two most important contributions of @xmath260 capture and annihilation ) .
we approximate the optical depth of the material to neutrinos by computing the cross - section for scattering off free nucleons and @xmath45 particles heavy nucleons .
these provide the dominant contribution , and are given by @xmath285 and @xmath286^{2},\ ] ] where @xmath287 @xmath1 , @xmath288 is the weinberg angle and @xmath289 is the energy of the neutrinos ( roughly equal to the fermi energy @xmath290 of the mildly degenerate electrons ) .
the optical depth is estimated as @xmath291 , where @xmath131 is the local disk scale height and @xmath292 is the neutrino mean free path .
we find that in the inner regions , @xmath293 in high mass disks .
the cooling is then suppressed by a factor @xmath284 to mimic the effects of diffusion , so the total neutrino luminosity is @xmath294 the transition from a transparent to an opaque fluid occurs at @xmath295 , which corresponds roughly to @xmath296 g @xmath99 ( this happens largely in the high mass disks , with their low mass counterparts remaining mostly transparent ) .
photodisintegration cools the gas at a rate @xmath297 erg s@xmath2 @xmath99 , which is included in the energy equation .
finally , angular momentum is transported through the disk by shear stresses , and we use the @xmath45 prescription to vary the magnitude of the effective viscosity ( all terms in the stress tensor are included in the momentum and energy equations @xcite ) . clearly , although we have attempted to give an accurate thermodynamic description of the gas , several approximations remain . in the first place
, we have assumed that all reactions considered reach equilibrium .
this is quite accurate for photodisintegration , but not necessarily true for weak interactions .
second , our prescription for neutrino optical depth effects is obviously a far cry from the proper boltzmann transport used in one dimensional sn calculations ( or even diffusion ) , but we believe it qualitatively captures the nature of the transparent opaque transition , at approximately the correct density . finally , we have _ not _ considered energy deposition from neutrinos back to the gas through absorption or incoherent scattering
. this could be quite important in the outer regions of the disk ( see [ sec : pathways ] and [ sec : winds ] ) and relevant for the baryon loading of a possible grb producing outflow .
( g @xmath99 ) and cooling rate through neutrinos @xmath298 ( erg g@xmath2 s@xmath2 ) are plotted 100 and 200 ms after the start of the calculation , for disks with an initial mass m@xmath299 ( left panels ) and 0.01m@xmath299 ( right panels ) , where @xmath300 .
the @xmath45 viscosity coefficient is 10@xmath301 .
note the different bounds for the color scale between the two cases , in density as well as cooling.,width=576 ] here we present results for our own 2d calculations in azimuthal symmetry for the evolution of hypercritical accretion flows around a black hole following compact object mergers .
the initial conditions are essentially the same as those in @xcite , which were taken from 3d simulations of black hole neutron star mergers , and are evolved in the pseudo newtonian potential of pacyski & wiita @xcite , which reproduces the existence of a marginally stable orbit at @xmath302 for non spinning black holes .
the standard disk mass initially is @xmath303 .
this is probably close to the highest mass that will be found in such disks , and we have also investigated conditions in lighter disks , with @xmath304 .
the microphysics included is the same as in our previous calculations ( and described above in [ sec : thermo ] ) thus allowing for a clear identification of the effects of strong field gravity .
@xmath305 plane at 50 and 100 ms for disks with initial mass m@xmath299 ( left panels ) and 0.1m@xmath299 ( right panels ) .
the curved solid line marks the photodisintegration threshold between @xmath45 particles and free nucleons ( by mass ) , and the straight solid line marks the electron degeneracy threshold , given by @xmath306 mev . as the disk drains into the black hole , it becomes more transparent , cooler and less degenerate .
the @xmath45 viscosity coefficient is 10@xmath301.,width=480 ] the spatial structure of the disk is shown in figure [ rhodudt ] , where the cooling is also indicated .
the overall structure is similar to that seen in the newtonian simulations , but the maximum density is reduced , and only in the high mass disks and at early times is the presence of an optically thick region ( along the equator , and close to the black hole ) evident .
the thermodynamic structure is plotted in figure [ rhotau ] , where the regions in the density temperature plane occupied by the disk are shown , along with the optical depth .
the flow remains almost entirely photo dissociated in the inner regions , thus giving rise to the intense neutrino emission from pair capture onto free nucleons .
it is also clear that the electron gas is neither ideal nor fully degenerate , and finite degeneracy effects must be considered to compute its evolution properly . in our newtonian calculations we found that neutronization was important in the inner regions of the flow
this is still the case for the relativistic calculations , although the effect is reduced , particularly for low disk masses .
figure [ ye ] shows one snapshot of height integrated radial profiles of the electron fraction in the disk for different disk masses .
the most massive disk achieves higher densities and neutronization is substantially enhanced .
the negative gradient in @xmath307 at small radii was also observed in newtonian calculations , where it was a consequence of the transition to the optically thick regime . in the relativistic case
it occurs even for low
mass disks , and is related to the appearance of a plunging region , where the radial velocity rapidly increases , once the fluid reaches the marginally stable orbit ( see the bottom panel in figure [ ye ] ) .
the net effect is also to produce an inversion of the radial lepton gradient , indicating that convection is possible ( differential rotation has a stabilizing effect on this , and the full solberg hoiland criterion must be considered , see below ) . for disks with varying mass in a pseudo newtonian relativistic potential .
bottom : meridional velocity magnitude , @xmath308 , for the same calculation .
the marginally stable orbit is located at @xmath309})\approx 6.6 $ ] . in the outer disk ,
the gas consists mostly of @xmath45 particles , and @xmath259 .
as the fluid nears the black hole it rises slightly , then becomes smaller as neutronization takes place .
the trend is then reversed and @xmath310 increases again .
the minimum is a sensitive function of the disk mass ( or equivalently , the density).,width=288 ] the accretion rate and neutrino luminosity are plotted as functions of time in figure [ lnu ] for @xmath311 and disk masses covering two orders of magnitude .
for comparison , the result for high disk mass in the newtonian case is also shown .
clearly the interval during which a large luminosity is maintained is lower in the relativistic case , simply because the disk drains more rapidly onto the black hole .
the accretion rates scale roughly linearly with the disk mass , whereas the power output in neutrinos shows a slightly steeper dependence ( it also depends sensitively on the temperature ) , particularly evident at late times .
( @xmath312 ) as a function of time for calculations with three different initial disk masses , covering two orders of magnitude ( @xmath300 ) in the pseudo newtonian potential of pacyski & wiita . for reference ,
the curve labeled [ n ] is the result of a calculation evolved in a newtonian @xmath313 potential .
the viscosity parameter is set to @xmath311 in all cases.,width=288 ] one possibility for the driving of a relativistic outflow and the powering of a grb is @xmath314 annihilation , and another is the often quoted magnetic outflow @xcite .
if we assume a 1% efficiency for the first case at @xmath315 erg s@xmath2 ( which scales with the square of the neutrino luminosity ) , and using the present set of calculations , we infer a scaling with disk mass for the total energy release as @xmath316 this has the same dependence as in the newtonian case , but with a reduced intensity , by about a factor of five . for the magnetic scenario , the total energy release ( assuming the magnetic field energy density is in equipartition with the internal energy density @xmath317 ) scales as @xmath318 again in the same way as in the newtonian case , but also reduced ( this time by about a factor twenty ) . and duration ( @xmath20 ) for @xmath314 annihilation and blandford znajek powered events , respectively , as computed from our 2d disk evolution calculations in a pseudo newtonian potential @xcite .
the range in initial disk masses covers two orders of magnitude , and the effective viscosities are ( from left to right ) , @xmath319 .
the stars are estimates from @xmath314driven outflows in double neutron star mergers @xcite.,width=384 ] a comparison of these scalings and the actual data for the prompt emission of four sgrbs detected in 2005 is shown in figure [ lum ]
. clearly neutrinos are probably not the best way to produce a burst lasting more than a few tens of milliseconds , unless the disk mass is quite high .
the energy release is somewhat insensitive to the burst duration in this case . on the other hand
, magnetic energy extraction seems like a plausible mechanism , and would , at a given disk mass ( if this is a standard quantity in any sense ) , reproduce the trend that longer bursts seem to have a greater fluence .
it is also apparent that a reduced disk mass is less of a problem for magnetic mechanisms than for neutrinos as regards the energetics .
the outer regions and surface of the accretion flow are subject to various effects that can unbind substantial quantities of mass and drive powerful winds , with important implications for nucleosynthesis @xcite , grbs and supernovae @xcite and the possible collimation of relativistic outflows @xcite .
one of these is energy deposition by neutrino heating , another is thermonuclear burning and the corresponding energy release .
the former has been studied extensively by a number of authors , in the supernova as well as grb contexts .
dynamical disk evolution calculations do not always include this explicitly ( our own simulations fall in this category ) , but reasonable estimates can nevertheless be derived in some cases .
we include the latter in its crudest approximation , considering only the transition from @xmath45 particles to free nucleons , whose mass fraction is given in equation ( [ xnuc ] )
. of free nucleons drops appreciably ( a threshold of 0.9 has been set in the plot ) , indicating the creation of @xmath45 particles.,width=384 ] now the gravitational binding energy per nucleon is @xmath320 , and the production of one @xmath45 particle releases 7.7 mev / nucleon into the flow , so enough energy is available to produce a large scale wind where the binding energy becomes sufficiently small .
this is in fact seen in our calculations ( figure [ xwind ] ) , where the outward ( mostly vertical ) velocity is @xmath321 and @xmath322 s@xmath2 .
the corresponding power is @xmath323 erg s@xmath2 for a disk with @xmath324 initially .
the mass outflow from a neutrino driven wind in spherical symmetry was estimated by qian & woosley @xcite as @xmath325 so at the disk neutrino luminosity of @xmath326 erg s@xmath2 , it is interesting to note that the power of each mechanism is comparable ( table [ winds ] ) .
@lcccccc @xmath45 & @xmath327 & @xmath328 & @xmath329 & @xmath330 & @xmath331 & potential@xmath332 + & 10@xmath301m@xmath164 s@xmath2 & g @xmath99 & & 10@xmath178erg s@xmath2 & m@xmath164 & + 0.1 & 2.5 & @xmath333 & 0.08 & 1.5 & 0.3 & n + 0.01 & 2.5 & @xmath334 & 0.1 & 2.25 & 0.3 &
n + 0.01 & 1.5 & @xmath334 & 0.1 & 1.3 & 0.3 & pw@xmath332 + 0.001 & 1 & @xmath335 & 0.1 & 0.9 & 0.3 & n + 0.1 & 0.015 & @xmath336 & 0.08 & 0.009 & 0.003 & n + 0.01 & 0.04 & @xmath337 & 0.05 & 0.0094 & 0.003 & n + 0.01 & 0.025 & @xmath337 & 0.03 & 0.009 & 0.003 & pw + 0.001 & 0.02 & @xmath338 & 0.03 & 0.002 & 0.003 & n + @xmath332 [ n ] : newtonian ; [ pw ] : pacynski & wiita @xcite an important question which can not be addressed fully through steady - state calculations is that of stability .
several factors need to be considered , we address each in turn .
the `` runaway radial '' instability discovered by abramowicz et al .
@xcite applies to configurations with constant specific angular momentum as a function of radius .
this is fully an effect of general relativity , dependent on the existence of an inner roche lobe for disk accretion , akin to the @xmath339 lagrange point in binary stellar evolution .
essentially , if the fluid overflows this lobe , two things occur : ( i ) the black hole mass increases , pushing this critical point outward and ( ii ) material of the same angular momentum value is now in a position to be accreted .
the effect is a runaway , in which the disk is accreted in a matter of a few dynamical time scales .
it turn out that even a small positive gradient of the specific angular momentum , @xmath340 , is enough to stabilize this condition , because matter with greater angular momentum will not be easily accreted , and this damps the instability quite rapidly .
however , none of the post merger accretion disks obtained through merger calculations have angular momentum distributions that are even remotely close to being constant , and so this turns out to be most likely irrelevant for the evolution of grb central engines .
likewise , the self gravity of the disk is probably unimportant , particularly if one considers the results for disk masses from general relativistic calculations to be more realistic .
this has been raised as a possibility concerning the late time x ray flares in a few of the sgrbs discovered recently , and will be addressed in the following section . as for the standard photon
cooled thin disk solution , neutrino cooled flows were shown to be thermally unstable if radiation pressure dominates @xcite .
our numerical solutions which fully resolve the vertical as well as radial structure of the disk in time
dependent fashion show this to be the case as well .
the difference is that the disks response to this condition is a substantial thickening in the opaque regions , because of ( i ) the rise in pressure and ( ii ) the suppression of cooling . rather than having a thin disk , the scale height @xmath341 .
: the dark ( light ) areas represent regions where the solberg hoiland criterion implies instability ( stability ) .
the largest part of the unstable region clearly coincides with the optically thick region of the flow . as the disk drains into the black hole ,
the disk becomes more transparent and this region becomes smaller .
the filament like regions of instability close to the equator are due to the irregularities in the flow .
these particular snapshots are taken from run a2 m in @xcite , with a shakura sunyaev viscosity coefficient @xmath342 and 0.3 m@xmath164 in the disk initially.,width=336 ] we have also recently found that a convective instability is present in neutrino cooled accretion flows at high densities ( or equivalently , accretion rates ) .
this is analogous to what occurs above proto neutron stars following core collapse @xcite , and is directly related to the fact that at @xmath343 g @xmath99 the fluid becomes opaque to neutrinos ( i.e. , the equivalent optical depth , @xmath344 , is of order unity ) . a further rise in density is thus accompanied by an increase in entropy per baryon , @xmath345 , and in equilibrium electron fraction @xmath310 .
the classical solberg hoiland requirement for convective stability can be written as two simultaneous conditions ( e.g. , @xcite ) : @xmath346 and @xmath347 > 0,\ ] ] where @xmath145 is the pressure , @xmath305 is the temperature and @xmath348 is the specific entropy .
we find that the first of these is satisfied over most of the disk volume in dynamical calculations , and marginally so in the inner , opaque regions .
recall that established convection will erase the conditions which led to its occurrence in a characteristic turnover time @xmath349 , unless some external condition tries to maintain the instability , in which case ( for efficient convection ) marginal stability will ensue ( see e.g. , @xcite ) .
the second condition is clearly _ not _ satisfied in the optically thick regime ( see figure [ hoiland ] ) and drives vigorous motions continuously , as can be seen from inspection of the corresponding velocity field ( figure 4 in @xcite ) .
although some circulations are also apparent in the optically thin region , these are of much smaller strength , and are essentially driven by the overshooting of fluid elements out of the convective region and into the outer disk , and their subsequent damping .
it is possible that the mass and accretion rate shape many of the engine activity s visible manifestations . if we are to distinguish those properties of the activity which can be regulated by the engine itself and those which may varied independently , we must look for guidance from observations of the emitted radiation . a crucial observational development which must be noted is the discovery of late time x - ray flaring in a large number of bursts , both long and short @xcite
this has proven difficult to interpret in terms of refreshed shocks , because of their rapid rise and decay .
a likely possibility is that it reflects renewed activity ( although see @xcite for an alternative view ) in the central engine itself @xcite . for short bursts , with durations of the order of half a second ,
these flares occur tens , or even hundreds of seconds after the main burst .
there is also independent support that x - ray emission on these time scales is detected when light curves of many bursts are stacked @xcite . if the grb itself is powered by a hypercritical accretion flow around a stellar mass black hole , the dynamical time scale is only milliseconds , while the viscous time scale , which is usually related to the burst duration , is at most a few seconds if the effective viscosity is equivalent to @xmath350 .
it is thus in principle a problem to account for a resurgence of activity ten to one hundred viscous time scales ( or equivalently , up to ten thousand dynamical time scales ) later . over the past year
, a number of specific suggestions have been made concerning the production of flares , both for long and short events - in some case , suggesting also that their occurrence in both types of events should indicate a common origin : interaction with a binary companion @xcite ; fragmentation of a rapidly rotating core through non - axisymmetric instabilities @xcite ; magnetic regulation of the accretion flow @xcite ; fragmentation of the accretion disk through gravitational instabilities and subsequent accretion @xcite ; differential rotation in post
merger millisecond pulsars @xcite and magnetar like activity .
despite all these suggestions , it remains unclear how exactly the regulation ( in the case of magnetic mechanisms ) or the fragmentation ( in the case of self gravity ) would come about .
magnetic regulation was actually addressed by van putten & ostriker @xcite as a possible way to produce both long and short bursts from the same central engine , depending only on the spin of the black hole .
slow spinning holes would be unable to halt accretion and thus a short burst would ensue . in rapidly spinning holes , the transfer of angular momentum through electromagnetic torques would stop or delay accretion , giving a long event .
there are tantalizing clues in numerical simulations indicating that this might actually occur .
krolik et al .
@xcite find a large difference in the mass accretion rate onto the black hole in mhd simulations depending on the black hole spin .
they report a calculation with a dimensionless kerr parameter of @xmath351 , in which the accretion rate is strongly suppressed due to electromagnetic stresses , and no stationary state is achieved through the end of the computation ( although they note that the magnetic field geometry is fundamentally different from that suggested by van putten & ostriker ) .
this effect is absent for lower values of the kerr parameter .
it is not clear how to extrapolate these results to the time scales involved in sgrb flares , however , since they occur at much later times ( the calculations by krolik et al .
span about half a second for a black hole of four solar masses ) .
one of the basic unknowns of accretion disk theory is the physical mechanism ultimately responsible for angular momentum transport and energy dissipation in the disk .
it is well known that classical hydrodynamical viscosity can not drive accretion at the rates inferred from observations in almost every astrophysical context where accretion disks are thought to play a crucial role . the usual way to overcome
this difficulty is to assume that transport is dominated by some _ anomalous _ viscous phenomenon , possibly related to collective instabilities in the disk , and to give some adhoc parameterizations for its magnitude .
it has been recently recognized that accretion disks threaded by a weak magnetic field are subject to magneto hydrodynamic instabilities ( see balbus & hawley @xcite and references therein ) , which can induce turbulence in the disk and thereby transport angular momentum , promoting the accretion process .
a possible alternative source of transport in cold disks is provided by gravitational instabilities @xcite - although the outcome strongly depends on the thermodynamics of the disk . in particular
, the fragmentation of a gravitationally unstable disk requires that the disk be able to cool very efficiently @xcite , with a cooling time @xmath352 where @xmath353 is the local angular velocity in the disk .
gravitational instabilities would then lead to a self - regulating process : if the disk is initially cold , in the sense that @xmath354 , then gravitational instabilities would heat it up on the short dynamical time - scale , bringing it toward stability ; on the other hand , if the disk is hot enough to begin with , radiative cooling will drive it toward an unstable configuration . as a result of these competing mechanisms ,
the _ switch _ associated with the onset of gravitational instabilities will act as a thermostat . for self - regulated disks , a simple relationship holds between the disk aspect ratio @xmath355 and the mass ratio @xmath356 .
the disk is in fact marginally stable when its temperature is small enough that @xmath357 of course , this relationship can only hold in the limit where @xmath358 .
disk evolution calculations of neutrino cooled flows have so far failed to show that the disk is close to instability @xcite , but this eventually may be the case when enough time has elapsed to allow the disk to significantly cool .
it is not clear how to extrapolate these results to the time scales involved in sgrb flares , since current calculations span at most a few seconds .
it remains to be discussed under which conditions the development of a gravitational spiral structure would be indeed able to transport angular momentum in the disk efficiently , hence favoring accretion , and whether this might be able to account for flares .
there is an additional concern with the the use of a viscous formalism in self - gravitating accretion disks : balbus & papaloizou @xcite have shown that in general the energy transport provided by gravitational instabilities contains global terms , associated with wave energy transport , that can not be directly associated with an effective viscosity . their relative importance , and the conditions under which they may play a significant role are as yet poorly understood .
of the neutron star fluid , with the scale indicated in each panel .
the thick black line across the tail for @xmath206 divides material that is bound to the central black hole from that which has enough energy to escape the system .
note the condensations at regular intervals in the tail for the stiff equation of state , produced by gravitational instability.,width=432 ] in this section we explore the possibility that the long tidal tails formed during compact object mergers and/or collisions may provide the prolonged mass inflow necessary for the production of late flares in sgrbs .
the general idea of extended emission from compact object mergers has been advanced before @xcite , relying on the gradual disruption of the neutron star core over a time scale that is much longer than the dynamical one @xcite .
as we have argued above , we do not believe this will actually occur , but find that the possibility of injecting matter into the central engine at late times remains , in a modified form .
s. ( b ) the collision of a neutron star with a black hole ( the initial mass ratio is @xmath195 ) , using the pseudo - potential of pacyski & wiita @xcite , reveals the formation of a single large tail , with condensations due to self gravity appearing at late times ( see figure [ mergertails ] ) . note the different spatial and temporal scales between the two cases.,width=528 ]
we noted already that large scale tidal tails are a common feature formed during mergers and collisions between compact objects .
these are typically a few thousand kilometers in size by the end of the disruption event in the case of neutron stars and black holes , and a thousand times larger for events involving white dwarfs ( see figures [ mergertails ] and [ coltails ] ) . the amount of mass contained in these structures is significant , and typically @xmath359 .
depending on the details of the equation of state , a small fraction of this ( @xmath360 ) is actually unbound , and will escape to the surrounding medium .
the rest will eventually fall back onto the leftover disk , and probably be accreted onto the central mass . once the initial dynamical interaction is over
, the fluid in the tails is practically on ballistic trajectories , moving in the potential dominated by the central mass .
it is thus possible to compute the rate at which this matter will fall back and accrete onto the existing disk or black hole . to compare with the simple case of a supermassive black hole mentioned in [ sec : colls ] ,
we have also computed the differential distribution of mass with energy for this fluid : @xmath236 .
although it appears at first glance to be roughly constant , there are in fact significant deviations from this .
the corresponding accretion rate thus differs from the @xmath245 law derived by rees @xcite , and is somewhat shallower , with @xmath361 .
the fall back accretion rate illustrated in figure [ mdottails ] indicates that the bulk of the matter in the tail returns to the black hole in roughly one second . with energy , computed for black hole neutron star mergers and one parabolic collision ( labeled `` c '' ) .
each curve is marked with the initial mass ratio .
the material with negative energies will fall back onto the central mass after following essentially ballistic trajectories . ,
width=432 ] .
initially roughly constant , it rapidly assumes a power law decay , with index @xmath362 .
this is roughly independent of the type of interaction ( although we note that at late times numerical noise makes an accurate determination of the slope more difficult ) .
the lower limit indicated at 100 seconds marks the accretion rate required to account for the typical energy seen in a late x
ray flare in a sgrb.,width=432 ] now , clearly one second is much too short to account for flares , which typically occur after 30 - 100 s. however , the fluid does not directly fall onto the black hole , since it has considerable angular momentum .
in fact , its circularization radius is comparable to the size of the disk which was formed around the black hole in the first place ( typically 200 - 500 km ) .
this is the point at which the gas will settle if enough of its energy is dissipated , circularizing its orbit .
we thus have a situation in which in a short time scale , @xmath363 s , a mass @xmath364 returns to a radius @xmath365 cm .
if dissipation occurs ( e.g. , through shocks with what is left of the initial disk ) , a new ring of matter may form at @xmath366 .
the evolution of this newly injected mass will depend on the previous history of the original accretion disk , and on how much of it is left after the fall back time @xmath367 .
evidently the component with the most energy will dominate the behavior of the system , and if @xmath368 it is hard to see how the infalling tail could produce a substantial alteration of the overall flow properties and an accompanying observable signal .
if the reverse is true and @xmath369 , however , perhaps the inner accretion disk and the true accretion rate onto the black hole can be modified and a secondary episode of energy release be provoked .
if we naively estimate the viscous time scale for the newly formed ring at @xmath370 as @xmath371 where @xmath372 is the viscosity coefficient , @xmath373 is the gas sound speed and @xmath374 is the keplerian angular frequency , then typical values from the calculations indicate that @xmath375 viscous time scales of @xmath376 seconds are thus in principle plausible for @xmath377 , and are much longer than the injection interval itself .
this is akin to the instantaneous injection of a discrete amount of matter at a given radius , and will produce an accretion episode with a delay proportional to its duration , @xmath378 ( e.g. , @xcite ) , naturally accounting for this observational fact in sgrb flares . an additional constraint comes from consideration of the total energy release in the observed flares .
taking neutrino emission as an example , the accretion efficiency is @xmath379 based on our own calculations @xcite and those reported by other groups @xcite , where @xmath380 is the actual accretion rate that feeds the central black hole , and is reduced from the rate at which the disk itself is fed by a factor @xmath381 , the remainder going into outflows ( see also @xcite ) .
the accretion efficiency is predominantly dependent upon the flow being optically thin to its own emission , and can thus be extrapolated down to the average accretion rate produced by the fall back material , @xmath382 thus the neutrino luminosity would be @xmath383 accounting for a total fluence @xmath384erg s@xmath2 and a total energy release @xmath385 erg over @xmath386 s ( typical numbers for observed events ) would require a @xmath387 annihilation efficiency of 10@xmath388 .
this is close to being unrealistically high at this luminosity level , and would thus argue against neutrinos as an ultimate source for the flaring behavior . on the other hand , magnetic energy extraction
could presumably also operate , since the injection can deposit enough mass in the disk to anchor sufficiently strong magnetic fields .
it is hard to see , however , how either option could account for extremely delayed flares occurring after @xmath66 seconds , but perhaps those can be due to refreshed shocks @xcite .
the hydrodynamic stability of the ejected tail is a question that deserves special comment .
the fluid moves globally and to a good approximation in the potential of the central object .
it is , nevertheless , prone to an instability due to self - gravity , which may affect its structure on small scales .
this is the _ varicose _ or
_ sausage _
instability , and it can be shown ( see ch .
xii in ref .
@xcite ) that for a cylinder of _ incompressible _ fluid , gravitational instability sets in for perturbations with wavelength @xmath389 where @xmath390 is the radius of the cylinder and @xmath391 .
the fastest growing mode , which determines the size of the fragments upon manifestation of this effect has @xmath392 , corresponding to a wavelength @xmath393 , and a growth time @xmath394 binary merger calculations do not employ incompressible fluids , but for high enough adiabatic indices one can see that this behavior does indeed occur .
the left panel in figure [ mergertails ] plots the large scale structure formed during a black hole - neutron star merger in a simulation with @xmath175 .
the knots in the tail are spaced at regular intervals , with wavelength @xmath395 , as predicted by the incompressible analysis .
the estimated time scale for their formation is @xmath396 ms , in good agreement with what is seen in the calculations .
binary neutron star mergers with stiff polytropic pressure density relations performed by rasio et al .
@xcite show exactly the same structures .
once fragmentation occurs , the individual clumps continue to move in the overall gravitational potential , and those that are bound to the central object will return to its vicinity at discrete intervals to inject matter . for lower values of the adiabatic index this behavior is no longer seen .
instead , the fluid expands smoothly over large volumes , and the tails are less well defined . at the densities encountered in these structures , one would not expect a compressibility low enough to allow this instability to operate , and indeed it is not seen in merger simulations using realistic equations of state @xcite
. nevertheless , it is worthy to consider their possible formation and subsequent evolution as a general feature . note that if one substitutes the neutron star for a quark ( strange ) star , such droplets are a generic feature upon tidal disruption @xcite .
in this final section we present a short summary of what we have learned so far about the physics of sgrb sources .
though the field is far from being mature , sufficient progress has been made in identifying some of the essential ingredients .
we also describe the observational and theoretical prospects for the near future .
the progress in our understanding of sgrbs has been a story of consolidation and integration , and there is every indication that this progression will continue . until recently ,
sgrbs were known predominantly as bursts of @xmath0-rays , largely devoid of any observable traces at any other wavelengths . however , a striking development in the last several months has been the measurement and localization of fading x - ray signals from several sgrbs , making possible the optical and radio detection of afterglows , which in turn enabled the identification of host galaxies at cosmological distances .
the presence in old stellar populations e.g. , of an elliptical galaxy for grb 050724 , rules out a source uniquely associated with recent star formation .
in addition , no bright supernova is observed to accompany sgrbs , in distinction from most nearby long - duration grbs .
it is now clear that short and long events are not drawn from the same parent stellar population .
even with a handful of sgrbs detected to date , it has become apparent that they are far from standard , both in their energetics and cosmological niche .
this hints at the underlying possibility that the progenitor itself may be quite different from burst to burst , and not entirely restricted to the most discussed scenario involving the merger of compact binaries such as the hulse - taylor pulsar ( although see @xcite ) .
the most fundamental problem posed by sgrb sources is how to generate over @xmath397 erg in the burst nucleus and channel it into collimated plasma jets .
the cumulative evidence insistently suggests that the more powerful sgrbs must have `` processed '' upwards of @xmath63 through a compact entity - only this hypothesis accounts for the high luminosity and compactness inferred from @xmath0-ray variability .
we believe that accretion onto a compact object , be it a neutron star or a stellar mass black hole , offer the best hope of understanding the `` prime mover '' in all types of sgrb sources although a possible attractive exemption includes a rapidly spinning neutron star with a powerful magnetic field . consequently , most theoretical work has been directed towards describing the possible formation channels for these systems , and evaluate those which are likely to produce a viable central engine . the only way to cool the resulting hypercritical accretion flow ( other than through direct advective transport of the energy into a gravitational sink hole on a dynamical time scale ) is by neutrino emission , circumventing the classical eddington limit for photons and allowing for the conversion of gravitational binding energy into outflowing radiation .
consideration of the associated energy release thus leads us down the path of detailed thermodynamical and microphysical processes unlike those encountered in most areas of astrophysics , except supernovae , where the physical conditions are quite similar in terms of density , entropy and internal energy . as of this writing
, it is fairly clear from the concerted efforts of many groups working on different aspects of these problems that there is in principle no problem in accounting for the global energy budget of a typical sgrb from the class of systems here considered .
the devil is in the details , of course , and the actual modes of energy extraction have yet to be worked out carefully .
it would appear , however , that neutrino emission is more confined to be a competitive energy source in the early stages of the dynamical evolution , while magnetically powered events may be able to offer longer staying power .
still , various alternative ways of triggering the explosions responsible for sgrbs remain : ns - ns , ns - bh , bh - wd or wd - wd binary mergers , recycled magnetars , spun - down supra - massive ns and accretion induced collapse of a ns .
can we decide between the various alternatives ?
the progenitors of sgrbs are essentially masked by afterglow emission , largely featureless synchrotron light , which reveals little more than the basic energetics and micro physical parameters of relativistic shocks . in the absence of a supernova - like feature
, the interaction of burst ejecta with a stellar binary companion @xcite or with its emitted radiation @xcite may be the only observable signature in the foreseeable future shedding light on the identity of the progenitor system .
a definitive understanding will , however , come with the observations of concurrent gravitational radiation or neutrino signals arising from the dense , opaque central engine .
our understanding of sgrbs has come a long way since their discovery almost forty years ago , but these enigmatic sources continue to offer major puzzles and challenges .
sgrbs provide us with an exciting opportunity to study new regimes of physics .
as we have described , our rationalization of the principal physical considerations combines some generally accepted features with some more speculative and controversial ingredients .
when confronted with observations , it seems to accommodate their gross features but fails to provide us with a fully predictive theory .
what is more valuable , though considerably harder to achieve , is to refine models like the ones advocated here to the point of making quantitative predictions , and to assemble , assess and interpret observations so as to constrain and refute these theories .
what we can hope of our present understanding is that it will assist us in this endeavour .
high energy astrophysics is a young field .
it owns much of the remaining unexplored `` discovery space '' in contemporary astronomy .
two examples of this discovery space are at the extremes of observation of the electromagnetic spectrum . at the high end , there are already a few tens of tev sources , while at the low end of @xmath398 mhz radio astronomy there are essentially no sources .
neutrino astronomy claims only two cosmic sources so far , the sun and sn1987a . finally , as many of the most interesting high energy sources are ultimately black holes and neutron stars , the exciting field of gravitational wave astronomy perhaps a decade away from birth
is inextricably linked to high energy astrophysics .
these are the next frontiers and we consider the role of sgrbs within them in turn .
atmospheric cerenkov techniques are being used to detect @xmath0-rays in the gev tev range .
these are important as both sources and as probes .
persistent sources have been identified with pulsars , blazars and supernova remnants and in each case are likely to provide the best approach we have to understanding the fundamental nature of these sources .
a tentative @xmath399 tev detection of a long grb has been reported with the water cherenkov detector milagrito @xcite . here
the big question is to determine whether the sgrb jets comprise ultrarelativistic protons , that interact with either the radiation field or the background plasma , or if they are @xmath400 pairs .
tev sources should be far more plentiful in the latter case . the combination of glast and telescopes like hess and veritas ought to be able to sort this out .
the same shocks which are thought to accelerate the electrons responsible for the non - thermal @xmath0-rays in sgrbs should also accelerate protons @xcite .
both the internal and the external reverse shocks are mildly relativistic , and are expected to lead to relativistic protons @xcite .
the maximum proton energies achievable in sgrb shocks are @xmath401 ev , comparable to the highest energies measured with large cosmic ray ground arrays @xcite . for this
, the acceleration time must be shorter than both the radiation or adiabatic loss time and the escape time from the acceleration region @xcite .
the accelerated protons can interact with the fireball photons , leading to charged pions , muons and neutrinos . for internal shocks producing
observed 1 mev photons this implies @xmath402 ev protons , and neutrinos with @xmath403 of that energy , @xmath404 ev @xcite .
another copious source of target photons in the uv is the afterglow reverse shock , for which the resonance condition requires higher energy protons leading to neutrinos of @xmath405 ev @xcite .
whereas photon - pion interactions lead to higher energy neutrinos and provide a direct probe of the shock proton acceleration as well as of the photon density , inelastic proton - neutron collisions may occur even in the absence of shocks , leading to charged pions and neutrinos @xcite with lower energies than those from photon - pion interactions .
the typical neutrino energies are in the @xmath6 1 - 10 gev range , which could be detectable in coincidence with observed sgrbs .
this is the province of projects like amanda , icecube and antares .
success in the former will suggest that ultrarelativistic outflows comprise mainly protons .
neutrino astronomy has the advantage that we can see the universe up to @xmath6 eev energies .
by contrast , the universe becomes opaque to @xmath0-rays above @xmath6 tev energies through absorption by the infrared background .
finally the last and most challenging frontier is that of gravitational radiation , which is largely unknown territory .
a time - integrated luminosity of the order of a solar rest mass ( @xmath406 erg ) is predicted from merging ns - ns and ns - bh models , while that from collapsar models is less certain , but estimated to be lower .
ground - based facilities , like ligo , tama and virgo , will be seeking such stellar sources .
the observation the associated gravitational waves would be facilitated if the mergers involve observed sgrb sources ; and conversely , it may be possible to strengthen the case for ( or against ) ns - ns or ns - bh progenitors of sgrbs if gravitational waves were detected ( or not ) in coincidence with some bursts .
the technical challenge of achieving the sensitivities necessary to measure waves from assured sources should not be understated .
it may well take more than another decade to reach them . neither , however , should the potential rewards .
gravitational waves offer the possibility of observing in an entirely _ new _ spectrum , not merely another window in the same ( electromagnetic ) spectrum .
furthermore , they will by their very nature tell us about events where large quantitites of mass move in such small regions that they are utterly opaque and forever hidden from direct electromagnetic probing , and are the only way ( except perhaps for neutrinos ) through which we can learn about them
. there have been few regions of the electromagnetic spectrum where observation conformed to previous expectation .
so it would be indeed remarkable if gravitational wave astronomy , or any of the other frontiers , turned out to be as we have described .
indeed , if past experience is any guide , they will surely provide us with new surprises , challenges and potential loose threads through which we can unravel another piece of the fabric of the universe .
although some of the features now observed in grb sources ( especially afterglows ) were anticipated by theoretical discussions , the recent burst of observational discovery has left theory lagging behind .
there are , however , some topics on which we do believe that there will be steady work of direct relevance to interpreting observations .
one of the most important is the development and use of hydrodynamical codes for numerical simulation of sgrb sources with detailed physics input . existing two and three dimensional codes
have already uncovered some gas - dynamical properties of relativistic flows unanticipated by analytical models ( e.g. , @xcite ) , but there are some key questions that they can not yet address . in particular , higher resolution is needed because even a tiny mass fraction of baryons loading down the outflow severely limits the maximum attainable lorentz factor .
we must wait for useful and affordable three dimensional simulations before we can understand the nonlinear development of instabilities .
well - resolved three dimensional simulations are becoming increasingly common and they rarely fail to surprise us . the symmetry - breaking involved in transitioning from two to three dimensions is crucial and can lead to qualitatively new phenomena .
a particularly important aspect of this would be to link in a self - consistent manner the flow within the accretion disk to that in the outflowing gas , allowing for feedback between the two components .
the key to using simulations productively is to isolate questions that can realistically be addressed and where we do not know what the outcome will be , and then to analyse the simulations so that we can learn what is the correct way to think about the problem and to describe it in terms of elementary principles .
simulations in which the input physics is so circumscribed that they merely illustrate existing prejudice are of limited value ! a second subject ready for a more sophisticated treatment is related to the intensity and shape of the intrinsic spectrum of the emitted radiation .
few would dispute the statement that the photons which bring us all our information about the nature of grbs are the result of particle acceleration in relativistic shocks ( e.g. , @xcite ) . since charged particles radiate only when accelerated , one must attempt to deduce from the spectrum _ how _ the particles are being accelerated , _
why _ they are being accelerated , and to identify the macroscopic source driving the microphysical acceleration process .
collisionless shocks are among the main agents for accelerating ions as well as electrons to high energies whenever sufficient time is available ( e.g. , @xcite ) .
particles reflected from the shock and from scattering centres behind it in the turbulent compressed region have a good chance of experiencing multiple scattering and acceleration by first - order fermi acceleration when coming back across the shock into the turbulent upstream region .
second - order or stochastic fermi acceleration in the broadband turbulence downstream of collisionless shocks will also contribute to acceleration .
in addition , ions may be trapped at perpendicular shocks .
the trapping is a consequence of the shock and the lorentz force exerted on the particle by the magnetic and electric fields in the upstream region . with each reflection at the shock
the particles gyrate parallel to the motional electric field , picking up energy and surfing along the shock surface .
all these mechanisms are still under investigation , but there is evidence that shocks play a most important role in the acceleration of cosmic rays and other particles to very high energies .
there is no in situ information available from astrophysical plasmas .
so one is forced to refer to indirect methods and analogies with accessible plasmas , found only in near - earth space . actually ,
most of the ideas about and models of the behaviour of astrophysical plasmas have been borrowed from space physics and have been refitted to astrophysical scales .
however , the large spatial and long temporal scales in astrophysics and astrophysical observations do not allow for the resolution of the collisionless state of the plasmas .
for instance , in the solar wind the collisional mean free path is of the order of a few au . looked at from the outside , the heliosphere , the region which is affected by the solar wind , will thus be considered collision dominated over time scales longer than a typical propagation time from the sun to jupiter . on any smaller and shorter scales
this is wrong , because collisionless processes govern the solar wind here .
similar arguments apply to stellar winds , molecular clouds , pulsar magnetospheres and the hot gas in clusters of galaxies .
one should thus be aware of the mere fact of a lack of small - scale observations .
collisionless processes generate anomalous transport coefficients .
this helps in deriving a more macroscopic description .
however , small - scale genuinely collisionless processes are thereby hidden .
this implies that it will be difficult , if not impossible , to infer anything about the real structure , for instance , of collisionless astrophysical shock waves .
the reader is refer to @xcite , and @xcite for an excellent presentation of the basic kinetic collisionless ( space ) plasma theory .
the most interesting problem remains , however , in the nature of the central engine and the means of extracting power in a useful collimated form . in all observed cases of relativistic jets ,
the central object is compact , either a neutron star or black hole , and is accreting matter and angular momentum .
in addition , in most systems there is direct or indirect evidence that magnetic fields are present detected in the synchrotron radiation in galactic and extragalactic radio sources or inferred in collapsing supernova cores from the association of remnants with radio pulsars .
this combination of magnetic field and rotation may be very relevant to the production of relativistic jets ( e.g. , @xcite ) .
much of what we have summarized in this respect is conjecture and revolves largely around different prejudices as to how three dimensional flows behave in strong gravitational fields .
there are serious issues of theory that need to be settled independently of the guidance we obtain from observations of astrophysical black holes .
+ we are indebted to a considerable number of colleagues for their help and advice .
we acknowledge , in particular , the contributions of m. aloy , e. berger , k. belczyski , l. bildsten , d. burrows , d. fox , c. fryer , j. fynbo , n. gehrels , d. guetta , j. hjorth , p. hut , th .
janka , c. kouveliotou , s. kulkarni , e. nakar , p. mszros , t. piran , d. pooley , m. prakash , f. rasio , m.ruffert , a. soderberg , v. usov , g. van de ven , d. watson , e. waxman , s. woosley and z. zheng .
our views on the topics discussed here have been clarified through discussions with j. bloom , j. granot , w. kluniak , a. macfadyen , d. page , x. prochaska , m. rees , a. socrates and especially s. rosswog .
we are particularly grateful to y. kaneko for providing the time histories and spectra of batse sgrbs .
major portions of this review were written at dark cosmology centre , copenhagen ; institute for advanced study , princeton ; institute of astronomy , cambridge ; instituto de astronoma , mxico .
we thank the directors of these institutions for their generous hospitality .
financial support for this work was provided in part by nsf ( phy0503584 ) , conacyt ( 36632e,45845e ) and papiit ( 110600,119203 ) .
+ nakar e 2006 _ physics reports submitted _ fishman g and meegan c 1995 _ ann .
astrophys . _ * 33 * 415 kouveliotou c et al 1993 _ astrophys .
j. _ * 413 * l101 van paradijs j , kouveliotou c and wijers r a m j 2000 _ ann . rev .
astrophys . _
* 38 * 379 gehrels n et al 2005 _ nature _ * 437 * 851 bloom j s et al 2006 _ astrophys .
j. _ * 638 * 354 hjorth j et al 2005 _ nature _ * 437 * 859 covino s et al 2006 _ astron . and
astrophys . _
* 447 * l5 berger e et al 2005 _ nature _ * 438 * 988 fox d b et al 2005 _ nature _ * 437 * 845 barthelmy s d et al 2005 _ nature _ * 438 * 994 prochaska j x et al 2006 _ astrophys . j. _ * 642 * 989 villasenor j s et al 2005 _ nature _ * 437 * 855 band d et al 1993 _ astrophys
. j. _ * 413 * 281 mallozzi r s et al 1995 _ astrophys . j. _ * 454 * 597 ghirlanda g , ghisellini g and celotti a 2004 _ astron . and astrophys .
_ * 422 * l55 kaneko y et al 2006 _ astrophys .
j. suppl . _
* 166 * 298 bhat p n et al 1992 _ nature _ * 359 * 217 nakar e and piran t 2002 _ mon . not .
soc . _ * 330 * 920 donaghy t q et al 2006 _ astrophys .
j. submitted _ ( astro - ph/0605570 ) bloom j s et al 2001 _ astron .
j. _ * 121 * 2879 soderberg a et al .
2006 _ astrophys .
j. _ * 650 * 261 gorosabel j et al 2006 _ astron . and
astrophys . _ * 450 * 8 bloom j s et al 2006 _ astrophys .
j. _ * 654 * 878 roming p w a et al .
2006 _ astrophys .
j. _ * 651 * 985 castro - tirado a j et al 2005 _ astron . and astrophys . _ * 439 * l15 lee w h , ramirez - ruiz e and granot j 2005 _ astrophys.j . _ * 630 * l165 grupe d et al 2006 _ astrophys .
j. _ * 653 * 462 burrows d n et al .
2006 _ astrophys .
j. _ * 653 * 468 la parola v et al .
2006 _ astron . and astrophys .
_ * 454 * 753 watson d et al 2006 _ astron . and astrophys .
_ * 454 * l123 panaitescu a 2006 _ mon . not .
* 367 * l42 lazzati d , ramirez - ruiz e. and ghisellini g. 2001 _ astron . and astrophys .
_ * 379 * l39 montanari e et al 2005 _ astrophys
. j. _ * 625 * l17 rees m j and mszros p 1998 _ astrophys .
j. _ * 496 * l1 ramirez - ruiz e , merloni a and rees m j 2001 _ mon . not .
soc . _ * 324 * 1147 fan y and xu d. 2006 _ mon . not .
soc . _ * 372 * l19 dai z g et al .
2006 _ science _ * 311 * 1127 bloom j s and prochaska j x 2006 _ aip conference series _ * 836 * 473 berger e et al 2006 _ astrophys .
j. submitted _ ( astro - ph/0608498 ) berger e 2006 _ aip conference series _ * 836 * 33 pedersen k et al 2005 _ astrophys .
j. _ * 634 * l17 hjorth j et al 2005b _ astrophys .
j. _ * 630 * l117 christensen l , hjorth j and gorosabel j 2004 astron . and astrophys .
* 425 * 913 trentham n et al .
2002 _ mon . not .
soc . _ * 334 * 983 tanvir n
r et al 2004 _ mon . not .
soc . _ * 352 * 1073 le floch e et al .
2003 astron . and astrophys .
* 400 * 499 bornancini c g et al 2004 _ astrophys
. j. _ * 614 * 84 hurley k et al 2005 _ nature _ * 434 * 1098 palmer d m et al 2005 _ nature _ * 434 * 1107 fenimore e e et al 1996 _ astrophys .
j. _ * 460 * 964 hurley k et al 1999 _ nature _ * 397 * 41 duncan r c and thompson c 1992 _ astrophys
. j. _ * 392 * l9 nakar e , gal - yam a , piran t and fox d b 2006 _ astrophys . j. _ * 640 * 849 lazzati d , ghirlanda g and ghisellini g 2005 _ mon . not .
* 362 * l8 gaensler b m et al 2005 _ nature _ * 434 * 1104 granot j et al 2006 _ astrophys . j. _ * 638 * 391 galama t et al 1998 _ nature _ * 387 * 479 richmond m w et al 1996 _ astron
. j. _ * 111 * 327 zeh a , klose s and hartmann d h 2004 _ astrophys
. j. _ * 609 * 952 malesani d et al 2004 _ astrophys .
j. _ _ 609 _ l5 hjorth j et al 2003 _ nature _ * 423 *
847 garnavich p m et al 2003 _ astrophys .
j. _ * 582 * 924 fynbo j p u et al 2006 _ nature _ * 444 * 1047 guetta d and piran t 2005 astron . and astrophys .
* 435 * 421 guetta d and piran t 2006 astron . and astrophys .
* 453 * 823 nakar e et al 2006 _ astrophys . j. _ * 650 * 281 gal - yam a et al 2006 _ astrophys .
j. in submitted _ ( astro - ph/0509891 ) zheng z and ramirez - ruiz e 2006 _ astrophys .
j. submitted _ ( astro - ph/0601622 ) cavallo g and rees m j 1978 _ mon . not .
soc . _ * 183 * 359 paczynski b 1986 _ astrophys .
j. _ * 308 * l43 goodman j 1986 _ astrophys . j. _ * 308 * l47 shemi a and piran t 1990 _ astrophys .
j. _ * 365 * l55 rees m j and mszros p 1994 _ astrophys .
j. _ * 430 * l93 sari r and piran t 1997 _ astrophys . j. _ * 485 * 270 ramirez - ruiz e and lloyd - ronning n m 2002 _ new .
ast . _ * 7 * 197 kobayashi s , piran t and sari r 1997 _ astrophys
. j. _ * 490 * 92 granot j and sari r 2002 _ astrophys . j. _ * 568 * 820 sari r , piran t and narayan r 1998 _ astrophys .
j. _ * 497 * l17 ramirez - ruiz e 2005 _ mon . not .
_ 363 l61 ramirez - ruiz e and fenimore e e 2000 _ astrophys .
j. _ * 539 * 712 zeldovich ya b 1964 _ sov .
dokl . _ * 9 * 195 salpeter e e 1964 _ astrophys
. j. _ * 140 * 796 ramirez - ruiz e and socrates a 2005 _ astrophys .
j. submitted _ ( astro - ph/0504257 ) chevalier r a 1989 _ astrophys .
j. _ * 346 * 847 houck j c and chevalier r a 1991 _ astrophys .
j. _ * 376 * 234 goodman j , dar a and nussinov s 1987 _ astrophys .
j. _ * 314 * l51 eichler d , livio m , piran t and schramm dn 1989 _ nature _ * 340 * 126 paczyski b 1991 _ acta astron . _ * 41 * 257 narayan r , paczyski b and piran t 1992 _ astrophys . j. _ * 395 * l83 mszros p and rees m j 1992 _ astrophys .
j. _ * 397 * 570 mochkovitch r , hernanz m , isern j and martin x 1993 _ nature _ * 361 * 236 jaroszynski m 1993 _ acta astron . _ * 43 * 183 jaroszynski m 1996 _ astron . and astrophys . _ * 305 * 839 hulse r a and taylor j h 1975 _ astrophys .
j. _ * 195 * l51 lee w h and kluniak w 1995 _ acta astron . _ * 45 * 705 kluniak w and lee w h 1998 _ astrophys . j. _ * 494 * l53 janka h th , eberl t , ruffert m and fryer c l 1999 _ astrophys . j. _ * 527 * l39 fryer c l et al 1999 _ astrophys . j. _ * 520 * 650 katz n and canel l m 1996 _ astrophys .
j. _ * 471 * 915 levan a j et al 2006 * 368 * l1 vietri m and stella l 1999_astrophys . j. _ * 527 * l43 macfadyen a i , ramirez - ruiz e and zhang w 2006 _ astrophys .
j. submitted _ ( astro - ph/0510192 ) hansen b m and murali c 1998 _ astrophys .
j. _ * 505 * l15 grindlay j , portegies zwart s and mcmillan s 2006 _ nature phys . _
* 2 * 116 lee w h , ramirez - ruiz e and van de ven g 2006 _ astrophys
. j. submitted _ fryer c l , woosley s e and hartmann d h 1999 _ astrophys .
j. _ * 526 * 152 clark j p a , van den heuvel e p j and sutantyo w 1979 _ astron . and astrophys . _
* 72 * 120 lipunov v m , postnov k a and prokhorov m e 1987 _ astron . and astrophys .
_ * 176 * l1 hills d , bender p l and webbink r f 1991 _ astrophys .
j. _ * 369 * 271 phinney e s 1991 _ astrophys .
j. _ * 380 * l17 tutukov a v and yungelson l r 1993 _ mon . not .
soc . _ * 260 * 675 lipunov v m et al 1995 _ astron . and astrophys .
_ * 298 * 677 portegies - zwart s f and spreeuw h n 1996 _ astron . and astrophys .
_ * 312 * 670 fryer c l , burrows a and benz w 1998 _ astrophys .
j. _ * 496 * 333 bethe h and brown g e 1998 _ astrophys .
j. _ * 506 * 780 belczyski k and bulik t 1999 _ astron . and astrophys .
_ * 346 * 91 bloom j s , sigurdsson s and pols o r 1999 _ mon . not .
soc . _ * 305 * 763 belczyski k and kalogera v 2001 _ astrophys .
j. _ * 550 * l183 belczyski k et al 2002 _ astrophys .
j. _ * 571 * 394 belczyski k et al 2006 _ astrophys .
j. _ * 648 * 1110 izzard r g et al 2004 _ mon . not .
soc . _ * 348 * 1215 habets g m h j 1986 _ astron . and astrophys .
_ * 167 * 61 schutz b 1986 _ nature _ * 323 * 310 cutler c and flanagan e e 1994 _ phys .
d _ * 49 * 2658 bloom , j. s. 2002 _ ph.d .
thesis _ spruit h c 1999 _ astron . and astrophys .
_ * 341 * l1 rosswog s , ramirez - ruiz e and davies m b 2003 * 345 * 1077 rosswog s and ramirez - ruiz e _ aip conf
* 727 * 361 price d j and rosswog s 2006 _ science _ * 312 * 719 pacini f 1967 _ nature _ * 216 * 567 gold t 1968 _ nature _ * 218 * 731 usov v v 1994 _ mon . not .
soc . _ * 267 * 1035 thompson c 1994 _ mon . not .
soc . _ * 270 * 480 penrose r and floyd g r 1971 _ nature phys .
sci . _ * 229 * 177
ruffini r and wilson j r 1975 _ phys .
* 12 * 2959 lee w h and ramirez - ruiz e 2006 _ astrophys .
j. _ * 641 * 961 bondi h 1952 _ mon . not .
soc . _ * 112 * 195 shakura
n i and sunyaev r a 1973 _ astron . and astrophys . _ * 24 * 337 ichimaru s 1977 _ astrophys .
j. _ * 214 * 840 narayan r and yi i 1994 _ astrophys
. j. _ * 428 * l13 abramowicz m a , chen x , kato y , lasota j p and regev o 1995 _ astrophys . j. _ * 438 * l37 rosswog s and ramirez - ruiz e 2002 _ mon . not .
soc . _ * 336 * l7 blandford r d 2002 lighthouses of the universe : the most luminous celestial objects and their use for cosmology ed .
m gilfanov , r sunyaev and e churazov springer - verlag 2002 * 381 * usov v v 1992 _ nature _ * 357 * 472 mszros p and rees m j 1997 _ astrophys .
j. _ 482 l29 blandford
r d and znajek r l 1977 _ mon . not .
* 179 * 433 spruit h c , daigne f and drenkhahn g 2001 _ astron . and astrophys . _ * 369 * 694 lyutikov m and blandford r ( astro - ph/0312347 ) qian y z and woosley s e 1996 _ astrophys .
j. _ * 471 * 331 panaitescu a , kumar p and narayan r 2001 _ astrophys . j. _ * 561 * l171 perna r and belczyski k 2002 _ astrophys . j. _ * 570 * 252 lee w h and ramirez - ruiz e 2002 _ astrophys . j. _ * 577 * 893 knigl a and granot j 2002 _ astrophys . j. _ * 574 * 134 rosswog s 2006 _ revista mexicana de astronoma y astrofsica in press _ weisberg j m and taylor j h 2003 _ asp conf .
ser . 302 : radio pulsars _ * 302 * 93 stairs i h , thorsett s e , taylor j h and wolszczan a 2002 _ astrophys
. j _ * 581 * 501 lyne a g 2004 _ science _ * 303 *
1153 kramer m et al .
2006 _ science _ * 314 * 97 tauris t m and van den heuvel e 2006 _ cambridge astrophysics series _ * 39 * nice d j et al 2005 _ astrophys .
j. _ * 634 * 1242 bodmer a r 1971 _ phys .
d _ * 4 * 1601 witten e 1984 _ phys .
d _ * 30 * 272 fahri e and jaffe r l 1984 _ phys .
d _ * 30 * 2379 alcock c , fahri e and olinto a 1986 _ astrophys . j. _ * 310 * 261 jaikumar p , reddy s and steiner a w 2006 _ mod .
* 21 * 1965 page d and cumming a 2005 _ astrophys .
j. _ * 635 * l157 cumming a , macbeth j , int zand j j m and page d 2006 _ astrophys .
j. _ * 646 * 429 stairs i h 2004 _ science _ * 304 * 547 mcclintock j e and remillard r a 2006 _ ann . rev
. astron . astrophys . in press _
heger a , fryer c l , woosley s e , langer n and hartmann d h 2000 _ astrophys . j. _ * 591 * 288 rhoades c e and ruffini r 1974 _ astrophys .
j. _ * 32 * 324 mochkovitch r and livio m 1989 _ astron . and astrophys .
_ * 209 * 111 mochkovitch r and livio m 1990 _ astron . and astrophys .
_ * 236 * 378 nomoto k and kondo y 1991 _ astrophys .
j. _ * 367 * l19 segretain l , chabrier g and mochkovitch r 1997 _ astrophys . j. _ * 481 * 355 fryer c , benz w , herant m and colgate s a 1999 _ astrophys .
j. _ * 516 * 892 king a r , pringle j e and wickramasinghe d t 2001 _ mon . not .
soc . _ * 320 * l45 dessart l , burrows a , ott c d , livne e , yoon s y andlanger n 2006 _ astrophys
. j. _ * 644 * 1063 kidder l e , will c m and wiseman a g 1992 _ class . quant .
grav . _ * 9 * l125 cutler c et al 1993 _ phys
lett . _ * 70 * 2984 blanchet l et al 1995 _ phys .
lett . _ * 74 * 3515 buonanno a , chen y and vallisneri m 2003 _ phys .
* 4016 blanchet l , damour t , esposito - farese g , iyer b r 2004 phys . rev . lett . * 93 * 1101 blanchet l , damour t , esposito - farese g , iyer b r 2004 _ phys .
d _ * 71 * 4004 pati m e and will c m 2000 _ phys .
d _ * 62 * 4015 pati m e and will c m 2002 _ phys .
d _ * 65 * 4008 will c m 2005 _ phys
d _ * 71 * 4027 blanchet l 2006 _ living reviews in relativity _ ( www.livingreviews.org/lrr-2006-4 )
burgay m et al 2003 _ nature _ * 426 * 531 abramowici m , et al 1992 _ science _ * 256 * 325 bradaschia c et al 1990 _ nucl .
instrum . meth .
sect . a _ * 289 * 518 kokkotas k and schmidt b g 1999 _ living reviews in relativity _ ( www.livingreviews.org/lrr-1999-2 ) faber j a , grandclment p and rasio f a 2002 _ phys . rev .
letters _ * 89 * 231102 rees m j 1999 _ astron . and
astrophys . supp . _
* 138 * 491 centrella j 2006 _ sixth international lisa symposium proceedings in press aip _ ( astro - ph/0609172 ) lattimer j m and schramm d n 1974 _ astrophys .
j. _ * 192 * l145 lattimer j m and schramm d n 1976 _ astrophys .
j. _ * 210 * 549 burbidge e m , burbidge g r , fowler w a and hoyle f 1975 _ rev
* 29 * 547 wallerstein g et al 1997 _ rev .
phys . _ * 69 * 995 klebesadel r w , strong i b and olson r a 1973 _ astrophys
. j _ * 182 * l85 wheeler j a 1971 _ pontificae acad .
scripta varia _ * 35 * 539 mashhoon b 1973 _ astrophys .
j. _ * 181 * l65 glendenning n k 2000 _ compact stars _
a&a library springer chandrasekhar s 1969 _ ellipsoidal figures of equilibrium _ yale university press new haven connecticut lai d , rasio f a and shapiro s l 1993 _ astrophys . j. _ * 406 * l63 lai d , rasio f a and shapiro s l 1993 _ astrophys . j. supp _ * 88 * 205 lai d , rasio f a and shapiro s l 1994 _ astrophys . j. _ * 420 * 811 lai d , rasio f a and shapiro s l 1994 _ astrophys . j. _ * 423 * 344 lai d , rasio f a and shapiro s l 1994 _ astrophys
. j. _ * 437 * 752 frank j , king a and raine d _ accretion power in astrophysics _ 1992 cambridge university press cambridge oohara k and nakamura t 1989 _ prog . theor .
* 82 * 535 oohara k and nakamura t 1990 _ prog .
phys . _ * 83 * 906 oohara k and nakamura t 1992 _ prog .
phys . _ * 88 * 307 nakamura t and oohara k 1989 _ prog .
phys . _ * 82 * 1066 nakamura t and oohara k 1991 _ prog .
phys . _ * 86 * 73 zhuge x , centrella j m and mcmillan s l w 1994 _ phys .
d _ * 50 * 6247 zhuge x , centrella j m and mcmillan s l w 1996 _ phys .
d _ * 54 * 7261 shibata m , nakamura t and oohara k 1992 _ prog .
phys . _ * 88 * 1079 shibata m , nakamura t and oohara k 1993 _ prog .
phys . _ * 89 * 809 shibata m 1997 _ phys .
d _ * 55 * 6019 wilson j r mathews g j and marronetti p 1996 _ phys .
d _ * 54 * 1317 marronetti p , mathews g j and wilson j r 1998 _ phys .
d _ * 58 * 107503 mathews g j and wilson j r 2000 _ phys .
d _ * 61 * 127304 rasio f a and shapiro s l 1992 _ astrophys .
j. _ * 401 * 226 rasio f a and shapiro s l 1994 _ astrophys .
j. _ * 432 * 242 rosswog s et al 1999 _ astron . &
_ * 341 * 499 rosswog s , davies m b , thielemann f k and piran t 2000 _ astron . and astropys . _
* 360 * 171 rosswog s and davies m b 2002 _ mon . not .
* 334 * 481 davies m b , benz w , piran t and thielemann f k 1994 _ astrophys . j. _ * 431 * 742 janka h t and ruffert m 1996 _ astron . and astrophys . _ * 307 * l33 ruffert m , janka h th and schafer g 1996 _ astron . and astrophys . _ * 311 * 532 ruffert m , janka h th , takahashi k and schafer g 1997 _ astron . and astrophys . _ * 319 * 122 ruffert m and janka h th 2001 _ astron . and astrophys . _ * 380 * 544 rosswog s and liebendorfer m 2003 _ mon . not .
* 342 * 673 lee w h and kluniak w 1999a _ astrophys .
j. _ * 526 * 178 lee w h and kluniak w 1999b _ mon . not .
soc . _ * 308 * 780 lee w h 2000 _ mon . not .
soc . _ * 318 * 606 lee w h 2001 _ mon . not .
soc . _ * 328 * 583 rasio f a and shapiro s l 1995 _ astrophys .
j. _ * 438 * 887 duez m , baumgarte t w , shapiro s l , shibata m and uryu k 2002 _ phys . rev .
d _ * 65 * 024016 bildsten l and cutler c 1992 _ astrophys .
j. _ * 400 * 175 kochanek c 1992 _ astrophys .
j. _ * 398 * 234 faber j a and rasio f a 2002 _ phys
d _ * 65 * 084042 benz w , bowers r l , cameron a g w and press w h 1990 _ astrophys
. j _ * 348 * 647 freiburghaus c , rosswog s and thielemann f - k 1999 _ astrophys .
j. _ * 525 * l121 davies m b , levan a j and king a r 2005 _ mon . not .
soc . _ * 356 * 54 rosswog s , speith r and wynn g a 2004 _ mon . not .
soc . _ * 351 * 1121 rosswog s 2005 _ astrophys .
j. _ * 634 * 1202 faber j a et al 2006a _ phys .
d _ * 73 * 4012 faber j a et al 2006b _ astrophys .
j. _ * 641 * l93 miller m c 2005 _ astrophys .
j. _ * 626 * l41 prakash m , ratkovic s and lattimer j m 2004 _ j. phys .
g _ * 30 * s1279 rasio f a et al 2005 _ proceedings of jgrg14 yukawa institute for theoretical physics _
( astro - ph/0503007 ) paczynsky b and wiita p j 1980 _ astron . and astrophys .
_ * 88 * 23 shibata m 1999 _ phys
d _ * 60 * 104052 shibata m and uryu k 2000 _ phys .
d _ * 61 * 064001 shibata m , taniguchi k and uryu k 2003 _ phys .
d _ * 68 * 084020 shibata m , taniguchi k and uryu k 2005 _ phys .
d _ * 71 * 084021 shibata m and taniguchi k 2006 _ phys .
d _ * 73 * 064027 oechslin r and janka h th 2006 _ mon . not .
soc . _ * 368 * 1489 lattimer j m and swesty
f d 1991 _ nuclear physics a _ 1991 * 535 * 331 shen h et al 1998 _ nuclear phys .
1998 * 637 * 435 akmal a , pandharipande d r and ravenhall v g 1998 _ phys .
c _ * 58 * 1804 cook g b , shapiro s l and teukolsky s a 1994 _ astrophys .
j. _ * 424 * 823 baumgarte t w , shapiro s l and shibata m 2000 _ astrophys .
j. _ * 528 * l28 stella l and vietri m 1998 _ astrophys . j. _ * 507 * l45 duez m d , liu y t , shapiro s l , shibata m and stephens b c 2006 _ phys . rev .
lett . _ * 96 * 03101 shibata m , duez m d , liu y t , shapiro s l and stephens b c _ astrophys . j. _ * 96 * 031102 duez m d , liu y t , shapiro s l , shibata m and stephens b c 2006 _ phys
d _ * 73 * 104015 frank j _ mon . not .
soc . _ * 184 * 87 lacy j h , townes c h and hollenbach d j 1982 _ astrophys .
j. _ * 262 * 120 carter b and luminet j p _ astron . and astrophys . _ * 121 * 97 rees m j 1988 _ nature _ * 333 * 523 evans c r and kochanek c s 1989 _ astrophys .
j. _ * 346 * l13 laguna p , miller w a , zurek w h and davies m b 1993 _ astrophys .
j. _ * 410 * l83 ayal s , livio m and piran t 2000 _ astrophys .
j. _ * 545 * 772 michel f c 1988 _ nature _ * 333 * 644 macfadyen a i and woosley s e 1999 _ astrophys . j. _ * 524 * 262 woosley s e 1993 _ astrophys . j. _ * 405 * 273 popham r , woosley s e and fryer c l 1999 _ astrophys . j. _ * 518 * 356 narayan r , piran t and kumar p 2001 _ astrophys .
j. _ * 557 * 949 kohri k and mineshige s 2002 _ astrophys .
j. _ * 577 * 311 dimatteo t , perna r and narayan r 2002 _ astrophys .
j. _ * 579 * 706 chen w x and beloborodov a m 2006 _ astrophys .
j. in press _
( astro - ph/0607145 ) janiuk a , perna r , dimatteo t and czerny b 2004 _ mon . not .
* 355 * 950 lee w h , ramirez - ruiz e and page d 2004 _ astrophys . j. _ * 608 * l5 lee w h , ramirez - ruiz e and page d 2005 _ astrophys . j. _ * 632 * 421 setiawan s , ruffert m and janka h th 2004 _ mon . not .
soc . _ * 352 * 753 setiawan s , ruffert m and janka h th 2005 _ astron . and astrophys . _ * 458 * 553 ruffert m and janka h th 1999 _ astron . and astrophys . _ * 344 * 573 janka h th and mller e 1996 _ astron . and astrophys . _ *
306 * 167 tassoul j l 1978 _ theory of rotating stars princeton university press _
artemova i v , bjrnsson g and novikov i d 1996 _ astrophys .
j. _ * 461 * 565 salmonson j d and wilson g r 1999 _ astrophys . j. _ * 517 * 859 asano k and fukuyama t 2000 _ astrophys .
j. _ * 531 * 949 asano k and fukuyama t 2001 _ astrophys .
j. _ * 546 * 1019 miller w a , george n d , kheyfets a and mcghee j m 2003 _ astrophys . j. _ * 583 * 833 birkl r , aloy m a , janka h th and mueller e 2006 _ astron . and astrophys . submitted _ ( astro - ph/0608543 ) blinnikov s i , dunina
barkovskaya n v and nadyozhin d k 1996 _ astrophys
. j. supp _ * 106 * 171 beloborodov a m 2003 _ astrophys .
j. _ * 588 * 931 langanke j and martnez pinedo g 2001 _ at .
. data . tables _ * 79 * 1 itoh n , hayashi s , nishikawa a and kohyama y 1996 _ astrophys .
j. supp . _
* 102 * 411 hannestad s and raffelt g 1998 it astrophys .
j. * 507 * 339 paciesas w s 1999 _ astrophys .
j. supp _ * 122 * 465 rosswog s and liebendorfer m 2003 _ mon . not .
soc . _ * 342 * 673 rosswog s and ramirez - ruiz e 2003 _ mon . not .
soc . _ * 343 * l36 pruet j , woosley s e and hoffman r d 2003 _ astrophys .
j. _ * 586 * 1254 pruet j , thompson t a and hoffman r d 2004 _ astrophys . j. _ * 606 * 1006 kohri k , narayan r and piran t 2005 _ astrophys . j. _ * 629 * 341 levinson a and eichler d 2000 _ phys .
letters _ * 85 * 236 aloy m a , janka h th and mller e 2005 _ astron . and astrophys .
_ * 436 * 273 abramowicz m a , calvani m and nobili l 1983 _ nature _ * 302 * 597 quataert e and gruzinov a 2000 _ astrophys
. j. _ * 539 * 809 burrows d n et al 2005 _ science _ * 309 * 1833 nousek j a. et al 2006 _ astrophys . j. _ * 642 * 389 ramirez - ruiz e and merloni a 2001 _ mon . not .
soc . _ * 320 * l25 zhang b 2006 _ aip conf .
* 836 * 392 liang e w et al 2006 _ astrophys .
j. _ * 646 * 351 giannios d 2006 _ astron . and astrophys .
_ * 455 * l5 dermer c d and atoyan a 2006 _ astrophys . j. _ * 643 * l13 king a r et al 2005 _ astrophys . j. _ * 630 * l113 proga d and zhang b 2006 _ mon . not .
soc . _ * 370 * l61 perna r , armitage p j and zhang b 2005 _ astrophys . j. _ * 636 * l29 gao w h and fan y z 2006 _ chin . j. astron . and astrophys . _
* 6 * 513 van putten m h p m and ostriker ec 2001 _ astrophys .
j. _ * 552 * l31 krolik j h , hawley j f and hirose s 2005 _ astrophys .
j. _ * 622 * 1008 balbus s a and hawley j f 1998 _ reviews of modern physics _ * 70 * 1 lin d n c and pringle j e 1987 _ mon . not .
soc . _ * 225 * 607 gammie c f 2001 _ astrophys .
j. _ * 553 * 174 balbus s a and papaloizou j c b 1999 _ astrophys .
j. _ * 521 * 650 chandrasekhar s 1961 _ hydrodynamic and hydromagnetic stability _
oxford press lee w h , kluniak w and nix j 2001 _ acta astron . _ * 51 * 331 ramirez - ruiz e 2004 _ mon . not .
_ 349 l38 atkins r et al 2000 _ astrophys .
j. _ * 533 * l119 waxman e 2004 _ new journal of physics _ * 6 * 140 dermer c 2005 _ nuovo cimento _ * 28 * 789 waxman e 2006 _ aip conf
* 836 * 589 waxman e 2004 _ astrophys . j. _ * 606 * 988 hayashida n et al 1999 _ astrophys . j. _ * 522 * 225 waxman e 1995 _ phys . rev . lett . _ * 75 * 386 waxman e and bahcall j n 1997 _ phys .
lett . _ * 78 * 2292 waxman e and bahcall j n 1999 _ phys.rev .
d. _ * 59 * 023002 derishev e v , kocharovsky v v and kocharovsky vl v 1999 _ astrophys .
j. _ * 521 * 640 mckinney j c 2006 _ mon . not .
soc . _ * 368 * 1561 granot j. 2006 _
revista mexicana de astronomia y astrofisica in press _ blandford r and eichler d 1987 _ physics reports _ * 154 * 1 achterberg a et al 2001 _ mon . not .
soc . _ * 328 * 393 baumjohann w. and treumann r a 1996 _ basic space plasma physics _
imperial college press london treumann r a and baumjohann w 1997 _ advanced space plasma physics _
imperial college press london | recent months have witnessed dramatic progress in our understanding of short @xmath0-ray burst ( sgrb ) sources .
there is now general agreement that sgrbs or at least a substantial subset of them are capable of producing directed outflows of relativistic matter with a kinetic luminosity exceeding by many millions that of active galactic nuclei . given the twin requirements of energy and compactness , it is widely believed that sgrb activity is ultimately ascribable to a modest fraction of a solar mass of gas accreting onto a stellar mass black hole or to a precursor stage whose inevitable end point is a stellar mass black hole .
astrophysical scenarios involving the violent birth of a rapidly rotating neutron star , or an accreting black hole in a merging compact binary driven by gravitational wave emission are reviewed , along with other possible alternatives ( collisions or collapse of compact objects ) .
if a black hole lies at the center of this activity , then the fundamental pathways through which mass , angular momentum and energy can flow around and away from it play a key role in understanding how these prime movers can form collimated , relativistic outflows .
flow patterns near black holes accreting matter in the _ hypercritical _ regime , where photons are unable to provide cooling , but neutrinos do so efficiently , are discussed in detail , and we believe that they offer the best hope of understanding the _ central engine_. on the other hand , statistical investigations of sgrb niches also furnish valuable information on their nature and evolutionary behavior .
the formation of particular kinds of progenitor sources appears to be correlated with environmental effects and cosmic epoch .
in addition , there is now compelling evidence for the continuous fueling of sgrb sources .
we suggest here that the observed late flaring activity could be due to a secondary accretion episode induced by the delayed fall back of material dynamically stripped from a compact object during a merger or collision .
some important unresolved questions are identified , along with the types of observation that would discriminate among the various models .
many of the observed properties can be understood as resulting from outflows driven by hyperaccreting black holes and subsequently collimated into a pair of anti - parallel jets .
it is likely that most of the radiation we receive is reprocessed by matter quite distant to the black hole ; sgrb jets , if powered by the hole itself , may therefore be one of the few observable consequences of how flows near nuclear density behave under the influence of strong gravitational fields . |
charmonium spectroscopy has become a challenging topic in hadron physics and qcd , because of the recent findings of possible new charmonium states ( for recent experimental and theoretical reviews and discussions , see e.g. @xcite and references therein ) . among others , for the puzzling state x(3872 ) , possible assignments of e.g. the @xmath8 and @xmath9 charmonium states and the charm - molecule
have been suggested ( see , e.g.@xcite for a comprehensive review ) , and it will be helpful to search for those states in other experiments and to clarify these assignments ; the measured mass splitting between @xmath10 and @xmath11 is about 50 mev , which is smaller than some theoretical predictions , and it is certainly useful to search for the @xmath12 to see what will be the mass splitting between the @xmath13 , which could be the observed @xmath14 , and the @xmath12 .
this may be particularly interesting since according to some potential model calculations the @xmath12 could lie above 4040 mev ( see , e.g. in @xcite the mass of @xmath12 is predicted to be 4060 mev ) . and the @xmath13 mass could actually be lowered by coupling to the nearby @xmath15 decay channels ( note that the energy level spacing between @xmath16 and @xmath17 is smaller than that between @xmath18 and @xmath16 , which is in contradiction with potential model calculations unless the coupled channel effects are considered or the assignments for @xmath16 and @xmath18 are incorrect
. the mass spectrum of excited charmonium states will certainly provide important information on interquark forces and color confinement .
in addition , studies of the decay and production of these states will also be very important in understanding the underlying theory of strong interaction perturbative and nonperturbative qcd in view of many outstanding puzzles in charmonium physics .
@xmath2 meson decays have proven to be very useful processes to find new charmonium states . aside from the @xmath2 meson decay , @xmath0 annihilation at @xmath1 gev
could also be a very useful process in finding the excited charmonium states , since the recent belle experiments @xcite have found unusually strong signals for the double charmonuim production from the @xmath0 continuum , e.g. , @xmath19 , @xmath20 and @xmath21 , @xmath22 .
theoretically , the calculated cross sections for these processes based on the leading order non - relativistic qcd(nrqcd ) ( or more generally perturbative qcd ( pqcd ) ) are about an order of magnitude smaller than the experiments @xcite .
this is a big issue in charmonium physics and nrqcd , and it still remains to be further clarified though many considerations are suggested to understand the large production rates in both exclusive and inclusive charmonium production via double charm pairs in @xmath0 annihilation @xcite ( the theoretical predictions for the inclusive @xmath23 production cross section with the color - singlet @xcite as well as color - octet@xcite contributions are also much smaller than the belle data ) . despite of these disagreements , however , we find that the calculated relative rates of the double charmonium production processes are roughly compatible with the belle data ( e.g. the production cross sections of @xmath24 , @xmath11 , and @xmath25 associated with @xmath26 and @xmath10 are much larger than that of @xmath27 and @xmath28 ) .
so , we may use the same method as in our previous work to calculate the production rates for the excited charmonium states in @xmath0 annihilation into double charmonia , but mainly pay attention to the relative rates for these production processes .
we hope the calculation will make sense in predicting the relative production rates for those excited charmonium states , and can be tested by experiments
. this will be useful not only in the search for those excited charmonium states , but also in understanding the production mechanism itself .
if the predicted relative production rates turn out to be consistent with experiments , it is likely that the nrqcd factorization treatment for these processes probably still makes sense and only an overall enhancement factor is needed and should be clarified in further theoretical considerations ( including qcd radiative corrections , relativistic corrections , and other nonperturbative qcd effects ) . in the last section
we will have a discussion on recent developments in this regard . in the following
, we will calculate the leading order production cross sections for various excited charmonium states in @xmath0 annihilation at @xmath1 gev in the same way as in @xcite .
following the nrqcd factorization formalism@xcite , the scattering amplitude of double charmonia production can be described as @xmath29 where @xmath30 , @xmath31 , @xmath32 , @xmath33 , @xmath34 and @xmath35 are respectively the color - su(3 ) , spin - su(2 ) , and angular momentum clebsch - gordon coefficients for @xmath36 pairs projecting out appropriate bound states .
@xmath37 is the scattering amplitude for double @xmath36 production and @xmath38 is the derivative of the amplitude with respect to the relative momentum between the quark and anti - quark in the bound state .
the coefficients @xmath39 and @xmath40 can be related to the radial wave function of the bound states or its derivative with respect to the relative spacing as @xmath41 we introduce the spin projection operators @xmath42 as@xcite @xmath43 expanding @xmath44 in terms of the relative momentum @xmath45 , we get the projection operators and their derivatives , which will be used in our calculation , as follows @xmath46 @xmath47 @xmath48 @xmath49.\ ] ] we then get the following expressions and numerical results for various processes of double charmonium production in @xmath0 annihilation at @xmath1 gev . in the calculation of the short distance coefficients ,
the quark and anti - quark are all on mass shell , and the meson masses are taken to be @xmath50 .
the input parameters are @xmath51 ( corresponding to @xmath52 and @xmath53=338 mev ) , and the wave functions at the origin are taken from a potential model calculation ( see the qcd ( bt(buchmller - tye ) ) model in ref.@xcite ) : @xmath54 , @xmath55 , @xmath56 , @xmath57 , @xmath58 , and @xmath59 .
in the two s - wave ( ns and ms ) case , the cross section for @xmath61 is given by @xmath62 where @xmath63 is the scattering angle between @xmath64 and @xmath65 , @xmath66 is as follows @xmath67 here the mandelstam variables are defined as @xmath68 @xmath69 @xmath70 the cross sections for the double s - wave charmonium production are listed as follows @xmath71)=5.5[3.6 , 3.1]~\rm{fb},\ ] ] @xmath72)=3.6[2.3 , 2.0]~\rm{fb},\ ] ] where the values in the brackets are the cross sections respectively for @xmath11 and @xmath12 with recoiling @xmath73 mesons . in the one spin - triplet s - wave ( ns ) and one spin - triplet p - wave ( mp ) case
, the cross section for @xmath75 process reads @xmath76 here for the production of spin - triplet states @xmath77 , @xmath66 is given in eq .
( [ kc0 ] ) for @xmath25 , in eq .
( [ kc1])for @xmath27 and eq . ( [ kc2 ] ) for @xmath28 , @xmath78 @xmath79 @xmath80 for the known spin - triplet p - wave states @xmath81 we find @xmath82)=6.7[1.1 , 1.6]~\rm{fb},\ ] ] @xmath83)=4.4[0.74 , 1.1]~\rm{fb},\ ] ] for the exited spin - triplet 2p states @xmath84 , which are to be searched for , we find @xmath85)=9.1[1.6 , 2.2]~\rm{fb},\ ] ] @xmath86)=5.9[1.0 , 1.4]~\rm{fb}.\ ] ] in eqs . (
1922 ) , we see that the cross sections for the excited @xmath87 states @xmath88 are somewhat larger than that for the corresponding @xmath89 states @xmath90 in the nonrelativistic limit .
this numerical result is due to the fact that we have chosen the first derivative of the wave function at the origin for the @xmath87 states to be larger than that for the @xmath89 states , and actually the former could be slightly smaller than the latter , depending on the potentials that are used ( see the values in the qcd ( bt ) model and other models in ref.@xcite ) .
furthermore , another important effect comes from the relativistic corrections , which may lower the cross sections for the @xmath88 states .
e.g. , if we take the charm quark mass to be @xmath91 for the 2p states , then the cross sections will be substantially lower . despite of these uncertainties , we expect that the cross sections for the @xmath87 states should be comparable to that for the @xmath89 states . to search for the spin singlet p - wave charmonium @xmath7 is certainly interesting .
recently , cleo has found evidence for @xmath7 in the @xmath93 decay followed by @xmath94 with a mass of @xmath95 and hyperfine splitting of about 1.0 mev measured in both the @xmath24 exclusive and inclusive analysis@xcite .
it will also be interesting to search for the @xmath7 in @xmath0 annihilation at @xmath1 gev in the recoil spectra of charge parity c=+1 states such as @xmath96 .
for the production of one spin - singlet s - wave state @xmath97 and one spin - singlet p - wave state @xmath98 , differing from @xcite , we find @xmath66 to be not vanishing but given by eq .
( [ echc ] ) : @xmath99 for the spin - singlet p - wave states @xmath100 and @xmath101 we find @xmath102)=0.73[0.99]~\rm{fb},\ ] ] @xmath103)=0.48[0.65]~\rm{fb},\ ] ] @xmath104)=0.41[0.56]~\rm{fb}.\ ] ] in the two p - wave ( np and mp ) case , the cross section for @xmath106 is @xmath107 for the @xmath108 production @xmath66 reads @xmath109 @xmath110 @xmath111 and the corresponding cross sections are @xmath112)=0.22[0.31]~\rm{fb},\ ] ] @xmath113)=1.0[1.4]~\rm{fb},\ ] ] @xmath114)=0.063[0.085]~\rm{fb}.\ ] ] the cross section for @xmath116 process is formulated as @xmath117 @xmath118 and the numerical result is @xmath119 + ^1d_2)=0.19[0.12]~\rm{fb}.\ ] ] the cross section for @xmath121 process is formulated as @xmath122 @xmath123 @xmath124 @xmath125 @xmath126 @xmath127 @xmath128 the numerical results are listed as follows , where @xmath129 , @xmath130 + \delta_1)=0.080[0.041]~\rm{fb},\ ] ] @xmath130 + \delta_2)=0.099[0.084]~\rm{fb},\ ] ] @xmath130 + \delta_3)=0.041[0.0099]~\rm{fb},\ ] ] as a summary of the above results , we show the differential cross sections ( the angular distribution functions ) for different double charmonium production processes at leading order in nrqcd ( qed contributions are not included ) in table i , and the corresponding graphs in fig . 1 - 7 .
since the @xmath132 can be produced via a single photon , the qed contribution may be significant or even comparable to the qcd contribution in some exclusive processes involving one @xmath133 charmonium state .
these qed effects are considered in @xcite .
there are six feynman diagrams for the qed process at @xmath134 order , and only two represent @xmath135 , which is dominant and has been calculated in @xcite . in this paper
we include all the eight diagrams to get the full result at order @xmath134 , though the contributions of other six diagrams are numerically small . using the notation in section
ii , we re - express below the analytical formulas of the exclusive processes including both qcd and qed contributions .
the cross section for @xmath61 is now changed to @xmath136 where @xmath137 and @xmath63 is the scattering angle between @xmath138 and @xmath139 .
and @xmath140is @xmath141 the numerical results become @xmath71)=6.6[4.3 , 3.7]~\rm{fb},\ ] ] @xmath72)=4.3[2.8 , 2.4]~\rm{fb},\ ] ] where the values in the brackets are the cross sections respectively for @xmath11 and @xmath12 with recoiled @xmath73 mesons .
the cross section for @xmath61 is changed to @xmath143 and now @xmath66 for @xmath144 production , are given in eq .
( [ kc0e ] ) , eq .
( [ kc1e ] ) , eq .
( [ kc2e ] ) respectively .
@xmath146 @xmath147 the numerical results of spin - triplet p - wave states @xmath81 are @xmath148)=6.9[1.0 , 1.8]~\rm{fb},\ ] ] @xmath149)=4.5[0.7 , 1.1]~\rm{fb},\ ] ] and for the exited spin - triplet 2p states @xmath84 , the results turn to be @xmath150)=9.4[1.4 , 2.4]~\rm{fb},\ ] ] @xmath86)=6.2[0.9 , 1.6]~\rm{fb}.\ ] ] the cross section for @xmath152 process is @xmath153 and @xmath154 and the numerical result becomes @xmath119 + ^1d_2)=0.21[0.13]~\rm{fb}.\ ] ] the cross section for @xmath152 process is formulated as @xmath156 and @xmath157 and the numerical result becomes @xmath158 in this section we have considered the qed contribution to the exclusive processes involving one @xmath159 charmonium state ( e.g. , @xmath160 and @xmath161 ) .
we find that in general by adding qed contribution the cross section can be changed by the order of ten percent .
we list the results below with @xmath1gev , @xmath162 .
the cross sections of @xmath163 are increased by 20 percent , that of @xmath164 are increased by 4 , -5 , 8 percent for @xmath165 respectively , that of @xmath166 is increased by @xmath167 percent , and that of @xmath168 is increased by @xmath169 percent . in @xcite
the authors also considered the qed process but only the two dominant diagrams were included .
our results are in agreement with their results in most processes when we choose the same parameters as theirs , except for the process of @xmath170 production , for which they obtained 41% enhancement with qed effects whereas we get 19% with the same parameters . here
our analytical expression also differs from theirs .
we show the differential cross sections ( the angular distribution functions ) for different double charmonium production processes including both qcd and qed contributions in table ii , and the corresponding graphs in fig .
8 . ratios of production cross sections of various double charmonia to that of @xmath171 in @xmath172 annihilation at @xmath1 gev are listed in table iii .
in this paper , we make predictions for various double charmonia production processes in @xmath0 annihilation at @xmath1 gev with @xmath2 factories , based on a complete leading order calculation including both qcd and qed contributions . in particular , we aim at searching for excited charmonium states in these processes . from the obtained results we make the following observations : the calculated relative production rates for @xmath173 , @xmath174 are roughly compatible with the new belle measurements @xcite ( see also @xcite ) , assuming the decay branching ratios into charged tracks are comparable for @xmath175 and @xmath176 .
the calculated relative production rates for @xmath177 and @xmath178 are large , and these two states may be observable in the mass range @xmath179 gev and @xmath180 gev respectively ( see , e.g. @xcite ) .
both of them are above the ozi ( okubo - zweig - iizuka ) allowed thresholds , but unlike the @xmath181 , the @xmath12 can not decay to @xmath182 pair , which may distinguish between these two states experimentally .
the calculated relative production rate for @xmath183 is not zero .
this differs from the result given in @xcite , but agrees with @xcite .
hence the @xmath7 might be observable via this channel with high statistics in the future . aside from this process
, @xmath184 is also very hopeful in finding the @xmath7 meson , since the @xmath27 has a large branching ratio ( larger than 30% ) decaying into @xmath185 , and the @xmath26 can be easily detected by the @xmath186 signal .
moreover , the calculated cross section for @xmath187 is about 1.0 fb , not very small and much larger than that for @xmath188 and @xmath189 production rates .
we show the differential cross sections ( the angular distribution functions ) for different double charmonium production processes in table i ( not including qed contribution ) and table ii ( including both qcd and qed contributions ) , and the corresponding graphs in fig . 1 - 7 , and fig . 8
. as a whole
, our results agree with those in @xcite and @xcite ( also agree with our previous result for @xmath190@xcite ) , if using the same parameters . for the qed part , in @xcite
only two dominant diagrams are taken into account , while in this paper all diagrams are considered , but the numerical contributions of the remaining four diagrams are small . however , there still exists a difference in the result for the @xmath191 production . as for numerical results , since we use a larger value of the strong coupling constant @xmath192 ( corresponding to @xmath193 and @xmath194=338 mev ) than that in @xcite , our predicted cross sections are in general larger than that in @xcite .
moreover , we use the charmonium wave functions and their derivatives at the origin ( including the ground state and excited states ) from the bt potential model calculation@xcite but not from the experimental values of leptonic decay widths , etc .
, as in @xcite , and this may further enlarge our predicted cross sections .
these parameters may not be the best choice , but , at present , since we do not have enough available data for higher excited charmonium states , using the potential model calculation may still be a reasonable and tentative choice .
we view these as theoretical uncertainties in our approach for the leading order calculations .
as emphasized above , this paper aims at searching for excited charmonium states in @xmath0 annihilation at @xmath1 gev with @xmath2 factories .
after the main part of the results were presented in @xcite ( with a more complete leading order calculation including both qcd and qed contributions for various double charmonium production processes being added in its present form ) , a number of new experimental and theoretical results have appeared recently .
the double charmonium production in @xmath0 annihilation at b factories has been confirmed by the babar collaboration ( see ref.@xcite ) with comparable cross sections to that observed by belle .
theoretically , the large gap between experiment and theory could be largely narrowed by the next to leading order qcd radiative corrections @xcite and relativistic corrections ( see , refs.@xcite and references therein ) in the framework of nonrelativistic qcd , and other possible approaches ( see , e.g.@xcite ) .
belle has observed the x(3940 ) , a new charmonium state or charmonium - like state , in @xmath195 with @xmath196@xcite . while its main decay mode is @xmath197 , no signal is found for @xmath198 .
this rules out the possibility of x(3940 ) being the @xmath181 state .
furthermore , the x(3940 ) is unlikely to be the @xmath199 state , since no signal for the @xmath200 is found .
this is in line with our calculation , which shows that the cross section for @xmath201 is much smaller than @xmath202 .
finally , the x(3940 ) could be the @xmath12 state . according to our calculation ( see eq.(13 ) ) , the production cross sections for @xmath203 are respectively 5.5 , 3.6 , 3.1 fb , and the relative rates are roughly consistent with the signal yields @xmath204 events for @xmath205@xcite .
this might be viewed as a support to interpreting the x(3940 ) as the @xmath12 .
however , the remaining problem is how to understand its low mass of x(3940 ) if it is the @xmath12 , which is lower than potential model predictions by 50 - 120 mev . but this could be explained by the coupled channel effects that the coupling of @xmath12 to the @xmath206 and @xmath207 charmed meson pair ( in s - wave ) will lower the mass of @xmath12@xcite .
belle has not observed the x(3872 ) in @xmath208 .
this implies that the x(3872 ) is unlikely to be a conventional @xmath209 or @xmath210 charmonium , since in our calculation they may have relatively large rates to be observed in double charmonium production . on the other hand
, a @xmath211 or @xmath212 charmonium could be possible for x(3872 ) , since they have relatively small production rates .
of course , the nature of x(3872 ) needs clarifying by other more relevant experiments ( see , e.g. @xcite ) , aside from the @xmath0 annihilation processes .
belle has very recently observed a new state , the x(4160 ) , in the process of double charm production @xmath213 followed by @xmath214@xcite .
possible interpretations for the x(4160 ) are discussed in@xcite with emphasized possible assignments of @xmath215 charmonium states and related problems . in conclusion , we find that the double charmonium production processes in @xmath0 annihilation at @xmath2 factories are very useful tools in searching for excited charmonium states or charmonium - like states . the leading order nrqcd calculation might hopefully provide a useful guide for the relative production rates , but not the absolute rates themselves . a systematical study for the qcd radiative corrections and relativistic corrections for different processes
are apparently needed . on the other hand , studies of charmonium spectroscopy including charmonium masses ,
decays , and coupled channel effects , and the new type charmonium - like states are also very desirable .
we thank p. pakhlov for discussions on the belle data and possible implications , and e. braaten and j. lee for communications on some calculations .
this work was supported in part by the national natural science foundation of china ( no .
10421503 , no .
10675003 ) , the key grant project of chinese ministry of education ( no . 305001 ) , and the research found for doctorial program of higher education of china .
belle collaboration , k. abe _
et al . _ , phys .
lett . * 89 * , 142001 ( 2002 ) .
belle collaboration , k. abe _
et al . _ , hep - ex/0407009 , phys .
d70 , 071102 ( 2004 ) .
e. braaten and j. lee , phys . rev .
* d67 * , 054007 ( 2003 ) . b. aubert _
et al _ , the babar collaboration , phys.rev .
* d*72 , 031101(2005 ) .
y.j . zhang , y.j .
gao and k.t .
chao , phys .
96 , 092001(2006 ) ; see also y.j .
zhang and k.t .
chao , phys .
lett.98 , 092003 ( 2007 ) .
j.p . ma and z.g .
si , arxiv : hep - ph/0608221 ; v.v .
braguta , a.k .
likhoded , and a.v .
luchinsky , phys .
d72 , 074019 ( 2005 ) ; arxiv : hep - ph/0602047 ; arxiv : hep - ph/0611021 ; a.e . bondar and v.l .
chernyak , phys .
b612 , 215 ( 2005 ) ; d. ebert and a.p .
martynenko , phys .
d74 , 054008 ( 2006 ) ; h .-
choi and c .- r .
ji , arxiv : 0707.1173 [ hep - ph ] . | we suggest searching for excited charmonium states in @xmath0 annihilation via double charmonium production at @xmath1 gev with @xmath2 factories , based on a more complete leading order calculation including both qcd and qed contributions for various processes .
in particular , for the c=+ states , the @xmath3 ( n=2,3 ) and @xmath4 ( m=3,4 ) may have appreciable potentials to be observed ; while for the c=- states , the @xmath5 production and especially the @xmath6 production might provide opportunities for observing the @xmath7 with higher statistics in the future . a brief discussion for the x(3940 ) observed in the double charmonium production is included . |
billiards were introduced by birkhoff @xcite .
let @xmath7 be a smooth strictly convex curve in the plane , oriented counterclockwise , and let @xmath8 be the billiard table enclosed by @xmath7 .
billiard trajectories inside @xmath8 consist of polygonal lines inscribed in @xmath7 whose consecutive sides obey to the rule `` the angle of reflection is equal to the angle of incidence . ''
see @xcite for a general description .
a @xmath0-periodic billiard trajectory forms a closed polygon with @xmath6 sides that makes @xmath9 turns inside @xmath7 .
birkhoff @xcite proved that there are at least two different birkhoff @xmath0-periodic billiard trajectories inside @xmath8 for any relatively prime integers @xmath9 and @xmath6 such that @xmath10 .
let @xmath11 be the supremum of the absolute values of the differences of the lengths of all such trajectories .
the quantities @xmath11 were already studied by marvizi and melrose @xcite and colin de verdire @xcite for smooth tables .
the former authors produced an asymptotic expansion of the lengths for @xmath0-periodic billiard trajectories approaching @xmath7 when @xmath9 is fixed and @xmath12 .
they saw that there exists a sequence @xmath13 , depending only on @xmath9 and @xmath7 , such that , if @xmath14 is the length of any @xmath0-periodic trajectory , then @xmath15 + \sum_{k \ge 1 } \frac{l_k}{q^{2k } } , \qquad q\to\infty,\ ] ] where @xmath16 , and @xmath17 is the curvature of @xmath7 as a function of the arc - length parameter @xmath18 .
the symbol @xmath19 means that the series in the right hand side is asymptotic to @xmath20 .
the assymptotic coefficients @xmath21 can be explicitly written in terms of the curvature @xmath17 .
for instance , the explicit formulas for @xmath22 , @xmath23 , @xmath24 , and @xmath25 can be found in @xcite . since the expansion of the lengths in powers of @xmath26 coincides for all these @xmath0-periodic trajectories , @xmath27 for smooth strictly convex tables when @xmath9 is fixed and @xmath12 .
colin de verdire studied the lengths of periodic trajectories close to an elliptic @xmath28-periodic trajectory on a smooth symmetric billiard table , and found that the quantities @xmath11 are again beyond all order with respect to @xmath6 .
these works suggest that the supremum length differences @xmath11 are exponentially small in the period @xmath6 for analytic strictly convex tables .
indeed , we have proved that if @xmath7 is analytic and @xmath9 is a fixed positive integer , then there exists @xmath29 such that @xmath30 for all integer @xmath31 relatively prime with @xmath9 .
the exponent @xmath32 is related to the width of a complex strip where a certain @xmath33-periodic angular coordinate is analytic .
a more precise statement is given in theorem [ thm : classicalboundary ] .
similar exponentially small upper bounds hold in other billiard problems .
we mention two examples .
first , for @xmath0-periodic billiard trajectories inside strictly convex analytic tables of constant width when @xmath34 .
second , for @xmath0-periodic billiard trajectories inside strictly convex analytic tables in surfaces of constant curvature when @xmath35 .
the billiard dynamics close to the boundary has also been studied from the point of view of kam theory .
lazutkin @xcite proved that there are infinitely many caustics inside any @xmath36 strictly convex table .
these caustics accumulate at the boundary of the table , and have diophantine rotation numbers .
douady @xcite improved the result to @xmath37 billiard tables .
a special remark on the relevance of these results is the following .
kac @xcite formulated the inverse spectral problem for planar domains .
that is , to study how much geometric information about @xmath8 can be obtained from the laplacian spectrum with homogeneous dirichlet conditions on @xmath7 .
andersson and melrose @xcite gave an explicit relation between the length spectrum and the laplacian spectrum .
the length spectrum of @xmath8 is the union of the lengths of all its @xmath0-periodic billiard trajectories and all the integer multiples of @xmath38 $ ] .
see also @xcite .
our results also apply to the dual billiards introduced by day @xcite and popularized by moser @xcite as a crude model for planetary motion . some general references are @xcite . let @xmath39 be unbounded component of @xmath40 .
the dual billiard map @xmath41 is defined as follows : @xmath42 is the reflection of @xmath43 in the tangency point of the oriented tangent line to @xmath7 through @xmath43 .
billiards and dual billiards are projective dual in the sphere @xcite .
a @xmath0-periodic dual billiard trajectory forms a closed circumscribed polygon with @xmath6 sides that makes @xmath9 turns outside @xmath7 .
the area of a @xmath0-periodic trajectory is the area enclosed by the corresponding polygon , taking into account some multiplicities if @xmath44 .
there are at least two different birkhoff @xmath0-periodic dual billiard trajectories outside @xmath7 for any relatively prime integers @xmath9 and @xmath6 such that @xmath45 and @xmath10 .
tabachnikov @xcite studied the supremum @xmath46 of the absolute value of the differences of the areas enclosed by all such @xmath5-periodic trajectories for smooth tables .
he proved that there is a sequence @xmath47 , depending only on @xmath7 , such that , if @xmath48 is the area enclosed by any @xmath5-periodic dual billiard trajectory , then @xmath49 + \sum_{k \ge 1 } \frac{a_k}{q^{2k } } , \qquad q\to\infty,\ ] ] where @xmath50 .
hence , the expansion of the areas in powers of @xmath26 coincides for all these @xmath5-periodic trajectories , and so , @xmath51 for smooth strictly convex dual tables when @xmath12 .
douady @xcite found the existence of infinitely many invariant curves outside any @xmath37 strictly convex dual table .
these invariant curves accumulate at the boundary of the dual table and have diophantine rotation numbers . in a completely analogous way to ( classical ) billiards , we have proved that , once fixed any positive integer @xmath9 , if @xmath7 is analytic , then there exists @xmath29 such that @xmath52 for all integer @xmath31 relatively prime with @xmath9 . once more , the exponent @xmath32 is related to the width of a complex strip where a certain @xmath33-periodic angular coordinate is analytic . the precise statement is given in theorem [ thm : dualboundary ] . still in the context of dual billiards , the points at infinity can be seen as @xmath28-periodic points , hence they form a @xmath28-resonant ric .
douady @xcite found the existence of infinitely many invariant curves outside any @xmath53 strictly convex dual table .
these invariant curves accumulate at infinity and have diophantine rotation numbers .
we have proved that , once fixed any constant @xmath54 , if @xmath7 is analytic , then there exist @xmath29 such that @xmath55 for all relatively prime integers @xmath9 and @xmath6 such that @xmath56 and @xmath31 .
see theorem [ thm : dualinfinity ] .
the three exponents @xmath32 that appear in the exponentially small upper bounds ( [ eq : bounddifferencelengths ] ) , ( [ eq : bounddifferenceareas ] ) , and ( [ eq : bounddifferenceareas_bis ] ) may be different , since each one is associated to a different analyticity strip width . besides , all of these upper bounds follow directly from a general upper bound about analytic area - preserving twist maps with analytic resonant rics .
let us explain it .
classical and dual billiard maps are exact twist maps defined on an open cylinder when written in suitable coordinates .
exact twist maps have been vastly studied .
they satisfy a lagrangian formulation and their orbits are stationary points of the action functional .
see for instance @xcite .
birkhoff @xcite showed that the minima and minimax points of the @xmath0-periodic action correspond to two different birkhoff @xmath0-periodic orbits of the twist map .
a birkhoff @xmath0-periodic orbit is an orbit such that , after @xmath6 iterates , performs exactly @xmath9 revolutions around the cylinder and its points are ordered in the base @xmath57 as the ones following a rigid rotation of angle @xmath2 .
since there exist at least two different birkhoff @xmath0-periodic orbits , we consider the supremum @xmath58 of the absolute value of the differences of the actions among all of them .
the quantity @xmath58 coincides with @xmath11 and @xmath59 for classical and dual billiards , respectively .
let @xmath60 be the difference of actions between the minimax and minima @xmath0-periodic orbits .
note that @xmath58 is an upper bound of @xmath60 .
mather @xcite used @xmath60 as a criterion to prove the existence of rics of given irrational rotation numbers .
more concretely , he proved that there exists a ric with irrational rotation number @xmath61 if and only if @xmath62 .
another criterion related to the destruction of rics , in this case empirical , was proposed by greene .
the destruction of a ric with diophantine rotation number @xmath61 under perturbation is related to a `` sudden change from stability to instability of the nearby periodic orbits '' @xcite .
the stability of a periodic orbit is measured by the residue .
mackay @xcite proved the criterion in some contexts .
in particular , for an analytic area - preserving twist map , the residue of a sequence of periodic orbits with rotation numbers tending to @xmath61 decays exponentially in a positive power of the distance between rotation numbers , that is , @xmath63 for some @xmath64 .
the same proof leads to a similar exponentially small bound of mather s @xmath60 as @xmath65 .
delshams and de la llave @xcite studied similar problems for analytic area - preserving nontwist maps .
generically rics with a rational rotation number break under perturbation @xcite . nevertheless , there are situations in which some distinguished resonant rics always exist .
see sections [ sec : lengthspectrum ] and [ sec : areaspectrum ] for several examples related to billiard and dual billiard maps .
let us assume that we have an analytic exact twist map with a @xmath1-resonant ric .
that is , a ric whose points are @xmath1-periodic . then there exist some variables @xmath66 in which the resonant ric is located at @xmath67 and the @xmath4-th power of the exact twist map is a small perturbation of the integrable twist map @xmath68 . to be precise , it has the form @xmath69 since the @xmath4-th power map is real analytic , it can be extended to a complex domain of the form @xmath70 the quantity @xmath71 plays a more important role than @xmath72 . to be precise ,
we have proved that , once fixed any @xmath73 and @xmath54 , there exist @xmath74 such that @xmath75 for any relatively prime integers @xmath9 and @xmath6 such that @xmath76 and @xmath77 .
see theorem [ thm : asymptotic ] for a more detailed statement .
this upper bound is optimal because @xmath73 .
that is , the exponent @xmath32 can be taken as close to the analyticity strip width @xmath71 as desired .
the constant @xmath78 may explode when @xmath32 tends to @xmath71 , so , in general , we can not take @xmath79 .
a similar optimal exponentially small upper bound was obtained in @xcite in the setting of the splitting of separatrices of weakly hyperbolic fixed points of analytical area - preserving maps .
the proof of this optimal bound adds some extra technicalities , but we feel that the effort is worth it .
the proof is based on two facts .
first , we write the @xmath4-th power of the exact twist map as the integrable twist map @xmath68 plus an exponentially small remainder on the distance to the ric . see theorem [ thm : remainder ] .
the size of the remainder is reduced by performing a finite sequence of changes of variables , but the number of such changes increases when we approach to the resonant ric .
this is a classical neishtadt - like argument @xcite .
second , we apply the mackay - meiss - percival action principle @xcite , in which the difference of actions of @xmath0-periodic actions is interpreted as an area on the phase space .
the structure of the paper is the following .
section [ sec : maintheorems ] is devoted to state our results in the general context of analytic exact twist maps . in sections [ sec :
lengthspectrum ] and [ sec : areaspectrum ] , we present the different billiard maps and show how the results in section [ sec : maintheorems ] apply .
sections [ sec : proofremainder ] and [ sec : proofthmasymptotic ] contain the technical proofs .
let us introduce some notions about exact twist maps defined on open cylinders .
for a general background on these maps , we refer to @xcite and @xcite .
let @xmath80 and @xmath81 , for some @xmath82 .
we will use the coordinates @xmath83 for both @xmath84 and its universal cover @xmath85 .
we refer to @xmath18 as the angular coordinate .
let @xmath86 be an _ exact symplectic form _ on the open cylinder @xmath84 such that @xmath87 and @xmath88 for some smooth function @xmath89 .
in particular , @xmath90 . if @xmath91 is a real - valued smooth function , @xmath92 denotes the derivative with respect to the @xmath93-th variable .
a smooth diffeomorphism @xmath94 is an _ exact twist map _ when it preserves the exact symplectic form @xmath86 , has zero flux , and satisfies the classical _ twist condition _
@xmath95 , where @xmath96 is a lift of @xmath97 .
henceforth , the exact symplectic form @xmath98 and the lift @xmath99 remain fixed .
we will assume that @xmath100 is analytic when we deal with analytic maps . if @xmath97 preserves @xmath101 and @xmath102 , then @xmath103 , @xmath104 , and @xmath97 preserves the canonical area in the _ global darboux coordinates _
it is worth to remark that certain billiard maps can be analytically extended to the boundaries of their phase spaces in @xmath83 variables , but not in global darboux coordinates . for this reason we consider the coordinates @xmath83 and the above exact symplectic forms .
in fact , we could deal with any exact symplectic form , but we do not need it for the problems we have in mind .
[ def : intersectionproperty ] let @xmath106 be a continuous map .
the map @xmath97 has the _ intersection property _ on @xmath84 if the image of any closed homotopically non trivial loop of the cylinder @xmath84 intersects the loop .
the intersection property is preserved under global changes of variables .
a _ rotational invariant curve ( ric ) _ of @xmath97 is a closed loop @xmath107 homotopically non trivial such that @xmath108 .
let @xmath109 be a ric of @xmath97 .
let @xmath110 and @xmath4 be two relatively prime integers such that @xmath111 .
we say that @xmath109 is @xmath1-_resonant _ when @xmath112 for all @xmath113 , and we say that @xmath109 is _ analytic _ when @xmath114 for some analytic function @xmath115 .
if @xmath94 is a diffeomorphism preserving the exact symplectic form @xmath116 and has a ric , both the zero flux condition and the intersection property are automatically satisfied .
let us study the dynamics of an analytic exact twist map in a neighbourhood of an analytic @xmath1-resonant ric .
first , we note that all points on a @xmath1-resonant ric of @xmath97 remain fixed under the power map @xmath117 .
second , we adapt a classical lemma that appears in several papers about billiards @xcite to our setting .
[ lem : order2 ] if @xmath106 is an analytic exact twist map with an analytic @xmath1-resonant ric @xmath107 , then there exist an analytic strip width @xmath118 , an analytic radius @xmath119 , and some analytic coordinates @xmath66 such that @xmath120 and the power map @xmath117 satisfies the following properties : 1 .
it is real analytic on @xmath121 and can be analytically extended to the complex domain @xmath122 2 .
it has the intersection property on the cylinder @xmath121 ; and 3 .
it has the form @xmath123 , with @xmath124 if @xmath125 and @xmath126 , then @xmath127 and @xmath128 , with @xmath129 for some real analytic @xmath33-periodic functions @xmath130 and @xmath131 .
the twist condition on the ric implies that @xmath130 is positive , since any power of a twist map is locally twist on its smooth rics ( * ? ? ?
* lemma 2.1 ) .
the preservation of @xmath101 implies that @xmath132 , where @xmath133 .
next , we consider the analytic coordinates @xmath66 defined by @xmath134 the constant @xmath135 has been determined in such a way that the new angular coordinate @xmath136 is defined modulus one : @xmath137 .
clearly , @xmath120 .
thus , the coordinates @xmath66 cover an open set containing @xmath138 , since they are defined in a neighbourhood of @xmath109 .
in particular , @xmath139 can be analytically extended to the complex domain @xmath140 for some @xmath141 .
besides , @xmath139 has the intersection property on @xmath121 because the integral of the area form @xmath101 over the region enclosed between the ric @xmath109 and any closed homotopically non trivial loop should be preserved .
finally , a straightforward computation shows that @xmath139 has the form ( [ eq : order2 ] ) in the coordinates @xmath66 .
lemma [ lem : order2 ] has , at a first glance , a narrow scope of application because resonant rics are destroyed under generic perturbations .
however , the boundaries of the cylinder can be considered resonant rics of the extended twist map in many interesting examples .
let us precise this idea .
let @xmath94 be a continuous map .
if @xmath142 is finite , we say that @xmath143 is a _
rigid rotation boundary _ when @xmath97 can be continuously extended to @xmath144 and its extended lift satisfies that @xmath145 for some _ boundary rotation number _ @xmath146 , and we say that @xmath147 is a @xmath1-_resonant boundary _ when @xmath148 .
let @xmath94 be a smooth diffeomorphism .
we say that the _ twist condition holds on the boundary _
@xmath147 when @xmath97 can be smoothly extended to @xmath144 and its extended lift satisfies that @xmath149 for all @xmath150 .
analogous definitions can be written for the upper boundary @xmath151 when @xmath152 is finite .
we recall that @xmath96 is a lift of @xmath97 .
next , we present a version of lemma [ lem : order2 ] for the boundaries of the cylinder .
the only remarkable difference is that the exact symplectic form @xmath98 may vanish on the boundaries .
[ lem : order2_boundaries ] if @xmath106 is an analytic exact twist map such that @xmath142 is finite , @xmath153 is a @xmath1-resonant boundary , @xmath97 can be analytically extended to @xmath144 , @xmath100 can be analytically extended to @xmath154 , and the twist condition holds on @xmath147 , then there exist an analytic strip width @xmath118 , an analytic radius @xmath119 , and some analytic coordinates @xmath66 such that @xmath155 and the power map @xmath117 satisfies the properties ( i)(iii ) given in lemma [ lem : order2 ] .
an analogous result holds for the upper boundary @xmath151 when @xmath152 is finite . if @xmath156 , then @xmath157 and the extended power map @xmath128 has the form ( [ eq : varphi_psi ] ) for some real analytic @xmath33-periodic functions @xmath130 and @xmath131 .
the twist condition on @xmath147 implies that @xmath130 is positive .
since @xmath100 is analytic in @xmath154 and @xmath158 is positive in @xmath159 , we deduce that @xmath160 for some @xmath161 and some integer @xmath162 .
the preservation of @xmath101 implies that @xmath163 .
next , we consider the analytic coordinates @xmath66 defined by @xmath164 where the constant @xmath135 is determined in such a way that the angular coordinate @xmath136 is defined modulus one .
the rest of the proof follows the same lines as in lemma [ lem : order2 ] .
we just note that the intersection property on a ( real ) neighbourhood of the boundary @xmath147 follows by analytic extension , since @xmath139 preserves the exact symplectic form @xmath101 for negative values of @xmath165 too .
we unify the resonant rics studied in lemma [ lem : order2 ] and the resonant boundaries studied in lemma [ lem : order2_boundaries ] as a single object for the sake of brevity .
[ defi : c ] let @xmath94 be an analytic map .
let @xmath110 and @xmath166 be two relatively prime integers .
let @xmath167 and @xmath119 .
we say that @xmath107 ( respectively , @xmath168 for a finite @xmath142 or @xmath169 for a finite @xmath152 ) is a _
@xmath170-analytic @xmath1-resonant ric _ when @xmath109 is an analytic @xmath1-resonant ric ( respectively , @xmath1-resonant boundary ) and there exist some coordinates @xmath66 such that @xmath171 and the power map @xmath117 satisfies the properties ( i)(iii ) stated in lemma [ lem : order2 ] .
the map ( [ eq : order2 ] ) can be viewed as a perturbation of the integrable twist map @xmath172 we want to reduce the size of the nonintegrable terms @xmath173 and @xmath174 as much as possible .
we can reduce them through normal form steps up to any desired order ; see lemma [ lem : beyondallorders ] .
thus , the nonintegrable part of the dynamics is beyond all order in @xmath175 ; that is , in the distance to the resonant ric ( or resonant boundary ) .
a general principle in conservative dynamical systems states that beyond all order phenomena are often exponentially small in the analytic category .
our goal is to write the map as an exponentially small perturbation in @xmath175 of the integrable twist map ( [ eq : integrablemap ] ) .
the final result is stated in the following theorem .
[ thm : remainder ] let @xmath176 be an analytic map with a @xmath170-analytic ric @xmath109 of fixed points such that @xmath107 , @xmath168 for a finite @xmath142 , or @xmath169 for a finite @xmath152 .
let @xmath177 be an arbitrary order .
let @xmath73 .
there exist constants @xmath178 and @xmath179 such that , if @xmath180 , then there exists an analytic change of variables @xmath181 such that : 1 .
it is uniformly ( with respect to @xmath182 ) close to the identity on @xmath183 .
that is , @xmath184 = 1 + \or(\eta),\ ] ] for all @xmath185 , where the @xmath186 and @xmath187 terms are uniform in @xmath182 ; and 2 .
the transformed map @xmath188 is real analytic and has the intersection property on the cylinder @xmath183 . besides , it has the form @xmath189 where @xmath190 and @xmath191 for all @xmath192 .
the proof can be found in section [ sec : proofremainder ] . consider a perturbed hamiltonian system which is close to an integrable system .
it is known that , under the appropriate nondegeneracy conditions , the measure of the set of tori which decompose under the perturbation can be bounded from above by a quantity of order @xmath193 , @xmath194 being the perturbation parameter @xcite .
neishtadt @xcite also considered a context where the perturbation becomes exponentially small in some parameter @xmath194 and hence the measure of the complementary set which is cut out from phase space by the invariant tori is of order @xmath195 , @xmath196 being a positive constant .
this argument could be applied to our context .
first , any neighbourhood of an analytic resonant ric ( or resonant boundary ) of an analytic exact twist map contains infinitely many rics .
second , the area of the complementary of the rics in any of such neighbourhoods is exponentially small in the size of the neighbourhood .
third , the gaps between the rics are exponentially small in their distance to the resonant ric .
the first result follows from the classical moser twist theorem @xcite .
the others follow from the ideas explained above .
let @xmath116 , with @xmath87 , be a fixed exact symplectic form on the open cylinder @xmath84 , with @xmath197 .
let @xmath106 be an exact twist map that can be extended as rigid rotations of angles @xmath198 and @xmath199 to the boundaries @xmath153 and @xmath151 , respectively .
we know that @xmath200 from the twist condition .
let @xmath201 be a fixed lift of @xmath97 .
let @xmath202 .
then there exists a function @xmath203 , determined modulo an additive constant , such that @xmath204 the function @xmath91 is called the _
lagrangian _ or _ generating function _ of @xmath97 .
let @xmath9 and @xmath6 be two relatively prime integers such that @xmath205 and @xmath206 .
a point @xmath207 is @xmath0__-periodic _ _ when @xmath208 .
the corresponding point @xmath209 is a periodic point of period @xmath6 by @xmath97 that is translated @xmath9 units in the base by the lift .
a @xmath0-periodic orbit is _ birkhoff _ when it is ordered around the cylinder in the same way that the orbits of the rigid rotation of angle @xmath2 .
see @xcite for details .
the poincar - birkhoff theorem states that there exist at least two different birkhoff @xmath0-periodic orbits @xcite .
let @xmath210 be a @xmath0-periodic orbit .
its @xmath0__-periodic action _ _ is @xmath211 = h(s_0,s_1 ) + h(s_1,s_2 ) + \cdots + h(s_{q-1},s_0 + p).\ ] ] our goal is to establish an exponentially small bound for the non - negative quantity @xmath212 - w^{(p , q)}[o ] \right|,\ ] ] where @xmath213 denotes the set of all birkhoff @xmath0-periodic orbits of the exact twist map @xmath97 .
the difference of @xmath0-periodic actions can be interpreted as the @xmath101-area of certain domains .
let us explain it .
let @xmath214 and @xmath215 be two @xmath0-periodic orbits .
we can assume , without loss of generality , that @xmath216 .
let @xmath217 be a curve from @xmath218 to @xmath219 contained in @xmath85 .
set @xmath220 .
the curves @xmath217 and @xmath221 have the same endpoints in @xmath84 .
let us assume that these two curves have no topological crossing on the cylinder @xmath84 and let @xmath222 be the domain enclosed between them .
observe that @xmath223 .
hence , @xmath224 , where the sign @xmath225 depends on the orientation of the closed path @xmath226 , but we do not need it , because we take absolute values in both sides of the previous relation : @xmath227 - w^{(p , q)}[o ] \right| = \int_b \omega = : { \mathop{\rm area}\nolimits}_\omega [ b].\ ] ]
these arguments go back to the _ mackay - meiss - percival action principle _ @xcite . if the curves @xmath217 and @xmath228 have some topological crossing , then the domain @xmath229 has several connected components , in which case @xmath230 - w^{(p , q)}[o ] \right| \le \int_b \omega = : { \mathop{\rm area}\nolimits}_\omega [ b]$ ] , because the sign in front of the integral of the area form @xmath101 depends on the connected component .
if the analytic exact twist map @xmath97 has a @xmath1-resonant ric , then @xmath231 indeed , we can take a segment of the ric as the curve @xmath217 used in the previous construction in such a way that @xmath232 and @xmath233 = 0 $ ] .
it turns out that the differences of @xmath0-periodic actions of @xmath97 are exponentially small when @xmath2 is `` sufficiently close '' to @xmath3 .
the meaning of `` sufficiently close '' is clarified in the following theorem .
see also remark [ rem : sufficientlyclose ] .
[ thm : asymptotic ] let @xmath234 be an analytic exact twist map that can be extended as rigid rotations of angles @xmath198 and @xmath199 to the boundaries @xmath153 and @xmath151 , respectively .
we also assume that @xmath97 has a @xmath170-analytic @xmath1-resonant ric @xmath107 , @xmath168 for a finite @xmath142 , or @xmath169 for a finite @xmath152 .
let @xmath73 and @xmath54 .
there exist @xmath235 such that @xmath236 for all relatively prime integers @xmath9 and @xmath6 with @xmath76 and @xmath31 .
the proof has been placed at section [ sec : proofthmasymptotic ] .
[ rem : tendingtoboundaryrics ] if @xmath168 ( respectively , @xmath169 ) , the bound ( [ eq : deltabound ] ) only makes sense as @xmath2 tends to the boundary rotation number @xmath148 ( respectively , @xmath237 ) from the right : @xmath238 ( respectively , from the left : @xmath239 ) .
we will see several examples of this situation in sections [ sec : lengthspectrum ] and [ sec : areaspectrum ] .
[ rem : infinitevalues ] infinite values of @xmath240 can also be dealt with .
for instance , if @xmath241 , then we consider the coordinate @xmath242 and we assume that : 1 ) @xmath243 is a @xmath1-resonant boundary ; 2 ) @xmath97 can be analytically extended to @xmath244 ; 3 ) @xmath245 with @xmath246 for some real number @xmath247 and some integer @xmath248 ; 4 ) @xmath117 has the form ( [ eq : varphi_psi ] ) for some real analytic @xmath33-periodic functions @xmath130 and @xmath131 ; and 5 ) @xmath130 is negative .
then the change @xmath249 transforms the infinite setup into a finite one , so theorem [ thm : asymptotic ] still holds .
we just mention two subtle points .
first , the change ( [ eq : analyticcoordinates_xy ] ) is not well - defined when @xmath250 .
thus , @xmath251 can have a pole at @xmath252 , provided it is not a double one .
second , @xmath130 is negative because the change @xmath249 reverses the orientation , and so , it changes the sign of the twist .
[ rem : sufficientlyclose ] condition @xmath253 implies that @xmath254 as @xmath12 . in many applications , @xmath109 is a ric of fixed points ; that is , a @xmath255-resonant ric .
then theorem [ thm : asymptotic ] implies that @xmath58 is exponentially small in the period @xmath6 when @xmath9 remains uniformly bounded . to be precise , if @xmath73 and @xmath54 , there exist @xmath235 such that @xmath256 for all relatively prime integers @xmath9 and @xmath6 with @xmath77 and @xmath257 .
we recall some well - known results about billiards that can be found in @xcite .
let @xmath7 be a closed strictly convex curve in the euclidean plane @xmath258 .
we assume , without loss of generality , that @xmath7 has length one .
let @xmath150 be an arc - length parameter on @xmath7 .
set @xmath259 .
let @xmath176 , @xmath260 , be the map that models the billiard dynamics inside @xmath7 using the _ birkhoff coordinates _
@xmath83 , where @xmath150 determines the impact point on the curve , and @xmath261 denotes the angle of incidence .
the map @xmath139 preserves the exact symplectic form @xmath262 and has the intersection property .
indeed , @xmath263 is an exact twist map with boundary rotation numbers @xmath264 and @xmath265 . besides , its lagrangian is the distance between consecutive impact points .
finally , @xmath139 is analytic when @xmath7 is analytic .
any @xmath0-periodic orbit on the billiard map forms a a closed inscribed polygon with @xmath6 sides that makes @xmath9 turns inside @xmath7 .
since the lagrangian of the billiard map is the distance between consecutive impact points , the periodic action of a periodic orbit is just the total length of the corresponding polygon .
therefore , the supremum action difference among @xmath0-periodic billiard orbits is the supremum length difference among inscribed billiard @xmath0-polygons .
let us check that the boundary @xmath266 satisfies the hypotheses stated in theorem [ thm : asymptotic ] .
[ pro : classicalextensiontoboundary ] if @xmath139 is the billiard map associated to an analytic strictly convex curve @xmath7 , the boundary @xmath147 is a @xmath170-analytic @xmath255-resonant ric of @xmath139 for some @xmath118 and @xmath119 .
the lower boundary @xmath147 is a @xmath255-resonant because @xmath264 .
clearly , @xmath267 is finite and the function @xmath268 can be analytically extended to @xmath269 .
hence , in order to end the proof we only need to prove that the billiard map @xmath139 can be analytically extended to @xmath270 and the twist condition holds on @xmath147 ; see lemma [ lem : order2_boundaries ] .
let @xmath271 , @xmath272 , and @xmath273 be an arc - length parametrization , the curvature , and a unit normal vector of the analytic strictly convex curve linked by the relation @xmath274 .
we write the angle of incidence as a function of consecutive impact points : @xmath275 . if @xmath276 , then @xmath277 is analytic , @xmath278 , and @xmath279 .
the last property follows from the twist character of the billiard map .
let us study what happens on @xmath147 ; or , equivalently , what happens in the limit @xmath280 .
we note that @xmath281 is the angle between @xmath282 and @xmath283 .
hence , @xmath284 this expression shows that @xmath277 is analytic when @xmath285 and @xmath286 . besides , @xmath287 for all @xmath150 , which implies that @xmath288 is an analytic diffeomorphism that maps a neighbourhood of the diagonal in @xmath289 to a neighborhood of the lower boundary @xmath266 .
next , we write the billiard map @xmath260 as the composition of three maps : @xmath290 we have already seen that the first map is analytic in a neighborhood of @xmath147 . clearly , the second map is analytic .
the third one is also analytic , because @xmath291 by definition of billiard map .
thus , @xmath139 can be analytically extended to @xmath270 .
the twist condition on @xmath147 follows from inequality ( [ eq : twistboundarybilliard ] ) .
we can compare the coordinates ( [ eq : analyticcoordinates_xy ] ) with the coordinates defined by lazutkin in @xcite .
we note that @xmath292 , so @xmath293 in the change ( [ eq : analyticcoordinates_xy ] ) .
if @xmath294 is the radius of curvature of @xmath7 , then the taylor expansion around @xmath295 of the billiard map is @xmath296 from this taylor expansion , lazutkin deduced that the billiard map takes the form @xmath297 in the analytic _ lazutkin coordinates _
@xmath66 defined by @xmath298 the constant @xmath135 has , once more , been determined in such a way that the new angular coordinate @xmath136 is defined modulus one .
lazutkin s results are more refined , he wrote the billiard map as a smaller perturbation of the integrable twist map @xmath68 , but we do not need it . by direct application of proposition [ pro : classicalextensiontoboundary ] and theorem [ thm : asymptotic ] , we get the following exponentially small upper bound of the quantities @xmath11 defined in the introduction .
[ thm : classicalboundary ] let @xmath7 be an analytic strictly convex curve in the euclidean plane .
let @xmath118 be the analyticity strip width of the lower boundary @xmath147 .
let @xmath73 and @xmath54 .
there exist a constant @xmath178 and a period @xmath299 such that @xmath300 for all relatively prime integers @xmath9 and @xmath6 with @xmath77 and @xmath301 .
the same exponentially small upper bound holds for analytic geodesically strictly convex curves on surfaces of constant curvature , where the billiard trajectories are just broken geodesics .
billiard maps on the klein model of the hyperbolic plane @xmath302 and on the positive hemisphere @xmath303 have been studied , for instance , in @xcite , where it is shown that they are exact twist maps with the same boundary rotation numbers as in the euclidean case .
therefore , by local isometry arguments , we can write a version of theorem [ thm : classicalboundary ] on any surface of constant curvature . a smooth closed convex curve is of _ constant width _ if and only if it has a chord in any direction perpendicular to the curve at both ends .
billiards inside convex curves of constant width have a nice property @xcite .
let us explain it .
the billiard map associated to a smooth convex curve of constant width has the horizontal line @xmath304 as a resonant @xmath28-ric .
any trajectory belonging to that ric is orthogonal to the curve at its two endpoints . due to the variational formulation ,
all the @xmath28-periodic orbits are extrema of the @xmath28-periodic action and , thus , all @xmath28-periodic trajectories have the same length , which is the reason we refer to them as constant width curves .
another characterization of constant width curves is the following .
we reparametrize the curve by using the angle @xmath305 between the tangent vector at a point in the curve and some fixed line .
let @xmath306 be the radius of curvature at this point .
the curve has constant width if and only if the fourier series of @xmath306 contains no other even coefficients than the constant term .
thus , the space of analytic constant width curves has infinite dimension and codimension .
theorem [ thm : asymptotic ] can be directly applied in this context and we have the following result .
let @xmath94 be the billiard map of an analytic strictly convex curve of constant width .
let @xmath118 be the analyticity strip width of the @xmath28-resonant ric of @xmath97 .
let @xmath73 and @xmath54 .
there exist a constant @xmath307 and a period @xmath299 such that @xmath308 for all relatively prime integers @xmath9 and @xmath6 such that @xmath56 and @xmath31 .
one could try to generalize constant width billiards , where @xmath309 is a @xmath28-resonant ric , to _ constant angle tables _
, where @xmath310 is assumed to be a @xmath1-resonant ric . however , the only table such that @xmath310 is a @xmath1-resonant ric , with @xmath311 , is the circle .
see @xcite . by the way , theorem [ thm : asymptotic ] applies to this case but , since the circular billiard is integrable , @xmath312 , for all @xmath1 .
in fact , the circular billiard map is globally conjugated to the integrable twist map ( [ eq : integrablemap ] ) .
there are more billiard tables with resonant rics , but their rics are not horizontal . for instance , the elliptic table has all possible @xmath1-resonant rics , but the @xmath28-resonant one . hence , in this case , @xmath313 .
baryshnikov and zharnitsky @xcite proved that an ellipse can be infinitesimally perturbed so that any chosen resonant ric will persist .
innami @xcite found a condition on the billiard table that guarantees the existence of a @xmath314-resonant ric .
however , theorem [ thm : asymptotic ] can not be applied in such cases , because both the baryshnikov - zharnitsky and the innami constructions are done in the smooth category , where we can only claim that @xmath11 is beyond all order in the difference between rotation numbers .
we recall some well - known facts about dual billiards that can be found in @xcite .
let @xmath7 be a strictly convex closed curve in the euclidean plane @xmath258 .
let @xmath39 be the unbounded component of @xmath315 .
the dual billiard map @xmath41 is defined as follows : @xmath42 is the reflection of @xmath43 in the tangency point of the oriented tangent line to @xmath7 through @xmath43 .
this map is area - preserving .
next , we introduce the _ envelope coordinates
_ @xmath316 of a point @xmath317 . in this section , @xmath318 and @xmath319 .
we recall that @xmath320 .
given a point @xmath317 , let @xmath321 be the angle made by the positive tangent line to @xmath7 in the direction of @xmath43 with a fixed direction of the plane , and let @xmath261 be the distance along this line from @xmath7 to @xmath43 .
see figure [ fig : dualenvelopecoordinates ] . and the dual billiard map @xmath322.,scaledwidth=80.0% ] the dual billiard map preserves the exact symplectic form @xmath323 , and it has the intersection property .
indeed , @xmath324 , @xmath325 , is an exact twist map with boundary rotation numbers @xmath264 and @xmath326 .
its lagrangian is the area enclosed by @xmath7 and the tangent lines through the points on @xmath7 with coordinates @xmath32 and @xmath327 . any @xmath0-periodic orbit on the dual billiard map forms a closed circumscribed polygon with @xmath6 sides that makes @xmath9 turns outside @xmath7 .
since the lagrangian of the dual billiard map is the above - mentioned area , the periodic action of a periodic orbit is just the area enclosed between the corresponding polygon and @xmath7 , taking into account some multiplicities when @xmath44 .
therefore , the supremum action difference among @xmath0-periodic dual billiard orbits is the supremum area difference among circumscribed dual billiard @xmath0-polygons .
we note that @xmath328 when the point @xmath317 approaches to the curve @xmath7 .
therefore , in order to study the dual billiard dynamics close to @xmath7 , we must study the dual billiard map @xmath139 in a neighbourhood of the lower boundary @xmath329 of @xmath330 .
let us check that @xmath147 satisfies the hypotheses stated in theorem [ thm : asymptotic ] .
[ pro : dualextensiontoboundary ] if @xmath139 is the dual billiard map associated to an analytic strictly convex curve @xmath7 , the lower boundary @xmath147 is an @xmath170-analytic @xmath255-resonant ric of @xmath139 for some @xmath118 and @xmath119 .
the lower boundary @xmath147 is a @xmath255-resonant because @xmath264 .
clearly , @xmath267 is finite and the function @xmath331 can be analytically extended to @xmath332 .
hence , in order to end the proof we only need to prove that the dual billiard map @xmath139 can be analytically extended to @xmath333 and the twist condition holds on @xmath147 ; see lemma [ lem : order2_boundaries ] .
we write the distance @xmath281 as a function of consecutive tangent points : @xmath334 .
we know that @xmath335 is analytic , @xmath336 , and @xmath337 when @xmath338 . and @xmath339 are parallel , so @xmath340 . ]
the last property follows from the twist character of the dual billiard map .
let us study what happens on @xmath147 ; or , equivalently , what happens in the limit @xmath341 .
let @xmath342 be the radius of curvature of @xmath7 in the angular coordinate @xmath32 .
set @xmath343 . from boyland @xcite
, we know that @xmath344 this expression shows that @xmath335 is analytic when @xmath345 and @xmath346 . by derivating relation ( [ eq : rfromalphas ] ) with respect to @xmath327 , we get that @xmath347 for all @xmath321 , which implies that @xmath348 is an analytic diffeomorphism that maps a neighbourhood of the diagonal in @xmath349 to a neighborhood of the lower boundary @xmath329 .
next , we write the billiard map @xmath325 as the composition of three analytic maps : @xmath350 thus , @xmath139 can be analytically extended to @xmath333 .
the twist condition on @xmath147 follows from inequality ( [ eq : twistboundarydualbilliard ] ) .
we can compare the coordinates ( [ eq : analyticcoordinates_xy ] ) with the coordinates defined by tabachnikov in @xcite .
we note that @xmath351 , so @xmath293 in the change ( [ eq : analyticcoordinates_xy ] ) .
if @xmath352 is the curvature of @xmath7 , then the taylor expansion around @xmath295 of the dual billiard map is @xmath353 ( tabachnikov wrote this taylor expansion in coordinates @xmath83 , where @xmath18 is an arc - length parameter , but his result can be easily adapted . ) from this taylor expansion , one deduces that the dual billiard map takes the form ( [ eq : order2 ] ) in the analytic _ tabachnikov coordinates _
@xmath354 the constant @xmath135 has been determined in such a way that @xmath136 is defined modulus one : @xmath137 .
we get the following exponentially small upper bound of the quantities @xmath59 defined in the introduction by direct application of proposition [ pro : dualextensiontoboundary ] and theorem [ thm : asymptotic ] .
[ thm : dualboundary ] let @xmath7 be an analytic strictly convex curve in the euclidean plane . let @xmath118 be the analyticity strip width of the lower boundary @xmath147 .
let @xmath73 and @xmath54 .
there exist a constant @xmath178 and a period @xmath299 such that @xmath355 for all relatively prime integers @xmath9 and @xmath6 with @xmath77 and @xmath301 .
tabachnikov @xcite studied the dual billiard map in the hyperbolic plane @xmath302 , and extended the asymptotic expansion ( [ eq : asymptoticarea ] ) to that new setting .
he also claimed that there exists an analogous formula for dual billiards on the unit sphere @xmath356 .
therefore , by local isometry arguments , we can write a version of theorem [ thm : dualboundary ] on any surface of constant curvature . we note that @xmath357 when the point @xmath317 moves away from the curve @xmath7 .
we use the coordinates @xmath358 to work at infinity , where @xmath242 and @xmath316 are the coordinates introduced in subsection [ ssec : dualbilliards ] .
the exact symplectic form @xmath359 becomes @xmath360 in coordinates @xmath358 .
tabachnikov @xcite realized that the dual billiard map at infinity can be seen as a map defined in a neighbourhood of the @xmath28-resonant ric @xmath361 . to be precise
, he saw that the dual billiard map can be analytically extended to @xmath244 , its square has the form @xmath362 for some real analytic @xmath33-periodic functions @xmath363 and @xmath364 , and @xmath363 is negative .
hence , the following result is deduced from remark [ rem : infinitevalues ] .
[ thm : dualinfinity ] let @xmath7 be an analytic strictly convex curve in the euclidean plane .
let @xmath118 be the analyticity strip width of the boundary @xmath365 .
let @xmath73 and @xmath54 .
there exist a constant @xmath178 and a period @xmath299 such that @xmath366 for all relatively prime integers @xmath9 and @xmath6 such that @xmath56 and @xmath31 .
let @xmath367 , with @xmath368 and @xmath369 , be the space of all analytic functions @xmath97 defined on the open set @xmath370 where @xmath371 denotes the @xmath135-th fourier coefficient of the @xmath33-periodic function @xmath372 and @xmath373 denotes its sup - norm on the complex open ball @xmath374 .
let @xmath375 be the sup - norm on @xmath376 .
let @xmath377 be the space of all vectorial functions @xmath378 of the form @xmath379 such that @xmath380 .
the space @xmath381 is a banach space with the fourier norm @xmath382 the sup - norm @xmath383 on @xmath381 is defined analogously .
let @xmath384 be the average of @xmath385 .
let @xmath386 be the direct decomposition where @xmath387 is the vectorial subspace of the elements of the form @xmath388 , whereas @xmath389 is the one of the elements with @xmath390 .
let @xmath391 and @xmath392 be the associated projections .
thus , any @xmath393 can be decomposed as @xmath394 , where @xmath395 obviously , @xmath396 and @xmath397 we will always denote the scalar functions in @xmath367 with lower - case letters , and the vectorial functions in @xmath377 with upper - case letters .
asterisk and bullet superscripts in upper - case letters stand for the @xmath398-projections and @xmath399-projections of vectorial functions in @xmath377 , respectively .
asterisk superscripts in lower - case letters denote averages of scalar functions in @xmath367 .
we will always write the couple of scalar functions associated to any given vectorial function of @xmath377 with the corresponding lower - case letter and the subscripts @xmath400 .
hat symbols denote fourier coefficients .
henceforth , let @xmath401 be the integrable twist map introduced in ( [ eq : integrablemap ] ) .
let @xmath402 be a map satisfying the properties listed in lemma [ lem : order2 ] , so @xmath403 for some @xmath404 , where @xmath405 is the analyticity strip width in the angular variable @xmath136 , and @xmath406 is the analyticity radius in @xmath175 .
hence , @xmath407 is a perturbation of @xmath408 of order two .
the following lemma allows us to increase that order as much as we want by simply losing as little analyticity strip width as we want .
it is based on classical averaging methods . in particular
, we see that @xmath409 is a perturbation beyond all order of @xmath408 .
[ lem : beyondallorders ] let @xmath403 , with @xmath404 and @xmath410 and @xmath411 , be a real analytic map with the intersection property on the cylinder @xmath412 .
let @xmath413 be an integer .
let @xmath414 be any analyticity strip width such that @xmath415 .
there exist an analyticity radius @xmath416 and a change of variables of the form @xmath417 for some @xmath418 such that the transformed map @xmath419 is real analytic , has the intersection property on the cylinder @xmath420 , and has the form @xmath421 for some @xmath422 .
besides , the change of variables @xmath423 is close to the identity on @xmath420 .
that is , @xmath424 = 1 + \or(y),\ ] ] uniformly for all @xmath425 .
the change @xmath423 is the composition of @xmath426 changes of the form @xmath427 where @xmath428 , @xmath429 , @xmath430 is a positive decreasing sequence , and @xmath431 is constructed as follows to increase the order of the perturbation from @xmath432 to @xmath433 . let us suppose that we have a real analytic map with the intersection property on @xmath434 of the form @xmath435 , for some @xmath436 with @xmath437 and @xmath438 .
we begin with a formal computation .
we write @xmath439 where the functions @xmath440 and @xmath441 are 1-periodic and analytic on the open complex strip @xmath442 .
we will see , by using an _ a posteriori _ reasoning , that the intersection property implies that @xmath441 has zero average ; that is , @xmath443 .
nevertheless , we can not prove it yet .
thus , we will keep an eye on @xmath444 in what follows . if we take the change of variables @xmath445 for some functions @xmath446 and @xmath447 , then , after a straightforward computation , the map @xmath448 has the form @xmath449 with @xmath450 and @xmath451 .
therefore , we take @xmath452 so that @xmath453 and @xmath454 . these functions @xmath447 and @xmath446 are 1-periodic , because @xmath455 and @xmath456 have zero average . besides , @xmath446 and @xmath447 are analytic in the open complex strip @xmath457 . indeed ,
@xmath458 with @xmath459 . next , we control the domain of definition of the map @xmath460 .
the inverse change is @xmath461 thus , the maps @xmath462 , @xmath463 , and @xmath464 have the form @xmath465 , since @xmath438 . consequently , if @xmath466 is small enough , then @xmath467 so @xmath468 is well - defined on @xmath469 .
now , let us check that @xmath470 .
at this moment , we only know that @xmath471 since the change of variables has not eliminated the average @xmath444 .
the map @xmath460 has the intersection property on the cylinder @xmath472 , because the intersection property is preserved by changes of variables .
this implies that @xmath470 . on the contrary ,
the image of the loop @xmath473 does not intersect itself when @xmath474 .
finally , properties ( [ eq : changephi_m ] ) follow directly from the fact that we have performed a finite number of changes , all of them satisfying these same properties .
next , the following theorem provides the exponentially small bound for the @xmath399-projection of the residue provided an initial order @xmath110 big enough .
it is the main tool to prove theorem [ thm : remainder ] .
[ thm : technical ] let @xmath475 be an integer , @xmath476 , @xmath477 , and @xmath478 . there exist constants @xmath479 and @xmath480 , @xmath481 , such that , if @xmath482 and @xmath483 then there exists a change of variables @xmath484 satisfying the following properties : 1 .
@xmath485 with @xmath486 ; and 2 .
the transformed map @xmath487 is real analytic , has the intersection property on the cylinder @xmath488 , and has the form @xmath489 , @xmath490 , @xmath491 theorem [ thm : technical ] is proved in subsection [ sec : proofthmtechnical ] . in order to present the main ideas of the proof ,
let us try to completely get rid of the remainder of the map of the form @xmath492 , for some @xmath393 , with a change of variables of the form @xmath493 , for some @xmath494 . concretely
, we look for @xmath495 such that @xmath496 , or , equivalently , we look for @xmath497 such that @xmath498 it is not possible to solve this equation in general . instead , we consider the linear equation @xmath499 this vectorial equation reads as @xmath500 therefore , we need to solve two linear equations of the form @xmath501 where @xmath502 is known . if the average of @xmath503 is different from zero : @xmath504 , then this equation can not be solved .
besides , it is a straightforward computation to check that , if @xmath505 , the formal solution of this equation in the fourier basis is @xmath506 whereas the zero - th coefficient @xmath507 can be chosen arbitrarily . from ( [ eq : fouriercoefficient ] ) , it is clear that ( [ eq : linearsingle ] ) can not be solved unless @xmath97 has only a finite number of harmonics and zero average . for this reason , given a function @xmath503 with zero average , we define its @xmath78__-cut off _ _ as @xmath508 let @xmath78 be such that @xmath509 for all @xmath510 and @xmath511 . hence , we will take @xmath512 for some fixed @xmath513 , and we will actually solve truncated linear equations of the form @xmath514 the fourier norm is specially suited to analyse this kind of equations ; see lemma [ lem : linearsingle ] . summarizing these ideas ,
we look for a change of variables of the form @xmath515 , where @xmath497 satisfies the truncated linear vectorial equation @xmath516 where @xmath517 denotes the @xmath78-cut off of @xmath518 .
the average of the first component of @xmath517 may be non - zero .
equation ( [ eq : truncatedlineardouble ] ) is studied in lemma [ lem : lineardouble ] .
this close to the identity change of variables @xmath515 does not completely eliminate the remainder .
however , if @xmath182 is small enough , it reduces the size of the @xmath399-projection of the remainder as the following lemma shows .
[ lem : iterative ] let @xmath475 be an integer , @xmath476 , @xmath477 , @xmath519 , and @xmath520 . there exist a constant @xmath521 such that if @xmath522 with @xmath523 then there exists a change of variables @xmath524 satisfying the following properties : 1 .
@xmath494 is a solution of the truncated linear equation ( [ eq : truncatedlineardouble ] ) such that @xmath525 where @xmath526 is defined in lemma [ lem : lineardouble ] ; and 2 .
the transformed map @xmath527 is real analytic , has the intersection property on the cylinder @xmath528 , and has the form @xmath529 , @xmath530 , @xmath531 where @xmath532 and @xmath533 . if @xmath532 , then @xmath534 .
the proof of this lemma is found in subsection [ ssec : proofofiterativelemma ] .
some technicalities in the proof require the use of the sup - norm , which forces us to deal with both the fourier norm and the sup - norm .
the relations between them are stated in lemma [ lem : normbounds ] .
finally , theorem [ thm : technical ] is obtained by means of a finite sequence of changes of variables like the ones described in the iterative lemma .
we want to perform as many of such changes as possible because each change reduces the size of the @xmath399-projection of the remainder by the factor @xmath535 .
since the loss of analyticity in the angular variable is @xmath536 , then we can at most perform a number @xmath537 of such changes .
this idea goes back to neishtadt @xcite .
the intersection property is used neither in the proof of the iterative lemma nor in the proof of theorem [ thm : technical ] , but will be essential to control the size of the @xmath398-projections of the remainders in terms of the size of their @xmath399-projections later on .
[ lem : normbounds ] let @xmath538 , @xmath539 , and @xmath502 .
let @xmath540 , with @xmath541 the @xmath78-cut off of @xmath97 , defined in ( [ eq : kcutoff ] ) . then : 1 .
@xmath542 , 2 .
@xmath543 , 3 .
@xmath544 , and 4 . @xmath545 .
if @xmath546 , then these bounds also hold for any vectorial function @xmath393 . first , the fourier norm of @xmath541 is a partial sum of the fourier norm of @xmath97 .
second , @xmath547 .
third , @xmath548 , for all @xmath549 .
fourth , we recall that the fourier coefficients of the analytic function @xmath97 satisfy the inequality @xmath550 for all @xmath551 . hence , @xmath552 where we have used that @xmath553 for all @xmath554 .
the last part follows from the definition of the norms @xmath555 and @xmath383 .
[ lem : linearsingle ] if @xmath556 , @xmath557 , and @xmath502 is a function with zero average , then the truncated linear equation ( [ eq : truncatedlinearsingle ] ) has a unique solution @xmath558 with zero average and @xmath559 , where @xmath560 the fourier coefficients of @xmath561 must satisfy ( [ eq : fouriercoefficient ] ) .
we note that @xmath562 for all @xmath556 , since the function @xmath563 is analytic on the open ball @xmath564 .
moreover , @xmath565 for all @xmath566 . finally , we recall that @xmath567 . then we obtain that @xmath568 .
we will denote by @xmath569 the linear operator that sends the independent term @xmath541 of the truncated linear equation ( [ eq : truncatedlinearsingle ] ) to the solution @xmath561 with zero average .
note that the solution @xmath561 has no harmonics of order @xmath570 .
[ lem : lineardouble ] if @xmath571 , @xmath556 , @xmath557 , and @xmath572 , then the truncated linear equation ( [ eq : truncatedlineardouble ] ) has a solution @xmath573 such that @xmath574 where @xmath575 , @xmath576 , and @xmath577 is defined in ( [ eq : omega ] ) .
let @xmath578 and @xmath579 .
then the vectorial equation @xmath580 reads as @xmath581 let @xmath582 and @xmath583 .
these operations are well - defined since both @xmath584 and @xmath585 have zero average . as for the bounds , @xmath586 where we have used lemma [ lem : linearsingle ]
[ lem : difference ] let @xmath587 , @xmath588 , @xmath589 , @xmath590 , and @xmath591 , such that @xmath592 let @xmath593 .
let @xmath594 .
let @xmath595 with @xmath596 .
let @xmath597 with @xmath598 . then , 1 .
@xmath599 , 2 .
@xmath600 , and 3 .
@xmath601 .
. then @xmath603 and @xmath604 let @xmath605 , with @xmath606 $ ] fixed .
we deduce from conditions ( [ eq : abcd ] ) that @xmath607 and @xmath608 belong to @xmath609 for all @xmath610
. therefore , @xmath611 by convexity of the domain , and so , the composition @xmath612 is well - defined on the domain @xmath613 .
a simple computation shows that the product @xmath614 is equal to @xmath615 let us bound the elements above . on the one hand , @xmath616 , @xmath617 , and @xmath618 for all @xmath610 . on the other hand ,
@xmath619 and the cauchy estimates imply that @xmath620 for all @xmath621 $ ] , where @xmath622 . thus , @xmath623 this proves the first item .
the other items follow from the bounds @xmath624 and @xmath625 , since @xmath626 .
[ lem : inverse ] let @xmath627 , @xmath588 , @xmath589 , and @xmath628 such that conditions ( [ eq : abcd ] ) hold with @xmath629 .
let @xmath515 , with @xmath630 and @xmath631 .
then @xmath632 for all @xmath633 and @xmath634 .
let @xmath635 , where @xmath636 is defined in lemma [ lem : difference ] .
if @xmath637 , then @xmath495 is invertible and the inverse change @xmath638 satisfies the following properties : 1 .
@xmath639 for some @xmath640 such that @xmath641 , 2 .
@xmath642 for all @xmath643 and @xmath644 , and 3 .
@xmath645 . note that @xmath646 .
conditions ( [ eq : abcd ] ) imply that @xmath647 and @xmath648 .
therefore , @xmath632 for all @xmath633 and @xmath634 .
analogously , if @xmath649 and @xmath650 , then @xmath651 for all @xmath643 , @xmath644 .
we denote by @xmath652 the closed ball in @xmath653 of radius @xmath9 in the sup - norm .
let us prove that the functional @xmath654 , @xmath655 , is a well - defined contraction with lipschitz constant @xmath656 first , we observe that @xmath657 so @xmath658 .
second , we bound @xmath659 as follows : @xmath660 the first inequality is direct , and the second comes from lemma [ lem : difference ] with @xmath661 , @xmath662 , @xmath663 , @xmath664 , @xmath665 , and @xmath666 .
this proves that @xmath667 is a contraction with lipschitz constant ( [ eq : lipschitzconstant ] ) .
thus , @xmath667 has a unique fixed point @xmath668 which satisfies that @xmath669 on @xmath670 .
therefore , the inverse map @xmath638 exists and equals @xmath671 .
furthermore , @xmath672 follows from ( [ eq : pnorm ] ) . finally , @xmath673 we have used the second inequality of equation ( [ eq : contraction ] ) with @xmath674 .
we recall that @xmath492 with @xmath675 , @xmath676 , @xmath677 , @xmath678 , and @xmath475 .
let @xmath679 .
let @xmath680 be the constant introduced in lemma [ lem : lineardouble ] and @xmath681 .
let @xmath682 be the change of variables where @xmath683 is the solution given in lemma [ lem : lineardouble ] of the truncated linear equation @xmath684 with @xmath685 , so that @xmath686 for all @xmath687 and @xmath688 .
let @xmath689 be the inverse change studied in lemma [ lem : inverse ] .
let @xmath690 be the transformed map .
let @xmath691 be the new remainder .
henceforth , we will assume that @xmath32 , @xmath182 , and @xmath692 are some positive constants such that @xmath693 we split the proof in four steps .
_ step 1 : control of the domains . _ note that @xmath694 for all @xmath695 and @xmath696 .
indeed , @xmath697 the behaviours of the changes @xmath495 and @xmath638 follow directly from lemma [ lem : inverse ] , which can be applied since conditions ( [ eq : conditions ] ) are more restrictive than the ones required in lemma [ lem : inverse ] when @xmath698 and @xmath699 .
we also need that @xmath700 and @xmath701 in order to control the inverse @xmath638 , which explains the restrictions on @xmath702 and @xmath703 .
the behaviour of the map @xmath704 follows from the bound @xmath705 and conditions @xmath706 and @xmath707 , which are also a consequence of ( [ eq : conditions ] )
. _ step 2 : decomposition of the new remainder . _
it turns out that @xmath708 , where @xmath709 indeed , @xmath710 and @xmath711 , so @xmath712 finally , let @xmath713 and @xmath714 .
_ step 3 : bounds of the projections of the new remainder .
_ lemma [ lem : normbounds ] and the bound ( [ eq : psigbullet ] ) will be used several times in what follows .
below , we apply lemma [ lem : difference ] ( twice ) and lemma [ lem : inverse ] ( once ) .
the required hypotheses in each case are satisfied due to conditions ( [ eq : conditions ] ) . * if @xmath715 and @xmath716 , then @xmath717 * if @xmath718 and @xmath719 , then @xmath720 the first inequality follows from lemma [ lem : difference ] with @xmath721 , @xmath722 , @xmath723 , @xmath724 , @xmath725 , and @xmath699 , so that @xmath726 . * if @xmath727 and @xmath728 , then @xmath729 .
indeed , @xmath730 and @xmath731 we have used that @xmath732 to bound @xmath733 . *
if @xmath718 and @xmath719 , then @xmath734 the first inequality follows from lemma [ lem : difference ] with @xmath735 , @xmath736 , @xmath737 , @xmath738 , @xmath739 , and @xmath699 , so that @xmath740 . * if @xmath741 and @xmath742 , then @xmath743 the second inequality uses the inclusion @xmath744 .
the fourth one follows from lemma [ lem : inverse ] with @xmath745 .
we need to verify the hypothesis @xmath746 in this last lemma .
it turns out that @xmath747 , so it suffices to take @xmath748 , with @xmath749 small enough . * if @xmath741 and @xmath742 , then @xmath750 where @xmath751 and the constants @xmath752 , @xmath753 , have been defined previously .
_ step 4 : choice of the loss of analyticity domain .
_ we set @xmath754 and @xmath755 .
if @xmath756 is small enough , then conditions ( [ eq : conditions ] ) hold for all @xmath748 .
in addition , @xmath757 where @xmath636 is the expression introduced in lemma [ lem : difference ] .
if we take @xmath758 and @xmath759 , then the bounds of the previous step imply that @xmath760 where @xmath761 .
we recall that @xmath762 , @xmath763 , and @xmath685 . hence ,
if @xmath764 and @xmath749 is small enough , then @xmath765 indeed , @xmath766 if we take a small enough value of @xmath749 .
this ends the proof of the iterative lemma .
set @xmath767 , @xmath768 , and @xmath526 , where the function @xmath769 is defined in lemma [ lem : lineardouble ] .
let @xmath749 be the positive constant associated to the integer @xmath475 in lemma [ lem : iterative ] , the numbers @xmath770 , and the exponent @xmath771 .
let @xmath772 , @xmath773 , and @xmath774 .
let us check that @xmath749 , @xmath775 , @xmath776 , and @xmath777 satisfy the properties given in theorem [ thm : technical ] .
let @xmath778 , @xmath779 , @xmath780 , @xmath781 , @xmath782 be the map given in ( [ eq : barfv ] ) , @xmath783 , and @xmath784 . by recursively applying lemma [ lem : iterative ]
, we obtain a sequence of changes of variables @xmath785 , with @xmath786 , and a sequence of maps @xmath787 , with @xmath788 , @xmath789 and @xmath790 , such that @xmath791 with @xmath792 , @xmath793 , @xmath794 , and @xmath795 .
let @xmath796 be the biggest integer satisfying @xmath797 .
the sequences @xmath798 , @xmath799 , and @xmath800 are decreasing .
the sequence @xmath801 is increasing .
indeed , @xmath802 and @xmath803 for all @xmath804
. we can apply @xmath796 times the iterative lemma .
let @xmath805 be the map obtained after those @xmath796 steps .
then @xmath806 finally , let @xmath807 be the change of variables such that @xmath487 .
we want to check that @xmath484 for some @xmath485 such that @xmath486 .
we note that @xmath808 where each term of the above summation is evaluated at a different argument . nevertheless , those arguments are not important when computing the sup - norm : @xmath809 this ends the proof of theorem [ thm : technical ] .
let us begin with a simple , but essential , chain of inequalities associated to certain analyticity strip widths that will appear along the proof .
if @xmath73 , then there exists @xmath478 , @xmath521 , and some analyticity strip widths @xmath810 , @xmath811 , and @xmath812 , such that @xmath813 the first two reductions ( that is , from @xmath71 to @xmath810 and from @xmath810 to @xmath414 ) are as small as we want
. the third reduction ( from @xmath811 to @xmath814 ) should be a little bigger than @xmath32 in order to get the desired exponentially small upper bound with the exponent @xmath32 .
the fourth reduction ( that is , from @xmath812 to @xmath815 ) is also small , since @xmath749 can be taken as small as necessary .
this decreasing positive sequence of analyticity strip widths is associated to a similar sequence of analyticity radii . to be precise
, we will construct a sequence of the form @xmath816 the inequality @xmath817 does not correspond to a true reduction , but to a restriction on the size of @xmath749 .
note that we have consumed all the analyticity strip width after the last reduction , but we still keep a positive analyticity radius @xmath182 .
we split the proof in the eight steps .
_ step 1 : control of the fourier norm .
_ if the analytic map @xmath139 satisfies the properties ( i)(iii ) listed in lemma [ lem : order2 ] , then the map @xmath818 is real analytic and has the intersection property on the cylinder @xmath121 , can be extended to the complex domain @xmath140 , and has the form ( [ eq : order2 ] ) .
the fourier norm @xmath819 may be infinite , but @xmath820 for any @xmath821 and @xmath822 .
_ step 2 : application of the averaging lemma .
once fixed an integer @xmath475 and any @xmath415 , we know from lemma [ lem : beyondallorders ] that there exist an analytical radius @xmath416 and a change of variables of the form @xmath417 for some @xmath418 such that the transformed map @xmath419 is real analytic , has the intersection property on the cylinder @xmath420 , and has the form @xmath421 for some @xmath422 .
_ step 3 : application of theorem [ thm : technical ] .
_ let @xmath478 be the number that appears in ( [ eq : achain ] ) .
set @xmath823 , @xmath811 , and @xmath824 .
let @xmath479 be the constant stated in theorem [ thm : technical ] .
we can assume that @xmath817 and the condition ( [ eq : achain ] ) holds , by taking a smaller @xmath521 if necessary .
let @xmath179 be defined by @xmath825 .
fix any @xmath826 .
set @xmath814 and @xmath827 .
if @xmath828 and @xmath829 are the norms defined in ( [ eq : barfv ] ) , then @xmath830 . hence , we can apply theorem [ thm : technical ] to obtain a change of variables @xmath484 , with @xmath831 and @xmath832 , and a transformed map @xmath833 , with @xmath490 , @xmath834 , and @xmath835 .
_ step 4 : uniform estimates on the change @xmath836_. by construction , @xmath484 , with @xmath485 and @xmath837 , where the constant @xmath838 does not depend on @xmath182 .
thus , @xmath839 for some functions @xmath840 analytic on @xmath841 such that @xmath842 .
the cauchy estimates imply that @xmath843 and , in particular , for all @xmath844 .
hence , @xmath845 = 1 + \or(y^{m-2}),\ ] ] for all @xmath844 , where the @xmath846 , @xmath847 , and @xmath848 terms are uniform in @xmath182 .
we recall that @xmath475 and the change @xmath423 satisfies properties ( [ eq : changephi_m ] ) , so the complete change @xmath836 satisfies the properties stated in theorem [ thm : remainder ] .
_ step 5 : exponentially small bound on the remainder @xmath99 .
_ after all these changes of variables , we have the map @xmath849 , with @xmath850 , @xmath851 and @xmath852 .
we can bound @xmath853 by using the bound on @xmath518 and the intersection property of @xmath409 on the cylinder @xmath183 .
we recall that if @xmath854 for some @xmath380 , then @xmath855 where @xmath856 is the average of @xmath857 and @xmath858 .
fix any @xmath859 .
we know that @xmath860 therefore , there exists @xmath861 such that @xmath862 , and so @xmath863 this implies that @xmath864 for all @xmath185 .
_ step 6 : exponentially small bounds on some derivatives of the remainder .
_ we recall that @xmath865 for all @xmath866 .
thus , we get from @xmath867 and the cauchy estimates that @xmath868 for all @xmath185 . _ step 7 : a crude bound on the derivative of @xmath869 .
_ we recall that @xmath870 , so @xmath871 therefore , the cauchy estimates imply that @xmath872 for all @xmath873 and , in particular , for all @xmath874 .
_ step 8 : computation of the constant @xmath78 . _ by combining the inequalities obtained in steps 57 , we get that @xmath875 and @xmath191 for all @xmath183 , provided @xmath876 this ends the proof of theorem [ thm : remainder ] .
henceforth , let @xmath877 and @xmath878 with @xmath369 .
let @xmath879 be any compact subset of @xmath880 .
let @xmath881 .
let @xmath882 be the set of all matrix functions @xmath883 of the form @xmath884 for some continuous functions @xmath885 .
the set @xmath882 is a banach space with the norm @xmath886 [ lem : boundgammas ] let @xmath887 , @xmath888 , @xmath889 , and @xmath890 .
let @xmath891 and @xmath892 .
let @xmath893 be a map of the form ( [ eq : neishtadtform ] ) with @xmath894 for all @xmath895 .
then : 1 .
@xmath896 and @xmath897 .
2 . @xmath898 and @xmath899 .
3 . @xmath900 and @xmath901 .
4 . @xmath902 and @xmath903 .
it is a straightforward computation .
let @xmath139 be an analytic map of the form ( [ eq : neishtadtform ] ) .
let @xmath9 and @xmath6 be two relatively prime integers . there exist two curves @xmath904 and @xmath905 and two rics @xmath906 with diophantine rotation numbers @xmath907 , all four contained in a small neighborhood of @xmath908 , such that @xmath909 projects @xmath910 onto @xmath911 along the vertical direction and @xmath910 and @xmath911 are contained in the strip of the cylinder enclosed by the rics @xmath912 .
following birkhoff ( * ? ? ?
* section vi ) and arnold ( * ? ? ?
* section 20 ) , all @xmath0-periodic points of @xmath139 are contained in @xmath913 .
besides , we will see later on that the ( geometric ) area enclosed between @xmath910 and @xmath911 is an upper bound of the quantities @xmath58 .
these are the reasons for the study of @xmath910 and @xmath911 .
let us prove that these four curves exist for big enough periods @xmath6 . in this case , `` big enough '' only depends on the size of the nonintegrable terms of @xmath139 , the size of the neighborhood of @xmath908 , the exponent @xmath110 , and the winding number @xmath9 . on the contrary
, it _ does not depend _ on the particular map at hand .
therefore , every time that we ask @xmath6 to be `` big enough '' along the proof of the following lemma , it only depends on the quantities @xmath914 , @xmath915 , @xmath916 , @xmath917 , and @xmath918 fixed at the first line of the next statement .
[ lem : preliminary ] let @xmath919 , @xmath915 , @xmath916 , and @xmath920 .
let @xmath921 be an integer relatively prime with @xmath9 . set @xmath922 .
let @xmath923 be an analytic map of the form ( [ eq : neishtadtform ] ) such that @xmath924 and @xmath925 for all @xmath926 .
let @xmath927 .
let @xmath928 , @xmath929 , and @xmath930 .
there exists @xmath931 such that , if @xmath932 , the following properties hold : 1 .
the map @xmath139 has two rics @xmath933 whose internal dynamics is conjugated to a rigid rotation of angles @xmath934 , respectively ; 2 . if @xmath879 is the compact subset of @xmath880 enclosed by @xmath935 and @xmath936 , then @xmath937 3 .
[ item : zeta ] there exist two unique analytic functions @xmath938 and @xmath939 such that @xmath940 and all the @xmath0-periodic points of the restriction @xmath941 are contained in @xmath942 .
the same statement holds if @xmath9 is a negative integer , @xmath943 , @xmath944 , @xmath945 , and @xmath946 .
let us assume @xmath628 .
the case @xmath947 is analogous .
first , the existence of the rics @xmath935 and @xmath936 follows from some quantitative estimates in kam theory established by lazutkin ( * ? ? ?
* theorem 2 ) . to be precise
, lazutkin proved that there exists @xmath948 such that if @xmath949 satisfies the diophantine condition @xmath950 for all integers @xmath951 and @xmath93 , then @xmath139 has a ric @xmath952 whose internal dynamics is @xmath953-conjugated to a rigid rotation of angle @xmath101 , for a suitable @xmath954 .
the conjugation is @xmath955-close to the identity .
( i ) follows directly from this estimate , because there exist some real numbers @xmath956 and @xmath957 satisfying the diophantine condition ( [ eq : diophantine ] ) , provided @xmath6 is big enough .
second , let us check that the power map @xmath909 satisfies ( [ eq : twistcondition ] ) . the compact subset @xmath887 is invariant by @xmath139 , because it is delimited by rics .
thus , all powers @xmath958 are well defined on @xmath879 .
we write @xmath959 , where @xmath408 was introduced in lemma [ lem : boundgammas ] .
next , we compute the differential of the power map : @xmath960 where @xmath961 and @xmath962 is the sum of all the products of the form @xmath963 , with @xmath964 , @xmath965 , and @xmath966 .
these products are elements of @xmath967 , because @xmath968 . indeed , if @xmath969 , then @xmath970 where we have used lemma [ lem : boundgammas ] , inequality @xmath971 for @xmath972 , @xmath973 , @xmath974 , @xmath975 , @xmath922 , and @xmath916 .
we have also defined @xmath976 .
the matrix @xmath962 is the sum of the products with precisely @xmath432 factors @xmath977 .
this shows that there are @xmath978 terms inside @xmath962 .
therefore , @xmath979 and @xmath980 the element of the first row and second column of @xmath981 is equal to @xmath6 , so @xmath982 for all @xmath983 , which implies the twist condition ( [ eq : twistcondition ] ) provided that @xmath6 is big enough . here , we have used relation ( [ eq : dfq ] ) , bound ( [ eq : bounddelta ] ) , @xmath922 , and @xmath984 .
we have also used that @xmath985 and @xmath986 , provided @xmath6 is big enough .
third , we establish the existence of the functions @xmath987 .
we know from lazutkin @xcite that @xmath906 for some differentiable functions @xmath988 .
we work with the lifts @xmath409 , @xmath989 , and @xmath990 of the objects @xmath139 , @xmath991 , and @xmath992 .
the rics are invariant , so @xmath993 for some differentiable functions @xmath994 .
if we prove that there exist two unique analytic 1-periodic functions @xmath995 such that @xmath996 then item ( iii ) follows .
since the dynamics of @xmath997 on @xmath912 is @xmath953-conjugated to a rigid rotation of angle @xmath998 through a @xmath955-close to the identity conjugation , @xmath999 provided that @xmath6 is big enough .
analogously , we obtain @xmath1000 .
that is , @xmath1001 since @xmath1002 is analytic and strictly increasing for @xmath1003 , we deduce that there exists a unique function @xmath1004 such that @xmath1005 .
the function @xmath1006 is analytic and @xmath1007 , so @xmath1008 is analytic by the implicit function theorem .
the @xmath33-periodicity of @xmath1008 follows from the uniqueness and the property @xmath1009 .
function @xmath1010 is defined by means of relation ( [ eq : verticalcurveslift ] ) .
finally , functions @xmath1011 are the projections of @xmath1012 .
if @xmath1014 , by hypothesis , the map @xmath1015 , @xmath1016 , is an analytic exact twist map with a @xmath170-analytic @xmath255-resonant ric , such that @xmath1017 .
the map @xmath1018 satisfies the properties ( i)(iii ) listed in lemma [ lem : order2 ] in some suitable coordinates @xmath66 .
let @xmath1019 be the associated change of variables .
let @xmath1020 be the new map defined in the domain ( [ eq : dab ] ) .
note that the @xmath170-analytic @xmath255-resonant ric is @xmath1021 in the @xmath66 coordinates .
let @xmath9 be an integer such that @xmath1022 .
let @xmath1023 such that @xmath1024 .
we take @xmath1025 as the @xmath32 appearing in theorem [ thm : remainder ] , @xmath1026 , and @xmath1027 , provided that @xmath6 is relatively prime with @xmath9 and is large enough so that @xmath1028 .
that is , @xmath1029 . hence , there exist @xmath1030 , both independent of @xmath6 , and a change of coordinates @xmath1031 such that @xmath1032 is an analytic map of the form ( [ eq : neishtadtform ] ) such that @xmath1033 , @xmath1034 , and @xmath1035 \right| \ } \le k_1 $ ] for all @xmath926 .
the map @xmath1036 satisfies the hypotheses of lemma [ lem : preliminary ] for any @xmath1037 .
let @xmath1038 be the maximum value of @xmath1039 among the integers @xmath1040 .
let @xmath912 be the rics with rotation numbers @xmath1041 given in lemma [ lem : preliminary ] .
let @xmath879 be the compact subset of @xmath880 enclosed by @xmath935 and @xmath936 .
since @xmath139 is globally twist and @xmath1042 , all the birkhoff @xmath0-periodic orbits of @xmath139 are contained in @xmath879 . by lemma [ lem : preliminary ] ,
any @xmath0-periodic orbit in @xmath879 lies on @xmath904 .
let @xmath1043 be the domain enclosed by the curves @xmath904 and @xmath905 .
let @xmath1044 .
let @xmath1045 be the supremum of @xmath1046|$ ] in the compact set @xmath1047 $ ] .
let @xmath1048 .
then , following the arguments contained in subsection [ ssec : differenceofperiodicactions ] about the difference of periodic actions , we get that @xmath1049 \le k_1 k_2 { \mathop{\rm area}\nolimits}[\omega ] = k_1 k_2 \int_{\mathbb{t}}\left|\hat{\zeta}(\xi ) - \zeta(\xi)\right|\rmd \xi \\ \label{eq : deltaboundproof } & \le & k_1 k_2 q b^4 k_0 \rme^{-2\pi \alpha q /|p| } \le k
\rme^{-2\pi\alpha q /|p|},\end{aligned}\ ] ] for all relatively prime integers @xmath9 and @xmath6 with @xmath1022 and @xmath77
. we have used expression ( [ eq : neishtadtform ] ) , @xmath1050 , @xmath1051 , the bounds on the nonintegrable terms @xmath1052 , and the bounds on the jacobians of the changes of variables @xmath495 and @xmath1053 .
this ends the proof of theorem [ thm : asymptotic ] when @xmath1014 .
we reduce the general case to the previous one .
we split the argument in four steps .
_ step 1 : about the rational rotation numbers .
_ if @xmath109 is a @xmath1-resonant ric and @xmath83 is a @xmath0-periodic point of @xmath97 , then @xmath109 is a @xmath1054-resonant ric and @xmath83 is a @xmath1055-periodic point of the power map @xmath117 , where @xmath1056 by taking the suitable lift @xmath409 of @xmath139 , we can assume that @xmath109 is a @xmath255-resonant ric and @xmath83 is a @xmath1057-periodic point of @xmath139 , with @xmath1058 .
that is , @xmath1059 if @xmath9 and @xmath6 are relatively prime integers such that @xmath1060 and @xmath77 , @xmath1061 and @xmath1062 are relatively prime integers such that @xmath1063 and @xmath1064 .
_ step 2 : about the lagrangians .
_ let @xmath99 and @xmath1065 be the lifts of @xmath97 and @xmath1066 we are dealing with . if @xmath1067 , then @xmath1068 = \sum_{j=0}^{n-1 } \rmd ( h \circ g^j ) = \rmd \left ( \sum_{j=0}^{n-1 } h \circ g^j \right),\ ] ] so @xmath1069 is a lagrangian of @xmath139 .
this lagrangian is well defined in a neighboorhood of the resonant ric @xmath109 , because @xmath139 is twist on @xmath109 .
_ step 3 : about the periodic actions .
_ let @xmath1070 be the @xmath0-periodic orbit of @xmath97 through the point @xmath83 , being @xmath1071 $ ] its @xmath0-periodic action .
let @xmath1072 be the @xmath1057-periodic orbit of @xmath139 through the same point , being @xmath1073 $ ] its @xmath1057-periodic action .
we deduce from the previous steps and a straightforward computation that @xmath1074 = \frac{n}{\gcd(n , q ) } w^{(p , q)}[o].\ ] ] _ step 4 : final bound .
_ the result follows directly from the bound ( [ eq : deltaboundproof ] ) taking into account relations ( [ eq : newrotationnumber ] ) and ( [ eq : newperiodicaction ] ) . we just note that @xmath1075 this ends the proof of theorem [ thm : asymptotic ] . the authors were supported in part by cur - diue grant 2014sgr504 ( catalonia ) and mineco - feder grant mtm2012 - 31714 ( spain ) .
some of the results of this paper were obtained while a t was in the university of warwick as a visiting researcher .
thanks to vassily gelfreich , robert mackay , laurent niederman , snia pinto - de - carvalho , and tere seara for useful and stimulating conversations . | area - preserving twist maps have at least two different @xmath0-periodic orbits and every @xmath0-periodic orbit has its @xmath0-periodic action for suitable couples @xmath0 .
we establish an exponentially small upper bound for the differences of @xmath0-periodic actions when the map is analytic on a @xmath1-resonant rotational invariant curve ( resonant ric ) and @xmath2 is `` sufficiently close '' to @xmath3 .
the exponent in this upper bound is closely related to the analyticity strip width of a suitable angular variable .
the result is obtained in two steps .
first , we prove a neishtadt - like theorem , in which the @xmath4-th power of the twist map is written as an integrable twist map plus an exponentially small remainder on the distance to the ric .
second , we apply the mackay - meiss - percival action principle .
we apply our exponentially small upper bound to several billiard problems .
the resonant ric is a boundary of the phase space in almost all of them .
for instance , we show that the lengths ( respectively , areas ) of all the @xmath5-periodic billiard ( respectively , dual billiard ) trajectories inside ( respectively , outside ) analytic strictly convex domains are exponentially close in the period @xmath6 .
this improves some classical results of marvizi , melrose , colin de verdire , tabachnikov , and others about the smooth case .
_ keywords _ : twist maps , invariant curves , exponential smallness , billiards , dual billiards |
the purpose of this work is to develop a positivity - preserving version of the lax - wendroff discontinuous galerkin method for the compressible euler equations on unstructured meshes .
the compressible euler equations form a system of hyperbolic conservation law that can be written as follows : @xmath0 the _ conserved _ variables are the mass density , @xmath1 , the momentum density , @xmath2 , and the energy density , @xmath3 ; the _ primitive _ variables are the mass density , @xmath1 , the fluid velocity , @xmath4 , and the pressure , @xmath5 . the energy @xmath3 is related to the primitive variables through the equation of state , @xmath6 where the constant @xmath7 is the ratio of specific heats ( aka , the _ gas constant _ ) .
the compressible euler equations are an important mathematical model in the study of gases and plasma .
attempts at numerically solving the these equations has led to a plethora of important historical advances in the development of numerical analysis and scientific computing ( see e.g. , @xcite ) .
the focus of this work is on high - order discontinuous galerkin ( dg ) methods , which were originally developed for general hyperbolic conservation laws by cockburn , shu , et al . in series of papers @xcite .
the purpose of this section is to set the notation used throughout the paper and to briefly describe the dg spatial discretization .
let @xmath8 be a polygonal domain with boundary @xmath9 .
the domain @xmath10 is discretized via a finite set of non - overlapping elements , @xmath11 , such that @xmath12 .
let @xmath13 denote the set of polynomials from @xmath14 to @xmath15 with maximal polynomial degree @xmath16 .
let @xmath17 denote the _ broken _ finite element space on the mesh : @xmath18^{\meq } : \ , w^h \bigl|_{{\mathcal{t}}_i } \in \left [ { p}^{\ , \deg } \right]^{\meq } , \ , \forall { \mathcal{t}}_i \in { \mathcal{t}}^h \right\},\ ] ] where @xmath19 is the mesh spacing .
the above expression means that @xmath20 has @xmath21 components , each of which when restricted to some element @xmath11 is a polynomial of degree at most @xmath16 and no continuity is assumed across element edges ( or faces in 3d ) .
the approximate solution on each element @xmath11 at time @xmath22 is of the form q^h(t^n , * x * ( ) ) |__i = _
= 1^ ( ) q^ ( ) n_i ^ ( ) ( ) , where @xmath23 is the number of legendre polynomials and @xmath24 are the legendre polynomials defined on the reference element @xmath25 in terms of the reference coordinates @xmath26 .
the legendre polynomials are orthonormal with respect to the following inner product : _ _ 0 ^(k ) ( ) ^ ( ) ( ) d= 1 & k= , + 0 & k .
we note that independent of @xmath19 , @xmath27 , @xmath16 , and the type of element , the lowest order legendre polynomial is always @xmath28 .
this makes the first legendre coefficient the cell average : q^(1 ) n_i = _ _ 0 q^h(t^n , * x * ( ) ) |__i ^(1 ) ( ) d= : ^n_i .
the most common approach for time - advancing dg spatial discretizations is via explicit runge - kutta time - stepping ; the resulting combination of time and space discretization is often referred to as the `` rk - dg '' method @xcite .
the primary advantage for this choice of time stepping is that explicit rk methods are easy to implement , they can be constructed to be low - storage , and a subclass of these methods have the so - called strong stability preserving ( ssp ) property @xcite , which is important for defining a scheme that is provably positivity - preserving . however , there are no explicit runge - kutta methods that are ssp for orders greater than four @xcite .
the main difficulty with runge - kutta methods is that they typically require many stages ; and therefore , many communications are needed per time step .
one direct consequence of the communication required at each rk stage is that is difficult to combine rk - dg with locally adaptive mesh refinement strategies that simultaneously refine in both space and time .
the key piece of technology required in locally adaptive dg schemes is local time - stepping ( see e.g. , dumbser et al .
@xcite ) . local time - stepping is easier to accomplish with a single - stage , single - step ( lax - wendroff ) method than with a multi - stage runge - kutta scheme . for these reasons there
is interest from discontinuous galerkin theorists and practitioners in developing single - step time - stepping techniques for dg ( see e.g. , @xcite ) , as well as hybrid multistage multiderivative alternatives @xcite . in this work ,
we construct a numerical scheme that uses a lax - wendroff time discretization that is coupled with the discontinuous galerkin spatial discretization . in subsequent discussions in this paper
we demonstrate the advantages of switching to single - stage and single - stage time - stepping in regards to enforcing positivity on arbitrary meshes . in simulations involving strong shocks , high - order schemes ( i.e. more than first - order ) for the compressible euler equations generally create nonphysical undershoots ( below zero ) in the density and/or pressure
. these undershoots typically cause catastrophic numerical instabilities due to a loss of hyperbolicity .
moreover , for many applications these positivity violations exist even when the equations are coupled with well - understood total variation diminishing ( tvd ) or total variation bounded ( tvb ) limiters .
the chief goal of the limiting scheme developed in this work is to address positivity violations in density and pressure , and in particular , to accomplish this task with a high - order scheme . in the dg literature ,
the most widely used strategy to maintain positivity was developed by zhang and shu in a series of influential papers @xcite .
the basic strategy of zhang and shu for a positivity - preserving rk - dg method can be summarized as follows : * step 0 . * : : write the current solution in the form @xmath29 where @xmath30 is yet - to - be - determined .
@xmath31 represents the unlimited solution .
* step 1 . * : : find the largest value of @xmath30 , where @xmath32 , such that @xmath33 satisfies the appropriate positivity conditions at some appropriately chosen quadrature points and limit the solution .
* step 2 . * : : find the largest stable time - step that guarantees that with a forward euler time that the cell average of the new solution remains positive .
* step 3 . *
: : rely on the fact that strong stability - preserving runge - kutta methods are convex combinations of forward euler time steps ; and therefore , the full method preserves the positivity of cell averages ( under some slightly modified maximum allowable time - step ) . for a lax - wendroff time discretization ,
numerical results indicate that the limiting found in step 1 is insufficient to retain positivity of the solution , even for simple 1d advection .
therefore , the strategy we pursue in this work will still contain an equivalent step 1 ; however , in place of step 2 and step 3 above , we will make use of a parameterized flux , sometimes also called a flux corrected transport ( fct ) scheme , to maintain positive cell averages after taking a single time step . in doing so , we avoid introducing additional time step restrictions that often appear ( e.g. , in step 2 . above ) when constructing a positivity - preserving scheme based on runge - kutta time stepping .
this idea of computing modified fluxes by combining a stable low - order flux with a less robust high - order flux is relatively old , and perhaps originates with harten and zwas and their _ self adjusting hybrid scheme _ @xcite .
the basic idea is the foundation of the related _ flux corrected transport _ ( fct ) schemes of boris , book and collaborators @xcite , where fluxes are adjusted in order to guarantee that average values of the unknown are constrained to lie within locally defined upper and lower bounds .
this family of methods is used in an extensive variety of applications , ranging from seismology to meteorology @xcite .
a thorough analysis of some of the early methods is conducted in @xcite .
identical to modern maximum principle preserving ( mpp ) schemes , fct can be formulated as a global optimization problem where a `` worst case '' scenario assumed in order to decouple the previously coupled degrees of freedom @xcite .
here we do not attempt to use fct to enforce any sort of local bounds ( in the sense of developing a shock - capturing limiter ) , instead we leverage these techniques in order to retain positivity of the density and pressure associated to @xmath34 ; such approaches have recently received renewed interest in the context of weighted essentially non - oscillatory ( weno ) methods @xcite . to summarize , our limiting scheme draws on ideas from the two aforementioned families of techniques that are well established in the literature .
first , we start with the now well known ( high - order ) pointwise limiting developed for discontinuous galerkin methods @xcite , and second , we couple this with the very large family of flux limiters @xcite .
( developed primarily for finite - difference ( fd ) and finite - volume ( fv ) schemes ) .
the compressible euler equations can be written compactly as @xmath35 where the _ conserved variables _ are @xmath36 and the _ flux function _ is @xmath37 where @xmath38 and @xmath39 . the basic positivity limiting strategy proposed in this work is summarized below .
some important details are omitted here , but we elaborate on these details in subsequent sections . *
step 0 . * : : on each element we write the solution as @xmath40 where @xmath30 is yet - to - be - determined .
@xmath41 represents the _ unlimited _ solution . *
step 1 . * : : assume that this solution is positive in the mean .
that is , we assume for all @xmath42 that @xmath43 and @xmath44 a consequence of these assumptions is that @xmath45 .
* step 2 . * : : find the largest value of @xmath30 , where @xmath32 , such that the density and pressure are positive at some suitably defined quadrature points .
this step is elaborated upon in [ subsec : mpp - limiter ] .
* step 3 . * : : construct time - averaged fluxes through the lax - wendroff procedure .
that is , we start with the exact definition of the time - average flux : @xmath46 and approximate this via a taylor series expansion around @xmath47 : @xmath48 all time derivatives in this expression are replaced by spatial derivatives using the chain rule and the governing pde .
the approximate time - averaged flux is first evaluated at some appropriately chosen set of quadrature points in fact , the same quadrature points as used in step 2 and then , using appropriate quadrature weights , summed together to define a high - order flux , at both interior and boundary quadrature points .
thanks to step 2 , all quantities of interest used to construct this expansion are positive at each quadrature point .
* step 4 . * : : time step the solution so that cell averages are guaranteed to be positive .
that is , we update the _ cell averages _ via a formula of the form @xmath49 where @xmath50 is an outward - pointing ( relative to @xmath11 ) normal vector to edge @xmath51 with the property that @xmath52 is the length of edge @xmath53 , and the numerical flux on edge , @xmath54 , is a convex combination of a high - order flux , @xmath55 , and a low - order flux @xmath56 : @xmath57 the low - order flux , @xmath58 , is based on the ( approximate ) solution to the riemann problem defined by cell averages only , and the `` high - order '' flux , @xmath59 , is constructed after integrating via gaussian quadrature the ( approximate ) riemann solutions at quadrature points along the edge @xmath60 : @xmath61 where @xmath62 are the gaussian quadrature weights for quadrature with @xmath63 points and @xmath64 are the numerical fluxes at each of the @xmath63 quadrature points .
note that this sum has only a single summand in the one - dimensional case .
the selection of @xmath30 is described in more detail in [ subsec : mpp - limiter ] .
this step guarantees that the solution retains positivity ( in the mean ) for a single time step .
* step 5 . * : : apply a shock - capturing limiter .
the positivity - preserving limiter is designed to preserve positivity of the solution , but it fails at reducing spurious oscillations , and therefore a shock - capturing limiter needs to be added .
there are many choices of limiters available ; we use the limiter recently developed in @xcite because of its ability to retain genuine high - order accuracy , and its ability to push the polynomial order to arbitrary degree without modifying the overall scheme .
* step 6 . * : : repeat all of these steps to update the solution for the next time step .
each step of this process is elaborated upon throughout the remainder of this paper .
the end result is that our method is the first scheme to simultaneously obtain all of the following properties : * * high - order accuracy . *
the proposed method is third - order in space and time , and can be extended to arbitrary order . *
* positivity - preserving . *
the proposed limiter is provably positivity - preserving for the density and pressure , at a finite set of point values , for the entire simulation . * * single - stage , single - step .
* we use a lax - wendroff discretization for time stepping the pde , and therefore we only need one communication per time step . *
* unstructured meshes . *
because we use the discontinuous galerkin method for our spatial discretization and all of our limiters are sufficiently local , we are able to run simulations with dg - fem on both cartesian and unstructured meshes .
* * no additional time - step restrictions .
* because we do not rely on a ssp runge - kutta scheme , we do not have to introduce additional time - step restrictions to retain positivity of the solution .
this differentiates us from popular positivity - preserving limiters based on rk time discretizations @xcite .
the remainder of this paper has the following structure .
the lax - wendroff dg ( lxw - dg ) method is described in [ sec : lwdg ] , where we view the scheme as a method of modified fluxes .
the positivity - preserving limiter is described in [ sec : positivity ] , where the discussion of the limiter is broken up into two parts : ( 1 ) the moment limiter ( [ sec : zh - limiter ] ) and ( 2 ) the parameterized flux limiter ( [ subsec : mpp - limiter ] ) . in
[ sec : numerical - results ] we present numerical results on several test cases in 1d , 2d cartesian , and 2d unstructured meshes .
finally we close with conclusions and a discussion of future work in
[ sec : conclusions ] .
the lax - wendroff discontinuous galerkin ( lxw - dg ) method @xcite serves as the base scheme for the method developed in this work .
it is the result of an application of the cauchy - kovalevskaya procedure to hyperbolic pde : we start with a taylor series in time , then we replace all time derivatives with spatial derivatives via the pde .
finally , a galerkin projection discretizes the overall scheme , where a single spatial derivative is reserved for the fluxes in order to perform integration - by - parts .
we review the lax - wendroff dg scheme for the case of a general nonlinear conservation law that is autonomous in space and time in multiple dimensions @xcite .
the current presentation illustrates the fact that lax - wendroff schemes can be viewed as a method of modified fluxes , wherein higher - order information about the pde is directly incorporated by simply redefining the fluxes that would typically be used in an `` euler step . ''
we consider a generic conservation law of the form @xmath65 where the matrix @xmath66 is diagonalizable for every unit length vector @xmath67 and @xmath68 in the domain of interest .
formal integration of over an interval @xmath69 $ ] results in an exact update through @xmath70 where the _ time - averaged _ flux @xcite is defined as @xmath71 moreover , a taylor expansion of @xmath72 and a change of variables yields @xmath73 which can be inserted into . in a numerical discretization of
, the taylor series in is truncated after a finite number of terms . if @xmath74 , then reduces to a forward euler time discretization for hyperbolic conservation law .
this fact will allow us to incorporate positivity - preserving limiters into the lax - wendroff flux construction .
this observation allows us to incorporate the positivity - preserving limiters that are presented in
[ sec : zh - limiter ] and [ subsec : mpp - limiter ] , because we view the lxw - dg method as a method of modified fluxes .
we now describe how to compute the temporal derivative terms : @xmath75 that are required to define the time - averaged flux in .
this discussion is applicable to high - order finite difference methods , finite volume methods ( e.g. , ader ) , as well as discontinuous galerkin finite element methods .
a single application of the chain rule to compute the time derivative of the flux function yields @xmath76 where the flux jacobian is @xmath77 the matrix - vector products in can be compactly written using the einstein summation convention ( where repeated indices are assumed to be summed over ) , which produces a vector whose @xmath78-component is @xmath79 a second derivative of yields @xmath80 where @xmath81 is the hessian with elements given by @xmath82 equations and are generic formulae ; the equalities are appropriate for any two - dimensional hyperbolic system , and similar identities exist for three dimensions .
the first product in the right hand side of is understood as a hessian - vector product .
scripts that compute these derivatives , as well as the matrix , and hessian vector products that are necessary to implement a third - order lax - wendroff scheme for multidimensional euler equations can be found in the open source software finess @xcite .
finally , these two time derivatives are sufficient to construct a third - order accurate method by defining the time - averaged flux through @xmath83 and then updating the solution through @xmath84 in place of .
the final step is to construct a fully discrete version of .
the lxw - dg scheme follows the following process @xcite : * step 1 . * : : at each quadrature point evaluate the numerical flux , @xmath85 , and then integrate this numerical flux against basis functions to obtain a galerkin expansion of @xmath86 inside each element .
* step 2 . * : : using the galerkin expansions of @xmath87 and @xmath86 , evaluate all required spatial derivatives to construct the time expansion @xmath88 in at each quadrature point . * step 3 . * : : multiply by a test function @xmath89 , integrate over a control element @xmath11 , and apply the divergence theorem to yield @xmath90 where @xmath91 is the outward pointing unit normal to element @xmath11 , which reduces to @xmath92 by orthogonality of the basis @xmath93 . in practice ,
both the interior and edge integrals are approximated by appropriate numerical quadrature rules .
the flux values , @xmath94 , in the edge integrals still need to be defined
. * step 4 . * : : along each edge solve riemann problems at each quadrature point by using the left and right interface values . in this work we use the well - known lax - friedrichs flux : @xmath95,\ ] ] where @xmath96 is an estimate of the maximum global wave speed , @xmath97 is the approximate solution evaluated on the element boundary on the interior side of @xmath11 , and @xmath98 is the approximate solution evaluated on the element boundary on the exterior side of @xmath11 . in order to achieve high - order accuracy at the boundaries of the computational domain , a careful treatment of the solution in each boundary element
is required .
in particular , all simulations in this work require either reflective ( hard surface ) or transparent ( outflow ) boundary conditions .
\(o ) at ( 0,0 ) ; ( a ) at ( 2.5,1 ) ; ( b ) at ( 0,2 ) ; ( c ) at ( @xmath99 ) ; ( c ) at ( @xmath99 ) ; ( c ) at ( @xmath100 ) ; ( o)(a)(b ) ; ( b)(o ) ; suppose the solution takes on the value @xmath101 at a quadrature point @xmath102 on the boundary @xmath9 , and we wish to define a boundary value , @xmath103 , on the exterior side of the boundary @xmath9 that yields a flux with one of two desired boundary conditions .
this is depicted in figure [ fig : bcs ] .
let @xmath104 and @xmath91 be the unit tangent and unit normal vectors to the boundary at the boundary point @xmath102 , respectively . in both the reflective and transparent boundary conditions , we enforce continuity of the tangential components : @xmath105 the only difference between the two types of boundary conditions we consider lie in the normal direction .
we set @xmath106 where the minus sign corresponds to the reflective boundary condition and the plus sign corresponds to the transparent boundary condition .
from this we can easily write out the full boundary conditions at the point @xmath107 : @xmath108 where again the minus sign corresponds to the reflective boundary condition and the plus sign corresponds to the transparent boundary condition . in order to achieve high - order time accuracy
we need apply the above boundary conditions to the time derivatives on the boundary : @xmath109 where all time derivatives must be replaced by spatial derivatives using the pde .
the temporal evolution described in the previous section fails to retain positivity of the solution , even in the simple case of linear advection with smooth solutions that are near zero .
in fact , in extreme cases the projection of the initial conditions can fail to retain positivity of the solution due to the gibbs phenomena .
the positivity - preserving limiter we present follows a two step procedure : * step 1 . *
: : limit the moments in the expansion so that the solution is positive at each quadrature point .
* step 2 . * : : limit the fluxes so that the cell averages retain positivity after a single time step .
we now describe the first of these two steps .
this section describes a procedure that implements the following : if the cell averages are positive , then the solution is forced to be positive at a preselected and finite collection of quadrature points .
we select only the quadrature points that are actually used in the numerical update ; this includes internal gauss quadrature points as well as face / edge gauss quadrature points . unlike other positivity limiting schemes @xcite in the ssp runge - kutta framework ,
this step is not strictly necessary to guarantee positivity of the cell average at the next time - step ; however , the main reason for applying the limiter at quadrature points is to guarantee that each term in the update is physical , which will reduce the total amount of additional limiting of the cell average updated needed in section [ subsec : mpp - limiter ] . the process to maintain positivity at quadrature points
is carried out in a series of three simple steps . because this part of the limiter is entirely local ,
we focus on a single element @xmath110 , and therefore drop the subscript for ease of notation .
we assume that the cell averages for the density satisfies @xmath111 , where @xmath112 is a cutoff parameter that guarantees hyperbolicity of the system . in this work , we set @xmath113 in all simulations .
furthermore , we assume that the cell average for the pressure satisfies @xmath114 , where the ( average ) pressure @xmath115 is defined through the averages of the other conserved quantities in .
note that these two conditions are sufficient to imply that the average energy @xmath116 is positive . in this step
, we enforce positivity of the density at each quadrature point @xmath117 . because @xmath111 , there exists a ( maximal ) value @xmath118 $ ] such that @xmath119 for all quadrature points @xmath117 and all @xmath120 $ ] .
note that we drop the subscript that indicates the element number to ease the complexity of notation for the the ensuing discussion .
recall that the pressure is defined through the relation .
we seek to guarantee that @xmath121 for each quadrature point @xmath122 .
in place of working directly with the pressure , we observe that it suffices to guarantee that the product of the density and pressure is positive . to this end
, we expand the momentum and energy in the same free parameter @xmath30 .
similar to the density in , we write the momentum and energy as @xmath123 we define deviations from cell averages as @xmath124 and compactly write the expression for the limited variables as the cell average plus deviations : @xmath125 the product of the density and pressure at each quadrature point is a quadratic function of @xmath30 : @xmath126 where @xmath127 is defined in . after expanding each of these conserved variables in the _ same scaling parameter
_ @xmath30 , we observe that @xmath128 , \end{aligned}\ ] ] where @xmath129 and @xmath130 depend only on the quadrature point and higher - order terms of the expansions of density , energy , and momentum : @xmath131 the quadratic function defined by is non - negative for at least one value of @xmath30 , namely @xmath132 . however , if is positive at @xmath132 , then we are guaranteed that there exists a @xmath133 $ ] that guarantees @xmath134 for all @xmath135 $ ] .
in particular , we are interested in finding the largest such @xmath30 ( i.e. , the least amount of damping ) . instead of exactly computing the optimal @xmath30 , which could readily be done , but would require additional floating point operations , we make use of the following lemma @xcite to find an approximately optimal @xmath30 . the pressure function is a convex function of @xmath30 on @xmath136 $ ] .
that is , @xmath137 for all , and @xmath138 $ ] .
we observe that directly from the definition of the limiter for the conserved variables in that @xmath139 and therefore @xmath140 the final inequality follows because @xmath141 for all @xmath142 , and the pressure is a convex function ( of the conserved variables ) whenever the density is positive . as a consequence of we
can define @xmath143 which will guarantee that @xmath144 for all @xmath135 $ ] .
finally , we define the scaling parameter for the entire cell as @xmath145 and use this value to limit the higher order coefficients in the galerkin expansions of the density , momentum , and energy displayed in eqns . and .
this definition gives us the property that @xmath146 and @xmath147 at each quadrature point @xmath148 .
this process is repeated ( locally ) in each element @xmath11 in the mesh . as a side benefit to guaranteeing that the density and pressure are positive
, we have the following remark .
if @xmath127 and @xmath149 are positive at each quadrature point , then @xmath150 is also positive at each quadrature point .
divide by @xmath151 and add @xmath152 to both sides .
this concludes the first of two steps for retaining positivity of the solution .
we now move on to the second and final step , which takes into account the temporal evolution of the solver .
the procedure carried out for the flux limiter presented in this section is very similar to recent work for finite volume @xcite as well as finite difference @xcite methods . when compared to the finite difference methods ,
the main difference in this discussion is that the expressions do not simplify as much because quantities such as the edge lengths must remain in the expressions .
this makes them more similar to work on finite volume schemes @xcite .
overall however , there is little difference between flux limiters on cartesian and unstructured meshes , and between flux limiters for finite difference , finite volume ( fv ) , and discontinuous galerkin ( dg ) schemes .
this is because the updates for the cell average in a dg solver can be made to look identical to the update for a fv solver , and once flux interface values are identified , a conservative fd method can be made to look like a fv solver , albeit with a different stencil for the discretization .
all of the aforementioned papers rely on the result of perthame and shu @xcite , which states that a first - order finite volume scheme ( i.e. , one that is based on a piecewise constant representation with forward euler time - stepping ) that uses the lax - friedrichs ( lxf ) numerical flux is positivity - preserving under the usual cfl condition . similar to previous work , we leverage this idea and incorporate it into a flux limiting procedure . here , the focus is on lax - wendroff discontinuous galerkin schemes . in this work we write out the details of the limiting procedure only for the case of 2d triangular elements .
however , all of the formulas generalize to higher dimensions and cartesian meshes .
to begin , we consider the euler equations and a mesh that fits the description given in
[ sec : dg_spatial ] .
after integration over a single cell , @xmath11 , and an application of the divergence theorem , we see that the exact evolution equation for the cell average of the density is given by @xmath153 where @xmath91 is the outward pointing ( relative to @xmath11 ) unit normal to the boundary of @xmath11 . applying to this equation a first - order finite volume discretization using the lax - friedrichs flux yields @xmath154 and
the lax - friedrichs flux is @xmath155 where @xmath50 is an outward - pointing ( relative to @xmath11 ) normal vector to edge @xmath51 with the property that @xmath52 is the length of edge @xmath53 , and the @xmath156 index refers to the solution on edge @xmath51 on the exterior side of @xmath157 .
we use a _ global _ wave speed @xmath96 , because this flux defines a provably positivity - preserving scheme @xcite .
other fluxes can be used , provided they are positivity - preserving ( in the mean ) .
recall that the update for the lxw - dg method is given in .
the numerical edge flux , @xmath158 , is based on the temporal taylor series expansion of the fluxes .
the update for the cell averages takes the form : @xmath159 which in practice needs to be replaced by a numerical quadrature along the edges . applying guassian
quadrature along each edge produces the following edge value ( or face value in the case of 3d ) : @xmath160 where @xmath161 and @xmath162 are gaussian quadrature points and weights , respectively , for integration along edge @xmath51 .
this allows us to write the update for the cell average in the lax - wendroff dg method in a similar fashion to that of the lax - friedrichs solver , but this time we have higher - order fluxes : @xmath163 next we define a parameterized flux on edge @xmath51 by @xmath164 where @xmath165 $ ] is as free parameter that is yet to be determined .
we also define the quantity , @xmath166 , which by virtue of the positivity of the lax - friedrichs method is positive in both density and pressure : @xmath167 in order to retain a positive density in the high - order update formula , the following condition must be satisfied : @xmath168 where @xmath169 note that a @xmath1 superscript is introduced to the flux function in order to denote the first component of the flux , namely the mass flux .
positivity of the density is achieved if @xmath170 the basic procedure for positivity limiting is to reduce the values of @xmath171 until inequality condition is satisfied . for a triangular mesh
, there are three values of @xmath171 that contribute to each cell . in this case
there exists a three - dimensional feasible region that contains all admissible @xmath171 values , one for each @xmath60 , where the new average density , @xmath172 , is positive . finding the exact boundary of this set is computationally impractical ; and therefore , we make an approximation so that the problem becomes much simpler . in particular , we approximate the feasible region by a rectangular cuboid : @xmath173 \times \left[0,\varlambda_{e_{i2}}]\times[0,\varlambda_{e_{i3}}\right ] \subseteq [ 0,1]^3,\ ] ] where @xmath174 , @xmath175 , and @xmath176 are the three edges that make up element @xmath11 , over which @xmath177 once we have determined the feasible region for each element over which the average density , @xmath178 , remains positive , the next step is to rescale each @xmath179 for @xmath180 to also guarantee a positive average pressure , @xmath181 , on the same element : @xmath182 \times \left[0,\mu_{e_{i2 } } \ , \varlambda_{e_{i2 } } ] \times[0,\mu_{e_{i3 } } \ , \varlambda_{e_{i3}}\right ] \subseteq [ 0,1]^3,\ ] ] where @xmath183 for each @xmath42 and for each @xmath180 .
we leave the details of the procedure to determine the feasibility region to the next subsection , in which we also summarize the full algorithm
. an efficient implementation of the positivity preserving limiter should avoid communication with neighboring cells as much as possible .
indeed , this is one advantage of single - stage , single - step methods such as the lax - wendroff discontinuous galerkin method . in order to avoid additional communication overhead
, we suggest the implementation described below . in the formulas below we make use of the following quantities : @xmath184 these quantities are illustrated in figure [ fig : edge ] .
( a ) \(o ) at ( 0,0 ) ; ( a ) at ( 2.5,1 ) ; ( b ) at ( 0,2 ) ; ( c ) at ( -2.5,1 ) ; ( d1 ) at ( 0,1 ) ; ( d2 ) at ( 0.75,1 ) ; ( o)(a)node[rotate=0,xshift=-1.2 cm , yshift=0.0 cm ] @xmath185(b)(o ) ; ( c)(o)node[rotate=90,xshift=1 cm , yshift=0.225 cm ] edge : @xmath51 ; ( c)node[rotate=0,xshift=1.5 cm , yshift=0.0 cm ] @xmath186(b ) ; ( d1)node[xshift=0.5cm , yshift=-0.25cm]@xmath50(d2 ) ; ( b ) \(cc ) at ( 1.5,1 ) ; ( o ) at ( 0,0 ) ; ( a ) at ( 3,0 ) ; ( b ) at ( 1.5,2 ) ; ( d1 ) at ( 1.5,-0.4 ) ; ( d2 ) at ( 1.5,0.4 ) ; ( e1 ) at ( 1.9,0.7375 ) ; ( e2 ) at ( 2.6,1.2625 ) ; ( f2 ) at ( 1.1,0.7375 ) ; ( f1 ) at ( 0.4,1.2625 ) ; ( o)(a)(b)(o ) ; ( d1)node[xshift=0.4cm , yshift=0.65cm]@xmath187(d2 ) ; ( cc)node@xmath11 ; ( e1)node[xshift=0.9cm , yshift=0.2cm]@xmath188(e2 ) ; ( cc)node@xmath11 ; ( f1)node[xshift=-0.1cm , yshift=-0.3cm]@xmath189(f2 ) ; ( cc)node@xmath11 ; loop over all elements , @xmath190 : : : enforce positivity of the density and pressure at internal and boundary quadrature points using the limiter described in detail in section [ sec : zh - limiter ] .
: : compute the time - averaged fluxes , @xmath191 , defined in equation .
loop over all edges , @xmath192 : : : for each quadrature point , @xmath193 , on the current edge compute the high - order numerical flux : @xmath194 - \frac{s \| { \bf n}_e \|}{2 } \hspace{-0.6 mm } \left ( q^h \hspace{-0.8 mm } \left({\bf x}_{k}^{+}\right ) - q^h \hspace{-0.8 mm } \left({\bf x}_{k}^{-}\right ) \right),\end{aligned}\ ] ] where @xmath96 is an estimate of the maximum global wave speed . from these numerical flux values , compute the edge - averaged high - order flux : @xmath195 where @xmath62 are the weights in the gauss - legendre numerical quadrature with @xmath63 points .
still on the same edge , also compute the edge - averaged low - order flux : @xmath196 - \frac{s \| \vec{n}_e \|}{2 } \bigl ( { \overline{q}}^{\ , h}_{j_e } - { \overline{q}}^{\ , h}_{i_e } \bigr).\end{aligned}\ ] ] finally , set @xmath197 loop over all elements , @xmath190 : : : let @xmath198 for @xmath180 represent the high - order flux contribution on the three edges of element @xmath11 .
we choose a reordering of the indices 1 , 2 , 3 to the indices @xmath199 , @xmath200 , @xmath201 such that the three flux differences are ordered as follows : @xmath202 we now define the three values @xmath203 , @xmath204 , and @xmath205 and consider four distinct cases .
+ ; ; if @xmath206 , then @xmath207 ; ; if @xmath208 , then set @xmath209 * case 3 . *
; ; if @xmath210 , then set @xmath211 ; ; if @xmath212 , then set @xmath213 + note that in each of these cases the ratios used in the relevant formulas are found by setting the positive contributions on the left - hand side of inequality equal to zero , and then solving for the remaining elements .
this is equivalent to only looking at the worst case scenario where mass is only allowed to flow out of element @xmath11 .
: : define the lax - friedrichs average solution : @xmath214 and do the following .
+ * ( a ) * ; ; loop over all seven cases , @xmath215 , shown in figure [ fig : seven - cases ] and for each case construct the average solution : @xmath216 the seven cases studied here enumerate all of the possible values of @xmath217 , with the exception of @xmath218 , which reduces to updating the element with purely a lax - friedrichs flux . *
( b ) * ; ; for each @xmath201 , determine the largest value of @xmath219 $ ] such that the average pressure as defined by and based on the average state : @xmath220 is positive .
note that the average pressure is always positive when @xmath221 , and if the average pressure is positive for @xmath222 , then the average pressure will be positive for any @xmath223 .
this is because the pressure is a convex function of @xmath224 . *
( c ) * ; ; rescale the edge @xmath225 values ( when compared to the neighboring elements ) based on @xmath224 as follows : @xmath226 loop over all elements , @xmath190 : : : for each of the three edges that make up element @xmath11 determine the damping coefficients , @xmath227 for @xmath180 , as follows : @xmath228 update the cell averages : @xmath229,\ ] ] as well as the high - order moments @xmath230 for @xmath231 , where exact integration is replaced by numerical quadrature .
( a ) \(a ) at ( -1.5cm,-1.cm ) ; ( c ) at ( 0.5cm,-1.0 cm ) ; ( b ) at ( -1.5cm,1.0 cm ) ; ( a )
node[rotate=90,xshift=0.0 cm , yshift=0.2 cm ] @xmath232 ( b )
node[right , rotate=-45,xshift=-0.5 cm , yshift=0.2 cm ] @xmath233 ( c )
node[below ] @xmath234 ( a ) ; ( b ) \(a ) at ( -1.5cm,-1.cm ) ; ( c ) at ( 0.5cm,-1.0 cm ) ; ( b ) at ( -1.5cm,1.0 cm ) ; ( a )
node[rotate=90,xshift=0.0 cm , yshift=0.2 cm ] @xmath232 ( b )
node[right , rotate=-45,xshift=-0.5 cm , yshift=0.2 cm ] @xmath235 ( c )
node[below ] @xmath234 ( a ) ; ( c ) \(a ) at ( -1.5cm,-1.cm ) ; ( c ) at ( 0.5cm,-1.0 cm ) ; ( b ) at ( -1.5cm,1.0 cm ) ; ( a )
node[rotate=90,xshift=0.0 cm , yshift=0.2 cm ] @xmath232 ( b )
node[right , rotate=-45,xshift=-0.5 cm , yshift=0.2 cm ] @xmath233 ( c ) node[below ] @xmath236 ( a ) ; ( d ) \(a ) at ( -1.5cm,-1.cm ) ; ( c ) at ( 0.5cm,-1.0 cm ) ; ( b ) at ( -1.5cm,1.0 cm ) ; ( a )
node[rotate=90,xshift=0.0 cm , yshift=0.2 cm ] @xmath232 ( b )
node[right , rotate=-45,xshift=-0.5 cm , yshift=0.2 cm ] @xmath235 ( c )
node[below ] @xmath236 ( a ) ; ( e ) \(a ) at ( -1.5cm,-1.cm ) ; ( c ) at ( 0.5cm,-1.0 cm ) ; ( b ) at ( -1.5cm,1.0 cm ) ; ( a )
node[rotate=90,xshift=0.0 cm , yshift=0.2 cm ] @xmath237 ( b )
node[right , rotate=-45,xshift=-0.5 cm , yshift=0.2 cm ] @xmath233 ( c )
node[below ] @xmath234 ( a ) ; ( f ) \(a ) at ( -1.5cm,-1.cm ) ; ( c ) at ( 0.5cm,-1.0 cm ) ; ( b ) at ( -1.5cm,1.0 cm ) ; ( a )
node[rotate=90,xshift=0.0 cm , yshift=0.2 cm ] @xmath237 ( b )
node[right , rotate=-45,xshift=-0.5 cm , yshift=0.2 cm ] @xmath235 ( c )
node[below ] @xmath234 ( a ) ; ( g ) \(a ) at ( -1.5cm,-1.cm ) ; ( c ) at ( 0.5cm,-1.0 cm ) ; ( b ) at ( -1.5cm,1.0 cm ) ; ( a )
node[rotate=90,xshift=0.0 cm , yshift=0.2 cm ] @xmath237 ( b )
node[right , rotate=-45,xshift=-0.5 cm , yshift=0.2 cm ] @xmath233 ( c ) node[below ] @xmath236 ( a ) ; extensions to 2d cartesian , 3d cartesian , and 3d tetrahedral mesh elements follow directly from what is presented here , and require considering flux values along each of the edges / faces of a given element .
all of the results presented in this section are implemented in the open - source software package dogpack @xcite . in addition , the positivity limiter described thus far is not designed to handle shocks , and therefore an additional limiter needs to be applied in order to prevent spurious oscillations from developing ( e.g. , in problems that contain shocks but have large densities ) .
there are many options available for this step , but in this work , we supplement the positivity - preserving limiter presented here with the recent shock - capturing limiter developed in @xcite in order to navigate shocks that develop in the solution .
specifically we use the version of this limiter that works with the primitive variables , and we set the parameter @xmath238 . in our experience , this limiter with these parameters offers a good balance between damping oscillations while maintaining sharply refined solutions .
additionally , we point out that extra efficiency can be realized by locally storing quantities computed for the aforementioned positivity - limiter as well as this shock - capturing limiter .
unless otherwise noted , these examples use a cfl number of @xmath239 with a @xmath240 order lax - wendroff time discretization .
all of the examples have the positivity - preserving and shock - capturing limiters turned on . in this section we present some standard one - dimensional problems that can be found in @xcite and references therein .
these problems are designed to break codes that do not have a mechanism to retain positivity of the density and pressure , but with this limiter , we are able to successfully simulate these problems .
our first example is the double rarefaction problem that can be found in @xcite .
this is a riemann problem with initial conditions given by @xmath241 and @xmath242 .
the solution involves two rarefaction waves that move in opposite directions that leave near zero density and pressure values in the post shock regime .
we present our solution on a mesh with a course resolution of @xmath243 , as well as a highly refined solution with @xmath244 .
our results , shown in figure [ fig:1ddoublemach ] , are comparable to those obtained in other works , however the shock - capturing limiter we use is not very diffusive and therefore there is a small amount of oscillation visible in the solution at the lower resolution .
however this oscillation vanishes for the more refined solution .
this example is a simple one - dimensional model of an explosion that is difficult to simulate without aggressive ( or positivity - preserving ) limiting .
the initial conditions involve one central cell with a large amount of energy buildup that is surrounded by a large area of undisturbed air .
these initial conditions are supposed to approximate a delta function of energy .
as time advances , a strong shock waves emanates from this central region and they move in opposite directions .
this leaves the central post - shock regime with near zero density .
the initial conditions are uniform in both density and velocity , with @xmath245 and @xmath246 .
the energy takes on the value @xmath247 in the central cell and @xmath248 in every other cell .
this problem is explored extensively by sedov , and in his classical text gives an exact solution that we use to construct the exact solution underneath our simulation @xcite .
we show our solution in figure [ fig:1dsedov ] , and we point out that our results are quite good especially since we use such a coarse resolution of size @xmath249 .
[ fig:1dsedov ] here , we highlight the fact that this solver is able to operate on both cartesian and unstructured meshes . we first verify the high - order accuracy of the proposed scheme . for problems where
the density and pressure are far away from zero , the limiters proposed in this work `` turn off '' , and therefore have no effect on the solution . in order to investigate the effect of this positivity preserving limiter
, we simulate a smooth problem where the solution has regions that are nearly zero .
this is similar to the smooth test case considered by @xcite , @xcite and @xcite . to this end , we consider initial conditions defined by @xmath250 on a computational domain of @xmath251\times[0,1]$ ] .
we integrate this problem up to a final time of @xmath252 , and compute @xmath253-norm errors against the exact solution given by @xmath254 results for cartesian as well as unstructured meshes are presented in table [ table : convergence ] .
these indicate that the high - order accuracy of the method is not sacrificed when the limiters are turned on . as a final note
, we observe that it appears that in general much higher resolution is required on unstructured meshes before the numerical results enter the asymptotic regime . .
convergence results for the 2d euler problem in section [ subsubsec : convergencetable ] .
all errors are @xmath253 norm errors .
we see that the positivity preserving limiter does not affect the asymptotic convergence rate of the method .
also note that this solution was run only to short time , as in @xcite , @xcite and @xcite , because for this example we must resolve a positive , yet very small density leading to a very large wave - speed and thus a very small permissable time - step . [ cols=">,^,^,^,^,^,^ " , ] in this example we implement a two - dimensional version of the sedov blast wave on the circular domain @xmath255 the bulk of the domain begins with an undisturbed gas , @xmath256 , with uniform density @xmath257 , and near - zero energy @xmath258 .
only the cells at the center of the domain contain a large amount of energy that approximate a delta function . to simulate this
, we introduce a small region at the center of the domain that is radially symmetric ( because the simulation should be radially symmetric ) of the form @xmath259 where @xmath260 is a characteristic length of the mesh .
we present results in figure [ fig:2dsedov ] where we simulate our solution with a total of @xmath261 mesh cells .
this is a common example used to test positivity limiters @xcite .
it involves a mach @xmath262 shock located above a step moving into air that is at rest with @xmath263 and @xmath264 .
the domain this problem is typically solved on is @xmath251\times [ 6,11 ] \cup [ 1,13]\times[0,11]$ ] . the step is the region @xmath251 \times [ 0,6]$ ]
our boundary conditions are transparent everywhere except above the step where they are inflow and on the surface of the step where we used solid wall .
our initial conditions have the shock located above the step at @xmath265 .
the problem is typically run out to @xmath266 , and if a positivity limiter is not used the solution develops negative density and pressure values , which causes the simulation to fail . the solution shown in figure [ fig : shockdiffstruct ]
is run on a @xmath267 cartesian mesh .
next , we run the same problem from [ ex : shockdiffstruct ] , but we discretize space using an unstructured triangular mesh with @xmath268 cells . the results are shown in figure [ fig : shockdiffunstruct ] , and indicate that the unstructured solver behaves similarly to the cartesian one .
this final shock - diffraction test problem is very similar to the previous test problems , however it must be run on an unstructured triangular mesh because the wedge involved in this problem is triangular ( with a @xmath269 degree angle ) @xcite .
our domain for this problem is given by @xmath270\times[0,11 ] \setminus [ 0,3.4 ] \cup [ 0,3.4 ] \times \left[\frac{6.0}{3.4}x , 6.0\right].\ ] ] again the boundary conditions are transparent everywhere except above the step where they are inflow and on the surface of the step where they are reflective solid wall boundary conditions .
in addition , the initial conditions for this problem are also slightly different that those in the previous example . here
, we have a mach @xmath271 shock located above the step at @xmath272 , and undisturbed air in the rest of the domain with @xmath263 and @xmath264 .
this problem is run on an unstructured mesh with a total of @xmath273 cells .
these results are presented in figure [ fig : shockdiffwedge ] .
in this work we developed a novel positivity - preserving limiter for the lax - wendroff discontinuous galerkin ( lxw - dg ) method .
our results are high - order and applicable for unstructured meshes in multiple dimensions .
positivity of the solution is realized by leveraging two separate ideas : the moment limiting work of zhang and shu @xcite , as well as the flux corrected transport work of xu and collaborators @xcite .
the additional shock capturing limiter , which is required to obtain non - oscillatory results , is the one recently developed by the current authors @xcite .
numerical results indicate the robustness of the method , and are promising for future applications to more complicated problems such as the ideal magnetohydrodynamics equations .
future work includes introducing source terms to the solver , as well as pushing these methods to higher orders ( e.g. , @xmath274-order ) , but that requires either ( a ) an expedited way of computing higher derivatives of the solution , or ( b ) rethinking how runge - kutta methods are applied in a modified flux framework ( e.g. , @xcite ) .
p. bochev , d. ridzal , g. scovazzi , and m. shashkov .
formulation , analysis and numerical study of an optimization - based conservative interpolation ( remap ) of scalar fields for arbitrary lagrangian - eulerian methods .
, 230(13):51995225 , 2011 .
b. cockburn , g.e .
karniadakis , and c .- w .
the development of discontinuous galerkin methods . in _ discontinuous galerkin methods ( newport , ri , 1999 )
_ , volume 11 of _ lect .
notes comput .
sci . eng .
_ , pages 350 .
springer , berlin , 2000 .
m. dumbser , d.s .
balsara , e.f .
toro , and c .- d .
munz . a unified framework for the construction of one - step finite volume and discontinuous galerkin schemes on unstructured meshes .
, 227(18):82098253 , 2008 .
ruuth and r.j .
two barriers on strong - stability - preserving time discretization methods . in _ proceedings of the fifth international conference on spectral and high order methods ( icosahom-01 ) ( uppsala )
_ , volume 17 , pages 211220 , 2002 .
seal , q. tang , z. xu , and a.j . christlieb .
an explicit high - order single - stage single - step positivity - preserving finite difference weno method for the compressible euler equations . ,
pages 120 , 2015 .
titarev and e.f .
ader : arbitrary high order godunov approach . in _ proceedings of the fifth international conference on spectral and high order methods ( icosahom-01 ) ( uppsala ) _ , volume 17 , pages 609618 , 2002 . | this work introduces a single - stage , single - step method for the compressible euler equations that is provably positivity - preserving and can be applied on both cartesian and unstructured meshes .
this method is the first case of a single - stage , single - step method that is simultaneously high - order , positivity - preserving , and operates on unstructured meshes .
time - stepping is accomplished via the lax - wendroff approach , which is also sometimes called the cauchy - kovalevskaya procedure , where temporal derivatives in a taylor series in time are exchanged for spatial derivatives . the lax - wendroff discontinuous galerkin ( lxw - dg ) method developed in this work
is formulated so that it looks like a forward euler update but with a high - order time - extrapolated flux . in particular , the numerical flux used in this work
is a linear combination of a low - order positivity - preserving contribution and a high - order component that can be damped to enforce positivity of the cell averages for the density and pressure for each time step .
in addition to this flux limiter , a moment limiter is applied that forces positivity of the solution at finitely many quadrature points within each cell .
the combination of the flux limiter and the moment limiter guarantees positivity of the cell averages from one time - step to the next . finally , a simple shock capturing limiter that uses the same basic technology as the moment limiter is introduced in order to obtain non - oscillatory results .
the resulting scheme can be extended to arbitrary order without increasing the size of the effective stencil .
we present numerical results in one and two space dimensions that demonstrate the robustness of the proposed scheme . |
shannon s entropy quantifies information @xcite .
it measures how much uncertainty an observer has about an event being produced by a random system .
another important concept in the theory of information is the mutual information @xcite .
it measures how much uncertainty an observer has about an event in a random system * x * after observing an event in a random system * y * ( or vice - versa ) .
mutual information is an important quantity because it quantifies not only linear and non - linear interdependencies between two systems or data sets , but also is a measure of how much information two systems exchange or two data sets share . due to these characteristics
, it became a fundamental quantity to understand the development and function of the brain @xcite , to characterise @xcite and model complex systems @xcite or chaotic systems , and to quantify the information capacity of a communication system @xcite . when constructing a model of a complex system ,
the first step is to understand which are the most relevant variables to describe its behaviour .
mutual information provides a way to identify those variables @xcite .
however , the calculation of mutual information in dynamical networks or data sets faces three main difficulties@xcite .
mutual information is rigorously defined for random memoryless processes , only .
in addition , its calculation involves probabilities of significant events and a suitable space where probability is calculated .
the events need to be significant in the sense that they contain as much information about the system as possible .
but , defining significant events , for example the fact that a variable has a value within some particular interval , is a difficult task because the interval that provides significant events is not always known .
finally , data sets have finite size .
this prevents one from calculating probabilities correctly . as a consequence , mutual information
can often be calculated with a bias , only @xcite . in this work ,
we show how to calculate the amount of information exchanged per unit of time [ eq .
( [ mir_introduction ] ) ] , the so called mutual information rate ( mir ) , between two arbitrary nodes ( or group of nodes ) in a dynamical network or between two data sets .
each node representing a d - dimensional dynamical system with @xmath0 state variables .
the trajectory of the network considering all the nodes in the full phase space is called `` attractor '' and represented by @xmath1 .
then , we propose an alternative method , similar to the ones proposed in refs .
@xcite , to calculate significant upper and lower bounds for the mir in dynamical networks or between two data sets , in terms of lyapunov exponents , expansion rates , and capacity dimension .
these quantities can be calculated without the use of probabilistic measures . as possible applications of our bounds calculation
, we describe the relationship between synchronisation and the exchange of information in small experimental networks of coupled double - scroll circuits . in previous works of refs .
@xcite , we have proposed an upper bound for the mir in terms of the positive conditional lyapunov exponents of the synchronisation manifold . as a consequence , this upper bound could only be calculated in special complex networks that allow the existence of complete synchronisation . in the present work
, the proposed upper bound can be calculated to any system ( complex networks and data sets ) that admits the calculation of lyapunov exponents .
we assume that an observer can measure only one scalar time series for each one of two chosen nodes .
these two time series are denoted by @xmath2 and @xmath3 and they form a bidimensional set @xmath4 , a projection of the `` attractor '' into a bidimensional space denoted by @xmath5 . to calculate the mir in higher - dimensional projections @xmath5 , see supplementary information .
assume that the space @xmath5 is coarse - grained in a square grid of @xmath6 boxes with equal sides @xmath7 , so @xmath8 .
mutual information is defined in the following way @xcite . given two random variables , * x * and * y * , each one produces events @xmath9 and @xmath10 with probabilities @xmath11 and @xmath12 , respectively , the joint probability between these events is represented by @xmath13 .
then , mutual information is defined as @xmath14 @xmath15 = @xmath16}$ ] , @xmath17 = @xmath18}$ ] , and @xmath19}$ ] . for simplification in our notation for the probabilities ,
we drop the subindexes @xmath20 , @xmath21 , and @xmath22 , by making @xmath23 , @xmath24 , and @xmath25 . when using eq .
( [ is ] ) to calculate the mutual information between the dynamical variables @xmath2 and @xmath3 , the probabilities appearing in eq .
( [ is ] ) are defined such that @xmath26 is the probability of finding points in a column @xmath9 of the grid , @xmath27 of finding points in the row @xmath10 of the grid , and @xmath28 the probability of finding points where the column @xmath9 meets the line @xmath10 of the grid .
the mir was firstly introduced by shannon @xcite as a `` rate of actual transmission '' @xcite and later more rigorously redefined in refs .
it represents the mutual information exchanged between two dynamical variables ( correlated ) per unit of time . to simplify the calculation of the mir ,
the two continuous dynamical variables are transformed into two discrete symbolic sequences @xmath2 and @xmath3 .
then , the mir is defined by @xmath29 where @xmath30 represents the usual mutual information between the two sequences @xmath2 and @xmath3 , calculated by considering words of length @xmath31 .
the mir is a fundamental quantity in science .
its maximal value gives the information capacity between any two sources of information ( no need for stationarity , statistical stability , memoryless ) @xcite .
therefore , alternative approaches for its calculation or for the calculation of bounds of it are of vital relevance . due to the limit to infinity in eq .
( [ original_mir ] ) and because it is defined from probabilities , the mir is not easy to be calculated especially if one wants to calculate it from ( chaotic ) trajectories of a large complex network or data sets . the difficulties faced to estimate the mir from dynamical systems and networks are similar to the ones faced in the calculation of the kolmogorov - sinai entropy , @xmath32 @xcite , ( shannon s entropy per unit of time ) .
because of these difficulties , the upper bound for @xmath32 proposed by ruelle @xcite in terms of the lyapunov exponents and valid for smooth dynamical systems ( @xmath33 , where @xmath34 represent all the @xmath9 positive lyapunov exponents ) or the pesin s equality @xcite ( @xmath35 ) proved in ref .
@xcite to be valid for the large class of systems that possess a srb measure , became so important in the theory of dynamical systems .
our upper bound [ eq .
( [ i_c ] ) ] is a result equivalent to the work of ruelle .
one of the main results of this work ( whose derivation can be seen in sec .
[ methods_mir ] ) is to show that , in dynamical networks or data sets with fast decay of correlation , @xmath36 in eq .
( [ is ] ) represents the amount of mutual information between @xmath2 and @xmath3 produced within a special time interval @xmath37 , where @xmath37 represents the time for the dynamical network ( or data sets ) to lose its memory from the initial state or the correlation to decay to zero .
correlation in this work is not the usual linear correlation , but a non - linear correlation defined in terms of the evolution of spatial probabilities , the quantity @xmath38 in sec . [ mixing ] .
therefore , the mutual information rate ( mir ) , between the dynamical variables @xmath2 and @xmath3 ( or two data sets ) can be estimated by @xmath39 in systems that present sensitivity to initial conditions , e.g. chaotic systems , predictions are only possible for times smaller than this time @xmath37 .
this time has other meanings .
it is the expected time necessary for a set of points belonging to an @xmath7-square box in @xmath5 to spread over @xmath40 and it is of the order of the shortest poincar return time for a point to leave a box and return to it @xcite .
it can be estimated by @xmath41}. \label{t}\ ] ] where @xmath42 is the largest positive lyapunov exponent measured in @xmath40 .
chaotic systems present the mixing property ( see sec . [ mixing ] ) , and as a consequence the correlation @xmath43 always decays to zero , surely after an infinitely long time .
the correlation of chaotic systems can also decay to zero for sufficiently large but finite @xmath44 ( see supplementary information ) .
@xmath37 can be interpreted to be the minimum time required for a system to satisfy the conditions to be considered mixing .
some examples of physical systems that are proved to be mixing and have exponentially fast decay of correlation are nonequilibrium steady - state @xcite , lorenz gases ( models of diffusive transport of light particles in a network of heavier particles ) @xcite , and billiards @xcite .
an example of a `` real world '' physical complex system that presents exponentially fast decay of correlation is plasma turbulence @xcite .
we do not expect that data coming from a `` real world '' complex system is rigorously mixing and has an exponentially fast decay of correlation .
but , we expect that the data has a sufficiently fast decay of correlation ( e.g. stretched exponential decay or polynomially fast decays ) , implying that the system has sufficiently high sensitivity to initial conditions and as a consequence @xmath45 , for a reasonably small and finite time @xmath44 .
the other two main results of our work are presented in eqs .
( [ i_c_intro ] ) and ( [ icl_intro ] ) , whose derivations are presented in sec .
[ methods_bounds ] .
the upper bound for the mir is given by @xmath46 where @xmath42 and @xmath47 ( positive defined ) represent the largest and the second largest lyapunov exponent measured in @xmath40 , if both exponents are positive .
if the @xmath9-largest exponent is negative , then we set @xmath48 .
if the set @xmath40 represents a periodic orbit , @xmath49 , and therefore there is no information being exchanged .
the quantity @xmath50 is defined as @xmath51 where @xmath52 is the number of boxes that would be covered by fictitious points at time @xmath37 . at time @xmath53 , these fictitious points are confined in an @xmath7-square box .
they expand not only exponentially fast in both directions according to the two positive lyapunov exponents , but expand forming a compact set , a set with no `` holes '' . at @xmath44 , they spread over @xmath40 .
the lower bound for the mir is given by @xmath54 where @xmath55 represents the capacity dimension of the set @xmath40 @xmath56 , \label{tildad}\ ] ] where @xmath57 represents the number of boxes in @xmath5 that are occupied by points of @xmath40 .
@xmath50 is defined in a way similar to the capacity dimension , thought it is not the capacity dimension .
in fact , @xmath58 , because @xmath55 measures the change in the number of occupied boxes in @xmath5 as the space resolution varies , whereas @xmath50 measures the relative number of boxes with a certain fixed resolution @xmath7 that would be occupied by the fictitious points ( in @xmath5 ) after being iterated for a time @xmath37 . as a consequence , the empty space in @xmath5 that is not occupied by @xmath40 does not contribute to the calculation of @xmath55 , whereas it contributes to the calculation of the quantity @xmath50 .
in addition , @xmath59 ( for any @xmath7 ) , because while the fictitious points form a compact set expanding with the same ratio as the one for which the real points expand ( ratio provided by the lyapunov exponents ) , the real set of points @xmath40 might not occupy many boxes .
denote by @xmath60 a mixing transformation that represents how a point @xmath61 is mapped after a time @xmath37 into @xmath40 , and let @xmath62 to represent the probability of finding a point of @xmath40 in @xmath63 ( natural invariant density ) .
let @xmath64 represent a region in @xmath5 .
then , @xmath65 , for @xmath66 represents the probability measure of the region @xmath64 .
given two square boxes @xmath67 and @xmath68 , if @xmath69 is a mixing transformation , then for a sufficiently large @xmath37 , we have that the correlation @xmath70 - \mu[i^{\prime}_1]\mu[i^{\prime}_2]$ ] , decays to zero , the probability of having a point in @xmath64 that is mapped to @xmath71 is equal to the probability of being in @xmath64 times the probability of being in @xmath71 .
that is typically what happens in random processes . if the measure @xmath72 is invariant , then @xmath73=\mu(\sigma_{\omega})$ ] . mixing and ergodic systems produce measures that are invariant .
we consider that the dynamical networks or data sets to be analysed present either the mixing property or have fast decay of correlations , and their probability measure is time invariant . if a system that is mixing for a time interval @xmath37 is observed ( sampled ) once every time interval @xmath37 , then the probabilities generated by these snapshot observations behave as if they were independent , and the system behaves as if it were a random process .
this is so because if a system is mixing for a time interval @xmath37 , then the correlation @xmath38 decays to zero for this time interval . for systems that have some decay of correlation
, surely the correlation decays to zero after an infinite time interval .
but , this time interval can also be finite , as shown in supplementary information .
consider now that we have experimental points and they are sampled once every time interval @xmath37 .
the probability @xmath74 of the sampled trajectory to follow a given itinerary , for example to fall in the box with coordinates @xmath75 and then be iterated to the box @xmath76 depends exclusively on the probabilities of being at the box @xmath75 , represented by @xmath77 , and being at the box @xmath76 , represented by @xmath78 .
therefore , for the sampled trajectory , @xmath79 .
analogously , the probability @xmath80 of the sampled trajectory to fall in the column ( or line ) @xmath9 of the grid and then be iterated to the column ( or line ) @xmath10 is given by @xmath81 .
the mir of the experimental non - sampled trajectory points can be calculated from the mutual information @xmath82 of the sampled trajectory points that follow itineraries of length @xmath31 : @xmath83 due to the absence of correlations of the sampled trajectory points , the mutual information for these points following itineraries of length @xmath31 can be written as @xmath84 , \label{original_mir_sampled1}\ ] ] where @xmath85 = @xmath86}$ ] , @xmath87 = @xmath88}$ ] , and @xmath89}$ ] , and @xmath90 , @xmath91 , and @xmath77 represent the probability of the sampled trajectory points to fall in the line @xmath9 of the grid , in the column @xmath10 of the grid , and in the box @xmath75 of the grid , respectively .
due to the time invariance of the set @xmath40 assumed to exist , the probability measure of the non - sampled trajectory is equal to the probability measure of the sampled trajectory . if a system that has a time invariant measure is observed ( sampled ) once every time interval @xmath37 , the observed set has the same natural invariant density and probability measure of the original set . as a consequence ,
if @xmath40 has a time invariant measure , the probabilities @xmath26 , @xmath27 , and @xmath28 ( used to calculate @xmath36 ) are equal to @xmath90 , @xmath91 , and @xmath77 . consequently , @xmath92 , @xmath93 , and @xmath94 , and therefore @xmath95 . substituting into eq .
( [ original_mir_sampled ] ) , we finally arrive to @xmath96 where @xmath36 between two nodes is calculated from eq .
( [ is ] ) .
therefore , in order to calculate the mir , we need to estimate the time @xmath37 for which the correlation of the system approaches zero and the probabilities @xmath26 , @xmath27 , @xmath28 of the experimental non - sampled experimental points to fall in the line @xmath9 of the grid , in the column @xmath10 of the grid , and in the box @xmath75 of the grid , respectively .
consider that our attractor @xmath1 is generated by a 2d expanding system that possess 2 positive lyapunov exponents @xmath42 and @xmath47 , with @xmath99 .
imagine a box whose sides are oriented along the orthogonal basis used to calculate the lyapunov exponents .
then , points inside the box spread out after a time interval @xmath101 to @xmath102 along the direction from which @xmath42 is calculated . at @xmath44 , @xmath103 , which provides @xmath37 in eq .
( [ t ] ) , since @xmath104 .
these points spread after a time interval @xmath101 to @xmath105 along the direction from which @xmath47 is calculated .
after an interval of time @xmath44 , these points spread out over the set @xmath40 .
we require that for @xmath106 , the distance between these points only increases : the system is expanding .
imagine that at @xmath44 , fictitious points initially in a square box occupy an area of @xmath107 .
then , the number of boxes of sides @xmath7 that contain fictitious points can be calculated by @xmath108 . from eq .
( [ t ] ) , @xmath109 , since @xmath8 .
we denote with a lower - case format , the probabilities @xmath110 , @xmath111 , and @xmath112 with which fictitious points occupy the grid in @xmath5 .
if these fictitious points spread uniformly forming a compact set whose probabilities of finding points in each fictitious box is equal , then @xmath113 ( @xmath114 ) , @xmath115 , and @xmath116 .
let us denote the shannon s entropy of the probabilities @xmath112 , @xmath110 and @xmath111 as @xmath117 , @xmath118 , and @xmath119 .
the mutual information of the fictitious trajectories after evolving a time interval @xmath37 can be calculated by @xmath120 . since , @xmath121 and @xmath116 , then @xmath122 . at @xmath44 , we have that @xmath109 and @xmath123 , leading us to @xmath124 .
therefore , defining , @xmath125 , we arrive at @xmath126 .
we defining @xmath50 as @xmath127 where @xmath52 being the number of boxes that would be covered by fictitious points at time @xmath37 . at time
@xmath53 , these fictitious points are confined in an @xmath7-square box .
they expand not only exponentially fast in both directions according to the two positive lyapunov exponents , but expand forming a compact set , a set with no `` holes '' . at @xmath44 , they spread over @xmath40 . using @xmath128 and @xmath129 in eq .
( [ d ] ) , we arrive at @xmath130 , and therefore , we can write that @xmath131 to calculate the maximal possible mir , of a random independent process , we assume that the expansion of points is uniform only along the columns and lines of the grid defined in the space @xmath5 , i.e. , @xmath132 , ( which maximises @xmath133 and @xmath134 ) , and we allow @xmath28 to be not uniform ( minimising @xmath135 ) for all @xmath9 and @xmath10 , then @xmath136}. \label{is_lower}\ ] ] since @xmath137 , dividing @xmath138 by @xmath139 , taking the limit of @xmath140 , and reminding that the information dimension of the set @xmath40 in the space @xmath5 is defined as @xmath141=@xmath142}}{\log{(\epsilon)}}$ ] , we obtain that the mir is given by @xmath143 since @xmath144 ( for any value of @xmath7 ) , then @xmath145 , which means that a lower bound for the maximal mir [ provided by eq .
( [ almost_true1 ] ) ] is given by @xmath146 but @xmath58 ( for any value of @xmath7 ) , and therefore @xmath97 is an upper bound for @xmath98 . to show why @xmath97 is an upper bound for the maximal possible mir , assume that the real points @xmath40 occupy the space @xmath5 uniformly .
if @xmath147 , there are many boxes being occupied .
it is to be expected that the probability of finding a point in a line or column of the grid is @xmath148 , and @xmath149 .
in such a case , @xmath150 , which implies that @xmath151 . if @xmath152 , there are only few boxes being sparsely occupied .
the probability of finding a point in a line or column of the grid is @xmath153 , and @xmath149 .
there are @xmath57 lines and columns being occupied by points in the grid .
in such a case , @xmath154 . comparing with @xmath122 , and since @xmath155 and @xmath59 , then we conclude that @xmath156 , which implies that @xmath151 .
notice that if @xmath157 and @xmath158 , then @xmath159 . in order to extend our approach for the treatment of data sets coming from networks whose equations of motion
are unknown , or for higher - dimensional networks and complex systems which might be neither rigorously chaotic nor fully deterministic , or for experimental data that contains noise and few sampling points , we write our bounds in terms of expansion rates defined in this work by @xmath160 } , \label{define_exp_rates}\ ] ] where we consider @xmath161 .
@xmath162 measures the largest growth rate of nearby points . in practice
, it is calculated by @xmath163 , with @xmath164 representing the largest distance between pair of points in an @xmath7-square box @xmath9 and @xmath165 representing the largest distance between pair of the points that were initially in the @xmath7-square box but have spread out for an interval of time @xmath101 .
@xmath166 measures how an area enclosing points grows . in practice
, it is calculated by @xmath167 , with @xmath168 representing the area occupied by points in an @xmath7-square box , and @xmath169 the area occupied by these points after spreading out for a time interval @xmath101 .
there are @xmath57 boxes occupied by points which are taken into consideration in the calculation of @xmath170 . an order-@xmath171 expansion rate , @xmath172 , measures on average how a hypercube of dimension @xmath171 exponentially grows after an interval of time @xmath101 .
so , @xmath173 measures the largest growth rate of nearby points , a quantity closely related to the largest finite - time lyapunov exponent @xcite . and
@xmath174 measures how an area enclosing points grows , a quantity closely related to the sum of the two largest positive lyapunov exponents . in terms of expansion rates , eqs .
( [ t ] ) and ( [ i_c ] ) read @xmath175}$ ] and @xmath176 , respectively , and eqs .
( [ d ] ) and ( [ icl ] ) read @xmath177 and @xmath178 , respectively . from the way we have defined expansion rates
, we expect that @xmath179 . because of the finite time interval and the finite size of the regions of points considered , regions of points that present large derivatives , contributing largely to the lyapunov exponents , contribute less to the expansion rates . if a system has constant derivative ( hyperbolic ) and has constant natural measure , then @xmath180 .
there are many reasons for using expansion rates in the way we have defined them in order to calculate bounds for the mir .
firstly , because they can be easily experimentally estimated whereas lyapunov exponents demand huge computational efforts .
secondly , because of the macroscopic nature of the expansion rates , they might be more appropriate to treat data coming from complex systems that contains large amounts of noise , data that have points that are not ( arbitrarily ) close as formally required for a proper calculation of the lyapunov exponents .
thirdly , expansion rates can be well defined for data sets containing very few data points : the fewer points a data set contains , the larger the regions of size @xmath7 need to be and the shorter the time @xmath37 is .
finally , expansion rates are defined in a similar way to finite - time lyapunov exponents and thus some algorithms used to calculate lyapunov exponents can be used to calculate our defined expansion rates .
to illustrate the use of our bounds , we consider the following two bidirectionally coupled maps@xmath181 where @xmath182 $ ] . if @xmath183 , the map is piecewise - linear and quadratic , otherwise .
we are interested in measuring the exchange of information between @xmath184 and @xmath185 .
the space @xmath5 is a square of sides 1 .
the lyapunov exponents measured in the space @xmath5 are the lyapunov exponents of the set @xmath40 that is the chaotic attractor generated by eqs .
( [ network_maps ] ) .
the quantities @xmath186 , @xmath97 , and @xmath98 are shown in fig .
[ figure4 ] as we vary @xmath187 for @xmath183 ( a ) and @xmath188 ( b ) .
we calculate @xmath36 using in eq .
( [ is ] ) the probabilities @xmath28 in which points from a trajectory composed of @xmath189 samples fall in boxes of sides @xmath7=1/500 and the probabilities @xmath26 and @xmath27 that the points visit the intervals @xmath190 of the variable @xmath191 or @xmath192 of the variable @xmath193 , respectively , for @xmath194 . when computing @xmath186 , the quantity @xmath37 was estimated by eq .
( [ t ] ) .
indeed for most values of @xmath187 , @xmath195 and @xmath196 . for @xmath197
there is no coupling , and therefore the two maps are independent from each other .
there is no information being exchanged .
in fact , @xmath49 and @xmath198 in both figures , since @xmath199 , meaning that the attractor @xmath40 fully occupies the space @xmath5 .
this is a remarkable property of our bounds : to identify that there is no information being exchanged when the two maps are independent .
complete synchronisation is achieved and @xmath97 is maximal , for @xmath200 ( a ) and for @xmath201 ( b ) . a consequence of the fact that @xmath202 , and therefore , @xmath203 .
the reason is because for this situation this coupled system is simply the shift map , a map with constant natural measure ; therefore @xmath204 and @xmath28 are constant for all @xmath9 and @xmath10 . as usually happens when one estimates the mutual information by partitioning the phase space with a grid having a finite resolution and data sets possessing a finite number of points , @xmath36 is typically larger than zero , even when there is no information being exchanged ( @xmath197 ) .
even when there is complete synchronisation , we find non - zero off - diagonal terms in the matrix for the joint probabilities causing @xmath36 to be smaller than it should be .
due to numerical errors , @xmath205 , and points that should be occupying boxes with two corners exactly along a diagonal line in the subspace @xmath5 end up occupying boxes located off - diagonal and that have at least three corners off - diagonal .
the estimation of the lower bound @xmath98 suffers from the same problems .
our upper bound @xmath97 is calculated assuming that there is a fictitious dynamics expanding points ( and producing probabilities ) not only exponentially fast but also uniformly .
the `` experimental '' numerical points from eqs .
( [ network_maps ] ) expand exponentially fast , but not uniformly .
most of the time the trajectory remains in 4 points : ( 0,0 ) , ( 1,1 ) , ( 1,0 ) , ( 0,1 ) . that is the main reason of why @xmath97 is much larger than the estimated real value of the @xmath206 , for some coupling strengths . if a two nodes in a dynamical network , such as two neurons in a brain , behave in the same way the fictitious dynamics does , these nodes would be able to exchange the largest possible amount of information .
we would like to point out that one of the main advantages of calculating upper bounds for the mir ( @xmath186 ) using eq .
( [ i_c ] ) instead of actually calculating @xmath186 is that we can reproduce the curves for @xmath97 using much less number of points ( 1000 points ) than the ones ( @xmath189 ) used to calculate the curve for @xmath186 .
if @xmath183 , @xmath207 can be calculated since @xmath208 and @xmath209 .
we illustrate our approach for the treatment of data sets using a network formed by an inductorless version of the double - scroll circuit @xcite .
we consider four networks of bidirectionally diffusively coupled circuits .
topology i represents two bidirectionally coupled circuits , topology ii , three circuits coupled in an open - ended array , topology iii , four circuits coupled in an open - ended array , and topology iv , coupled in an closed array .
we choose two circuits in the different networks ( one connection apart ) and collect from each circuit a time - series of 79980 points , with a sampling rate of @xmath210 samples / s .
the measured variable is the voltage across one of the circuit capacitors , which is normalised in order to make the space @xmath5 to be a square of sides 1 .
such normalisation does not alter the quantities that we calculate .
the following results provide the exchange of information between these two chosen circuits .
the values of @xmath7 and @xmath101 used to course - grain the space @xmath5 and to calculate @xmath174 in eq .
( [ define_exp_rates ] ) are the ones that minimises @xmath211 and at the same time satisfy @xmath212 , where @xmath213 represents the number of fictitious boxes covering the set @xmath40 in a compact fashion , when @xmath44 .
this optimisation excludes some non - significant points that make the expansion rate of fictitious points to be much larger than it should be . in other words , we require that @xmath174 describes well the way most of the points spread .
we consider that @xmath101 used to calculate @xmath214 in eq .
( [ define_exp_rates ] ) is the time for points initially in an @xmath7-side box to spread to 0.8@xmath215 .
that guarantee that nearby points in @xmath40 are expanding in both directions within the time interval @xmath216 $ ] .
using @xmath217 produces already similar results . if @xmath218 , the set @xmath40 might not be only expanding .
@xmath37 might be overestimated .
results for experimental networks of double - scroll circuits . on the left - side upper corner pictograms represent how the circuits ( filled circles ) are bidirectionally coupled . @xmath219 as ( green online ) filled circles , @xmath97 as the ( red online ) thick line , and @xmath98 as the ( brown online ) squares , for a varying coupling resistance @xmath220 .
the unit of these quantities shown in these figures is ( kbits / s ) .
( a ) topology i , ( b ) topology ii , ( c ) topology iii , and ( d ) topology iv . in all figures , @xmath55 increases smoothly from 1.25 to 1.95 as @xmath220 varies from 0.1k@xmath5 to 5k@xmath5 .
the line on the top of the figure represents the interval of resistance values responsible to induce almost synchronisation ( as ) and phase synchronisation ( ps).,width=264,height=264 ] @xmath36 has been estimated by the method in ref .
since we assume that the space @xmath5 where mutual information is being measured is 2d , we will compare our results by considering in the method of ref .
@xcite a 2d space formed by the two collected scalar signals . in the method of ref .
@xcite the phase space is partitioned in regions that contain 30 points of the continuous trajectory . since that these regions do not have equal areas ( as it is done to calculate @xmath97 and @xmath98 ) , in order to estimate @xmath37 we need to imagine a box of sides @xmath221 , such that its area @xmath222 contains in average 30 points . the area occupied by the set @xmath40 is approximately given by @xmath223 , where @xmath57 is the number of occupied boxes .
assuming that the 79980 experimental data points occupy the space @xmath5 uniformly , then on average 30 points would occupy an area of @xmath224 .
the square root of this area is the side of the imaginary box that would occupy 30 points .
so , @xmath225 .
then , in the following , the `` exact '' value of the mir will be considered to be given by @xmath219 , where @xmath226 is estimated by @xmath227 .
the three main characteristics of the curves for the quantities @xmath219 , @xmath97 , and @xmath98 ( appearing in fig . [ figure4_letter00 ] ) with respect to the coupling strength are that ( i ) as the coupling resistance becomes smaller , the coupling strength connecting the circuits becomes larger , and the level of synchronisation increases followed by an increase in @xmath219 , @xmath97 , and @xmath98 , ( ii ) all curves are close , ( iii ) and as expected , for most of the resistance values , @xmath228 and @xmath229 .
the two main synchronous phenomena appearing in these networks are almost synchronisation ( as ) @xcite , when the circuits are almost completely synchronous , and phase synchronisation ( ps ) @xcite . for the circuits considered in fig .
[ figure4_letter00 ] , as appears for the interval @xmath230 $ ] and ps appears for the interval @xmath231 $ ] . within this region of resistance values the exchange of information between the circuits becomes large .
ps was detected by using the technique from refs .
@xcite . to analytically demonstrate that the quantities @xmath97 and @xmath186 can be well calculated in stochastic systems , we consider the following stochastic dynamical toy model illustrated in fig . [ toy_model ] . in
it points within a small box of sides @xmath7 ( represented by the filled square in fig .
[ toy_model](a ) ) located in the centre of the subspace @xmath5 are mapped after one iteration of the dynamics to 12 other neighbouring boxes .
some points remain in the initial box .
the points that leave the initial box go to 4 boxes along the diagonal line and 8 boxes off - diagonal along the transverse direction .
boxes along the diagonal are represented by the filled squares in fig .
[ toy_model](b ) and off - diagonal boxes by filled circles . at the second iteration ,
the points occupy other neighbouring boxes , as illustrated in fig .
[ toy_model](c ) , and at the time @xmath232 the points do not spread any longer , but are somehow reinjected inside the region of the attractor .
we consider that this system is completely stochastic , in the sense that no one can precisely determine the location of where an initial condition will be mapped .
the only information is that points inside a smaller region are mapped to a larger region . at the iteration @xmath31
, there will be @xmath233 boxes occupied along the diagonal ( filled squares in fig .
[ toy_model ] ) and @xmath234 ( filled circles in fig .
[ toy_model ] ) boxes occupied off - diagonal ( along the transverse direction ) , where @xmath235 for @xmath236=0 , and @xmath237 for @xmath238 and @xmath239 .
@xmath240 is a small number of iterations representing the time difference between the time @xmath37 for the points in the diagonal to reach the boundary of the space @xmath5 and the time for the points in the off - diagonal to reach this boundary .
the border effect can be ignored when the expansion along the diagonal direction is much faster than along the transverse direction . at the iteration @xmath31
, there will be @xmath241 boxes occupied by points . in the following calculations we consider that @xmath242 .
we assume that the subspace @xmath5 is a square whose sides have length 1 , and that @xmath100 , so @xmath104 . for @xmath243
, the attractor does not grow any longer along the off - diagonal direction .
the time @xmath232 , for the points to spread over the attractor @xmath1 , can be calculated by the time it takes for points to visit all the boxes along the diagonal .
thus , we need to satisfy @xmath244 . ignoring the 1 appearing in the expression for @xmath245 due to the initial box in the estimation for the value of @xmath37
, we arrive that @xmath246 .
this stochastic system is discrete . in order to take into consideration the initial box in the calculation of @xmath37 ,
we pick the first integer that is larger than @xmath247 , leading @xmath37 to be the largest integer that satisfies @xmath248 the largest lyapunov exponent or the order-1 expansion rate of this stochastic toy model can be calculated by @xmath249 , which take us to @xmath250 therefore , eq . ( [ tm_t ] ) can be rewritten as @xmath251 .
the quantity @xmath50 can be calculated by @xmath252 , with @xmath232 .
neglecting @xmath253 and the 1 appearing in @xmath254 due to the initial box , we have that @xmath255 $ ] . substituting in the definition of @xmath50
, we obtain @xmath256 . using @xmath37 from eq .
( [ tm_t ] ) , we arrive at @xmath257 where @xmath258 placing @xmath50 and @xmath42 in @xmath259 , give us @xmath260 let us now calculate @xmath186 . ignoring the border effect , and assuming that the expansion of points is uniform , then @xmath261 and @xmath262 . at the iteration @xmath232 , we have that @xmath263 . since @xmath264 $ ] , we can write that @xmath265 . placing @xmath37 from eq .
( [ tm_t ] ) into @xmath36 takes us to @xmath266 . finally , dividing @xmath36 by @xmath37
, we arrive that @xmath267 \nonumber \\ & = & \log{(2)}(1-r ) .
\label{tm_ist}\end{aligned}\ ] ] as expected from the way we have constructed this model , eq .
( [ tm_ist ] ) and ( [ tm_ic ] ) are equal and @xmath268 .
had we included the border effect in the calculation of @xmath97 , denote the value by @xmath269 , we would have typically obtained that @xmath270 , since @xmath47 calculated considering a finite space @xmath5 would be either smaller or equal than the value obtained by neglecting the border effect . had we included the border effect in the calculation of @xmath186 , denote the value by @xmath271 , typically we would expect that the probabilities @xmath28 would not be constant .
that is because the points that leave the subspace @xmath5 would be randomly reinjected back to @xmath5 .
we would conclude that @xmath272 .
therefore , had we included the border effect , we would have obtained that @xmath273 . the way we have constructed this stochastic toy model results in @xmath274 .
this is because the spreading of points along the diagonal direction is much faster than the spreading of points along the off - diagonal transverse direction .
in other words , the second largest lyapunov exponent , @xmath47 , is close to zero .
stochastic toy models which produce larger @xmath47 , one could consider that the spreading along the transverse direction is given by @xmath275 , with @xmath276 $ ] . in terms of the order-1 expansion rate , @xmath173 , our quantities read @xmath176 , @xmath277}$ ] , and @xmath178 .
in order to show that our expansion rate can be used to calculate these quantities , we consider that the experimental system is uni - dimensional and has a constant probability measure .
additive noise is assumed to be bounded with maximal amplitude @xmath278 , and having constant density .
our order-1 expansion rate is defined as @xmath279}. \label{define_exp_rates_sup}\ ] ] where @xmath280 measures the largest growth rate of nearby points .
since all it matters is the largest distance between points , it can be estimated even when the experimental data set has very few data points . since , in this example , we consider that the experimental noisy points have constant uniform probability distribution , @xmath281 can be calculated by @xmath282}. \label{define_exp_rates_sup1}\ ] ] where @xmath283 represents the largest distance between pair of experimental noisy points in an @xmath7-square box and @xmath284 represents the largest distance between pair of the points that were initially in the @xmath7-square box but have spread out for an interval of time @xmath101 . the experimental system ( without noise )
is responsible to make points that are at most @xmath164 apart from each other to spread to at most to @xmath165 apart from each other .
this points spread out exponentially fast according to the largest positive lyapunov exponent @xmath42 by @xmath285 substituting eq .
( [ delta ] ) in ( [ define_exp_rates_sup1 ] ) , and expanding @xmath286 to first order , we obtain that @xmath287 , and therefore , our expansion rate can be used to estimate lyapunov exponents .
as rigorously shown in @xcite , the decay with time of the correlation , @xmath43 , is proportional to the decay with time of the density of the first poincar recurrences , @xmath288 , which measures the probability with which a trajectory returns to an @xmath7-interval after @xmath101 iterations .
therefore , if @xmath288 decays with @xmath101 , for example exponentially fast , @xmath43 will decay with @xmath101 exponentially fast , as well .
the relationship between @xmath43 and @xmath289 can be simply understood in chaotic systems with one expanding direction ( one positive lyapunov exponent ) . as shown in @xcite , the `` local '' decay of correlation ( measured in the @xmath7-interval )
is given by @xmath290 , where @xmath291 is the probability measure of a chaotic trajectory to visit the @xmath7-interval .
consider the shift map @xmath292 .
for this map , @xmath293 and there are an infinite number of possible intervals that makes @xmath294 , for a finite @xmath101 .
these intervals are the cells of a markov partition . as recently demonstrated by [ p. pinto , i. labouriau , m. s. baptista ] , in piecewise - linear systems as the shift map , if @xmath7 is a cell in an order-@xmath101 markov partition and @xmath295 , then @xmath296 and by the way a markov partition is constructed we have that @xmath297 .
since that @xmath298 , we arrive at that @xmath299 , for a special finite time @xmath101 .
notice that @xmath297 can be rewritten as @xmath300 .
since for this map , the largest lyapunov exponent is equal to @xmath208 , then @xmath301 , which is exactly equal to the quantity @xmath37 , the time interval responsible to make the system to lose its memory from the initial condition and that can be calculated by the time that makes points inside an initial @xmath7-interval to spread over the whole phase space , in this case @xmath302 $ ] .
imagine a network formed by @xmath303 coupled oscillators .
uncoupled , each oscillator possesses a certain amount of positive lyapunov exponents , one zero , and the others are negative .
each oscillator has dimension @xmath0 .
assume that the only information available from the network are two @xmath304 dimensional measurements , or a scalar signal that is reconstructed to a @xmath304-dimensional embedding space .
so , the subspace @xmath40 has dimension @xmath305 , and each subspace of a node ( or group of nodes ) has dimension @xmath304 . to be consistent with our previous equations , we assume that we measure @xmath306 positive lyapunov exponents on the projection @xmath40 .
if @xmath307 , then in the following equations @xmath305 should be replaced by @xmath308 , naturally assuming that @xmath309 . in analogy with the derivation of @xmath97 and @xmath98 in a bidimensional projection , we assume that if the spreading of initial conditions is uniform in the subspace @xmath5 .
then , @xmath310 represents the probability of finding trajectory points in @xmath304-dimensional space of one node ( or a group of nodes ) and @xmath311 represents the probabilities of finding trajectory points in the @xmath305-dimensional composed subspace constructed by two nodes ( or two groups of nodes ) in the subspace @xmath5 . additionally , we consider that the hypothetical number of occupied boxes @xmath254 will be given by @xmath312 .
then , we have that @xmath313 , which lead us to @xmath314 similarly to the way we have derived @xmath98 in a bidimensional projection , if @xmath40 has more than 2 positive lyapunov exponents , then @xmath315 to write eq .
( [ ic_hd ] ) in terms of the positive lyapunov exponents , we first extend the calculation of the quantity @xmath50 to higher - dimensional subspaces that have dimensionality 2q , @xmath316 where @xmath317 are the lyapunov exponents measured on the subspace @xmath5 . to derive this equation
we only consider that the hypothetical number of occupied boxes @xmath254 is given by @xmath318 .
we then substitute @xmath50 as a function of these exponents ( eq .
( [ dk ] ) ) in eq .
( [ ic_hd ] ) .
we arrive at @xmath319 consider a network whose attractor @xmath1 possesses @xmath320 positive lyapunov exponents , denoted by @xmath321 , @xmath322 . for a typical subspace @xmath5 , @xmath42 measured on @xmath5 is equal to the largest lyapunov exponent of the network .
just for the sake of simplicity , assume that the nodes in the network are sufficiently well connected so that in a typical measurement with a finite number of observations this property holds , i.e. , @xmath323 .
but , if measurements provide that @xmath324 , the next arguments apply as well , if one replaces @xmath325 appearing in the further calculations by the smallest lyapunov exponent , say , @xmath326 , of the network that is still larger than @xmath42 , and then , substitute @xmath327 by @xmath328 , and so on . as before , consider that @xmath306 .
then , for an arbitrary subspace @xmath5 , @xmath329 , since a projection can not make the lyapunov exponents larger , but only smaller or equal
. defining @xmath330 since @xmath331 , it is easy to see that @xmath332 so , @xmath97 , measured on the subspace @xmath40 and a function of the @xmath305 largest positive lyapunov exponents measured in @xmath40 , is an upper bound for @xmath333 , a quantity defined by the @xmath305 largest positive lyapunov exponents of the attractor @xmath1 of the network . therefore , if the lyapunov exponents of a network are know , the quantity @xmath333 can be used as a way to estimate how much is the mir between two measurements of this network , measurements that form the subspace @xmath5 . notice that @xmath97 depends on the projection chosen ( the subspace @xmath5 ) and on its dimension , whereas @xmath333 depends on the dimension of the subspace @xmath40 ( the number 2q of positive lyapunov exponents ) . the same happens for the mutual information between random variables that depend on the projection considered .
equation ( [ icu_new ] ) is important because it allows us to obtain an estimation for the value of @xmath97 analytically .
as an example , imagine the following network of coupled maps with a constant jacobian @xmath334 where @xmath335 $ ] and @xmath336 represents the connecting adjacent matrix . if node @xmath10 connects to node @xmath9 , then @xmath337 , and 0 otherwise .
assume that the nodes are connected all - to - all .
then , the @xmath303 positive lyapunov exponents of this network are : @xmath338 and @xmath339}$ ] , with @xmath340 .
assume also that the subspace @xmath5 has dimension @xmath305 and that @xmath305 positive lyapunov exponents are observed in this space and that @xmath341 .
substituting these lyapunov exponents in eq .
( [ icu_new ] ) , we arrive at @xmath342 we conclude that there are two ways for @xmath333 to increase
. either one considers larger measurable subspaces @xmath5 or one increases the coupling between the nodes .
this suggests that the larger the coupling strength is the more information is exchanged between groups of nodes . for arbitrary topologies
, one can also derive analytical formulas for @xmath333 in this network , since @xmath321 for @xmath343 can be calculated from @xmath327 @xcite .
one arrives at @xmath344 where @xmath345 is the @xmath9th largest eigenvalue ( in absolute value ) of the laplacian matrix @xmath346 .
concluding , we have shown a procedure to calculate mutual information rate ( mir ) between two nodes ( or groups of nodes ) in dynamical networks and data sets that are either mixing , or present fast decay of correlations , or have sensitivity to initial conditions , and have proposed significant upper ( @xmath97 ) and lower ( @xmath98 ) bounds for it , in terms of the lyapunov exponents , the expansion rates , and the capacity dimension . since our upper bound is calculated from lyapunov exponents or expansion rates , it can be used to estimate the mir between data sets that have different sampling rates or experimental resolution ( e.g. the rise of the ocean level and the average temperature of the earth ) , or between systems possessing a different number of events .
additionally , lyapunov exponents can be accurately calculated even when data sets are corrupted by noise of large amplitude ( observational additive noise ) @xcite or when the system generating the data suffers from parameter alterations ( `` experimental drift '' ) @xcite .
our bounds link information ( the mir ) and the dynamical behaviour of the system being observed with synchronisation , since the more synchronous two nodes are , the smaller @xmath47 and @xmath347 will be
. this link can be of great help in establishing whether two nodes in a dynamical network or in a complex system not only exchange information but also have linear or non - linear interdependences , since the approaches to measure the level of synchronisation between two systems are reasonably well known and are been widely used .
if variables are synchronous in a time - lag fashion @xcite , it was shown in ref .
@xcite that the mir is independent of the delay between the two processes .
the upper bound for the mir could be calculated by measuring the lyapunov exponents of the network ( see supplementary information ) , which are also invariant to time - delays between the variables . *
acknowledgments * m. s. baptista was partially supported by the northern research partnership ( nrp ) and alexander von humboldt foundation .
m. s. baptista would like to thank a. politi for discussions concerning lyapunov exponents .
rubinger , e.r .
v. junior and j.c .
sartorelli thanks the brazilian agencies capes , cnpq , fapemig , and fapesp . | the amount of information exchanged per unit of time between two nodes in a dynamical network or between two data sets is a powerful concept for analysing complex systems .
this quantity , known as the mutual information rate ( mir ) , is calculated from the mutual information , which is rigorously defined only for random systems . moreover ,
the definition of mutual information is based on probabilities of significant events .
this work offers a simple alternative way to calculate the mir in dynamical ( deterministic ) networks or between two data sets ( not fully deterministic ) , and to calculate its upper and lower bounds without having to calculate probabilities , but rather in terms of well known and well defined quantities in dynamical systems . as possible applications of our bounds
, we study the relationship between synchronisation and the exchange of information in a system of two coupled maps and in experimental networks of coupled oscillators . |
bose - einstein condensation of quasiparticles in solid - state systems has been observed in excitons in quantum hall bilayers , @xcite exciton - polaritons in semiconductor microcavities , @xcite gases of magnons , @xcite cavity photons @xcite and indirect excitons .
@xcite exciton - polaritons , mixed light - matter quasiparticles behaving as bosons , form condensates which exhibit not only the fundamental properties of quantum gases , but also new fascinating phenomena related to their out - of - equilibrium character .
@xcite the photonic component of polaritons is responsible for their light mass , which makes condensation possible up to room temperature @xcite , and for their easy creation , manipulation and detection by using simple optical - microscopy setups . on the other hand ,
their excitonic component yields strong coulomb repulsive interactions that make them promising candidates for future non - linear optical technologies .
the peculiar quantum fluid properties of polariton condensates are under intense research nowadays .
recent findings include : robust propagation of coherent polariton bullets @xcite and elucidation of the validity of the landau criterion for frictionless flow in the presence of weak structural defects , @xcite persistent quantized superfluid rotation , @xcite and solitary waves resulting from compensation between dispersion and particle interaction .
@xcite moreover , the intrinsic out - of - equilibrium character of polariton condensates has motivated recent theoretical studies on how to describe properly the energy flow from an optically - injected hot exciton reservoir to the coherent polariton modes , @xcite which we carefully address in this work .
the functionalities of microcavities in the strong coupling regime , as integrated optical elements , promote polaritons as an undreamt platform to create new logical devices .
@xcite thanks to their interactions with non - condensed excitons , polaritons can be easily accelerated , propagating over macroscopic distances in high finesse microcavities .
@xcite in this case , new interferometric devices can be built by properly shaping the excitation profile @xcite as well as the microcavity etching .
@xcite extra confinement can be achieved by lateral bounding the optical modes through patterning the microcavity , @xcite by sculpting the pumping profile creating blueshift - induced traps , @xcite or by a combination of both methods .
@xcite this paves the way for studies of atom - like scenarios in which the energy spectrum becomes discrete . in a recent work using quasi 1d - microwire ridges , a polariton condensate transistor switch has been realized through optical excitation with two beams .
@xcite one of the beams creates a polariton condensate which serves as a source ( _ s _ ) of polaritons ; their propagation is gated using a second weaker gate beam ( _ g _ ) that controls the polariton flow by creating a local blueshifted barrier ( a list of symbols used in the manuscript are given in the appendix a ) .
the on state of the transistor ( absence of _ g _ ) corresponds to forming a trapped condensate at the edge of the ridge ( collector , _
c _ ) labelled as @xmath0 .
the presence of _ g _ hinders the propagation of polaritons towards _
c _ , remaining blocked between _
s _ and _ g _ ( off state ) .
an insight of the energy relaxation and dynamics of the condensed polariton propagation in this system has been obtained lately by a time - resolved study of the on / off states .
@xcite in the present work , we make a systematic study of the influence of the density of polaritons created in _ s _ and _ _ g__on the propagation and the gating of polariton bullets , of their energy and density relaxation and of the optimal conditions for realizing an all - optical polariton condensate transistor switch .
our experiments are compared with simulations of the polariton condensate dynamics based on a generalized gross - pitaevskii equation , modified to account for incoherent pumping , decay and energy relaxation within the condensate .
we investigate a high - quality @xmath1 algaas - based microcavity with 12 embedded quantum wells , with a rabi splitting of @xmath2 mev .
ridges have been sculpted through reactive ion etching with dimensions @xmath3m@xmath4 ( further information about this sample is given in refs . and ) .
figure [ fig : fig0 ] ( a ) shows a scanning electron microscopy image of such a ridge , including the excitation scheme ; a temporal scheme of the excitation and emission processes is given in panel ( b ) . in our sample lateral confinement is insignificant as compared to much thinner , 1d polariton wires .
@xcite the chosen ridge is in a region of the sample corresponding to resonance ( detuning between the bare exciton and bare cavity mode is @xmath5 0 ) .
the sample , mounted in a cold - finger cryostat and kept at 10 k , is excited with 2 ps - long light pulses from a ti : al@xmath6o@xmath7 laser , tuned to the first high - energy bragg mode of the microcavity ( 1.612 ev ) .
we split the laser beam into two independent beams , whose intensities , spatial positions and relative time delay ( zero for these experiments ) can be independently adjusted .
we focus both beams on the sample through a microscope objective to form 5 @xmath8m-@xmath9 spots spatially separated by @xmath1040 @xmath8 m along the ridge .
the same objective is used to collect ( angular range @xmath11 ) and direct the emission towards a spectrometer coupled to a streak camera obtaining energy- , time- and spatial - resolved images , with resolutions of 0.4 mev , 15 ps and 1 @xmath8 m , respectively . in our experiments
polaritons propagate along the @xmath12 axis of the ridge .
there is also some diffusion of polaritons in the @xmath13 direction , but it is not relevant for the operation of our device .
all the images in the manuscript show the emission collected along the @xmath12 axis from a 10-@xmath8 m wide , central region of the ridge .
the power threshold for condensation of polaritons is @xmath14 mw .
figure [ fig : fig1 ] shows , as an example , under cw conditions , the intensity distribution of the polariton emission as a function of energy and of the position in the ridge : when we only use the _ s _ beam , fig . [ fig : fig1 ] ( a ) , we place it @xmath1075 @xmath8 m away from the right ridge border ; in the _ _ s__+_g _ beam excitation , _ g _ is placed @xmath1035 @xmath8 m away from the border , fig .
[ fig : fig1 ] ( b ) . in fig .
[ fig : fig1 ] ( a ) , exciting with a power @xmath15 , the blue - shifted ( @xmath16 mev ) emission at the source , together with a weak condensate emission @xmath0 , are clearly observed .
@xmath0 emits from an energy lower than that of the propagating polaritons as a result of an unintentional modification of the microcavity structure created by the etching process at the edge of the ridge .
polaritons propagate at a constant energy towards the left and right sides of the ridge .
exciting with both laser beams , under different excitation conditions for a typical off state , @xmath17 and @xmath18 , the gating state of the switch is readily seen with the stopped , condensed polaritons just before the _
g _ position .
in this work , we time - resolve the different excitation configurations presented in ref . :
fig . 2 ( where the @xmath19 is varied whilst @xmath20 ) and fig .
3 ( @xmath21 and @xmath22 is varied ) . our study obtains intensity- and energy - dynamics of exciton and polariton emission in the ridge . in the former case varying @xmath19
, we fully characterize the on state response of a polariton transistor switch ; in the latter one , we modulate the polariton condensate trapping potential .
@xcite in this section we present three time - resolved cases for different @xmath19 pump powers , figs .
[ fig : fig2]-[fig : fig4 ] .
figure [ fig : fig2 ] shows the dynamics of the emission when @xmath15 . for each panel
the time is displayed at the right upper corner , being the temporal origin set at the instant when the _ s _ intensity is maximum .
panel ( a ) shows that the emission from _ s _ at -24 ps occurs at 1.545 ev ; at 31 ps , ( b ) , the emission redshifts and a small spatial expansion around 0 @xmath8 m is observed ; propagating polaritons , expanding more rapidly towards the border , at an energy of1.542 ev , are detected at 194 ps ( c ) , eventually reflecting backwards , interfering coherently and creating the @xmath0 condensate , weakly emitting at 1.539 ev at later times , 483 ps ( d ) .
since the pump power is at threshold , the emission intensity of polaritons is weak and slightly higher than the noise level in all panels of fig .
[ fig : fig2 ] .
let us note that at early times , the emission observed in fig .
[ fig : fig2 ] appears blueshifted from the lower polariton minimum by an amount comparable to one - half of the rabi splitting .
this suggests that the emission at the source comes from polaritons with a strong excitonic character .
for this reason we will refer to the emission from the source as arising from excitons , although the decrease in the blueshift over time corresponds to a continuous transition from excitonic polariton states to those with roughly equal excitonic and photonic fractions .
the emission from the propagating states and collector region , at lower energy , is clearly a polaritonic emission .
it is also important to note that the duration over which the condensate is present greatly exceeds the polariton lifetime .
this is because the condensate is continuously fed by high energy excitons ( not visible in the spectrum and distinct from the excitonic states emitting at the source ) excited by the non - resonant pulse .
the emission at _ s _ is determined by repulsive interactions with hot excitons which contribute a blueshift to the potential energy . as hot excitons decay from the system , either through recombination or condensation ,
this potential energy decreases over time .
once the polariton condensate has formed , due to the low density of polaritons , there are minimal energy relaxation processes , such that the propagating polaritons tend to conserve their energy as they spread out from the source ( see fig .
[ fig : fig2 ] ( d ) ) .
the dynamics of the emission increasing @xmath19 to @xmath23 is shown in fig .
[ fig : fig3 ] : the initial excitonic emission at the source takes place at 1.546 ev ( a ) , slightly higher than before , due to increased blueshift due to a larger hot - exciton repulsion . at @xmath24 ps ( b ) , an essential difference with respect to the case of fig .
[ fig : fig2 ] ( b ) is revealed : polaritons emit from a lower energy than that of the source , which is @xmath102 mev blueshifted ; this situation holds during the first @xmath10200 ps of the decay process .
panel ( c ) shows the arrival of polaritons at the ridge border at 140 ps , and the eventual condensation of @xmath0 ( d ) .
this final relaxation phase in the dynamics takes place into a state defined in a minimum of the wire structural potential located at the wire edge . a clear indication of the polariton coherence is evidenced by the interference at 1.540 ev between counter - propagating wave - packets .
the source population at _ s _ , still 1 mev blueshifted with respect to propagating polaritons , expands around @xmath25 as it decays in energy , and continuously feeds the propagating polariton condensate , increasing its effective lifetime ( e ) .
finally , as shown in ( f ) at 617 ps , polaritons at _ s _ merge with those propagating along the ridge .
the emission is still observed for times as large as @xmath26 ns ( not shown ) .
the case of the highest source power used in our experiments is shown in fig .
[ fig : fig4 ] : at @xmath27 ps the excitonic source population emits at 1.547 ev ( a ) .
the progressive spatial expansion of the excitonic population and the fast relaxation of the polariton condensate , as it propagates towards the right side , at 1.541 ev , reaching the ridge edge at 35 ps , is shown in panels ( b ) and ( c).@xmath0 is now slightly blueshifted , with respect to its energy at lower @xmath19 conditions , to 1.540 ev , due to the higher density condensate population at this place of the ridge ( d ) . at later times , as those shown in ( e , f ) for 300 and 508 ps , the population at _ s _ decreases and expands in space whilst @xmath0 redshifts its energy emission due to its reduced occupancy .
movies corresponding to figs .
[ fig : fig2]-[fig : fig4 ] are provided as supplementary material .
@xcite the introduction of a new secondary pulse , dubbed before as gate ( _ g _ ) , between _ s _ and _ c _ , adds new interaction phenomena .
the existence of two condensates becomes very clear in this case : one of them located initially between _
s_-_g _ , @xmath28 , which eventually becomes propagating , and a second one , already labelled as @xmath0 .
the polariton propagation towards _
c _ along the ridge can be hindered with a below - threshold intensity gate beam , see fig .
[ fig : fig1 ] ( b ) and fig .
[ fig : fig5 ] , rendering the @xmath0 switch - off and creating the trapped condensate @xmath28 . as the _
g_-repulsive potential gradually decreases in time , a tiny fraction of @xmath28 is able to tunnel through the barrier and it spreads between 40 and 80 @xmath8 m , see fig .
[ fig : fig5 ] ( d ) .
this configuration has been already discussed in detail in ref . .
figure [ fig : fig6 ] shows the dynamics for @xmath17 and @xmath29 .
the visibility of the emission at _ g _ is delayed by @xmath1020 ps with respect to that at _ s _ , despite of the fact that both beams reach the sample simultaneously , panels ( a , b ) .
this delay is due to the power dependence of the emission rise time , which increases with decreasing power .
the @xmath28 condensate lies at a constant energy , 1.541 ev , remaining trapped , ( c ) . on their own account , the emission energy of _ s _ and _ g _ decay , at a rate determined by the carrier density and the carrier - carrier interactions , until they reach 1.542 ev for both .
the population between _
g _ and _ c _ , mainly created by the _
g _ pulse , propagates towards the border ( c ) , being reflected ( d ) and forming the @xmath0 condensate ( e ) .
when the _ g _ barrier further decays the @xmath28 condensate becomes propagating and coherent interference patterns are generated from counter - propagation , see panel ( f ) . for completeness , fig .
[ fig : fig7 ] depicts the case corresponding to large values of @xmath22 .
panel ( a ) depicts the excitonic emission at 1.547 ev when the laser beams arrive at _ s _ and _
g_. @xmath28 is trapped around @xmath12=20 @xmath8 m , blueshifted up to 1.542 ev , due to repulsive interactions ( b ) , whilst polaritons between _ g _ and _ c _ propagate towards the border . at 76 ps ,
a new condensate , @xmath0 , becomes trapped at 1.540 ev , and the emission energy of _ s _ and _ g _ reaches that of @xmath28 ( c ) . due to the barrier reduction at _ g _ , @xmath28 propagates along the ridge from 0 to 60 @xmath8 m ( d ) .
@xmath0 remains confined for later times at a constant energy , whereas @xmath28 decays and interferes with itself , panels ( e , f ) .
movies corresponding to figs .
[ fig : fig5]-[fig : fig7 ] are provided as supplementary material .
@xcite in our sample , the emission above 1.544 ev is coming from excitonic states .
the polariton emission lies at lower energies , down to 1.538 ev at the collector region . in this section ,
we analyze the dynamics of the energy and population relaxation along the full region of propagation of the condensates between _ s _ and _ c _ both in the presence or absence of _ g _ , obtaining quantitative values for the energy time - decays and the optimal working conditions for the on - state . figures [ fig : fig8 ] and [ fig : fig9 ] show spatial - temporal maps of the energy ( a - c)/intensity ( d - f ) evolution of the emission for the same power values as those used in figs .
[ fig : fig2]-[fig : fig4 ] and [ fig : fig5]-[fig : fig7 ] , respectively .
panels ( a - c ) have been obtained identifying the time at which the maximum emission intensity takes place , at a given _ x _ position on the ridge , for every energy : this gives a point in the map whose energy is coded with the false - color scale shown on the right - hand side of the upper row .
note that all the information concerning the strength of the emission , and therefore the polariton population , is lost in this representation .
@xcite the complementary information is encoded in the second row in figs .
[ fig : fig8 ] and [ fig : fig9 ] , giving in this case the polariton population from integrating all emission energies .
these plots provide a straight and precise insight on the energy / intensity decay of the population at every position along the ridge .
let us start by considering the one - beam excitation compiled in fig .
[ fig : fig8 ] . in panels ( a , b ) the energy trap created at @xmath30 m due to the potential discontinuity close to the border of the ridge
is clearly observed : @xmath0 , emitting at 1.539 ev , is separated by a small energy gap from the polaritons propagating above at 1.540 ev .
the horizon , @xmath31 , on the right upper corners , between white and colored points is given by the arrival of polaritons at different positions along the ridge .
another discontinuity is observed between the decay of carriers at _ s _ and the propagating polaritons , @xmath32 .
the power dependence of both discontinuities is evident in these panels and gives information about the speed of propagation of different emitting species . at @xmath33 , the border , @xmath32 , between carriers at _
s _ and polaritons , whose propagation is seen for @xmath34 m , is absent ( a ) , because the energy of excitons and polaritons decay at the same rate , but it becomes very clear in panels ( b , c )
. the speed of propagation of the carriers can be obtained from the slope of @xmath32 and @xmath31 lines . for the carriers at _ s _ in panel ( b ) , @xmath32 is almost straight , therefore a mean speed value , @xmath35 , can be obtained amounting to @xmath36m / ps . at the highest power ( c ) , @xmath37 initially has increased by a factor of @xmath103 as compared to @xmath35 , but the strong non - linearities associated with the high carrier densities lead to the appearance of deceleration rendering a gradual decrease of @xmath38 .
the spatial extension of the carriers around _
s _ also widens with increasing power , almost doubling its value from @xmath1030 to @xmath1060@xmath39 m at 400 ps as seen in ( b ) and ( c ) panels .
it is also noticeable that the energy decay of the carriers is spatially flat in the region enclosed by @xmath32 .
the acceleration / deceleration of the propagating polaritons is distinct in the slope changes of @xmath31 , panels ( a - c ) . for @xmath15 , a rough estimation of the speed obtains @xmath40@xmath41m / ps ; @xmath42 increases to @xmath43 and @xmath44m / ps for @xmath45 and @xmath46 , respectively . in the later case @xmath42
amounts to @xmath47 of the speed of light in vacuum .
the formation of @xmath0 at threshold , ( a ) , is seen by the purple ( 1.539 ev ) oval shape at ( @xmath30 m , 400 - 600 ps ) .
the enhancement of @xmath42 together with that of stimulated scattering processes with power give rise to an earlier appearance of @xmath0 at @xmath10280 ps lasting for 400 ps , almost doubling its spatial extent , at @xmath48 ( b ) .
the values for @xmath42 are in agreement with others reported in the literature ( see , for example , ref . ) .
the much smaller values for @xmath38 are due to the larger exciton mass compared to that of polaritons .
the energy gap between @xmath0 and the propagating polaritons dissolves at @xmath49 due to the very large number of polaritons and the very fast formation of this condensate .
finally , let us remark that the ballistic propagation of polaritons is evidenced in panel ( b ) by the constant energy ( same color ) , for a given time , seen in the region enclosed by the @xmath32-border and _
c_. however , in case ( c ) a gradual change in energy ( color ) is observed , indicating the energy loss during the polariton propagation towards _
we briefly discuss now the density maps for different power excitation shown in the lower row of fig .
[ fig : fig8 ] , in a normalized , logarithmic false - color scale shown on their right - hand side .
panels ( d - f ) show that the main emission intensity arises from the population at _ s _ , with a gradual expansion towards _
c _ with a much lower polariton population .
the emission - intensity decay becomes faster with increasing @xmath19 power . for @xmath48 , ( b ) , an enhanced emission following
the @xmath32 is apparent ; interferences of polaritons in the region between @xmath1020 and @xmath1060 @xmath8 m are visible ; the formation of @xmath0 appears at 280 ps .
panel ( f ) shows several reflections of condensed polaritons between the ridge edge and the left - bouncing positions marked with white bars , which are determined by the potential delimited by @xmath32 and the energy of the bouncing condensates : the longer the time , the larger the energy loss of the polaritons , which become less able to climb the barrier side , as borne out by the progressively increasing distance between the bars and the @xmath32 line , obtained from panel ( c ) and depicted with a white dotted line . at @xmath1060 @xmath8 m and 100 ps a considerable amount of population forms the @xmath0 condensate .
we turn now to the two - beam excitation compiled in fig .
[ fig : fig9 ] .
panel ( a ) displays the energy decay of the polaritons in the off - state for @xmath17 and @xmath18 .
the @xmath28 condensate , extending 20 @xmath8 m , reveals an almost constant energy emission in time .
the contrast of the off - state is high as assessed by the negligible amount of polaritons that goes through the _
g _ potential ( d ) ; only a hint of the polaritons that were able to tunnel through is seen at ( 80 @xmath8 m , 400 ps ) in panel ( a ) , that codifies the energy but not the intensity of the signal .
the ratio @xmath50is much larger than @xmath51 , obtained in the one beam case since @xmath28 is trapped closer to _ s _ and its feeding process is more efficient . increasing @xmath22 to @xmath52 both _
s _ and _ g _ beams contribute to the formation and trapping of polariton condensates ( b ) , @xmath28 and @xmath0 .
figure [ fig : fig9 ] ( c ) , for @xmath53 , shows that , for the first @xmath54 ps , the energy decays at _ s _ and _ g _ are much faster than those of the polariton condensates . for longer times , @xmath55 ps , the energy decay of the populations at _ s _ , @xmath28 and _ g _ is almost identical ; however , @xmath0 is always at a lower energy due to the trapping at _ c_. a further inspection of the energy - integrated intensity maps shows that in panel ( e ) , at 300 ps , when the _ g_-barrier has considerably decayed , so that its energy coincides with that of @xmath28 , the @xmath28 condensate starts expanding along the ridge ; concomitantly a slanted interference pattern is obtained , revealing the dynamics of merging counter - propagating polaritons .
the @xmath28 formation time , @xmath1070 ps , is much shorter than of @xmath0 , @xmath10350 ps , due to the fact that @xmath19 is much larger than @xmath22 and that both beams contribute to feed @xmath28 while only the population at
_ g _ refills the @xmath0 condensate , which reaches its maximum intensity emission at 400 ps . in panel
( f ) , the high _ s_- and _ g_-pump powers make the @xmath28 condensate very intense at 40 ps . the confluence of the _ s_- and _ g_-population with @xmath28 takes place at 100 ps and 1.542 ev .
then @xmath28 doubles its spatial width , as observed by the spreading cone of polaritons extending 20 @xmath8 m at 40 ps to 40 @xmath8 m at 250 ps . a clear back and forth bouncing of the @xmath0 condensate between the _ g_-barrier and the ridge edge
is observed during the first 100 ps . after losing its kinetic energy at @xmath56 ps , @xmath0stops and emits for more than 600 ps , as its population is continuously fed by propagating polaritons at @xmath57 ev .
the energy maps shown in figs .
[ fig : fig8]-[fig : fig9 ] ( a - c ) allow to quantitatively analyze the energy decay at every @xmath12-position ; in particular we present in fig .
[ fig : fig10 ] this decay at the _ s _ position for @xmath58 . the solid white line in fig . [ fig : fig10 ] ( a ) corresponds to its best fit to the sum of two exponentially decaying functions , shown separately by the dashed and dot - dashed lines .
the double fashion decay is attributed to two different physical processes : a fast decay due to relaxation driven by exciton - exciton interactions and a slow one , attributed to the decreasing blueshift caused by the diminishing exciton and polariton populations .
the rate of condensation can be expected to be faster at early times due to larger densities of carriers resulting in stronger stimulated scattering processes .
a fast condensation rate results in an initial fast drop in the exciton density since excitons condense rapidly into polaritons that quickly decay .
this drop in the exciton population gives a corresponding drop in the polariton population and so both blueshifts , due to polariton - exciton and polariton - polariton repulsion , drop sharply at early times . at longer times
polariton condensation proceeds slower , due to weaker stimulated scattering and the exciton populations decay with a slow exponential dependence due to exciton recombination .
figure [ fig : fig10 ] ( b ) compiles the power dependence of the decay times : both decrease with increasing power , more markedly for @xmath59 ( circles ) , which decreases by @xmath60 for a 20 fold increase of power , whilst @xmath61 ( squares ) only diminishes by @xmath62 , revealing the larger influence of density in exciton - exciton scattering processes than in exciton - polariton ones .
the spatial integration of the data shown in figs . [ fig : fig2]-[fig : fig7 ] reveals the total energy and intensity decay dynamics for the different configurations under study as shown in fig .
[ fig : fig11 ] : panels ( a - c)/(i - iii ) , correspond to one / two beam excitation under different @xmath19/@xmath22 powers .
the addition of contributions from different population species gives rise to a very rich dynamics .
figures [ fig : fig11 ] ( a - c ) exhibit a critical difference in the power dependence of the total decay : panel ( a ) shows a collective energy decay for * all * spatial positions along the ridge . panels ( b , c ) show a low energy streak corresponding to a polariton condensate drop that propagates along the ridge with an almost constant energy , unveiling the ballistic propagation of the condensate .
the two streaks presented in panel ( i ) correspond to the decay of population at _ s _ ( high energy one ) and the emission of @xmath28 for a typical switch off state ( low energy one ) : the dynamics of both streaks is similar to those shown in panel ( c ) , with the difference , not appreciated in the figure , that polaritons now are stopped just before the _ g _ barrier .
the three traces appearing in panel ( ii ) , ordered by decreasing energy , compile the emission from : the population at the source and the gate ( _ s+g _ ) , the @xmath28 condensate and the @xmath0 condensate , respectively .
it is worthwhile noting the identical decay dynamics of the _ s _ and _ g _ populations , observed by the existence of only one streak for both populations .
the @xmath0 condensate shows an emission at @xmath101.539 ev , with a dynamics similar to that shown in panel ( b ) . as the _ s _ power
is kept constant in this subset of experiments , the _ g _ power permits manipulating on demand the amount of condensed polaritons at @xmath28 : if it would have been formed only by the _ s _ pulse , its energy should decay slightly ; however , the extra population injected by the _ g _
pulse contributes with an additional blueshift giving rise to an increase of the @xmath28 emission energy , hinted at @xmath10 300 ps in panel ( ii ) , which becomes clearly visible in panel ( iii ) . in this latter panel ,
s+g _ decays are also superimposed and @xmath0 emits at a constant energy of 1.540 ev for @xmath63 ps .
it is important to note that in panels ( ii , iii ) , the additional polaritons provided by the _
g _ pulse make the @xmath28 condensate the highest populated state in the device with an emission intensity even larger than that of @xmath64 together .
let us consider the optimal power conditions for the on state for a _ s - c _ spatial separation of @xmath65 @xmath8 m .
we present in fig .
[ fig : fig12 ] the main effects of the @xmath19 power on the transistor switch on state , which was illustrated before in figs .
[ fig : fig2]-[fig : fig4 ] . figure [ fig : fig12 ] ( a ) plots the normalized intensity dynamics of the source , at @xmath25 , ( shadowed traces ) and that of @xmath0 , at @xmath66 @xmath8 m , ( full lines ) , for different @xmath67 values .
we define the switch on time , @xmath68 , as the temporal delay between the _ s _ maximum intensity and that of the @xmath0 . it is clearly observed that @xmath68 decreases , and the shape of the @xmath0 time - evolutions becomes more asymmetric with increasing @xmath19 . the asymmetry of the @xmath0 temporal evolution ( see fig . [
fig : fig12 ] ( a ) ) , which strongly depend on @xmath19 , is characterized in fig .
[ fig : fig12 ] ( b ) , where we define a raise time , @xmath69 ( up triangles ) , given by the time spent to raise from an intensity of 0.5 up to the maximum value of 1 .
similarly , @xmath70 ( down triangles ) is given by the time interval in which the intensity falls from a value of 1 to 0.5 .
a non monotonic dependence of @xmath70 on power is observed with a sharp raise at low @xmath19 values and a gradual fall for high ones : if the aim is to create a long lived on state , the optimal power corresponds to @xmath71 , where @xmath72 ps . on its own hand , the raise time , @xmath69 , decreases monotonically with increasing @xmath19 , reaching a minimum value of @xmath73 ps : a marked dependence at small powers , followed by an almost negligible decay at high ones , results in an optimum power to create a fast response transistor at similar powers than those required for a long lived on state .
figure [ fig : fig12 ] ( c ) shows the power dependence of @xmath68 ( full circles ) together with the initial energy shift of the emission at _ s _
( @xmath74 , open diamonds ) . the monotonous decrease of @xmath68 with power ( increase of switching rate ) is linked to the increase of @xmath74 , due to the enhanced polariton acceleration from _ s _ to _ c _ produced by the augmented photo - generated repulsive excitonic potential , but other contributions as , for example , increase of stimulated scattering processes in the creation of @xmath0 are also responsible for the quickening of @xmath68 .
the minimum value of @xmath68 , @xmath75 ps , corresponds to a polariton propagation speed of @xmath76 @xmath8m / ps in agreement with the values of @xmath42 obtained from the horizon established by @xmath31 in the energy maps of fig [ fig : fig8 ] ( a - c ) .
finally , we should mention that our results indicate that the optimal conditions for gating are obtained for @xmath77 , in agreement with the previous results of ref . .
at this power , the maximum attenuation of the @xmath0 condensate is obtained yielding the highest contrast for the off state ; at lower values of @xmath22 the traveling polaritons are not gated efficiently and at higher values the gate starts feeding @xmath0 .
to model our experimental results theoretically , we make use of a phenomenological treatment of polariton energy - relaxation processes taking place in the system . such processes are not only responsible for the relaxation of hot excitons ( injected by the pump ) into polaritons in the form of a condensate , but also for the further relaxation in energy of the polariton condensate as it propagates .
this latter energy relaxation process can be strongly influenced by a spatially dependent potential coming from repulsion from the hot excitons .
let us first introduce the description of the polariton condensate , which we will later couple to a description of higher energy excitons .
a fundamental feature of bose - einstein condensates is their spatial coherence that allows them to be well described with a mean - field approach .
@xcite the gross - pitaevskii equation has been developed to describe the non - equilibrium dynamics of condensed polaritons , where losses due to the short polariton lifetime @xcite and gain due to non - resonant pumping @xcite were included phenomenologically . in such form , a variety of recent experiments can be modeled , including , for example , experiments on polariton transport @xcite , spatial pattern formation @xcite and spin textures .
@xcite the gross - pitaevskii equation for the polariton wave - function , @xmath78 , is : @xmath79\psi(x , t)\notag\\ & \hspace{5mm}+i\hbar\mathfrak{r}\left[\psi(x , t)\right]\label{eq : gp}\end{aligned}\ ] ] here @xmath80 represents the kinetic energy dispersion of polaritons , which at low wavevectors can be approximated as @xmath81 , with @xmath82 the polariton effective mass .
@xmath83 represents the strength of polariton - polariton interactions .
being repulsive ( @xmath84 ) , these interactions allow both a spatially dependent blueshift of the polariton condensate energy and energy - conserving scattering processes .
our analysis shows , however , that neither of these effects play a dominant role in our experiments .
the effective potential acting on polaritons caused by repulsive interactions between polaritons and higher energy excitons is more significant , and is responsible for the blocking of polariton propagation in the presence of a gate pump .
the effective potential @xmath85 can be divided into a contribution from three different types of hot exciton states , which will be described shortly , as well as a static contribution due to the wire structural potential , @xmath86 : @xmath87\notag\\ & \hspace{5mm}+v_0(x)\end{aligned}\ ] ] @xmath88 , @xmath89 and @xmath90 correspond to density distributions of active " , inactive " and dark excitons , respectively , as described below .
experimental characterization has revealed that the static potential , @xmath86 , is non - uniform along the wire and exhibits a slight dip in the potential near the wire edge .
@xmath91 , @xmath92 and @xmath93 define the strengths of interaction with the various hot exciton states .
@xmath88 represents the density distribution of an `` active '' hot exciton reservoir .
@xcite these excitons have the correct energy and momentum for direct stimulated scattering into the condensate and so appear as an incoherent pumping term in eq .
( [ eq : gp ] ) with @xmath94 the condensation rate . to describe the dynamics of the system , it is important to note that not all excitons in the system are in this active form .
in fact , the non - resonant pumping creates excitons with very high energy and they must first relax in energy before becoming active .
we can thus identify an `` inactive '' reservoir of hot excitons that is excited by the non - resonant pump but not directly coupled to the condensate .
the dynamics of the exciton densities are described by rate equations : @xmath95 @xmath96 @xmath97 when solving the equations we start from the initial condition @xmath98 , @xmath99 and introduce a density proportional to the pump intensity profile in the inactive reservoir , @xmath100 .
this represents an instantaneous injection by the non - resonant ultra - short pulse used in the experiment .
the inactive reservoir is coupled by both linear and non - linear terms to the active reservoir , described by @xmath101 and @xmath102 , respectively
. we also account for a linear coupling to a dark exciton reservoir , @xmath103 , described by coupling rate @xmath104 .
dark excitons are long - lived states that are optically inactive yet can nevertheless be populated as high energy excitations from the non - resonant pump relax in energy .
the dark excitons introduce a long - lived repulsive contribution to the effective polariton potential , @xmath85 , and are thus efficient at gating propagating polaritons at long - times .
@xmath105 , @xmath106 and @xmath107 describe the decay rates of each of the reservoirs .
the feeding of dark excitons from the inactive reservoir represents processes where higher energy electron - hole pairs relax in energy forming dark exciton states .
we have neglected any further conversion between bright and dark excitons .
nonlinear conversion has been shown to generate oscillations between bright and dark excitons .
@xcite however , these processes require coherent excitation of exciton - polaritons near the dark exciton resonance . in our case
, we do not expect accumulation of exciton - polaritons at such an energy .
furthermore , the fact that no oscillations in the polariton condensate density were observed , suggests that any coupling between bright and dark states is slower than the condensation rate .
it is worthwhile mentioning that we have also considered hot - exciton diffusion along the wire when solving eqs .
( [ eq : na]-[eq : nd ] ) , using typical exciton diffusion rates ; however no noticeable effect on polariton dynamics was observed .
returning to eq .
( [ eq : gp ] ) , the decay of polaritons is accounted for by the decay rate @xmath108 .
the final term in eq .
( [ eq : gp ] ) accounts for energy relaxation processes of condensed polaritons .
polaritons are expected to condense at the source into the lowest energy state , where they have zero kinetic energy and potential energy given by @xmath85 ( and an additional blueshift due to polariton - polariton interactions ) .
while this is the lowest energy state available at the source , one notes that the potential energy can be reduced if polaritons propagate away from the source ( @xmath85 decreases away from the source , where the reservoir densities are weaker ) .
if polaritons were to conserve their energy , then they would convert this potential energy into kinetic energy as they move away from the source , accelerating down the potential gradient .
however , the polariton kinetic energy can be lost as polaritons scatter with acoustic phonons @xcite or hot excitons .
@xcite surface scattering could be also responsible for this loss ; however since we have considered a phenomenological energy relaxation , the actual mechanism that causes that relaxation does not play a direct role in our calculations .
previous methods to introduce energy relaxation into a description of polariton condensates have been based on the introduction of an additional decay of particles depending on their energy @xcite ( occasionally known as the landau - khalatnikov approach ) .
the polariton number can be conserved in such a process via the introduction of an effective chemical potential .
@xcite the energy relaxation term is : @xmath109=-\left(\nu+\nu^\prime|\psi(x , t)|^2\right)\left(\hat{e}_\mathrm{lp}-\mu(x , t)\right)\psi(x , t),\label{eq : relax}\ ] ] where @xmath110 and @xmath111 are phenomenological parameters determining the strength of energy relaxation .
@xcite we do not attempt here a microscopic derivation of the energy relaxation terms , we only note that we can expect some energy relaxation at low polariton densities ( described by the parameter @xmath110 ) as well as a stimulated component of the relaxation proportional to the polariton density , @xmath112 ( described by the parameter @xmath111 ) .
note that the energy relaxation rate is assumed proportional to the kinetic energy of polaritons ; polaritons will relax in energy until they decay from the system or until their kinetic energy is zero ( such that they have zero in - plane wave - vector )
. the local effective chemical potential , @xmath113 , can be obtained from the condition : @xmath114 where , @xmath115 for @xmath116 real , and : @xmath117\notag\\ & = \left.\frac{\partial\sqrt{n(x , t)}}{\partial t}\right|_\mathfrak{r}e^{i\theta(x , t)}\notag\\ & \hspace{10mm}+\left.i\sqrt{n(x , t)}e^{i\theta(x , t)}\frac{\partial\theta(x , t)}{\partial t}\right|_\mathfrak{r}.\end{aligned}\ ] ] @xmath118 denotes the components of the derivatives due to the term @xmath119 $ ] in eq .
( [ eq : gp ] ) .
note that other terms in eq .
( [ eq : gp ] ) , namely the pumping and loss terms , do not conserve the number of condensed polaritons . although it would be desirable to define a mean free path between scattering events , this is not trivial because the energy - relaxation rate is both energy and density dependent .
equations ( [ eq : gp]-[eq : nd ] ) were solved numerically for different initial density profiles @xmath100 , corresponding to the different source and gate configurations studied experimentally .
we used the following parameters in the theory : @xmath120 ( obtained from fits to the dispersions measured in ref . ; @xmath121 is the free electron mass ) , @xmath122mev@xmath123 ( ref . ) , @xmath124 ps@xmath125 , @xmath126 ps@xmath125 , @xmath127 ps@xmath125 , @xmath128 ps@xmath125 , @xmath129 ps@xmath125 , @xmath130 ps@xmath131m@xmath4 , @xmath132 ps@xmath131m@xmath4 , @xmath133 ps@xmath131m@xmath4 , @xmath134 , @xmath135m@xmath4 .
figure [ fig : time_1 ] shows the evolution of the spectrum in real space for @xmath15 . as in the experimental case ( fig .
[ fig : fig2 ] ) , condensation initially takes place into a state blueshifted due to the repulsive interactions from the hot exciton reservoirs contributing to the effective potential @xmath85 . over time
, this blueshift decays resulting in progressively lower energy of the source state .
in addition , a propagating polariton state can be observed , which forms interference fringes due to reflection from the end of the ridge .
energy relaxation is slow , appearing only at very long times due to the lack of stimulation by the low polariton density . for higher power ( @xmath45 ) , fig .
[ fig : time_2 ] shows the onset of stimulated energy relaxation processes .
as in the experimental case ( fig .
[ fig : fig3 ] ) the relaxation takes place in two subsequent stages : first there is relaxation from the source state into the extended state with energy set by the ridge potential , followed by relaxation into the @xmath0 condensate at low energy . at @xmath46 , fig .
[ fig : time_3 ] shows that the energy relaxation occurs rapidly . as in the experimental case ( fig .
[ fig : fig4 ] ) the collector state is rapidly populated .
it is interesting to note that , as shown for the experiments in fig .
[ fig : fig12 ] ( c ) , the blueshift of the condensate at the source position does not increase linearly with the pump power .
this is because even though the injected hot exciton population can be expected to increase linearly , the increased carrier density results in a faster condensation rate due to the stimulation of scattering processes ( hot exciton relaxation processes as well as processes that cause excitons to relax into condensed polaritons ) .
polaritons decay much faster than uncondensed hot excitons , such that a high intensity pumping of hot excitons is quickly depleted giving rise to a limited blueshift of polaritons at the source .
in the presence of the gating pulse , the propagation of the @xmath28 condensate is blocked , as shown in fig .
[ fig : time_4 ] .
this is due to the injected hot exciton density at the gate position that adds to the polariton effective potential profile , @xmath85 . at long times
, the theory predicts a small transmission across the gate pulse , due to the decay of the potential barrier .
spatial - temporal maps of the peak emission energy with one - beam excitation are shown in fig .
[ fig : peak ] . in panel
( a ) there is a fast propagation of a high energy mode from the source followed by a decrease in energy of the emission over the whole space . at longer times
, one identifies relaxation into @xmath0 , near the wire edge .
this behaviour is in qualitative agreement with the experimental result , however , it can be noted that the speed of propagation appears overestimated in the theory .
this is because the theory neglects changes in the shape of the polariton dispersion caused by the hot - exciton induced blueshift , which can be particularly important at early times when the particles are strongly excitonic with a larger effective mass and slower group velocity .
figures [ fig : peak ] ( b ) and ( c ) show the peak emission energy maps for increasing source intensity , where the relaxation into an extended state with lower energy than the source can be identified . the relaxation is stronger at the highest pump power , due to increased stimulated energy relaxation processes .
this is also evidenced by the shorter time required for @xmath0 to appear with increasing pump power .
we discuss now the simulated energy- and intensity maps under two - beam excitation conditions compiled in fig .
[ fig : peak2 ] . panel ( a ) shows the case when a gate pulse , @xmath18 , is present . at short times
the energy of the collector state is low , although it should be noted that it is populated with a negligible density , panel ( d ) .
the weak tunneling of particles across the gate is better evidenced by the increase of the collector state energy , since the tunneling particles have higher energy than the collector ground state . at these low gate powers , the theory appears to predict a high number of polaritons passing the gate .
these polaritons have relatively high momentum and are expected to be less visible experimentally due to reduced photonic fractions .
an increase of the gate power above threshold , panels ( b , c , e , f ) , leads to an enhanced collector signal , as in the experimental case , see figs .
[ fig : fig9 ] ( b , c , e , f ) , and the device leaves the off state .
this is expected as additional excited polaritons move directly from the gate to the collector .
figure [ fig : it_all ] shows the time evolution of the spatially integrated spectra . in agreement with the experimental results ( fig .
[ fig : fig11 ] ) there is a two timescale decay of the emission energy . at early times
, the fast drop is due to the fast condensation rate in the presence of strong stimulated scattering .
as mentioned earlier , this fast condensation rapidly depletes the hot exciton reservoir and the total particle density quickly drops as polaritons quickly decay . at longer times ,
reduced relaxation between the active and inactive reservoirs limits the effective condensation rate .
the condensate is continuously fed while the reservoir intensities slowly decay .
the short lifetime of the condensate for pumping at threshold ( fig .
[ fig : it_all ] ( a ) ) is expected from the theoretical definition of threshold where the incoming rate @xmath136 is slightly larger than the polariton decay rate @xmath108 in eq .
( [ eq : gp ] ) .
as soon as condensation starts , the reservoir density @xmath88 drops below threshold such that continued condensation can not take place . for higher pump powers ,
[ fig : it_all ] ( b , c ) show that both emission into a high energy mode , corresponding to the source , and a lower energy emission coexist .
this fact is in close agreement with the experimental data shown in figs .
[ fig : fig11 ] ( b , c ) and ( i ) , with the relaxation into the lower energy state occurring earlier for increased pumping power .
in summary , we have time - resolved the energy and intensity relaxation processes of excitons and polaritons in a microcavity ridge .
two different excitation configurations have been studied with one and two non - resonant , pulsed laser beams , permitting polariton condensate trapping on demand .
a detailed analysis of the decay processes has been accomplished by mapping the energy and intensity emission along the ridge .
decay times of the source emission are reported under one - beam excitation , where we show the acceleration of the decaying processes as a function of increasing @xmath19 .
the time response of the polariton transistor switch is characterized and optimized.we used a generalized gross - pitaevskii model to describe the spatial dynamics of our propagating polariton condensates , which includes a phenomenological treatment of energy - relaxation processes that cause condensates to further thermalize as they travel in a non - uniform effective potential .
the nonlinearity of energy - relaxation processes throughout the system , those causing relaxation between polariton states as well as relaxation between higher energy exciton states , is necessary to explain features of the experimental results . approximating the system as a 1d system , we are able to describe the main qualitative features of the experiment .
while the system is essentially 1d , lateral expansion could result in a lower propagation speed than that predicted theoretically .
we intend to investigate lateral propagation effects in future work .
the optimization of individual condensate transistor elements , as we have reported here , is an essential step towards developing information processing devices with the present scheme . in the future ,
an important goal is the achievement of cascadability and fan - out of multiple elements for the construction of extended circuits .
such a feat was very recently achieved in polariton based systems with coherent near - resonant excitation .
@xcite achieving the same with the gating of incoherently generated polariton condensates , as we study here , would be particularly promising as it would open up routes toward electrically injected devices and consequently hybrid electro - optical processing systems . c.a . and g.t .
acknowledge financial support from spanish fpu and fpi scholarships , respectively .
p.s . acknowledges greek gsrt program
aristeia " ( 1978 ) for financial support .
the work was partially supported by the spanish mec mat2011 - 22997 , cam ( s-2009/esp-1503 ) and fp7 itn s
clermont4 " ( 235114 ) , spin - optronics " ( 237252 ) and index ( 289968 ) projects .
in this appendix we define the symbols used as abbreviations along the manuscript .
1@|c|>p0.8| * symbol & * meaning @xmath74 & energy shift of the emission at the source position@xmath59 & energy fast decay time at the source@xmath61 & energy slow decay time at the source__c _ _ & collector @xmath137 & polariton condensate trapped at the collector@xmath28 & polariton condensate trapped between source and gate positions@xmath138 & horizon of propagating polaritons along the ridge@xmath32 & horizon given by the interface between carries at the source and propagating polaritons__g _ _ & gate @xmath139 & emission intensity of the condensate trapped at the collector@xmath140 & emission intensity of the condensate trapped between source and gate@xmath141 & emission intensity at the source position@xmath22 & gate beam power@xmath19 & source beam power@xmath33 & pump power threshold for polariton condensation__s _ _ & source @xmath70 & time time for the collector intensity to drop from 1 to 0.5@xmath68 & time delay between maxima of the source and collector intensities@xmath69 & time time for the collector intensity to rise from 0.5 to 1@xmath42 & mean speed of propagating polaritons @xmath38 & mean speed of carriers at the source position * * 1@|c|>p0.8| * symbol & * meaning@xmath83 & polariton - polariton interaction strength@xmath108 & polariton decay rate@xmath105 & decay rate of the active exciton reservoir@xmath107 & decay rate of the dark exciton reservoir@xmath106 & decay rate of the inactive exciton reservoir@xmath110/@xmath111 & phenomenological parameters for the strength of energy relaxation@xmath78 & polariton wavefunction@xmath142 & lower polariton branch energy dispersion@xmath93 & polariton - exciton interaction strength for dark excitons @xmath92 & polariton - exciton interaction strength for indirect excitons @xmath91 & polariton - exciton interaction strength for reservoir excitons @xmath82 & polariton effective mass@xmath121 & free electron mass@xmath88 & density distribution of active excitons@xmath89 & density distribution of inactive excitons@xmath90 & density distribution of dark excitons * * @xmath94 & polariton condensation rate@xmath104 & linear coupling to the dark exciton reservoir@xmath101 & linear coupling to the inactive exciton reservoir@xmath102 & non - linear coupling to the inactive exciton reservoir@xmath143 & effective polariton potential@xmath144 & wire structural potential j. kasprzak , m. richard , s. kundermann , a. baas , p. jeambrun , j. m. j. keeling , f. m. marchetti , m. h. szymanska , r. andre , j. l. staehli , v. savona , p. b. littlewood , b. deveaud , and l. s. dang , nature * 443 * , 409 ( 2006 ) .
s. christopoulos , g. baldassarri hoger von hogersthal , a. grundy , p. g. lagoudakis , a. v. kavokin , j. j. baumberg , g. christmann , r. butte , e. feltin , j .- f .
carlin , and n. grandjean , phys .
* 98 * , 126405 ( 2007 ) .
d. sanvitto , f. m. marchetti , m. h. szymanska , g. tosi , m. baudisch , f. p. laussy , d. n. krizhanovskii , m. s. skolnick , l. marrucci , a. lemaitre , j. bloch , c. tejedor , and l. via , nature phys . * 6 * , 527 ( 2010 ) .
m. galbiati , l. ferrier , d. d. solnyshkov , d. tanese , e. wertz , a. amo , m. abbarchi , p. senellart , i. sagnes , a. lematre , e. galopin , g. malpuech , and j. bloch , phys .
* 108 * , 126403 ( 2012 ) . | we present a time - resolved study of energy relaxation and trapping dynamics of polariton condensates in a semiconductor microcavity ridge . the combination of two non - resonant , pulsed laser sources in a gaas ridge - shaped microcavity gives rise to profuse quantum phenomena where the repulsive potentials created by the lasers allow the modulation and control of the polariton flow .
we analyze in detail the dependence of the dynamics on the power of both lasers and determine the optimum conditions for realizing an all - optical polariton condensate transistor switch .
the experimental results are interpreted in the light of simulations based on a generalized gross - pitaevskii equation , including incoherent pumping , decay and energy relaxation within the condensate . |
the role played by disorder on the nature of collective excitations in condensed matter physics has been the subject of intensive studies due to its relevance in defining general transport characteristics @xcite .
usually , disorder induces localization of collective excitations thus degrading transport properties , an effect that is largely pronounced in low dimensions .
in particular , the one - electron eigen - states in the one - dimensional anderson model with site - diagonal uncorrelated disorder are exponentially localized for any degree of disorder @xcite .
however , several one - dimensional models with correlated disorder have been proposed which exhibit delocalized states @xcite .
recently , it has been shown that the one - dimensional anderson model with long - range correlated disorder presents a phase of extended electronic states @xcite .
these results have been confirmed by microwave transmission spectra of single - mode waveguides with inserted correlated scatters @xcite .
the above results have motivated the study of further model systems that can be mapped onto the anderson model and , therefore , expected to present a similar transition between localized and extended collective excitations .
recently , a study concerning the one - dimensional quantum heisenberg ferromagnet with exchange couplings exhibiting long - range correlated disorder reported some finite - size scaling evidences of the emergence of a phase of extended low - energy excitations @xcite . by using a renormalization group calculation the existence of such phase of extended spin - waves
was confirmed and the scaling of the mobility edge with the degree of correlation was obtained @xcite .
it was also shown that , associated with the emergence of extended spin - waves in the low - energy region , the wave - packet mean - square displacement exhibits a long - time ballistic behavior . the collective vibrational motion of one - dimensional disordered harmonic chains of @xmath5 random masses can also be mapped onto an one - electron tight - binding model @xcite .
in such a case , most of the normal vibrational modes are localized .
however , there are a few low - frequency modes not localized , whose number is of the order of @xmath6 , in which case the disordered chains behaves like the disorder - free system @xcite .
futher , it was shown that correlations in the mass distribution produce a new set of non - scattered modes in this system @xcite .
non - scattered modes have also been found in disordered harmonic chain with dimeric correlations in the spring constants @xcite . by using analytical arguments , it was also demonstrated that the transport of energy in mass - disordered ( uncorrelated and correlated ) harmonic chains is strongly dependent on non - scattered vibrational modes as well as on the initial excitation @xcite . for impulse initial excitations ,
uncorrelated random chains have a superdiffusive behavior for the second moment of the energy distribution [ @xmath7 , while for initial displacement excitations a subdiffusive spread takes place [ @xmath8 .
the dependence of the second moment spread on the initial excitation was also obtained in ref .
moreover , correlations induced by thermal annealing have been shown to enhance the localization length of vibrational modes , although they still present an exponential decay for distances larger than the thermal correlation length @xcite .
recently the thermal conductivity on harmonic and anharmonic chains of uncorrelated random masses @xcite , as well as of the chain of hard - point particles of alternate masses @xcite , has been numerically investigated in detail .
the main issue here is whether the systems display finite thermal conductivity in the thermodynamic limit , a question that remains controversial @xcite . in this paper
we extend the study of collective modes in the presence of long - range correlated disorder for the case of vibrational modes .
we will consider harmonic chains with long - range correlated random masses assumed to have spectral power density @xmath9 . by using a transfer matrix calculation
, we obtain accurate estimates for the lyapunov exponent , defined as the inverse of the degree of localization @xmath10 of the vibrational modes .
we show that , for @xmath11 , this model also presents a phase of extended modes in the low frequency region .
this result is confirmed by participation ratio measurements from an exact diagonalization procedure and finite size scaling arguments .
the spatial evolution of an initially localized excitation is also studied by computing the spread of the second moment of the energy distribution , @xmath3 .
we find that , associated with the emergence of a phase of delocalized modes , a ballistic energy spread takes place .
we consider a disordered harmonic chain of @xmath5 masses , for which the equation of motion for the displacement @xmath12 of the _ n_-th mass with vibrational frequency @xmath13 is @xcite @xmath14 here @xmath15 is the mass at the _ n_-th site and @xmath16 is the spring constant that couples the masses @xmath15 and @xmath17 .
we use units in which @xmath18 . in the present harmonic chain model , we take the masses @xmath15 following a random sequence describing the trace of a fractional brownian motion @xcite : @xmath19^{1/2 } \cos{\left ( \frac{2\pi nk}{n } + \phi_k\right)},\ ] ] where @xmath1 is the wave - vector of the modulations on the random mass landscape and @xmath20 are @xmath21 random phases uniformly distributed in the interval @xmath22 $ ] . the exponent @xmath23 is directly related to the hurst exponent @xmath24 ( @xmath25 ) of the rescaled range analysis . in order to avoid vanishing masses
we shift and normalize all masses generated by eq .
( 2 ) such to have average value @xmath26 and variance independent of the chain size ( @xmath27 ) . using the matrix formalism , eq .
( 1 ) can be rewritten as obtained using dean s method .
the chain length is @xmath28 for all cases .
the dos becomes less rough as @xmath23 is increased .
for @xmath29 it displays a non - fluctuating part near the bottom of the band.,scaledwidth=45.0% ] @xmath30 for a specific frequency @xmath13 , a @xmath31 transfer matrix @xmath32 connects the displacements at the sites @xmath33 and @xmath34 to those at the site @xmath35 : @xmath36 once the initial values for @xmath37 and @xmath38 are known , the value of @xmath12 can be obtained by repeated iterations along the chain , as described by the product of transfer matrices @xmath39 as a function of @xmath40 for @xmath41 ( uncorrelated random chain ) and @xmath42 sites .
the lyapunov coefficient is finite for non - zero frequencies ( localized states ) and vanishes as @xmath43 , @xmath44 ( inset).,scaledwidth=45.0% ] the localization length of each vibrational mode is taken as the inverse of the lyapunov exponent @xmath45 defined by @xcite @xmath46 where @xmath47 is a generic initial condition .
typically , @xmath48 matrix products were used to calculate the lyapunov exponents .
the nature of the vibrational modes can also be investigated by computing the participation ratio @xmath49 , since it displays a dependence on the chain size for extended states and is finite for exponentially localized ones .
@xmath49 is defined by @xcite @xmath50 where the displacements @xmath12 are those associated with an eigenmodes @xmath13 of a chain of @xmath5 masses and are obtained by direct diagonalization of the @xmath51 secular matrix @xmath52 defined by @xmath53 , @xmath54 , and all other @xmath55 @xcite .
the participation ratio calculations were averaged over @xmath56 samples .
we compute the lyapunov exponents and the participation ratio for several values of the correlation exponent @xmath23 , and obtain the density of states ( dos ) using the numerical dean s method @xcite .
strong fluctuations in the dos are related to the presence of localized states , whereas smooth a dos is usually connected with the emergence of delocalized states @xcite . in fig . 1 we show the normalized dos for chains with @xmath28 sites , and
notice that it becomes less rough as @xmath23 is increased . in fig . 2 we display the plot of @xmath45 versus @xmath57 for @xmath58 ( uncorrelated random masses ) .
the lyapunov coefficient is finite for all frequencies and vanishes at @xmath59 as @xmath60 , in agreement with ref .
@xcite ( see inset of fig . 2 ) .
the scaled participation ratio @xmath61 as a function of @xmath40 is shown in fig . 3 :
@xmath62 remains finite in the thermodynamic limit , whereas for any finite frequency the vibrational modes are localized with @xmath63 as @xmath64 , in agreement with the above results obtained from the lyapunov coefficient calculations . to investigate the effect of weak long - range correlated disorder , we present in fig .
4(a ) the lyapunov coefficient as a function of @xmath40 for @xmath65 and @xmath42 . in spite of @xmath45 being very small in the bottom of the band , the scaled participation ratio for @xmath66 vanishes in the thermodynamic limit [ see fig .
therefore , all modes with @xmath66 are still localized , a feature that holds for any @xmath67 . however , the nature of the low - frequency modes changes qualitatively for @xmath2 . in fig .
5(a ) we show @xmath45 versus @xmath57 for @xmath68 and @xmath42 sites .
the lyapunov coefficient vanishes within a finite range of frequency values , thus revealing the presence of extended vibrational modes .
the scaled participation ratio @xmath61 [ see fig . 5(b ) ] displays a well defined data collapse , confirming that the phase of extended low - frequency vibrational modes is stable in the thermodynamic limit . as a function of @xmath40 for @xmath41 ( uncorrelated disorder ) . from top
to bottom , @xmath69 . for vibrational modes with @xmath70 , @xmath63 as @xmath5 diverges , thus confirming their localized nature.,scaledwidth=45.0% ] versus @xmath40 for @xmath71 and @xmath72 sites .
( b ) scaled participation ratio @xmath61 as a function of @xmath40 for @xmath71 . from top to bottom , @xmath73 . in spite of @xmath45 being very small in the bottom of the band , all modes with @xmath66
are localized.,scaledwidth=45.0% ] versus @xmath40 for @xmath74 and @xmath72 sites .
the lyapunov coefficient vanishes within a finite range of frequency values , thus revealing the presence of extended vibrational modes .
( b ) scaled participation ratio @xmath61 as a function of @xmath40 for @xmath74 . from top to bottom , @xmath73 .
the phase of extended vibrational modes is confirmed by the size independent plateau in the low - frequency region.,scaledwidth=45.0% ]
in order to study the time evolution of a localized energy pulse , we calculate the second moment of the energy distribution @xcite . this quantity is related to the thermal conductivity by kubo s formula @xcite .
the classical hamiltonian @xmath24 for an harmonic chain can be written as @xmath75 where the energy @xmath76 at the site @xmath34 is given by @xmath77~.\ ] ] here @xmath78 and @xmath79 define the momentum and displacement of the mass at the _ n_-th site .
the hamilton s equations are @xmath80\ ] ] and @xmath81 the fraction of the total energy @xmath24 at the site @xmath34 is given by @xmath82 and the second moment of the energy distribution , @xmath3 , is defined by @xcite @xmath83,\ ] ] versus time @xmath84 for @xmath85 ( dotted line ) and @xmath86 ( dashed line ) with initial impulse excitation . for @xmath67 only superdiffusive behavior is found for long times.,scaledwidth=45.0% ] where an initial excitation is introduced at the site @xmath87 at @xmath88 . using the fourth - order runge - kutta method
, we solve the differential equations for @xmath89 and @xmath90 and calculate @xmath3 .
the second moment of the energy distribution @xmath3 has the same status of the mean - square displacement of the wavepacket of an electron in a crystal @xcite . in harmonic chains with an initial impulse excitation ,
the energy spread is faster than that in chains with an initial displacement excitation @xcite .
we calculate @xmath3 for several @xmath23 values and two kinds of initial excitation : impulse excitation and displacement excitation . in fig .
6 we present the scaled second moment @xmath91 versus time @xmath84 for @xmath85 ( dotted line ) , which corresponds to the uncorrelated random chain , and @xmath86 ( dashed line ) .
these results have been obtained after an initial impulse excitation , @xmath92 . in our calculations for @xmath85 , the self - expanded chain method with initial chain size @xmath93 was used to minimize end effects . throghout the numerical integration process we kept the fraction of the total energy @xmath24 at the ends of the chain [ @xmath94 and @xmath95 smaller than @xmath96 for all times .
as shown in fig .
6 , we find a long - time superdiffusive behavior for @xmath41 , in agreement with previous analytical and numerical results for energy transport in harmonic chains with uncorrelated random masses under an impulse initial excitation @xcite .
in contrast , for @xmath97 we can not use the self - expanded chain method due to the long - range character of the mass correlations .
therefore , chains with @xmath98 masses were considered , and the runs stopped whenever the fraction of the total energy at the chain ends achieved @xmath96 .
for @xmath71 the time - dependence of the scaled energy second moment @xmath91 typically represents a weak long - range correlated case .
in such a case we also find superdiffusive behavior for long times . on the other hand , in the strong correlated regime , @xmath2 , a breakdown in the superdiffusive behavior
7 shows the time - dependence of the scaled energy second moment , @xmath99 , for @xmath29 ( dotted line ) and @xmath100 ( dashed line ) .
associated with the emergence of extended vibrational modes in the low - energy region , the second moment @xmath3 displays a long - time ballistic behavior . versus time @xmath84 for @xmath29 ( dotted line ) and @xmath100 ( dashed line ) with initial impulse excitation .
results were obtained by numerical integration in chains with @xmath98 sites . a ballistic behavior is found after an initial transient.,scaledwidth=45.0% ] the long time behavior of the second moment @xmath3 in uncorrelated random chains with initial displacement excitation is significantly different from the corresponding behavior with impulse initial excitation @xcite .
analytical calculations predict that @xmath101 , a result that has been corroborated by numerical techniques @xcite . for @xmath85
we indeed reproduce this behavior , as shown in fig . 8 for the scaled second moment @xmath102 versus time @xmath84 with initial displacement excitation @xmath103 .
again , we find that this asymptotic subdiffusive behavior remains true for @xmath104 ( dashed line in fig .
for strong correlations ( @xmath2 ) , which induce the emergence of new extended vibrational modes in the low - energy region , the energy transport is faster than in the subdiffusive regime and again assumes a ballistic nature , as shown in fig .
9 . versus time @xmath84 for @xmath85(dotted line ) and
@xmath86 ( dashed line ) with initially given displacement excitation .
the subdiffusive behavior is found for long times.,scaledwidth=45.0% ]
in this paper we have studied the nature of collective vibrational modes in harmonic chains with long - range correlated random masses @xmath15 , with spectral power density @xmath9 . by using a transfer matrix method and exact diagonalization ,
we have computed the localization length and the participation ratio of all normal modes .
our results indicate that in the strong correlations regime , @xmath4 , there is a phase of extended low - energy vibrational modes . in this sense ,
long - range correlations in the mass distribution induce the emergence of a delocalization transition in harmonic chains similar to the one observed to occur with one - magnon excitations in ferromagnetic chains with random couplings @xcite and with one - electron eigen - states in the random hopping anderson model @xcite .
we have also studied the energy transport in this harmonic chain model .
the spread of the energy second moment @xmath3 is shown to be strongly dependent on the existence of non - scattered vibrational modes and initial excitation .
we have also found that , associated with the emergence of a phase of low - energy extended collective excitations , @xmath3 displays a crossover from an anomalous sub- or super - diffusive regime ( depending on the initial impulse or displacement excitation , respectively ) to an asymptotic ballistic behavior .
the above findings indicate that the thermal conductivity can be strongly influenced by the presence of long - range correlations in the random distribution of masses and we hope that the present work will stimulate further studies along this direction .
this work was partially supported by cnpq , capes and finep ( brazilian agencies ) .
mll also acknowledges the partial support of fapeal ( alagoas state agency ) . versus time @xmath84 for @xmath29 ( dotted line ) and @xmath100 ( dashed line ) with initially given displacement excitation .
results obtained by numerical integration of diferential equation for chain with @xmath98 sites .
a ballistic behavior is found for all times.,scaledwidth=45.0% ] 40 t. a. l. ziman , phys .
lett . * 49 * , 337 ( 1982 ) . for a review
see , e.g. , b. kramer and a. mackinnon , rep .
. phys . * 56 * 1469 , ( 1993 ) .
e. abrahams , p. w. anderson , d. c. licciardello , and t. v. ramakrishnan , phys .
* 42 * , 673 ( 1979 ) . for a review
see , e.g. , i. m. lifshitz , s. a. gredeskul and l. a. pastur , _ introduction to the theory of disordered systems _ ( wiley , new york , 1988 ) . j. c. flores , j. phys . : condens . matter * 1 * , 8471 ( 1989 ) .
d. h. dunlap , h. l. wu , and p. w. phillips , phys .
65 * , 88 ( 1990 ) ; h .-
l . wu and p.
phillips , phys .
lett . * 66 * , 1366 ( 1991 ) ; p. w. phillips and
wu , science * 252 * , 1805 ( 1991 ) ; s. n. evangelou and d. e. katsanos , phys .
a * 164 * , 456 ( 1992 ) .
see also , s. n. evangelou , a. z. wang , and s. j. xiong , j. phys . :
matter * 6 * , 4937 ( 1994 ) .
f. a. b. f. de moura and m. l. lyra , phys .
lett . * 81 * , 3735 ( 1998 ) .
f. a. b. f. de moura and m. l. lyra , physica a * 266 * , 465 ( 1999 ) .
f. m. izrailev and a. a. krokhin , phys .
lett . * 82 * , 4062 ( 1999 ) ; f. m. izrailev , a. a. krokhin , and s. e. ulloa , phys . rev .
b * 63 * , 41102 ( 2001 ) .
u. kuhl , f. m. izrailev , a. krokhin , and h. j. stckmann , appl .
77 * , 633 ( 2000 ) .
r. p. a. lima , m. l. lyra , e. m. nascimento , and a. d. de jesus , phys .
b * 65 * , 104416 ( 2002 ) .
f. a. b. f. de moura , m. d. coutinho - filho , e. p. raposo and m. l. lyra , phys .
b , * 66 * , 014418 ( 2002 ) .
p. dean , proc .
* 84 * , 727 ( 1964 ) .
h. matsuda and k. ishii , prog .
phys . suppl .
* 45 * , 56 ( 1970 ) ; k. ishii , prog .
phys . suppl .
* 53 * , 77 ( 1973 ) .
p. k. datta and k. kundu , j. phys .
: condens .
matter * 6 * , 4465 ( 1994 ) .
f. dominguez - adame , e. maci and a. snchez , phys .
6054 ( 1993 ) . p. k. datta and k. kundu , phys .
b * 51 * , 6287 ( 1995 ) .
m. wagner , g. zart , j. vazquez - marquez , g. viliani , w. frizzera , o. pilla , and m. montagna , philos . mag .
b * 65 * , 273 ( 1992 ) .
j. c. cressoni and m. l. lyra , phys .
b * 53 * , 5067 ( 1996 ) ; j. phys . : condens .
matter * 8 * , l83 ( 1996 ) .
b. li , h. zhao and b. hu , phys .
lett . * 86 * , 63 ( 2001 ) ; a. dhar , phys .
lett . * 86 * , 5882 ( 2001 ) ; p. l. garrido , p. i. hurtado and b. nadrowski , phys .
* 86 * , 5486 ( 2001 ) ; a. v. savin , g. p. tsironis and a. v. zolotaryuk , phys .
* 88 * , 154301 - 1 ( 2002 ) .
a. dhar , phys .
lett . * 88 * , 249401 - 1 ( 2002 ) ; p. l. garrido and p. i. hurtado , phys . rev
. lett . * 88 * , 249402 - 1 ( 2002 ) .
j. feder , _ fractals _ ( plenum press , new york , 1988 ) ; a. tsonis , _ chaos : from theory to applications _ ( plenum press , new york , 1992 ) .
a. r. osborne and a. provenzale , physica d * 35 * , 357 ( 1989 ) .
n. p. greis and h. s. greenside , phys .
a * 44 * , 2324 ( 1991 ) .
s. n. evangelou and e. n. economou , j. phys .
* 26 * , 2803 ( 1993 ) .
p. dean , rev .
* 44 * , 127 ( 1972 ) .
p. b. allen and j. l. feldman , phys .
b * 48 * , 12581 ( 1993 ) . | we study the nature of collective excitations in harmonic chains with masses exhibiting long - range correlated disorder with power spectrum proportional to @xmath0 , where @xmath1 is the wave - vector of the modulations on the random masses landscape . using a transfer matrix method and exact diagonalization ,
we compute the localization length and participation ratio of eigenmodes within the band of allowed energies .
we find extended vibrational modes in the low - energy region for @xmath2 . in order to study the time evolution of an initially localized energy input ,
we calculate the second moment @xmath3 of the energy spatial distribution .
we show that @xmath3 , besides being dependent of the specific initial excitation and exhibiting an anomalous diffusion for weakly correlated disorder , assumes a ballistic spread in the regime @xmath4 due to the presence of extended vibrational modes . |
in the recent years there has been much interest in various loop models .
loop models are graphical models defined by drawing closed loops along the bonds of the underlying lattice .
the loops may come in @xmath4 different flavours ( colours ) .
no two loops can share a bond , while sharing a vertex is generally allowed .
explicitly , the bond configurations are such that each vertex houses an even number possibly zero of bonds of each colour .
each loop configuration is assigned a `` weight '' that depends on the number of participating vertices of each type . in the cases of interest
these weights are actually positive hence , at least in finite volume , they define a _ probability measure _ on the set of all loop configurations . thus , for a finite lattice the loop partition function may be written as : @xmath5 with the sum running over all allowed loop configurations @xmath6 . here @xmath7 is the total number of participating bonds , @xmath8 ( @xmath9 ) is the number of vertices of type @xmath10 and @xmath11 is the corresponding vertex factor . where @xmath12 is a `` loop fugacity '' and @xmath13 is the number of loops of the @xmath10-th colour .
although the objects @xmath13 are unambiguous when self - intersections are forbidden , in the general case they are not easily defined .
nevertheless , the essence of such a term at least in the case of integer @xmath12 is captured by the introduction of additional colours . ]
this definition is slightly different from the one typically found in literature ( _ cf .
_ refs .
@xcite ) since it also includes the bond fugacity @xmath14 .
although strictly speaking it is not needed ( since the bond fugacity can always be incorporated into the vertex factors ) , we find it convenient to keep @xmath14 as a separate parameter .
we remark that by relabeling the empty bonds as an additional colour , these models may be formally regarded as `` fully packed '' .
the reason loop models have been extensively studied is because they appear quite naturally as representations ( often approximate ) of various statistical - mechanical models .
these include , among others , the ising model ( this approach dates back to kramers and wannier @xcite and was later used to solve the model exactly @xcite ) , the potts model ( polygon expansion @xcite ) , @xmath15 spin models @xcite , 1-d quantum spin models @xcite , a supersymmetric spin chain @xcite , the @xmath16-colouring problem @xcite and polymer models @xcite .
here we consider the loop models explicitly related to the high - temperature expansions of the standard @xmath15 , corner - cubic ( aka diagonal - cubic ) and face - cubic spin models .
this is , in fact , the same set of models that was treated in ref .
however , in this paper , we provide a careful treatment of the large @xmath4 cases and we treat the standard @xmath0-dimensional lattices . as a result , we arrive at quite unexpected results concerning the behaviour of these models in the high fugacity region . in particular , despite the considerable attention
the subject has received , most authors ( with certain exceptions , e.g. @xcite ) chose to consider models where only loops of _ different _ colours are allowed to cross each other ( if at all ) . on the other hand spin systems ( in the high - temperature approximation )
naturally generate self - intersecting loops . in order to avoid this issue ,
an exorbitant amount of work has been done on lattices with coordination number @xmath17 ( e.g. the honeycomb lattice ) , where loop intersections simply can not occur .
overall this approach appears to be justified since one is usually interested in the critical properties of the underlying spin systems .
indeed , consider the archetypal @xmath4-component spin system with @xmath18 and let us write @xmath19 .
although as a spin system the right hand side makes strict sense only if @xmath20 ( the `` physical regime '' ) , the associated loop model turns out to be well defined for all @xmath21 . since the systems can be identified for @xmath22 it can be argued that the critical properties of the spin system and those of the loop model are the same and are independent of the underlying lattice . notwithstanding , for @xmath3 any phase transition in the actual spin system is not anticipated until temperatures of order @xmath23 ( i.e. @xmath24 ) , which we note is well outside the physical regime of the loop model . at first glance
this appears to be borne out : the natural parameter in the loop model ( as well as in the spin system ) seems to be @xmath25 .
thus , the loop model could , in principle , capture the essential features of the spin system up to and including the critical point .
we have found such a picture to be overly optimistic . indeed , depending on the specific details , e.g. the lattice structure
, there may be a phase transition in the region @xmath26 ( specifically , @xmath27 ) , well outside the physical regime but well before the validity of the approximation was supposed to break down .
furthermore , it would seem that both the temperature scale and the nature of the transition ( not to mention the existence of the transition ) depend on such details .
finally , we shall demonstrate that in contrast to their spin system counterparts , the large-@xmath4 models have _ no _ phase transition for any value of bond fugacity associated with the formation of large loops ( i.e. divergent loop correlations ) .
the structure of this paper is as follows .
section [ sec : models ] is dedicated to the description of the spin models and their connection to the loop models .
specific results for those models with the two - dimensional spin variable ( @xmath1 ) are presented in section [ sec : n=2 ] .
finally , section [ sec : n_large ] contains the discussion of reflection positivity as well as some results concerning phase transitions in the large @xmath4 case .
let us start by considering the @xmath15 model on some finite lattice @xmath28 _ defined _ by the following partition function : @xmath29 with @xmath30 , @xmath31 and @xmath32 denoting normalised summation ( integration ) over all possible spin configurations .
the corresponding loop model is readily obtained along the lines of a typical `` high - temperature '' expansion .
we write @xmath33 and define @xmath4 different colours ( each associated with a coordinate direction of the @xmath15-spins ) . expanding the product
, we have @xmath4 choices for each bond plus a possibility of a vacant bond .
thus , various terms are represented by @xmath4-coloured bond configurations : @xmath34 with @xmath35 denoting those bonds where the term @xmath36 has been selected .
clearly , the various @xmath35 s are pairwise ( bond ) disjoint .
thus , for each @xmath37 we obtain the weight @xmath38 on the basis of elementary symmetry considerations it is clear that @xmath39 if and only if each vertex houses an even number ( which could be zero ) of bonds of each colour .
once this constraint is satisfied , we get an overall factor of @xmath40 with @xmath41 being the total number of participating bonds times the product of the _ vertex factors _ obtained by performing the appropriate @xmath15 integrals . the details and results of these calculations are presented in appendix [ sec : kmf ] . in general
, it is seen that the vertex factors depend only on the number of participating colours and the number of bonds of each colour emanating from a given vertex , i.e. not on the particular colours that were involved nor on the directions of these bonds . in the case of a square lattice ( @xmath2 ) we have only three main types of ( non - empty ) vertices : those where two bonds of the same colour join together , those with two pairs of bonds of two different colours and those with four bonds of the same colour .
these have weights of @xmath23 , @xmath42 and @xmath43 correspondingly .
rescaling the bond fugacity from @xmath44 to @xmath45 we arrive at the vertex weights @xmath46 , @xmath47 and @xmath48 .
the factor of 3 relating @xmath49 to @xmath50 has an interesting interpretation ( which , as shown in appendix [ sec : kmf ] , turns out to be quite general ) .
indeed , each vertex of the third type may be decomposed into three different vertices as shown in fig .
[ o(n)_vertices ] .
each of the new vertices is now assigned equal weight , which is also that of @xmath49 .
we thus split each @xmath37 into @xmath51 different graphs
each of equal weight in which every vertex with four bonds now provides explicit instructions relating outgoing and incoming directions of an individual walk . hence in every such graph ( now defined with the walking instructions encoded at every vertex ) the individual loops are well defined .
furthermore , changing the colour of any loop does not change the weight of the graph .
thus we may write @xmath52 where the summation now takes place over all configurations @xmath53 of colourless loop graphs in which every vertex housing four bonds is resolved by `` walking instructions '' , and @xmath54 is the number of such loops ( being now defined completely unambiguously ) .
in addition to the advantages of a manifestly colourless expression , the above permits continuation to non - integer @xmath4 .
we conclude this subsection with the following series of remarks and observations . *
as shown in appendix [ sec : kmf ] , such vertex decomposition works , in fact , for an arbitrary lattice in an arbitrary number of spatial dimensions ( with the proper weights for vertices housing 6 , 8 , etc .
bonds ) . *
notice that only @xmath20 region of the parameter space is `` physical '' ( in a sense of the underlying hamiltonian : @xmath55 $ ] ) , while for @xmath56 one presumes , no spin hamiltonian can be written at all .
the corresponding loop model , however , makes perfect sense in the entire parameter space . *
if we consider a 2d xy model ( @xmath1 ) , we notice that the factor @xmath57 in eq .
( [ o(n)_loops ] ) can be obtained by assigning directions to the colourless loops .
the above decomposition of type 3 vertices makes this procedure unambiguous . having done that , we can turn this model into a random surface model by assigning heights to the plaquettes in such a way that a plaquette to the right of a directed bond is always one step higher than the plaquette to the left .
not surprisingly , this random surface model turns out to be identical to the one obtained by the standard means of fourier - transforming the original weights in eq .
( [ o(n)_part ] ) ( _ cf
. _ refs .
@xcite ) . *
finally , it is worth mentioning that in the fully packed limit @xmath58 , the @xmath1 loop model on the square lattice ( with the colour degrees of freedom being replaced by assigning directions to the loops ) turns out to be nothing but the square ice model ( i.e. the six - vertex model with all six weights being equal see ref .
@xcite for a definition of this model ) .
the mapping between the vertices of these models is shown in fig .
[ 6-vertex ] .
we remark that the perspective of the ice model ( and , for that matter , other six - vertex models ) as a two colour loop model provides additional flexibility in the analysis of these systems .
these issues will be pursued in a future publication .
model onto the square ice model.,width=491 ] we now consider the following `` discretised '' modification of the above @xmath15 model ( given by eq .
( [ o(n)_part ] ) ) : @xmath59 \label{corner_part}\ ] ] with @xmath60 . for small values of @xmath21
this model may be viewed as a high - temperature limit of a corner - cubic model .
indeed , it describes an interaction of the type in eq .
( [ o(n)_part ] ) where spins @xmath61 are allowed to point at the corners of an @xmath4-dimensional hypercube ( with the origin being placed at the centre of the cube ) . mapping it onto an @xmath4-colour loop model is almost identical to the @xmath15 case , with the only difference being the vertex factor : @xmath62 .
we can choose to associate the weight of @xmath63 with each bond , thus making all vertex weights @xmath11 to be equal to unity .
in other words , the resulting loops in this model do not interact with each other via vertices ( there is still a hard core bond repulsion , however ) .
the partition function is then simply @xmath64 ) for a loop model of _ corner - cubic _ type are represented here as ( a ) , ( b ) and ( c ) . for the _ face - cubic _ type ,
only monotonous clusters like ( a ) remain allowed.,width=491 ] finally , let us examine a different model with cubic symmetry given by the following partition function : @xmath65\ , .
\label{face - cubic}\ ] ] here @xmath66 , and for a given site @xmath10 exactly one of @xmath67 ( @xmath68 ) has a non - zero value .
in fact , one may think of @xmath69 s as components of an @xmath4-dimensional unit vector that is only allowed to point along the coordinate axes ( or from the centre to the faces of an @xmath4-dimensional hypercube - thus the name face - cubic ) .
while the corner - cubic model described earlier had @xmath70 degrees of freedom per site ( the number of corners of a hypercube ) , the present model has only @xmath71 such degrees of freedom ( the number of faces ) . once again , the corresponding loop model is obtained by performing multiplication in eq .
( [ face - cubic ] ) and then summing the resulting terms over all possible values of @xmath69 s .
but since for each site @xmath10 only one of the spin components @xmath72 at a time , we notice that no terms that mix different @xmath73 s are allowed . in terms of resulting loops
this means hard - core repulsion of different colours : only loops of the same colour can share a vertex .
the vertex factors are now : zero for any vertex with multiple colours and @xmath23 for vertices with two or more bonds of the same colour ; the bond fugacity is given by @xmath74 .
in this section we shall restrict our attention to the models with cubic symmetries .
firstly , let us slightly change the notations for convenience : let @xmath75 and @xmath76 for the corner - cubic , while @xmath77 and @xmath78 for the face - cubic model .
the corresponding partition functions are then written as @xmath79 \label{cc2}\ ] ] and @xmath80\ , .
\label{fc2}\ ] ] while we have seen that the loop models generated by these partition functions are very different , the spin models themselves turn out to be identical . indeed , eq . ( [ fc2 ] )
is obtained from eq .
( [ cc2 ] ) by the following transformation : @xmath81 , @xmath82 .
this is equivalent to a @xmath83 rotation in the spin space ( along with a @xmath84 rescaling ) and is very specific to the @xmath1 case . in turn , both models are equivalent to the ashkin teller model @xcite with a particular choice of parameters that will be detailed in section [ sec : at_duality ] .
the two - colour loop models generated by eq .
( [ cc2 ] ) and eq .
( [ fc2 ] ) are given by the following sets of parameters in eq .
( [ part ] ) : @xmath85 , with all vertices having weight one , and @xmath44 , with all multi - colour vertices given weight zero and all other non - empty vertices given weight one - half respectively .
turning our attention to the particular case of two spatial dimensions , we remark that in the former model one can sum over all possible colourings to obtain the following result for the partition function : @xmath86 with @xmath87 being the total number of occupied ( colourless ) bonds and @xmath88 being the total number of _ faces _ in the clusters they form .
the number of faces is the minimum number of bonds that one must remove in order for the remaining clusters to be tree - like .
for example , the cluster in fig .
[ cross - cluster ] has five _ faces _ , while it can at most consist of four _ loops_. curiously enough , this result appears to have no simple generalisation for @xmath89 .
it appears that the two - dimensional loop model derived from the corner - cubic model can not be mapped directly onto a random surface model .
however , the other loop representation ( the one obtained via expansion of the face - cubic model ) does correspond to a random surface model .
indeed , consider the following `` recipe '' : take a loop configuration generated by a particular term in the expansion of eq .
( [ fc2 ] ) .
let red be the colour of loops originating from @xmath69 s , while blue corresponds to @xmath90 s .
take all plaquettes at the outermost region to be at height zero ( these plaquettes are said to form a substrate ) .
on this substrate we have clusters of loops .
the outermost boundaries of these clusters are themselves closed loops .
the plaquettes immediately adjacent to these boundaries are assigned the hight of @xmath91 if the loop forming a boundary is red , or @xmath92 if it is blue .
these plaquettes , along with any other plaquette accessible from them without crossing coloured bonds are said to form a plateau and thus have the same height .
inside such a plateau region there may be other loop clusters , that may or may not touch the boundary of a plateau ( only corners are allowed to touch , since no bond sharing between the loops is possible ) .
every such cluster is now treated in the same way : its boundary defines the `` secondary '' plateau with the height being that of the `` primary '' plateau @xmath93 depending on the colour of the boundary .
this procedure is repeated until all plaquettes are assigned their heights . as an example , consider the cluster in fig .
[ cross - cluster](a ) , which may now only consist of the bonds of a single colour .
if this were a red cluster , the heights would be @xmath91 for the plaquettes 1 - 4 and @xmath94 for the plaquette 5 .
in fact , this description is essentially identical to that given in ref .
@xcite in the context of wetting transition with the only difference that we allow for two - coloured clusters instead of single - coloured , and therefore the heights in our case may be both positive and negative .
firstly , due to an irreducible four - leg vertex factor , it can not be described by a nearest - neighbour hamiltonian ( i.e. a hamiltonian that depends only on the height difference of neighbouring plaquettes ) .
secondly , since no directions are assigned to the loops separating the plateaux , there is no way of deciding on the _ sign _ of their relative height difference without going through the necessary construction steps starting from the _ outside_. by contrast , the @xmath95-related random surface model can be constructed starting from _ any _ plaquette a particular choice simply determines the overall additive constant . in this sense
the mapping between the present random surface and the loop model is _
nonlocal_. ] the important feature of this random surface model is that it must have a phase transition whenever the underlying ashkin teller model undergoes a transition .
let us derive yet another graphical representation for the @xmath1 ( ashkin teller ) model considered in the previous section , this time it will be a _ random cluster _ representation closely resembling the fk representation for the potts model .
we start from eq .
( [ corner_part ] ) ( with @xmath1 ) , and with the help of the identity @xmath96 rewrite it as follows : @xmath97 \label{at_part}\ ] ] with @xmath98 .
the random cluster representation is generated by evaluating the product over all bonds in eq .
( [ at_part ] ) and then summing over the possible values of @xmath99 s and @xmath100 s .
if we think of the bonds originating from the @xmath99 variables as green ( g ) , and the bonds originating from the @xmath100 variables as orange ( o ) , then each of the resulting terms in the partition function can be graphically represented as a collection of green and orange clusters as well as empty sites .
the clusters of different colours may share sites , but not bonds . denoting the configurations of green and orange bonds as @xmath101 and @xmath102 respectively , we can then write the partition function as @xmath103 with @xmath104 being the total number of bonds ( of a specified colour ) , and @xmath105 being the number of corresponding connected components .
the rule for counting the connected components is as follows : every site that is not a part of a green cluster is considered to be a separate connected component for the purposes of @xmath106 , even if this site is a part of an orange cluster , and vise versa . in particular , the quantities @xmath107 and @xmath108 are to be interpreted exactly as in the usual random cluster models . the duality relations for such random cluster representation of the standard at model in two dimensions were established in refs .
@xcite and @xcite .
firstly , let us write the generic at hamiltonian as @xmath110 .
\label{at_hamlt}\ ] ] the graphical representation for the partition function is then obtained along the lines of the previous section .
the only difference is that this time double - coloured ( i.e. green and orange at the same time ) bonds are also allowed .
the graphical weight of a given bond configuration @xmath111 is @xmath112 where @xmath113 observe that eq .
( [ at_fk ] ) describes a particular case of this model provided that @xmath114 while @xmath115 .
[ sec : at_duality ] the dual model is obtained by placing orange bonds between the sites of a dual lattice every time when it does not cross a green bond of the original lattice .
correspondingly , the green bonds on the dual lattice are dual to the original orange bonds .
the duality relations are given by : @xmath116 and the model becomes self - dual when @xmath117 .
it is therefore suggestive that our model becomes self - dual at @xmath109 ( or @xmath118 ) . in order to show that it is indeed _ exactly _ self - dual
, we shall perform the above duality transformation to the orange bonds only ( @xmath119 ) , leaving the green bonds intact .
this results in having green bonds on both original and the dual lattices .
the green bonds on the dual lattice can be then split into those traversal to the original green bonds and those traversal to the previously vacant bonds , or symbolically @xmath120 . here
@xmath121 is the configuration of ( green ) bonds _ dual _ to the orange bonds while @xmath122 and @xmath123 are the configurations of bonds _ transversal _ to the original green and vacant bonds correspondingly .
the corresponding weight is now given by : @xmath124 we now observe that @xmath125 , and also that @xmath126 as @xmath127 ( the original random cluster model becomes fully packed according to eq .
( [ at_fk ] ) ) .
then the weight in this limit becomes simply @xmath128 the model described by such weights is manifestly self - dual since @xmath129 and @xmath130 as @xmath127 , and there is a symmetry between the green and the orange bonds .
it is tempting to speculate that a phase transition occurs exactly at the self - dual point .
although this is plausible , it is not the only possibility . in particular , there may be a phase transition at some @xmath131 .
however , we can say the following : if at @xmath132 there is no magnetisation ( i.e. percolation of green or orange bonds ) then the theorem proved by two of us @xcite applies ; @xmath132 is a critical point in the sense of infinite correlation length and infinite susceptibility .
the only other possibility is positive magnetisation at @xmath132 which implies a magnetic transition which could be continuous or first order at some @xmath133 .
( in particular , this is shown to happen , with a first order transition for the large-@xmath16 versions of these models @xcite ) .
although we find these alternative scenarios unlikely , we have , in any case , established the existence of a transition in this model for some value of @xmath21 between zero and one .
as mentioned above , in our opinion the most likely scenario is that a phase transition occurs precisely at @xmath132 .
the interesting question then is that of the universality class . without any supporting mathematical statements
, we suggest that _ at _ this point our model behaves similarly to the 4-state potts ferromagnet at its critical point . in order to substantiate this claim ,
let us first recall the random cluster representation for a @xmath16-state potts model : @xmath134 for @xmath135 this model is universally accepted to have a continuous transition at the self - dual point @xmath136 .
thus for the @xmath137 model at the self - dual point we have @xmath138 on the other hand , we can use eq .
( [ limiting_weight ] ) to rewrite the partition function of our model at @xmath132 as follows : @xmath139 the difference between the two models is in the last factor of @xmath140 in eq .
( [ stupid ] ) .
it is , however , reasonable to speculate that it can be neglected . indeed , on average @xmath141 , and
therefore one would expect the typical value of the difference @xmath142 to be sublinear in the system size .
by contrast , the individual terms @xmath143 and @xmath144 indeed scale linearly with the size of the system so this correction may be `` unimportant '' .
this , however , does not mean that the two models approach the self - dual point in a similar fashion .
in other words , we expect the exponents associated with the critical point itself ( such as @xmath145 and @xmath146 ) of the two models to be the same , while this needs not be true for the exponents associated with the _ approach _ to the critical point ( such as @xmath147 , @xmath148 and @xmath149 ) .
in fact , the @xmath132 point of our model may well be an edge of a critical kosterlitz - thouless phase in which case the approach exponents would take on extreme values ( zero or infinity ) .
this section concerns the reflection positivity property of the loop models defined by eq .
( [ part ] ) which in turn permits the analysis of their large @xmath4 limit .
let @xmath150 denote a @xmath0dimensional torus . here , and for the remainder of this paper , it will be assumed that the linear dimensions of @xmath150 are all the same , and are of the form @xmath151 .
we denote by @xmath152 the number of sites in the torus .
let @xmath153 denote the set of all possible loop configurations on @xmath150 . finally let @xmath154 denote a hyperplane perpendicular to one of the coordinate axes which cuts through the bonds parallel to this axis dividing the torus into two equal parts .
let @xmath155 and @xmath156 be the bond configurations on the two sides of the `` cut '' , with the bonds intersected by @xmath154 belonging to _ both _ sets .
thus @xmath157 , while @xmath158 contains only the intersected bonds .
we now define a map @xmath159 such that it simply reflects the configuration on the `` left '' to that on the `` right '' .
let @xmath160 be a function that depends only on the bond configuration on the right and define @xmath161 for any @xmath162 .
similarly , we can use @xmath163 to map @xmath164 ; with this convention @xmath165 is the identity .
a probability measure @xmath166 on the set @xmath37 is said to be _ reflection positive _ if for every such @xmath167 and any functions @xmath168 and @xmath169 as described above @xmath170 and @xmath171 .
[ theorem_4_1 ] the measures @xmath166 determined by the weights in eq .
( [ part ] ) are reflection positive on any even @xmath0-dimensional torus . proof .
let @xmath150 denote one such torus and @xmath167 denote one of the above described planes .
let @xmath172 be a configuration of bonds going through this plane .
assuming that @xmath173 , let us consider the measure @xmath174 .
our claim is that this splits into two measures , which we will call @xmath175 and @xmath176 , defined on @xmath155 and @xmath156 which are independent and identical under the reflection @xmath163 .
indeed , it is not hard to see that @xmath177 if and only if @xmath178 has an even number of bonds of each colour .
in each half of the torus , the endpoints of these bonds serve as `` source / sinks '' for bond configurations . in other words , a configuration in , say , @xmath179 must contain lines of the appropriate colour that pair up these sources .
but , aside from having to satisfy these `` boundary conditions '' , the weights are the same as given in eq .
( [ part ] ) .
these two measures defined accordingly on @xmath179 and @xmath180 are the above mentioned @xmath175 and @xmath176 respectively .
it is clear that if @xmath181 then @xmath182 furthermore if @xmath178 is a configuration and @xmath183 is any configuration that agrees with @xmath178 and has non - zero weight then for every @xmath184 we see that @xmath185 thence , for every @xmath168 that is determined on @xmath156 we have @xmath186 which can not be negative .
similarly we get that @xmath187 .
one of the important consequences of reflection positivity is a cauchy schwartz - type inequality : @xmath188 which in turn leads to the _ chessboard estimates _ to be described below .
( the reader interested in a more detailed description of reflection positivity is referred to the review @xcite and the references therein ) .
in this subsection we will consider some @xmath4colour models with @xmath189 and vertex factors that are uniformly bounded above and below independently of @xmath4 : @xmath190 .
examples include the @xmath15-type models and the corner - cubic models discussed in section [ sec : models ]
. however , the face - cubic model does not fall into this category since all the multi - coloured vertex factors vanish .
it is no coincidence that we can not treat these models since , as is obvious such models have a colour - symmetry broken phase for high enough value of bond fugacity ( for brevity we omit a formal proof ) .
the suppression of long contours will be established by showing that long lines of any particular colour are exponentially rare in the length of the line . to prove this
we will need the so called chessboard estimate which in the present context reads as follows : for @xmath191 let @xmath192 denote indicator functions for bond events that are determined by the bonds emanating from the site @xmath193 .
( the @xmath194 need not all be distinct . ) for any of these @xmath194 , cover the lattice with ( multiple ) reflections of the corresponding event and let @xmath195 denote the partition function constrained so that at each site , the appropriately reflected event is satisfied
. then @xmath196 proof .
see section 2.4 of ref .
@xcite . our principal result of this subsection : [ theorem_4_3 ] consider an @xmath4-colour loop model as described by eq .
( [ part ] ) on the torus @xmath150 ( which is taken to be `` sufficiently large '' ) and suppose that the vertex factors are bounded below by @xmath197 and above by @xmath198 uniformly in @xmath4 . for sites @xmath193 and @xmath199 in @xmath150 , let @xmath200 denote the probability that these sites belong to the same loop . then , provided @xmath4 is sufficiently large ,
there is a @xmath201 such that for all values of @xmath14 , @xmath202 where @xmath203 denotes the minimum length of a walk between @xmath193 and @xmath199 and @xmath204 is a constant . for conceptual clarity
, we will start with a proof of the case @xmath205 ; all of the essential ideas are contained in this case .
the problems in @xmath206 involve some minor technicalities and the general proof can be omitted on a preliminary reading .
proof of theorem [ theorem_4_3 ] ( @xmath205 ) .
let us focus on a particular colour red and show the statement is true for red loops ; this only amounts to a factor of @xmath4 in the prefactor .
we define the `` red event '' as an event where at least two red bonds are connected to the site in question .
it is clear that there are two main types of red events : those where the red bonds attached to a given site line up along the straight line and those where they form the right angle .
these two types are shown in figures [ reflection](i - a ) and [ reflection](ii - a ) respectively .
we will denote by @xmath207 , @xmath208 the events of the first type and @xmath209 , @xmath210 those of the second type .
obviously there are only two distinct constrained partition functions which we respectively denote by @xmath211 and @xmath212 .
figure [ reflection ] shows these single events ( a ) , the results of the first two reflections with respect to the dashed lines ( b ) , ( c ) and finally the configurations obtained by applying each reflection @xmath213 times in order to completely tile the surface of the torus .
the grey lines correspond to yet unidentified bonds
these are the degrees of freedom left after the process of tiling has been completed .
let us perform the estimate on @xmath214 first .
we claim that if @xmath215 is any legitimate configuration that contributes to @xmath212 each of the red squares of which there are @xmath216 can be independently replaced by vacant bonds or a square loop of any other colour or left as red .
of course this may cost us an exchange of the `` best '' for the `` worst '' vertex factor but even so , the result is @xmath217 let us now turn attention to the @xmath218 estimate .
we start with a factor of @xmath219 for the red bonds already in place as well as another worst case scenario of @xmath220 . as for the lines that are orthogonal to ( horizontal in fig .
[ reflection](i - d ) ) once started in any colour they must continue until they wrap the torus , a total length of @xmath221 .
there are @xmath4 possible choices of colour for each line as well as the possibility of no bonds at all .
since there are a total of @xmath222 lines altogether , this gives @xmath223 to obtain our estimate on @xmath224 , we simply pick the even ( or odd ) sublattice of dual sites and surround each site with one of @xmath4 coloured elementary loops or with a `` loop '' of vacant bonds .
folding in the worst case scenario for the vertex factors this gives @xmath225 the ratio may be expressed as a product of two terms namely @xmath226^n$ ] and @xmath227 times an additional @xmath228 . clearly the first term is bounded by @xmath229 .
as for the second ratio , if @xmath230 we may neglect @xmath231 for @xmath232 and the ratio is bounded by one . on the other hand ,
if @xmath233 , we may neglect the 1 and we get , modulo a factor of @xmath234 , another @xmath229 ; we will settle for the bound of 1 .
thus we have @xmath235 with the same upper bound for @xmath236 valid for all @xmath237 .
we thus denote the mutual upper bound by @xmath238 .
the desired result now follows from a standard peierls argument : if @xmath193 and @xmath199 are part of the same loop , some subset of this loop must be a self - avoiding walk of length at least @xmath239 .
we enumerate all such walks and use the chess board estimate on each particular walk
. then if @xmath240 where @xmath241 is the two dimensional connectivity constant we write @xmath242 ( the factor of 2 because we must go there and back ) and the stated result follows . proof of theorem [ theorem_4_3 ] ( @xmath243 ) .
the preliminary steps are the same as the two - dimensional case : there are again only two types of @xmath111 s ( but with more indices ) and two constrained partition functions which we again denote by @xmath218 and @xmath212 .
the pattern for @xmath212 is the two - dimensional pattern in fig .
[ reflection](ii - d ) reflected in all directions orthogonal to the plane visualised . noting that each reflection doubles the number of red squares
, we see that in the @xmath212 patterns there are a total of @xmath244 squares altogether . repeating the argument leading to eq .
( [ eq : z_ii ] ) we end up with exactly the same bound .
what is a little harder is the estimates on @xmath218 and the partition function itself .
we first claim that @xmath224 can be estimated by @xmath245 to achieve this , we assert that the following holds : there is a set of plaquettes of the lattice with the property that each bond of the lattice belongs to exactly one plaquette .
once this claim is established it is clear that eq .
( [ eq : z_bound ] ) holds ; indeed there are just @xmath246 \times nd$ ] plaquettes in question , we consider those configurations in which each of them is independently left vacant or traversed with an elementary loop in one of the @xmath4 possible colours .
let us turn to a proof of the above assertion .
let @xmath247 denote the elementary unit vectors .
we adapt the following notation for plaquettes : if , starting at @xmath248 we first move in the @xmath249 direction then in the @xmath250 direction and then complete the circuit we will denote this plaquette by @xmath251_\mathbf{x}$ ] in general we can have @xmath252_\mathbf{x}$ ] and it is noted that @xmath253_\mathbf{x}$ ] = @xmath254_\mathbf{x}$ ] starting with the origin , consider the following list of instructions for plaquettes : @xmath255_{0 } , \ [ \mathbf{\hat e}_2 \diamond -\mathbf{\hat e}_3]_{0 } , \dots , [ \mathbf{\hat e}_d \diamond -\mathbf{\hat e}_1]_{0}. \label{eq : plaquette_instructions}\ ] ] so far so good each bond emanating from the origin belongs to exactly one plaquette . if @xmath256 let us define @xmath257 to be the parity of the @xmath258 coordinate . then at the site @xmath259 , select the following : @xmath260_\mathbf{x } , \dots [ \pi_d(\mathbf{x})\mathbf{\hat e}_d \diamond -\pi_1(\mathbf{x})\mathbf{\hat e}_1]_\mathbf{x}$ ] . taking the union of all these lists
, it is clear that indeed each bond belongs to _ at least _
one plaquette , the only question is whether there has been an over - counting . to settle this issue
we establish the following assertion : let @xmath259 denote a site
. then a plaquette is specified by the instructions at @xmath259 if and only if the same plaquette is specified by the neighbours of @xmath259 on that plaquette .
indeed , consider one such plaquette namely @xmath261_\mathbf{x}$ ] .
( if necessary we use the convention @xmath262 ) .
one of the relevant neighbours is @xmath263 .
but at @xmath264 , we have instructions for the plaquette @xmath265_\mathbf{x^\prime}$ ] .
since @xmath266 and @xmath267 this is seen to be the same plaquette .
the other neighbour , at @xmath268 , follows from the same argument . reversing the rles of @xmath259 and @xmath264 ( as well as @xmath259 and @xmath269 )
establishes the ` only if ' part of the assertion .
it is thus evident that no bond can belong to two plaquettes : if @xmath259 and @xmath264 are neighbours and there are instructions coming from @xmath259 to make some particular plaquette that includes the bond @xmath270 then : \(a ) these are the only instructions coming from @xmath259 pertaining to this bond .
\(b ) the list at @xmath264 ( and the other corners of this plaquette ) also include this plaquette .
\(c ) if @xmath271 is another neighbour of @xmath259 which is _ not _ in this plaquette there can not be instructions at @xmath271 to include the plaquette containing @xmath259 , @xmath264 and @xmath271 .
items ( a ) and ( b ) are immediate . to see item ( c ) ,
note that by the assertion , such instructions would also have to appear on the list at @xmath193 .
but , by item ( a ) , they do nt .
equation ( [ eq : z_bound ] ) is now established .
finally , let us estimate @xmath211 from above .
first it is noted that the @xmath211 patterns consist of parallel lines ( pointing in the direction of the original red bonds ) that pass through every site .
hence the only loops we can draw are confined to the various @xmath272dimensional hyperplanes orthogonal to these lines .
modulo vertex factors which we estimate by @xmath273
each hyperplane yields the @xmath272dimensional partition function .
we thus have @xmath274 where @xmath275 is the @xmath272dimensional partition function with all vertex factors equal to unity .
( this observation is not of crucial importance
the key ingredient is the number of available bonds but it serves to compartmentalise the argument . ) let us tend to the estimate of @xmath275 .
we denote by @xmath276 the contribution to @xmath275 that comes about when there are exactly @xmath277 bonds in the configuration . given the placement of the bonds , let us count ( estimate ) the number of ways that they can be organised into self - returning walks and then independently colour each walk .
throwing in a combinatoric factor for the placement of the @xmath277 bonds and the fact that there can not conceivably be more than @xmath278 loops to colour we arrive at @xmath279 where @xmath280 is defined as follows : for @xmath277 bonds , consider their placement on the lattice chosen so as to maximise the number of ways in which the resulting ( colourless ) graph can be decomposed into distinct loops .
( such decompositions are done by `` resolving '' all intersecting vertices into connected pairs of bonds , similar to fig . [ o(n)_vertices ] . )
then the quantity @xmath280 denotes this maximum possible number of decompositions .
let @xmath281 ( which equals @xmath282 ) denote the greatest possible number of local options for an on - going self - avoiding walk .
we claim that @xmath283 indeed , take this optimal placement of @xmath277 bonds and , starting at some predetermined bond ( and moving in some predetermined direction ) draw a loop of length @xmath284 .
there are no more than @xmath285 possibilities for this loop . then what is left over can not yield a total better than @xmath286
this must be done for every possible value of @xmath284 and summed .
defining @xmath287 , we arrive at the recursive inequality @xmath288 and the bound in eq .
( [ eq : y_bound ] ) can be established inductively .
putting together equations ( [ eq : xi_bound ] ) and ( [ eq : y_bound ] ) we find @xmath289 and thus @xmath290^l = [ cr(1 + 2{\mathfrak}zrn^{\frac 14})^{(d-1)}]^n .
\label{eq : z_i_bound}\ ] ] an estimate for the straight segments is nearly complete .
we write @xmath291^{\frac 1n } \leq \frac cc\frac { r(1 + 2{\mathfrak}zrn^{\frac 14})^{(d-1 ) } } { ( 1 + nr^4)^{d/4}}\ , .
\label{eq : z_i / z}\ ] ] let us split the power of the denominator : @xmath292 ; the first term will handle the @xmath14 in the numerator and this ratio is less than @xmath293 .
as for what is left over , it is easy to see that @xmath294 thus we have @xmath291^{\frac 1n } \leq \frac cc ( 1 + 2{\mathfrak}z)^{d-1 } \frac 1{n^{1/4 } } \label{eq : z_i / z_bound}\ ] ] so again all of the @xmath111 s have estimates with the scaling of @xmath293 .
the remainder of the argument follows the same course as the two - dimensional argument with modifications where appropriate .
so far we have shown that for a sufficiently large @xmath4 , there are no phase transitions associated with the divergent loop correlation length . however
, this does not rule out the possibility of phase transitions of a totally different nature in fact we shall see that such transitions do indeed take place .
the type of these transitions happens to be sensitive to the lattice structure and the vertex factors that correspond to loop intersections .
this leads to a variety of critical phenomena which has not been observed in the case of a honeycomb lattice @xcite .
we shall consider the case of all vertex factors being positive and uniformly bounded .
let the number of colours @xmath4 be very large and consider the high fugacity limit of @xmath58 .
the model becomes fully - packed in this limit , i.e. every bond is occupied .
but since @xmath4 is large , according to the estimates of the previous subsection the partition function is dominated by configurations in which there are as many loops as possible . in two dimensions ,
it is clear that these `` maximum entropy states '' break the translational symmetry : the loops of different colours may run around either odd or even plaquettes of the lattice .
( we remark that in @xmath243 , the situation is considerably more complicated and is currently under study . ) on the other hand , if the bond fugacity is low , the system must be translationally invariant .
therefore one would anticipate that at some large but finite value of @xmath14 a translational symmetry - breaking transition takes place . due to a doubly positionally degenerate nature of the resulting high - density state
, we find it natural to expect that such transition is of the ising type . indeed
, it appears to be very similar to the transition in the ising antiferromagnet .
let us now turn to the rigorous arguments supporting this qualitative picture .
[ thm : symmetry - breaking ] consider an @xmath4-colour loop model on @xmath295 with vertex factors bounded below by @xmath296 and above by @xmath198 uniformly in @xmath4 .
then , if @xmath4 is sufficiently large there is a phase transition characterised by the breaking of translational symmetry . in particular
, there exist @xmath297 ( sufficiently small ) and @xmath298 ( sufficiently large ) such that for @xmath299 all states that emerge as limits of torus states are translation invariant . on the other hand ,
if @xmath300 , there are ( at least ) two states with broken translational symmetry . proof
let us start with the formal proof of translation - invariance at low fugacity .
any given site belongs to zero , two or four bonds .
let @xmath301 , @xmath302 and @xmath303 denote the corresponding events and @xmath304 , @xmath305 and @xmath306 the constrained partition functions in which every site is of the stated type
. we will estimate @xmath307 and @xmath308 ; obviously @xmath309 . in calculating @xmath306
we notice that in all configurations , each bond must be coloured and there are at most @xmath310 separate loops .
this gives @xmath311^n y(2n)\ ] ] where @xmath280 is the same quantity that appears in eq .
( [ eq : xi_bound ] ) .
similarly , in @xmath305 half the bonds are used ( resulting in a factor of @xmath219 ) and there are no more than @xmath216 loops .
we get @xmath312^n y(n)\ ] ] essentially the square root of the above . in order to show that there is translation invariance it is sufficient to establish the following :
let @xmath313 and @xmath314 denote local events , let @xmath315 be the translation operator to the point @xmath191 and let @xmath316 denote any unit vector .
then we must show that for all @xmath317 with @xmath318 large the probabilities of @xmath319 and @xmath320 are essentially the same .
we note that either the supports of @xmath313 and @xmath321 are attached by a @xmath322-connected path of sites of type @xmath305 and @xmath306 , or they are separated by a connected circuit of cites of type @xmath304 .
( two sites are considered @xmath322-connected if they share the same plaquette . ) however , in case the support of a is surrounded by a connected circuit of empty sites , it is clear that for any @xmath316 , @xmath313 and @xmath323 have the same probability .
thus we arrive at @xmath324 where @xmath325 is the event that there is a @xmath322-connected path of non-@xmath301 sites between the support of a and that of @xmath321 .
now , provided @xmath326 is small , @xmath327 with @xmath328 for some constant @xmath147 .
thence translation invariance ( among the states that emerge from the torus ) is established .
let us now turn attention to the opposite limit , namely @xmath329 .
we claim that with high probability the plaquettes on the lattice are of one of the two types : they are either surrounded by four bonds of the same colour , or by bonds of four different colours .
we will again define site events representative of the two purported ground states .
if a site is on the even sublattice , we say it is of the a - type if it has four bonds of two different colours , one colour occupying the positive @xmath330 and @xmath331 directions and the other , the negative @xmath330 and @xmath331 directions .
the odd site of the a - type is defined as the mirror reflection of an even a - type site .
a site is said to be of b - type if the rles of the even and the odd sublattices are reversed see fig .
[ a - b_sites ] .
any site which is not of one of these two types will be called `` bad '' .
there are three choices for such bad sites , namely housing no bonds , only two bonds and a `` + '' configuration in which the @xmath330 and @xmath332 directions have the same colour and similarly for the @xmath333 directions .
the first two are the old @xmath301 and @xmath302 from the preceding portion of this proof , while the constrained partition function for the latter event , denoted ( for the lack of better notations ) by @xmath334 can be bounded by @xmath335 on the other hand , we can always estimate @xmath224 from below by @xmath336 ^ 2n$ ] as discussed previously .
thus we have ( as @xmath337 ) [ eq : mu_estimates ] @xmath338 thus almost all sites are either a - type or b - type . however , there is no constraint that a and b sites can not be neighbours .
but , as we show below , this possibility is also suppressed for @xmath3 .
firstly , let us notice that the reflections through bonds map the even into the odd sites ( and vice versa ) . by definition ,
the a - even and a - odd events are mirror images , and similarly for b. thus , under multiple reflections , a - sites map to a - sites and b s to b s . hence ,
if ( a - b ) is the event that the origin is of the a - type and its right nearest neighbour is of the b - type , we have @xmath339 where @xmath340 is the partition function constrained according to the pattern shown in fig .
[ a - b_pattern ] .
are the even sites of the a - type etc .
the zigzag lines that form loops wrapping the torus and separating the columns of a- and b - sites are shown black.,width=264 ] now , the interior of the a or b columns still consists of the four - bond loops just as a pure a or pure b - type tiling would .
however , as is easily seen in fig .
[ a - b_pattern ] , a boundary between the columns is formed by a zigzag line of a single colour running around the torus . thus we arrive at @xmath341 and hence , in the large @xmath237 limit , @xmath342 .
the rest of the proof follows easily : a connected cluster of a - type sites has , on its boundary a `` bad '' site or an a - b pair ; similarly for the clusters of b - sites .
thus , if @xmath343 and @xmath4 are large enough , the possibility that any given site is isolated from the origin is small .
thence , given that the origin is of the a - type which has probability very near one half the population of b - type sites is small and vice versa .
evidently , there are two states , one with an abundance of a - sites and the other with an abundance of b s .
it is not hard to see that in these two states , translation symmetry is broken both are dominated by the appropriate staggered patterns .
so far we have succeeded in proving the following general statements : * a generic multi - coloured loop model with _ all _ vertex factors @xmath344 uniformly bounded from above and below does not have a phase transition corresponding to the divergence of the loop size in _ any _ dimension _ provided _ that the number of colours @xmath4 is sufficiently large . * in two dimensions these models undergo a different phase transition , presumably of the ising type , that is associated with breaking the _ translational _ symmetry . while the `` ground states '' ( the states at @xmath345 ) of the system are _ not _ translationally invariant , their degeneracy appears to grow very fast with @xmath0 , and it is not obvious that the resulting `` entropy '' will not destroy the transition at any finite value of @xmath14 . ] .
while the examples of the models that have been considered in this paper all have their vertex factors independent of the particular arrangements of different colours entering the vertex , all these results remain valid even if this is not the case as long as all different vertex factors are still reflection - symmetric and bounded from above and below by some positive numbers . a clear example of a model that does not follow this rule is the loop model derived from the face - cubic spin model in section [ sec : face - cubic ] .
here the bonds of different colours are simply not allowed to share a vertex , making the corresponding vertex factor vanish .
it is clear that as the bond fugacity @xmath14 increases , the system will have a phase transition ( possibly of the first order ) associated with breaking the _ colour _ symmetry .
indeed , the only way to `` pack '' more bonds into the system is to force all of them to be of the same colour .
this transition is very similar to the widom - rowlinson transition .
another interesting observation can be made about the two - dimensional loop model with all two- and four - leg vertices having the same weight ( this is the model originating from the corner - cubic spin model see section [ sec : corner - cubic ] ) . comparing the states of the system at @xmath345 when all bonds are occupied ( the fully packed limit ) , and at @xmath346 when no additional weight is associated with placing extra loops
, we conclude that the @xmath347-coloured model at @xmath345 is identical to the @xmath4-coloured model at @xmath346 .
indeed , one should simply consider the vacant bonds at @xmath346 as being coloured grey to turn the @xmath4-coloured loop system into a fully - packed @xmath347-coloured system . now
start with the large enough @xmath4 at @xmath345 .
the result of section [ sec : transl_sym ] guarantees the existence of a broken translational symmetry in this case . by the argument presented here
this means the existence of a broken symmetry in @xmath348-coloured , @xmath346 case .
now let us continuously increase the value of the bond fugacity in this @xmath348-coloured model .
as @xmath14 grows , there are only two possible scenarios : the broken translational symmetry is either lost via a phase transition or is retained all the way to @xmath345 .
the former case corresponds to the _ intermediate _ symmetry - broken phase surrounded by the phase transitions at @xmath349 and @xmath350 .
if the latter scenario is realised , then go to @xmath345 and repeat the process of mapping it onto a lower @xmath4 model .
notice that this latter scenario can not continue all the way to @xmath351 because the @xmath351 , @xmath346 case is nothing but a loop representation for the ising magnet at @xmath352 which does not have broken translational symmetry .
`` high temperature '' expansion are also the domain walls of the ising model on the dual lattice at @xmath353 .
these dual spins are assigned their values independently with probability @xmath354 , and the contours separating regions of opposite type are manifestly translation invariant . ]
therefore one _ must _ find an intermediate phase at some not very large value of @xmath4 ( although we can not rule out a possibility of it being just the point @xmath346 ) . from the above
we also learn that the @xmath1 ( ashkin teller - like ) model is _ not _ critical in its fully - packed limit ( since it is just the @xmath352 ising model ) . on the other hand the @xmath95 loop model discussed in section [ sec : on ] _ is _ critical in this limit ( it maps onto the square ice model ) .
but the only difference between the two is the factor of 3 in the same - colour four - leg vertex factor . on the basis of this observation ( along with the first two remarks of this section )
we reiterate that the vertex factors and lattice details are often important for determining the phase diagram of a particular model .
in this appendix we present the calculation of the generic vertex factor appearing in the loop expansion of the @xmath15-type model featured in [ sec : on ] .
as it turns out , this calculation is independent of the lattice details ( such as the dimensionality or coordination number ) which is reflected in our notation .
thus let @xmath355 denote the @xmath4-dimensional solid angle and let @xmath356 denote the components of a @xmath4-dimensional unit vector . since we only consider the spin at one particular site , the actual site index is conveniently dropped for now .
the vertex factor we need to calculate is @xmath357 where @xmath358^{-1 } } $ ] is the total solid angle and @xmath359 is the haar measure .
the object @xmath360 is , of course , the `` vertex factor '' originating from @xmath361 terms from the first component ,
@xmath362 terms from the @xmath4-th component associated with these numbers of these types of bonds entering into a vertex .
[ proposition_a1 ] let @xmath363 which is equal to @xmath364 denote the number of different ways of dividing @xmath365 objects into @xmath366 distinct pairs .
then @xmath367 where @xmath368 and @xmath369 the calculation of the value for @xmath360 has appeared before , _ cf . _
@xcite and references therein .
however , the interpretation that emerges from this formula has , to our knowledge , not been discussed previously ( _ cf . _ the remark following the proof ) .
the proof boils down to the calculation of the above integral .
we write @xmath370 where @xmath371 .
also , @xmath372 where @xmath373 is an @xmath374-dimensional unit vector . finally , we note the fact that @xmath375 as a consequence of the expression ( [ eq : proposition_a1 ] ) we may decompose the coloured bond configurations to obtain a colourless loop model ( which generalises the result of eq .
( [ o(n)_loops ] ) obtained for the square lattice ) .
in particular , at each vertex we have a factor which depends only on the total number of bonds entering the vertex ( and the dimensionality of the spin @xmath4 ) times the number of ways that the bonds of the various colours can be paired . by choosing a particular pairing scheme at each vertex , the bond configuration breaks , unambiguously , into a collection of
self - returning walks ( loops ) .
configurations with the same set of loops that is to say the same set of bonds and the same pairing schemes ( walking instructions ) at each vertex but which differ in the colours of the loops are seen to have identical weights .
thus we consider configurations @xmath53 of loops ( bonds + walking instructions ) and the partition function becomes the analog of the expression [ o(n)_loops ] : @xmath380 with @xmath381 being the factor for a vertex with @xmath382 bonds given by @xmath383 we observe that the right hand side of eq .
( [ o(n)_loops_gen ] ) provides a well defined model for all @xmath4 and can be continued essentially with no ambiguity to non - integer values . as usual , we may _ define _ correlation functions via loop probabilities or be content with the numerators containing strings divided by denominators without . in the previous discussions of these issues ( see ref .
@xcite and references therein ) the goal has been to derive the self - avoiding walk as the @xmath384 limit of the @xmath15 models . thus the vertex weights themselves get continued without reference to intermediate models .
( and indeed , the self - avoiding walk does emerge as @xmath385 . ) although we have only treated the case where the original spin models were defined by @xmath386 as opposed to @xmath387 we claim that the latter case also leads to graphical models that are well defined for non - integer @xmath4 .
( the derivation is somewhat intricate and will eventually appear in a future publication .
) however , we remark that in either case , the @xmath384 limit is , in this context , quite simple to understand .
the authors would like to thank j. kondev , r. kotecky and t. spencer for the interesting discussions .
lc and ks were supported in part by nsa grant no .
mda904 - 98 - 1 - 0518 and nfs grant no .
99 - 71016 ; lpp was supported in part by doe grant no .
de - fg02 - 90er40542 . | we consider a general class of ( intersecting ) loop models in @xmath0 dimensions , including those related to high - temperature expansions of well - known spin models .
we find that the loop models exhibit some interesting features often in the `` unphysical '' region of parameter space where all connection with the original spin hamiltonian is apparently lost .
for a particular @xmath1 , @xmath2 model , we establish the existence of a phase transition , possibly associated with divergent loops .
however , for @xmath3 and arbitrary @xmath0 there is no phase transition marked by the appearance of large loops .
furthermore , at least for @xmath2 ( and @xmath4 large ) we find a phase transition characterised by broken translational symmetry .
loop models , reflection positivity , phase transitions . |
the magneto - rotational instability ( mri ) is a candidate for driving turbulence and enabling the accretion of matter onto the central object @xcite .
dependent on the ionization degree , the mri generated turbulence will be reduced , down to a low turbulence regime , called the dead zone @xcite .
various studies showed that a certain level of resistivity suppresses mri activity @xcite .
one of the most important dimensionless number , characterizing the coupling between gas and magnetic fields , is the magnetic reynolds number @xmath7 which relates the timescale of magnetic advection to magnetic diffusion .
we consider here the ohmic diffusion term which is most important at the midplane . up to now
, there is no clear prescription for which values of @xmath7 the mri turbulence is sustained in disks .
@xcite found sustained mri with a zero - net flux magnetic field for @xmath8 . a recent study by @xcite , using stratified local box simulations , investigate the turbulence level for different values of the magnetic reynolds and prandtl number .
they found a so - called low state , a state where turbulence is partly suppressed , but sustained due to a dynamo process .
in addition , they predict a critical reynolds numbers @xmath9 in the range between @xmath10 .
a similar region was investigated by @xcite in which they found @xmath11 .
this critical reynolds number is important to model the surface density of active layers in proto - planetary disk as it was recently done by @xcite .
+ in our study , we will search for the critical magnetic reynolds number in global zero - net flux stratified mri simulations . here
the mri turbulence criterion , elsasser number @xmath12 , should become unity .
in contrast to the magnetic reynolds number , the elsasser number gives one clear threshold independent of the magnetic geometry or the stratification .
we will also investigate the hydrodynamical motions which become important in the dead - zone region @xcite .
we concentrate on the magnetic reynolds number regime proposed by @xcite and @xcite . for our simulations we use only explicit resistivity .
@xcite found out that well ionized mri turbulence scales independently of @xmath13 if @xmath14 .
in addition , as the molecular viscosity is very small in proto - planetary disks we expect prandtl numbers of @xmath15 and we focus on this low prandtl number regime . ] in this paper we will first describe the numerical and physical setup .
then we will present the results , discussion and the conclusion .
the initial conditions for density , pressure and azimuthal velocity follow hydrostatic equilibrium .
we set the density @xmath16 to @xmath17 with @xmath18 , the scale height to radius @xmath19 , @xmath20 .
the pressure follows locally an isothermal equation of state : @xmath21 with the sound speed @xmath22 .
the azimuthal velocity is set to @xmath23 the initial velocities @xmath24 and @xmath25 are set to a white noise perturbation amplitude of @xmath26 .
we start the simulation with a pure toroidal magnetic seed field with constant plasma beta @xmath27 . to obtain a range of magnetic reynolds number
we keep the magnetic dissipation value constant in the disk .
we use three different values of @xmath28 , @xmath29 , @xmath30 and @xmath31 .
@xmath32 to estimate the numerical magnetic reynolds number having mri turbulence simulations we run also ideal mhd simulations with different resolutions as reference .
the radial domain extends from 1 to 10 au .
the @xmath33 domain covers @xmath34 4.3 disk scale heights , or @xmath35 . for the azimuthal domain we use @xmath36 for the l models and @xmath37 for the h models .
we use a uniform grid in spherical coordinates .
models l have a resolution of @xmath38 , @xmath39 , @xmath40 and @xmath41 for the h models .
+ all models resolve the radial scale height with 9 to 22 grid cells for the inner and outer radius .
the vertical scale height is resolved by 22 grid points . in models
l the azimuthal scale height is resolved by 9 grid cells .
the h models have a higher resolution of 17 per scale height in the azimuth .
their are calculated with the fargo mhd to reduce even more the numerical dissipation @xcite .
the simulation models are summarized in table 1 .
we note that model @xmath42 is special . here
the numerical dissipation can not be neglected . by comparing with the results of h models , the @xmath42 model shows magnetic reynolds number below the value used in @xmath43 .
this model establishes a large dead - zone region . here hydro - dynamical motions become important .
.from left to right : model name , resolution , domain size , fargo - mhd , range of magnetic reynolds number . in model @xmath44 the used explicit resistivity
is not resolved and we estimated the magnetic reynolds number . [ cols="<,<,<,<,<,<",options="header " , ] [ mri - t ] buffer zones extent from 1 to 2 au as well as from 9 to 10 au . in the buffer zones we use a linearly increasing resistivity ( up to @xmath45 ) reaching the boundary .
this damps the magnetic field fluctuations and suppresses boundary interactions . for our analysis
we use the range between 3 and 8 au , which is not affected by the buffer zones .
our outflow boundary condition projects the radial gradients in density , pressure and azimuthal velocity into the radial boundary and the vertical gradients in density and pressure at the @xmath33 boundary .
for all runs we employ the second order scheme in the pluto code with the hlld riemann solver @xcite , piece - wise linear reconstruction and @xmath46 order runge kutta time integration .
we treat the induction equation with the `` constrained transport '' ( ct ) method in combination with the upwind ct method described in @xcite , using explicit resistivity .
a more detailed description of the physical setup can be found in @xcite .
according to eq .
3 , we obtain a specific value of magnetic reynolds number at each radius , resulting in a specific value of turbulence . to compare the results at different specific radii
the timescale of radial mixing is important . due to the low magnetic dissipation
, the timescale of magnetic diffusion is very long compared to the turbulent mixing timescale . the maximum radial mixing scale for an alpha value of 0.01 after 1000 years at the outer radius at 8 au is around @xmath47 . for our analysis
we use specific radial positions which are separated enough . in chapter 3.5
we will investigate how the radial transport of magnetic field effects the local evolution .
the @xmath48 value at a specific radial position is calculated with @xmath49 the integral is done for each radius separately @xmath50 with @xmath51 and @xmath52 is the left and right radial cell boundary .
the @xmath53 is mass weighted to recover the correct total value . in chapter 3.1 we compare results having the same local time period . here
we average @xmath54 , the turbulent velocities @xmath55 and the dynamo @xmath56 between 25 and 60 local orbits .
this ensures that we compare the same dynamical evolution at a specific radius and reynolds number .
the space average of @xmath57 and @xmath56 is done at the midplane region between @xmath58 scale heights . in chapter 3.2
we concentrate on the longterm evolution for different values of @xmath7 .
there we include the two mri criteria , the elsasser number @xmath59 as well as the factor q @xmath60 the q factor is the ratio of the mri fastest - growing azimuthal wavelength to the azimuthal cell size
. the value should be larger than 8 @xcite . in chapter 3.3
we concentrate on the run @xmath42 having the largest dead - zone . here
we present results of the dominant hydro - dynamical motions . due to the change of the rotation period @xmath61 with radius , the comparison of the turbulence statistics at different radii is limited to the number of rotations at the outer radius .
we therefore present the comparison for the different reynolds numbers between a time average of @xmath62 local orbits .
we note that in this period the initial net magnetic flux has already vanished and we have a zero - net flux mri turbulence .
we note also that the results from model @xmath42 should be excluded .
they indicate that here the explicit dissipation is not resolved and the numerical dissipation dominates
. we still overplot the results obtained by this lower resolution ( l models , fig .
1 blue color ) to estimate the total magnetic reynolds number . this model will be still useful as it shows very similar conditions as present in dead - zones .
1 combines the results of @xmath48 , the turbulent rms velocity @xmath57 and the dynamo @xmath56 value as a function of the magnetic reynolds number .
the results show for all turbulent quantities a saturation for magnetic reynolds numbers around 5000 and above . below @xmath63 the mri
is damped and the turbulence decreases . we observe a slightly higher turbulence level for high magnetic reynolds numbers compared to fully ideal mhd run .
+ there is a saturation of @xmath48 at around @xmath64 value for magnetic reynolds numbers greater than 5000 , see fig
below @xmath65 , @xmath66 drops down to @xmath67 at around @xmath68 .
even one could see a flattening at around @xmath69 the values are still dominated by maxwell stress which indicates that the mri is still operating .
the turbulent velocity scales roughly with the square root of @xmath70 , found in recent local box simulations of dead - zones @xcite .
we observe a saturation around @xmath71 ( fig .
1 , middle ) . below @xmath65
, @xmath72 drops down to @xmath73 at around @xmath74 .
we note again that these values represent the turbulent motions at the midplane , in the corona the turbulent velocity increases also in the dead - zone models .
the dynamo action could be expressed by showing the correlations @xmath75 with @xmath76 .
we follow the analysis described in chapter 3.7 in @xcite and @xmath77 is normalized over @xmath78 .
the dynamo action in the northern hemisphere is shown in fig .
1 , bottom .
@xmath77 saturates at around 0.004 and decreases only down to 0.002 at @xmath79 . in all
runs the sign of @xmath77 in the southern hemisphere is negative with similar amplitude as the corresponding value in the northern hemisphere .
+ the results of the ideal mhd runs show convergence . here
the fargo mhd method plays an important role in decreasing numerical dissipation as it is presented in mignone et al .
2012 , a&a accepted .
@xmath48 converge around 0.01 .
we note again that the results presented in fig .
1 reflect mainly the turbulence level at the midplane .
in addition we see here no decay of the mri turbulence due to the short time period . in the next chapter
we will concentrate on the longterm evolution over height for specific reynolds number .
the previous chapter has shown the transition to saturated turbulence in well ionized disk regions . in this chapter
we focus on the longterm evolution , including the turbulence over height . for each model
we perform the analysis at a specific reynolds number which are located in the middle of the domain . here , @xmath80 has @xmath81 , @xmath82 for model @xmath83 and @xmath84 for model @xmath43 . in model
@xmath42 we expect to have a reynolds number around @xmath85 by comparison with model @xmath43 .
+ we plot the time evolution of @xmath48 in fig .
the analysis show that both ideal runs as well as the resistive runs down to @xmath82 ( @xmath80 and @xmath86 ) , show a steady turbulent evolution with @xmath2 .
the models @xmath43 and @xmath42 with @xmath87 show a steady decrease of the turbulence until they oscillate in the range between @xmath88 . due to the decrease we note that the results of the models @xmath43 and @xmath42 plotted in fig . 1 depend on the averaged time period . for these models ,
an extended simulation runtime was needed .
we mark the regions with dominating reynolds stress in thick ( blue and green solid lines ) .
the plots show only a short - term domination which indicate that the total integral over height suggest still a operating mhd turbulence .
the configuration of the maxwell stress @xmath89 over height is presented in fig .
3 . after around 200 inner orbits ( years ) the initial net azimuthal flux is lost and the flux start to oscillate around zero @xcite . for a magnetic reynolds number above @xmath81
the ionization is well enough to sustain the mri turbulence at each height . in the midplane region
we found a saturation of @xmath2 at around @xmath90 . in model @xmath80 ,
3 , top , the elsasser number is between 1 and 10 in the midplane region .
if the ionization is decreased below @xmath91 the elsasser number reaches unity at the midplane ( see also fig .
4 ) . in model @xmath86 ,
3 , second from top , there is a small region at the midplane with @xmath92 .
the total @xmath48 is still around 0.01 ( compare fig .
2 ) but with higher fluctuations . locally the magnetic turbulence starts to vanish at the midplane . at around @xmath93 ( model @xmath43 )
the total @xmath53 gets affected and decreases ( compare fig .
2 ) . in model
@xmath43 the elsasser number @xmath94 point is around 1 scale height . at the midplane
@xmath95 drops below 0.1 .
we confirmed the @xmath96 region equivalent with the @xmath97 definition by @xcite , fig . 1 therein .
at this point also the azimuthal mri wavelength becomes unresolved ( @xmath98 , red solid line ) . at the midplane
the reynolds stress dominate over the maxwell stress ( fig .
3 , yellow solid line ) . at @xmath74 ,
the mri gets strongly damped .
for model @xmath42 and model @xmath43 after around 500 years , the poloidal magnetic fields decreases until the elsasser number drops below 0.1 as well as the azimuthal mri wavelength becomes unresolved ( fig .
3 , red solid line ) . if the turbulence is able to survive depends also on the surrounding layers .
even we choose a constant resistivity , the turbulence in the runs @xmath43 and @xmath42 is similar as they are present in dead - zones .
turbulent upper layers of the disk are usually active @xcite . in this active layers
the turbulent mixing is also stronger and active channels can reach into the dead - zone . fig .
4 presents a 2d @xmath99 slice of radial magnetic field overplotted with the @xmath94 line for model @xmath80 and @xmath86 after 1000 years . at around @xmath63
the dissipation starts to damp the mri at the midplane while there is still mri turbulence in the corona of the disk .
a similar shape could be expected at the outer edge of dead - zones ( dzyurkevich et al .
2012 , apj subm . ) .
+ we summarize that for magnetic reynolds numbers above @xmath100 we observe a saturated and converged mri turbulence in zero - net flux stratified simulations down to the midplane region .
the maxwell stress dominates in each region the reynolds stress .
below @xmath65 the turbulence starts to decay at the midplane .
the elsasser number drops below 1 and there regions with low magnetic fields and dominating reynolds stress . at around @xmath101
, the elsasser number drops below 0.1 at the midplane .
long - period oscillations becomes visible with mri activation and decays .
below @xmath102 the mri can not operate . here
, reynolds stress dominates over the maxwell stress .
dependent on the surroundings , the upper layers are still active having mri turbulence .
the results show that for magnetic reynolds numbers down to @xmath103 , the mri driven by a zero - net flux azimuthal magnetic field , can sustained the turbulence with @xmath104 .
models @xmath43 and @xmath42 with @xmath102 show a decrease of the turbulence and eventually a dominating reynolds stress . from this results
we conclude that the critical magnetic reynolds number should around @xmath105 .
@xcite also found a critical magnetic reynolds number around @xmath74 in local box simulations . in the next chapter we concentrate on the run @xmath42 which present similar conditions as they are present in dead - zones .
before we move to the hydro - dynamical motions in dead - zones we want to investigate the long - term oscillations of the maxwell stress we observed in the previous chapter .
this oscillations could be connected to some sort of dynamo process .
@xcite presented in local box simulations a sudden transition from a low turbulent state to a high turbulent state .
we observe in model @xmath43 and @xmath42 for @xmath74 a similar effect .
the results are combined in fig 5 showing the mean toroidal magnetic field overplotted with the turbulent @xmath106 ( times 1000 ) , the elsasser number and the q value . for this analysis
we choose the southern hemisphere between 0 and 1.5 scale heights .
after around 200 years the initial mean toroidal magnetic field ( black solid line ) vanishes and starts to oscillate .
there is a clear correlation between the turbulent emf ( red dotted line ) and the mean field which indicates a working @xmath107 dynamo . in both models , after around 500 years the mri from azimuthal magnetic fields becomes unresolved ( @xmath108 blue dotted line below black dotted line ) . here
the elsasser number drops below unity which still allows for damped mri growth .
after 800 years the elsasser number drops below 0.1 . at this stage
there is no mri working and we are in the dead - zone stage . here also the oscillations of mean field and turbulent emf stop . in model
@xmath43 the mri switch on again after less then 200 years ( 27 local orbits ) with a larger oscillation period . in model @xmath42
we observe a late switch on at 1100 years for only 150 years ( 20 local orbits ) .
afterwards the elsasser number reaches nearly @xmath109 and we do nt expect a relaunch again .
the sudden increase of @xmath95 is due to a sudden increase of poloidal magnetic field .
this sudden increase could be connected with the accumulation of toroidal magnetic fields ( black solid line in fig . 5 , bottom ) in the dead - zone as well as due to radial transport of magnetic field
( see chapter 3.5 ) .
in fig . 6 , we present an azimuthal slice of toroidal magnetic field in the @xmath99 plane after 1500 years for model @xmath42 . in the dead - zone ( @xmath110 )
we observe locally a q factor of 8 and above which correspond to a plasma beta below 50 , still the dissipation is too high to relaunch the mri growth by azimuthal fields . in general the line of
resolved azimuthal field @xmath108 matches the one of decaying mri @xmath110 ( compare fig .
+ this dead - zone region is also interesting in terms of the hydro - dynamical motions .
a close looks at the radial velocity in the @xmath111 midplane in fig .
7 reveals velocity amplitudes around @xmath112 . on the first
look these hydro - dynamical waves appear to be linear waves with a single mode . but
a closer look at the fourier spectra of the radial velocity reveals a more detailed picture .
we calculated the spectra in the dead - zone region between 4 and 5 au , time averaged between 1300 and 1500 years in model @xmath42 . in fig .
8 , we compare this spectra with the one obtained in fully ionized disk from model @xmath113 . in the dead - zone
there is a peak at @xmath114 .
the turbulence at lower scales ( green solid line ) is around 1 order of magnitude below the value in the fully mri turbulent region . beside the density waves we observe anti - cyclonic vortices ( fig .
7 between 6 and 7 au ) which create large extended spiral arms .
they are produced at around 8 au where we have a positive density slope due to the buffer zones .
we calculated the relative vorticity @xmath115 between @xmath116 and @xmath117 in the large vortex .
the growth of vortices at the border of the dead zone can be due to the rossby wave instability as proposed by @xcite and recently investigated by @xcite . the vortex structure
is presented in fig .
the vertical extension is around @xmath118 sh , the radial extension is @xmath119 and the azimuthal one is around @xmath120 .
the vortex is dragged by the hydro - dynamical surroundings which gives him the concave ( fig .
9 , top , dominant radial inward velocity ) , convex ( fig . 9 , bottom , @xmath121 ) shape respectively .
the magnetic fields ( fig .
9 , bottom black vectors ) are not present inside the vortex but in the surrounding layers .
we measure a plasma beta of @xmath122 inside the vortex compared to @xmath123 in the surrounding region .
we already mentioned the effect of radial transport due to turbulent mixing . in this subsection
we want to focus on the importance of radial transport of magnetic fields .
we calculate the divergence of the radial poynting flux @xmath124 integrated over time for the total @xmath33 and @xmath125 domain at @xmath126 using one scale height in radius .
we integrate the divergence of the radial poynting flux over every output ( @xmath127 local orbits ) and compare it with the total magnetic energy .
10 shows how much magnetic energy is transported at each output ( 0.1 local orbits ) . in the fully turbulent layers the net radial transport is around zero with fluctuations of around 5 % . in the transition zone and the dead - zone
the radial transport becomes more important and these fluctuations reach peaks of 10 to 20 % .
in our work we showed that the mri turbulence saturates around @xmath128 with @xmath2 . here
the elsasser number is around @xmath129 .
this turbulence level is sustained for elsasser number values of @xmath130 .
stratified converged ideal mhd local simulations @xcite presented similar values of @xmath48 .
still our results indicate a higher turbulence level using small explicit resistivity .
high resolution simulations using explicit resistivity are needed to confirm the convergence of the accretion stress .
in addition , there are mean - field mechanism , which can temporal increase the value of @xmath48 @xcite . + for magnetic reynolds numbers around the critical value @xmath131 we observe long - period oscillations of the accretion stress as well as irregular switch off and on of mri activity .
such oscillations were also reported by @xcite in local box simulations .
we see indications that these oscillations are trigger by some dynamo process as locally there are accumulations of strong mean toroidal fields in the dead - zone .
also radial transport of magnetic field could here play a role . in such configurations
the corona is still active while in the midplane the mri is switched of . here
the reynolds stress dominate over the maxwell stress at @xmath110 .
a similar classification of turbulent regions in proto - planetary disks was done by comparing different heights , see also in @xcite , fig . 1
. in such turbulent simulations there are 3 kinds of important dissipation sources . the numerical dissipation , due to the finite grid size , the explicit resistivity and the turbulent dissipation . in model
@xmath42 the numerical dissipation plays an important role .
in addition one has to note that using a uniform grid , the numerical dissipation will be larger at the inner part then in the outer part . here a logarithmic increasing grid would help .
then there is of course the explicit dissipation which is included in the induction equation in the code
. here one should be sure that the explicit one is higher than the numerical one .
the turbulent dissipation scales with the level of the turbulence .
here it is still unclear how this dissipation process works in detail , see also fig . 9 in @xcite .
we performed 3d global ideal and non - ideal mhd simulations to study mri turbulence in low - ionized proto - planetary disks using an initial toroidal magnetic field for a range of magnetic reynolds numbers . with our global simulations we are able to investigate the transition regime which is present between the active and the dead - zone in proto - planetary disks .
we define 3 different disk regimes dependent on the magnetic reynolds numbers : * above a reynolds number of @xmath132 . here
mri turbulence saturates and is sustained . for this region
we find steady and converged @xmath48 values around 0.01 in ideal and non - ideal simulations .
the maxwell stress dominates down to the midplane region .
the turbulent velocities reach maximum values around @xmath71 .
the @xmath133 dynamo is operating and we find @xmath134 with a positive sign in the northern hemisphere .
the elsasser number @xmath95 stays above 1 in the midplane . *
reynolds number between @xmath135 . here
, the mri starts to switch of at the midplane region .
the elsasser number @xmath95 drops below 1 .
simulations with @xmath136 show a still sustained turbulence , supported by radial transport of magnetic fields .
for @xmath101 the mri turbulence starts to switch off .
we expect the critical magnetic reynolds number around @xmath137 and below .
* reynolds number around @xmath138 and below . here
, there is no mri at the midplane anymore .
the elsasser number @xmath95 drops below 0.1 .
the turbulence is dominated by the reynolds stress .
we observe long - period oscillations of mri activity and mri decay .
there is an accumulation of toroidal magnetic fields below @xmath139 in the dead - zone .
the turbulence at the midplane is supported by active channels in the corona which pump sound wave into the midplane region .
the velocity spectra at the midplane reveals a drop of one order of magnitude at the scale of h in the dead - zone compared to the fully ionized turbulent region .
we observe long lived anti - cyclonic vortices in the transition regime , creating spiral arms in the dead - zone .
magnetic fields are not present in the inner part of the vortex .
we thank sebastien fromang for the helpful comments on the global models .
we thank natalia dzyurkevich for her suggestions and comments during this work .
we thank also neal turner , satoshi okuzumi and hlose meheut for their revision .
we thank andrea mignone for supporting us with the pluto code .
the research leading to these results has received funding from the european research council under the european union s seventh framework programme ( fp7/2007 - 2013 ) / erc grant agreement n 258729 .
parallel computations have been performed on the theo cluster of the max - planck institute for astronomy heidelberg located at the computing center of the max - planck society in garching . | we investigate the characteristic properties of self - sustained mri turbulence in low - ionized proto - planetary disks .
we study the transition regime between active and dead - zone , performing 3d global non - ideal mhd simulations of stratified disk covering range of magnetic reynolds number between @xmath0 .
we found converged and saturated mri turbulence for @xmath1 with a strength of @xmath2 .
below @xmath3 the mri starts to decay at the midplane , having elsasser numbers below one .
we find a transition regime between @xmath4 where the mri turbulence is still sustained but damped . at around @xmath5 the mri turbulence decays but could reestablished due to the accumulation of toroidal magnetic field or the radial transport of magnetic field from the active region . below @xmath6 the mri can not be sustained and is decaying . here hydro - dynamical motions , like density waves dominate .
we observe long - living anti - cyclonic vortices in the transition between dead - zone and active zone . |
spontaneous parametric down - conversion ( spdc ) is the basic source of non - classical light in experimental quantum optics @xcite , testing foundations of the quantum theory @xcite , and implementing protocols for quantum information information processing and communication @xcite .
the essential feature of spdc is the guarantee that the photons are always produced in pairs , and suitable arrangements allow one to generate various types of classical and quantum correlations within those pairs .
the physics of spdc depends strongly on optical properties of nonlinear media in which the process is realized .
this leads to an interplay between different characteristics of the source and usually imposes trade - offs on its performance .
for example , many experiments require photon pairs to be prepared in well - defined single spatio - temporal modes .
in contrast , photons generated in typical media diverge into large solid angles and are often correlated in space and time , as shown schematically in fig .
[ fig : source ] .
specific modes can be selected afterwards by coupling the output light into single - mode fibers and inserting narrowband spectral filters .
however , it is usually not guaranteed that both the photons in a pair will always have the matching modal characteristics , and in many cases only one of the twin photons will get coupled in @xcite .
this effect , which can be modelled as a loss mechanism for the produced light , destroys perfect correlations in the numbers of twin photons .
these losses come in addition to imperfect detection , and can be described jointly using overall efficiency parameters .
is pumped with a laser beam @xmath0 .
generated photons are highly correlated and useful modes @xmath1 and @xmath2 are typically selected by narrow spatial and frequency filters @xmath3 . ]
the effects of losses become more critical when the spdc source is pumped with powers so high that it is no longer possible to neglect the contribution of events when multiple pairs have been simultaneously produced @xcite .
such a regime is necessary to carry out multiphoton interference experiments , it can be also approached when increasing the production rate of photon pairs .
one is then usually interested in postselecting through photocounting the down - conversion term with a fixed number of photon pairs and observing its particular quantum statistical features @xcite . in the presence of losses
the same number of photocounts can be generated by higher - order terms when some of the photons escape detection .
however , the statistical properties of such events can be completely different , thus masking the features of interest .
although some quantum properties may persist even in this regime , with a notable example of polarization entanglement @xcite , their extraction and utilization becomes correspondingly more difficult .
the present paper is an experimental study of multiphoton events in spontaneous parametric down - conversion with particular attention paid to the effects of filtering and losses .
the multiple - pair regime is achieved by pumping the nonlinear crystal by the frequency - doubled output of a 300 khz titanium - sapphire regenerative amplifier system . the kilohertz repetition rate has allowed us to count the number of the photons at the output with the help of the loop detector @xcite . using a simplified theoretical description of the spdc source we introduce effective parameters that characterize its performance in multiphoton experiments .
the obtained results illustrate trade - offs involved in experiments with multiple photon pairs and enable one to select the optimal operation regime for specific applications .
this paper is organized as follows .
first we describe a theoretical model for spdc statistics in sec .
[ sec : spdcstat ] .
[ sec : parameters ] introduces effective parameters to characterize spdc sources .
the experimental setup and measurement results are presented in sec .
[ sec : exp ] . finally , sec .
[ sec : conclusions ] concludes the paper .
we will start with a simple illustration of the effects of higher - order terms in spdc .
suppose for simplicity that the source produces a two - mode squeezed state which can be written in the perturbative expansion as @xmath4 , where @xmath5 measures squeezing and is assumed to be real . for two - photon experiments , the relevant term is @xmath6 and the contribution of the higher photon number terms
can be neglected as long as @xmath7 .
this enables postselecting the two - photon term and observing associated quantum effects , such as hong - ou - mandel interference .
suppose now that each of the modes is subject to losses characterized by @xmath8 , where @xmath9 is the overall efficiency .
losses may transform the term @xmath10 into @xmath11 or @xmath12 , whose presence will lower the visibility of the hong - ou - mandel interference .
the two - photon term now occurs with the probability @xmath13 , while the four - photon term effectively produces one of the states @xmath11 or @xmath12 with the total probability equal to @xmath14 .
this constitutes a fraction of @xmath15 of the events that come from single pairs produced by the source .
this fraction can easily become comparable with one , especially when the losses are large .
let us now develop a general model of photon statistics produced by an spdc source . in the limit of a classical undepleted pump the output field is described by a pure multimode squeezed state . by a suitable choice of spatio - temporal modes , called characteristic modes , such a state can be brought to the normal form @xcite in which modes are squeezed pairwise . denoting the annihilation operators of the characteristic modes by @xmath16 and @xmath17 ,
the non - vanishing second - order moments can be written as : @xmath18 where @xmath19 is the squeezing parameter for the @xmath20th pair of modes .
because the state of light produced in spdc is gaussian , these equations , combined with the fact that first - order moments vanish @xmath21 define fully quantum statistical properties of the output field .
let us first consider the case when the spatial and spectral filters placed after the source select effectively single field modes .
the annihilation operators @xmath22 and @xmath23 are given by linear combinations of the characteristic modes : @xmath24 where @xmath25 and @xmath26 describe amplitude transmissivities of the filters for the characteristic modes and @xmath27 . because the complete multimode state is gaussian , the reduced state of the modes @xmath22 and @xmath23 is also gaussian , and it is fully characterized by the average numbers of photons @xmath28 and @xmath29 , and the moment @xmath30 .
these quantities can be written in terms of the multimode moments given in eq .
( [ eq:<aa>,<ab > ] ) , but we will not need here explicit expressions .
it will be convenient to describe the quantum state of the modes @xmath22 and @xmath23 with the help of the wigner function : @xmath31 where @xmath32 and @xmath33 is the correlation matrix composed of symmetrically ordered second order moments : @xmath34 given the wigner function of the reduced state for the modes @xmath22 and @xmath23 in the gaussian form , the calculation of the joint count statistics @xmath35 is straightforward .
it will be useful to introduce an operator representing the generating function of the joint count statistics @xmath36 whose expectation value over the quantum state expanded into the power series yields the joint count statistics @xmath37 : @xmath38 because the operator @xmath39 is formally equal , up to a normalization constant , to product of density matrices describing thermal states of modes @xmath22 and @xmath23 with average photon numbers @xmath40 and @xmath41 respectively , the corresponding wigner function has a gaussian form : @xmath42 using the above expression it is easy to evaluate the generating function for the joint count statistics by integrating the product of the respective wigner functions : @xmath43 where we introduced the following three parameters : @xmath44 these three parameters have a transparent physical interpretation in the regime when @xmath45 . then , it is easy to check that the generating function given in eq .
( [ eq : xi(x , y)=squeezedwlosses ] ) describes the count statistics of a plain two - mode squeezed state whose two modes have been sent through lossy channels with transmissivities @xmath9 and @xmath46 , and the average number of photons produced in each of the modes was equal to @xmath47 .
note that the inequality @xmath45 implies nonclassical correlations between the modes @xmath22 and @xmath23 .
if the opposite condition is satisfied , the state can be represented as a statistical mixture of coherent states in modes @xmath22 and @xmath23 with correlated amplitudes . in this case , it is not possible to carry out an absolute measurement of losses .
in a realistic situation , the spectral filters employed in the setup are never sufficiently narrowband to ensure completely coherent filtering
. therefore a sum of counts originating from multiple modes will be observed .
we will model this effect by assuming that the detected light is composed of a certain number of @xmath48 modes with identical quantum statistical properties described in the preceding section .
the generating function @xmath49 for count statistics is therefore given by the @xmath48-fold product of the expectation value @xmath50 calculated in eq .
( [ eq : xi(x , y)=squeezedwlosses ] ) : @xmath51 the parameter @xmath48 , which we will call the equivalent number of modes , can be read out from the variance of the count statistics in one of the arms characterized by the generating function @xmath49 . for a single mode source
the variance is equal to that of a thermal state with @xmath52 .
it is easily seen that for @xmath48 equally populated modes the variance becomes reduced to the value @xmath53 .
solving this relation for @xmath48 leads us to a measurable parameter that will help us to characterize the effective number of detected modes : @xmath54 note that @xmath48 is closely related to the inverse of the mandel parameter @xcite .
the second parameter we will use to characterize the spdc source measures the overall losses experienced by the produced photons . let us first note that in the perturbative regime , when all the squeezing parameters @xmath55 , each one of the quantities @xmath56 , @xmath57 , and @xmath58 appearing in eq .
( [ eq : r2,eta = onemode ] ) is proportional to the pump intensity . in this regime
the efficiencies can be approximated by the ratios @xmath59 and @xmath60 that are independent of the pump intensity .
on the other hand , the average photon numbers @xmath61 and @xmath62 in both the arms are linear in the pump power . following early works on squeezing @xcite , we will introduce here a parameter that quantifies the subpoissonian character of correlations between the counts @xmath63 and @xmath64 in the two arms of the setup .
first we define a stochastic variable : @xmath65 this definition takes into account the possibility of different losses in the two arms through a suitable normalization of the count numbers .
the subpoissonian character of the correlations can be tested by measuring the average @xmath66 .
the semiclassical theory predicts @xmath67 , while for two beams with count statistics characterized by the generating function @xmath68 one obtains : @xmath69 in the case of equal efficiencies @xmath70 this expression reduces simply to @xmath71 .
we will use the last relation as a method to measure the average overall efficiency of detecting the state produced by the spdc source . as discussed at the beginning of sec .
[ sec : spdcstat ] , the initial quantum state of light used for many experiments should ideally be in a state @xmath6 or @xmath10 .
in the perfect case of two - mode squeezing and negligible losses , such states can be isolated through postselection of the spdc output on the appropriate number of counts . in practice
, the postselected events will also include other combinations of input photon numbers .
because typical detectors have limited or no photon number resolution , one needs to take into account also the deleterious contribution of higher total photon numbers .
this leads us to the following definition of the parameter measuring the contamination of photon pairs with other terms that can not be in general removed through postselection : @xmath72 an analogous definition can be given for quadruples of photons , which ideally should be prepared in a state @xmath10 : @xmath73 in fig .
[ fig:2imp ] we depict contour plots of the contamination parameters @xmath74 and @xmath75 as a function of the production rates and the overall efficiency .
it is clearly seen that the non - unit efficiency imposes severe bounds on the production rates that guarantee single or double photon pair events sufficiently free from spurious terms .
the graphs also imply that strong pumping is not a sufficient condition to achieve high production rates , but it needs to be combined with a high efficiency of collecting and detecting photons .
( upper plot ) and double photon pairs @xmath74 ( lower plot ) as a function of the overall efficiency @xmath9 and the respective production rates @xmath76 and @xmath77 . the calculations have been carried out in the regime of single selected modes when @xmath78.,title="fig : " ] ( upper plot ) and double photon pairs @xmath74 ( lower plot ) as a function of the overall efficiency @xmath9 and the respective production rates @xmath76 and @xmath77 .
the calculations have been carried out in the regime of single selected modes when @xmath78.,title="fig : " ]
the experimental setup is depicted in the fig . [ fig : setup ] .
the master laser ( rega 9000 from coherent ) produces a train of 165 fs fwhm long pulses at a 300 khz repetition rate centered at the wavelength 774 nm , with 300 mw average power .
the pulses are doubled in the second harmonic generator xsh based on a 1 mm thick beta - barium borate ( bbo ) crystal cut for a type - i process .
ultraviolet pulses produced this way have 1.3 nm bandwidth and 30 mw average power .
they are filtered out of the fundamental using a pair of dichroic mirrors dm and a color glass filter bg ( schott bg39 ) , and imaged using a 20 cm focal length lens il on a downcoversion crystal x , where they form a spot measured to be 155 @xmath79 m in diameter .
the power of the ultraviolet pulses , prepared in the polarization perpendicular to the plane of the setup , is adjusted using neutral density filters nd and a motorized half waveplate .
the type - i down - conversion process takes place in a 1 mm thick bbo crystal x cut at @xmath80 to the optic axis , and oriented for the maximum source intensity .
the down - converted light emerging at the angle of @xmath81 to the pump beam is coupled into a pair of single mode fibers placed at the opposite ends of the down - conversion cone .
the fibers and the coupling optics define the spatial modes in which the down - conversion is observed @xcite . the coupled photons enter the loop detector @xcite in which light from either arm can propagate towards one of the detectors through eight distinct paths .
the minimal delay difference between two paths is 100 ns , more than twice the dead time of the detectors .
finally the photons exit the fiber circuit , go through interference filters if and are coupled into multimode fibers which route them directly to single photon counting modules spcm ( perkinelmer spcm - aqr-14-fc ) connected to fast coincidence counting electronics ( custom - programmed virtex4 protype board ml403 from xilinx ) detecting events in a proper temporal relation to the master laser pulses .
the measurement series are carried out with pairs of interference filters of varying spectral widths .
the measurement proceeds as follows : for each pair of interference filters the loop detector is calibrated using data collected at a very low pump light intensities , when the chance of more than one photon entering the fiber circuit is negligible compared to the rate single photons appear .
this allows one to calculate the complete matrix of conditional probabilities @xmath82 of observing @xmath20 detector clicks with @xmath63 initial photons @xcite for one arm , and analogously @xmath83 for the second arm .
the losses in the detectors and in the fiber circuit are assumed to contribute to the overall efficiencies @xmath9 and @xmath46 .
thus @xmath82 and @xmath83 describe lossless loop detectors for which @xmath84 . after the calibration ,
the counts are collected for approximately @xmath85 master laser pulses for each chosen intensity of the pump , which yields the probabilities @xmath86 of observing @xmath20 clicks in one arm and @xmath87 in the other one .
these probabilities are related to the joint count probability @xmath35 corrected for combinatorial inefficiencies of the loop detector through the formula : @xmath88 the probabilities @xmath35 can be retrieved from experimentally measured @xmath86 using the maximum likelihood estimation technique @xcite .
an exemplary joint photon number distribution reconstructed from the experimental data is shown in fig .
[ fig : exprho ] . the results of the reconstruction
are subsequently used to calculate the parameters of the source discussed in the preceding section .
= 0.15 and @xmath89=0.18 . ] in a single arm as a function of the average photon number @xmath90 , for measurements carried out without interference filters ( circles ) , with 10 nm fwhm filters ( squares ) and 5 nm fwhm filters ( crosses ) . ] in fig .
[ fig : m ] we depict the reconstructed equivalent number of modes @xmath48 in single arm for different filtering and pump intensities .
this quantity can be also understood as a measure of how many incoherent modes a photon from the source occupies .
naturally , @xmath48 drops with application of narrowband spectral filtering since it erases the information on the exact time the photon pair was born in the nonlinear crystal .
it also is seen that @xmath48 is practically independent on the pumping intensity , as predicted in sec .
[ sec : parameters ] . as a function of the total average photon number @xmath90 , measured without interference filters ( circles ) , with 10 nm fwhm filters ( squares ) and 5 nm fwhm filters ( crosses ) . ] the average overall efficiency of the squeezed state calculated as @xmath91 is shown in fig .
[ fig : eta ] .
it is seen that the efficiency decreases with an application of narrowband filtering , which again is easily understood .
in addition , @xmath9 exhibits a very weak dependence on the pump intensity which again agrees with theoretical predictions for the regime when the average number of photons is much less than one . a relatively large difference in @xmath9 between 10 nm and 5 nm interference filters
can be explained by the fact that in the latter case the selected bandwidth becomes narrower than the characteristic scale of spectral correlations within a pair .
consequently , the filter in one arm selects only a fraction of photons conjugate to those that have passed through the filter placed in the second arm .
this observation is consistent with the determination of the parameter @xmath92 , which shows that for 5 nm filters effectively single spectral modes are selected . for single photon pairs as a function of the pair production rate , measured without interference filters ( circles ) , with 10 nm fwhm filters ( squares ) and 5 nm fwhm filters ( crosses ) .
the parameters of the fitted theoretical curves are : @xmath93 and @xmath94 ( red solid line ) , @xmath95 and @xmath96 ( red dashed line ) and @xmath97 and @xmath98 ( red dash - dot line ) . ] for double photon pairs as a function of the production rate , measured without interference filters ( dots ) .
the parameters of the fitted theoretical curve are : @xmath99 and @xmath100 ( red dashed line ) .
measurements with interference filters did not provide statistically significant data . ] finally , in figs .
[ fig:2impexp ] and [ fig:4impexp ] we plot the respective contamination parameters for single @xmath74 and double @xmath75 photon pairs , that describes non - postselectable contributions of other photon terms to the output . as expected , the contamination grows with decreasing overall efficiency @xmath9 that corresponds to narrowing the spectral bandwidth , as well as with the increasing pair production rates .
it is noteworthy that the measurement results agree well with fitted theoretical curves whose parameters match those that can be read out from figs .
[ fig : m ] and [ fig : eta ] .
it is seen that in the regime the experiment was carried out the contamination is rather significant , and it would affect substantially effects such as two - photon interference .
this limits in practice the energy of pump pulses and consequently the production rates that result in photon pairs sufficiently free from spurious terms .
the effect is particularly dramatic in the case of @xmath75 , where genuine double pairs appear only in approximately half of all the events .
despite the simplicity of the basic concept , the application of spdc sources in more complex quantum optics and quantum information processing experiments requires a careful choice of operating conditions .
the actual output state is a result of a subtle interplay between the pump strength , spatial and spectral filtering of the output , and the losses experienced by the signal . in this paper we concentrated on the features of the photon number distribution and developed an effective theoretical description .
starting from the fact that the spdc output processed by arbitrary passive linear optics is given by a gaussian state , we introduced effective parameters characterizing the joint photon number distribution .
the first parameter is the effective number of orthogonal modes that impinge on each of the detectors .
this parameter carries information about the modal purity of the photons produced in each arm .
the second parameter characterizes average losses experienced by the spdc output , and can be calculated as a suitably normalized variance of the count difference .
finally , we also proposed and determined experimentally a measure of the contamination of photon pairs by other spurious terms that in general can not be rejected by postselection and may contribute to an unwanted background .
we showed that the presence of such a background puts stringent requirements on the detection efficiency if a bright source with high pair generation probability is desired .
these constraints become very challenging when designing multiple - pair experiments .
we acknowledge helpful discussions with c. silberhorn and i. a. walmsley . this work has been supported by the polish budget funds for scientific research projects in years 2005 - 2008 and the european commission under the integrated project qubit applications ( qap ) funded by the ist directorate as contract number 015848 .
the experiment has been carried out in the national laboratory for atomic , molecular , and optical physics in toru , poland .
p. g. kwiat , e. waks , a. g. white , i. appelbaum , and p. h. eberhard , phys .
a * 60 * , r773 ( 1999 ) ; t. paterek , a. fedrizzi , s. grblacher , t. jennewein , m. ukowski , m. aspelmeyer , and a. zeilinger , phys .
* 99 * , 210406 ( 2007 ) ; c. branciard , a. ling , n. gisin , c. kurtsiefer , a. lamas - linares , and v. scarani , _ ibid . _ * 99 * , 210407 ( 2007 )
. j. l. obrien , g. j. pryde , a. g. white , t. c. ralph , d. branning , nature * 426 * , 264 ( 2003 ) ; p. walther , k. j. resch , t. rudolph , e. schenck , h. weinfurter , v. vedral , m. aspelmeyer , and a. zeilinger , _ ibid . _ * 434 * , 169 ( 2005 ) ; m. halder , a. beveratos , n. gisin , v. scarani , c. simon , and h. zbinden , nature physics * 3 * , 692 ( 2007 ) ; g. vallone , e. pomarico , f. de martini , and p. mataloni , phys .
lett . * 100 * , 160502 ( 2008 ) .
arvind , b. dutta , n. mukunda , and r. simon , pramana - j .
* 45 * , 471 ( 1995 ) ; r. s. bennink and r. w. boyd , phys .
a * 66 * , 053815 ( 2002 ) ; a. botero and b. reznik , _ ibid . _
* 67 * , 052311 ( 2003 ) ; s. l. braunstein , _ ibid . _ * 71 * , 055801 ( 2005 ) ; w. wasilewski , a. i. lvovsky , k. banaszek , and c. radzewicz , _ ibid . _ * 73 * , 063819 ( 2006 ) .
f. a. bovino , p. varisco , a. m. colla , g. castagnoli , g. di giuseppe , and a. v. sergienko , opt . comm . * 227 * , 343 ( 2003 ) ; s. castelletto , i. p. degiovanni , a. migdall , and m. ware new j. phys . * 6 * , 87 ( 2004 ) ; a. dragan , phys .
a * 70 * , 053814 ( 2004 ) . | we present an experimental characterization of the statistics of multiple photon pairs produced by spontaneous parametric down - conversion realized in a nonlinear medium pumped by high - energy ultrashort pulses from a regenerative amplifier .
the photon number resolved measurement has been implemented with the help of a fiber loop detector .
we introduce an effective theoretical description of the observed statistics based on parameters that can be assigned direct physical interpretation .
these parameters , determined for our source from the collected experimental data , characterize the usefulness of down - conversion sources in multiphoton interference schemes that underlie protocols for quantum information processing and communication . |
landau s fermi liquid theory @xcite can be considered as the `` standard theory '' of solid state physics .
it predicts that the electronic properties of a ( normal ) metal follow , in the presence of coulomb correlations , a renormalized ( quasiparticle ) version of those for non - interacting electrons . among the most fundamental physical properties
is the heat capacity which , at low temperatures , is predominated by the electronic degrees of freedom . for a normal metal ,
landau s fermi liquid theory @xcite predicts a linear increase of the specific heat capacity with temperature @xmath0 and a cubic term , more precisely a term @xmath1 , as the leading correction @xcite . for free electrons ,
the prefactor @xmath2 is directly proportional to the electronic density of states at the fermi level since , due to the pauli principle , only these electrons contribute . in the case of a fermi liquid
, we merely need to introduce a quasiparticle renormalization factor @xmath3 to account for the coulomb interaction , which enhances the specific heat of a correlated metal by @xmath4 .
however , this electronic contribution prevails only at temperatures much lower than the debye temperature .
otherwise the cubic phononic term , which has a much higher prefactor because of its bosonic nature , dominates .
as we have shown in a recent letter @xcite , this longstanding conception needs to be supplemented : for strongly correlated electrons , a clear kink in the temperature dependence of the specific heat appears , marking the abrupt change from one linear behavior to a second one with a reduced slope at higher temperatures .
this can be shown numerically solving the hubbard model @xcite within dynamical mean field theory @xcite , using exact diagonalization as an impurity solver .
quantitatively the same results can also been obtained analytically , following the derivation by abrikosov _
@xcite for the specific heat , taking the recently observed kinks in the self energy as a starting point @xcite .
the experimental confirmation of these theoretical results is somewhat problematic since typically , at the kink temperature , the phononic contribution to the specific heat is already dominating .
an exception in this respect is liv@xmath5o@xmath6 @xcite , the first heavy fermion system with @xmath7-electrons , since the kink temperature is here particularly low . and ,
indeed , the precise inspection of the experimental data shows a kink in the correct temperature range @xcite . in this proceeding
, we will review the previous theoretical results and include additional data .
in the presence of strong electronic correlations , the @xmath8-integrated spectral function @xmath9 shows a typical three peak structure with a lower and an upper hubbard band and a quasiparticle peak in between . using the dmft formula @xmath10 which relates green function and self energy at frequency @xmath11 , one can show @xcite that such a strongly correlated three peak structure necessarily entails a kink .
note that the hybridization function @xmath12 in eq .
( [ eq : sigmacontributions ] ) is for a simple semi - circular density of states just @xmath13 ( @xmath14 : half the bandwidth ) . for other lattices , @xmath12 depends on the moments of the non - interacting density of states with a similar term proportional @xmath15 dominating .
the kink argument is now as follows @xcite : the first three terms of the r.h.s of eq .
( [ eq : sigmacontributions ] ) yield an almost linear frequency dependence for the real part of the self energy over the entire region of the central quasiparticle peak .
the derivative hence gives a quasiparticle renormalization factor @xmath16^{-1}\label{eq : z}\ ] ] the last ( hybridization ) term however yields an additional contribution to @xmath17 which basically is proportional to @xmath18 .
this real part can be directly obtained from the spectral function of the central quasiparticle peak @xmath19 via a kramers - kronig transformation .
_ inside _ the overall width of the quasiparticle peak , there is a turning point in @xmath9 .
hence , @xmath18 has a linear frequency dependence up to an maximum at the turning point @xmath9 .
after this maximum changes in @xmath18 are minor .
altogether this means that at low frequencies we have to add the slope obtained from the @xmath12 in eq .
( [ eq : z ] ) yielding an altogether smaller fermi liquid renormalization factor @xmath20 at low energies . in between the two regimes
@xmath20 and @xmath21 , there is a kink at a frequency @xmath22 . in summary
we hence have : @xmath23 with the constant @xmath24 providing for a continuous function .
before we turn to the specific heat in the next section , let us note that the kink in the self energy directly leads to a kink in the energy - momentum dispersion relation of the correlated electrons .
let us also remark here , that the result of generic kinks in the self energy of strongly correlated systems should be robust beyond dmft , as similar effects are to be expected in cluster @xcite and diagrammatic extensions @xcite of dmft .
based on eq . ( [ eq : sigma ] ) , we have developed an analytical theory for the specific heat on the basis of a formula by abrikosov , gorkov and dzyaloshinski ( agd ) for the entropy of a fermionic system at low temperatures@xcite : @xmath25 , \label{eq : abr2}\ ] ] this agd formula is ( through a low - temperature diagrammatic expansion ) based on the low frequency behavior of the self energy @xmath26 at zero temperature [ or the corresponding retarded and advanced green function in eq .
( [ eq : abr2 ] ) ] so that we can directly apply it with @xmath26 from eq .
( [ eq : sigma ] ) .
let us note that @xmath27 is the density of states and the frequency integral has been rewritten through a dimensionless variable @xmath28 ( @xmath29 ) . the agd formula ( [ eq : abr2 ] )
can be easily differentiated w.r.t .
@xmath30 , so that the specific heat @xmath31 can be computed directly . with some algebra
( see @xcite for more details ) and substituting substituting @xmath32 by @xmath33-functions , we obtain the final result @xmath34 d y \ , \frac{y^2 \mbox{e}^y } { ( \mbox{e}^y+1)^2}. \label{eq : final}\end{aligned}\ ] ] knowing the two renormalization factors @xmath35 and @xmath36 and the kink frequency @xmath22 as well as the non - interacting density of states @xmath27 , eq .
( [ eq : final ] ) allows us to calculate the specific heat through a simple integral .
if vice versa , the specific heat is known , e.g. , experimentally , we can employ eq .
( [ eq : final ] ) as a fit formula for obtaining @xmath35 , @xmath36 and @xmath22 .
the latter ( fit formula ) approach had been taken for the experimental validation of the specific heat kinks in liv@xmath5o@xmath6 since lda+dmft calculations @xcite for liv@xmath5o@xmath6 @xcite could not provide for a fine enough resolution of the low frequency dependence of the self energy , even when using the projective quantum monte carlo method @xcite which is most appropriate for this purpose . as fig .
3 of @xcite shows the three fitting parameters provide for an excellent agreement of eq .
( [ eq : final ] ) and the specific heat of @xcite , including a clearly visible _ kink_. let us note that we did not subtract a phonon contribution for the specific heat since ( i ) it does not alter the existence or non - existence of a kink , ( ii ) it is still an order of magnitude smaller than the electronic contribution at the kink temperature and ( iii ) there is a considerable arbitrariness depending on the reference system chosen .
[ fig:1 ] shows results of our final equation ( [ eq : final ] ) applied to the hubbard model with a semi - circular density of states with half bandwidth @xmath37 . as an input to eq .
( [ eq : final ] ) , @xmath35 , @xmath36 and @xmath22 have been fitted to the numerical renormalization data for the dmft self energy from ref .
besides , we have also calculated the dmft specific heat directly , using exact diagonalization ( ed ) with 7 levels in the impurity bath .
both results agree very well , and the minor deviations can be attributed to the numerical inaccuracy of the exact diagonalization study for which @xmath38 had to be obtained via a numerical differentiation of the total energy . in the inset , we show the specific heat over a slightly extended temperature range .
as the quasiparticle peak eventually ends , the specific heat has a maximum after the kink temperature .
note that this maximum is not contained in our formula ( [ eq : final ] ) since it would require an extension of the self energy description eq .
( [ eq : sigma ] ) to higher energies where @xmath17 has a maximum at the frequencies between hubbard and quasiparticle bands and @xmath39 becomes important .
obtained via the agd based equation ( [ eq : final ] ) [ solid lines ] and numerically by exact diagonalization of the dmft equations ( crosses ) .
inset : behavior over an extended temperature range which shows a maximum after the kink temperature marking the end of the quasiparticle band .
[ fig:1],width=340 ]
landau s fermi liquid theory predicts a linear temperature dependence of the electronic specific heat capacity with temperature . we have shown this longstanding conception needs to be supplemented : for strongly correlated electrons , a clear kink in the temperature dependence appears , marking the abrupt change from one linear behavior to a second one with a reduced slope at higher temperatures .
recent experiments on liv@xmath5o@xmath6 , an ideal material for studying the electronic specific heat , confirm our theory .
a consequence of our findings is that materials with correlated electrons are more resistive to cooling at low temperatures ( where cooling is particularly difficult ) than expected from the behavior at intermediate temperatures .
we thank r. bulla and m. nohara for making available their raw data .
this work has been supported in part by the european - indian cooperative network monami . | the textbook knowledge of solid state physics is that the electronic specific heat shows a linear temperature dependence with the leading corrections being a cubic term due to phonons and a cubic - logarithmic term due to the interaction of electrons with bosons .
we have shown that this longstanding conception needs to be supplemented since the generic behavior of the low - temperature electronic specific heat includes a kink if the electrons are sufficiently strongly correlated . |
the research fields of magnonics and , on a broader scale , magnon spintronics , which includes interfacing to electric charge transport phenomena , aim at the utilization of magnons , the quanta of spin - wave excitation in a magnetic material , as information carriers in a next generation of logic devices beyond conventional cmos@xcite .
this research area has attracted prominent interest in the recent years due to the demonstration of various game - changing devices such as magnonic all - linear time - inverters@xcite , magnon transistors@xcite and reconfigurable magnonic waveguides@xcite , which are just a few , yet instructive , examples . however , for a successful application of magnonics on the micro- and nanometer scale , the research field still faces several challenges .
one of the main issues is the spin - wave damping , which is substantial in most conventionally used , cmos - compatible ferromagnetic metals@xcite .
while recent studies show the large potential to decrease this damping by material engineering@xcite , a reduction of the spin - wave damping comes together with a restriction of the maximum tolerable amplitude of the spin - wave excitation if nonlinear phenomena are to be avoided@xcite .
this is , for example , the case if several spin waves are conveying information encoded in their phase in parallel in a magnonic waveguide .
hence , in future magnonic devices , spin waves with small amplitudes and energies in the @xmath0- to @xmath1-range are envisioned to propagate through extended magnonic networks on a chip , conveying information through spin - wave conduits which feature a small , yet finite damping
. therefore , spin - wave amplifiers are of crucial importance to restore the spin - wave amplitude during the course of their propagation as well as to increase the spin - wave intensity to allow for a convenient detection at the output of the network and interconnects to conventional electronics . in the context of a phase - conserving , mode- and frequency - selective amplification of traveling spin waves ,
the process of parallel pumping has , so far , proven superior to other approaches of spin - wave amplification , such as the reduction of the spin - wave damping by spin transfer torque ( stt)@xcite .
parallel pumping , or parallel parametric amplification , results from the interaction of selected spin waves with a sufficiently large dynamic effective magnetic field , which acts in parallel to the static magnetization and features two times the frequency of these spin waves .
any dynamic effective field modulation with proper frequency can act as source of parametric excitation .
classically , this effective field , referred to as the pumping field , has been provided by a microwave field . in this case , the parametric process creates pairs of magnons by the splitting of microwave photons .
alternatively , the pumping field can be provided by a variation of any magnetic control parameter , such as , e. g. , a dynamic variation of the magnetic anisotropy or an incident spin current .
one general key feature of the process of parallel pumping is the creation of pairs of magnons which are mutually connected in their phase@xcite . due to this ,
the application of parallel pumping goes way beyond a mere amplification of spin waves .
since the relative phase between the created waves can be externally controlled , this allows for a phase - sensitive amplification , which can discriminate logic values encoded in the spin - wave phase@xcite . in macroscopic systems ,
the practical application of parallel pumping has suffered from the fact that the preferably amplified mode , referred to as the dominant mode , is generally not the signal carrying fundamental waveguide mode@xcite .
the latter , which can be conveniently excited externally , features the largest group velocity and net magnetic moment of all modes available in a spin - wave waveguide .
hence , it is the most interesting mode for magnonic applications .
however , an efficient amplification of this mode could only be achieved under specific conditions .
recently , it was demonstrated that in microstructured magnonic waveguides and elliptical elements , the dominant mode can be tuned to be the fundamental mode@xcite . in this review
, we will discuss this phenomenon as well as its consequences on the spin - wave amplification in these microscopic systems .
we briefly review linear spin waves in thin films and microstructured waveguides and introduce the basic equations to describe the effect of parallel parametric amplification on the spin - wave dynamics in section [ theoback ] .
consequently , in section [ pgen ] , we present examples of spin - wave generation by parallel parametric amplification of thermal spin waves in several geometries , showing the versatility and the advantages of this technique in comparison to other excitation schemes . in section [ pamp ]
, we review the amplification of traveling , externally excited spin - wave packets in magnonic waveguides , addressing the potential and the limitations of this amplification scheme .
finally , we provide an overview over the recent applications and novel concepts of spin - wave manipulation and characterization by means of parallel parametric amplification in section [ nov ] , and , ultimately , a brief conclusion of the review is provided in section [ concl ] .
we begin with the presentation of the theory of linear spin waves .
it describes the unperturbed eigenstates of the small - amplitude spin - wave excitation in a magnetic material .
the spatial confinement in magnetic microstructures leads to the formation of quantized , laterally standing spin - wave modes in these systems .
however , already in a thin , laterally extended ferromagnetic film , the finite thickness gives rise to quantized standing spin - wave modes across its thickness , referred to as perpendicular standing spin - wave ( pssw ) modes@xcite .
in very thin films , experiments involving propagating spin waves and their theoretical description are restricted to the lowest , fundamental thickness mode , which features a quasi - uniform magnetization distribution across the thickness of the film .
this is due to the fact that higher modes , which feature a non - uniform magnetization distribution across the thickness , exhibit a large exchange energy and , thus , are found at frequencies beyond the frequency - range typically studied in these experiments .
for instance , for a @xmath2 thick film from the magnetic material ni@xmath3fe@xmath4 ( permalloy ) featuring standard parameters , exchange constant @xmath5 , gyromagnetic ratio @xmath6 . ] , the first standing spin wave modes lies at frequencies above @xmath7 and , thus , more than @xmath8 above the frequency of the uniform ferromagnetic resonance at @xmath9 for an effective magnetic field of @xmath10 within the waveguide . for the uniform mode in an isotropic , in - plane magnetized film ,
the spin - wave dispersion relation for spin waves with @xmath11 is of the compact form@xcite @xmath12 here , @xmath13 is the magnitude of the spin - wave wavevector @xmath14 in the film plane , @xmath15 with the effective field @xmath16 inside the waveguide , @xmath17 with the saturation magnetization @xmath18 , @xmath19 with the exchange constant @xmath20 , @xmath21 denotes the angle between the spin - wave wavevector @xmath14 and the magnetization direction @xmath22 and @xmath23 .
if the thin film is structured into a spin - wave waveguide , i.e. , a thin rectangular slab , an additional quantization of the wavevector @xmath14 is introduced across the width of the waveguide .
this gives rise to the formation of waveguide modes .
in the geometries reported in this review , these modes are described reasonably well by a continuous wavevector @xmath24 along the waveguide long axis and an approximately cosinusoidal profile across the width of the waveguide .
the latter is sketched in fig .
[ fig : theo - sw1 ] , where the amplitude distribution of the first four waveguide modes is schematically shown .
an approximate derivation of the spin - wave dispersion of the waveguide modes can be obtained by using eq .
[ eq : disp - tf ] with a fixed wavevector component @xmath25 along the short axis of the waveguide , which results in a continuous variation of the angle @xmath21 as the wavevector component @xmath24 along the propagation direction is changed . here
, @xmath26 denotes the mode number of the waveguide modes and @xmath27 the effective width of the waveguide .
this effective width can deviate from the geometrical width of the waveguide due to the boundary conditions of the static and the dynamic magnetization at the waveguide edges@xcite .
this approach has been successfully applied in , e.g. , refs .
@xcite to obtain the approximate spin - wave frequencies . in the two commonly studied geometries , namely in the transversely magnetized waveguide @xmath28
( often also referred to as _ damon - eshbach_-geometry ) and the longitudinally magnetized waveguide @xmath29 ( _ backward - volume_-geometry ) , the angle @xmath21 and the resulting dispersion relations are then calculated using @xmath30 and @xmath31 , respectively . figure [ fig : theo - sw1 ] ( b ) shows the dispersion of the first four waveguide modes @xmath32 to @xmath33 in a @xmath34 wide , @xmath2 thick transversely magnetized ni@xmath3fe@xmath4 waveguide for an effective field of @xmath35 in the waveguide .
figure [ fig : theo - sw1 ] ( c ) shows the resulting dispersion in a @xmath36 wide , @xmath2 thick longitudinally magnetized waveguide .
both effective widths correspond to a geometrical width of @xmath37 and the same material parameters of ni@xmath3fe@xmath4 as stated above have been used .
the process of parallel parametric amplification or parallel pumping can be realized by the interaction of a microwave pumping field @xmath39 , which is applied along the static magnetization , with the longitudinal component of the dynamic magnetization in a magnetic material . in an in - plane magnetized magnetic thin film ,
this dynamic longitudinal component is non - zero as a consequence of the shape anisotropy , which results in an elliptic magnetization trajectory , and the fact that the length of the magnetization vector is preserved on the timescale of the magnetization precession@xcite ( cf .
[ fig : theo - pamp1 ] ( a ) ) . in the particle picture ,
the process of parallel pumping is described by the splitting of microwave photons with frequency @xmath40 into pairs of magnons with frequency @xmath41 and @xmath42@xcite under frequency and momentum conservation ( cf .
[ fig : theo - pamp1 ] ( b ) ) .
hereby , the decaying microwave photons provided by the pumping field constitute an energy flow into the magnon system which counteracts the magnon damping .
if the inserted energy per unit time exceeds the loss rate @xmath43 of the pumped magnon group , the intensity of the magnon group rises exponentially in time .
this compensation is referred to as the parametric instability , which , consequently , is a threshold process which requires a minimum threshold field @xmath44 . in the following
, a short summary of the main findings of the basic @xmath38-theory is given@xcite .
this theory provides a description of the thresholds of the parametric instability as well as of the temporal evolution of the spin - wave intensity when the pumping field is applied .
moreover , it has successfully provided a description of the steady state reached in macroscopic systems under parallel pumping if the applied pumping field @xmath39 does not exceed the threshold field @xmath44 significantly . assuming a simple harmonic time dependence of the pumping field @xmath45 with frequency @xmath46 without a spatial confinement , and assuming that this field is coaligned with the static magnetization and an external bias field @xmath47 ( i.e. , @xmath48 , the hamiltonian associated with the pumping process is given by @xcite . ] .
: @xmath49 with the coupling parameter @xmath50 , which describes the coupling of the pumping field @xmath39 to the spin - wave mode @xmath51 .
the sum is taken over all modes @xmath14 interacting with the pumping field .
the fact that the pumping mainly interacts with wave pairs with @xmath52 is a consequence of linear momentum conservation : @xmath53 here , @xmath54 denotes the wavevector of the pumping field , which , in the absence of a strong spatial confinement of the pumping field , is approximately zero due to the long wavelength of the microwave .
in addition to the interaction of the magnons with the pumping field , the interaction of the created magnons with each other has a large influence on the spin - wave dynamics under parallel pumping .
this interaction is described by the four - wave interaction between the generated spin - wave pairs . in many cases
, this interaction provides the main mechanism for a saturation of the amplification . in general , it is described by the hamiltonian @xmath55 where the matrix element @xmath56 describes the four wave interaction between a pair of waves with @xmath57 and @xmath58 with a pair of waves featuring @xmath59 and @xmath60 . by solving the hamilton equations including the hamiltonian @xmath61 , which describes the spin - wave interaction with the pumping field , together with the common hamiltonian @xmath62 describing the linear spin - wave dynamics , one obtains the basic equations of the @xmath38-theory@xcite : @xmath63,\\ \frac{\partial \psi(\mathbf{k},t)}{\partial t } & = \omega_\mathrm{nl}(\mathbf{k})-\frac{\omega_\mathrm{p}}{2}+\mathrm{re}\left\{p^*(\mathbf{k},t)\mathrm{e}^{\mathrm{i}\psi(\mathbf{k},t)}\right\},\label{eq : s - psi } \\
p(\mathbf{k},t ) & = \mu_0h_\mathrm{p } v(\mathbf{k})+\sum_{\mathbf{k}_1}s(\mathbf{k},\mathbf{k}_1)n(\mathbf{k}_1,t)\mathrm{e}^{-\mathrm{i}\psi(\mathbf{k},t)},\label{eq : s - p } \\
\omega_\mathrm{nl } & = \omega(\mathbf{k})+\sum_{\mathbf{k}_1}t(\mathbf{k},\mathbf{k}_1)n(\mathbf{k}_1,t).\label{eq : s - nl}\end{aligned}\ ] ] here , @xmath64 denotes the density of the magnon group @xmath51 , @xmath65 represents the phase of the wave pairs with @xmath52 , i.e. @xmath66 , @xmath67 is the complete pumping . the latter is constituted of the energy flux by the external pumping @xmath68 and the energy flux by the _ internal pumping _ , which describes the energy exchange between the wave pairs , and which is given by the sum in eq .
[ eq : s - p ] .
the fourth equation describes the nonlinear frequency @xmath69 , which can deviate from the linear frequency @xmath70 of the created spin waves due to the creation of spin - wave pairs during the pumping . in the following ,
this shift is neglected since , in the discussed experiments in microstructures , only a low supercriticality @xmath71 is treated and , consequently , the nonlinear frequency shift is low .
the summation is performed over all wave - pairs interacting with the pumping field .
the parameter @xmath72 , which is the name - giving parameter of the @xmath38-theory , describes the four - magnon coupling between waves with @xmath52 , i.e. , @xmath73 .
it should be noted that the phase @xmath65 of the wave pairs is well defined by eq .
[ eq : s - psi ] , while the individual phases @xmath74 and @xmath75 can be selected spontaneously from the thermal spin - wave background or predefined by an external excitation .
initially , i.e. , before the parametric amplification starts , the system can be treated as linear and the four - wave interaction can be neglected , i.e. , @xmath76 . under these conditions , eq .
[ eq : s - n ] reduces to a simple rate equation , where the pumping is acting as a source of magnons which are dissipated due to their interaction with the electron and phonon system . from this equation
the threshold condition can be obtained@xcite : @xmath77 here , @xmath78 accounts for the incremental change of the magnon density @xmath79 in the form of @xmath80 which holds on short timescales for powers in the vicinity of the threshold ( i.e. for small @xmath81 ) .
@xmath82 denotes the starting density of the waves .
thus , the threshold field @xmath44 is given by @xmath83 in general , the last term in the numerator becomes minimum for @xmath84 and , consequently , the pumping is degenerate , unless large variations of @xmath50 or @xmath85 around @xmath84 make a non - degenerate splitting favorable .
thus , in the general case , the magnon pairs with the minimal threshold feature half the frequency of the pumping field reads @xmath86 and similar changes also apply to the other equations stated above . ] . in this case , the threshold field is simply given by @xmath87 therefore , the mode with the lowest relaxation frequency @xmath85 and the highest coupling @xmath50 features the lowest threshold .
this mode is referred to as the critical or the dominant mode . from the spin - wave dispersion ,
the relaxation frequency @xmath43 caused by the gilbert damping , which is dominating in the case of metallic microstructures , can be derived via@xcite @xmath88 and the coupling parameter @xmath89 follows to@xcite @xmath90,\end{aligned}\ ] ] which is basically a measure for the ellipticity of the magnetization trajectory .
since the ellipticity is maximal for the uniform precession and decreases with increasing wavevector , the coupling decreases as the mode - number and the wavevector component in the propagation direction increase .
the resulting threshold fields using the parameters stated above for the calculation of the spin - wave dispersion relations and a gilbert damping parameter of @xmath91 are presented in figs .
[ fig : theo - pamp2 ] ( a ) and [ fig : theo - pamp2 ] ( b ) for the first four waveguide modes of a transversely and a longitudinally magnetized ni@xmath3fe@xmath4 waveguide , respectively . as can be seen from the panels , the threshold fields are fairly similar in magnitude and , generally , the lowest available waveguide mode @xmath26 features the lowest threshold field ( cf .
inset in fig .
[ fig : theo - pamp2 ] ( a ) for the transversely magnetized waveguide ) .
hence , it is expected that the fundamental waveguide mode @xmath92 is the dominant mode in these scenarios .
this is a unique feature of microstructured elements and a major difference to the previously studied macroscopic samples ( see , e.g. , refs .
@xcite ) . since this mode is the most interesting mode for magnonics as it features the largest net magnetic moment and the largest group velocity among all waveguide modes ( compare slopes of dispersion in fig .
[ fig : theo - sw1 ] ) , this renders parallel pumping particularly interesting in microstructured spin - wave conduits .
an alternative way to look at the threshold is the energy balance of the spin - wave system . for this , one should compare the energy flux @xmath93 into the spin - wave system by the external pumping field with the energy flux @xmath94 , which describes the losses of this spin - wave mode per unit - time .
initially , i.e. , before wave pairs are generated , for a pair @xmath52 with the assumptions @xmath95 and @xmath96 , these are given by@xcite : @xmath97 with the pumping phase @xmath98 , which is given by @xmath99 at the threshold .
again , @xmath66 denotes the phase of the spin - wave pair . from this
, it follows directly there is no temporal dynamic of the phase difference . ]
that the energy flux into the mode @xmath100 is maximal if @xmath101 i.e. , the phase between the wave pair and the pumping field is shifted by @xmath102 . under this condition
, the threshold , which is defined by @xmath103 , is reached if : @xmath104 .
another very important feature can be seen from eq .
[ eq : w+ ] : the energy flux into the mode @xmath14 is proportional to its density @xmath105 .
thus , the mode @xmath14 needs to be initially populated in order to be amplified . at room temperature ,
this initial population is generally provided by a homogeneous thermal excitation of practically all spin - wave modes @xmath14 in the investigated frequency range , since @xmath106 .
alternatively , a ( in general significantly larger ) population of spin waves at a desired frequency can be realized by an external excitation with driving frequency @xmath107 . as depicted in fig .
[ fig : theo - pamp1 ] ( b ) , due to the negligible linear momentum of the microwave photons @xmath108 , parallel pumping a priori leads to the formation of counter - propagating waves with @xmath109 in order to fulfill momentum conservation . in the context of parametric amplification of
externally excited , signal carrying waves , these two waves are commonly referred to as the signal ( @xmath110 ) and idler ( @xmath111 ) wave , which share a mutual phase relation according to eq .
[ eq : phaserel ] . hereby , the signal wave is the wave running in parallel to the input waves whereas the idler wave runs in the opposite direction .
this process is commonly referred to as the adiabatic regime of parallel pumping . however , by providing a spatial confinement to the pumping field , it is possible to obtain a pumping source with an effective wavevector spectrum @xmath112 .
in this so called nonadiabatic regime of parallel pumping@xcite , the finite momentum provided by the localization allows for a splitting of microwave photons with @xmath113 into pairs of co - propagating waves following @xmath114 several mechanisms can limit the amplitude of the amplified spin waves .
the most important ones are the dephasing mechanism which has been originally introduced by zhakarov , lvov and starobinets and the nonlinear damping of the generated waves .
in particular , the dephasing mechanism has proven to be responsible for most observations made in the material yttrium iron garnet ( yig ) , where the phenomenon of parallel pumping was extensively studied in bulk and thin - film samples ( see , e.g. , refs .
the dephasing mechanism is already incorporated into eqs .
[ eq : s - n]-[eq : s - nl ] and it is based on the fact that the generated waves provide a mutual pumping field by themselves , named the internal pumping can deviate from @xmath115 ( cf .
. [ eq : s - psi ] ) . ] . in a simple picture ,
if a certain population @xmath105 is already established , a new magnon pair @xmath52 which is supposed to be generated feels the influence of the external pumping field as well as the one of the internal pumping field , which is proportional to @xmath100 . due to the fact that the phase of the already generated wave - pairs differs from the phase of the external pumping initially by @xmath102 , the newly generated magnons now have to adjust their phase to the phase of the effective pumping field . by doing so , the energy flux @xmath93 from the pumping field into the magnon group decreases until , at some point , at a pumping field above the threshold field @xmath44 , @xmath93 matches the flux @xmath94 and the spin - wave density @xmath105 saturates . a numerical demonstration of the influence of this process under the assumption of the presence of two modes @xmath57 and @xmath58
can , for example , be found in refs .
@xcite . in particular , these works address the coexistence of two modes in the system of which one is the signal mode @xmath116 and the other one is the dominant mode @xmath117 . here
, the dominant mode is supposed to be the thermally excited exchange wave featuring the lowest instability threshold while the signal mode is excited initially at a much larger level , i.e. , @xmath118 . in return
, it features a larger damping @xmath119 .
if now a pumping field , which is sufficiently large to exceed the threshold of the signal mode , is applied , the signal group is initially favored due to its higher starting population . however , after a certain period , the dominant mode will end up with a larger population due to its larger intensity increment @xmath120 per unit time ( cf .
[ eq : thresh1 ] ) .
once the amplitudes of the two waves have increased beyond a certain critical level , the dephasing mechanism limits the magnon density in the steady state .
in particular , in the steady state the internal pumping field has decreased the total pumping field to the point where @xmath121 , where the last inequality is based on the fact of the dominant group is largest and/or @xmath94 is smallest , since it features the lowest @xmath43 and/or the largest @xmath89 . ] that @xmath119 .
thus , the losses of the signal group are smaller than the inserted energy and it will start to decay in time .
hence , on the long run , only the dominant mode survives in the steady state , which is a general feature of this mode that also holds if more than two modes are investigated .
this concept puts important restrictions on the amplification of traveling spin waves by the mechanism of parametric amplification , especially if the signal mode is not the dominant mode .
the second important limiting mechanism of the spin - wave amplitude is the nonlinear damping of the spin waves themselves , i.e. , a change of the spin - wave relaxation frequency @xmath43 due to magnon - magnon interaction .
this can be included by an additional term @xmath122 in the rate equations@xcite .
while in most experiments performed in bulk materials and thin films this is only a side effect which appears at a rather large supercriticality @xmath123 , the nonlinear damping can play a more decisive role in microstructures such as transversely magnetized waveguides .
in particular , this can be the case because , depending on the geometry , the waveguide modes in these waveguides can exhibit very similar relaxation frequencies @xmath43 and coupling parameters @xmath50 and , thus , feature a very similar threshold ( cf .
[ fig : theo - pamp2 ] ) .
hence , the usual assumption that only one spin - wave mode is excited by the parametric amplification is not valid anymore . even at very low supercriticalities a whole set of waveguide modes
is potentially amplified . before the aforementioned saturation due to the phase - mechanism sets in , where only the dominant mode is maintained by the action of the pumping field
, this increases the amplitude of all of these modes . for a four - magnon process , the coupling between the involved modes is proportional to the product of their amplitudes ( cf .
, for example , @xcite ) and the threshold for such processes is inversely proportional to the spin - wave damping .
thus , the fact that the damping of several modes @xmath124 , @xmath125 , ... is compensated by the pumping field also increases the nonlinear damping of each of these modes due to the four - magnon interaction with all other modes which are excited in this transition phase .
consequently , it is expected that nonlinear damping can also have a significant influence on the amplitude limitation in microstructures .
it should be noted that both of these limitation mechanisms imply that the intensity level at which the parametrically amplified spin waves saturate depends on the supercriticality .
the higher the applied pumping power above the threshold , the higher the saturation level which can be maintained in the steady state under the act of parallel pumping .
thus , to achieve a large spin - wave density of the amplified waves , a large supercriticality is generally favorable . as mentioned above
, the fact that in the investigated waveguides generally more than one mode is excited for very small supercriticalities already limits the simple application of the @xmath38-theory described above .
moreover , the @xmath38-theory has been derived explicitly for spatially homogeneous samples and pumping fields .
therefore , it can not take into account the finite sizes of the analyzed waveguides or the resulting lateral quantization effects .
the prediction of the thresholds , however , is expected to be correct since it comes from the very general energy balance which can - at least via the spin - wave dispersion - account for the effects arising from the quantization .
in fact , it has been demonstrated in @xcite and @xcite that the general hamilton approach sketched above can describe the thresholds and the critical modes for spin - wave instabilities in waveguides reasonably well . since the treatment of the parallel - pumping instability is very similar to the treatment of these spin - wave interactions , one can expect that the prediction of the thresholds is valid . still , the predictions of the @xmath38-theory for the steady state and the limitation mechanisms have to be considered rather as a tool for a qualitative understanding of the occurring phenomena .
an additional side effect of the quantization should be noted at this point : only a limited number of modes exists at a given frequency @xmath70 as a consequence of the lateral quantization and due to the low film thicknesses .
thus , it is possible to realize the case in which the fundamental waveguide mode constitutes the dominant mode as well as the signal mode .
as mentioned above , this is hardly ever the case in macroscopic samples . in these samples
typically some exchange mode features the lowest threshold@xcite .
thus , after a certain time , this mode diminishes the intensity of the signal wave and suppresses the information stored in this mode . in microstructures
, this implies that , at least according to the phase mechanism , the steady - state can be reached for the signal waves and the information carried by these waves can , this way , be stored in the spin - wave system by the pumping field .
parametric generation is referred to as the amplification of thermal spin waves@xcite .
spin waves are always thermally excited at room temperature due to their @xmath0 excitation energy .
the pumping field couples to the longitudinal dynamic component provided by these thermal waves and , if the instability threshold is overcome , they can be amplified . due to the fact that the parallel pumping process always allows for the creation of counter - propagating waves with a vanishing sum of wavevectors , this gives access to the excitation of spin waves with large wavevectors ,
even with spatially extended pumping sources@xcite . this way , parametric generation has enabled the excitation of short - wavelength spin waves in macroscopic systems , an important prerequisite for the observation of , for instance , the bose einstein condensation of magnons@xcite .
moreover , unlike conventional antennae@xcite , it is not primarily sensitive to the net magnetic moment of the excited modes and , thus , their symmetry .
in contrast , the wavevector dependence of the threshold fields is mainly determined by the above mentioned characteristics and , thus , a function of the ellipticity of the mode pairs and their relaxation frequency@xcite .
hence , parametric generation is a powerful tool to study the mode spectra in magnetic systems . due to the fact that on the long run the amplification of the dominant mode suppresses all other modes , a knowledge of the dominant mode spectrum of the investigated structures is mandatory in order to design the optimum amplification of externally excited spin waves . in this section
, we will discuss the dominant mode spectra in the most commonly studied geometries in microstructures , namely in microscopic elliptical elements and longitudinally as well as transversely magnetized spin - wave waveguides .
we will show that in all these systems a favored generation of the fundamental mode can be achieved .
microstructured elliptical elements are widely used as basic elements in prototypes of magnetic random access memories@xcite and spin - transfer torque driven magnetic nano - oscillators@xcite .
hence , an understanding of their fundamental spin - wave spectrum , including odd as well as even modes , its relaxation characteristics and its coupling to an rf magnetic field are of vital importance for the optimization of these devices . in these microscopic elliptical elements ,
the boundaries give rise to a spectrum of discrete resonance frequencies associated with laterally standing spin - wave modes@xcite .
the parametric generation of these modes was studied in ref .
@xcite by means of microfocussed brillouin light scattering ( bls)@xcite as well as in ref .
@xcite by means of ferromagnetic resonance force microscopy@xcite . in both of these works ,
magnetic elliptical elements have been fabricated on top of a micro - strip which provides the rf pumping field in the form of the oersted field created by a microwave current in the strip .
the configuration studied in ref .
@xcite is depicted in fig .
[ fig : ulrichs ] ( a ) , where a scanning electron microscopy ( sem ) micrograph of the investigated structure is shown . in this experiment , as well as in most other experiments regarding parametric generation and amplification in microscopic systems
, the pumping field has been applied in pulses .
this minimizes sample heating and the time the parametric amplifier is saturated .
moreover , it allows for a time - resolved investigation of the action of the pumping field on the spin - wave dynamics . as has been demonstrated in refs . @xcite and
@xcite , the discrete spectrum of standing modes give rise to the observation of distinct resonance peaks in the spectra of parametrically generated spin waves , which feature similar , yet slightly different , threshold powers .
this can be comprehended from fig .
[ fig : ulrichs ] ( b ) , where the bls intensity is shown as a function of one - half of the applied pumping frequency , which corresponds to the detection frequency of the parametrically generated spin waves . for these measurements ,
the bls laser has been positioned in the center of the elliptical element , which is magnetized along its short axis by an external bias field , and the frequency of the pumping field has been varied for various fixed peak powers of the microwave pumping field . at first , only one peak is observed ( @xmath126 , frequency @xmath127 ) .
an increase of the pumping power leads first to an increased number of visible modes at higher frequencies ( @xmath128 . at larger powers ,
nonlinear interactions lead to a broad excitation spectrum .
an analysis of the spatial intensity distribution associated with the observed peaks at low powers yields the profiles of the corresponding modes .
figure [ fig : ulrichs ] ( c ) shows space - resolved bls maps of the mode profiles associated with the peaks @xmath127 , @xmath129 and @xmath130 .
hereby , it should be noted that the bls intensity is directly proportional to the spin - wave intensity at a given position@xcite . ] .
the mode associated with the excitation with the lowest threshold is the fundamental mode @xmath92 , featuring one intensity maximum in the center of the dot
. since the higher modes feature an increasing number of nodes , their dynamic dipolar stray field and , consequently , their ellipticity of precession decreases .
as a consequence , the higher modes feature a reduced coupling to the microwave pumping field .
since they additionally feature a higher frequency , they exhibit a larger relaxation rate ( see eq .
[ eq : taucomp ] ) .
hence , from eq .
[ eq : threshold ] it is expected that the first mode has the lowest parametric generation threshold , as it is visible in fig . [
fig : ulrichs ] ( b ) .
the peak that appears at the frequency @xmath131 is associated with the parametric generation of an edge mode@xcite . due to the fact that the changes in the relaxation frequency and in the coupling strengths are not very pronounced , the differences in the thresholds are rather subtle .
hence , parametric generation allows to excite modes of odd and even symmetry as well as edge modes with a comparable excitation strength and is well suited for the study of the entire mode spectrum in such microstructured elements .
this makes this method an interesting alternative to conventional excitation schemes , which use a dynamic oersted field perpendicular to the static magnetization@xcite , since these can not access even modes and whose excitation efficiency for higher modes decreases with @xmath132 due to the decreasing magnetic moment of the modes with increasing mode number .
in addition to its application as a tool to study or amplify the fundamental excitations in elliptical elements , the technique of parallel pumping is also well suited for the generation and amplification of spin waves in waveguides . as has been demonstrated in ref .
@xcite , a convenient way to realize a local spin - wave generation in a waveguide magnetized along its long axis can be realized by using a simple micro - strip - antenna . in a geometry like the one depicted in fig .
[ fig : locbv1 ] ( a ) , where such an antenna is patterned on top of such a waveguide , the in - plane component of the ac oersted fields provides the microwave pumping field interacting with the longitudinal component of the dynamic magnetization .
similar to the application to elliptical elements , this allows for the generation of odd and even spin - wave modes in the waveguide .
in contrast to the out - of - plane component of the antenna field , which is the dominant conventional means of spin - wave excitation by a direct torque in this geometry , the localization of the in - plane component results in a local spin - wave excitation by parametric generation . moreover , due to the creation of spin - wave pairs with @xmath133 , such an excitation source can practically excite arbitrarily short wavelengths , independent of its width . for a better comparison to the experiment in ref .
@xcite , in which the pumping frequency is fixed while the externally applied magnetic field is varied , fig .
[ fig : locbv1 ] ( b ) shows the spin - wave wavevector of the first four waveguide modes @xmath92 to @xmath134 as a function of the externally applied bias field for a fixed spin - wave frequency of @xmath135 . for the calculations , a @xmath136 wide , @xmath137 thick waveguide made from ni@xmath3fe@xmath4 has been assumed , as has been investigated in ref .
they have been performed assuming a saturation magnetization of @xmath138 , an exchange constant of @xmath5 and an effective waveguide width of @xmath139@xcite . in the figure
, the solid lines correspond to the dipolar branch of the dispersion relation and the dashed lines to the exchange branch .
it is important to note , that for each field , the lowest available waveguide mode features the lowest available wavevector in the dipolar branch . hence ,
it is supposed to feature the largest ellipticity and , consequently , the largest coupling to the pumping field . figure [ fig : locbv1 ] ( c ) shows the experimentally determined thresholds of parametric generation together with the theoretically expected threshold powers . to achieve parametric generation , @xmath140
long pumping pulses with a carrier frequency of @xmath141 have been applied to the antenna .
the theoretical threshold powers have been calculated following eqs .
[ eq : taucomp ] and [ eq : v ] by using eq .
[ eq : threshold ] and @xmath142 , with the same parameters used to derive the spin - wave dispersion relation in fig .
[ fig : locbv1 ] ( b ) and a gilbert damping parameter of @xmath143 .
the calculated thresholds have been normalized to their minimum value at @xmath144 for the mode @xmath92 , corresponding to @xmath145 .
the experimental threshold values have been normalized to the nominal peak power @xmath146 at @xmath147 , corresponding to @xmath148 for the mode @xmath92 .
the shading represents the observed dominant spin - wave mode , for which exemplary mode profiles are shown in fig .
[ fig : locbv1 ] ( d ) . as can be seen from fig .
[ fig : locbv1 ] c ) , the modes @xmath92 to @xmath134 can be parametrically generated with a similar threshold .
in agreement with the theoretical prediction , the experimentally determined threshold is approximately constant within the experimental resolution in a large field range , while it increases at large fields when no resonant excitation is possible anymore .
however , there is also a clear discrepancy from the simple analytical theory described in section [ theoback ] at low fields .
this less accurate description is likely mediated by the fact that in this field - range the assumption @xmath149 , i.e. , that the dynamic fields are much smaller than the static effective fields , is not satisfied anymore .
nevertheless , the analytical formalism is able to reproduce the observed dominant mode , as is evident in the inset of fig .
[ fig : locbv1 ] ( c ) , where the threshold powers are shown with an adequate scale .
as expected , the lowest available waveguide mode features the lowest expected threshold , a finding also predicted in ref . @xcite . as can be comprehended from comparing figs .
[ fig : locbv1 ] ( b ) and ( c ) , the changes of the dominant mode are associated with the respective origin of the dipolar dispersion relations where @xmath150 .
the resonance field and frequency of these modes as well as their frequency - splitting can be tuned by changing the geometric extents of the spin - wave waveguide .
hence , parametric generation is a viable means to excite propagating spin waves in longitudinally magnetized waveguides , where the dominant mode can be tuned to be the desired odd or even waveguide mode by adjusting the geometry and the frequency and field combination .
it should be noted that the jumps from one waveguide mode to another are analogous to the transitions between the different standing modes across the film thickness in thin , but not ultra - thin , films ( cf . ,
e.g. , refs .
@xcite ) .
a particularly interesting application of this excitation geometry is the study of waveguides in their remanent state , i.e. , in the absence of external bias fields .
this paves the way for magnonic applications without the need for an applied magnetic field and allows , for instance , to study the interaction of spin waves with domain walls in the remanent state@xcite . as mentioned above , by varying the dimensions of the waveguide or the pumping frequency , the fundamental waveguide mode @xmath92 can be shifted to be the dominant mode at zero field .
this is demonstrated in fig .
[ fig : locbv2 ] ( a ) , where the calculated spin - wave dispersion relations of the first three waveguide modes in a @xmath151 wide and @xmath137 thick ni@xmath3fe@xmath4 spin - wave waveguide are shown in the absence of an external bias field ( material parameters as stated above ) . due to the small width of the waveguide ,
the splitting between the adjacent waveguide modes is fairly large and , consequently , each mode exists basically in a different frequency range .
thus , by changing the working frequency , the desired mode can be chosen .
figure [ fig : locbv2 ]
( b ) shows the parametric excitation spectrum as a function of frequency in the absence of an external bias field in a waveguide with these dimensions for different pumping powers .
the parametric generation has been performed by means of a @xmath152 wide micro - strip antenna .
if the threshold is exceeded , spin waves are excited in a wide frequency and , thus , wavevector range .
hereby , the excitation with the lowest threshold is associated with the generation of the fundamental waveguide mode @xmath92 . according to the spin - wave dispersion ,
the smallest excited wavelength in this scenario corresponds to a spin - wave wavelength of ca .
this wavelength is below the linear excitation range of the micro - strip - antenna , which is limited to wavelengths larger than the antenna width .
this demonstrates the feasibility of parametric generation to excite a large wavevector range .
figure [ fig : locbv2 ] ( c ) shows the exponential spin - wave decay lengths extracted from the space - resolved bls intensity scanning away from the antenna@xcite .
the experimentally determined decay lengths are compared to the theoretically estimated decay length @xmath154 derived from eqs .
[ eq : disp - tf ] and [ eq : taucomp ] , which is represented by the red solid line .
as can be seen from the figure , there is a good agreement between the observed and expected spatial decay of the waves emitted from the antenna , underlining nicely the propagating character of the excitation . for larger powers , the threshold for the parametric generation of the mode @xmath155 is overcome and , consequently , the generation of this mode is observed for frequencies @xmath156 in a large wavevector range as well . due to the larger group velocity@xcite and the large sensitivity of the spin - wave spectrum for the interface conditions@xcite , spin waves in transversely magnetized waveguides are of great interest for the realization of magnonic devices .
this geometry , commonly also referred to as the damon - eshbach geometry , is the geometry in which most experiments in microstructured spin - wave waveguides have been performed@xcite .
figure [ fig : gende1 ] ( a ) shows the calculated spin - wave wavevector as a function of the effective magnetic field in a @xmath157 wide , @xmath158 thick ni@xmath3fe@xmath4 spin - wave waveguide for the first 10 waveguide modes for a fixed frequency of @xmath159 . as can be seen from the figure , the separation between these modes is much less pronounced than in a longitudinally magnetized waveguide .
nevertheless , the lowest available waveguide mode features the lowest threshold of all modes , as can be comprehended from fig .
[ fig : gende1 ] ( b ) , where the corresponding calculated threshold fields are shown assuming the same parameters and a gilbert damping parameter of @xmath91 .
hence , also in transversely magnetized waveguides , the fundamental waveguide mode can be tuned to be the dominantly amplified spin - wave mode .
this has been experimentally verified in ref .
@xcite by a study of the parametric generation in such a ni@xmath3fe@xmath4 waveguide . using the sample layout depicted in fig .
[ fig : gende2 ] ( a ) , the oersted field created by a microwave current flowing through the micro - strip underneath the transversely magnetized ni@xmath3fe@xmath4 waveguide provides the pumping field interacting with the dynamic magnetization .
this way , a homogeneous parametric generation along the waveguide could be achieved , similar to the homogeneous parametric generation in the microscopic elliptical elements discussed in section [ pgen - ell ] .
figure [ fig : gende2 ] ( b ) shows the color - coded bls intensity at @xmath159 as a function of the externally applied bias field ( @xmath160-axis ) and the pumping microwave peak power ( @xmath161-axis ) . in the experiment ,
the microwave pumping has been applied in @xmath162 long pulses with a carrier frequency of @xmath141 . as can be seen , parametric generation can be observed in a wide range of magnetic fields , again with no large variation in the threshold power .
figure [ fig : gende2 ] ( c ) presents the observed profiles of the dominantly generated modes . below an external field of @xmath163 , which can be associated with an effective field of @xmath164 within the waveguide@xcite ,
the fundamental waveguide mode @xmath92 is generated , as is expected from fig .
[ fig : gende1 ] ( b ) . for @xmath163 ,
the second waveguide mode is the lowest available waveguide mode and , thus , dominantly amplified .
a further increase of the field leads to a further increase of the mode number until at magnetic fields exceeding @xmath165 , no more waveguide modes can be resonantly excited .
consequently , at higher fields one observes the parametric generation of edge modes .
thus , also in extended magnonic waveguides , parametric generation gives access to the excitation of odd and even spin - wave modes with finite wavevector .
in particular , the fact that the dominantly generated waveguide mode is the fundamental waveguide mode in a large wavevector range is the basis of the large potential for the application of parallel pumping in microstructures , as will be discussed in the following sections .
however , it should be noted that due to the very small differences in the threshold fields for waveguides with widths in the @xmath166 range , small changes in the spin - wave lifetime due to wavevector dependent damping@xcite or changes of the dispersion , such as avoided crossings due to dipolar interaction between the waveguide modes@xcite , can render the generated mode spectra more complicated .
one of the main applications of parallel pumping of spin waves is the amplification of externally excited , traveling spin - wave packets carrying information in form of their amplitude , phase or both . in the following , these waves
are referred to as signal spin waves .
the technique of parallel pumping has been successfully applied to traveling spin - wave packets in macroscopic spin - wave waveguides made from yig . among others ,
the amplification of solitons@xcite , the cloning and trapping of spin - wave packets@xcite , as well as the storage and recovery in the spin - wave system@xcite have been realized on the macroscopic scale .
however , in these systems , the parametric generation of the dominant spin - wave group , which is generally not the externally excited , signal - carrying spin wave , always leads to a suppression of the signal spin waves after a finite time .
hence , the phase - information carried by the signal spin waves is lost .
the aforementioned fact , that the preferably generated spin - wave mode in microscopic spin - wave waveguides is the fundamental waveguide mode is a new paradigm for the applicability of parallel pumping for the spin - wave amplification . since , consequently , in these systems the externally excited
, signal carrying spin wave coincides with the dominantly generated spin - wave mode , a particularly efficient parametric amplification should be possible .
due to the beneficial properties of parametric amplification , such as frequency- and mode - selectivity and the conservation of the spin - wave phase , it has several advantages over alternative methods , such as the amplification via the stt effect induced by an incident spin current , which lack such selectivity . in the following section ,
the potential and the limits of parallel pumping as a mere means of spin - wave amplification are presented , highlighting the role of the parametric generation , which , at room temperature , is still interfering with the amplification of the externally excited spin waves .
it will be shown that its role is quite different from macroscopic systems : the fact that the same wave as the signal mode is predominantly generated means that its generation is not generally suppressing the signal wave . however , the fact that the thermal spin waves with exactly the same spin - wave frequency and wavevector are amplified gives rise to two problems : 1
. it can lead to an undesired creation of _ noise _ , since - at least for a detection only sensitive to the spin - wave intensity - the generated waves can not be distinguished from the signal waves .
one feature of the parametric amplifier is its saturation after a certain spin - wave intensity is reached within the amplifier ( cf .
section [ theo - ppa - sat ] ) .
thus , as will be discussed in the following , the parametric generation can lead to an undesired _ saturation _ of the amplifier before the traveling signal spin waves arrive .
figure [ fig : pamp - loc1 ] ( a ) shows the layout of a local spin - wave amplifier which has been experimentally investigated in ref .
a spin - wave waveguide has been patterned on top of a pumping micro - strip .
in addition , a micro - strip - antenna has been patterned across the spin - wave waveguide to allow for a local and coherent spin - wave excitation at frequency @xmath167 .
similar to the experimental realization for the parametric generation , a microwave with frequency @xmath168 provides the microwave pumping field for the parametric amplification .
the spin - wave excitation and the parallel pumping field are applied in pulses .
in addition to the aforementioned minimization of heating , this also allows to study the influence of the relative timing of the amplification with respect to the arrival of the spin - wave packet within the amplifier .
this is experimentally controlled by the delay between the arrival of the externally excited spin wave packet and the application of the pumping pulse @xmath169 , where @xmath170 implies that the pumping starts before the spin - wave packet arrives .
the amplifier shown in fig .
[ fig : pamp - loc1 ] ( a ) features a general width of @xmath171 , which is reduced to a width of @xmath172 over a length of @xmath173 at a distance of @xmath174 from the antenna . the larger current density in this region with smaller cross - section provides a locally enhanced microwave oersted field .
the combination of this local enhancement with the threshold character of parametric amplification results in the actual localization of the parametric spin - wave amplifier@xcite . in the experiment in ref .
@xcite , the spin - wave excitation and the parallel pumping have been applied in @xmath175 and @xmath176 long microwave pulses , respectively .
the microwaves at @xmath167 and @xmath177 have been created by two independent microwave sources .
the output of the amplifier has been studied as a function of the power of the pumping as well as the delay @xmath169 between the two pulses .
consequently , the time - integrated spin - wave intensity behind the amplifier , which has been measured by bls microscopy , can be constituted of the traveling spin - wave packet excited by the antenna , a packet of parametrically generated waves and , if the parallel pumping interacts with the traveling waves , a packet of amplified signal spin waves . the spatial evolution of the spin - wave intensity arising form the interplay of an externally excited , propagating spin - wave packet with such a localized parallel pumping field is shown in fig . [ fig : pamp - loc1 ] ( b ) by the black squares .
for the applied moderate pumping power of @xmath178 and for an arrival of the spin - wave packet within the amplifier shortly before the pumping pulse is applied ( @xmath179 ) , a strong spin - wave amplification is observed .
this can be comprehended by a comparison with the dashed blue line , which is the sum of the intensity arising from the antenna excitation only ( red dots ) and the intensity arising from parametric generation only ( green triangles ) .
in contrast to the bls intensity arising from the antenna excitation only , the amplifier is able to restore the spin - wave amplitude to its initial intensity level in the vicinity of the antenna .
this way , it greatly enhances the propagation distance of the traveling waves , which , in the absence of parallel pumping , are found to decay exponentially with a decay length on the order of a few microns until their intensity falls below the experimental noise level of the bls microscope .
the green triangles demonstrate that such a local parametric amplifier can also be used as a source for local parametric generation , an aspect which has been studied in more detail in ref .
@xcite .
figure [ fig : pamp - loc1 ] ( c ) shows the output of the amplifier measured at a distance of @xmath180 from the antenna .
the output has been normalized to the intensity measured at a delay of @xmath181 , i.e. , for a start of the pumping pulse @xmath182 before the traveling spin - wave packet arrives . in this case
there is no temporal overlap between the two pulses , no parametric amplification is observed and the output is simply given by the sum of the parametrically generated spin - wave packet and the incoming spin - wave packet , i.e. , it corresponds to the blue dashed line in fig .
[ fig : pamp - loc1 ] ( b ) .
this way , the output is a measure of the amplification , defined by the ratio of the output of the amplifier for the interaction of the signal spin waves with the pumping field over the output obtained if there is no interaction , i.e. , no temporal overlap .
it should be noted that in the latter case , the spin - wave packet excited at the antenna has already traveled over a distance of @xmath180 . consequently , its intensity has already decayed significantly and , in fact , the main contribution to the measured intensity for a delay of @xmath181 shown in fig .
[ fig : pamp - loc1 ] ( c ) arises from the generated packet .
this can also be comprehended from fig .
[ fig : pamp - loc1 ] ( b ) , since the signal spin - wave intensity in the absence of parametric amplification is already below the noise level of the bls at @xmath183 . as the delay is shifted and the pulses overlap , a pronounced increase of the bls intensity
is observed , which is a fingerprint of the parametrically amplified spin waves . as can be seen from fig .
[ fig : pamp - loc1 ] ( c ) , the strength and the delay - dependence of the amplification depend strongly on the applied peak power of the pumping pulse . for a comparably low pumping power of @xmath184 , which exceeds the parametric instability threshold by @xmath185 ,
the largest amplification by a factor of about @xmath186 is achieved and the amplification is rather symmetric as a function of the delay . as the pumping power
is increased , the amplification reduces and the delay - dependence becomes increasingly asymmetric . at high powers , such as @xmath187 ,
the curve is strongly asymmetric and a notable amplification is only observed , if the spin - wave packet enters the amplifier before the pumping starts ( delay @xmath188 ) .
this delay- and power - dependence are caused by the interaction of the pumping field with the thermal spin waves .
the fact that the overall amplification decreases is due to the fact that the amplifier saturates very fast at large pumping powers far beyond the instability threshold .
consequently , the relative number of parametrically generated waves in comparison to amplified signal waves increases strongly with increasing pumping power .
consequently , the noise , i.e. , the amount of undesired parametrically generated waves , created by the amplifier increases and the overall amplification drops according to the definition given above . to comprehend the observed dependence on the delay at large pumping powers , it is instructive to analyze the bls intensity as a function of time .
figure [ fig : pamp - loc1 ] ( d ) shows the time - resolved bls intensity arising for different delays at a pumping power of @xmath187 .
the shading represents the temporal extents of the spin - wave packet , whose temporal position is not shifted during the experiment . due to the aforementioned strong decay of the traveling spin waves ,
the externally excited spin - wave packet is not visible in the time - resolved measurement if it is not amplified by the pumping field .
the red curve shows the bls intensity arising for a negative delay of @xmath189 , i.e. , for a complete displacement of the spin - wave packet and the pumping pulse in time .
the detected spin - wave intensity is , thus , due to parametric generation only . after a initial fast increase of the bls intensity ,
the intensity saturates quickly due to the nonlinear interaction of the generated spin waves and the pumping field .
the parametrically generated packet is easily visible , due to the strong applied pumping field . in analogy to the time resolved measurements , the observed time - integrated intensity behind the amplifier at large pumping powers is ( cf .
[ fig : pamp - loc1 ] ( c ) ) , thus , as mentioned before , mainly originating from parametric generation for this delay . as can be seen from the green curves , even for a substantial overlap of the spin - wave packet with the pumping pulse , the time - dependent output of the amplifier is unchanged , as well as the time - integrated intensity .
this is due to the fact that the parametric generation has already saturated the amplifier before the traveling signal spin - wave packet arrives .
hence , no amplification of the externally excited packet is observed and the measured time - integrated intensity is still only due to parametric generation . only if the spin - wave packet arrives first ( delay @xmath188 , black and blue curve ) , the shape of the output pulse is changed and an amplification of the spin - wave packet in time takes place .
consequently , the time - integrated output changes and a net amplification is observed .
this highlights the aforementioned two effects of parametric generation in microstructures : it can lead to the creation of undesired background noise and to an unwanted saturation of the amplifier . therefore , in particular for applications where a strong pumping is needed to achieve a fast operation and large output powers
, it is mandatory that the pumping starts after the signal spin waves have arrived at the amplifier .
moreover , short pumping pulses should be used to minimize the time the amplifier is saturated .
in contrast , at low pumping powers , where the increase of the spin - wave intensity with time is slower ( cf .
eq . [ eq : intvicth ] )
and it takes a longer time until saturation is reached , the delay between the signal spin - waves and the pumping becomes less important . in this case
the delay - dependence is not pronounced and a substantial amplification can be achieved , even for negative delays ( cf .
violet upward triangles in fig .
[ fig : pamp - loc1 ] ( c ) ) .
as will be discussed in the following subsection , these findings have important implications for the global amplification of traveling waves . for many applications ,
it is desired to compensate the spin - wave damping in the entire magnonic network .
the technique of parallel parametric amplification is well suitable to achieve this aim .
however , for such a _ global _ parametric amplifier , parametric generation can impose limitations on the overall functionality@xcite , which will be discussed in this subsection .
the layout of an exemplary global parametric amplifier for a transversely magnetized magnonic waveguide is schematically shown in fig .
[ fig : pamp - glob1 ] ( a ) .
similar to the parametric generation in a spin - wave waveguide , a micro - strip under the waveguide provides the microwave pumping field at frequency @xmath168 .
an additional micro - strip - antenna patterned across the waveguide provides the local , coherent spin - wave excitation at frequency @xmath167 .
a direct way to distinguish the amplified signal spin wave from the parametrically generated waves can be realized by a phase - locking of the spin - wave excitation and the microwave pumping .
this can be achieved technically by generating them mutually using a single microwave source , whose output is split and frequency - doubled in one of the microwave paths . by the use of a phase - shifter in one of the microwave paths , the relative phase between the signal spin waves and the microwave pumping field can be shifted . due to the fact that the parallel pumping process creates pairs of counter - propagating waves with a fixed phase relation determined by eq .
[ eq : phaserel ] , the amplification of a coherent wave with a well - defined phase results in a stationary interference pattern with a spatial periodicity of one half of the spin - wave wavelength .
in contrast , the amplification of thermal waves results in the formation of a spatially homogeneous intensity distribution .
this way , the amplification of thermal and signal waves can be distinguished directly by analyzing the stationary intensity distribution .
an example of this is shown in fig .
[ fig : pamp - glob1 ] ( b ) , where the stationary intensity distribution along the waveguide ( measured along the waveguide center - axis ) is shown for only the spin - wave excitation , only parametric generation and for the interplay of the parallel pumping field with the externally excited spin waves .
the spin - wave excitation and the pumping field have been provided by @xmath190 long microwave pulses with a carrier frequencies of @xmath191 and @xmath192 , respectively .
the waveguide made from ni@xmath3fe@xmath4 features a geometrical width of @xmath193 and a thickness of @xmath158 and is magnetized along its short axis by an applied external bias field of @xmath194 .
the microwave pulse have either been applied independently or , in case of the demonstration of parametric amplification , the spin - wave packet has been launched a few nanoseconds before the start of the pumping pulse . as can be seen from the figure ( blue curve ) , the interplay of the pumping field with the traveling spin waves leads to a compensation of the otherwise significant decay of the traveling waves ( cf . violet curve ) and gives rise to a spatial modulation of the spin - wave intensity .
this spatial modulation is the standing wave pattern which results from the interference of the created counter - propagating signal and idler waves with the same frequency @xmath167 but featuring opposite wavevectors @xmath195 .
in contrast , in the absence of a coherent spin - wave excitation , parametric generation leads to a constant intensity level ( cf .
green curve ) , since the phase of the parametrically generated magnons is random . despite the compensation of the spin - wave damping , the intensity modulation in the case of parametric amplification diminishes for distances larger than @xmath196 as well and the intensity drops to the level of the parametric generation .
the reason for this is analog to the delay dependence of the amplification in a local amplifier : for larger distances , the parametric generation has already locally saturated the amplifier before the traveling spin - wave packet arrives . to visualize this , fig .
[ fig : pamp - glob2 ] ( a ) shows the rising edge of the time - resolved bls intensity measured at different distances from the antenna . for these measurements ,
the length of the pumping pulses has been increased to @xmath197 , the external field has been changed to @xmath198 and the microwave frequency @xmath199 has been changed to @xmath200 ( i.e. @xmath201 ) .
this results in a significantly smaller spin - wave wavelength of @xmath202 in comparison to the waves excited in the measurements shown in fig .
[ fig : pamp - glob1 ] .
the long pumping pulses ensure a sufficient interaction time between the spin - wave packet and the pumping pulses . as can be seen from the figure , for distances below @xmath203 ,
the bls intensity is emerging from the noise level at different times .
this is a direct consequence of the amplification of the propagating waves , which need longer times to reach larger distances .
however , for distances beyond @xmath203 , no further displacement is observed .
in fact , the temporal evolution at a distance of @xmath203 is identical with the temporal evolution of the parametric generation only .
thus , for larger distances the amplifier is already saturated prior to the arrival of the propagating waves .
this is the reason for the fact that the stationary interference pattern in fig .
[ fig : pamp - glob1 ] ( b ) is only visible up to a certain point in space . beyond that point in space ,
the amplifier is saturated by parametric generation and no standing interference pattern , which is a fingerprint of the amplification of the coherently excited signal spin waves , is observed .
time - resolved bls measurements like the one shown in fig .
[ fig : pamp - glob2 ] ( a ) can be used to extract the propagation range of the traveling spin - wave packet : the maximum distance that the amplified wave front travels is at a distance in front of the point in space where the generation has already saturated the amplifier and beyond which no further displacement of the spin - wave intensity in time is observed .
the minimum range of the traveling waves is , in return , associated with the last point in space where the displacement of the traveling wave fronts is still notable .
it should be noted that due to this , the range of the parametric amplification waves is not only determined by the influence of the pumping field : also the group velocity of the traveling waves plays a decisive role .
faster spin waves can achieve higher amplified propagation ranges due to the delay - dependence of the amplification .
the propagation range of a @xmath190 long spin - wave packet interacting with such @xmath197 long pumping pulse is shown in fig .
[ fig : pamp - glob2 ] ( b ) as a function of the applied pumping peak power .
hereby , the pumping power is given as the supercritical pumping power with respect to the instability threshold power @xmath204 in db ( spin - wave excitation frequency @xmath200 , @xmath205 ) . as can be seen from the figure , for low supercriticalities ,
very large ranges exceeding @xmath206 can be observed , exceeding more than 50 times the spin - wave wavelength of @xmath207 .
with increasing pumping power the range decreases , due to the increasingly faster saturation of the amplifier by the faster parametric generation .
the two data sets shown in fig . [ fig : pamp - glob2 ] ( b ) differ in the relative timing between the spin - wave excitation and the beginning of the pumping pulse .
for the first data set ( black ) , the pumping starts at the same time than the spin - wave packet is launched whereas for the second data set ( red ) , the spin waves are launched a few nanoseconds before the pumping starts . in the latter scenario
, the spin waves can already travel along the waveguide and reach a finite , stationary amplitude at distances @xmath208 before the pumping is applied at a larger distance from the antenna when the pumping pulse is applied . ] .
consequently , the propagation range of the waves is increased , as can be seen for a supercriticality of @xmath209 by comparing the two series . from the time - resolved measurements ,
it is also possible to evaluate the velocity with which the signal carried by the amplified wave front travels through the spin - wave waveguide and becomes detectable after reaching a certain threshold - intensity . for a measure of this , we define the effective signal velocity , which is given by the inverse of the time @xmath169 by which the rising edges of the amplified wave fronts visible in fig .
[ fig : pamp - glob2 ] ( b ) are displaced with respect to each other over a certain distance @xmath210 , i.e. , @xmath211 .
if a constant intensity is needed for detection , this velocity determines how long it takes until this intensity is reached at a given distance from the spin - wave source .
this velocity is determined by the interplay of the spin - wave group velocity and the amplification rate , which determine how long it takes to build up the needed intensity at a given point in space . for our determination of this effective velocity ,
we extrapolate the time at which the spin - wave intensity has risen to @xmath212 of its saturation value from the rising edge , which is associated with the arrival of the amplified spin - wave packet at the detector , given by the bls laser spot in this experiment .
if this arrival time is plotted as a function of the distance to the antenna , the slope @xmath213 is the inverse of the effective signal velocity defined above .
the corresponding values extracted this way are shown in fig .
[ fig : pamp - glob2 ] ( c ) as a function of the supercritical pumping power .
as can be seen , for a low supercriticality , i.e. , where large ranges are achieved , the effective signal velocity is slow in comparison to the spin - wave group velocity of about @xmath214 . for the first series
, the signal velocity stays always below the group velocity and approaches it with increasing pumping power .
as mentioned above , this is mediated by the fact that the amplified wave - front needs a certain time until it reaches the intensity level needed for detection .
due to the fact that the supercriticality is weak , the incremental change of the spin - wave intensity per unit time is small ( cf .
eq . [ eq : intvicth ] )
and it takes a comparable large time until the waves have been amplified to the detection limit .
this leads to a seemingly slow propagation of the detectable signal .
in contrast , for the second series , where the spin - waves have been launched before the pumping starts , the observed signal velocity is always higher and can even exceed the spin - wave group velocity .
this is a consequence of the stationary spin - wave intensity present in the waveguide due to the earlier launch of the spin - wave packet . for large supercriticalities
the rise from this stationary level - which is not connected to the spin - wave group velocity itself - is very fast and , consequently , the velocity with which the amplified wave - front becomes detectable within the waveguide is very high .
thus , where the spin waves are already present with a weak amplitude below the detection limit but above the thermal level , their information propagates trough the waveguide seemingly fast . it should be noted that this is not connected to a propagation of the wave packet faster than the group velocity itself - the detected waves do not need to travel , they are basically already present at the point of detection before the pumping starts and they are merely amplified to the detection limit . comparing figs .
[ fig : pamp - glob2 ] ( b ) and ( c ) , it becomes evident that - if a single pumping pulse is used for the spin - wave amplification - a trade - off between a long propagation range and a high effective signal velocity of the amplified waves has to be made .
this way , propagation ranges beyond @xmath215 with signal velocities exceeding @xmath216 can be achieved .
however , the fact that the signal and the dominant mode are the same opens an alternative path : figure [ fig : pamp - glob3 ] ( a ) shows the time - resolved spin - wave intensity of the amplified and the generated waves exemplarily , @xmath217 , @xmath218 long spin - wave excitation pulses , @xmath219 long pumping pulses . ] . due to the fact that the signal wave is the dominant mode , the two spin - wave packets feature identical rise and fall times .
the solid blue lines in the figures are an analytical reconstruction of the measured spin - wave intensity taking into account the changes of the temporal profile due to the bls detection and assuming the same rise and fall times for the amplified as well as the generated spin waves .
as can be seen , they are both in good agreement with the measured profiles .
the only difference in the temporal dynamics of these two spin waves is that the amplified spin - wave packet is detected at an earlier time because it starts with a larger starting amplitude than the thermal waves .
consequently , if a pulse scheme as sketched in fig .
[ fig : pamp - glob3 ] ( b ) is used , the parametric amplifier can be operated with a comparably large pumping power , resulting in a fast effective signal propagation as well as a large peak spin - wave intensity and , at the same time , it can operate with a very low noise .
if the amplification pulses are short enough to avoid that the intensity of the parametrically generated waves always stays well below the initial level of the signal spin waves and if they are applied with a repetition that ensures that the externally excited , amplified packet always recovers to its initial amplitude during the amplification , a loss - less propagation without a notable creation of parametrically generated noise can be realized . as can be seen from fig .
[ fig : pamp - glob3 ] ( b ) , where the rise and fall times of the experiment in ni@xmath3fe@xmath4 have been assumed , this scheme would require a high repetition rate of the pumping pulses .
even though a fast enough on- and off - switching of the pumping might be feasible by a temporal modulation of the pumping power , for a practical application , a material with a lower spin - wave damping is favorable .
such materials , like the half - metallic heusler compound comnfesi@xcite or microstructured yig@xcite will allow for a smaller repetition rate and a larger tolerance of the pulses .
in the previous chapter , the prospects and the limitations of the application of parallel pumping as an exclusive means of spin - wave amplification have been discussed . in the following chapter ,
some of its applications that go beyond a mere amplification will be presented . from eq .
[ eq : phaserel ] it follows that the phase of the two created spin waves , the signal as well as the idler spin waves , are mutually connected to the phase of the pumping field . since the phase of the pumping field and the phase of the signal spin waves are externally determined quantities , the idler wave needs to adjust its phase in order to fulfill this phase - relation .
if the phase evolution between the signal spin waves and the parallel pumping field is externally locked , this can be used to convert the phase of the signal spin waves into an intensity information . in a global amplifier such as
the one discussed above , this implies that the standing wave pattern arsing from the interference of the counter - propagating signal and idler waves can be shifted in space by changing the relative phase @xmath220 between the signal spin waves and the pumping field .
consequently , the phase of the idler wave has to adjust and , hence , the relative phase @xmath221 between signal and idler waves changes , which displaces the standing wave pattern in space .
therefore , within the amplifier , the intensity of the standing wave is connected to the relative phase between the signal spin waves and the pumping field .
thus , in the case of the creation of counter - propagating waves , which is commonly referred to as adiabatic parametric amplification@xcite , the use of a detector in a fixed position allows to deduce on the phase of the signal spin waves if the phase of the pumping field is employed as a fixed reference phase . within the amplifier it is possible to connect the phase of the signal spin waves to an intensity at a given position .
if the amplifier is not extending along the entire magnetic structure but is still large in comparison to the spin - wave wavelength , the output of the amplifier in the forward direction is , however , not affected by the relative phase between signal spin waves and pumping field .
hence , it is impossible to obtain a phase information which is not sensitive to the readout position or to pass on only spin waves with a desired phase within a magnonic network . in turn , this can be achieved by making the amplifier smaller than the wavelength of the spin waves which are to be amplified . in this
_ nonadiabatic regime _ of parametric amplification , the amplification process is quite different .
the basic features of this regime are shown in fig .
[ fig : nov - pti1 ] .
figure [ fig : nov - pti1 ] ( a ) shows the fourier spectrum of a @xmath173 long rectangular amplifier in wavevector space . as can be seen from the figure , the spatial confinement leads to the presence of an extensive effective wavevector spectrum of the parametric amplifier .
these finite wavevectors allow for the splitting process schematically shown in fig .
[ fig : nov - pti1 ] ( b ) : a photon with an effective wavevector of @xmath222 splits into two co - propagating signal and idler waves with wavevector @xmath110 .
hereby , the finite effective wavevector @xmath223 does not imply that the pumping field consists of a propagating wave .
this just accounts for the fact that , throughout the amplifier , the pumping field can provide the necessary momentum to fulfill the momentum conservation condition stated in eq .
[ eq : conservation - nonadiabatic ] at all times .
thus , in this nonadiabatic regime , the creation of co - propagating wave pairs is possible . the consequence of this wave - pair formation is shown in fig . [ fig : nov - pti1 ] ( c ) .
if the phase of the pumping field is considered as reference and the signal and the idler waves are initially in phase ( @xmath224 ) , this leads to a large output of the amplifier .
however , if the phase of the signal wave is shifted by @xmath102 , the idler wave has to adjust its phase by @xmath225 in order to fulfill eq .
[ eq : phaserel ] .
consequently , the created spin waves are out - of - phase and the output of the amplifier is minimal ( @xmath226 ) .
[ fig : nov - pti1 ] ( d ) shows the time - averaged output intensity of the amplifier arising from this interference mechanism which , in the absence of the creation of noise due to parametric generation , is given by the interference of the incoming signal spin wave with the signal and idler spin waves created by the parallel pumping process . for amplitudes @xmath227 of the signal wave and @xmath228 of the idler wave , the time - averaged output intensity of the amplifier at a fixed readout position is given by : @xmath229 with the constant @xmath230 . here
, @xmath231 denotes the amplitude of the incoming signal wave and @xmath232 the amplitude of the created signal as well as idler waves .
hence , the output of the amplifier can be tuned continuously by changing the relative phase @xmath233 .
it should be noted that the dependence of the output intensity on @xmath220 implies that the output of the amplifier is invariant to a change of the spin - wave phase of @xmath234 .
consequently , the spin - wave phase has to be encoded in values between 0 and @xmath102 for the application of such an amplifier to a logic architecture relying on the phase of the spin waves as continuous variable@xcite . for the experimental realization of a phase - dependent amplification in microstructured spin - wave waveguides , a local spin - wave amplifier such as the one schematically shown in fig .
[ fig : pamp - loc1 ] can be used . if the spin - wave excitation and the parallel pumping are phase - locked , it is possible to observe the influence of the relative phase between the signal spin waves and the pumping field @xmath233 on the output of the amplifier@xcite if the spin - wave wavelength is comparable to or larger than the size of the local amplifier .
figure [ fig : nov - pti2 ] ( a ) shows the spin - wave intensity measured along the center - line of a @xmath193 wide , @xmath158 thick , transversely in - plane magnetized ni@xmath3fe@xmath4 waveguide for different scenarios . in this experiment , which is described in more detail in ref .
@xcite , the signal spin waves are excited with a @xmath235 long microwave pulse with a carrier frequency of @xmath236 and their interaction with @xmath235 long pumping pulses with twice this frequency is studied .
the pumping pulse starts when the spin - wave packet enters the amplifier , and the used peak pumping power of @xmath237 is sufficiently low to avoid a strong parametric generation during the duration of the pumping pulse . at the applied external bias field of @xmath238
, the spin - wave frequency corresponds to a spin - wave wavelength of @xmath239 , which is larger than the local parametric amplifier created by a @xmath173 long and @xmath172 wide geometric constriction .
consequently , the output of the amplifier will depend on the relative phase between the signal spin wave and the pumping field .
the green squares show the spin - wave intensity which arises from the parametric amplification of the input spin waves for a constructive interference between the signal and the idler spin waves ( @xmath224 , which corresponds to @xmath240 ) . as can be seen from fig .
[ fig : nov - pti2 ] ( a ) , in this scenario , the spin waves are amplified during their propagation through the amplifier , similar to the adiabatic case shown in fig .
[ fig : pamp - loc1 ] ( b ) .
if the relative phase of the signal spin waves is shifted by @xmath102 with respect to the phase of the pumping field via a phase - shifter localized in the microwave setup providing the pumping pulses , the interference between the signal and the idler waves becomes destructive ( @xmath226 , @xmath241 ) .
the resulting spin - wave intensity is shown by the blue upward triangles .
in contrast to the amplification with a constructive interference , a much smaller spin - wave intensity is leaving the amplifier and , in agreement with eq .
[ eq : output ] , the intensity within and after the amplifier basically coincides with the intensity arising from the unamplified spin - wave packet , which is shown by the violet downward triangles . for comparison
the red dots show the bls intensity if only the pumping field is applied . due to the low pumping power in combination with the finite duration of the pumping pulse , no parametrically generated noise is visible , showing that such a parametric amplifier can work with a low operational noise .
figure [ fig : nov - pti2 ] ( b ) shows the output of the amplifier as a function of the relative phase between the signal spin waves and the pumping field together with a cosinusoidal fit according to eq .
[ eq : output ] .
the output is measured at a distance of @xmath242 from the center of the amplifier .
as can be seen from the figure , the output can be continuously tuned as a function of the phase shift .
thus , if the phase of the pumping field is kept fixed with a certain reference value , the phase of the signal spin waves is directly transferred into an amplified intensity .
this intensity can be readout conveniently by conventional detection schemes such as the inverse spin hall effect in combination with spin pumping or the tunneling magnetoresistance effect@xcite , which convert the ac spin - wave intensity into a dc voltage .
hereby , the inherent amplification of the spin waves with proper phase facilitates the readout and allows to operate with a low spin - wave intensity in the conduit , minimizing energy costs and nonlinear spin - wave interaction .
it should be noted that only an amplitude of amplified signal and idler waves @xmath232 comparable to the amplitude of the incident spin waves @xmath243 is needed to achieve such a large modulation ( @xmath244 in the figure ) .
if the pumping power is increased , a lager number of amplified waves @xmath232 can easily be achieved ( i.e. , high saturation level ) . to minimize the noise and to maximize the intensity difference between constructive and destructive interference , a proper pulse duration to reduce the effect of the stronger parametric generation at larger pumping powers should be used ( cf .
discussion at the end of section [ pamp - glob ] ) . for comparison ,
[ fig : nov - pti2 ] ( c ) shows the bls intensity measured along the spin - wave waveguide if the magnetic bias field is changed to @xmath245 for two different phase - relations between the signal waves and the pumping field . for this bias field ,
the spin - wave wavelength is reduced to @xmath246 and , consequently , the parametric amplifier operates in the adiabatic regime . as mentioned above , in this regime , a change of the relative phase should only change the interference pattern within or in front of the amplifier and should not affect the output of the amplifier . as can be seen from the two measurements shown in fig .
[ fig : nov - pti2 ] ( c ) , between which the relative phase between the signal spin waves and the pumping field has been shifted by @xmath247 , this behavior is well reproduced in the experiment .
thus , with an operation in the adiabatic regime , spin waves are amplified in the forward direction independently of their phase , while a readout in front of the amplifier allows for a deduction on their phase based on the detected spin - wave intensity .
due to the fact that the transition from the adiabatic to the nonadiabatic operation regime can be achieved by simply changing the spin - wave wavelength , a local parametric amplifier is , thus , a very versatile tool in any magnonic network .
one instructive example of how a nonadiabatic parametric amplifier can be implemented into a magnonic network has been demonstrated in ref .
@xcite . in this work
, the application of the phase - to - intensity conversion by parallel pumping to a magnonic majority gate@xcite has been studied numerically by micromagnetic simulations using mumax3@xcite .
a majority gate@xcite is a logic element with an odd number of inputs whose output represents the majority of the input values .
it can act as the basic element of novel logic architectures .
figure [ fig : nov - pti3 ] ( a ) shows the layout of the simulated device : three input waveguides are merged into an output waveguide in a combiner region ( see also ref .
all waveguides operate in the backward - volume geometry and are longitudinally magnetized by an external field of @xmath248 .
each input can be connected to another magnonic device or a spin - wave excitation source . in the output of the majority gate ,
the dynamic field of a @xmath249 wide micro - strip antenna is simulated , which acts as local , nonadiabatic source of the pumping field ( cf .
section [ pgen - bv ] ) .
if a logic zero is encoded in a spin - wave phase - shift of @xmath250 and a logic one is encoded in a phase - shift of @xmath251 , the majority operation is preserved , if all phase - shifts @xmath252 are considered as majority 1 .
the output of the majority gate with parametric amplification is shown in fig .
[ fig : nov - pti3 ] ( b ) , together with a fit according to eq .
[ eq : output ] . as can be seen from the figure , for a fixed phase of the pumping field , the pumping results in four distinct intensity levels , representing the different combinations of the input waveguides . by defining the majority by a threshold intensity situated at one - half of the maximum output intensity , the parametric amplifier preserves the majority function .
moreover , from the absolute value of the intensity it can be concluded whether the majority is full ( e.g. , input 1 in all input waveguides , abbreviated as 111 ) or partial ( e.g. , 101 , i.e. , 1 in input 1 and 3 and 0 in input 2 ) . moreover ,
if the phase of the pumping field is shifted by @xmath234 , the desired majority which should be amplified can be selected , allowing for a flexible operation of the majority gate .
this is just one illustrating example on how the use of the technique of nonadiabatic parallel pumping can enhance the functionality of a magnonic circuit .
a caveat of micro - focused bls is the loss of direct information about the wavevector of the investigated spin waves .
this information can be recovered by phase - resolved bls@xcite , i.e. , the interference of the inelastically scattered photons with photons with a well - defined frequency modulation , or by investigating the interference of counter - propagating spin waves@xcite .
parametric amplification provides an interesting alternative to these techniques , which does not require the use of an electro - optical modulator , which limits the usable frequency - range significantly , or the use of a second antenna .
the inherent interference of the counter - propagating waves in an adiabatic amplifier provides the wavevector information with some additional advantages : on the one hand , the parametric amplification enhances the propagation range of the waves remarkably , which gives access to the observation of the interference of longer wavelengths closer to @xmath253 . on the other hand , due to its weak wavevector selectivity , the process of parallel pumping can amplify even a very weakly excited spin - wave intensity beyond the efficient excitation range of a micro - strip antenna and , this way , gives experimental access to a larger wavevector span .
[ fig : nov - disp ] ( a ) and ( b ) show the spin - wave wavevector obtained from interference patterns such as the one shown in fig .
[ fig : pamp - glob1 ] ( b ) as a function of the applied external bias - field for a fixed frequency of @xmath254 ( a ) and as a function of the applied frequency for a fixed field of @xmath255 ( b ) together with the analytically calculated dispersion relations of the fundamental mode according to eq . [
eq : disp - tf ] .
the measurements have been performed on a @xmath193 wide , @xmath256 thick magnonic waveguide made from ni@xmath3fe@xmath4 grown on a thermally oxidized si wafer on top of a ru / mgo buffer layer and the waveguide has been capped with mgo .
the waveguide has been magnetized transversely by an external bias field and the signal spin waves have been excited with a @xmath136 wide antenna , which is able to excite spin waves efficiently in a range with @xmath257 . the pumping field and the spin - wave excitation have been performed by @xmath190 long microwave pulses and the microwaves have been created by the same frequency generator together with a frequency doubler for the pumping field to ensure a fixed phase - relation ( cf .
section [ pamp - glob ] ) .
the analytical calculations have been performed using @xmath258 , @xmath5 , @xmath259 and @xmath260 ( @xmath261 ) .
as can be seen from the figure , the analytical calculations provide a reasonable agreement for wavevectors @xmath262 , whereas they provide a poor description for smaller wavevectors .
in fact , the frequency vs. wavevector dependence can be approximated by a linear dependence , a finding that has also been made in ref .
this is related to the fact that eq .
[ eq : disp - tf ] is only an approximation , which , among others , neglects the dipolar interaction between the waveguide modes . since , in addition , the use of eq .
[ eq : disp - tf ] requires the knowledge of a large set of material parameters , the convenient characterization of the spin - wave dispersion by the adiabatic parallel pumping process provides a useful tool for experimentalists .
the ni@xmath3fe@xmath4 waveguide investigated in the study of the spin - wave dispersion depicted in figs .
[ fig : nov - disp ] ( a ) and ( b ) features growth - induced differences between the bottom and the top interface@xcite .
this gives rise to different surface anisotropy - constants at the individual surfaces which , in turn , leads to a splitting of the spin - wave dispersion for spin waves running with @xmath263 along the waveguide@xcite .
however , the observed frequency - splitting is on the order of a few mhz and , thus , way below the frequency resolution of the used bls spectrometer .
one way to work around this limitation in the frequency resolution is the use of parametric amplification of coherently excited traveling spin waves together with micro - focused bls or electrical detection .
the concept of the detection of such small frequency - non - reciprocities in transversely magnetized waveguides is schematically shown in fig .
[ fig : nov - disp ] ( c ) : the presence of different interfaces leads to different boundary conditions for waves traveling to the left and to the right along the waveguide . with respect to the dispersion relation @xmath264 of the fundamental mode in the case of identical interfaces , which is indicated by the dashed lines ,
this leads to a splitting of the dispersion for @xmath265 which is an odd function of the spin - wave wavevector@xcite . in the process of adiabatic parallel pumping ,
the two created waves need to satisfy @xmath266 . here , @xmath267 denotes the wave traveling to the right and @xmath268 the wave traveling to the left . if the frequency - splitting is an odd function of @xmath269 , the system satisfies momentum and energy conservation by amplifying wave pairs featuring @xmath270 and @xmath271 and @xmath272 with @xmath273 , which is possible in many situations due to approximate linearity of the spin - wave dispersion relation ( cf .
[ fig : nov - disp ] ( a ) and ref .
@xcite ) . hence ,
the interference pattern resulting from the interference of the counter - propagating signal and the idler waves will be non - stationary , since now the difference between the two frequencies does not vanish .
instead , the created interference pattern , which features a spatial periodicity of @xmath274 , is now constantly moving in space with the difference of the frequencies @xmath275 since @xmath276 .
thus , by measuring the time - resolved bls intensity in a fixed position , one observes a temporal modulation of the bls intensity with a frequency in the mhz range .
this modulation can easily be resolved with the temporal resolution of bls . ] , which lies in the order of @xmath277 .
this is demonstrated in fig .
[ fig : nov - disp ] ( d ) , where the time - resolved bls intensity shows the interplay of a @xmath278 long spin - wave packet ( @xmath205 , @xmath254 , @xmath202 ) with the microwave pumping field together with a cosinusoidal fit .
it should be noted that the intensity modulation can be maintained on these comparably long time scales only because the dominantly generated spin - wave mode in the waveguide is the fundamental mode . the fit yields @xmath279 , a value in the order of magnitude expected for a ni@xmath3fe@xmath4 film in this thickness range with adjacent oxide interfaces@xcite .
figure [ fig : nov - disp ] ( e ) shows the wave - vector dependent frequency splitting obtained from similar measurements performed as a function of the externally applied bias field for several fixed frequencies and as a function of frequency for a fixed field of @xmath280 . for all these measurements , the wavevector @xmath269 has been extracted from the spatial modulation from measurements with short pulses .
the frequency shifts have been determined from the fourier transformation of the time - resolved bls measurements . as can be seen from the figure
, the non - reciprocity primarily depends on the wavevector @xmath24 , as the values for @xmath281 obtained from the different measurements mostly coincide within the error bars .
the orange curve is a theoretical prediction of the frequency non - reciprocity based on ref .
@xcite assuming nonidentical surface constants of @xmath282 and @xmath283 , values which are in the expected parameter range @xcite .
as can be seen from the figure , the model provides a reasonable description for wavevectors larger than @xmath284 , whereas there is a significant deviation for smaller wavevectors .
thus , the region at very small wavevectors , which is not accessible to conventional bls measurements , indicates that a more involved theoretical model is necessary to explain all the ( nonreciprocal ) features of the spin - wave dispersion in such layer systems .
another origin of frequency non - reciprocities which are an odd function of the spin - wave wavevector is the interfacial dzyaloshinskii - moriya interaction ( idmi)@xcite , which can arise in thin magnetic films in a layer system featuring a structural inversion asymmetry ( sia)@xcite . the idmi has recently attracted a large research interest due to its intriguing consequences such as , among others , the stabilization of magnetic skyrmions@xcite , its impact on the domain wall dynamics @xcite or the deterministic magnetization switching driven by spin orbit torques@xcite .
conventional measurements of the idmi - strength in ultra - thin films are either performed by analyzing the domain - wall expansion@xcite or by analyzing the non - reciprocal spin - wave dispersion by conventional , wavevector resolved brillouin light scattering spectroscopy in extended films@xcite .
more recently , also an electrical characterization of the idmi in an individual microstructure has been reported in ref .
@xcite using similar means as have been used in ref .
@xcite for the characterization of influence of unequal surface anisotropies .
while the analysis of the domain - wall - motion in the presence of idmi can be rather involved , the bls measurements suffer from the rather small frequency resolution of the bls of a few @xmath285 , which , as mentioned above , hampers the determination of the idmi - constant ( or anisotropy - differences ) for small wavevectors or for very small idmi - strengths .
moreover , since micro - focused bls comes along with a loss of wavevector resolution , this method can not be used to probe individual micro- or nanostructures unless phase - resolved bls is applied or spin - wave interference is studied .
the demonstrated electrical detection schemes can be applied to individual microstructures . however , the demonstration in ref .
@xcite has been carried out on pt / co / mgo , a system that features a very high dmi constant , and the method is already at the detection limit in this system .
since the impact of the idmi on the spin - wave dispersion scales with the inverse of the thickness of the ferromagnetic layer , its determination is best performed in very thin films , which also helps to reduce the influence of the likely unequal surface anisotropies in the presence of sia .
due to the large induced damping caused by the contact to the heavy metals in common systems featuring idmi such as co / pt , ni@xmath3fe@xmath4/pt or cofeb / pt and the low group velocity in very thin films , propagating spin - wave spectroscopy becomes a challenging task . here ,
the presented method to use the interference pattern created by the parametric amplification of traveling waves offers a good alternative : it can detect very tiny shifts and it is not limited by the small propagation range .
thus , together with the determination of the spin - wave dispersion using short microwave pulses to determine the spin - wave wavevector from the spatial interference pattern , it constitutes a valuable alternative method for the characterization of idmi in individual microstructures . besides the amplification of traveling waves ,
parallel parametric amplification has also been applied to spin - transfer torque driven nano - oscillators ( stno)@xcite to study the underlying physical phenomena and to improve their output characteristics . in an stno ,
the magnetization of an active layer , typically referred to as the free layer , is subjected to an incident spin current .
this spin current can be provided , for instance , by a direct current in a magnetoresistive device such as a spin valve or a magnetic tunneling junction@xcite or by the inverse spin hall effect in an adjacent heavy - metal layer@xcite . despite the various differences in geometry ,
magnetization configuration and material systems , all stnos are based on the common principle that the incident spin current results in a spin transfer torque ( stt)@xcite acting on the magnetization of the free layer which counteracts the damping . similar to the impact of parametric amplification , this leads to a growth of the oscillation amplitude of the stno if the inserted energy per unit time exceeds the loss rate of the stt - driven mode .
spin - wave amplification by stt is , thus , also a threshold process .
this mechanism allows for the generation of an rf current by the conversion of a dc current .
it should , however , be noted that it is hardly possible to select the fundamental mode to be the dominant mode amplified by stt in a spin - wave waveguide .
therefore , it is difficult to implement it as a means of spin - wave amplification in extended systems such as spin - wave waveguides@xcite .
in contrast , it was demonstrated that it can constitute an efficient means to excite standing waves in a highly - quantized geometry in stnos .
the manifold of device geometries for stnos leads to a large span of frequencies which can be covered by such rf sources .
this renders stnos highly attractive for device applications , such as rf communications , where miniaturized rf sources with a large tunability are needed .
however , due to the lack of a frequency - selectivity , the dc stt interacts with the entire spin - wave spectrum . from a fundamental point of view
this gives rise to many interesting nonlinear phenomena and creates a large playground to study non - linear spin - wave dynamics .
for a device application , however , this hampers the functionality of the stnos , since the interaction of the stt with the entire spin - wave spectrum leads to an undesired increase of the noise present in the oscillator . the resulting spin - wave spin - wave interaction can result in a loss of phase - stability and increases the linewidth of the oscillator .
thus , typically stnos are systems where the spin waves are strongly quantized and the spin - wave spectrum should be as simple and diluted as possible to ensure good device properties . since this is not always easy to achieve , many stnos suffer from a rather large linewidth .
in addition , the output power of individual stnos is too low for most practical applications .
one proposed way to overcome these issues is the synchronization of an array of oscillators .
such an array with oscillators working in parallel can produce a large output power where the mutual phase - locking of the oscillators provides the necessary phase - stability and reduces the linewidth@xcite .
first experimental evidence of phase - locked pairs of point contacts@xcite has been followed by the demonstration of the synchronization of up to 5 oscillators@xcite . however , the creation of large synchronized arrays is a challenging task .
this is , for instance , due to the limited reproducibility of the exact device shape , resulting in a frequency detuning between individual devices , and the large joule heating created in an array@xcite .
moreover , due to the limited spin - wave propagation range , only oscillators situated close - by can be directly synchronized via spin - wave beams . in order to analyze the full potential and the limitations of phase - locking experimentally , as well as to explore the optimum operation conditions , large efforts have also been devoted to study the interaction of an stno with an external locking - source , such as a microwave generator . among these experiments ,
the parametric interaction of oscillators with a microwave current at twice their free - running auto - oscillation frequency has proven a vital technique , since it allows to decouple the driving frequency from the detected fundamental frequency of the oscillator . as will be briefly discussed in the following , these experiments have , among others , allowed to study the influence of noise and the applied pumping power on the locking bandwidth , have enabled an alternative quantification of the effect of the stt on the damping by analyzing its influence on the parametric instability threshold , and have proven that the coupling to an external pumping field can compensate the frequency mismatch of different oscillators in the locked state .
the first observations of the excitation of spin waves by parallel pumping in a magnetic point contact has been made in ref .
@xcite , even though the proper identification of the parallel pumping process has been only made later in ref .
@xcite by means of numerical simulations . in ref .
@xcite , the frequency - response of a ni@xmath286fe@xmath287 layer to a rf current which is injected by a point - contact of a diameter of about @xmath249 has been studied by means of bls microscopy .
the ni@xmath286fe@xmath287 is the free layer in a co@xmath288fe@xmath289/cu / ni@xmath286fe@xmath287 spin valve .
the measurements have been performed at room temperature .
it has been found that beyond a certain threshold value of the injected rf power a strong generation of spin waves at a frequency equal to one half of the frequency of the driving microwave current is observed .
figure [ fig : nov - ps1 ] ( a ) shows the bls intensity measured at the microwave frequency @xmath290 and at one half of the externally applied frequency @xmath291 as a function of the applied microwave power in the absence of a dc bias . during the measurement , an external field of @xmath292 is applied in - plane , perpendicular to the top electrode of the point contact ( @xmath161-direction in ref .
as can be seen from the figure , a clear threshold behavior is observed for the spin waves at @xmath293 .
moreover , the intensity of the spin waves at frequency @xmath199 and , in particular , its increase with increasing microwave power , are not affected by the onset of the instability which leads to the formation of spin waves at @xmath293 . this is strong evidence that the observed instability is indeed the parametric instability due to longitudinal parallel pumping : if the growth of the spin - wave intensity at @xmath293 would be related to a spin - wave instability - as has initially been assumed in ref .
@xcite - this should result in a kink in the dependence of the spin - wave intensity at @xmath199 as a function of the driving microwave power at the threshold of the instability@xcite .
after the onset of the three - magnon interaction , the intensity is supposed to grow with a smaller slope with respect to @xmath199 .
the absence of a three - magnon splitting in the experiment is further supported by the fact that this splitting process is impossible in the experimental conditions due to momentum conservation @xcite . in ref .
@xcite , the influence of an additional dc current with current densities below the threshold of stt - driven auto - oscillations on the threshold of the parametric instability has been investigated . as can be seen from fig .
[ fig : nov - ps1 ] ( b ) , a strong modulation of the threshold power as a function of the applied bias current is observed .
the observed quadratic dependence of the threshold power on the applied dc current is consistent with the parametric instability : since the stt is resulting in a linear change of the relaxation frequency@xcite , it leads to a linear change in the threshold field @xmath294 ( cf .
. [ eq : threshold ] ) . since @xmath295 ,
a quadratic change of the threshold power is expected . in ref .
@xcite , the experimental configuration of ref .
@xcite has been investigated by means of micromagnetic simulations@xcite . in these simulations ,
the excitation mechanism has been indeed identified with the parametric instability and the oersted field created by the current flow through the top contact was identified as the origin of the pumping field .
this can be comprehended from fig .
[ fig : nov - ps1 ] ( c ) , where the simulated excitation spectra are shown if the excitation is only created by the transverse oersted field @xmath296 , which can exert a direct torque on the magnetization , or by the longitudinal oersted field @xmath297 , which fulfills the requirements for parallel parametric amplification .
the distribution of the oersted field has been computed numerically . as can be seen from the figure
, the transverse field only leads to an excitation at frequency @xmath199 and a higher harmonic at @xmath298 .
in contrast , the longitudinal component , which can not exert a direct torque on the magnetization , leads to the formation of spin waves at @xmath293 . since the spin waves at @xmath293 are observed exclusively in simulations where the longitudinal component is included and since no spin waves at frequency @xmath199 are excited under these conditions , a spin - wave instability can indeed be excluded as origin of the instability observed in the experiment .
the fact that the excitation at @xmath199 and at @xmath298 are mediated by the different components of the dynamic oersted field also explains there complete independence in terms of their intensity - vs .- power characteristics ( cf .
[ fig : nov - ps1 ] ( a ) ) .
thus , the experiment in ref .
@xcite constitutes the first observation of parametric generation in a nanocontact and provides a first insight of the interplay of the damping change mediated by the dc - stt and the parallel parametric amplification in such a device .
the parametric generation in a stno below the threshold of the instability due to the stt has been labeled as _ parametric excitation _ in ref .
@xcite . in this work ,
the parallel parametric amplification in a stno was studied more systematically and the region below and beyond the threshold for the dc - stt driven auto - oscillations has been investigated experimentally as well as theoretically .
the study was conducted on a ni@xmath286fe@xmath287/cu / co@xmath299fe@xmath300 spin - valve , where the polarizing top co@xmath299fe@xmath300 and a part of the cu layer were patterned into an ellipse of dimensions @xmath301 .
the microwave pumping field was provided by an additional micro - strip patterned on top of the nanocontact . in the study ,
an external bias field is applied in the sample plane at an angle of @xmath302 with respect to the long axis of the ellipse .
the dynamic magnetization in the free layer is detected via the magnetoresistance .
all measurements were performed at @xmath303 .
the applied microwave power is always below the parametric generation threshold in the absence of stt . below the threshold of stt - driven auto - oscillations ,
parametric excitation can be used to study the spectral characteristics of the stno and to study the damping change inflicted by the stt .
this way , the parametric excitation of an stno provides a powerful alternative to spin torque ferromagnetic resonance spectroscopy ( st - fmr)@xcite , since the microwave excitation at frequency @xmath199 and the dynamic response of the oscillator at @xmath293 are decoupled .
figure [ fig : nov - ps2 ] ( a ) shows the dynamic response of the stno for different microwave frequencies for a fixed microwave field @xmath304 and dc bias @xmath305 . as can be seen from the figure , a sizable parametric generation can be observed even for microwave frequencies different from the resonance frequency of the stno of @xmath306 for the used magnetic bias field .
this frequency detuning , which corresponds to the locking - range of a frequency - locked stno according to ref .
@xcite , depends on the applied supercriticality of the pumping field . ] .
this can be comprehended from fig .
[ fig : nov - ps2 ] ( c ) , where the locking - range is demonstrated as a function of the applied pumping field .
the locking range is given by the area enclosed by the black circles , which correspond to the minimum and maximum frequency at which parametric excitation can still be observed experimentally , respectively . as can be comprehended from fig .
[ fig : nov - ps2 ] ( a ) and ( b ) , the most efficient generation is achieved for @xmath307 , which is also associated with the minimum linewidth of the parametrically excited spin waves in the stno . due to the fact that in this scenario the stt has not compensated the spin - wave damping
, a finite threshold field always remains for parametric excitation .
this is different in the case of parametric synchronization .
if the dc bias is increased beyond the threshold of auto - oscillations , the oscillator can be locked without a threshold rf current ( cf .
[ fig : nov - ps2 ] ( d ) , where the locking range is depicted as a function of the microwave pumping field above the dc stt - driven auto - oscillation threshold ) .
thus , beyond this threshold , the oscillator can easily be phase - locked to the parametric excitation source and a convenient _ parametric synchronization _ to the external source can be observed , which has also been demonstrated by the same authors in ref .
@xcite . in the regime of parametric synchronization
, the supercriticality of the pumping field mainly affects the locking - range of the oscillator .
these first experiments were followed by a multitude of experiments in different devices and geometries and the development of the necessary theoretical framework for their description . besides studies on point - contacts@xcite , where , for instance , the parametric generation and emission of propagating spin waves due to a frequency mismatch between the oscillator and the microwave pumping was demonstrated@xcite , experiments were extended to stno incorporating magnetic tunneling junctions ( mtj)@xcite , free layers in the magnetic vortex configuration@xcite and , more recently , to spin - hall - effect driven stnos@xcite .
for instance , in the early work on vortex - oscillators in ref .
@xcite , it was demonstrated that parallel pumping can lock two oscillators which feature different free running auto - oscillation frequencies within the locking range .
moreover , in the more recent experiment in ref .
@xcite , the perfect phase - locking by parametric synchronization of a double - vortex oscillator , such as the one schematically shown as an inset in the lower panel in fig . [ fig : nov - ps3 ] ( a ) , is demonstrated .
the oscillator features a vortex in the free as well as the fixed layer , which mutually precess in the dc - stt driven auto - oscillating regime .
such devices can feature remarkably low linewidths in the free running regime at room temperature , such as the reported @xmath308 relative to the free running frequency of @xmath309 .
the microwave pumping in this experiment is provided by a micro - strip antenna patterned on top of the stno , which provides a microwave pumping field with frequency @xmath199 in the plane of the discs . as demonstrated by the blue curve in fig .
[ fig : nov - ps3 ] ( a ) , where the frequency of the stno is shown relative to one half of the applied driving frequency @xmath199 of the pumping field , this enables the robust perfect phase - locking over a range @xmath310 over a locking - range greatly exceeding the free - running linewidth with a frequency linewidth @xmath311 .
the importance of the phase - stability of the free - running oscillator is highlighted by the red curve : if the excitation field is changed , the mode - spectrum of the oscillator changes and the linewidth increases to a rather poor value of @xmath312 at a free running frequency of @xmath313 . as is evident from the figure , this leads to a drastic decrease of robustness of the locking as well as of the best achievable linewidth , which increases to about @xmath314 .
this behavior was theoretically explained with the model presented in ref .
@xcite ( cf .
bold lines in fig .
[ fig : nov - ps3 ] ( a ) ) .
more recent experiments on mtj - based oscillators have demonstrated the robustness of parametrically synchronized stno to a modulation of the applied microwave current if the modulation frequency is too fast for the nonlinearity of the oscillator to follow@xcite .
other experiments have demonstrated the possible use of a stno to detect microwaves at the pumping frequency via spin rectification@xcite .
all of these device geometries have in common that parametric synchronization does not show a threshold behavior and , in general , is connected to a significant increase of the peak spectral density ( psd ) ( i.e. , of the intensity of the maximum of the peak ) .
they also show that the linewidth of the peak under synchronization benefits from an initial low phase - noise of the free running oscillator ( compare , e. g. , refs . @xcite and @xcite ) .
further light was shed on this in the recent experiments on parametric synchronization of a spin - hall - effect driven stno@xcite . in this experiment ,
the microwave emission of such an stno based on a ni@xmath286fe@xmath287/pt bilayer , subjected to parametric excitation as well as parametric synchronization was characterized by means of bls and by measuring the dynamic anomalous magnetoresistance ( amr)@xcite .
the device design is shown in fig .
[ fig : nov - ps3 ] ( b ) . the rf field for the parallel pumping
was provided by the part of the rf current sent through the oscillator flowing inside the pt - layer . as shown in fig .
[ fig : nov - ps3 ] ( c ) , where the frequency of the oscillator is shown as a function of the pumping frequency , it was observed that parametric synchronization of the stno is possible .
it was also found by the electrical measurements that it leads to a significant decrease of the linewidth of the oscillator .
however , in this type of stno , parametric synchronization requires a threshold microwave pumping power , has a comparably small locking - bandwidth and does not lead to a significant increase of the psd .
these findings have been explained in the framework of the theories provided in refs . @xcite and @xcite and could be ascribed to the presence of a prominent thermal noise in the spin - hall - effect driven stno .
it should be noted that the parametric excitation , or rather the effect of the stt on the parametric generation , has been studied by the same authors in ref .
@xcite . in this work
, they demonstrate that the stt can be used to lower the generation threshold in a microscopic ni@xmath286fe@xmath287/pt disc similar to the ones studied in ref .
@xcite ( cf
. section [ pgen - ell ] ) and that , in turn , parametric generation can be used to quantify the stt - induced damping change .
this approach has also been successfully applied to study the influence of the spin - hall effect driven stt in a macroscopic bilayer of yig / pt@xcite .
another interesting prospect of parametric synchronization in spin - hall - effect driven stno was demonstrated in ref .
@xcite . in this work ,
a single as well as a pair of nano - constriction based oscillators@xcite were analyzed by micromagnetic simulations .
a key to this numerical study is the lateral displacement of the two oscillators , which can be comprehended from fig .
[ fig : nov - ps3 ] ( d ) , where a color map shows the spin - wave intensity at @xmath298 and the outline shows the extents of the oscillators .
these spin waves at @xmath298 are created as a higher harmonic in the oscillator and lie within the band of propagating spin waves .
they are created in the individual stnos and are emitted at an oblique angle . if the design of the device is adapted and the spin waves emitted from one stno can reach the second stno and vice versa , an efficient mutual parametric synchronization of the stnos mediated by these spin waves is possible .
this is demonstrated in fig .
[ fig : nov - ps3 ] ( e ) , where the spectrum of the oscillator pair is shown as a function of their distance .
if they are far apart , the spin waves have decayed significantly during their propagation from one oscillator to the other .
consequently , their intensity is small to lock the oscillators and the individual spectra of the two stno are visible . since the geometric constrictions are slightly different , the stnos feature different resonance frequencies .
nevertheless , for a still substantial separation of about @xmath315 , an efficient spin - wave mediated locking of the oscillators is observed and they oscillate with a mutual frequency . for smaller distances , the locking becomes even more robust .
the oscillatory behavior of the characteristics has been attributed to the interference of the short - wavelength spin waves at @xmath298 , which is visible in fig .
[ fig : nov - ps3 ] ( d ) and , at small separations , to the influence of the evanescent waves at frequency @xmath199 created in the stnos .
hence , the parametric locking of the oscillators to the propagating spin waves at @xmath298 allows for a mutual phase - locking of stnos , even in an array with rather large device separations exceeding @xmath316 .
in this review we have given an overview over the recent advances in the application of the technique of parallel parametric amplification to magnetization dynamics in micro- and nanostructures .
parametric generation allows to study the dominant mode spectra and enables the efficient excitation of spin waves with very short wavelengths , practically independent of the size of the pumping source .
this way , it gives access to plenty of nonlinear spin - wave phenomena and paves the way for the observation of bose einstein condensation of magnons in micro- and nanostructures made from materials with ultra low damping .
moreover , due to the fact that the frequency- and mode - selectivity of parametric amplification allow for the dominant amplification of the fundamental mode in these quantized systems , this method provides a very efficient means of spin - wave amplification . this new paradigm , which is very different from the previous applications of parallel pumping in macroscopic structures , allows for new applications of parallel pumping to create , amplify and manipulate information processed by magnons in microscopic devices .
its selectivity makes parallel pumping a superior method for the amplification of traveling waves , since the energy from the pumping field is not distributed over the entire spin - wave spectrum
. however , there are still some issues to be addressed in this field : most of the current demonstrations of parallel pumping have been based on the utilization of oersted fields created by the rf currents sent through an adjacent micro - strip or through the structures itself .
this approach is limited in scalability and requires quite large microwave fields . on the one hand , since the absorption of the individual structures is low and the applied microwave field can interact with a large number of structures , the field can be used as a global reference phase on the chip and it can be implemented as an active part in the logic circuits - benefiting in particular from the development of new materials with large saturation magnetization ( i.e. , large coupling ) and low damping . nevertheless , the interplay with individual nanostructures will require alternative approaches to create the pumping field in a local and energy - efficient way .
since any interaction which alters the effective field in parallel to the alignment of the static magnetization can be used , there are , however , plenty of options to achieve this .
first hint that a part of the parametric excitation in stnos are promoted at least partially by the ac - stt are just one instructive example of where the journey is heading .
an identification of alternative driving mechanisms such as the variation of the anisotropy field by electric fields or mechanical stress in magnetostrictive layers , the utilization of field - like spin - orbit torques or of stt in the current - perpendicular - to - plane geometry will constitute the next milestones for the application of parallel parametric amplification .
the feasibility of these means for the parametric spin - wave amplification needs to be explored experimentally as well as theoretically , and the corresponding coupling mechanisms need to be identified and optimized . in addition , the interplay of parametric amplification with a control of the magnetic damping by a dc - stt needs to be further explored and its full potential on the amplification of traveling waves needs to investigated . in the end
, the journey will be a rewarding one : a decay - free propagation of spin waves in extended magnonic circuits becomes possible by parallel parametric amplification .
the use of adiabatic and nonadiabatic amplifiers allows to satisfy all the needs for phase - dependent and phase - independent spin - wave amplification , adding vital ingredients to the magnonic toolbox.
financial support by the graduate school materials science in mainz ( mainz ) through dfg - funding of the excellence initiative ( gsc 266 ) and by the dfg ( trr49 ) as well as the nachwuchsring of the tu kaiserslautern is gratefully acknowledged .
a. v. chumak , v. i. vasyuchka , a. a. serga , and b. hillebrands , magnon spintronics , nat . phys . *
11 * , 453 - 461 ( 2015 ) .
s. neusser and d. grundler , magnonics : spin waves on the nanoscale , advanced materials * 21 * , 2927 ( 2009 ) .
a. khitun , m. bao , and k. l. wang , magnonic logic circuits , j. phys .
* 43 * , 264005 ( 2010 ) .
a. khitun , m. bao , and k. l. wang , spin wave magnetic nanofabric : a new approach to spin - based logic circuitry , ieee trans .
mag.,*44 * , 9 ( 2008 ) .
b. lenk , h. ulrichs , f. garbs , m. mnzenberg , the building blocks of magnonics , phys .
rep . * 507 * , 107 ( 2011 ) .
p. pirro , t. brcher , k. vogt , b. obry , h. schultheiss , b. leven , and b. hillebrands , interference of coherent spin waves in micron - sized ferromagnetic waveguides , phys .
status solidi b * 248 * , 2404 ( 2011 ) .
t. sebastian , y. ohdaira , t. kubota , p. pirro , t. brcher , k. vogt , a. a. serga , h. naganuma , m. oogane , y. ando , and b. hillebrands , low - damping spin - wave propagation in a micro - structured co@xmath317mn@xmath318fe@xmath319si heusler waveguide , appl .
100 * , 112402 ( 2012 ) p. pirro , t. brcher , a. v. chumak , b. lgel , c. dubs , o. surzhenko , p. grnert , b. leven , and b. hillebrands , spin - wave excitation and propagation in microstructured waveguides of yttrium iron garnet / pt bilayers , appl .
. lett . * 104 * , 012402 ( 2014 ) .
a. hamadeh , o. dallivy kelly , c. hahn , h. meley , r. bernard , a. h. molpeceres , v. v. naletov , m. viret , a. anane , v. cros , s. o. demokritov , j. l. prieto , m. mu@xmath320 oz , g. de loubens , and o. klein , full control of the spin - wave damping in a magnetic insulator using spin - orbit torque , phys .
lett . * 113 * , 197203 ( 2014 ) . c. hahn , v. v. naletov , g. de loubens , o. klein , o. dallivy kelly , a. anane , r. bernard , e. jacquet , p. bortolotti , v. cros , j. l. prieto , and m. mu@xmath320 oz , measurement of the intrinsic damping constant in individual nanodisks of y@xmath321fe@xmath322o@xmath323 and y@xmath321fe@xmath322o@xmath323|pt , appl .
* 104 * , 152410 ( 2014 ) .
p. pirro , t. sebastian , t. brcher , a. a. serga , t. kubota , h. naganuma , m. oogane , y. ando , and b. hillebrands , non - gilbert - damping mechanism in a ferromagnetic heusler compound probed by nonlinear spin dynamics , phys . rev
. lett . * 113 * , 227601 ( 2014 ) o. gladii , m. collet , k. garcia - hernandez , c. cheng , s. xavier , p. bortolotti , v. cros , y. henry , j .- v .
kim , a. anane , and m. bailleul , spin wave amplification using the spin hall effect in permalloy / platinum bilayers , appl .
. lett . * 108 * , 202407 ( 2016 ) .
k. an , d. r. birt , c .- f .
pai , k. olsson , d. c. ralph , r. a. buhrman , and x. li , control of propagating spin waves via spin transfer torque in a metallic bilayer waveguide , phys .
b * 89 * , 140405(r ) ( 2014 ) .
a. a. serga , v. s. tiberkevich , c. w. sandweg , v. i. vasyuchka , d. a. bozhko , a. v. chumak , t. neumann , b. obry , g. a. melkov , a. n. slavin , and b. hillebrands , bose - einstein condensation in an ultra - hot gas of pumped magnons , nat . commun . * 5 * , 3452 ( 2014 ) a. v. chumak , a. a. serga , g. a. melkov , a. n. slavin , v. tiberkevich , and b. hillebrands , parametric recovery of a microwave signal using standing spin - wave modes of a magnetic film , phys .
b * 79 * ( 1 ) , 014405 ( 2009 ) .
s. schfer , v. kegel , a. a. serga , b. hillebrands , and m. p. kostylev , variable damping and coherence in a high - density magnon gas , phys .
b * 83 * 184407 ( 2011 )
. g. a. melkov , a. a. serga , a. n. slavin , v. s. tiberkevich , a. n. oleinik , and a. v. bagada , parametric interaction of magnetostatic waves with a nonstationary local pump , jetp * 89 * , 1189 - 1199 ( 1999 ) .
t. brcher , p. pirro , a. a. serga , and b. hillebrands , localized parametric generation of spin waves in a longitudinally magnetized ni@xmath3fe@xmath4 waveguide , appl .
* 103 * , 142415 ( 2013 ) .
f. guo , l. m. belova , and r. d. mcmichael , parametric pumping of precession modes in ferromagnetic nanodisks , phys .
b * 89 * , 104422 ( 2014 ) .
b. a. kalinikos and a. n. slavin , theory of dipole - exchange spin wave spectrum for ferromagnetic films with mixed exchange boundary conditions , j. phys .
c : solid state * 19 * , 7013 ( 1986 ) . c. bayer , j. jorzick , s. demokritov , a. slavin , k. guslienko , d. berkov , n. gorn , m. kostylev , b. hillebrands , spin - wave excitations in finite rectangular elements , in b. hillebrands , a. thiaville ( hg . ) , spin dynamics in confined magnetic structures iii , band 101 , topics in applied physics , 57103 , springer berlin heidelberg ( 2006 ) , isbn 978 - 3 - 540 - 20108 - 3 .
k. vogt , h. schultheiss , s. j. hermsdoerfer , p. pirro , a. a. serga , and b. hillebrands , all - optical detection of phase fronts of propagating spin waves in a ni@xmath3fe@xmath4 microstripe , appl .
95 * , 182508 ( 2009 ) .
g. a. melkov , a. a. serga , v. s. tiberkevich , yu . v. kobljanskij , and a. n. slavin , nonadiabatic interaction of a propagating wave packet with localized parametric pumping , phys .
e * 63 * , 066607 ( 2001 ) . c. w. sandweg , y. kajiwara , a. v. chumak , a. a. serga , v. i. vasyuchka , m. b. jungfleisch , e. saitoh , and b. hillebrands , spin pumping by parametrically excited exchange magnons , phys .
lett . * 106 * , 216601 ( 2011 ) .
s. schfer , parametrisch angeregte kohrente wechselwirkungen in rumlich eingeschrnkten magnonengasen , dissertation , technische universitt kaiserslautern ( 2010 ) , available online at http://www.physik.uni-kl.de/fileadmin/hillebrands/diplom-_und_doktorarbeiten/phd-schaefer.pdf .
p. pirro , lineare und nichtlineare spinwelleninteraktionen in magnetischen mikrostrukturen , dissertation , technische universitt kaiserslautern ( 2014 ) , available online at https://www.physik.uni-kl.de/fileadmin/hillebrands/diplom-_und_doktorarbeiten/phd_pirro.pdf .
a. a. serga , m. p. kostylev , b. a. kalinikos , s. o. demokritov , b. hillebrands , and h. benner , parametric generation of solitonlike spin - wave pulses in ring resonators based on ferromagnetic films , jetp lett . * 77 * , 300 ( 2003 ) .
b. a. kalinikos and n. v. kozhus , effect of magnetic anisotropy of ferromagnetic films on the parametric instability of spin waves for the case of longitudinal excitation , sov .
phys . tech .
phys . * 34 * , 1105 ( 1989 ) .
b. a. kalinikos , b. v. kovshikov , and n. v. kozhus , determination of the instability threshold of standing spin waves in yttrium iron garnet films subjected to longitudinal pumping , sov .
phys . solid state * 26 * , 1735 ( 1985 ) .
g. wiese , l. buxman , p. kabos , and c. e. patton , parallel pumping fine structure at 9.4 ghz for in - plane magnetized yttrium iron garnet thin films , j. appl .
75 * , 1041 ( 1994 ) .
t. brcher , parallel parametric amplification of spin waves in micro - structures , dissertation , technische universitt kaiserslautern ( 2015 ) , available online at https://www.physik.uni-kl.de/fileadmin/hillebrands/diplom-_und_doktorarbeiten/phd_braecher.pdf . c. w. sandweg , y. kajiwara , a. v. chumak , a. a. serga , v. i. vasyuchka , m. b. jungfleisch , e. saitoh , b. hillebrands , spin pumping by parametrically excited exchange magnons , phys .
rev . lett .
* 106 * ( 21 ) , 216601 ( 2011 ) .
s. o. demokritov , v. e. demidov , o. dzyapko , g. a. melkov , a. a. serga , b. hillebrands , and a. n. slavin , bose - einstein condensation of quasi - equilibrium magnons at room temperature under pumping , nature * 443 * , 430 ( 2006 ) .
p. clausen , d. a. bozhko , v. i. vasyuchka , b. hillebrands , g. a. melkov , and a. a. serga , stimulated thermalization of a parametrically driven magnon gas as a prerequisite for bose - einstein magnon condensation , phys .
b * 91 * , 220402(r ) ( 2015 ) .
r. ditizio , p.
werbaneth , and j .-
zhu , cell shape and patterning considerations for magnetic random access memory ( mram ) fabrication , ( 2008 ) .
available at http://repository.cmu.edu/cgi/viewcontent.cgi?article=1000&context=ece .
m. cubukcu , o. boulle , m. drouard , k. garello , c. o. avci , i. m. miron , j. langer , b. ocker , p. gambardella and g. gaudin , spin - orbit torque magnetization switching of a three - terminal perpendicular magnetic tunnel junction , appl .
* 104 * , 042406 ( 2014 ) .
k. vogt , o. sukhostavets , h. schultheiss , b. obry , p. pirro , a. a. serga , t. sebastian , j. gonzalez , k. y. guslienko , and b. hillebrands , optical detection of vortex spin - wave eigenmodes in microstructured ferromagnetic disks , phys . rev .
b * 84 * , 174401 ( 2011 ) .
j. jorzick , s. o. demokritov , b. hillebrands , b. bartenlian , c. chappert , d. decanini , f. rousseaux , and e. cambril , quantized spin wave modes in micron size magnetic disks , j. appl . phys . * 87 * , 5082 ( 2000 ) .
r. mock , b. hillebrands , and r. sandercock , construction and performance of a brillouin scattering set - up using a triple - pass tandem fabry - perot interferometer , j. phys .
instrum . * 20 * , 656 ( 1987 ) .
o. klein , g. de loubens , v. v. naletov , f. boust , t. guillet , h. hurdequint , a. leksikov , a. n. slavin , v. s. tiberkevich , and n. vukadinovic , ferromagnetic resonance force spectroscopy of individual submicron - size samples , phys . rev .
b * 78 * , 144410 ( 2008 ) .
j. jorzick , c. krmer , s. o. demokritov , b. hillebrands , b. bartenlian , c. chappert , d. decanini , f. rousseaux , e. cambril , e. sndergard , m. bailleul , c. fermon , and a. n. slavin , spin wave quantization in laterally confined magnetic structures , j. appl . phys . * 89 * , 7091 ( 2001 ) .
j. jorzick , s. o. demokritov , b. hillebrands , m. bailleul , c. fermon , k. y. guslienko , a. n. slavin , d. v. berkov , and n. l. gorn , spin wave wells in nonellipsoidal micrometer size magnetic elements , phys .
lett . * 88 * , 47204 ( 2002 ) .
p. pirro , t. koyama , t. brcher , t. sebastian , b. leven , and b. hillebrands , experimental observation of the interaction of propagating spin waves with nel domain walls in a landau domain structure , appl .
106 * , 232405 ( 2015 ) .
a. v. chumak , p. pirro , a. a. serga , m. p. kostylev , r. l. stamps , h. schultheiss , k. vogt , s. j. hermsdoerfer , b. laegel , p. a. beck , b. hillebrands , spin - wave propagation in a microstructured magnonic crystal , appl .
lett . * 95 * , 262508 ( 2009 ) .
p. a. kolodin , p. kabos , c. e. patton , b. a. kalinikos , n. g. kovshikov , and m. p. kostylev , amplification of microwave magnetic envelope solitons in thin yttrium iron garnet films by parallel pumping , phys .
lett . * 80 * , 1976 ( 1998 ) .
a. a. serga , a. v. chumak , a. andr , g. a. melkov , a. n. slavin , s. o. demokritov , and b. hillebrands , parametrically stimulated recovery of microwave signal stored in standing spin - wave modes of a magnetic film , phys .
99 , 227202 ( 2007 ) .
t. brcher , f. heussner , p. pirro , t. fischer , m. geilen , b. heinz , b. lgel , a. a. serga , and b. hillebrands , time- and power - dependent operation of a parametric spin - wave amplifier , appl .
. lett . * 105 * , 232409 ( 2014 ) .
t. brcher , p. pirro , f. heussner , a. a. serga , and b. hillebrands , localized parallel parametric generation of spin waves in a ni@xmath3fe@xmath4 waveguide by spatial variation of the pumping field , appl .
. lett . * 104 * , 092418 ( 2014 ) .
t. brcher , p. pirro , t. meyer , f. heussner , b. lgel , a. a. serga , and b. hillebrands , parallel parametric amplification of coherently excited propagating spin waves in a microscopic ni@xmath3fe@xmath4 waveguide , appl .
. lett . * 104 * , 202408 ( 2014 ) .
t. brcher , f. heussner , p. pirro , t. meyer , t. fischer , m. geilen , b. heinz , b. lgel , a. a. serga , and b. hillebrands , phase - to - intensity conversion of magnonic spin currents and application to the design of a majority gate , submitted to : scientific reports .
y. tserkovnyak , a. brataas , and g. e. w. bauer , spin pumping and magnetization dynamics in metallic multilayers , phys .
b * 66 * , 224403 ( 2002 ) .
j. e. hirsch , spin hall effect , phys .
lett . * 83 * , 1834 - 1837 ( 1999 ) .
l. feiler , k. sentker , m. brinker , n. kuhlmann , f .- u .
stein , and g. meier , inverse spin - hall effect voltage generation by nonlinear spin - wave excitation , phys .
b * 93 * , 064408 ( 2016 ) . a. v. chumak , a. a. serga , m. b. jungfleisch , r. neb , d. a. bozhko , v. s. tiberkevich , and b. hillebrands , direct detection of magnon spin transport by the inverse spin hall effect , appl . phys
. lett . * 100 * , 082405 ( 2012 ) . c. bi , x. fan , l. pan , x. kou , j. wu , q. yang , h. zhang , and j. q. xiao , spin wave resonance detection using magnetic tunnel junction structure , appl .
. lett . * 99 * , 192501 ( 2011 ) .
k. walus , g. schulhof , g. a. jullien , r. zhang , and w. wang , circuit design based on majority gates for applications with quantum - dot cellular automata , proceedings of the 38th asilomar conference on signals , systems , and computers , 2004 , vol .
2 , p. 1354 .
s. klingler , p. pirro , t. brcher , b. leven , b. hillebrands , and a. v. chumak , design of a spin - wave majority gate employing mode selection , appl .
. lett . * 105 * , 152410 ( 2014 ) .
s. klingler , p. pirro , t. brcher , b. leven , b. hillebrands , and a. v. chumak , spin - wave logic devices based on isotropic forward volume magnetostatic waves , appl .
* 106 * , 212406 ( 2015 ) .
i. p. radu , o. zografos , a. vaysset , f. ciubotaru , j. yan , j. swerts , d. radisic , b. briggs , b. soree , m. manfrini , m. ercken , c. wilson , p. raghavan , s. sayan , c. adelmann , a. thean , l. amar , p .- e .
gaillardon , g. de micheli , d. e. nikonov , s. manipatruni , and i. a. young , spintronic majority gates , 2015 ieee international electron devices meeting ( iedm ) , washington , dc , pp .
32.5.1 - 32.5.4 , dec .
2015 .
a. ruiz - calaforra , t. brcher , v. lauer , p. pirro , b. heinz , m. geilen , a. chumak , a. conca , b. leven , and b. hillebrands , the role of the non - magnetic material in spin pumping and magnetization dynamics in nife and cofeb multilayer systems , j. appl . phys . * 106 * , 163901 ( 2015 ) . k. di , v. l. zhang , h. s. lim , s. c. ng , m. h. kuok , j. yu , j. yoon , x. qiu , and h. yang , direct observation of the dzyaloshinskii - moriya interaction in a pt / co / ni film , phys .
lett . * 114 * , 047201 ( 2015 ) .
moon , s .-
seo , k .- j .
lee , k .- w .
kim , j. ryu , h .- w .
lee , r. d. mcmichael , m. d. stiles , spin - wave propagation in the presence of interfacial dzyaloshinskii - moriya interaction , physical review b * 88 * , 184404 ( 2013 ) .
m. kostylev , interface boundary conditions for dynamic magnetization and spin wave dynamics in a ferromagnetic layer with the interface dzyaloshinskii - moriya interaction , j. appl . phys . * 115 * , 233902 ( 2014 )
. h. t. nembach , j. m. shaw , m. weiler , e. ju , and t. j. silva , linear relation between heisenberg exchange and interfacial dzyaloshinskii
moriya interaction in metal films , nat .
phys . * 11 * , 825 ( 2015 ) .
a. a. stashkevich , m. belmeguenai , y. roussign , s. m. cherif , m. kostylev , m. gabor , d. lacour , c. tiusan , and m. hehn , experimental study of spin - wave dispersion in py / pt film structures in the presence of an interface dzyaloshinskii - moriya interaction , phys .
b * 91 * , 214409 ( 2015 ) . k. von bergmann , a. kubetzka , o. pietzsch , and r. wiesendanger , interface - induced chiral domain walls , spin spirals and skyrmions revealed by spin - polarized scanning tunneling microscopy , j. phys : condens .
matter * 26 * , 394002 ( 2014 ) .
o. boulle , j. vogel , h. yang , s. pizzini , d. de souza chaves , a. locatelli , t. onur mente , a. sala , l. d. buda - prejbeanu , o. klein , m. belmeguenai , y. roussign , a. stashkevich , s. mourad chrif , l. aballe , m. foerster , m. chshiev , s. auffret , i. m. miron , and g. gaudin , room - temperature chiral magnetic skyrmions in ultrathin magnetic nanostructures , nat .
nano . * 11 * , 449 ( 2016 ) .
o. boulle , s. rohart , l. d. buda - prejbeanu , e. ju , i. m. miron , s. pizzini , j. vogel , g. gaudin , and a. thiaville , domain wall tilting in the presence of the dzyaloshinskii - moriya interaction in out - of - plane magnetized magnetic nanotracks , phys .
lett . * 111 * , 217203 ( 2013 ) .
e. martinez , s. emori , and g. s. d. beach , current - driven domain wall motion along high perpendicular anisotropy multilayers : the role of the rashba field , the spin hall effect , and the dzyaloshinskii - moriya interaction , appl .
. lett . * 103 * , 072406 ( 2013 ) .
n. mikuszeit , o. boulle , i. m. miron , k. garello , p. gambardella , g. gaudin , and l. d. buda - prejbeanu , spin - orbit torque driven chiral magnetization reversal in ultrathin nanostructures , phys .
b * 92 * , 144424 ( 2015 ) .
a. hrabec , n. a. porter , a. wells , m. j. benitez , g. burnell , s. mcvitie , d. mcgrouther , t. a. moore , and c. h. marrows , measuring and tailoring the dzyaloshinskii - moriya interaction in perpendicularly magnetized thin films , phys .
b * 90 * , 020402(r ) ( 2014 ) .
j. m. lee , c. jang , b .-
min , s .- w .
lee , k .- j .
lee , and j. chang , all - electrical measurement of interfacial dzyaloshinskii - moriya interaction using collective spin - wave dynamics , nano lett .
* 16 * ( 1 ) , 62 ( 2016 ) .
s. i. kiselev , j. c. sankey , i. n. krivorotov , n. c. emley , r. j. schoelkopf , r. a. buhrman , and d. c. ralph , microwave oscillations of a nanomagnet driven by a spin - polarized current , nature * 425 * , 380 ( 2003 ) .
w. h. rippard , m. r. pufall , s. kaka , s. e. russek , and t. j. silva , direct - current induced dynamics in co@xmath288fe@xmath324/ni@xmath286fe@xmath325 point contacts , phys .
lett . * 92 * , 027201 ( 2004 ) .
a. dussaux , b. georges , j. grollier , v. cros , a. v. khvalkovskiy , a. fukushima , m. konoto , h. kubota , k. yakushiji , s.
yuasa , k. a. zvezdin , k. ando , and a. fert , large microwave generation from current - driven magnetic vortex oscillators in magnetic tunnel junctions , nat . commun . * 1 * , 8 ( 2010 ) .
m. quinsat , v. tiberkevich , d. gusakova , a. slavin , j. f. sierra , u. ebels , l. d. buda - prejbeanu , b. diny , m .- c .
cyrille , a. zelster , and j. a. katine , linewidth of higher harmonics in a nonisochronous auto - oscillator : application to spin - torque nano - oscillators , phys .
b * 86 * , 104418 ( 2012 ) .
m. evelt , v. e. demidov , v. bessonov , s. o. demokritov , j. l. prieto , m. mu@xmath320 oz , j. ben youssef , v. v. naletov , g. de loubens , o. klein , m. collet , k. garcia - hernandez , p. bortolotti , v. cros , and a. anane , high - efficiency control of spin - wave propagation in ultra - thin yttrium iron garnet by the spin - orbit torque , appl .
108 * , 172406 ( 2016 ) .
a. ruotolo , v. cros , b. georges , a. dussaux , j. grollier , c. deranlot , r. guillemet , k. bouzehouane , s. fusil and a. fert , phase - locking of magnetic vortices mediated by antivortices , nat . nano .
* 4 * , 528 ( 2009 ) .
s. sani , j. persson , s. m. mohseni , y. pogoryelov , p. k. muduli , a. eklund , g. malm , m. kll , a. dmitriev , and j. kerman , mutually synchronized bottom - up multi - nanocontact spin
torque oscillators , nat .
commun . * 4 * , 2731 ( 2013 ) .
h. schultheiss , f. ciubotaru , a. laraoui , s. j. hermsdoerfer , b. obry , a. a. serga , x. janssens , m. van kampen , l. lagae , a. n. slavin , b. leven , and b. hillebrands , direct current control of three magnon scattering processes in spin - valve nanocontacts , phys .
lett . * 103 * , 157202 ( 2009 ) .
a. a. tulapurkar , y. suzuki , a. fukushima , h. kubota , h. maehara , k. tsunekawa , d. d. djayaprawira , n. watanabe , and s.
yuasa , spin - torque diode effect in magnetic tunnel junctions , nature ( london ) * 438 * , 339 ( 2005 ) . j. c. sankey , p. m. braganca , a. g. f. garcia , i. n. krivorotov , r. a. buhrman , and d. c. ralph , spin - transfer - driven ferromagnetic resonance of individual nanomagnets , phys . rev . lett . * 96 * , 227601 ( 2006 ) .
v. e. demidov , s. urazhdin , v. tiberkevich , a. n. slavin , and s. o. demokritov , control of spin - wave emission from spin - torque nano - oscillators by microwave pumping , phys .
b * 83 * , 060406(r ) ( 2011 ) .
a. hamadeh , n. locatelli , v. v. naletov , r. lebrun , g. de loubens , j. grollier , o. klein , and v. cros , perfect and robust phase - locking of a spin transfer vortex nano - oscillator to an external microwave source , appl .
. lett . * 104 * , 022408 ( 2014 ) .
a. dussaux , a. v. khvalkovskiy , j. grollier , v. cros , a. fukushima , m. konoto , h. kubota , k. yakushiji , s.
yuasa , k. ando , and a. fert , phase locking of vortex based spin transfer oscillators to a microwave current , appl .
. lett . * 98 * , 132506 ( 2011 ) .
p. bortolotti , e. grimaldi , a. dussaux , j. grollier , v. cros , c. serpico , k. yakushiji , a. fukushima , h. kubota , r. matsumoto , and s.
yuasa , parametric excitation of magnetic vortex gyrations in spin - torque nano - oscillators , phys .
b * 88 * , 174417 ( 2013 ) .
s. y. martin , n. de mestier , c. thirion , c. hoarau , y. conraux , c. baraduc , and b. diny , parametric oscillator based on nonlinear vortex dynamics in low - resistance magnetic tunnel junctions , phys .
b * 84 * , 144434 ( 2011 ) .
r. lebrun , a. jenkins , a. dussaux , n. locatelli , s. tsunegi , e. grimaldi , h. kubota , p. bortolotti , k. yakushiji , j. grollier , a. fukushima , s. yuasa , and v. cros , understanding of phase noise squeezing under fractional synchronization of a nonlinear spin transfer vortex oscillator , phys .
lett . * 115 * , 017201 ( 2015 ) .
b. georges , j. grollier , m. darques , v. cros , c. deranlot , b. marcilhac , g. faini , and a. fert , coupling efficiency for phase locking of a spin transfer nano - oscillator to a microwave current , phys .
* 101 * , 017201 ( 2008 ) .
m. quinsat , j. f. sierra , i. firastrau , v. tiberkevich , a. slavin , d. gusakova , l. d. buda - prejbeanu , m. zarudniev , j .-
michel , u. ebels , b. dieny , m .- c .
cyrille , j. a. katine , d. mauri , and a. zeltser , injection locking of tunnel junction oscillators to a microwave current , appl .
. lett . * 98 * , 182503 ( 2011 ) .
d. tiwari , n. sisodia , r. sharma , p. drrenfeld , j. kerman , and p. k. muduli , enhancement of spin - torque diode sensitivity in a magnetic tunnel junction by parametric synchronization , appl .
. lett . * 108 * , 082402 ( 2016 ) .
v. e. demidov , h. ulrichs , s. v. gurevich , s. o. demokritov , v. s. tiberkevich , a. n. slavin , a. zholud , and s. urazhdin , synchronization of spin hall nano - oscillators to external microwave signals , nat . commun . * 5 * , 3179 ( 2014 ) .
v. lauer , d. a. bozhko , t. brcher , p. pirro , v. i. vasyuchka , a. a. serga , m. b. jungfleisch , m. agrawal , yu .
v. kobljanskyj , g. a. melkov , c. dubs , b. hillebrands , and a. v. chumak , spin - transfer torque based damping control of parametrically excited spin waves in a magnetic insulator , appl .
. lett . * 108 * , 012402 ( 2016 ) . | magnonics and magnon spintronics aim at the utilization of spin waves and magnons , their quanta , for the construction of wave - based logic networks via the generation of pure all - magnon spin currents and their interfacing with electrical charge transport . the promise of efficient parallel data processing and low power consumption renders this field one of the most promising research areas in spintronics . in this context , the process of parallel parametric amplification , i.e. , the conversion of microwave photons into magnons at one half of the microwave frequency , has proven to be a versatile tool to excite and to manipulate spin waves .
its beneficial and unique properties such as frequency and mode - selectivity , the possibility to excite spin waves in a wide wavevector range and the creation of phase - correlated wave pairs , have enabled the achievement of important milestones like the magnon bose einstein condensation and the cloning and trapping of spin - wave packets .
parallel parametric amplification , which allows for the selective amplification of magnons while conserving their phase is , thus , one of the key methods of spin - wave generation and amplification .
the application of parallel parametric amplification to cmos - compatible micro- and nanostructures is an important step towards the realization of magnonic networks .
this is motivated not only by the fact that amplifiers are an important tool for the construction of any extended logic network but also by the unique properties of parallel parametric amplification .
in particular , the creation of phase - correlated wave pairs allows for rewarding alternative logic operations such as a phase - dependent amplification of the incident waves .
recently , the successful application of parallel parametric amplification to metallic microstructures has been reported which constitutes an important milestone for the application of magnonics in practical devices .
it has been demonstrated that parametric amplification provides an excellent tool to generate and to amplify spin waves in these systems in a wide wavevector range . in particular
, the amplification greatly benefits from the discreteness of the spin - wave spectra since the size of the microstructures is comparable to the spin - wave wavelength .
this opens up new , interesting routes of spin - wave amplification and manipulation . in this review
, we will give an overview over the recent developments and achievements in this field .
spin waves , parametric amplification , micro- and nanostructures |
the einstein s general relativity ( gr ) is a relativistic theory of gravity where the graviton is a massless spin-2 particle @xcite .
gr is the current description of gravitation in modern physics and has significantly astrophysical implications .
nevertheless , whether there exist a consistent extension of gr by a mass term is a basic challenge of classical field theory , since the open questions such as the old cosmological constant problem and the origin of the late - time acceleration of the universe remain behind the puzzles at the interface between gravity / cosmology and particle physics . in general , by adding generic mass terms for the gravitons on the given background usually brings various instabilities for the gravitational theories , sometimes on the nonlinear level .
a new nonlinear massive gravity theory was proposed by de rham , gabadadze and tolley ( drgt ) @xcite , where the boulware - deser ghost @xcite was eliminated by introducing higher order interaction terms in the action . until now
, a nontrivial black hole solution with a ricci flat horizon in four - dimensional drgt massive gravity has been constructed by vegh @xcite .
the spherically symmetric solutions for drgt were also addressed in @xcite , the corresponding charged black hole solution was found in @xcite , and its bi - gravity extension was found in @xcite . moreover , the charged ads black hole solution in the higher - dimensional drgt massive gravity , and its corresponding thermodynamics and phase structure in the grand canonical and canonical ensembles were also presented in @xcite .
recently , the study of thermodynamics in ads black holes has been generalized to the extended phase space , where the cosmological constant is regarded as a variable and also identified with thermodynamic pressure @xcite @xmath3 in the geometric units @xmath4 . here
@xmath5 stands for the spacetime dimensions and @xmath6 denotes the ads radius . in this case , the variation of the cosmological constant is included in the first law of black hole thermodynamics , which ensures the consistency between the first law of black hole thermodynamics and the smarr formula . in @xcite , it was pointed out that the extended phase space can be interpreted as an rg - flow in the space of field theories , where isotherm curves codify how the number of degrees of freedom @xmath7 ( or the central charge @xmath8 ) runs with the energy scale .
moreover , the variation of cosmological constant could be corresponded to variation of number of the colors in yang - mills theory residing on the boundary spacetime @xcite . in the extended phase space ,
the charged ads black hole admits a more direct and precise coincidence between the first order small / large black holes ( sbh / lbh ) phase transition and the van der waals liquid - gas phase transition , and both systems share the same critical exponents near the critical point @xcite . as a result , the analogy between the charged ads black hole and the van der waals system becomes more complete .
more discussions were found as well in @xcite . in this direction , some investigations for thermodynamics of ads black holes in the drgt massive gravity have been generalized to the extended phase space @xcite . in these models
, it revealed the existence of first - order sbh / lbh phase transition reminiscent of the liquid / gas transition of the van der waals fluid . on the other hand , the four - dimensional born - infeld - ads @xcite , and higher - dimensional singly spinning ads black holes @xcite have been demonstrated the peculiar behavior of intermediate / small / large black hole phase transitions reminiscent of reentrant phase transitions ( rpts ) observed for multicomponent fluid systems , ferroelectrics , gels , liquid crystals , and binary gases , e.g.,@xcite .
a system undergoes an rpt if a monotonic variation of any thermodynamic quantity results in two ( or more ) phase transitions such that the final state is macroscopically similar to the initial state . in this paper
, we will concentrate on the critical behaviors of higher - dimensional ads black holes in the drgt massive gravity , and also report the finding of interesting reentrant phase transition .
this paper is organized as follows . in sect
. [ 2s ] , we review the thermodynamics of massive gravity black holes in the extended phase space . in sect .
[ 3s ] , we study the critical behavior of higher - dimensional ads black hole in context of @xmath0 criticality and phase diagrams .
we end the paper with closing remarks in sect .
we start with the action of higher - dimensional massive gravity in presence of a negative cosmological constant @xmath9},\label{action}\end{aligned}\ ] ] where the last four terms are the massive potential associate with graviton mass , @xmath10 are constants and @xmath11 is a fixed rank-2 symmetric tensor .
moreover , @xmath12 are symmetric polynomials of the eigenvalues of the @xmath13 matrix @xmath14 @xmath15,\nonumber\\ { \cal u}_{2}&=&[{\cal k}]^2-[{\cal k}^2],\nonumber\\ { \cal u}_{3}&=&[{\cal k}]^3 - 3[{\cal k}][{\cal k}^2]+2[{\cal k}^3],\nonumber\\ { \cal u}_{4}&=&[{\cal k}]^4 - 6[{\cal k}^2][{\cal k}]^2 + 8[{\cal k}^3][{\cal k } ] + 3[{\cal k}^2]^2 - 6[{\cal k}^4].\label{ac}\end{aligned}\ ] ] the square root in @xmath16 is understood as the matrix square root , ie . ,
@xmath17 , and the rectangular brackets denote traces @xmath18={\cal k}^{\mu}_{~\mu}$ ] .
consider the metric of @xmath5-dimensional spacetime in the following form @xmath19 where @xmath20 is the line element for an einstein space with constant curvature @xmath21 .
the constant @xmath22 characterizes the geometric property of hypersurface , which takes values @xmath23 for flat , @xmath24 for negative curvature and @xmath25 for positive curvature , respectively . by using the reference metric @xmath26 the metric function @xmath27 is obtained as @xcite @xmath28 here
@xmath29 is a positive constant , @xmath30 is the volume of space spanned by coordinates @xmath31 , and @xmath32 is the black hole mass .
it is worth to note that the terms @xmath33 and @xmath34 only appear in the black hole solutions for @xmath35 and @xmath1 , respectively @xcite . in terms of the horizon radius @xmath36 , the mass @xmath32 , hawking temperature @xmath37 and entropy @xmath38 of black holes
can be written as @xmath39,\nonumber\\ t&=&\frac{f'(r_+)}{4\pi } = \frac{1}{4\pi r_+}\left[(d-3)k+\frac{16\pi p}{d-2}r_+^2+c_0c_1m^2r_{+ } + ( d-3)c_0 ^ 2c_2m^2\right.\nonumber\\ & & \left.+\frac{(d-3)(d-4)c_0 ^ 3c_3m^2}{r_+ } + \frac{(d-3)(d-4)(d-5)c_0 ^ 4c_4m^2}{r_+^2}\right],\nonumber\\ s&=&\frac{v_{d-2}}{4 } r_+^{d-2}. \label{eq:5a}\end{aligned}\ ] ] the black hole mass @xmath32 can be considered as the enthalpy rather than the internal energy of the gravitational system . moreover ,
the first law of black hole thermodynamics and smarr relation are given by @xcite @xmath40 where @xmath41 denotes the thermodynamic volume and equals to @xmath42 .
for further convenience , we would like to denote @xmath43 here @xmath44 denotes the shifted temperature and can be negative according to the value of @xmath45 . then , the equation of state of the black hole can be obtained from eq .
( [ eq:5a ] ) @xmath46.\label{eos}\end{aligned}\ ] ] to compare with the van der waals fluid equation , we can translate the
geometric " equation of state to physical one by identifying the specific volume @xmath47 of the fluid with the horizon radius of the black hole as @xmath48 .
evidently , the specific volume @xmath47 is proportional to the horizon radius @xmath36 , therefore we will just use the horizon radius in the equation of state for the black hole hereafter in this paper .
we know that the critical point occurs when @xmath49 has an inflection point , @xmath50 where the subscript stands for the quantities at the critical point .
the critical shifted temperature is obtained as @xmath51,\label{teos}\end{aligned}\ ] ] and the equation for the critical horizon radius @xmath52 is given by @xmath53 one can easily find that in four - dimensional spacetime @xmath54 , the absence of positive solution of eq .
( [ eq:7a ] ) indicates that no criticality can happen @xcite .
we further discuss the critical behaviors of higher - dimensional @xmath55 ads black hole when @xmath56 and @xmath57 . when setting @xmath58 , one have @xmath59 note that the critical behavior occurs only when @xmath60 and @xmath61 .
we can easily find an interesting relation among the critical pressure @xmath62 , temperature @xmath63 and horizon radius @xmath52 @xmath64 for @xmath65 , the expressions ( eqs .
( [ eq:8a])([eq:9a ] ) ) reduce to that described in @xcite with regard to the case of @xmath66 , which only appears for @xmath1 , the direct solution of eq .
( [ eq:7a ] ) reads as @xmath67\equiv\frac{\chi_{\pm}}{2w_2}\label{eq:18a}\end{aligned}\ ] ] when @xmath68 . in this case ,
@xmath69 and @xmath70 correspond to the @xmath71 and @xmath72 branches , respectively .
the condition of @xmath73 crucially depends on the dimension of spacetime and values of @xmath74 , @xmath75 and @xmath76 . for @xmath60
, the positivity of solution @xmath69 leads to @xmath77 and @xmath78 or @xmath61 . in order that @xmath70 be positive
, it requires an additional constraint @xmath61 and @xmath79 .
moreover , by substituting the solutions @xmath80 ( eq . ( [ eq:18a ] ) ) into eqs .
( [ eos])([teos ] ) , we obtain @xmath81}{\chi_{\pm}^3},\label{tw2}\\ & & p_{c1,2}=\frac{\pm4(d-2)(d-3)(d-4)w_2 ^ 3\left[6(d-5)w_2w_4 + w_3\chi_{\pm}\right]}{\chi_{\pm}^4}.\label{pw2}\end{aligned}\ ] ] due to the shifted temperature @xmath82 can be negative , here we only evaluate the results of @xmath83 .
when @xmath84 , we shall keep @xmath77 and @xmath78 or @xmath61 . on the other hand , taking @xmath61 , @xmath79 and @xmath85 lead to @xmath86 . as a result , two critical points ( @xmath73 and @xmath87 ) will appear in the range of @xmath85 with @xmath61 and @xmath79 . in case of @xmath88 , there is only one critical point ( @xmath89 and @xmath84 ) for @xmath77 , @xmath78 and @xmath90 .
we summarize the critical points in the table [ sphericcase ] .
the corresponding @xmath91 diagrams for @xmath92 are displayed in figs.[fig1 ] and [ fig2 ] .
.the behaviors of critical points for different values of coupling constants in higher - dimensional spacetime . [ cols= " < , < , < , < ,
< , < " , ] to study the possible phase transitions in the system , let us now turn to the expression for the gibbs free energy . the behavior of free energy @xmath93 is important to determine the thermodynamic phase transition .
the free energy @xmath93 obeys the thermodynamic relation @xmath94.\label{eq:20a}\end{aligned}\ ] ] here @xmath36 is understood as a function of pressure and temperature , @xmath95 , via equation of state [ eq .
( [ eos ] ) ] . in the range of @xmath85 with @xmath61 and @xmath79 ,
the behavior of @xmath93 in the six dimensional spacetime is depicted in fig.[fig3](a ) .
we has one physical ( with positive pressure ) critical point and the corresponding first order sbh / lbh phase transition .
this phase transition occurs for @xmath96 and terminates at @xmath97 .
there also exists a certain range of temperatures , @xmath98 , for which the global minimum of @xmath93 is discontinuous , see fig.[fig3](b ) . in this range of temperatures , two separate branches of intermediate size and small size black holes
co - exist .
they are separated by a finite jump in @xmath93 , which is so - called `` zeroth - order phase transition '' .
this phenomenon is also seen in superfluidity and superconductivity @xcite .
this novel situation can be also clearly illustrated in the @xmath99 diagrams in fig.[fig4 ] .
there is the expected sbh / lbh line of coexistence , which ends in a critical point @xmath100 and terminates at @xmath101 .
especially , a `` triple point '' between the small , intermediate , and large black holes appears in the point @xmath101 . for @xmath98
, there is a new ibh / sbh line of coexistence that terminates in another critical point @xmath102 .
the range for the rpt is quite narrow and must be determined numerically .
taking @xmath103 , @xmath104 , @xmath105 and @xmath92 , we obtain @xmath106 in fig .
[ fig5 ] , we also plot the behaviors of gibbs free energy of six dimensional ads black hole for three other cases , showed in table [ sphericcase ] .
one can see that the @xmath93 surface demonstrates the characteristic `` swallow tail '' behavior , which indicates that there is a van der waals like first order sbh / lbh phase transition when @xmath107 in the corresponding system .
moreover , the corresponding @xmath108 diagram ( not shown ) is reminiscent of what was observed for charged black holes in @xcite and is analogous to the van der waals @xmath108 diagram in each case .
in this paper we have studied the thermodynamic behaviors of higher - dimensional ads black hole in the drgt massive gravity .
we discussed this issue in the extended phase space where the cosmological constant appears as the pressure of the thermodynamic system and its conjugate quantity is the thermodynamic volume of the black holes . following the standard thermodynamic techniques , we have written out the equations of state and examined the phase structures . when the coupling coefficients of massive potential satisfy @xmath85 with @xmath61 and @xmath79 , we find that a monotonic lowering of the temperature yields a large - small - large black hole transition , where we refer to the latter `` large '' state as an intermediate black hole ( ibh ) , which is reminiscent of reentrant phase transitions .
moreover , this process is also accompanied by a discontinuity in the global minimum of the gibbs free energy , referred to as a zeroth - order phase transition . for three other cases in table [ sphericcase ] ,
the usual van der waals like small / large black hole phase transition occur when coupling coefficients of massive potential adopt some proper values in the higher - dimensional spacetime .
in addition , the charged black hole @xcite , born - infeld black hole @xcite , and black hole in the maxwell and yang - mills fields @xcite have been recently constructed in gauss - bonnet massive gravity .
it has also showed the existence of van der waals like first order sbh / lbh phase transition in these models .
it would be interesting to extend our discussion to these black hole solutions and find out whether the reentrant phase transition can appear .
the work is supported by the national natural science foundation of china ( grant no.11605152 , no.11275099 , and no.11647050 ) , and natural science foundation of jiangsu province under grant no.bk20160452 .
s. n. gupta , `` gravitation and electromagnetism , '' phys .
rev . * 96 * , 1683 ( 1954 ) .
s. weinberg , `` photons and gravitons in perturbation theory : derivation of maxwell s and einstein s equations , '' phys .
rev . * 138 * , b988 ( 1965 ) .
r. p. feynman , f. b. morinigo , w. g. wagner and b. hatfield , `` feynman lectures on gravitation , '' reading , usa : addison - wesley ( 1995 ) 232 p. ( the advanced book program ) c. de rham and g. gabadadze , `` generalization of the fierz - pauli action , '' phys .
d * 82 * , 044020 ( 2010 ) [ arxiv:1007.0443 [ hep - th ] ] . c. de rham , g. gabadadze and a. j. tolley , `` resummation of massive gravity , '' phys .
lett . * 106 * , 231101 ( 2011 ) [ arxiv:1011.1232 [ hep - th ] ] .
k. hinterbichler , `` theoretical aspects of massive gravity , '' rev
phys . * 84 * , 671 ( 2012 ) [ arxiv:1105.3735 [ hep - th ] ] .
d. g. boulware and s. deser , `` can gravitation have a finite range ?
, '' phys .
d * 6 * , 3368 ( 1972 ) .
d. vegh , `` holography without translational symmetry , '' arxiv:1301.0537 [ hep - th ] .
a. adams , d. a. roberts and o. saremi , `` hawking - page transition in holographic massive gravity , '' phys .
d * 91 * , no . 4 , 046003 ( 2015 ) [ arxiv:1408.6560 [ hep - th ] ] .
t. m. nieuwenhuizen , `` exact schwarzschild - de sitter black holes in a family of massive gravity models , '' phys .
d * 84 * , 024038 ( 2011 ) [ arxiv:1103.5912 [ gr - qc ] ] .
r. brito , v. cardoso and p. pani , `` black holes with massive graviton hair , '' phys .
d * 88 * , 064006 ( 2013 ) [ arxiv:1309.0818 [ gr - qc ] ] .
l. berezhiani , g. chkareuli , c. de rham , g. gabadadze and a. j. tolley , `` on black holes in massive gravity , '' phys .
d * 85 * , 044024 ( 2012 ) [ arxiv:1111.3613 [ hep - th ] ] .
e. babichev and a. fabbri , `` a class of charged black hole solutions in massive ( bi)gravity , '' jhep * 1407 * , 016 ( 2014 ) [ arxiv:1405.0581 [ gr - qc ] ] .
r. g. cai , y. p.
hu , q. y. pan and y. l. zhang , `` thermodynamics of black holes in massive gravity , '' phys .
d * 91 * , 024032 ( 2015 ) [ arxiv:1409.2369 [ hep - th ] ] .
b. p. dolan , `` pressure and volume in the first law of black hole thermodynamics , '' class .
* 28 * , 235017 ( 2011 ) [ arxiv:1106.6260 [ gr - qc ] ] .
b. p. dolan , `` the cosmological constant and the black hole equation of state , '' class .
* 28 * , 125020 ( 2011 ) [ arxiv:1008.5023 [ gr - qc ] ] .
e. caceres , p.
h. nguyen and j. f. pedraza , `` holographic entanglement entropy and the extended phase structure of stu black holes , '' jhep * 1509 * , 184 ( 2015 ) [ arxiv:1507.06069 [ hep - th ] ] .
j. d. e. creighton and r. b. mann , `` quasilocal thermodynamics of dilaton gravity coupled to gauge fields , '' phys .
d * 52 * , 4569 ( 1995 ) [ gr - qc/9505007 ] .
g. w. gibbons , r. kallosh and b. kol , `` moduli , scalar charges , and the first law of black hole thermodynamics , '' phys .
lett . * 77 * , 4992 ( 1996 ) [ hep - th/9607108 ] .
d. kubiznak and r. b. mann ,
`` p - v criticality of charged ads black holes , '' jhep * 1207 * , 033 ( 2012 ) [ arxiv:1205.0559 [ hep - th ] ] .
s. gunasekaran , r. b. mann and d. kubiznak , `` extended phase space thermodynamics for charged and rotating black holes and born - infeld vacuum polarization , '' jhep * 1211 * , 110 ( 2012 ) [ arxiv:1208.6251 [ hep - th ] ] . s. h. hendi and m. h. vahidinia , `` p - v criticality of higher dimensional black holes with nonlinear source , '' phys .
d * 88 * , 084045 ( 2013 ) [ arxiv:1212.6128 [ hep - th ] ] .
s. chen , x. liu , c. liu and j. jing , `` @xmath0 criticality of ads black hole in @xmath109 gravity , '' chin .
. lett . * 30 * , 060401 ( 2013 ) [ arxiv:1301.3234 [ gr - qc ] ] .
r. g. cai , l. m. cao , l. li and r. q. yang , `` p - v criticality in the extended phase space of gauss - bonnet black holes in ads space , '' jhep * 1309 * , 005 ( 2013 ) [ arxiv:1306.6233 [ gr - qc ] ] .
s. dutta , a. jain and r. soni , `` dyonic black hole and holography , '' jhep * 2013 * , 60 ( 2013 ) [ arxiv:1310.1748 [ hep - th ] ] .
w. xu , h. xu and l. zhao , `` gauss - bonnet coupling constant as a free thermodynamical variable and the associated criticality , '' eur .
j. c * 74 * , 2970 ( 2014 ) [ arxiv:1311.3053 [ gr - qc ] ] .
d. c. zou , s. j. zhang and b. wang , `` critical behavior of born - infeld ads black holes in the extended phase space thermodynamics , '' phys .
d * 89 * , 044002 ( 2014 ) [ arxiv:1311.7299 [ hep - th ] ] .
h. h. zhao , l. c. zhang , m. s. ma and r. zhao , `` p - v criticality of higher dimensional charged topological dilaton de sitter black holes , '' phys .
d * 90 * , 064018 ( 2014 ) .
n. altamirano , d. kubiznak , r. b. mann and z. sherkatghanad , `` thermodynamics of rotating black holes and black rings : phase transitions and thermodynamic volume , '' galaxies * 2 * , 89 ( 2014 ) [ arxiv:1401.2586 [ hep - th ] ] . j. -x .
mo and w. -b .
liu , `` p - v criticality of topological black holes in lovelock - born - infeld gravity , '' eur .
j. c * 74 * , 2836 ( 2014 ) [ arxiv:1401.0785 [ gr - qc ] ] .
d. c. zou , y. liu and b. wang , `` critical behavior of charged gauss - bonnet ads black holes in the grand canonical ensemble , '' phys .
d * 90 * , 044063 ( 2014 ) [ arxiv:1404.5194 [ hep - th ] ] .
h. xu , w. xu and l. zhao , `` extended phase space thermodynamics for third order lovelock black holes in diverse dimensions , '' eur .
j. c * 74 * , 3074 ( 2014 ) [ arxiv:1405.4143 [ gr - qc ] ] .
w. xu and l. zhao , `` critical phenomena of static charged ads black holes in conformal gravity , '' phys .
b * 736 * , 214 ( 2014 ) [ arxiv:1405.7665 [ gr - qc ] ] .
m. h. dehghani , s. kamrani and a. sheykhi , `` @xmath0 criticality of charged dilatonic black holes , '' phys .
d * 90 * , 104020 ( 2014 ) . c. o. lee , `` the extended thermodynamic properties of taub - nut / bolt - ads spaces , '' phys .
b * 738 * , 294 ( 2014 ) [ arxiv:1408.2073 [ hep - th ] ] .
j. l. zhang , r. g. cai and h. yu , `` phase transition and thermodynamical geometry for schwarzschild ads black hole in ads@xmath110s@xmath111 spacetime , '' jhep * 1502 * , 143 ( 2015 ) [ arxiv:1409.5305 [ hep - th ] ] .
m. zhang , z. y. yang , d. c. zou , w. xu and r. h. yue , `` @xmath0 criticality of ads black hole in the einstein - maxwell - power - yang - mills gravity , '' gen .
* 47 * , 14 ( 2015 ) [ arxiv:1412.1197 [ hep - th ] ] .
m. zhang , r. h. yue and z. y. yang , `` critical behaviors of black holes in an einstein - maxwell gravity with a conformal anomaly , '' chin .
* 32 * , 020401 ( 2015 ) [ arxiv:1412.1201 [ hep - th ] ] .
j. l. zhang , r. g. cai and h. yu , `` phase transition and thermodynamical geometry of reissner - nordstrom - ads black holes in extended phase space , '' phys .
d * 91 * , 044028 ( 2015 ) [ arxiv:1502.01428 [ hep - th ] ] .
y. liu , d. c. zou and b. wang , `` signature of the van der waals like small - large charged ads black hole phase transition in quasinormal modes , '' jhep * 1409 * , 179 ( 2014 ) [ arxiv:1405.2644 [ hep - th ] ] .
b. r. majhi and s. samanta , `` p - v criticality of ads black holes in a general framework , '' arxiv:1609.06224 [ gr - qc ] .
s. h. hendi , b. eslam panah , s. panahiyan and m. s. talezadeh , `` geometrical thermodynamics and p - v criticality of black holes with power - law maxwell field , '' arxiv:1612.00721 [ hep - th ] .
j. xu , l. m. cao and y. p. hu , `` p - v criticality in the extended phase space of black holes in massive gravity , '' phys .
d * 91 * , 124033 ( 2015 ) [ arxiv:1506.03578 [ gr - qc ] ] .
s. h. hendi , b. eslam panah and s. panahiyan , `` einstein - born - infeld - massive gravity : ads - black hole solutions and their thermodynamical properties , '' jhep * 1511 * , 157 ( 2015 ) [ arxiv:1508.01311 [ hep - th ] ] .
m. zhang and w. b. liu , `` coexistent physics of massive black holes in the phase transitions , '' arxiv:1610.03648 [ gr - qc ]
. s. g. ghosh , l. tannukij and p. wongjun , `` a class of black holes in drgt massive gravity and their thermodynamical properties , '' eur .
j. c * 76 * , 3 ( 2016 ) [ arxiv:1506.07119 [ gr - qc ] ] .
p. prasia and v. c. kuriakose , `` quasi normal modes and p - v criticallity for scalar perturbations in a class of drgt massive gravity around black holes , '' gen .
48 * , 89 ( 2016 ) [ arxiv:1606.01132 [ gr - qc ] ] .
n. altamirano , d. kubiznak and r. b. mann , `` reentrant phase transitions in rotating ads black holes , '' phys .
d * 88 * , 101502 ( 2013 ) [ arxiv:1306.5756 [ hep - th ] ] .
t. narayanan and a. kumar , `` reentrant phase transitions in multicomponent liquid mixtures , '' physics reports * 249 * ( 1994 ) .
v. p. maslov , `` zeroth - order phase transitions , '' _ mathematical notes _ * 76 * 697 - 710 ( 2004 ) s. h. hendi , s. panahiyan and b. eslam panah , `` charged black hole solutions in gauss - bonnet - massive gravity , '' jhep * 1601 * , 129 ( 2016 ) [ arxiv:1507.06563 [ hep - th ] ] .
s. h. hendi , g. q. li , j. x. mo , s. panahiyan and b. eslam panah , `` new perspective for black hole thermodynamics in gauss - bonnet - born - infeld massive gravity , '' eur .
j. c * 76 * , 10 ( 2016 ) [ arxiv:1608.03148 [ gr - qc ] ] . k. meng and j. li , `` black hole solution of gauss - bonnet massive gravity coupled to maxwell and yang - mills fields in five dimensions , '' europhys . lett . * 116 * , 10005 ( 2016 ) . | we study the @xmath0 criticality and phase transition in the extended phase space of ads black holes in higher - dimensional drgt massive gravity , treating the cosmological constant as pressure and the corresponding conjugate quantity is interpreted as thermodynamic volume . besides the usual small / large black hole phase transitions
, the interesting thermodynamic phenomena of reentrant phase transitions are observed for all @xmath1 dimensions when the coupling coefficients @xmath2 of massive potential satisfy some certain conditions . |
we undertake the first numerical study of a four dimensional asymptotically free gauge theory with dynamical fermions transforming according to the two index symmetric representation of the gauge group . to be more precise we consider the simplest of these theories with interesting dynamics and phenomenological applications , i.e. a two color gauge theory with two dirac fermions transforming according to the two index symmetric representation . for two colors the two index symmetric coincides with the adjoint representation and the reality of the representation
enhances the quantum flavor symmetry group to @xmath0 .
remarkably this theory coincides with the fermionic sector of the @xmath1 supersymmetric theory with two colors .
recently it has been argued that this theory lies close to a non trivial infrared fixed point @xcite . in this case the coupling constant will run very slowly from the infrared to the ultraviolet ; the coupling is said to _ walk_. by analyzing the phase diagram as function of the number of colors and flavors in @xmath2 gauge theories with dirac fermions in a given arbitrary representation of the gauge group we have shown that this theory is _ minimal _ in the sense that it is the theory with the smallest number of flavors ( above one ) which exhibits such walking dynamics @xcite .
the walking dynamics is expected to be dramatically different than in the qcd case .
this is so since the presence of a nearby i.r .
fixed point should generate an anomalously small mass scale and ensure that long distance quantities are insensitive to the short distance coupling .
the physics of the fixed point theory per se is very interesting and when coupled to non - conformal theories ( such as the standard model ) in the way described recently by georgi @xcite it leads to interesting experimental signatures .
this is so since the presence of a conformal symmetry signals itself experimentally in a way that _ formally _ resembles the production of a non - integer number of massless invisible particles .
the non - integer number is nothing but the scale dimension of the operator , coming from the underlying conformal theory , coupled weakly to the standard model operators .
however , as also stressed by georgi , very little is known about conformal or near - conformal theories in four dimensions because of the complicated nonperturbative dynamics .
our work should be considered as a first step in this direction .
in addition , the emergence of a walking property for just two dirac flavors renders this theory an ideal candidate to break the electroweak theory dynamically @xcite .
this extension of the standard model passes the stringent electroweak precision constraints @xcite .
moreover , it allows for a successful unification of the standard model couplings @xcite and allows us to construct different types of dark matter candidates @xcite .
the lattice results presented in this work support the theoretical expectations , i.e. that the present theory walks . to arrive to this conclusion
we analyze various physical quantities on the lattice and then compare them to the ones for a two color gauge theory with two dirac flavors in the fundamental representation of the gauge group . the latter theory , as it is clear from the phase diagram reviewed later , is very far away from an infrared fixed point .
although our simulations employ dynamical quarks the lattices we use are small so these results should only be taken as indicative of qualitative agreement with the theoretical expectations and encourage one to embark on a more serious study on larger lattices . in the next section we review the expected phase diagram as function of flavors and colors for non supersymmetric asymptotically free @xmath2 gauge theories with @xmath3 dirac fermions in a given representation of the gauge group @xcite .
here it is shown that the model we study in this paper is indeed the theory with the lowest number of dirac flavors ( above one ) able to feature walking .
this feature makes this theory also an ideal candidate when used for breaking the electroweak theory dynamically .
we review the salient features and the notation here .
we then describe the details of our lattice theory and simulation algorithm .
this is followed by a detailed description of our numerical results .
we compare them to the theory with fermions in the fundamental representation .
our results show clear differences between the two theories the symmetric quark theory has a behavior similar to the fundamental quark theory at strong coupling but deviates substantially for weak coupling where we observe substantially lighter hadronic masses .
the phase diagram of strongly coupled theories is relevant both theoretically and phenomenologically . by comparing the dynamics of various strongly coupled theories in different regimes
we acquire a deeper understanding of non - perturbative dynamics .
recently we have completed the analysis of the phase diagram of asymptotically free non supersymmetric gauge theories with at least two dirac fermions in a given arbitrary representation of the gauge group as function of the number of flavors and colors @xcite . with the exceptions of a few isolated higher dimensional representations below nine colors ( fully investigated in @xcite ) the main phase diagram taken from @xcite
is sketched in figure [ ph ] .
the analysis exhausts the phase diagram for gauge theories with dirac fermions in arbitrary representations and it is based on the ladder approximation presented in @xcite .
further studies of the conformal window and its properties can be found in @xcite . in the plot
the shaded areas represent the conformal windows for the fundamental representation ( grey ) , two - index antisymmetric ( blue ) , two - index symmetric ( red ) and adjoint representation ( green ) . for each representation
the upper solid curve represents the loss of asymptotic freedom , the lower curve loss of chiral symmetry breaking .
the dashed curves show the existence of a banks
zaks fixed point @xcite . note how consistently the various representations merge into each other when , for a specific value of @xmath4 , they are actually the same representation .
remarkably the adjoint and the two index - symmetric representation need a very low number of flavors , for an arbitrary number of colors , to be near an infrared fixed point . for any number of colors , in the case of the adjoint representation and for two colors in the case of the symmetric representation the estimated critical lower number of flavors above which the theory is already inside the conformal window and hence no chiral symmetry breaking occurs is @xmath5 @xcite .
since this theory walks already for two flavors and hence has also a nontrivial chiral dynamics it will be denoted the minimal walking theory .
the theoretical estimates of the conformal window presented above need to be tested further .
the very low number of flavors needed to reach the conformal window makes the minimal walking theories amenable to lattice investigations .
we also note that the theory with two colors and two dirac flavors in the fundamental representation is very much below the critical number of flavors needed to develop a non trivial infrared fixed point and hence it features qcd - like dynamics .
new strongly interacting theories can emerge in extensions of the standard model .
for example , to avoid unnaturally large quantum corrections to the mass squared of the elementary higgs one can replace it by a new strongly coupled fermionic sector .
this is the _ technicolor _ mechanism @xcite .
the generation of the masses of the standard model fermions requires extended technicolor interactions . to avoid large flavor - changing neutral currents technicolor theories possessing a sufficient amount of walking @xcite
are needed .
the simplest of such models which also passes the electroweak precision tests ( like , for example the experimental bounds on the oblique parameters ) has fermions in higher dimensional representations of the technicolour gauge group @xcite . we have shown in @xcite that the minimal walking theory is also the minimal walking technicolor theory .
such a theory has also a number of desirable features .
for example , together with a minimal modification of the sm fermionic matter content it yields a high degree of unification , at the one loop level , of the sm couplings @xcite .
straightforward extensions of the minimal walking theory able to accommodate extended technicolor interactions can be constructed @xcite .
the lattice action we employ consists of the usual wilson plaquette term @xmath6 together with the wilson action for two dirac quarks in the symmetric representation @xmath7 where the symmetric links are given by @xmath8 and the matrices @xmath9 are a basis for the symmetric representation @xmath10 we have simulated this theory over a range of gauge couplings @xmath11 and bare quark masses @xmath12 ranging from @xmath13 on @xmath14 lattices using the usual hybrid monte carlo algorithm @xcite .
we have focused on determining the critical line @xmath15 needed to approach the continuum limit and a variety of meson masses and decay constants .
typically we generate between @xmath16 @xmath17 hmc trajectories .
periodic boundary conditions were used for all fields . in tandem with these symmetric quark
runs we have also simulated the theory with fundamental quarks .
this allows us to make comparisons at identical lattice volumes and comparable couplings and masses and helps highlight the essential differences associated with the walking dynamics .
we estimate the hadron masses by suitable fits to corresponding time sliced averaged correlation functions @xmath18 where @xmath19 for the pion and @xmath20 for the rho ( the latter being averaged over spatial directions @xmath21 ) . in practice
we use point sources located at ( odd , odd , odd ) lattice sites on the @xmath22 timeslice .
since our lattices are so small we have simply thrown out the @xmath22 data point and fitted the remaining correlator to a simple hyperbolic cosine of the form @xmath23 to estimate the corresponding meson mass @xmath24 .
consider first the usual case corresponding to taking the quarks to lie in the fundamental representation of the gauge group .
figure [ pion_fund ] shows curves of the pion mass squared ( lattice units ) for fundamental quarks at several values of @xmath25 as a function of the bare quark mass @xmath26 . at each coupling @xmath25
we observe the usual linear variation of @xmath27 with bare quark mass @xmath12 .
we see that the pion mass attains a minimum value for some critical quark mass @xmath15 which moves towards @xmath28 as @xmath25 increases .
this is similar to the situation in qcd .
notice that the minimal pion mass _ increases _ with @xmath25 as a result of finite volume effects .
contrast this behavior with the analogous curves for the symmetric representation in fig .
[ pion_sym ] for the same bare couplings @xmath25 and over the same range of bare quark mass @xmath12 .
again , a goldstone behavior is seen over some range of bare quark mass with a critical quark mass @xmath15 that runs towards the origin as @xmath25 increases .
notice though that this linear regime appears distorted at larger @xmath25 which we interpret to be the result of strong finite volume effects .
this is consistent with the small values for the pion mass @xmath29 observed there . indeed
the minimal pion mass is a factor of @xmath30 times smaller for the symmetric pions than pions built from fundamental quarks .
the appearance of this light scale is the first indication that the dynamics of this theory could be quite different from qcd . to make this clearer notice that the meson mass in lattice units is given in general by a pure number times a non - perturbative lattice scale parameter @xmath31 which is related to the beta - function in the usual way @xmath32 this scale is small if the theory lies close to a zero of the beta - function
furthermore , finite size effects are governed by the quantity @xmath33 . for fixed @xmath34 the theory with the smaller @xmath31
might be expected to exhibit larger finite size effects as we observe for the symmetric quark theory .
however , perhaps the most striking difference between these symmetric plots and their fundamental cousins is the _ non - monomtonic _ behavior of the pion mass with increasing @xmath25 . for small @xmath25 the minimal pion mass increases with @xmath25 in a way which is similar to the fundamentals .
however for @xmath35 the pion mass _ decreases _ with further increases in @xmath25 .
this is consistent with the existence of a beta - function ( at zero quark mass ) which resembles that of qcd at strong coupling but is small for couplings less than some characteristic value .
such a beta - function corresponds to the case of walking dynamics .
fig [ walkbeta ] gives a schematic picture of the beta - function and related coupling constant evolution for such a walking theory .
[ cols="^,^ " , ] these differences are also seen in the rho meson .
figure [ rho_fund ] and fig . [ rho_sym ] show the rho mass ( in lattice units ) as a function of quark mass over the same range of couplings @xmath25 for both fundamental and symmetric quarks . again a minimal rho mass is seen which coincides approximately with the minimal pion mass and moves towards the origin as @xmath25 increases . in our later analysis
we have used the minimal rho mass rather than that of the pion when trying to get an estimate of the critical line . in the case of fundamental quarks the observed rho mass increases montonically with @xmath25 but again for symmetrics the minimal rho mass shows a non - monotonic behavior with @xmath36 initially increasing and then falling as @xmath25 increases .
the similar behavior of the rho mass to the pion mass for symmetric quarks lends strong support to the walking hypothesis since the near conformal dynamics will suppress the mass of all hadron states - not just the pi meson . to compute the pion decay constant
we have also measured the axial correlator defined by @xmath37 as for the hadron correlators we discard the @xmath22 data point and fit the remaining function to a hyperbolic sine function @xmath38 .
we observe that the pion mass extracted from this fit is indeed consistent with that derived from the pion correlator in the physical regime @xmath39 .
an expression for pion decay constant ( in lattice units ) is then given by @xcite @xmath40 we do not attempt in this paper to determine the renormalization constant @xmath41 and hence all results for @xmath42 are only derived up this factor . figure [ fpi_fund ] shows this quantity for a range of @xmath25 and quark masses @xmath43 for the theory with quarks in the fundamental .
we see that @xmath42 increases as the critical line is approached with @xmath42 taking on a maximal value there before falling rapidly to zero in the phase with @xmath44 .
the value of @xmath42 along the critical line falls as @xmath25 is increased and the lattice spacing reduced .
equivalent data for symmetric quarks is shown in fig .
[ fpi_sym ] .
notice that the maximal value of @xmath42 is about twice that found for fundamentals ( where both are measured in lattice units ) but otherwise the picture is qualitatively similar . presumably the fact that @xmath42 is _ larger _ rather than smaller for symmetric quarks as compared to fundamentals , contrary to naive expectations , could be related to the differing renormalization constants @xmath41 in the two cases . to examine the continuum limit in greater detail
it is useful to examine dimensionless quantities as we scan in @xmath25 and @xmath12 .
thus we have examined the ratio @xmath45 which is plotted in figs .
[ fpi_rho_ratio_fund ] and [ fpi_rho_ratio_sym ] over the same range of couplings and masses . for both types of representations
we see that @xmath45 increases as the quark mass is tuned towards the critical line .
furthermore there is evidence that for large enough quark mass the curves for different bare coupling @xmath25 lie on top of one another - this was not true for the bare @xmath42 .
notice also that the values of @xmath46 for the two representations are much closer in this region of parameter space . to extract continuum physics we should look for scaling
dimensionless quantities should become independent of coupling along the critical line as @xmath47 .
we see some evidence for this away from the critical line where the sets of curves for different bare coupling @xmath25 lie on top of one another .
this approximate collapse of the data onto a single scaling curve at intermediate quark masses happens for both representations of dynamical quark .
however , if we look at the values of @xmath46 for fundamental quarks near the critical quark mass @xmath48 we see that they _
decrease _ with increasing @xmath25 .
this is to be attributed to finite volume effects as the lattice spacing decreases the physical volume also decreases since we use a fixed lattice volume .
the rising rho mass with physical volume then leads to a falling value for this ratio .
notice also that the @xmath45 falls _ below _ the scaling curve as @xmath49 .
this should be contrasted with the case of the symmetric quarks where the situation is reversed the value of the ratio @xmath45 increases above the scaling curve as the quark mass is tuned towards its critical value . figs .
[ pi_rho_ratio_fund ] and [ pi_rho_ratio_sym ] show plots of the dimensionless ratio of pion to rho mass for the two theories .
both show values which decrease from unity as the critical line is approached .
there is some evidence that the theory with symmetric quarks exhibits a value of the ratio closer to unity as the critical line is approached and that this limiting value is rather insensitive to the bare coupling for large @xmath25 a hint perhaps of walking dynamics .
in this paper we report on the results of lattice simulations of @xmath50 color qcd coupled to @xmath50 flavors of quarks in the symmetric representation of the gauge group .
these studies are motivated by the idea that this theory lies close to a strongly coupled conformal field theory associated to a new i.r .
attractive zero of the beta function .
the resultant theory is expected to exhibit non - qcd like dynamics and a slow evolution of the coupling constant over some range of scales . to search for these effects
we have simulated both the symmetric quark model and its counterpart employing fundamental quarks on identical lattice volumes and for comparable couplings and quark masses . in our simulations
we observe several features which distinguish the symmetric from the fundamental representation .
the most obvious of these can be seen in the hadron masses .
while the symmetric pion and rho behave in a qualitatively similar way to their fundamental counterparts at strong coupling they depart from this behavior for @xmath51 . in this regime
their ( lattice ) masses decrease with further increases in the bare coupling @xmath25 attaining values substantially smaller than those seen in the theory with fundamental quarks .
this observation is consistent with the existence of a light scale @xmath31 attributable to walking dynamics .
however , our results can also be interpreted as evidence that for sufficiently light symmetric quarks and large @xmath25 the 2 color theory has _
already _ entered a conformal phase in which the quarks are deconfined at zero temperature .
this could explain the lack of linear scaling of @xmath52 with @xmath12 at large @xmath25 since in such a phase chiral symmetry would be restored .
if so , it is unclear whether this is the result of strong finite volume effects and will disappear on larger lattices or is a genuine bulk transition occuring for some @xmath53 .
if the latter scenario turns out to be true it would imply that the critical number of flavors needed to access the conformal phase is smaller than the perturbative estimate @xmath54 .
further work on larger lattices will be needed to answer this question definitively . in the case of fundamental quarks
a much larger number of flavors is needed to approach the conformal window .
nevertheless , a small suppression of chiral symmetry breaking effects was observed in @xcite .
more recently a study of qcd with fundamental quarks showed evidence for an intermediate conformal phase for @xmath55 @xcite .
clearly our lattices are very small and so our results should be seen primarily as providing motivation for a study on larger lattices .
going to larger lattices will allow us to understand better the finite volume effects clearly visible in this work , will allow for a more careful study of the hadron spectrum and should allow simulations to be undertaken at smaller cut - off . with a larger lattice it should also be possible to see direct evidence for walking by measuring the running of a renormalized coupling extracted from either the static quark potential or using schrdinger functional techniques .
this work is currently underway .
we are happy to thank l. del debbio , m. t. frandsen and l. giusti for helpful discussions . the work of s.c .
is support in part by doe grant de - fg02 - 85er40237 while f.s .
is supported by the marie curie excellence grant under contract mext - ct-2004 - 013510 as well as the danish research agency .
f. sannino and k. tuominen , phys .
d * 71 * ( 2005 ) 051901 [ arxiv : hep - ph/0405209 ] .
d. d. dietrich and f. sannino , phys .
d * 75 * , 085018 ( 2007 ) [ arxiv : hep - ph/0611341 ] .
h. georgi , arxiv : hep - ph/0703260 .
d. d. dietrich , f. sannino and k. tuominen , phys .
d * 72 * ( 2005 ) 055001 [ arxiv : hep - ph/0505059 ] ; _ ibid _ phys . rev .
d * 73 * ( 2006 ) 037701 .
note that we have also checked that our results do not change when taking into account the recent drop of the top mass from the tevatron results .
we are still within the 68% confidence level even in the event of a 1 tev composite higgs mass . c. kouvaris , arxiv : hep - ph/0703266 . s. b. gudnason , c. kouvaris and f. sannino , phys .
d * 74 * , 095008 ( 2006 ) [ arxiv : hep - ph/0608055 ] .
t. appelquist , k. d. lane and u. mahanta , phys .
* 61 * ( 1988 ) 1553 .
a. g. cohen and h. georgi , nucl .
b * 314 * ( 1989 ) 7 .
t. banks and a. zaks , nucl .
b * 196 * ( 1982 ) 189 .
s. weinberg , phys .
d * 19 * , 1277 ( 1979 ) ; l. susskind , phys .
d * 20 * , 2619 ( 1979 ) .
e. eichten and k. d. lane , phys .
b * 90 * , 125 ( 1980 ) .
b. holdom , phys .
d * 24 * , 1441 ( 1981 ) . k. yamawaki , m. bando and k. i. matumoto , phys.rev .
* 56 * , 1335 ( 1986 ) .
t. w. appelquist , d. karabali and l. c. r. wijewardhana , phys .
lett . * 57 * , 957 ( 1986 ) . | we provide the first evidence of a walking dynamics for two color lattice yang - mills theory with two dirac flavors in the symmetric representation of the gauge group . |
the nonlinear gross - pitaevskii equation ( gpe ) , as a mean - field theory , has been the dominant tool in describing the dynamics of bose - einstein condensates ( becs ) in ultracold atomic gases @xcite .
however , we face a quandary when the mean - field dynamics of a bec becomes dynamically unstable or chaotic @xcite : on one hand , one may regard this instability as an unphysical artifact resulted from the mean - field approximation , since the exact dynamics of a bec is governed by the many - body schrdinger equation , which is linear and thus does not allow chaos ; on the other hand , the dynamical instability was observed in experiments @xcite and it has been proved with mathematical rigor that the gpe describes correctly not only the ground state but also the dynamics of a bec in the large @xmath2 limit ( @xmath2 is the number of bosons ) @xcite . our aim in this work is to resolve this fundamental dilemma .
our study shows that the mean - field theory ( the gpe ) is only valid up to time _
h = n + o(n ) , [ eq : timescale ] where @xmath1 is the lyapunov exponent of the mean - field dynamics and @xmath3 is a constant that depends only on systems . with this time scale ,
the dilemma is resolved : on one hand , in the large @xmath2 limit ( @xmath5 ) , @xmath4 goes to infinity and thus the gpe is always valid just as proved rigorously in ref .
@xcite ; on the other hand , the time @xmath4 increases with @xmath2 only logarithmically and it is not a long time for a typical bec experiment .
for example , for the system studied in ref .
@xcite , the lyapunov time @xmath6 ms .
as the number of atoms in a bec prepared in a typical experiment is around @xmath7 , we have @xmath8 ms . as a result , the dynamical instability or
the breakdown of the mean - field dynamics can be easily observed in a typical experiment as reported in ref .
@xcite .
this time scale @xmath4 is essentially the ehrenfest time , which is the time that the correspondence between the classical and quantum dynamics breaks down @xcite .
the usual ehrenfest time @xmath9 , where @xmath1 is the lyapunov exponent of the classical motion and @xmath10 is a typical action @xcite .
the similarity is due to that the gpe can be regarded as a classical equation in the large @xmath2 limit @xcite .
therefore , our result paves a way to experimental investigation of a fundamental relation in the quantum - classical correspondence the logarithmic behavior of the ehrenfest time as @xmath2 can be varied in experiments .
we cast the quantum dynamics onto the particle - number phase space ( pnps ) , which is a rearrangement of fock states . in this phase space , for a nearly coherent state and in the large @xmath2 limit , quantum many - body dynamics is equivalent to an ensemble of mean - field dynamics .
when the mean - field motion is regular , mean - field trajectories will stay together and the bose gas remains coherent .
if the mean - field motion is unstable or chaotic , mean - field trajectories will separate soon from each other exponentially , leading to decoherence of bose gas and breakdown of the mean - field theory .
so , there are two distinct types of quantum dynamics , whose difference can be characterized by the quantum fidelity for one - particle reduced density matrices .
we investigate the ehrenfest breakdown numerically in the system of a bec in a triple - well potential @xcite , which may be the simplest bec model that embraces chaotic mean - field dynamics . with this model , we verify numerically the ehrenfest time and show that our quantum fidelity can well capture the characteristics of two different types of quantum dynamics .
the mean - field instability or breakdown has been discussed in literature @xcite . however , a general and explicit relation between mean - field chaos , number of particles and breakdown time is still lacking . and in pnps
not only such breakdown can be understood intuitively and quantitatively , but the significance of a local phase structure is also apparent , distortion of which leads to decoherence .
@xcite , it is shown that many quantum systems become classical in the large @xmath2 limit .
a dilute bose gas belongs to this class of quantum systems : its dynamics becomes classical and it is well described by the mean - field gpe in the large @xmath2 limit . in this section ,
we introduce pnps , where this quantum - classical correspondence in the large @xmath2 limit becomes transparent .
any quantum state @xmath11 of a system of @xmath2 identical bosons with @xmath12 single - particle states can be regarded as a wavefunction @xmath13 over an @xmath14-dimensional lattice space , which we call particle number phase space ( pnps ) , via ( ) _ i = 1^m , where @xmath15 s are entries of the @xmath12-dimensional vector @xmath16 , @xmath17 for @xmath18 and @xmath19 . and @xmath20 and @xmath21 are the creation and annihilation operators for the @xmath22-th single - particle state , with @xmath23 = \delta_{ij}$ ] and @xmath24 .
the continuous limit of pnps is a hyperplane in @xmath25^m$ ] ( defined by constraint @xmath26 ) , where we can define ( for @xmath22 from @xmath27 to @xmath12 ) & & x_i x_i |()|^2 + & & ( x_i)^2 ( x_i - x_i ) ^2 |()|^2 to characterize the average position and spread of the distribution @xmath28 over pnps , given @xmath11 normalized .
of course for any finite @xmath2 , the integral should be interpreted as summations over all @xmath16 in pnps . as an example of our particular interest
, we examine an su(@xmath12 ) coherent state @xmath29 in pnps : _ c ( _
i = 1^m _ i a_i^)^n , [ eq : coherent ] where @xmath30 .
in such case , we say @xmath31 ( an @xmath12-dimensional vector with @xmath32 as its entries ) is the mean - field state of the su(@xmath12 ) coherent state @xmath33 .
it is straightforward to show for this coherent state @xmath33 x_i = |_i|^2,(x_i)^2 = |_i|^2 ( 1 - |_i|^2 ) / n,[eq : width ] which indicate that the coherent state @xmath33 corresponds to a localized distribution @xmath28 in pnps that peaks around @xmath34 with a vanishing spread at large @xmath2 .
and the wavefunction @xmath13 in pnps has a phase structure . for any @xmath16 and @xmath35 in pnps , ( ) - ( )
= n _ i = 1^m ( x_i - y_i)_i ( 2 ) , [ eq : phase ] which shows a wavevector @xmath36 : @xmath37 .
this phase structure is important as it will give us an estimate of the time @xmath4 in our later discussion .
it is worth noting that when @xmath38 , there is no limit of the wavefunction @xmath13 because its wavevector @xmath36 diverges .
overall , we find that the coherent state corresponds to a single - peaked wavepacket with plane - wave phase structure in pnps . in the following ,
we shall discuss quantum dynamics in pnps and its relation to the mean - field dynamics .
note that the formalism of pnps was also used in other contexts @xcite , where phase structure and dynamics , however , were not discussed .
consider a quite general hamiltonian of a bose gas = _
i , j = 1^m\ { h^0_ij _ i^_j + _ i^_j^_j _ i } , [ eq : hamiltonian ] where @xmath39 and @xmath40 .
corresponding to the schrdinger equation @xmath41 , there is an equation of motion ( eom ) for @xmath42 in pnps ( eq.([schrodinger ] ) in the appendix ) .
we are especially interested in the dynamics of a nearly coherent state @xmath42 , which satisfies the following two conditions : ( _ i _ ) the distribution @xmath43 is localized such that @xmath44 for all @xmath45 ; ( _ ii _ ) a local wavevector @xmath46 exists in pnps and varies insignificantly over a scale of @xmath47 , i.e. , @xmath48 for all @xmath49 . with these two conditions and keeping only finite terms in the large @xmath2 limit , an approximate ( to @xmath50 ) eom for @xmath42 in pnps can be derived ( see eq.([approx ] ) in the appendix ) .
mathematically , there are @xmath51-function solutions to this eom ( eq.([approx ] ) ) : (; t ) = _ i = 1^m ( x_i - x^0_i(t ) ) .[eq : delta ] in these @xmath51-function solutions , @xmath52 , @xmath53 satisfy the following equation i _
t = [ _ , ] , [ eq : rhomotion ] where @xmath54 and _ ij(t ) ^ i ( k^0_i(t ) - k^0_j(t))/n .
this is just the mean - field eom for the one - particle reduced density matrix .
conditions ( _ i _ ) and ( _ ii _ ) reflect our expectations of nearly coherent states ( see eqs.([eq : width ] ) and ( [ eq : phase ] ) ) .
the existence of @xmath51-function solutions corresponds to the established result that for any time @xmath55 , when @xmath38 , coherent states at @xmath56 stay coherent when @xmath57 @xcite .
the results above can be interpreted as follows : at large @xmath2 , for any initial state satisfying the two conditions , its time evolution may be regarded as the superposition of mean - field dynamics of @xmath51-functions , since any function in pnps can be decomposed into a superposition of a cloud of @xmath51-functions ! this is similar to the quantum dynamics of a single - particle wavepacket in real space : it can be regarded as a cloud of classical particles and each of them follows the newton s eom . as the quantum - classical correspondence between a quantum wavepacket and a classical particle will break down at the ehrenfest time , the correspondence between one state in pnps and its mean - field description one @xmath51-function solution ( see eq.([eq : delta ] ) )
will also fail when the mean - field trajectories of the @xmath51-functions in the cloud diverge .
the breakdown time @xmath4 can be estimated using a conventional strategy in quantum chaos as in ref .
@xcite . essentially , before the breakdown the wavepacket of nearly coherent states in pnps expands in the form of @xmath58 , where @xmath1 is the lyapunov exponent of the mean - field dynamics .
according to eq.([eq : width ] ) , for @xmath59 , ( t ) .
[ eq : delta ] and there is a consistent mean - field description only if local wavevectors across the wavepacket are almost equal , that is , t n^- ( t ) 1 , [ eq : rate ] where @xmath60 is the average rate of growth of curvature @xmath61 and the @xmath2 dependence is written explicitly . substituting ( [ eq : delta ] ) into ( [ eq : rate ] ) ,
we have t + t + ( + ) n.[eq : check ] the ehrenfest time @xmath4 in eq.([eq : timescale ] ) is obtained with @xmath62 , which is independent of @xmath2 or @xmath1 .
numerical verification of this relation will be presented later .
note that it is well - known that the quantum - classical correspondence may last far beyond the ehrenfest time ( see , e.g. , ref.@xcite ) .
similarly , it is possible that the mean - field theory remains valid even after our first estimate @xmath63 ; this interesting and special topic will be left for future study .
we now illustrate our results with an example . consider a bec in a ring - shaped triple - well potential @xcite . under tight - binding approximation ,
the second - quantized hamiltonian is ( as a specific case of eq.([eq : hamiltonian ] ) ) = -_1 i , j 3^i j _
i^_j + _ i = 1 ^ 3 _ i^_i^_i _ i , where @xmath64 is the on - site interaction strength . for this system @xmath65 . its corresponding nonlinear mean - field eom is i ( c _ 1 + _ 2 + _ 3 + ) = ( ccc c |_1|^2 & -1 / 2 & -1 / 2 + -1 / 2 & c |_2|^2 & -1 / 2 + -1 / 2 & -1 / 2 & c |_3|^2 ) ( c _ 1 + _ 2 + _ 3 + ) . [ eq : mf ] shown in fig.[poin ] is a poincar section of the above mean - field dynamics , where two kinds of motion are evident : the central regular region is surrounded by a chaotic sea .
the conjugate variables used in plotting fig.[poin ] are @xmath66 , @xmath67 , which are defined as @xmath68 , @xmath69 , @xmath70 , @xmath71 .
the quantum dynamics of this model can also be computed rather easily .
the evolution of @xmath28 in pnps is plotted in fig.[fig : schematic ] , where two types of quantum dynamics are clearly observed . in fig.[fig : schematic ] ( a , b ) , an initial coherent state , which is a gaussian - like wavepacket in pnps , shows no significant expansion or distortion during dynamical evolution . in fig.[fig : schematic ] ( c , d ) , the situation is drastically different : a similar - looking initial coherent state expands and becomes dramatically distorted after a certain time .
the difference is caused by the fact that the initial state in fig.[fig : schematic ] ( a ) corresponds to a mean - field state in the regular region in fig.[poin ] while the one in fig.[fig : schematic ] ( c ) corresponds to a mean - field state in the chaotic region .
it is obvious that the mean - field theory can not describe the dramatic quantum dynamics shown in fig.[fig : schematic ] ( c , d ) .
such a failure or breakdown of the mean - field theory due to rapid decoherence has long been noticed in literature @xcite . in ref .
@xcite , a remedy was tried unsuccessfully to bridge the gap between the mean - field theory and the exact quantum theory . in this work
we have shown that there exists a general time scale @xmath4 in terms of lyapunov exponent and number of bosons beyond which the mean - field theory fails . in the following
, we shall introduce a quantum fidelity to distinguish the two types of quantum dynamics shown in fig.[fig : schematic ] without using mean - field formalism , and confirm the time scale @xmath4 numerically . to quantify the loss of coherence in the quantum evolution as shown in fig.[fig : schematic ] ( d ) ,
we introduce the following quantum fidelity @xmath72 for one - particle reduced density matrix ( rdm ) @xmath73 and @xmath74 : _ q ( , ) ^. [ def : many ] for a quantum state @xmath75 , its one - particle rdm can be explicitly written as _
ij ( t ) | _ i^_j | ( t ) .there are three reasons to use this quantum fidelity : \1 ) experimentally we are often interested in the one - particle rdm .
\2 ) it allows us to define coherence @xmath76 : ( ) _ q ( , ) , where @xmath73 is the one - particle rdm for @xmath11 . the coherence @xmath76 can quantify how coherent the state @xmath11 is : @xmath77 if and only if @xmath11 is a coherent state as in eq.([eq : coherent ] ) .
\3 ) it returns to the mean - field fidelity for coherent states , i.e. , @xmath78 if @xmath73 , @xmath74 are one - particle rdm for coherent states @xmath33 and @xmath79 , and @xmath80 , @xmath81 are mean - field states of @xmath33 and @xmath79 ( see discussion under eq.([eq : coherent ] ) ) .
therefore , before the ehrenfest breakdown @xmath72 essentially captures mean - field characteristics , especially the lyapunov exponent , which distinguishes regular and chaotic mean - field trajectories .
the numerical simulation aims at verifying our theoretical understanding as discussed : for a coherent initial state , at the beginning the mean - field dynamics agrees with the quantum evolution , producing even the same growth of discrepancy between states ; however , long - time exponential growth is not allowed by quantum mechanics , so there exists an ehrenfest time @xmath4 beyond which the mean - field and quantum correspondence fails .
such a failure is due to the decoherence of quantum states ; the breakdown time @xmath4 is given in eq.([eq : timescale ] ) . in the following
we provide numerical evidences for our theortical understanding .
we choose a coherent initial state @xmath82 with one - particle rdm @xmath83 , whose corresponding mean - field state is @xmath84 .
then we slightly perturb the mean - field state into @xmath85 , and generate the corresponding coherent state @xmath86 and rdm @xmath87 .
next we observe the evolution of quantum fidelity between these two states , which allows us to calculate the lyapunov exponent .
of course , @xmath88 and @xmath89 evolve according to the mean - field equations eq.([eq : mf ] ) , @xmath75 and @xmath90 evolve according to the quantum hamiltonian in eq.([eq : hamiltonian ] ) , @xmath91 and @xmath92 are obtained from @xmath75 and @xmath90 , respectively . @xmath93 and @xmath94 are shown in fig.[fig : lyapunov ] ( a ) , where we see that the mean - field fidelity @xmath95 coincides with @xmath96 for small @xmath97 , as expected .
however , we also observe in fig.[fig : lyapunov ] ( a ) that there is an ehrenfest time @xmath4 , when @xmath72 and @xmath98 start to visibly disagree .
cases for different @xmath2 and @xmath1 are plotted in fig.[fig : lyapunov ] ( a ) , where we can see that as @xmath2 increases or @xmath1 decreases , @xmath4 gets longer .
this qualitatively agrees with the scaling of the ehrenfest time . and in fig.[fig :
lyapunov ] ( b ) , it is observed that although @xmath4 is different for different @xmath2 and @xmath1 , @xmath4 is approximately the time when the coherence @xmath99 drops below 98% .
this confirms our understanding that the failure of correspondence between the mean - field and quantum descriptions is the result of decoherence of quantum states . for the integrable @xmath100 and the chaotic @xmath101 trajectories .
@xmath102 , @xmath103 , @xmath104 , @xmath105 and time step in simulation is @xmath106 .
@xmath76 of integrable @xmath100 ( the solid line in the top ) remains high while @xmath76 of chaotic @xmath101 ( other solid lines ) drops quickly ( @xmath107 from left to right ) .
the dashed line is @xmath108 .
( b ) linear fitting of @xmath109 for the chaotic initial state in ( a ) with @xmath110 , 1.25 and 2 , @xmath2 from 20 to 80 .
data points are calculated when @xmath111 drops to @xmath112 .
the slope @xmath113 is found independent of @xmath64 or @xmath1 .
, title="fig : " ] for the integrable @xmath100 and the chaotic @xmath101 trajectories .
@xmath102 , @xmath103 , @xmath104 , @xmath105 and time step in simulation is @xmath106 .
@xmath76 of integrable @xmath100 ( the solid line in the top ) remains high while @xmath76 of chaotic @xmath101 ( other solid lines ) drops quickly ( @xmath107 from left to right ) .
the dashed line is @xmath108 .
( b ) linear fitting of @xmath109 for the chaotic initial state in ( a ) with @xmath110 , 1.25 and 2 , @xmath2 from 20 to 80 .
data points are calculated when @xmath111 drops to @xmath112 .
the slope @xmath113 is found independent of @xmath64 or @xmath1 .
, title="fig : " ] based on such understanding , we can quantitatively define the ehrenfest time in this example as the time when the coherence @xmath99 drops below 98% .
examples of decay of @xmath99 is illustrated in fig.[fig2 ] ( a ) , where the ehrenfest time @xmath4 is measured when @xmath99 drops below the dashed line . by varying @xmath2 and @xmath64
, we verify the relation eq.([eq : check ] ) , which leads to eq.([eq : timescale ] ) , in fig.[fig2 ] ( b ) .
a linear fitting between @xmath114 and @xmath115 is found with a constant slope @xmath116 ( see eq.([eq : check ] ) ) , suggesting @xmath113 in eq.([eq : timescale ] ) .
note here @xmath117 is replaced by @xmath118 for numerical convenience .
in sum we have answered an intriguing question when does mean - field approximation of a dilute bose gas remain valid as the system evolves ?
our answer is the mean - field dynamics breaks down at the ehrenfest time @xmath119 .
the study is facilitated by introducing particle number phase space , where one can see easily that the correspondence between many - body quantum dynamics and mean - field dynamics is similar to the usual quantum - classical correspondence . as @xmath2 can be varied in bec experiments ,
it is now possible to experimentally measure the logarithmic behavior of the ehrenfest time .
one can compare physical observables in the experiment with their theoretical mean - field values , and measure the ehrenfest time when their discrepancy exceeds a threshold .
becs with unstable or chaotic mean - field descriptions are suitable for such experiments ; for example , spinor becs @xcite may be a good candidate system . +
this work is supported by the national basic research program of china ( 2013cb921903 , 2012cb921300 ) and the national natural science foundation of china ( 11274024 , 11334001 , 11429402 ) .
for the hamiltonian in eq.([eq : hamiltonian ] ) , the schrdinger equation in pnps reads ( @xmath120 ) i _ t (; t ) & = & _ i h^0_ii n x_i (; t ) + & + & _ i u_ii n x_i ( x_i - ) (; t ) + & + & _ i j h^0_ij n ( + ^ij ; t ) + & + & _ i j u_ij n x_i x_j (; t ) , [ schrodinger ] where @xmath121 , @xmath122 is an @xmath12-dimensional vector @xmath123 , @xmath124 .
we are especially interested in the dynamics of a nearly coherent state . with conditions ( _ i _ ) and ( _ ii _ ) in sect .
ii b and @xmath38 , eq.([schrodinger ] ) becomes i _
t & = & _ i h^0_ii n x_i + _
i u_ii n x_i ( x_i - ) + & + & _ i j h^0_ij n + & + & _ i j h^0_ij + & + & _ i j h^0_ij [ ( _ j - _ i ) || ] + & + & _ i j u_ij
n x_i x_j + o(1 ) , [ approx ] where @xmath125 and @xmath126 is the local wavevector of wavefunction @xmath127 at @xmath128 , as discussed in condition ( _ ii _ ) . the argument of all @xmath129 and @xmath127 is @xmath128 and omitted .
now we assume a @xmath51-function solution as in eq.([eq : delta ] ) . by equalling the coefficients before @xmath51 , @xmath130 and the derivatives of coefficients before @xmath51 ( which is necessary to reflect the plane - wave phase structure ) on both sides , keeping finite terms in the large @xmath2 limit , we obtain _
t x^0_i & = & 2 _ j h^0_ij i ( k^0_j - k^0_i)[eq : x ] _
t k^0_i & = & - _ j h^0_ij i ( k^0_j - k^0_i ) + & & - 2 _ j u_ij x^0_j , [ eq : k ] where the argument @xmath97 of all @xmath131 and @xmath132 is omitted for brevity .
lengthy but straightforward calculations will verify that eqs.([eq : x ] ) and ( [ eq : k ] ) are equivalent to eq.([eq : rhomotion ] ) , which is same as the mean - field eom for the one particle rdm @xmath73 . | the mean - field dynamics of a bose gas is shown to break down at time @xmath0 where @xmath1 is the lyapunov exponent of the mean - field theory , @xmath2 is the number of bosons , and @xmath3 is a system - dependent constant .
the breakdown time @xmath4 is essentially the ehrenfest time that characterizes the breakdown of the correspondence between classical and quantum dynamics .
this breakdown can be well described by a quantum fidelity defined for one - particle reduced density matrices .
our results are obtained with the formalism in particle - number phase space and are illustrated with a triple - well model .
the logarithmic quantum - classical correspondence time may be verified experimentally with bose - einstein condensates . |
scarce doubts remain that dark energy ( de ) exists . not only snia data indicate an accelerated cosmic expansion ( perlmutter et al . 1997 , 1998 , riess et al .
1998 , foley et al . 2007 ) ; also cmb and deep sample data show a clear discrepancy between the total density parameter @xmath5 approaching unity , and the matter density parameter @xmath6@xmath7 ( see , _
e.g. _ , spergel et al .
de covers this gap ; its state parameter @xmath8 must approach @xmath9 today , so apparently excluding that de is made of free particles ( @xmath10 : de pressure , energy density ) .
the true nature of de is however still elusive ; a false vacuum and a self interacting scalar field are among the most popular hypotheses for it ( wetterich 1988 , ratra & peebles 1988 ) .
in this paper we explore some possible consequences of our poor knowledge of de nature .
in particular we test the risk that other cosmological parameter estimates are biased by a inadequate parametrization of the de component .
we shall see that this risk is real .
theoretical predictions had an astonishing success in fitting cmb data .
for instance , the sw effect , predicting low@xmath11 @xmath12 data , or primeval compression waves , predicting @xmath12 peaks and deeps , were clearly detected .
there is little doubt that we are exploring the right range of models .
when we investigate de nature through cmb data , we must bear in mind that they were mostly fixed at a redshift @xmath13 when the very de density should be negligible , and so affects peak and deep positions indirectly , through the values of @xmath14 and @xmath15 ( here @xmath16km / s / mpc is the present hubble parameter ; @xmath17 are the present total , cdm , baryon density parameters ) .
later information on de state equation , conveyed by the isw effect , is seriously affected by cosmic variance and often relies on the assumption that a single opacity parameter @xmath18 can account for reionization , assumed to be ( almost ) instantaneous . accordingly ,
if we assume dynamical de ( dde ) , due to a scalar field @xmath19 self interacting through a potential @xmath20 cmb data allow to exclude some interaction shape , _
e.g. _ ratra peebles ( 1988 ) potentials with significantly large @xmath21 energy scales , but hardly convey much information on potential parameters . in spite of that , when we choose a de potential or a specific scale dependence of the de state parameter @xmath22 we risk to bias the values of other cosmological parameters , sometimes leading to premature physical conclusions .
an example is the value of the primeval spectral index for scalar fluctuation @xmath23 . using wmap3 data ( spergel et al .
2007 ) and assuming a @xmath21cdm cosmology , the value @xmath24 is `` excluded '' at the 2@xmath25 confidence level . on the contrary , colombo & gervasi ( 2007 ) showed that this is no longer true in a dde model based on a sugra potential ( brax & martin 1999 , 2001 ; brax , martin & riazuelo 2000 ) , whose likelihood was the same of @xmath21cdm .
the risk that our poor knowledge of de nature biases parameter determination is even more serious if dm de coupling is allowed . coupled de ( cde ) cosmologies were studied by various authors ( see , _ e.g. _ , wetterich 1995 , amendola 2000 , bento , bertolami & sen 2002 , macci et al .
2004 ) .
while dde was introduced in the attempt to ease de _ fine tuning _ problems , cde tries to ease the _ coincidence _ problem .
let us then parametrise the strength of dm de coupling through a parameter @xmath26 defined below .
when @xmath27 is large enough , dm and de scale ( quasi ) in parallel since a fairly high redshift . in turn
this modifies the rate of cosmic expansion whenever dm and/or de contributions to the total energy density are non
negligible , so that limits on @xmath27 can be set through data .
the range allowed ( @xmath280.12 ; mainini , colombo & bonometto 2005 , mainini & bonometto 2007 , see also majerotto , sapone & amendola 2004 , amendola , campos & rosenfeld , 2006 ) , unfortunately , is so limited that dm and de are doomed to scale differently , but in a short redshift interval .
clearly , this spoils the initial motivation of coupling , but , once the genie is outside the lamp , it is hard to put him back inside : even though the coupling solves little conceptual problems , we should verify that no bias arises on the other parameters , for the neglect of @xmath27 s consistent with data .
this is far from being just a theoretical loophole , the still unknown physics of the dark components could really imply the presence of a mild dm de coupling , and its discovery could mark a step forwards in the understanding of their nature . here
we test this possibility by performing some numerical experiments .
we assume dde and cde due to a sugra potential and use mcmc techniques to fit the following parameter set : @xmath29 , @xmath30 , @xmath18 ,
@xmath31 , @xmath23 , @xmath32 and ( constant ) @xmath0 ; @xmath31 is the angular size of the sound horizon at recombination ( see however below ) , @xmath23 and @xmath32 are spectral index and amplitude of scalar waves , no tensor mode is considered .
the plan of the paper is as follows : in section 2 we discuss how artificial data are built , outlining the models selected , the dde potential used and the sensitivity assumed . in section 3
we briefly debate the features of the mcmc algorithm used and illustrate a test on its efficiency , also outlining the physical reasons why some variables are more or less efficiently recovered . in section 4 we discuss the results of an analysis of dde artificial data , against the @xmath33 assumption . in section 5
we briefly summarize why and how cde models are built and do the same of sec . 4 for cde models .
this section yields the most significant results of this work . in section 6
we draw our conclusions .
in order to produce artificial data we use camb , or a suitable extension of it ( see below ) , to derive the angular spectra @xmath34 , @xmath35 , @xmath36 for various sets of parameter values .
artificial data are then worked out from spectra , according to perotto et al .
( 2006)@xmath37 .
the analysis we report is however mostly based on a single choice of parameters , wmap5 inspired ( see komatsu et al . , 2008 , spergel et al .
2007 ) , in association with three different @xmath27 s .
two other parameter choices will also be used : ( i ) to test the efficiency of the mcmc algorithm ; ( ii ) to add results for a still lower @xmath27 value , so strengthening our conclusions .
the parameter sets for the most general case and the ( ii ) case are shown in table 1 . in these cases
de is due to a scalar field self interacting through a sugra potential @xmath38 ( @xmath39 planck mass ) which has been shown to fit cmb data at least as well as @xmath21cdm ( colombo & gervasi 2007 ) ; as outlined in the table , we input the value of the energy scale @xmath40 the corresponding @xmath41 value is determined by the program itself ; the choice @xmath42gev is consistent with colombo & gervasi ( 2007 ) findings and is however scarcely constrained by data .
the meaning of the coupling constant @xmath27 is discussed at the beginning of section 5 .
.__cosmological parameters for artificial cmb data .
_ _ [ cols="^,^,^",options="header " , ] -.8truecm starting from the spectral components and assuming that cosmic fluctuations are distributed according to a gaussian process , we generate _ realizations _ of the coefficients of the spherical harmonic expansions for the temperature and e polarization fields , according to the expressions @xmath43
@xmath44 \sqrt{c_l^{(tt ) } } g_{lm}^{(1 ) } + \sqrt{c_l^{(ee ) } - \left[\left(c_l^{(te)}\right)^2/c_l^{(tt ) } \right ] } g_{lm}^{(2)}~ , \label{alme}\ ] ] where both @xmath45 for any @xmath11 and @xmath46 , are casual variables , distributed in a gaussian way with null averages and unit variance , so that the equality @xmath47 is approached when the number of realizations increases . together with eqs .
( [ almt ] ) and ( [ alme ] ) this grants that @xmath48 @xmath49 if averages are taken over an `` infinite '' set of sky realizations .
> from a single realization , we can define estimators of the power as @xmath50 @xmath51 taken independently from each other , @xmath52 and @xmath53 at a given @xmath11 are the sum of the squares of @xmath54 gaussian random variables , so they are distributed according to a @xmath55 with @xmath54 , which approaches a gaussian distribution centered around the fiducial @xmath56 , @xmath57 , values , as @xmath11 increases .
if we consider @xmath58 and @xmath59 simultaneously , the set of estimators @xmath52 , @xmath60 , @xmath61 follow a wishart distribution ( see , e.g. , percival & brown 2006 ) .
we consider here an idealized full sky experiment characterized by : ( i ) finite resolution , that we shall set through a full width half maximum angle @xmath62 , where the antenna sensitivity is reduced to @xmath63 assuming a circularly symmetric gaussian beam profile ; ( ii ) background noise due to the apparatus , that we shall assume to be _ white _ , _
i.e. _ @xmath11independent with an assigned variance @xmath64 under these simplified assumptions , the characteristics of an experiment are completely defined by the values of @xmath62 and @xmath77 . here
we shall take @xmath78 while @xmath79 and @xmath80 .
these can be considered conservative estimates of sensitivity in the forthcoming planck experiment .
when the number @xmath81 of the parameters to be determined from a given data set is large , the whole @xmath81dimensional parameter space can not be fully explored within a reasonable computing time .
montecarlo markov chain ( mcmc ) algorithms are then used , whose efficiency and reliability have been widely tested . in this work we used the mcmc engine and statistical tools provided by the cosmomc package ( lewis & bridle 2002 ; _
cosmologist.info/cosmomc_ ) .
this tool set allows us to work out a suitable set of @xmath82 markov chains and analyze their statistical properties , both in the full @xmath81dimensional parameter space , and in lower dimensional subspaces .
of particular interest are the _
marginalized _ distributions of each parameter , obtained by integrating over the distribution of the other @xmath83 parameters .
the above algorithms implement the following steps .
( i ) let @xmath84 be a point model in the parameter space .
the spectra @xmath12 of such model are computed and their corresponding likelihood @xmath85 is evaluated according to @xmath86 + { \rm const . } ~,\ ] ] where @xmath87 stands for either fiducial or realized spectra .
the above expression holds for a single @xmath88 or @xmath89 spectrum ; for its generalization to 3 spectra and a discussion of the effects of anisotropic noise , sky cuts , etc .
see , e.g. percival & brown ( 2006 ) .
( ii ) the algorithm then randomly selects a different point @xmath90 , in parameter space , according to a suitable _ selection function_. the probability of accepting @xmath90 is given by @xmath91 .
if @xmath90 is accepted , it is added to the chain , otherwise the multiplicity @xmath92 of @xmath84 increases by 1 .
( iii ) the whole procedure is then iterated starting either from @xmath90 or from @xmath84 again , until a stopping condition is met .
the resulting chain is defined by @xmath93 , where @xmath94 are the points explored and the multiplicity @xmath95 is the number of times @xmath94 was kept .
the whole cycle is repeated until the chains reach a satisfactory _ mixing _ and a good
_ convergence_. the first requisite essentially amounts to efficiently exploring the parameter space , by quickly moving through its whole volume .
in particular , the fact that the algorithm sometimes accepts points with lower likelihood than the current point , avoids permanence in local minima , while a careful choice of parametrization and of the selection function minimize the time spent exploring degenerate directions .
a good convergence instead guarantees that the statistical properties of the chains ( or of a suitably defined subset of the chains ) , correspond to those of the underlying distribution .
here we implement the convergence criterion of gelman & rubin ( 1992 ) , based on @xmath70ratio computation : convergence and mixing are reached when @xmath96 . fulfilling such criterion requires @xmath97 points , in the cases considered here .
the algorithm was preliminarily tested by using @xmath21cdm as true cosmology and the set of parameter values shown in figure [ lcdm ] .
such figure reports the likelihood distributions found in this case , both for input parameters ( with asterisk ) and for several derived parameters as well . besides confirming that input values are suitably recovered , mostly with fairly small errors , fig .
[ lcdm ] shows a broad non gaussian distribution for @xmath0 values , with little skewness but significant kurtosis .
this is a well known consequence of the geometrical degeneracy present in cmb spectra , and clearly shows the difficulty of cmb data to yield results on de nature .
the non gaussian behavior is even more accentuated for some derived parameter , as @xmath98 , @xmath2 and the universe age @xmath99 we shall recover analogous features in the next plots ; here we want to stress that they do not derive from fitting a partially unsuitable parameter set .
figure [ lcdm ] also shows how errors increase when passing to some secondary parameter , from primary parameters specifically devised to break degeneracies .
for instance , the errors on @xmath31 and @xmath1 are @xmath100 and @xmath101 , respectively . from these parameters , @xmath2 and @xmath98
are derived , whose errors are @xmath102 .
the point here is that the primary variable is @xmath103 @xmath104 setting the peak of the _
quasi_gaussian last scattering band ( lsb ) .
it is then easy to see that @xmath105 @xmath106 so that @xmath31 is apparently independent from @xmath2 and @xmath3 as , in the former equation @xmath107 while , in the latter one @xmath108 and @xmath3 only sets the normalization of the latter term at the r.h.s .
, so that @xmath109 as a matter of fact , these equations exhibit a mild dependence of @xmath31 on @xmath110 and @xmath111 , as @xmath112 $ ] and @xmath113 $ ] ( here @xmath114 is the present cmb temperature and @xmath115 is number of neutrino families in the radiative background ) .
information of @xmath3 and , thence , on @xmath2 can be obtained only if the factor @xmath116 , yielding the evolution of the ratio between de and matter at low redshift , is under control .
it is certainly so if a @xmath21cdm model is assumed ; but , if we keep @xmath0 as free parameter , the uncertainty on it , ranging around 60@xmath117 reflect the difficulty to obtain the secondary parameters @xmath3 and @xmath2 .
more in general , eq . ( [ hoh ] ) shows that cmb data are sensitive to a change of @xmath116 causing a variation of lsb distance @xmath118 . on the contrary , @xmath31 does not discriminate between different @xmath119 yielding the same @xmath118 . incidentally , all that confirms that it is unnecessary to fix @xmath104 with high precision , or to discuss whether it is a suitable indicator of the lsb depth .
knowing @xmath104 with @xmath120 approximation is quite suitable to this analysis .
let us now discuss what happens if de state equation , in the `` real '' cosmology , can not be safely approximated by a constant @xmath122 to build data , we use then a sugra cosmology and our own extension of camb , directly dealing with a sugra potential both in the absence and in the presence of dm de coupling .
artificial data worked out for model a with @xmath123 were then fit to the same parameters as in figure [ lcdm ] .
the whole findings are described in table [ tab1 ] ( second column ) . in figure [ dde ]
we add further information , comparing the value distributions in fits assuming either a cosmology with @xmath124 or a sugra cosmology , then including @xmath125 as parameter , instead of @xmath0 .
the two fits agree well within 1@xmath25 , among themselves and with input values . in the figure
we show the fiducial case . -.2truecm cccc + parameter & & @xmath123 & @xmath126 & @xmath127 + input value & av . value @xmath128 & av . value @xmath128 & av .
value @xmath128 + @xmath129 & @xmath130 & @xmath130 & @xmath131 + @xmath132 & @xmath133 & @xmath134 & @xmath135 + & @xmath133 & @xmath136 & @xmath137 + & @xmath138 & @xmath133 & @xmath139 + @xmath1 & @xmath140 & @xmath141 & @xmath142 + @xmath143 & @xmath144 & @xmath145 & @xmath142 + & @xmath146 & @xmath147 & @xmath148 + & @xmath149 & @xmath150 & @xmath151 + @xmath152 & @xmath153 & @xmath154 & @xmath155 + @xmath156 & @xmath157 & @xmath154 & @xmath158 + & @xmath159 & @xmath160 & @xmath155 + & @xmath157 & @xmath154 & @xmath155 + @xmath161 & @xmath162 & @xmath163 & @xmath164 + @xmath165 & @xmath163 & @xmath166 & @xmath167 + & @xmath168 & @xmath169 & @xmath163 + & @xmath170 & @xmath171 & @xmath172 + @xmath0 & @xmath173 & @xmath174 & @xmath175 + & @xmath176 & @xmath177 & @xmath178 + & @xmath179 & @xmath180 & @xmath181 + & @xmath182 & @xmath183 & @xmath184 + @xmath23 & @xmath185 & @xmath186 & @xmath187 + @xmath188 & @xmath189 & @xmath186 & @xmath190 + & @xmath191 & @xmath186 & @xmath185 + & @xmath192 & @xmath193 & @xmath192 + @xmath194 & @xmath195&@xmath196 & @xmath197 + @xmath198 for @xmath123 & @xmath199 & @xmath200 & @xmath201 + @xmath202 for @xmath126 & @xmath203&@xmath204&@xmath205 + @xmath206 for @xmath207 & @xmath208&@xmath209 & @xmath210 + @xmath211 & @xmath212 & @xmath213 & @xmath214 + @xmath215 & @xmath216 & @xmath217 & @xmath218 + & @xmath219 & @xmath220 & @xmath221 + & @xmath222 & @xmath223 & @xmath224 + @xmath3 & @xmath225 & @xmath226 & @xmath227 + @xmath228 & @xmath229 & @xmath230 & @xmath231 + & @xmath232 & @xmath233 & @xmath234 + & @xmath235 & @xmath236 & @xmath237 + -.1truecm
let us briefly comment the results in the absence of coupling , when the fitting parameter set includes @xmath0 instead of @xmath21 : ( i ) @xmath111 and @xmath1 are recovered without any bias .
( ii ) the same holds for @xmath238 if a numerically refined value of @xmath104 is used ; the expression of @xmath104 , as above outlined , has a precision @xmath120 .
errors on @xmath104 , however , are negligible in comparison with errors arising from the scarce knowledge of @xmath119 .
as outlined in the previous section , the main problem to fix @xmath3 , @xmath239 , and @xmath2 , resides in the difficulty to determine the state equation of the expansion source since de becomes substantially sub dominant or dominant .
this also makes clear why the formal error in the determination of the above three parameters is wider when the fit assumes a constant @xmath240 varying @xmath0 yields an immediate and strong effect on de contributions and state equation , extending up to now ; on the contrary , only huge displacements of the energy scale @xmath21 cause significant variations to de contributions and state equation within the family of sugra cosmologies . in other terms , when @xmath21 varies , the part of the functional space that @xmath119 covers is not so wide .
accordingly , the greater error bars on @xmath3 , @xmath239 , @xmath2 , found when we use @xmath0 as a parameter , do not arise because we fit data to a `` wrong '' cosmology , but because varying @xmath0 leads to a more effective spanning of the functional space of @xmath119 .
as a matter of fact , the reliability of the fit must be measured by comparing the likelihood of the best
fit parameter sets .
likelihood values , given in the next section , do not discriminate between the two fits .
otherwise , a direct insight into such reliability is obtained by looking at the width of the marginalized _ a posteriori _ likelihood distributions on primary parameter values .
figure [ dde ] then confirms that such distributions , although slightly displaced , have similar width for all primary parameters . a general conclusion this discussion allows to draw is to beware from ever assuming that error estimates on secondary parameters , as @xmath3 , @xmath241 , and @xmath2 , are safe
in cosmologies including cold dm and de , the equation obeyed by the stress energy tensors of the dark components @xmath242 yields @xmath243 if their pressure and energy densities are @xmath244 , @xmath245 and @xmath246 @xmath247 @xmath18 is the conformal time and dots indicate differentiation in respect to @xmath248 eqs .
( [ conti0 ] ) and ( [ conti ] ) state that no force , apart gravity , acts between standard model particles and the dark components .
( [ conti0 ] ) is fulfilled if dm and de separately satisfy the eqs .
@xmath249 where @xmath250 are traces of the stress energy tensors .
@xmath251 can be an arbitrary constant or , _
e.g. _ , a function of @xmath19 itself ; when @xmath252 , the two dark components are uncoupled and so are the cases we considered up to here .
it is however known ( wetterich 1995 , amendola 2000 , amendola & quercellini 2003 ) that self consistent theories can be built with @xmath253 so that @xmath254 is used to parametrize the strength of dm de coupling . by using mcmc techniques ,
we explored cosmologies with @xmath255 and 0.1 ( model a ) ; we also report some results for @xmath256 ( model b ) .
we aim to show that , when @xmath27 is excluded from the parameter budget , the values mcmc provide for some parameters can be significantly biased , even though @xmath254 is small . 0.5truecm
0.5truecm in table [ tab1 ] ( 3rd4th columns ) we report the results of fitting the parameter set including constant @xmath0 , in place of @xmath21 and @xmath27 .
the table allows an easy comparison with the uncoupled case , when all parameters unrelated to de are nicely recovered . on the contrary , already for @xmath255 ,
the input @xmath1 value is formally @xmath257 s away from best
fit , in the fiducial case , and more than 5 @xmath25 s in some realization . the effect is stronger for @xmath258 , yielding a discrepancy @xmath259 @xmath25 s . that coupling affects the detection of @xmath1 is not casual : in coupled models a continuous exchange of energy between dm and de occurs . several parameters exhibit a non gaussian distribution , however .
this is visible in figure [ couple ] , where we also provide a comparison between likelihood distributions when ( i ) the fit includes just @xmath0 or , instead , ( ii ) the parameter space includes @xmath21 and @xmath27 .
the comparison allows to appreciate that , in the ( i ) case , @xmath1 is apparently much better determined than in the ( ii ) case .
appreciable discrepancies concern also the @xmath31 parameter and they reflect onto the derived parameters @xmath241 , @xmath3 and @xmath2 , whose distribution however appears significantly non gaussian .
these results deserve to be accompanied by a comparison between the likelihood values in the different cosmologies .
the following table shows @xmath260 for the fiducial cases only : @xmath261 the likelihood values for realizations are systematically smaller , as expected , but confirm the lack of significance shown here .
given that coupled models have 1 additional parameter over the @xmath262 model , differences @xmath263 indicate that both models are an equally good fit of data .
the conclusion is that spectral differences , between input and best fit models , lay systematically below the error size . as a matter of fact
, we are apparently meeting a case of degeneracy .
this statistical observation is corroborated by a direct insight into angular spectra , provided by figures [ cla ] and [ clb ] .
we show the behaviors of the @xmath12 spectra for the model a , with @xmath123 , @xmath126 , @xmath127 , compared with the @xmath12 for the best fitting ( @xmath121 ) model .
differences are so small to be hardly visible .
figure [ cla ] refers just to anisotropy . in its lower panel
the @xmath264 ratio is compared with wmap5 1@xmath25 error size ( @xmath265 is the difference between the spectra of the input and best fit models ) .
notice that shifts by 1 or 2 units along the abscissa , at large @xmath11 values , would still badly cut off the apparent ratios .
clearly , degeneracy can be removed only if the error size is reduced by a factor @xmath266 or more , in the region between the first deep and the second acoustic peak . a further example we wish to add concerns a still smaller coupling intensity , @xmath256 ( model b , table [ tab02 ] ) . at this coupling level ,
the mcmc meet all input parameter values within a couple of @xmath25 s .
a closer inspection of plots similar to figure [ couple ] , not reported here , shows that the probability of realizations yielding @xmath1 formally more than 3@xmath25 s away from the true value , is still @xmath267 the genie is still not completely back inside the lamp .
cc + & input value & b.f .
value @xmath128 + @xmath268 & @xmath269 + & @xmath270 + & @xmath271 + & @xmath272 + @xmath273 & @xmath274 + & @xmath275 + & @xmath276 + & @xmath277 + @xmath152 & @xmath278 + & @xmath279 + & @xmath280 + & @xmath281 + @xmath282 & @xmath283 + & @xmath284 + & @xmath285 + & @xmath286 + @xmath0 ( at @xmath287 ) & @xmath288 + & @xmath289 + & @xmath290 + & @xmath291 + @xmath292 & @xmath293 + & @xmath294 + & @xmath295 + & @xmath296 + @xmath297 & @xmath298 + & @xmath299 + & @xmath300 + & @xmath301 + @xmath302 & @xmath303 + & @xmath304 + & @xmath304 + & @xmath305 + @xmath306 & @xmath307 + & @xmath308 + & @xmath309 + & @xmath310 +
the degeneracy observed in cmb spectra could be broken off through different observables .
we plan to deepen this aspect in a forthcoming work , by building detailed data sets accounting from the dependence on cosmology of the expansion rate and growth factor .
a first insight into the actual situations can be however gained through an inspection of hubble diagrams and transfer functions . in figure [ hubdia ]
we show the redshift dependence of the luminosity distances for models with @xmath123 , 0.05 , 0.1 and compare them with the corresponding best fit models with constant @xmath0 , as well as with scdm models with the same value of @xmath311 .
these plots clearly indicate that a fit with snia data would hardly allow any discriminatory signal : discrepancies from constant@xmath0 model increase with @xmath27 ; but , even for the @xmath312 case , they hardly exceed @xmath313 of the difference from scdm . in figure [ trans ]
we then exhibit the transfer functions @xmath314 ( multiplied by @xmath315 to improve the visibility of details ) for models a. in the cases @xmath316 and @xmath255 , we notice slight displacements for the bao system and the slope .
the actual setting of bao s is however subject to non linear effects and residual theoretical uncertainties are wider than the shifts in the plots .
the change of slope is also easily compensated by a shift of @xmath23 by @xmath317 , widely within expected observational errors .
notice then that the relative position of the input and best fit functions is opposite in the two cases .
accordingly , in the intermediate case @xmath318 , not shown in the plot , the overlap is almost exact and the observed scale dependence of the growth factor , at low redshift , does not break the degeneracy at all .
the situation is different when @xmath319 is considered . for @xmath258 ,
the bao displacement is indeed relevant ; to compensate for the change of slope , we would then require a shift of @xmath23 greater than 0.1@xmath37 .
a numerical analysis can therefore determine at which value , probably intermediate between 0.05 and 0.1 , the cmb spectra degeneracy is broken .
the nature of the dark cosmic components is unclear .
de could be a self interacting field , yielding a scale dependent state parameter @xmath320 which bias does then arise on cosmological parameter estimates , if performed by assuming @xmath321 ? this question regards also parameters which do not describe de ; their estimate could be biased because the _ true _ model is not directly explored .
a first conclusion of this work is that such bias exists but , in the cases we treated , yields acceptable displacements , within 1@xmath25 .
suppose however that future data allow to exclude @xmath322 , within 3@xmath25 s , when we assume @xmath321 .
setting @xmath323 discriminates among inflationary potentials .
it would be however legitimate to assess that , at that stage , such conclusions would still be premature .
our analysis was then extended to the case of dm de coupling .
the idea that dm and de have related origins or arise from the same field has been widely pursued ( see , _ e.g. _ kamenshchik , moschella & pasquier 2001 , bento , bertolami sen 2002 , 2004 ; mainini & bonometto 2004 ; for a review , see copeland , sami & tsujikawa 2006 ) .
coupling causes dm de energy exchanges and this option was first explored in the attempt to ease the _ coincidence _ problem .
unfortunately , when a dm de coupling , strong enough to this aim , is added to models , predictions disagree with data .
this does not forbid , however , that the physics of the dark components includes a weaker coupling or that a stronger coupling is compensated by other features ( see , _ e.g. _ , la vacca et al .
2008 ) . in this work we pointed out that a significant degeneracy exists , so that we can find an excellent fit of cmb data for coupled de cosmologies just by using constant@xmath0 uncoupled cosmologies .
the fit is so good that even likelihood estimates do not allow to distinguish between the `` true '' cosmology and the best
fit constant@xmath0 model , at the present or foreseeable sensitivity levels .
_ unfortunately , however , the values obtained for several parameters are then widely different from input ones . _ in fact
, if we ignore the coupling degree of freedom , when data are analyzed , we can find biased values for some primary parameter as @xmath1 , for which input values lay @xmath324@xmath25 s away from what is `` detected '' .
also @xmath325 and @xmath2 , which are secondary parameters , are significantly biased . in particular , with a coupling as low as @xmath255 , we found model realisations yielding @xmath2 estimate @xmath326@xmath25 s away from the input value . if we keep to the fiducial case , however , the probability to find @xmath327 its input value is 36.5@xmath328 for @xmath255 and 26.5@xmath328 for @xmath329 this outlines the tendency to find smaller values than `` true '' ones . as a matter of fact , however , because of the uncertainty induced by our ignorance on de state equation , the width of errors on secondary parameter significantly exceeds the width for primary ones , and a 5@xmath25 discrepancy is partially hidden by such ignorance .
this agrees with the fact that data analysis shows a significant increase of errors on secondary parameters when the set of models inspected passes from @xmath21cdm to generic@xmath0 cosmologies .
for instance , in the very wmap5 analysis , the error on @xmath330 increases by a huge factor @xmath331 , when one abandons the safety of @xmath21cdm to explore generic constant@xmath0 models .
a general warning is then that errors obtained assuming @xmath21cdm are to be taken with some reserve .
an important question is whether the bias persists when different observables , besides cmb , are used .
a preliminary inspection shows that snia data would hardly provide any discrimination . on the contrary
an analysis of fluctuation growth can be discriminatory if @xmath3320.08 .
an even more discriminatory signal could be found in the @xmath333 dependence of the growth factor is suitably analyzed .
observational projects aiming at performing a tomography of weak lensing , like dune euclid ( see , _
e.g. _ , refregier et al .
2006 , 2008 ) , therefore , can be expected to reduce this degeneracy case . | when cmb data are used to derive cosmological parameters , their very choice does matter : some parameter values can be biased if the parameter space does not cover the `` true '' model .
this is a problem , because of the difficulty to parametrize dark energy ( de ) physics .
we test this risk through numerical experiments .
we create artificial data for dynamical or coupled de models and then use mcmc techniques to recover model parameters , by assuming a constant de state parameter @xmath0 and no dm de coupling . for the de potential considered ,
no serious bias arises when coupling is absent . on the contrary , @xmath1 , and thence @xmath2 and @xmath3 ,
suffer a serious bias when the `` true '' cosmology includes even just a mild dm de coupling . until the dark components keep an unknown nature ,
therefore , it can be important to allow for a degree of freedom accounting for dm de coupling , even more than increasing the number of parameters accounting for the @xmath4 behavior . |
efron and morris ( 1975 ) demonstrated the benefit of simultaneous estimation using a simple example of using the batting outcomes of 18 players in the first 45 at - bats in the 1970 season to predict their batting average for the remainder of the season . essentially , improved batting estimates shrink the observed averages towards the average batting average of all players .
one common way of achieving this shrinkage is by means of a random effects model where the players underlying probabilities are assumed to come from a common distribution , and the parameters of this `` random effects '' distribution are assigned vague prior distributions . in modern sabermetrics research
, a batting average is not perceived to be a valuable measure of batting performance .
one issue is that the batting average assigns each possible hit the same value , and it does not incorporate in - play events such as walks that are beneficial for the general goal of scoring runs .
another concern is that the batting average is a convoluted measure that combines different batting abilities such as not striking out , hitting a home run , and getting a hit on a ball placed in - play .
indeed , it is difficult to really say what it means for a batter to `` hit for average '' .
similarly , an on - base percentage does not directly communicate a batter s different performances in drawing walks or getting base hits .
a deeper concern about a batting average is that chance plays a large role in the variability of player batting averages , or the variability of a player s batting average over seasons .
albert ( 2004 ) uses a beta - binomial random effects model to demonstrate this point .
if a group of players have 500 at - bats , then approximately half the variability in the players batting average is due to chance ( binomial ) variation the remaining half is due to variability in the underlying player s hitting probabilities .
in contrast , other batting rates are less affected by chance . for example
, only a small percentage of players observed home run rates are influenced by chance
much of the variability is due to the differences in the batters home run abilities .
the role of chance has received recent attention to the development of fip ( fielding independent pitching ) measures .
mccracken ( 2001 ) made the surprising observation that pitchers had little control of the outcomes of balls that were put in - play .
one conclusion from this observation is that the babip , batting average on balls put in - play , is largely influenced by luck or binomial variation , and the fip measure is based on outcomes such as strikeouts , walks , and home runs that are largely under the pitcher s control .
following bickel ( 2004 ) , albert ( 2004 ) illustrated the decomposition of a batting average into different components and discussed the luck / skill aspect of different batting rates .
in this paper , similar decompositions are used to develop more accurate predictions of a collection of batting averages . essentially , the main idea is to first represent a hitting probability in terms of component probabilities , estimate groups of component probabilities by means of random effects models , and use these component probability estimates to obtain accurate estimates of the hitting probabilities .
sections 3 , 4 , 5 illustrate the general ideas for the problem of simultaneously estimating a collection of `` batting average '' probabilities and section 8 demonstrates the usefulness of this scheme in predicting batting averages for a following season .
section 7 illustrates this plan for estimating on - base probabilities .
section 8 gives a historical perspective on how the different component hitting rates have changed over time , and section 9 illustrates the use of this perspective in understanding the career trajectories of hitters and pitchers .
the fip measure is shown in section 10 as a function of particular hitting rates and this representation is used to develop useful estimates of pitcher fip abilities .
section 11 concludes by describing several modeling extensions of this approach .
since efron and morris ( 1975 ) , there is a body of work finding improved measures of performance in baseball .
tango et al ( 2007 ) discuss the general idea of estimating a player s true talent level by adjusting his past performance towards the performance of a group of similar players and the appendix gives the familiar normal likelihood / normal prior algorithm for performing this adjustment .
brown ( 2008 ) , mcshane et al ( 2011 ) , neal et al ( 2010 ) , and null ( 2009 ) propose different `` shrinking - type '' methods for estimating batting abilities for major league batters
. similar types of methods are proposed by albert ( 2006 ) , piette and james ( 2012 ) , and piette et al ( 2010 ) for estimating pitching and fielding metrics .
albert ( 2002 ) and piette et al ( 2012 ) focus on the problem of simulataneously estimating player hitting and fielding trajectories .
albert ( 2004 ) , bickel ( 2004 ) and bickel and stotz ( 2003 ) describe decomposition of batting averages .
baumer ( 2008 ) performs a similar decomposition of batting average ( @xmath0 ) and on - base percentage ( @xmath1 ) with the intent of showing mathematically that @xmath0 is more sensitive than @xmath1 to the batting average on balls in - play .
the basic decomposition of a batting average is illustrated in figure 1 .
suppose one divides all at - bats into strikeouts ( so ) and not - strikeouts ( not so ) . of the ab that are not strikeouts , we divide into the home runs ( hr ) and the balls that are put `` in - play '' .
finally , we divide the balls in - play into the in - play hits ( hip ) and the in - play outs ( oip ) .
this representation leads to a decomposition of the batting average @xmath2 .
we first write the proportion of hits as the proportion of ab that are not strikeouts times the proportion of hits among the non - strikeouts .
@xmath3 continuing , if we breakdown these @xmath4 opportunities by @xmath5 , then we write the hit proportion as the proportion of non - strikeouts that are home runs plus the proportion of non - strikeouts that are singles , doubles , or triples @xmath6 .
@xmath7 finally , we write the proportion of non - strikeouts that are singles , doubles , or triples as the proportion of non - strikeouts that are not home runs times the proportion of balls in - play ( @xmath8 ) that are hits
. @xmath9 putting it all together , we have the following representation of a batting average @xmath10 : @xmath11 where the relevant rates are : * the strikeout rate @xmath12 * the home run rate @xmath13 * the batting average on balls in play rate @xmath14 . instead of simply recording hits and outs
, we are regarding the outcomes of an at - bat as multinomial data with the four outcomes so , hr , hip , and oip .
the decomposition of a batting average leads to a multinomial sampling model for the hitting outcomes , and the multinomial sampling leads to an attractive representation of the likelihood of the underlying probabilities .
there are four outcomes of an at - bat : so , hr , hip , and oip ( strikeout , home run , hit - in - play , and out - in - play ) .
let @xmath15 denote the probability that an at - bat results in a strikeout , let @xmath16 denote the probability a non - strikeout results in a home run , and @xmath17 denotes the probability that a ball - in - play ( not so or hr ) results in an in - play hit .
if this player has @xmath18 at - bats , the vector of counts of so , hr , hip , and oip is multinomial with corresponding probabilities @xmath19 , and @xmath20 these expressions are analogous to the breakdowns of the hitting rates . for example , the probability a hitter gets a home run in an at - bat is equal to the probability that the person does not strike out ( @xmath21 ) , times the probability that the hitter gets a home run among all the non - strikeouts ( @xmath16 ) .
likewise , the probability a batter gets a hit in - play is the probability he does not strikeout times the probability he does not get a home run in a non - strikeout times the probability he gets a hit in a ball put into play .
denote the multinomial counts for a particular player as @xmath22 where @xmath18 is the total number of at - bats .
the likelihood of the associated probabilities is given by @xmath23 with some rearrangement of terms , one can show that the likelihood has a convenient factorization : @xmath24 \times \left[p_{hr}^{y_{hr } } ( 1 - p_{hr})^{n - y_{so } - y_{hr}}\right]\\ & & \times \ , \
, \left[p_{hip}^{y_{hip } } ( 1 - p_{hip})^{n - y_{so } - y_{hr } - y_{hip}}\right ] \\ & = & l_1 \times l_2
\times l_3\end{aligned}\ ] ] from the above representation , we see * @xmath25 is the likelihood for a binomial(@xmath26 ) distribution * @xmath27 is the likelihood for a binomial(@xmath28 ) distribution * @xmath29 is the likelihood for a binomial(@xmath30 ) distribution above we consider the multinomial likelihood for a single player , when in reality we have @xmath31 hitters with unique hitting probabilities . for the @xmath32th player , we have associated probabilities @xmath33 .
following the same factorization , it is straightforward to show that the likelihood of the vectors of probabilities @xmath34 , and @xmath35 is given by @xmath36
the factorization of the likelihood motivates an attractive way of simultaneously estimating the multinomial probabilities for all players .
suppose that one s prior belief about the vectors @xmath37 , @xmath38 , and @xmath39 are independent , and we represent each set of probabilities by an exchangeable model represented by a multilevel prior structure .
in particular , suppose that the strikeout probabilities are believed to be exchangeable .
one way of representing this belief is by the following mixture of betas model .
* @xmath40 are independent @xmath41 , where a @xmath42 density has the form @xmath43 and @xmath44 is the beta function .
the parameter @xmath45 is the prior mean and @xmath46 is a `` precision '' parameter in the sense that the prior variance @xmath47 is a decreasing function of @xmath46 . * the beta parameters @xmath48 )
are assigned the vague prior @xmath49 similarly , we represent a belief in exchangeability of the home run probabilities in @xmath38 by assigning a similar two - stage prior with unknown hyperparameters @xmath50 , and @xmath51 . likewise , an exchangeable prior on the hit - in - play probabilities @xmath39 is assigned with hyperparameters @xmath52 , and @xmath53 .
we saw that the likelihood function factors into independent components corresponding to the so , hr , and hip data .
since the prior distributions of the probability vectors @xmath37 , @xmath38 , and @xmath39 are independent , it follows that these probability vectors also have independent posterior distributions .
we summarize standard results about the posterior of the vector of strikeout probabilities @xmath54 with the understanding that similar results follow for the other two vectors .
the posterior distribution of the vector @xmath54 can be represented by the product @xmath55 where @xmath56 is the posterior distribution of the parameters of the random effects distribution , and @xmath57 is the posterior of the probabilities conditional on the random effects distribution parameters .
we discuss each distribution in turn .
the random effects distribution @xmath58 represents the `` talent curve '' of the players with respect to strikeouts , and the posterior mode @xmath59 of this distribution tells us about the center and spread of this distribution .
in particular , @xmath60 represents the average strikeout rate among the players , and the estimated standard deviation @xmath61 measures the spread of this talent curve .
given values of the parameters @xmath62 and @xmath63 , the individual strikeout probabilities @xmath64 , ...
, @xmath65 have independent beta distributions where @xmath66 is beta with shape parameters @xmath67 and @xmath68 , where @xmath69 and @xmath70 represent the number of strikeouts and at - bats for the @xmath32th player . using this representation , the posterior mean of the strikeout rate for the @xmath32th player
is given by @xmath71 plugging in the posterior estimates for @xmath62 and @xmath63 , we get the posterior estimate of the @xmath32th player striking out : @xmath72 the same methodology was used to estimate the home run probabilities and the hit - in - play probabilities for all players denote the three sets of estimates as @xmath73 , @xmath74 , and @xmath75 , respectively .
one can use these estimates to estimate the hitting probabilities using the representation @xmath76 for an individual estimate , by substituting the `` component '' estimates @xmath77 , @xmath78 , and @xmath79 into this expression , we obtained a `` component '' estimate of @xmath80 that we denote by @xmath81 . to illustrate the use of these exchangeable models , we collect hitting data for all players with at least 100 at - bats in the 2011 season .
( we used 100 ab as a minimum number of at - bats to exclude pitchers from the sample . ) three exchangeable models are fit , one to the collection of strikeout rates \{@xmath82 } , one to the collection of home run rates \{@xmath83 } , and one to the collection of in - play hit rates \{@xmath84}. table 1 displays values of the random effect parameters @xmath45 and @xmath46 for each of the three fits .
the average strikeout rates is about 20% , the average home run rate ( among ab removing strikeouts ) is 3.6% , and the average in - play hit rate is @xmath85 .
the estimated values of @xmath46 are informative about the spreads of the associated probabilities .
the relatively small estimated value of @xmath46 for @xmath86 reflects a high standard deviation , indicating a large spread in the player strikeout probabilities .
the estimated value of @xmath46 for home runs is also relatively small , indicating a large spread in home run probabilities .
in contrast , the estimated value of @xmath46 for hits in play is large , indicated that players abilities to get in - play hits are more similar .
.estimates of random effect parameters for strikeout data , home run data , and in - play hit data from the 2011 season . [ cols=">,>,>,>",options="header " , ] his three observed rates are @xmath87 , @xmath88 , and @xmath89 from the fitted model , these observed rates are shrunk or adjusted towards average values using the formulas : @xmath90 @xmath91 @xmath92 ( note that beltran s strikeout and home runs are slightly adjusted towards the average values due to the small estimated values of @xmath46 . in contrast
, the consequence of the large estimated @xmath93 is that beltran s in - play hit rate is adjusted about half of the way towards the average value . ) using these estimates , beltran s hit probability is estimated to be @xmath94 which is smaller than his observed batting average of 156 / 520 = 0.300 .
much of this adjustment in his batting average estimate is due to the adjustment in beltan s in - play hitting rate .
if one is primarily interested in estimating the hitting probabilities , there is a well - known simpler alternative approach based on an exchangeable model placed on the probabilities .
if the hitting probability of the @xmath32th player is given by @xmath95 , then one can assume that @xmath96 are distributed from a common beta curve with parameters @xmath46 and @xmath45 , and assign @xmath97 a vague prior .
then the posterior mean of @xmath95 is approximated by @xmath98 where the @xmath32th player is observed to have @xmath99 hits in @xmath70 at - bats and @xmath100 and @xmath101 are estimates from this exchangeable model .
the following prediction contest is used to compare the proposed component estimates \{@xmath81 } with the batting average estimates \{@xmath102 } * first we collect hitting data for all players with at least 100 at - bats in both 2011 and 2012 seasons .
* fit the component model on the hitting data for 2011 season get estimates of the strikeout , home run , and in - play hit probabilities for all hitters and use these three sets of estimates to get the component estimates \{@xmath81}. * use the hit / at - bat data for all players in the 2011 season and the single exchangeable model to compute the estimates . *
use both the component estimates and the single exchangeable estimates to predict the batting averages of the players in the 2012 season .
let @xmath103 and @xmath104 denote the number of hits and at - bats of the @xmath32th player in the 2012 season .
compute the root sum of squared prediction errors for both methods .
@xmath105 + @xmath106 the improvement in using the components estimates is @xmath107 .
a positive value of @xmath108 indicates that the component estimates are providing closer predictions than the single exchangeable estimates .
this prediction contest was repeated for each of the seasons 1963 through 2012 . batting data for all players in seasons @xmath109 and @xmath110
were collected , y = 1963 , ... , 2012 .
the component and single exchangeable models were each fit to the data in season @xmath109 and the improvement in using the component method over the single exchangeable model in predicting the batting averages in season @xmath110 was computed .
figure 2 graphs the prediction improvement as a function of the season .
it is interesting that the component estimates were not uniformly superior to the estimates from a single exchangeable model .
however the component method appears generally to be superior to the standard method , especially for seasons 1963 - 1980 and 1995 - 2012 .
we have focused on the decomposition of an at - bat . in a similar manner , one can decompose a plate appearance as displayed in figure 3 .
if we ignore sacrifice hits ( both sh and sf ) , then one can express an on - base percentage as @xmath111 if one combines walks and hit by pitches and defines the `` walk rate '' @xmath112 then one can write @xmath113 where @xmath10 is the batting average
. this representation makes it clear that an obp is basically a function of a hitter ability to draw walks , as measured by the walk rate and his batting average .
also , following the logic of the previous section , this representation suggests that one may accurately estimate a player s on - base probability by combining separate accurate estimates of his walk probability and his hitting probability . in this
setting , one can simultaneously estimate on - percentages of a group of players by separately estimating their walk probabilities and their hit probabilities .
one represents a probability that a player gets on - base @xmath114 as @xmath115 this suggests a method of estimating a collection of on - base probabilities . 1 .
estimate the walk probabilities \{@xmath116 } by use of an exchangeable model .
2 . estimate the hitting probabilities \{@xmath95 } by use of an exchangeable model .
3 . estimate the on - base probabilities by use of the formula @xmath117 where @xmath118 and @xmath81 are estimates of the walk probability and the hit probability for the @xmath32th player .
figure 4 demonstrates the value of this method in providing better predictions . as in the `` prediction contest '' of section 5.2
, we are interested in predicting the on - base probabilities for one season given hitting data from the previous season .
two prediction methods are compared the `` single exchangeable '' method fits one exchangeable data using the on - base fractions , and the `` component '' method separately estimates the walk rates and hitting rates for the players .
one evaluates the goodness of predictions by the square root of the sum of squared prediction errors and one computes the improvement in using the component procedure over the single exchangeable method .
these methods are compared for 50 prediction contests using data from each of the seasons 1963 through 2012 to predict the on - base proportions for the following season . as most of the points fall above the horizontal line at zero
, this demonstrates that the component method generally is an improvement over the one exchangeable method .
to obtain a historical perspective of the change in hitting rates , the basic exchangeable model was fit to rates for all batters with at least 100 ab for each of the seasons 1960 through 2012 .
for each season , we estimate the mean talent @xmath119 and associated precision parameter @xmath93 the associated estimated posterior standard deviation of the talent distribution is @xmath120 figure 5 displays the pattern of mean strikeout rates for all batters with at least 100 ab .
note that the average strikeout rate among batters initially showed a decrease from 1970 through 1980 but has steadily increased until the current season .
if we performed fits of the exchangeable model for all pitchers for each season from 1960 through 2012 , one would see a similar pattern in the mean strikeout rates .
figure 6 displays the estimated standard deviations of the strikeout abilities of all batters with at least 100 ab across seasons and overlays the estimated season standard deviations of the strikeout abilities of all pitchers .
first , note that among batters , the spread of the strikeout abilities shows a similar pattern to the mean strikeout rate
there is a decrease from 1970 to 1980 followed by a steady increase to the current day .
the spread of strikeout abilities among pitchers shows a different pattern .
the standard deviations for pitchers have steadily increased over seasons , and the spread in the talent distribution for pitchers is significantly smaller than the spread of the talents for batters .
one way of measuring the effectiveness of a batter or a pitcher is to look at the vector of rates @xmath121 for a particular season . plotting these rates over a player s career , one gains
a general understanding of the strengths of the batter or pitcher and learns when these players achieved peak performances .
albert ( 2002 ) demonstrates the value of looking at career trajectories to better understand the growth and deterioration of player s batting abilities .
the four observed rates have different averages and spreads , and as we see from figures 5 and 6 , the averages and spreads can change dramatically over different seasons .
we use residuals from the predictive distribution to standardize these rates .
let @xmath109 denote the number of successes in @xmath18 opportunities for a player in a particular season and suppose the underlying probabilities of the players follow a beta curve with mean @xmath45 and precision @xmath46 .
the predictive density of the rate @xmath122 has mean @xmath45 and standard deviation @xmath123 when the exchangeable model is fit , one obtains estimates of the random effects parameters @xmath119 and @xmath93 , and obtains an estimate of the standard deviation @xmath124 .
define the standardized residual @xmath125 in the following plots of the standardized residuals of the walk / hit - by - pitch rates , strikeout rates , home run rates , and hit - in - play rates will be displayed to show special strengths of hitters and pitchers .
the graphs of the standardized rates are displayed for the careers of mickey mantle in figure 7 and ichiro suzuki in figure 8 .
looking at the four graphs of figure 7 in a clockwise manner from the upper - left , one sees * mantle drew many walks / hbp and his walk / hbp rate actually increased during his career .
* mantle had an above - average strikeout rate . *
his home run rate hit a peak during the middle of his career .
* his in - play hit rate decreased towards the end of his career .
in contrast , by looking at figure 8 , one sees that suzuki had consistent low walk / hbp , strikeout , and home run rates throughout his career .
he was especially good in his hit - in - play rate , although there was much variability in these rates and showed a decrease towards the end of his career .
these displays of standardized rates are also helpful for understanding the strengths of pitchers in the history of baseball .
figures 9 and 10 display the standardized rates for the hall of fame pitchers greg maddux and steve carlton .
maddux was famous for his low walk rate and generally low era .
looking at the trajectories of his rates in figure 9 , one sees that maddux s best walk rates occurred during the last half of his career .
his best strikeout rate , home run rate , and hip rate occurred about 1995 and all three of these rates deteriorated from 1995 until his retirement in 2008 .
in contrast , one sees from figure 10 that carlton had a slightly below average walk rate and a high strikeout rate during his career .
since all of these rates significantly deteriorated towards the end of his career , perhaps carlton should have retired a few years earlier .
based on these graphs , carlton s peak season in terms of performance was about 1980 , the season when the phillies won the world series .
recently , there has been an increased emphasis on the use of fielding - independent - performance ( fip ) measures of pitchers .
the idea is to construct a measure based on the outcomes such as walks , hit - by - pitches , strikeouts , and home runs that a pitcher directly controls . the usual definition of fip is given by @xmath126 where @xmath5 , @xmath127 , @xmath128 , and @xmath86 are the counts of these different events
, @xmath129 is the innings pitched , and @xmath130 is a constant defined to ensure that the average fip is approximately equal to the league era . although @xmath131 is defined in terms of counts , it is straightforward to write it as a function of the four rates @xmath132 , @xmath133 , @xmath134 , and @xmath135
. let @xmath136 denote the count of batters faced , then @xmath137\end{aligned}\ ] ] substituting these expressions into the formula and ignoring the constant term , the @xmath131 measure is expressed solely in terms of these four rates .
although on face value , the @xmath131 measure seems to depend on the sample size ( the number of batters faced ) , the value of @xmath136 cancels out in the substitution .
all of the observed rates are estimates of the underlying probabilities of those events .
if we take the expression of @xmath131 , ignoring the constant , and replace the rates with probabilities , we get an expression for a pitcher s @xmath131 ability denoted by @xmath138 : @xmath139 using data for a single season , we can use separate exchangeable models to estimate the walk probabilities \{@xmath116 } , the strikeout probabilities \{@xmath140 } , the home run probabilities \{@xmath141 } , and the hit - in - play probabilities \{@xmath142 } for all pitchers .
if we substitute the probability estimates into the @xmath138 formula , we get new estimates at the observed @xmath131 measures for all pitchers in a particular season . based on our earlier work , one would anticipate that our new estimates of @xmath131 ability would be superior to usual estimates in predicting the @xmath131 values of the pitchers in the following season . as in our evaluation of the performance of the improved batting probabilities , the new estimates
can be compared with exchangeable estimates based on the standard representation of the @xmath131 statistic . for a given pitcher , suppose one collects the measurement @xmath143 for each inning pitched .
if the pitcher pitches for @xmath144 innings , then the measurements can be denoted by @xmath145 and the @xmath131 statistic is simply the sample mean @xmath146 .
it is reasonable to assume that @xmath147 is normal with mean @xmath138 and variance @xmath148 , where @xmath149 reflects the variability of the values of @xmath150 within innings .
based on this representation , one can estimate the @xmath131 abilities \{@xmath151 } by use of an exchangeable model where the abilities are assigned a normal curve with mean @xmath152 and standard deviation @xmath153 , and a vague prior is assigned to @xmath154 . by fitting this model , one shrinks the observed @xmath131 values for the pitchers towards an average value . again
a prediction experiment is used to predict the @xmath131 values for all pitchers from a season given these measures from the previous season .
the `` standard '' method predicts the @xmath131 values using the single exchangeable model , and the `` component '' method first separately estimates the four sets of probabilities with exchangeable models , and then substitutes these estimates in the formula to obtain @xmath131 predictions . as might be expected , the component method results in a smaller prediction error for practically all of the seasons of the study .
this again demonstrates the value of this `` divide and conquer '' approach to obtain superior estimates of pitcher characteristics that are functions of the underlying probabilities .
in the sabermetrics literature , the regression effect is well known ; to predict a batter s hitting rate for a given season , one takes one s previous season s hitting average and move this estimate towards an average .
this paper extends this approach to estimating a batting measure that is a function of different rates .
apply the random effects model to get accurate estimates at the component rates for all players , and then substitute these estimates into the function to get improved predictions of the batting measures .
this approach was easy to apply for the batting probability and on - base probabilities situations due to the convenient factorization of the likelihood and use of independent exchangeable prior distributions .
the choice of a single beta random effects curve was chosen for convenience due to attractive analytical features , but this `` component '' approach can be used for any choice of random effects model .
for example , one may wish to use covariates in modeling the probabilities that hitters get a hit on balls put in play .
if @xmath155 is the probability that the @xmath32th player gets a hit , then one could assume that @xmath156 are independent from beta@xmath157beta@xmath158 distributions where the prior means satisfy the logistic model @xmath159 where @xmath160 is a relevant predictor such as the speed of the ball off the bat .
as before , the prior parameters @xmath161 would be assigned a weakly informative prior to complete the model .
the fip measure was motivated from the basic observation that a team defense , not just a pitcher , prevents runs , and one wishes to devise alternative measures that isolate a pitcher s effectiveness . in a similar fashion , the goal here is to isolate the different components of a hitter s effectiveness .
these component estimates are useful by themselves , but they are also helpful in estimating ensemble measures of ability such as the probability of getting on base .
mcshane , b. , braunstein , a. , piette , j. and jensen , s. ( 2011 ) , `` a hierarchical bayesian variable selection approach to major league baseball hitting metrics , '' _ journal of quantitative analysis in sports _ , 7 , issue 2 . | standard measures of batting performance such as a batting average and an on - base percentage can be decomposed into component rates such as strikeout rates and home run rates .
the likelihood of hitting data for a group of players can be expressed as a product of likelihoods of the component probabilities and this motivates the use of random effects models to estimate the groups of component rates .
this methodology leads to accurate estimates at hitting probabilities and good predictions of performance for following seasons .
this approach is also illustrated for on - base probabilities and fip abilities of pitchers . |
understanding biodiversity and coevolution is a central challenge in modern evolutionary and theoretical biology @xcite . in this context , for some decades
much effort has been devoted to mathematically model dynamics of competing populations through nonlinear , yet deterministic , set of rate equations like the equations devised by lotka and volterra @xcite or many of their variants @xcite .
this heuristic approach is often termed as population - level description . as a common feature
, these deterministic models fail to account for stochastic effects ( like fluctuations and spatial correlations ) .
however , to gain some more realistic and fundamental understanding on generic features of population dynamics and mechanisms leading to biodiversity , it is highly desirable to include internal stochastic noise in the description of agents kinetics by going beyond the classical deterministic picture .
one of the main reasons is to account for discrete degrees of freedom and finite - size fluctuations @xcite .
in fact , the deterministic rate equations always ( tacitly ) assume the presence of infinitely many interacting agents , while in real systems there is a _
large _ , yet _
finite _ , number of individuals ( recently , this issue has been addressed in refs .
@xcite ) . as a consequence ,
the dynamics is intrinsically stochastic and the unavoidable finite - size fluctuations may have drastic effects and even completely invalidate the deterministic predictions .
interestingly , both _ in vitro _
@xcite and _ in vivo _
@xcite experiments have recently been devoted to experimentally probe the influence of stochasticity on biodiversity : the authors of refs .
@xcite have investigated the mechanism necessary to ensure coexistence in a community of three populations of _ escherichia coli _ and have numerically modelled the dynamics of their experiments by the so - called ` rock - paper - scissors ' model , well - known in the field of game theory @xcite .
this is a three - species cyclic generalization of the lotka - volterra model @xcite . as a result ,
the authors of ref .
@xcite reported that in a well - mixed ( non - spatial ) environment ( i.e. when the experiments were carried out in a flask ) two species got extinct after some finite time , while coexistence of the populations was never observed .
motivated by these experimental results , in this work we theoretically study the stochastic version of the cyclic lotka - volterra model and investigate in detail the effects of finite - size fluctuations on possible population extinction / coexistence . for our investigation , as suggested by the flask experiment of ref .
@xcite , the stochastic dynamics of the cyclic lotka - volterra model is formulated in the natural language of urn models @xcite and by adopting the so - called individual - based description @xcite . in the latter , the explicit rules governing the interaction of a _
finite number _ of individuals with each other are embodied in a master equation .
the fluctuations are then specifically accounted for by an appropriate fokker - planck equation derived from the master equation via a so - called van kampen expansion @xcite .
this program allows us to quantitatively study the deviations of the stochastic dynamics of the cyclic lotka - volterra model with respect to the rate equation predictions and to address the question of the extinction probability , the computation of which is the main result of this work . from a more general perspective
, we think that our findings have a broad relevance , both theoretical and practical , as they shed further light on how stochastic noise can dramatically affect the properties of the numerous nonlinear systems whose deterministic description , like in the case of the cyclic lotka - volterra model , predicts the existence of neutrally stable solutions , i.e. cycles in the phase portrait @xcite .
this paper is organized as follows : the cyclic lotka - volterra model is introduced in the next section and its deterministic rate equation treatment is presented . in section [ stoch_appr ] , we develop a quantitative analytical approach that accounts for stochasticity , a fokker - planck equation is derived from the underlying master equation within a van kampen expansion .
this allows us to compute the variances of the agents densities .
we also study the time - dependence properties of the system by carrying out a fourier analysis from a set of langevin equations .
section [ sect - ext - prob ] is devoted to the computation of the probability of having extinction of two species at a given time , which constitutes the main issue of this work . in the final section ,
we summarize our findings and present our conclusions .
, @xmath0 , and @xmath1 .
the latter may correspond to the strategies in a rock - paper - scissors game @xcite , or to different bacterial species @xcite .
[ cycle ] ] the cyclic lotka volterra model under consideration here is a system where three states @xmath2 , @xmath0 , and @xmath1 cyclically dominate each other : @xmath2 invades @xmath0 , @xmath0 outperforms @xmath1 , and @xmath1 in turn dominates over @xmath2 , schematically drawn in fig .
[ cycle ] .
these three states @xmath2 , @xmath0 , and @xmath1 allow for various interpretations , ranging from strategies in the rock - paper - scissors game @xcite over tree , fire , ash in forest fire models @xcite or chemical reactions @xcite to different bacterial species @xcite .
in the latter case , a population of poison producing bacteria was brought together with another being resistant to the poison and a third which is not resistant . as the production of poison as well as the resistance against it have some cost , these species show a cyclic dominance : the poison - producing one invades the non - resistant , which in turn reproduces faster than the resistant one , and the latter finally dominates the poison - producing . in a well - mixed environment , like a flask in the experiments @xcite , eventually only one species survives .
the populations are large but finite , and the dynamics of reproduction and killing events may to a good approximation be viewed as stochastic . ) , red or medium gray ( @xmath0 ) , and blue or dark gray ( @xmath1 ) . at each time step ,
two random individuals are chosen ( indicated by arrows in the left picture ) and react ( right picture ) .
, title="fig : " ] ) , red or medium gray ( @xmath0 ) , and blue or dark gray ( @xmath1 ) . at each time step ,
two random individuals are chosen ( indicated by arrows in the left picture ) and react ( right picture ) .
, title="fig : " ] motivated from these biological experiments , we introduce a stochastic version of the cyclic lotka - volterra model .
consider a population of size @xmath3 which is well mixed , i.e. in the absence of spatial structure .
the stochastic dynamics used to describe its evolution , illustrated in fig . [ urn ] , is referred to as `` urn model '' @xcite and closely related to the moran process @xcite . at every time step ,
two randomly chosen individuals are selected , which may at certain probability react according to the following scheme : @xmath4 with reaction rates @xmath5 , and @xmath6 .
we observe the cyclic dominance of the three species .
also , the total number @xmath3 of individuals is conserved by this dynamics ; this will of course play a role in our further analysis .
we now proceed with the analysis of the deterministic version of the system ( [ react ] ) .
this will prove insightful for building a stochastic description of the model , which is the scope of sec .
[ stoch_appr ] .
the deterministic rate equations describe the time evolution of the densities @xmath7 , @xmath8 , and @xmath9 for the species @xmath2 , @xmath0 , and @xmath1 ; they read @xmath10 where the dot stands for the time derivative . these equations
describe a well - mixed system , without any spatial correlations , as naturally implemented in urn models @xcite or , equivalently , infinite dimensional lattice systems or complete graphs . in the following , the eqs .
( [ re ] ) are discussed and , from their properties , we gain intuition on the effects of stochasticity . already from the basic reactions ( [ react ] ) we have noticed that the total number of individuals is conserved , which is a property correctly reproduced by the the rate equations ( [ re ] ) . setting the total density , meaning the sum of the densities @xmath11 , @xmath12 , and @xmath13 , to unity , we obtain @xmath14 for all times @xmath15 .
only two out of the three densities are thus independent , we may view the time evolution of the densities in a two - dimensional phase space .
( [ re ] ) together with ( [ total_dens ] ) admit three trivial ( absorbing ) fixed points : @xmath16 ; @xmath17 ; and @xmath18 .
they denote states where only one of the three species survived , the other ones died out .
in addition , the rate equations ( [ re ] ) also predict the existence of a fixed point @xmath19 which corresponds to a reactive steady state , associated with the coexistence of all three species : @xmath20 to determine the nature of this fixed point , we observe that another constant of motion exists for the rate equations ( [ re ] ) , namely the quantity @xmath21 does not evolve in time .
in contrast to the total density ( [ total_dens ] ) , this constant of motion is only conserved by the rate equations but does not stem from the reaction scheme ( [ react ] ) .
hence , when considering the stochastic version of the cyclic model , the total density remains constant but the expression ( [ const2 ] ) will no longer be a conserved quantity .
the above fixed point ( [ fp - c ] ) and constant of motion ( [ const2 ] ) have been derived and discussed also within the framework of game theory , see e.g. ref .
@xcite . in fig .
[ simplex_as ] , we depict the ternary phase space @xcite for the densities @xmath11 , @xmath12 , and @xmath13 : the solutions of the rate equations ( [ re ] ) are shown for different initial conditions and a given set of rates @xmath22 , @xmath23 , and @xmath6 . as the rate equations ( [ re ] ) are nonlinear in the densities , we can not solve them analytically , but use numerical methods .
due to the constant of motion , the solutions yield cycles around the reactive fixed point ( thus corresponding to case 3 in durrett and levin s classification @xcite ) . the three trivial steady states , corresponding to saddle points within the linear analysis , are the edges of the simplex .
the reactive stationary state , as well as the cycles , are neutrally stable , stemming from the existence of the constant of motion ( [ const2 ] ) . especially , the reactive fixed point is a _ center fixed point_. the boundary of the simplex denotes states where at least one of the three species died out ; as cyclic dominance is lost , states that have reached this boundary will evolve towards one of the edges , making the boundary _ absorbing_. .
the rate equations predict cycles , which are shown in black .
their linearization around the reactive fixed point is solved in proper coordinates @xmath24 ( blue or dark gray ) .
the red ( or light gray ) erratic flow denotes a single trajectory in a finite system ( @xmath25 ) , obtained from stochastic simulations .
it spirals out from the reactive fixed point , eventually reaching an absorbing state .
[ simplex_as ] ] the nonlinearity of eqs .
( [ re ] ) induces substantial difficulties in the analytical treatment .
however , much can already be inferred from the linearization around the reactive fixed point ( [ fp - c ] ) , which we will consider in the following .
we therefore introduce the deviations from the reactive fixed point , denoted as @xmath26 : @xmath27 using conservation of the total density ( [ total_dens ] ) , we can eliminate @xmath28 , and the remaining linearized equations ( [ re ] ) may be put into the form @xmath29 , with the vector @xmath30 and the matrix @xmath31 the reactive fixed point is associated to the eigenvalues @xmath32 of @xmath33 , where @xmath34 is given by @xmath35 oscillations with this frequency arise in its vicinity . in proper coordinates @xmath36 , these oscillations are harmonic , being the solution of @xmath37 , with @xmath38 .
for illustration , we have included these coordinates , in which the solutions take the form of circles around the origin , in fig .
[ simplex_as ] .
the linear transformation @xmath39 is given by @xmath40 the equations @xmath41 are easily solved : @xmath42 eventually , we obtain the solutions for the linearized rate equations : @xmath43\cos(\omega_0 t)\cr & + \omega_0\bigg\{\frac{1}{k_c}[a(0)-a^*]+\frac{k_b+k_c}{k_bk_c}[b(0)-b^*]\bigg\}\sin(\omega_0 t)\cr \label{re - lin - sol}\end{aligned}\ ] ] where @xmath8 and @xmath9 follow by cyclic permutations .
to establish the validity of the linear analysis ( [ a - lin])-([re - lin - sol ] ) , we compare the ( numerical ) solution of the rate equations ( [ re ] ) with the linear approximations ( [ re - lin - sol ] ) . as shown in fig .
[ dens_comp ] , when @xmath44 , the agreement between the nonlinear rate equations ( [ re ] ) and the linear approximation ( [ re - lin - sol ] ) is excellent , both curves almost coincide . on the other hand , the nonlinear terms appearing in eqs .
( [ re ] ) become important already when @xmath45 and are responsible for significant discrepancies both in the amplitudes and frequency from the predictions of eq .
( [ re - lin - sol ] ) .
( color online ) the deterministic time - evolution of the density @xmath11 for small and large amplitudes .
the prediction ( [ re - lin - sol ] ) , shown in black , is compared to the numerical solution ( red or gray ) of the rate equations ( [ re ] ) . for small amplitudes ( @xmath46 ) ,
both coincide .
however , for large amplitudes ( @xmath47 ) , they considerably differ both in amplitude and frequency .
we used reaction rates @xmath48 . ]
we now aim at introducing a measure of distance @xmath49 to the reactive fixed point within the phase portrait . in the next section , this quantity will help quantify effects of stochasticity . as it was recently proved useful in a related context @xcite , we aim at taking the structure of cycles predicted by the deterministic equations , see fig .
[ simplex_as ] , into account by requiring that the distance should not change on a given cycle . motivated from the constant of motion ( [ const2 ] ) ,
we introduce @xmath50 with the normalization factor ( see below ) @xmath51 being conserved by the eqs .
( [ re ] ) , @xmath49 remains constant on every deterministic cycle .
as it vanishes at the reactive fixed point and monotonically grows when departing from it , @xmath49 yields a measure of the distance to the latter . expanding the radius @xmath49 in small deviations @xmath52 from the reactive fixed point results in @xmath53\cr & + o({\bf x}^2)\quad .
\label{r - x}\end{aligned}\ ] ] in the variables @xmath36 , with our choice for @xmath54 , it simplifies to @xmath55 corresponding to the radius of the deterministic circles , which emerge in the variables @xmath56 .
the fact that the number @xmath3 of particles is _ finite _ induces fluctuations that are not accounted for by eqs .
( [ re ] ) . in the following ,
our goal is to understand the importance of fluctuations and their effects on the deterministic picture ( [ re ] ) .
we show that , due to the neutrally stable character of the deterministic cycles , fluctuations have drastic consequences . intuitively , we expect that in the presence of stochasticity , each trajectory performs a random walk in the phase portrait , interpolating between the deterministic cycles ( as will be revealed by considering @xmath49 ) , eventually reaching the boundary of the phase space .
there , the cyclic dominance is completely lost , as one of the species gets extinct .
of the two remaining ones , one species is defeating the other , such that the latter soon gets extinct as well , leaving the other one as the only survivor ( this corresponds to one of the trivial fixed points , the edges of the ternary phase space ) .
we thus observe the boundary to be absorbing , and presume the system to always end up in one of the absorbing states .
a first indication of the actual emergence of this scenario can be inferred from the stochastic trajectory shown in fig .
[ simplex_as ] .
in this section , we set up a stochastic description for the cyclic lotka - volterra model in the urn formulation with finite number @xmath3 of individuals . starting from the master equation of the stochastic process , we obtain a fokker - planck equation for the time evolution of the probability @xmath57 of finding the system in state @xmath58 at time @xmath15 .
it allows us to gain a detailed understanding of the stochastic system .
in particular , we will find that , as anticipated at the end of the last section , after long enough time , the system reaches one of the absorbing states .
our main result is the time - dependence of the extinction probability , being the probability that , starting at a situation corresponding to the reactive fixed point , after time @xmath15 two of the three species have died out . it is obtained through mapping onto a known first - passage problem .
we compare our analytical findings to results from stochastic simulations . for the sake of clarity and without loss of generality , throughout this section , the case of equal reaction rates @xmath59 is considered .
details on the unequal rates situation are relegated to appendix [ app_gen_rates ] .
we carried out extensive stochastic simulations to support and corroborate our analytical results .
an efficient simulation method originally due to gillespie @xcite was implemented for the reactions ( [ react ] ) .
time and type of the next reaction taking place are determined by random numbers , using the poisson nature of the individual reactions . for the extinction probability , to unravel the universal time - scaling , system sizes ranging from @xmath60 to @xmath61
were considered , with sample averages over @xmath62 realizations .
let us start with the master equation of the processes ( [ react ] ) .
we derive it in the variables @xmath63 , which were introduced in ( [ x ] ) as the deviations of the densities from the reactive fixed point . using the conservation of the total density , @xmath64 ,
@xmath28 is eliminated and @xmath30 kept as independent variables .
the master equation for the time - evolution of the probability @xmath65 of finding the system in state @xmath52 at time @xmath15 thus reads @xmath66 where @xmath67 denotes the transition probability from state @xmath52 to the state @xmath68 within one time step ; summation extends over all possible changes @xmath69 .
we choose the unit of time so that , on average , every individual reacts once per time step .
+ according to the kramers - moyal expansion of the master equation to second order in @xmath69 , this results in the fokker - planck equation @xmath70+\frac{1}{2}\partial_i\partial_j[\mathcal{b}_{ij}({\bf x})p({\bf x},t ) ] ~. \label{fokker_planck}\ ] ] here , the indices @xmath71 stand for @xmath2 and @xmath0 ; in the above equation , the summation convention implies summation over them .
the quantities @xmath72 and @xmath73 are , according to the kramers - moyal expansion : @xmath74 note that @xmath75 is symmetric . for the sake of clarity , we outline the calculation of @xmath76 : the relevant changes @xmath77 in the density @xmath11 result from the basic reactions ( [ react ] ) , they are @xmath78 in the first reaction , @xmath79 in the second and @xmath80 in the third .
the corresponding rates read @xmath81 for the first reaction ( the prefactor of @xmath3 enters due to our choice of time scale , where @xmath3 reactions occur in one unit of time ) , and @xmath82 for the third , resulting in @xmath83 .
the other quantities are calculated analogously , yielding + @xmath84 with @xmath85 .
van kampen s linear noise approximation @xcite further simplifies these quantities . in this approach
, the values @xmath86 are expanded around @xmath87 to the first order .
as they vanish at the reactive fixed point , we obtain : @xmath88 where the matrix elements @xmath89 are given by @xmath90 the matrix @xmath33 , already given in ( [ a - lin ] ) , embodies the deterministic evolution , while the stochastic noise is encoded in @xmath75 .
to take the fluctuations into account within the van kampen expansion , one approximates @xmath75 by its values at the reactive fixed point .
hence , we find : @xmath91 and the corresponding fokker - planck equation reads @xmath92+\frac{1}{2}\mathcal{b}_{ij}\partial_i\partial_jp({\bf x},t ) \quad .
\label{gen - f - p}\ ] ] for further convenience , we now bring eq .
( [ gen - f - p ] ) into a more suitable form by exploiting the polar symmetry unveiled by the variables * y*. as for the linearization of eqs .
( [ re ] ) , it is useful to rely on the linear mapping @xmath93 , with @xmath94 .
interestingly , it turns out that this transformation diagonalizes @xmath95 .
one indeed finds @xmath96 , with @xmath97 in the @xmath56 variables , the fokker - planck equation ( [ gen - f - p ] ) takes the simpler form @xmath98p({\bf y},t)\cr & + \frac{1}{12n}[\partial_{y_a}^2+\partial_{y_b}^2]p({\bf y},t)\quad .
\label{f - p - y}\end{aligned}\ ] ] to capture the structure of circles predicted by the deterministic approach , we introduce polar coordinates : @xmath99 note that in the vicinity of the reactive fixed point , @xmath100 denotes the distance @xmath49 , which is now a random variable : @xmath101 . the fokker - planck equation ( [ f - p - y ] )
eventually turns into @xmath102p(r,\phi , t)~ .\cr \label{pol - f - p}\end{aligned}\ ] ] the first term on the right - hand - side of this equation describes the system s deterministic evolution , being the motion on circles around the origin at frequency @xmath34 .
stochastic effects enter through the second term , which corresponds to isotropic diffusion in two dimensions with diffusion constant @xmath103 .
note that it vanishes in the limit of infinitely many agents , i.e. when @xmath104 .
if we consider a spherically symmetric probability distribution @xmath105 at time @xmath106 , i.e. independent of the angle @xmath107 at @xmath106 , then this symmetry is conserved by the dynamics according to eq .
( [ pol - f - p ] ) and one is left with a radial distribution function @xmath108 .
the fokker - planck equation ( [ pol - f - p ] ) thus further simplifies and reads @xmath109p(r , t)\quad .
\label{f_p_r}\ ] ] this is the diffusion equation in two dimensions with diffusion constant @xmath103 , expressed in polar coordinates , for a spherically symmetric probability distribution .
this is the case on which we specifically focus in the following . within our approximations around the reactive fixed point ,
the probability distribution @xmath110 is thus the same as for a system performing a two - dimensional random walk in the @xmath56 variables .
intuitively , such behavior is expected and its origin lies in the neutrally stable character of the cycles of the deterministic solution of the rate equations ( [ re ] ) . the cycling around the reactive fixed point does not yield additional effects when considering spherically symmetric probability distributions . furthermore , due to the existence of the constant of motion ( [ const2 ] ) , the neutral stability does not only hold in the vicinity of the reactive fixed point , but in the whole ternary phase space .
we thus expect the distance @xmath49 from the reactive fixed point , a random variable , to obey a diffusion - like equation in the whole phase space ( and not only around the reactive fixed point ) . actually , qualitatively identical behavior is also found in the general case of non - equal reaction rates @xmath22 , @xmath23 , and @xmath6 ( see appendix [ app_gen_rates ] ) , as the constant of motion ( [ const2 ] ) again guarantees the neutral stability of the deterministic cycles . in this subsection , we will use the fokker - planck equation ( [ f_p_r ] ) to investigate the time - dependence of fluctuations around the reactive fixed point .
in particular , we are interested in the time evolution of mean deviation from the latter . the average square distance @xmath111 will be found to grow linearly in time , rescaled by the number @xmath3 of individuals , before saturating .
the frequency spectrum arising from erratic oscillations around the reactive fixed point is of further interest to characterize the stochastic dynamics : we will show that the frequency ( [ freq_gen_rates ] ) predicted by the deterministic approach emerges as a pole in the power spectrum .
( color online ) the averaged square radius @xmath112 as a function of the rescaled time @xmath113 .
the blue ( or dark gray ) curve represents a single trajectory , which is seen to fluctuate widely .
the linear black line indicates the analytical prediction ( [ r_time_evol ] ) , the red ( or light gray ) one corresponds to sample averages over @xmath114 realizations in stochastic simulations .
hereby , we used a system size of @xmath61 . ] being interested in the vicinity of the reactive fixed point , in this subsection we ignore the fact that the boundary of the ternary phase space is absorbing . then
, the solution to the fokker - planck equation ( [ f_p_r ] ) with the initial condition @xmath115 , where the prefactor @xmath116 ensures normalization , is simply a gaussian : @xmath117 this result predicts a broadening of the probability distribution in time , as the average square radius increases linearly with increasing time @xmath15 : @xmath118 as the time scales linearly with @xmath3 , increasing the system size results in rescaling the time @xmath15 and the broadening of the probability distribution takes longer . to capture these findings
, we introduce the rescaled variable @xmath119 . in fig .
[ radius_time_evol ] , we compare the time evolution of the squared radius @xmath120 obtained from stochastic simulations with the prediction of eq .
( [ r_time_evol ] ) and find a good agreement in the linear regime around the reactive fixed point , where @xmath101 .
in fact , for short times , @xmath120 displays a linear time dependence , with systematic deviation from ( [ r_time_evol ] ) at longer times .
we understand this as being in part due to the linear approximations used to derive ( [ r_time_evol ] ) , but also , and more importantly , to the fact that so far we ignored the absorbing character of the boundary .
we conclude that the latter invalidates the gaussian probability distribution for longer times .
this issue , which requires a specific analysis , is the scope of section [ sect - ext - prob ] , where a proper treatment is devised .
( color online ) time evolution of the variances of the densities @xmath11 and @xmath12 when starting at the reactive fixed point .
the blue ( or dark gray ) curves correspond to a single realization , while the red ( or light gray ) ones denote averages over 1000 samples .
our results are obtained from stochastic simulations with a system size of @xmath61 .
the black line indicates the analytical predictions . ] from the finding @xmath121 together with the spherical symmetry of ( [ gaussian ] ) , resulting in @xmath122 , we readily obtain the variances of the densities @xmath11 and @xmath12 : @xmath123 according to these results , the average square deviations of the densities from the reactive fixed point grow linearly in the rescaled time @xmath124 , thus exhibiting the same behavior as we already found for the average squared radius ( [ r_time_evol ] ) . in fig .
[ fluct ] , these findings are compared to stochastic simulations for small times @xmath124 , where the linear growth is indeed recovered .
( color online ) variances of the densities @xmath11 and @xmath12 when starting at a cycle away from the fixed point .
the black lines denote our analytical results , while the red ( or gray ) ones are obtained by stochastic simulations as averages over 1000 realizations .
they are seen to agree for small times , while for larger ones the stochasticity of the system induces erratic oscillations .
these results were obtained when starting from @xmath125 . ]
we may consider fluctuations of the densities @xmath11 , @xmath12 , and @xmath13 not only around the reactive stationary state , as obtained in ( [ fluct_dens ] ) , but also as the deviations from the deterministic cycles .
we consider the latter in the linear approximation around the fixed point , given by eq .
( [ re - lin - sol ] ) .
the fluctuations around them are again described by the fokker - planck equation ( [ fokker_planck ] ) with @xmath126 and @xmath75 given by eqs .
( [ alpha_b ] ) , but now @xmath127 , @xmath128 , and @xmath28 are the deviations from the deterministic cycle : @xmath129 , where @xmath130 , @xmath131 , and @xmath132 obey eqs .
( [ re ] ) and characterize the deterministic cycles given by eq .
( [ re - lin - sol ] ) . again performing van kampen s linear noise approximation
, we obtain a fokker - planck equation of the type ( [ gen - f - p ] ) , now with the matrices @xmath133 note that the entries of the above matrices now depend on time @xmath15 via @xmath134 and @xmath135 . the fokker - planck equation ( [ gen - f - p ] ) yields equations for the time evolution of the fluctuations : @xmath136 they may be solved numerically for @xmath137 , @xmath138 , and @xmath139 , yielding growing oscillations . in fig .
[ fluct_as ] we compare these findings to stochastic simulations .
we observe a good agreement for small rescaled times @xmath113 , while the oscillations of the stochastic results become more irregular at longer times .
+ recently , it has been shown that the frequency predicted by the deterministic rate equations appears in the stochastic system due to a `` resonance mechanism '' @xcite .
internal noise present in the system covers all frequencies and induces excitations ; the largest occurring for the `` resonant '' frequency predicted by the rate equations . here ,
following the same lines , we address the issue of the characteristic frequency in the stochastic cyclic lotka - volterra model . .
it agrees with stochastic simulations ( solid ) .
the inset shows the erratic oscillations of the density of one of the species for one realization .
the system size considered is @xmath140 .
[ urn_power_spectrum ] ] in the vicinity of the reactive fixed point , the deterministic rate equations ( [ re ] ) predict density oscillations with frequency @xmath34 given in eq .
( [ freq_gen_rates ] ) . for the stochastic model ,
we now show that a spectrum of frequencies centered around this value @xmath34 arises . the most convenient way to compute this power spectrum @xmath141 from the fokker - planck equation ( [ gen - f - p ] ) is through the set of equivalent langevin equations @xcite : @xmath142 with the white noise covariance matrix @xmath75 : @xmath143 . from the fourier
transform of eq .
( [ langevin ] ) , it follows that @xmath144\cr & = \frac{4}{3n}\frac{1 + 3\omega^2}{(1 - 3\omega^2)^2 } \quad .\end{aligned}\ ] ] the power spectrum has a pole at the characteristic frequency already predicted from the rate equations ( [ re ] ) , @xmath145 . for increasing system size @xmath3 ,
the power spectrum displays a sharper alignment with this value . in fig .
[ urn_power_spectrum ] we compare our results to stochastic simulations and find an excellent agreement , except for the pole , where the power spectrum in finite systems obviously has a finite value .
these results were obtained in the vicinity of the reactive fixed point , where the linear analysis ( [ re - lin - sol ] ) applies . as already found in the deterministic description ( see fig .
[ dens_comp ] ) , when departing from the center fixed point , nonlinearities will alter the characteristic frequency .
so far , within the stochastic formulation , the fluctuations around the reactive steady state were found to follow a gaussian distribution , linearly broadening in time .
however , when approaching the absorbing boundary , the latter alters this behavior , see fig .
[ radius_time_evol ] . in the following
, we will incorporate this effect in our quantitative description .
it plays an essential role when discussing the extinction probability , which is the scope of this subsection .
the probability that one or more species die out in the course of time is of special interest within population dynamics from a biological viewpoint . when considering meta - populations formed of local patches , such questions were e.g. raised in ref .
@xcite . here
, we consider the extinction probability @xmath146 that , starting at the reactive fixed point , after time @xmath15 two of the three species have died out . in our formulation , this corresponds to the probability that after time @xmath15 the system has reached the absorbing boundary of the ternary phase space , depicted in fig .
[ simplex_as ] . considering states far from the reactive fixed point
, we now have to take into account the absorbing nature of the boundary .
this feature is incorporated in our approach by discarding the states having reached the boundary , so that a vanishing density of states occurs there . as the normalization
is lost , we do no longer deal with probability distributions : in fact , discarding states at the boundary implies a time decay of the integrated density of states [ which , for commodity , we still refer to as @xmath110 ] . ignoring the nonlinearities , the problem now takes the form of solving the fokker - planck equation ( [ f_p_r ] ) for an initial condition @xmath148 , and with the requirement that @xmath110 has to vanish at the boundary .
hereby , the triangular shaped absorbing boundary is regarded as the outermost ( degenerate ) cycle of the deterministic solutions . as the linearization in the @xmath149 variables around the reactive fixed point maps the cycles onto circles ( see above ) , the triangular boundary is mapped to a sphere as well .
although the linearization scheme on which these mappings rely is inaccurate in the vicinity of the boundary , it is possible to incorporate nonlinear effects in a simple and pragmatic manner , as shown below .
however , first let us consider the linearized problem , being a first - passage to a sphere of radius @xmath150 , which is , e.g. , treated in chapter 6 of ref . @xcite . the solution is known to be a combination of modified bessel functions of the first and second kind . actually
, the laplace transform of the density of states reads @xmath151 where @xmath103 is the diffusion constant ; @xmath152 and @xmath153 denote the bessel function of the first , resp .
second , kind and of order zero ; and @xmath150 is the radius of the absorbing sphere .
it is normalized at the initial time @xmath154 ; however , for later times @xmath15 , the total number of states @xmath155 will decay in time , as states are absorbed at the boundary .
equipped with this result , we are now in a position to calculate the extinction probability @xmath156
. it can be found by considering the probability current @xmath157 at the absorbing boundary @xcite , namely : @xmath158 whose laplace transform is given by @xmath159 the extinction probability at time @xmath15 is obtained from @xmath157 by integrating over time @xmath160 until time @xmath15 @xcite : @xmath161 .
therefore , the laplace transform of @xmath156 reads @xmath162 .
again , we notice that we can write this equation in a form that depends on the time @xmath15 only via the rescaled time @xmath113 : @xmath163 hence , for different system sizes @xmath3 , one obtains the same extinction probability @xmath164 provided one considers the same value for the scaling variable @xmath165 ( see fig .
[ urn_transition ] ) .
we can not solve the inverse laplace transform appearing in eq .
( [ ext_prob_u ] ) , but @xmath166 might be expanded according to @xcite as @xmath167 considering the first three terms in this expansion , i.e contributions up to @xmath168 , yields the approximate result @xmath169 \quad . \label{approx_ext_prob}\ ] ] numerically , we have included higher terms .
the results for contributions up to @xmath170 in ( [ exp_i0 ] ) are shown in fig .
[ urn_transition ] .
an estimate of the effective distance @xmath150 to the absorbing boundary is determined by plugging either @xmath171 , or @xmath172 , or @xmath173 into eq .
( [ r - gen ] ) , which yields @xmath174 . from fig .
[ urn_transition ] , we observe the extinction probability to be overestimated by ( [ approx_ext_prob ] ) with @xmath175 .
this stems from nonlinearities altering the analysis having led to ( [ approx_ext_prob ] ) .
however , adjusting @xmath150 on physical grounds , we are able to capture these effects .
another estimate of @xmath176 is obtained by considering the expression eq .
( [ r - x ] ) , which arises from a linear analysis . as
the extinction of two species requires @xmath177 , in the expression eq .
( [ r - x ] ) this leads to the estimate @xmath178 for @xmath150 . the comparison with fig .
[ urn_transition ] , shows that ( [ approx_ext_prob ] ) together with @xmath179 is a lower bound of @xmath176 .
a simple attempt to interpolate between the above estimates is to consider the mean value of these radii , @xmath180 , which happens to yield an excellent agreement with results from stochastic simulations , see fig .
[ urn_transition ] . for the latter ,
we have considered systems of @xmath181 , and @xmath182 individuals . rescaling the time according to @xmath183 ,
they are seen to collapse on a universal curve .
this is well described by eq .
( [ approx_ext_prob ] ) with @xmath184 as the radius of the absorbing boundary .
( color online ) the extinction probability when starting at the reactive fixed point , depending on the rescaled time @xmath113 .
stochastic simulations for different system sizes ( @xmath60 : triangles ; @xmath25 : boxes ; @xmath182 : circles ) are compared to the analytical prediction ( [ ext_prob_u ] ) .
the left ( blue or dark gray ) is obtained from @xmath174 , the right one ( red or light gray ) from @xmath178 , and the middle ( black ) corresponds to the average of both : @xmath185 . ] to conclude , we consider the mean time @xmath186 that it takes until one species becomes extinct . from @xcite we find for the rescaled mean absorption time @xmath187 which for our best estimate @xmath188 yields a value of @xmath189 .
motivated by recent _ in vitro _ and _ in vivo _ experiments aimed at identifying mechanisms responsible for biodiversity in populations of _ escherichia coli _
@xcite , we have considered the stochastic version of the ` rock - paper - scissors ' , or three - species cyclic lotka - volterra , system within an urn model formulation .
this approach allowed us to quantitatively study the effect of finite - size fluctuations in a system with a large , yet _ finite _ , number @xmath3 of agents .
while the classical rate equations of the cyclic lotka - volterra model predict the existence of one ( neutrally stable ) center fixed point , associated with the _ coexistence _ of all the species , this picture is _ dramatically _ invalidated by the fluctuations which unavoidably appear in a finite system .
the latter were taken into account by a fokker - planck equation derived from the underlying master equation through a van kampen expansion . within this scheme
, we were able to show that the variances of the densities of individuals grow in time ( first linearly ) until extinction of two of the species occurs .
in this context , we have investigated the probability for such extinction to occur at a given time @xmath15 . as a main result of this work
, we have shown that this extinction probability is a function of the scaling variable @xmath183 . exploiting polar symmetries displayed by the deterministic trajectories in the phase portrait and using a mapping onto a classical first - passage problem
, we were able to provide analytic estimates ( upper and lower bounds ) of the extinction probability , which have been successfully compared to numerical computation . from our results
, it turns out that the classical rate equation predictions apply to the urn model with a finite number of agents only for short enough time , i.e. in the regime @xmath190 . as time increases
, the probability of extinction grows , asymptotically reaching @xmath191 for @xmath192 , so that , for finite @xmath3 , fluctuations are _ always _ responsible for extinction and thus dramatically jeopardize the possibility of coexistence and biodiversity .
interestingly , these findings are in qualitative agreement with those ( both experimental and numerical ) reported in ref .
@xcite , where it was found that in a well mixed environment ( as in the urn model considered here ) two species get extinct . while this work has specifically focused on the stochastic cyclic lotka - volterra model , the addressed issues are generic .
indeed , we think that our results and technical approach , here illustrated by considering the case of a paradigmatic model , might actually shed further light on the role of fluctuations and the validity of the rate equations in a whole class of stochastic systems .
in fact , while one might believe that fluctuations in an urn model should always vanish in the thermodynamic limit , we have shown that this issue should be dealt with due care : this is true for systems where the rate equations predict the existence of an asymptotically stable fixed point , which is always reached by the stochastic dynamics @xcite .
in contrast , in systems where the deterministic ( rate equation ) description predicts the existence of ( neutrally stable ) center fixed points , such as the cyclic lotka - volterra model , fluctuations have dramatic consequences and hinder biodiversity by being responsible ( at long , yet finite , time ) for extinction of species @xcite . in this case , instead of a deterministic oscillatory behavior around the linearly ( neutrally ) stable fixed point , the stochastic dynamics always drives the system toward one of its absorbing states .
thus , the absorbing fixed points , predicted to be linearly unstable within the rate equation theory , actually turn out to be the _ only stable fixed points _
available at long time .
we would like to thank u. c. tuber and p. l. krapivsky for helpful discussions , as well as a. traulsen and j. c. claussen for having made manuscript @xcite available to us prior to publication .
m.m . gratefully acknowledges the support of the german alexander von humboldt foundation through the fellowship no .
iv - scz/1119205 stp .
in section [ stoch_appr ] , when considering the stochastic approach to the cyclic lotka - volterra model , for the sake of clarity we have specifically turned to the situation of equal reaction rates , @xmath59 . in this appendix , we want to provide some details on the general case with unequal rates . while the mathematical treatment becomes more involved , we will argue that the qualitative general situation still follows along the same lines as the ( simpler ) case that we have discussed in detail in sec .
[ stoch_appr ] .
the derivation of the fokker - planck equation ( [ gen - f - p ] ) is straightforward , following the lines of subsection [ master - f - p ] .
the matrix @xmath33 remains unchanged and is given in eq .
( [ a - lin ] ) , for @xmath75 we now obtain : @xmath193 the corresponding fokker - planck equation reads @xmath92+\frac{1}{2}\mathcal{b}_{ij}\partial_i\partial_jp({\bf x},t ) \quad .\ ] ] again , we aim at benefitting from the cyclic structure of the deterministic solutions , and perform a variable transformation to @xmath36 , with @xmath194 given in eq .
( [ s - matrix ] ) . as already found in subsection [ det - eq ]
, @xmath33 turns into @xmath195 , such that in the @xmath56 variables the deterministic solutions correspond to circles around the origin . for the stochastic part , entering via @xmath75 , a technical difficulty arises . it is transformed into @xmath196 being no longer proportional to the unit matrix as in the case of equal reaction rates .
we can do slightly better by using an additional rotation , @xmath197 , with rotation angle @xmath198 this variable transformation leaves @xmath199 invariant , but brings @xmath200 to diagonal form , with unequal diagonal elements .
the stochastic effects thus correspond to _
anisotropic diffusion_. however , for large system size @xmath3 , the effects of the anisotropy are washed out : the system s motion on the deterministic cycles , described by @xmath33 , occurs on a much faster timescale then the anisotropic diffusion , resulting in an averaging over the different directions . to calculate the time evolution of the average deviation from the reactive fixed point , @xmath201 , we start from the the fluctuations in @xmath56 , which fulfill the equations @xmath202 using eq .
( [ r - y ] ) , we obtain @xmath203 the dependence on the deterministic part has dropped out , and the solution to the above equation with initial condition @xmath204 is a linear increase in the rescaled time @xmath119 : @xmath205\frac{t}{n}\quad,\ ] ] valid around the reactive fixed point . as in the case of equal reaction rates , we have a linear dependence @xmath206 , corresponding to a two - dimensional random walk . as a conclusion of the above discussion , the general case of unequal reaction rates qualitatively reproduces the behavior of the before discussed simplest situation of equal rates ( confirmed by stochastic simulations ) .
the latter turns out to already provide a comprehensive understanding of the system . | cyclic dominance of species has been identified as a potential mechanism to maintain biodiversity , see e.g. b. kerr , m. a. riley , m. w. feldman and b. j. m. bohannan [ nature * 418 * , 171 ( 2002 ) ] and b. kirkup and m. a. riley [ nature * 428 * , 412 ( 2004 ) ] . through analytical methods supported by numerical simulations
, we address this issue by studying the properties of a paradigmatic non - spatial three - species stochastic system , namely the ` rock - paper - scissors ' or cyclic lotka - volterra model .
while the deterministic approach ( rate equations ) predicts the coexistence of the species resulting in regular ( yet neutrally stable ) oscillations of the population densities , we demonstrate that fluctuations arising in the system with a _ finite number of agents _ drastically alter this picture and are responsible for extinction : after long enough time , two of the three species die out . as main findings we provide analytic estimates and numerical computation of the extinction probability at a given time .
we also discuss the implications of our results for a broad class of competing population systems . |
since a photon has the same quantum numbers as vector mesons , a beam of high energy photons is a beam of quark - antiquark pairs .
their flavor is selected by the kind of vector meson which is detected : light quarks for @xmath0 or @xmath1 , strange quarks for @xmath2 and charmed quarks for @xmath3 mesons .
the life time of the fluctuation @xmath4 is given by the uncertainty principle and increases with the beam energy @xmath5 : for instance , a 5 gev beam of real photons fluctuates into a beam of @xmath0 mesons over 4 fm .
therefore , the interaction with matter of a beam of energetic photons is similar to the interaction of hadrons .
this is illustrated in fig .
[ vect_tot ] , which shows the variation , against the available c.m .
energy @xmath6 , of the cross section of the photoproduction of vector mesons @xcite , from threshold up to the hera energy range .
the rise with energy is due to the pomeron exchange ( @xmath7 ) , while the exchange of the @xmath8 regge trajectory ( @xmath9 ) is necessary to reproduce the trend of the cross section in the intermediate energy range of slac and fnal . at low energy , the exchange of the regge trajectories of the @xmath10 and @xmath11 mesons
account for the @xmath12 fall off of the cross section in the @xmath0 and @xmath13 channels respectively .
more details on the model are given in ref .
note that , due to the dominant @xmath14 nature of the @xmath2 meson , this channel is dominated by the exchange of the pomeron .
such a picture fails to reproduce the energy variation of the cross section of @xmath3 meson photoproduction .
the reason is that the large mass of the charmed quark sets the hard scale and prevents the exchanged gluons to reinteract and form a pomeron .
indeed , models based on the exchange of two gluons @xcite are able to relate the rapid rise of the cross section with energy to the evolution of the gluon distribution in the proton .
this finding was generalized to the description of exclusive vector meson electroproduction , at high virtuality @xmath15 , in terms of generalized parton distributions wich are discussed elsewhere @xcite in these proceedings . in this talk
, i will explore another direction : i will increase the mometum transfer @xmath16 to resolve the pomeron and reggeons , which are exchanged in exclusive photoproduction of vector mesons , into their simplest gluon or quark content .
@xmath2 meson photoproduction allows to prepare a @xmath17 pair of strange quarks and study its interaction with hadronic matter . at low momentum transfer @xmath18 ( small angle ) , its diffractive scattering is mediated by the exchange of the pomeron . at high momentum transfer ( large angle ) , the impact parameter is small and comparable to the gluon correlation length ( the distance over which a gluon propagates before hadronizing ) : the pomeron is resolved into its simplest component , two gluons which may couple to each of the quarks in the emitted vector meson or in the proton target .
this is illustrated in fig .
[ phi ] , which shows data recently recorded at desy @xcite and jlab @xcite . at low @xmath18
, the data confirm the shrinkage of the forward diffraction peak and the slow rise of the cross section with the energy , as expected from the exchange of the pomeron regge trajectory .
the two gluon exchange contribution matches the pomeron exchange contribution around @xmath19gev@xmath20 and reproduces the data at higher @xmath18 . in the jlab energy range , @xmath21-channel nucleon exchange
`` pollutes '' the highest @xmath18 bin : here the @xmath22 coupling constant @xmath23 is the same as in the analysis of the nucleon electromagnetic form factors @xcite .
more details may be found in ref .
@xcite .
such a finding is important as it tells us that , in the intermediate range of momentum transfer ( let s say @xmath24 gev @xmath20 ) , large angle exclusive meson production can be understood in a perturbative way at the level of effective parton degrees of freedom : dressed quark and gluon propagators , constituent quark wave functions of the nucleon and of the meson , etc . at low momentum transfer ( up to @xmath25 gev@xmath20 ) , the cross section is driven by their integral properties : any nucleon wave function which reproduces the nucleon form factor leads to the same result ; any gluon dressed propagator which reproduces the gluon correlation length ( @xmath26 fm ) leads to the same result . at higher momentum transfer , the cross section becomes more sensitive to the details of the wave function ( giving access the quark correlations ) and the shape of the gluon propagator .
i refer to the contribution of f. cano @xcite for a more detailed account .
suffice to say that the large momentum cross section is reduced when either a more realistic wave function ( than used in ref .
@xcite ) , or a gluon propagator with a running mass ( instead of a gaussian gluon propagator as in ref .
@xcite ) , is used . at jlab
, the experiment has been repeated at higher energy ( around e@xmath27 = 4.5 gev ) .
the @xmath21-channel backward peak is moved at higher values of @xmath28 , leaving more room to reveal and check the two gluon exchange contribution .
the preliminary results ( not shown ) confirm these predictions .
such studies provide us with a bridge with lattice qcd calculations .
the mass of the gluon , @xmath29 mev at vanishing virtuality , is close to the prediction of those calculations . in the near future
they may also provide us with an estimate of quark correlations in the wave functions of hadrons .
in contrast to the @xmath2 meson sector , quark interchanges are not forbidden in the @xmath0 and @xmath1 meson photoproduction sector .
[ rho_slac ] shows the latest data @xcite obtained at jlab . at low @xmath28 the good agreement
with the data is obtained adding , on top of the two gluon exchange amplitude , the exchange of the @xmath8 and @xmath10 regge trajectories , while the rise at high @xmath28 comes from the @xmath21-channel exchange of the @xmath30 and @xmath31 regge trajectories ( see ref .
@xcite ) . at intermediate @xmath28 ,
the two gluon exchange contribution misses the data by a factor three .
the agreement is restored when a saturating trajectory ( @xmath32 when @xmath33 gev@xmath20 ) is used instead of a linear trajectory , for mesons .
it leads to the asymptotic quark counting rules , which model - independently determine the energy dependency of the cross section at large -t . in a model
where the mesons are built from a potential , the linear confining potential leads to the linear long range part of the trajectory , which corresponds to the spectrum of the excited states of the meson , and the one gluon exchange short range potential leads to the saturating part of the trajectory @xcite . in that sense , the exchange of the linear part of the trajectory can be viewed as the exchange of two quarks strongly interacting in a non pertubative way , while the exchange of the saturating part can be viewed as the exchange of two quarks interacting by the exchange of a single perturbative hard gluon .
saturating trajectories provides us with an economic and effective way to incorporate hard scattering mechanisms @xcite .
such a model reproduces also the sparse data in the @xmath1 production channel ( fig .
[ omega_kroll ] ) . in this channel pion exchange dominates the cross section in the jlab energy range ( see fig .
[ vect_tot ] ) , and the pion saturating trajetory is the same as the one which reproduces the cross section of the @xmath34 reaction around @xmath35 @xcite .
the effect is more dramatic than in the @xmath0 channel , and i am eagerly waiting for the final analysis of the recent jlab data .
note that the dip at hight @xmath28 is due to the zero in the nucleon non degenerated regge trajectory @xcite : this is the only trajectory which can be exchanged , contrary to the @xmath0 channel .
such a model leads also to the @xmath36 asymptotic behavior of the @xmath37 cross section of the @xmath38 reaction depicted in fig .
[ rho_90 ] .
the two gluon exchange is too low , and the use of a massive gluon propagator ( @xmath39 mev ) , instead of a gaussian one , decreases further its contribution at large @xmath28 .
the bulk of the cross section comes from the exchange of the @xmath10 regge saturated trajectory .
at high energy , the real photon emitted in compton scattering has enough time to fluctuate into a vector meson .
therefore , the compton scattering cross section can be obtained by mutiplying the @xmath0 meson photoproduction cross section , which dominate over the @xmath1 and @xmath40 ones , by @xmath41 ( where @xmath42 is the radiative decay constant of the vector meson ) .
this leads to a good accounting of the sparse existing data : fig .
[ gam_90 ] shows the excitation function at @xmath37 , while fig .
[ gam_kroll ] shows the angular distribution at @xmath43 gev .
a more comprehensive set of data will soon come from jlab .
the threshold regime of charmonium production can provide a new window into multi - quark , gluonic , and hidden color correlations in hadronic wave functions .
in contrast to charm production at high energy , charm production near threshold requires all of the target s constituents to act coherently in the heavy quark production : only compact fock states with radius of the order of the compton wavelength of the charm quark can contribute to charm production at threshold . under these circumstances
, one can relate @xcite the energy dependency of the cross section to the behavior of its different gluonic components near @xmath44 : @xmath45 for two gluon exchange , @xmath46 for three gluon exchange .
[ charm ] shows that such an expectation is consistent with the sparse existing data and calls for a comprehensive dedicated measurement when the energy upgrade of jlab becomes a reality .
a consistent picture is emerging from the study of exclusive photoproduction of vector meson and compton scattering . at high momentum transfer ,
it relies on a partonic description of hard scattering mechanisms which provides us with a bridge with lattice gauge calculations .
the dressed gluon and quark propagators have already been estimated on lattice .
one may expect that correlated constituent quark wave functions ( at least their first moments ) will soon be available from lattice . to day , jlab is the only laboratory which allows to explore this regime , thanks to its high luminosity .
its current operation , at 46 gev , has already revealed a few jewels .
its energy uprade to 12 gev will permit a more comprehensive study of this field . among other topics
, this will allow to study the onset of asymptotic freedom , to determine the correlations between quarks and to single out the van der walls part of the force between hadrons . | at high energy , the photoproduction or electroproduction of vector mesons allow to prepare a beam of quark - antiquark pairs of a given flavor . at high momentum transfer ,
the study of the scattering of these pairs on a nucleon opens up an original window on the quark - gluon structure of hadronic matter , which may eventually shed light on its gluonic content , on correlations between quarks and on the van der walls part of the interaction between hadrons . |
metastable states and configurational entropy , are central concepts in our understanding of low temperature relaxation of disordered and glassy systems @xcite . in presence of time scale separation between `` fast '' and `` slow '' degrees of freedom relaxation
is often described as a walk between metastable states , intended as regions of configuration space where the fast degrees of freedom reach a state of quasi equilibrium before the relaxation of slow degrees of freedom can effectively take place . according to the adam and gibbs theory @xcite the configurational entropy , i.e. the logarithm of the multiplicity of metastable states is related to the relaxation time by an inverse proportionality relation .
the relation between metastability and relaxation is well understood in infinite range spin glasses with `` random first order transition '' @xcite .
the dynamics of these systems is exactly described by the equations of ideal mode coupling theory @xcite .
the mode coupling transition , spurious in liquid systems , is here driven by the existence of metastable states capable to trap the system for time scales diverging with the systems size .
the divergence is of course a mean field artifact , but features of mean - field metastable states could well have a reflex in finite dimensional glassy systems @xcite .
a well known problem in the study of metastable states is that while they are easy to define within mean field theory , their definition becomes elusive in finite dimensions .
this problem has been studied in large detail in the context of systems with first order phase transition , where the problem of metastability , can be considered as satisfactorily understood @xcite . unfortunately ,
this is not the case in glassy systems , where despite the appeal of a description of dynamics in terms of metastable states , only very rare contributions have tried to clarify the meaning of metastability beyond the phenomenological level @xcite .
the `` random first order transition scenario '' has led to phenomenological attempts to treat the problem of glassy relaxation , and ergodicity recovery in analogy with kinetics of ordinary first order phase transition @xcite .
liquids at low temperature would appear as `` mosaic states '' , in which locally frozen variables configurations could be composed combinatorially on a large scale .
the typical size of the rigid regions could be computed in a nucleation theory , with the configurational entropy playing the role of an ergodicity restoration bulk driving force , competing with some postulated interface tension .
these developments stimulated first principle calculations in microscopic disordered models . in ref .
@xcite and later in @xcite a statistical description of the ergodic state and a computation of the glassy coherent length below @xmath0 was proposed through the asymptotic analysis of a landau - like glassy effective free - energy derived from microscopic models .
unfortunately , in that papers it was not possible to make an explicit connection between the computed free - energy barrier and the relaxation time of the system .
scope of this letter is to discuss this connection . in order to do that ,
we start from an analysis of glassy relaxation based on separation of time scales .
we argue that finite dimensional relaxation is dominated by metastable states that can be characterized along the classical lines of lebowitz and penrose ( lp ) @xcite , first proposed to describe metastable phases of matter in presence of first order phase transitions . according to lp ,
metastable states can be considered as constrained equilibrium ensembles with : 1 ) an order parameter is homogeneous on a suitable mesoscopic length scale 2 ) a large time life and 3 ) a very small probability of return once abandoned . in analogy with the work of lp ,
we use in our analysis models with long - but - finite range kac kind of interactions , which in our case have a disordered character .
these offer the possibility of studying finite dimensional effects in an expansion around mean field , and the local mean - field character of correlations , postulated in the phenomenological mosaic description , appears as a consequence of the range of interaction @xcite .
let us consider a spherical kac p - spin glass model@xcite defined for real spins @xmath1 on the @xmath2-dimensional hypercubic lattice @xmath3 of linear size @xmath4 and hamiltonian @xmath5 where the couplings @xmath6 are i.i.d .
gaussian variables with zero average and variance @xmath7 where @xmath8 is an integer @xmath9 and @xmath10 , is a non - negative integrable function verifying the normalization @xmath11 . with this choice ,
the couplings @xmath6 are sensibly different from zero only if all pairs of variables @xmath12 @xmath13 , so that only variables that are at distances @xmath14\ ; { < \atop\sim}\ ; \gamma^{-1}$ ] effectively interact .
the effective interaction range @xmath15 will be assumed to be large throughout the paper .
we partition the lattice in boxes @xmath16 of a coarse graining length size @xmath17 and impose a local spherical constraint @xmath18 for all @xmath19 .
we are interested to the regime where the three defining lengths are supposed , as in lp , to verify the relation @xmath20 and for definiteness we will have in mind the situation where @xmath21 for some @xmath22 .
the model is chosen in such a way to reduce to the usual mean - field spherical @xmath8-spin model in the regime @xmath23 .
let us recall the physics of this case @xcite that will be useful in the finite @xmath24 case . on lowering the temperature from the paramagnetic region one encounters two transitions .
there is a first transition at the mode coupling temperature @xmath0 where ergodicity is broken .
below @xmath0 an exponential number of ergodic components @xmath25 dominates the thermodynamics .
the configurational entropy @xmath26 decreases as a function of the temperature , until it becomes zero at a `` kauzmann temperature '' @xmath27 , where a second transition is met .
this structure of states can be studied through the so - called `` quenched glassy potential '' that considers the free - energy @xmath28 of a system @xmath29 constrained to have a fixed global overlap @xmath30 with a typical reference equilibrium configuration @xmath31 @xcite .
this function has the shape depicted in figure [ pot ] , in the region @xmath32 one finds a two minimum structure analogous to the free - energy as a function of the order parameter for mean - field systems with a first - order phase transition .
the value of the overlap in the low minimum and high minima represent respectively the typical overlap among random configurations in different and the same ergodic components .
both minima are associated to the exponential multiplicity of ergodic components , and their difference gives directly the configurational entropy times the temperature @xmath33 .
we now address the question of restoration of ergodicity for finite @xmath24 and @xmath32 . the relaxation of model ( [ hp ] )
can be conveniently studied through langevin dynamics @xmath34 where @xmath35 are lagrange multipliers that enforce in average the local spherical constraint at each time and the @xmath36 are white noise variables with correlations @xmath37 .
an insight on the dynamics in the regime ( [ regime ] ) can be obtained studying the kac limit @xmath38 after the thermodynamic limit . through standard dynamical techniques
-martin - siggia - rose formalism @xcite or dynamical cavity method @xcite- , it is possible to conclude @xcite that the finite time local equilibrium spin - spin correlation function @xmath39 on the scale @xmath17 , tends to a space homogeneous function , @xmath40 , that for temperatures larger then the kauzmann temperature @xmath41 verifies the usual equilibrium mean field equation for the spherical p - spin model @xcite @xmath42 this equation predicts ergodicity breaking below @xmath0 .
the equilibrium correlation function in this regime approaches exponentially fast on a scale @xmath43 a non zero limit @xmath44 .
conversely , for any finite @xmath45 , for finite or infinite @xmath4 , the system should be able to escape from the metastable states and @xmath46 coincide with the equilibrium correlation function @xmath47 . in the regime ( [ regime ] ) , the behavior of @xmath48 can be schematically depicted as in figure [ fig1 ] . as a function of time ,
the small @xmath24 correlation function remains close to the mean field curve up to some characteristic time @xmath49 diverging when @xmath50 and it decays to zero on longer scales .
since for small enough @xmath24 this characteristic time is arbitrary larger than the mean field relaxation time @xmath43 , one finds that the system equilibrates in the region of phase space where the local overlaps with the initial condition are close to @xmath51 before relaxing further .
moreover , one can expect that on time scales of the order of @xmath43 , the system has homogeneous correlations on the scale of the box size @xmath17 , with typical fluctuation from cell to cell of @xmath52 scaling as @xmath53 . in turn , assuming independence of the fluctuations beyond some finite length , using extreme statistics one can estimate the maximum deviation @xmath54 $ ] to be at most of order @xmath55 .
this , according to our hypothesis , can be made arbitrarily small .
the previously discussed properties are enough to characterize the regions explored on scales smaller then @xmath49 as metastable states in the lebowitz - penrose sense .
the role of the `` order parameter '' , homogeneous in space , is played by the overlap of the configuration of the system with the initial condition @xmath56 .
formally , for any typical equilibrium configuration @xmath31 , we can define metastable states , as equilibrium ensembles where the configurations @xmath29 are such that in each box @xmath16 the overlap with the initial configuration @xmath57 is suitably close to @xmath51 .
more precisely , one defines an interval @xmath58 $ ] , with @xmath59 and the set @xmath60 .
the metastable state is the restricted ensemble @xmath61 as in the lp case , the definition of metastable state makes sense for a whole range of values of @xmath62 , such that a local fluctuation of the overlap larger then @xmath62 is with high probability pushed back to @xmath51 .
however , in order for the definition to be useful , the value of @xmath62 should be fixed a posteriori in such a way that a fluctuation larger then that size should be amplified , and eventually run the system out of the metastable state with finite probability .
we can now use this analysis to estimate the relaxation rate of our system as a decay rate of the metastable state .
we can then proceeds as in lp , and argue that for the local langevin dynamics we are using , the relaxation rate @xmath63 is proportional to @xmath64 where @xmath65 is the set of configuration @xmath29 belonging to @xmath66 , such that @xmath67 for at least one point @xmath68 .
@xmath69 is in principle a random variable that depends on the initial configuration @xmath31 and on the quenched random couplings @xmath70 .
however the relaxation rate should be self - averaging with respect to both sources of noise , and its typical value can be computed considering the average of its log @xmath71 where @xmath72 stands for the canonical average over @xmath31 and the quenched average over the distribution of the disorder @xmath73 .
( [ quattro ] ) is the main result of this paper .
it connects the relaxation rate to a free - energy barrier computable in principle in purely static terms .
for fixed disorder and given finite volume we can define an `` effective potential '' as the free - energy cost to keep the overlap with an equilibrium configuration @xmath31 in each of the boxes @xmath16 fixed to specified values @xmath74 : @xmath75)= \sum_\sigma \prod_x \delta(q_x(\sigma,\sigma^*)-q_x ) e^{-\beta h(\sigma , j)}\end{aligned}\ ] ] where in the expression , we have anticipated the scaling of the free - energy with the interaction range @xmath15 @xcite .
is is clear how @xmath76 and @xmath77 can be related to @xmath78 $ ] , in fact , for @xmath79 , with transparent notation @xmath80}. \label{sei}\end{aligned}\ ] ] by the previous discussion , the dominant profile in @xmath77 is estimated as the constant @xmath81 for all @xmath19 . for this profile ,
the extensive part of @xmath82 $ ] is a self - averaging quantity : @xmath83\approx { e}(w_{\sigma^*,j}[q_x = q_{ea}])$ ] . for generic overlap profiles
@xmath74 , we can consider @xmath84= { e}(w_{\sigma^*,j}[q_x])$ ] , and ask if the knowledge of this quantity is enough to reconstruct the relaxation rate @xmath85 for small @xmath24 . the restricted partition function @xmath86 is : @xmath87}.\ ] ] each term of the sum on @xmath68 will be dominated by a single profile , minimizing @xmath78 $ ] in presence of the constraints . these will be profiles with a localized excitation around @xmath68 , corresponding to a finite free - energy cost with respect to the uniform state @xmath81 .
for fixed @xmath31 and @xmath73 , different values of @xmath68 could give rise to different free - energy cost and different profile shapes .
however , we could argue again that on the mesoscopic scale @xmath15 almost all sites @xmath68 would give the contributions of the same order and these dominate the relaxation rate .
this is a stronger hypothesis than the other ones underlying our reasoning .
it implies that the incipient spatial heterogeneities on a scale @xmath15 responsible for relaxation , have equal probability of appearing in any point of space , independently of the reference configuration @xmath31 and the realization of the quenched disorder @xmath73 . without that , the computation of @xmath88 should start from an evaluation of ( [ quattro ] ) .
this is in principle possible and when done could allow the validation or refutation of the hypothesis .
previous work @xcite was implicitly based on that strong hypotheses .
within that framework , if we denote by @xmath89 the profile maximizing @xmath90 $ ] for @xmath91 , the relaxation rate will be given by @xmath92-v[q_x = q_{ea}]\right)}\ ] ] and @xmath93 $ ] will provide a landau free - energy functional that we can use in a nucleation - like theory .
it is clear that restricted to profiles constant in space , @xmath94 , for small @xmath24 we have : @xmath95\approx(l\gamma)^d \phi(q)$ ] , where @xmath96 is the mean - field potential . in order to compute the free - energy barrier within this self - averaging setting
, one should consider spherically symmetric solutions with boundary conditions @xmath97 .
formulated in this way the problem reduces to the computation of a critical droplet as suggested by phenomenological theories @xcite . according to conventional nucleation theory @xcite , close to @xmath41 where the two minima are close to degeneracy
, one can use the `` thin wall approximation '' and evaluate the interface tension through the 1d instantonic solutions connecting the two minima .
this competes with a bulk driving term given by the free - energy difference between the minima , which in our case is the mean - field configurational entropy .
this first principle formulation confirms the prediction of phenomenological theories , that the configurational entropy provides the bulk driving force to ergodicity restoration .
the actual computation of @xmath98 for fixed profile can be performed through a modified version of the replica method @xcite , needed to average over the spin reference configuration @xmath31 and the quenched disorder @xmath73 .
this has been done for the model ( [ hp],[wij ] ) in ref .
@xcite . without entering in the details of the discussion of that paper we just mention that within a specific replica ansatz it is possible to find a finite _ surface _ tension is found for @xmath99 approaching @xmath41 .
this implies a glassy coherent length behaving as @xmath100 and an adam - gibbs free energy barrier @xmath101 .
this result is supported by a more elaborated ansatz in @xcite . as discussed in @xcite , and in view of possible non self - averaging effects , these solution could in fact just provide approximations to the true barrier .
the `` right '' replica solutions could change the exponent @xmath102 appearing in the relation . however , the general adam - gibbs structure of the formula , that only depends on the fact that the bulk ergodicity restoring force is the configurational entropy , should remain unchanged .
summarizing , we have presented an argument that identifies as lebowitz - penrose metastable states the dominant configurations of relaxing glassy systems below the mode coupling transition .
this allows to compute the relaxation time in purely static term through the evaluation of the probability of equilibrium fluctuations .
the relation between relaxation time and free - energy barrier is summarized in eq .
( [ quattro ] ) which is the main result of this letter .
we have also discussed how previous results in the literature relate implicitly to a self - averaging hypothesis .
if this holds , the relevant fluctuations have the shape of overlap droplets , and the bulk driving force for ergodicity restoration is provided by the configurational entropy . in agreement with phenomenological theories the relation of the relaxation time with the configurational
entropy follows a generalized adam - gibbs equation .
the argument has been presented in the case of disordered spin model , where playing with the interaction range one can have time scales as much separated as wanted and a mesoscopic scale for which correlations are homogeneous on short time scales . on the other hand
our argument is actually only based on the hypothesis that time separation is such to allow partial equilibration of the fast degrees of freedom for fixed values of the slow one , and as such it should be generically valid when this condition is respected .
p. g. wolynes , jour .
nist , * 102 * 187 ( 1997 ) , x. xia , p. g. wolynes , proc .
sci . * 97 * , 2990 ( 2000 ) , x. xia , p. g. wolynes , phys .
lett * 86 * , 5526 ( 2001 ) , v. lubchenko and p. g. wolynes , j. chem .
119 * 9088 ( 2003 ) g. parisi , `` gauge theories , spin glasses and real glasses '' talk presented at the oskar klein centennial symposium , cond - mat/9411115 , proceedings of the xiv sitges conference `` complex behaviour of glassy systems '' 10 - 14 june 1996 .
wolrd scientific , singapore 1996 . | in this note we discuss metastability in a long - but - finite range disordered model for the glass transition .
we show that relaxation is dominated by configuration belonging to metastable states and associate an in principle computable free - energy barrier to the equilibrium relaxation time .
adam - gibbs - like relaxation times appear naturally in this approach . |
in the last few years , our understanding of the reionization history of the universe has greatly improved thanks to the detection of quasars at increasingly high redshift ( e.g. , becker et al .
2001 ; fan et al .
2002 ; hu et al .
2002 ; white et al .
2003 ) and the new data on the cosmic microwave background ( cmb ) radiation from wmap ( bennet et al .
observations from high redshift quasars indicate a fast increase of the intensity of the ionizing background with cosmic time occurring at @xmath4 , probably signalling the end of reionization ( fan et al .
2002 ) , or the epoch when the low - density intergalactic medium ( hereafter , igm ) became fully ionized .
reionization is expected to have occurred over an extended period of time , during which sources gradually reionized every atom in the igm ( e.g. , gnedin 2000 ) .
indeed , the detection of ly@xmath5 emission from sources at @xmath6 ( hu et al . 2002 ; kodaira et al . 2003 ; cuby et al . 2003 ; kneib et al . 2004 ; rhoads et al .
2004 ) that are not highly luminous to have produced large regions around them implies that the igm could not be fully neutral at the redshift of these sources , otherwise such photons would have been scattered out of the line of sight ( miralda - escud & rees 1998 ; madau & rees 2000 ) .
another important probe to the epoch of reionization is the optical depth to electron scattering of cmb photons , @xmath7 .
the first measurement of @xmath7 was reported by the _ wilkinson microwave anisotropy probe _ mission ; although its value is still rather uncertain ( kogut et al . 2003 ; spergel et al .
2003 ) , the measurement favored an early start to reionization , such that the fractional ionization of the igm would have reached 50% at @xmath8 .
it is worth noting that this does not by itself contradict the appearance of the gunn - peterson trough ( which marks the end of reionization ) at @xmath4 , because reionization may advance gradually over a long period of time .
however , an early start of reionization presents a problem in the cold dark matter model of structure formation , in which only a small fraction of matter has collapsed into halos that can form stars at @xmath9 , and therefore one needs to assume a very high rate of emission of ionizing photons from this first generation of stars ( e.g. , haiman & holder 2003 ; chiu , fan , & ostriker 2003 ; onken & miralda - escud 2004 ) .
one possibility is that the first collapsed halos in which gas was able to cool were efficient in driving gas to accrete onto central black holes that were formed by the core collapse of the first massive stars .
however , as argued by dijkstra et al .
( 2004a ) , any such population of high-@xmath10 active galactic nuclei ( agn ) would likely contribute a smooth component to the present x - ray background at energies near @xmath11 kev that may be higher than is allowed by observational constraints ( although significant calibration uncertainties of the x - ray background intensity remain which could increase the upper limit on the fraction of the x - ray background in a smooth component ) .
the other possibility is that the first massive stars , which were metal - free and highly efficient at producing ionizing photons , were responsible for an early reionization ( oh et al .
2001 , yoshida et al .
2003 , sokasian et al .
2004 ) .
much work has been done to characterize the properties of these metal - free stars ( abel et al .
1998 , 2002 ; bromm , coppi , & larson 1999 , 2002 ; oh et al . 2001 ) .
the formation of the first stars was governed by molecular hydrogen radiative cooling , which becomes effective at gas temperatures above @xmath12 k ( yoshida et al .
this virial temperature is first reached in halos of total mass @xmath13 m@xmath14 , which start becoming substantially abundant at @xmath15 .
metal - free stars have higher effective surface temperatures than their metal - rich counterparts during their main - sequence phase . in addition , they are nearly fully convective and can burn most of their hydrogen content during their main - sequence lifetime .
this makes them highly efficient as producers of ionizing radiation ( emitting @xmath16 ionizing photons per baryon ; schaerer 2002 ) , most of which can escape into the igm due to the small amount of gas in the host halos . despite these advantages of metal - free stars as sources of ionizing radiation , the idea that these stars may have been responsible for an early reionization faces a number of difficulties .
first , it is unlikely that many massive stars might form simultaneously in the first halos where star formation took place , which have low velocity dispersion and can therefore easily lose their gas after it is heated by photoionization and supernova explosions .
numerical simulations suggest that one central massive star will initially form , with a mass of @xmath17 m@xmath14 ( e.g. , abel et al . 2002 ; bromm et al .
2002 ) , containing a fraction of only @xmath18 of the baryonic mass of the halo from which it formed .
the ionization and the supernova explosion resulting from this star can then eject the rest of the gas from the halo ( bromm , yoshida , & hernquist 2003 ; whalen , abel , & norman 2004 , kitayama et al .
2004 ) , yielding a very low efficiency to form stars . later ,
when the gas falls back into a halo with a total mass increased by the merging process , it will already be enriched by metals and will form the next generation of stars with properties that are presumably similar to present stellar populations in normal galaxies .
if the metal - free star releases about @xmath19 of metals , a total mass of @xmath20 of baryons may be polluted to a metallicity of @xmath21 , above which cooling is already modified by the presence of metals , while the emission properties of massive stars are also modified for even lower metallicities . in cases where a metal - free star collapses into a black hole releasing no metals ,
any cold gas reaching the halo center at a later time will also not form another metal - free star because the black hole will be in the center .
the second difficulty is that the radiation emitted by these first stars can act as a negative feedback mechanism to slow down the formation of other metal - free stars in new collapsed halos with pristine gas .
a widely studied effect is photodissociation of molecular hydrogen ( haiman , abel , & rees 2000 ) .
another feedback effect is the heating of the igm by x - rays emitted by supernovae or by the metal - free main - sequence stars themselves ( oh & haiman 2003 ) , which raises the entropy of gas accreting into halos , resulting in a more stringent requirement for the gas to cool .
this effect has not been so widely studied as the molecular photodissociation one , but as we shall see below it may be even more important . the aim of this paper is to determine the maximum possible contribution to reionization from zero metallicity stars , under the assumption that only one such star forms in each halo , with a fixed stellar mass .
special attention will be paid to the way that the radiative feedback mechanisms of photodissociation and x - ray heating limit the formation rate of these metal - free stars .
we shall use a semianalytic model based on merger trees ( similar to the approach of , e.g. , haiman et al . 2000 and wise & abel 2005 ) , including a new ingredient to compute the halo age distribution and derive the fraction of halos that contain a metal - free main - sequence star at any given time .
the paper is organized as follows . in
[ smethod ] , we present our method to calculate the formation rate of metal - free stars , the photodissociating and x - ray background they create , and how these radiation backgrounds affect the formation rate by altering the entropy profile in collapsed halos and the molecular hydrogen abundance .
the results are presented in
[ res ] , showing how each of the two feedback effects we consider impact the rate at which metal - free stars can form and their contribution to reionization .
a discussion of these results follows in [ sdisc ] .
our central assumption in this paper is that when a halo that contains metal - free gas ( i.e. , a halo for which no star has previously formed in any of its merger - tree progenitors ) collapses , a single massive star forms at the halo center after a cooling time .
the mass of this star is assumed to be constant for all halos with metal - free gas . as explained in detail later ( [ sstform ] ) , the star forms in our model if the cooling time of the gas at the center of the halo is shorter than the lifetime of the halo .
we develop an algorithm to define the lifetime of halos from the merger tree framework based on the extended press - schechter model , which is fully described in the appendix .
the cooling time depends on the central gas density and temperature , which depends at the same time on the initial entropy of the gas prior to the collapse of the halo ( e.g. , oh & haiman 2003 ) .
we compute the initial central gas density ( prior to the onset of cooling ) in halos at each redshift self - consistently , assuming hydrostatic equilibrium in a halo with the navarro , frenk , & white ( 1997 ) density profile for the dark matter ( hereafter , nfw ) .
we calculate the ionizing radiation emitted by the metal - free stars to infer the history of reionization ( taking into account an escape fraction of ionizing photons from halos determined by recombinations in the halo gas ) , as well as the feedback effects of x - ray heating ( and consequent entropy raise of the atomic medium , which affects the gas density profiles in halos ) and molecular photodissociation .
we also include a component of ionizing radiation from a normal stellar population assumed to form in metal - enriched halos , which have had previous star formation and ionization .
this section explains in detail how all these ingredients are implemented in our model .
our goal will be to find the maximum contribution that these metal - free stars could plausibly make to an early reionization , and how this is limited by the minimal expected feedback effects .
we use the halo mass function of sheth & tormen ( 2002 ) .
we assume that a first - generation star forms when the halo age exceeds the cooling time at the halo center .
we adopt for this purpose the definition of halo formation time of lacey & cole ( 1993 ) , which is the time when the halo mass was half the present one .
the probability distribution of the age of a halo as a function of its mass and redshift is calculated following lacey & cole , which we then use to compute the probability that a halo contains a metal - free star as described in the appendix . to determine the cooling time at the halo center , we assume the gas is in hydrostatic equilibrium in the halo gravitational potential , and that the dark matter follows a nfw density profile with concentration parameter @xmath22 . in reality
, @xmath23 depends on the halo mass and redshift ( bullock et al .
2001 ; eke et al .
2001 ) , but here we approximate it as constant and with the same value used by oh & haiman ( 2003 ) . under the assumption of hydrostatic equilibrium , the gas density profile depends on its temperature .
shock - heating of the gas during the collapse heats the gas to a roughly constant temperature , about equal to the halo virial temperature .
this results in a final gas density profile similar to the dark matter , although with a small , constant density core , with a size that is determined by the minimum entropy @xmath24 of the gas in the halo .
the gas with the lowest entropy is generally the one that is shock - heated to the virial temperature only after reaching a high density , and ends up near the halo center .
however , if the gas had already been heated prior to the collapse , acquiring a high initial entropy , the value of @xmath24 may be increased , resulting in a much more extended gas profile and reduced central gas density , and therefore a much longer central cooling time . to take this into account
we follow the procedure explained in oh & haiman ( 2003 ) .
we define the function @xmath25 as @xmath26 where @xmath27 is the adiabatic index ( equal to @xmath28 for monatomic gas ) , and @xmath29 and @xmath30 the gas temperature and number density .
note that the entropy of a gas is actually the logarithm of @xmath25 , although in this paper we refer to the function @xmath25 as `` entropy '' for convenience .
the entropy before the collapse is what we call the entropy floor , @xmath31 , which we calculate as a function of redshift as described in
[ xr ] . as an example
, we show in figure [ fprof ] the gas density profile in a halo with @xmath32 m@xmath14 formed at different redshifts .
the initial entropy is fixed to the primordial value given in oh & haiman ( 2003 ) , determined by early compton heating by the cmb on the adiabatically cooling gas , with the residual ionized fraction left over from recombination . at @xmath33 ,
the entropy of the igm is practically constant and the entropy floor decreases in importance with decreasing redshift because of the decreasing gas density , hence the gas density profile becomes steeper at low redshift .
we show also the effect of raising the igm temperature to @xmath34 k at @xmath35 , which leads to a large reduction in the central gas density .
we define the cooling timescale as the ratio of the internal energy of the gas to the cooling rate per unit of volume , @xmath36 : @xmath37 where @xmath38 is the boltzmann constant .
we include only molecular hydrogen cooling for metal - free gas . in practice ,
atomic cooling can be neglected because metal - free stars form very rarely in halos with virial temperature above @xmath39 k in our model .
the reason is that these halos usually have some progenitor of lower virial temperature that has already formed a star previously ( this will be seen in fig . 2 ) .
in this case , we consider that the gas is already enriched with metals and that the emissivity per unit mass is constant , as described in [ semis_met_rich ] . the cooling rate ( due to collisional excitation of molecular hydrogen roto - vibrational lines ) is @xmath40 , where @xmath41 is the molecular fraction .
we use the cooling function @xmath42 of galli & palla ( 1998 ) . the molecular fraction @xmath41 is determined by the continuous formation of molecules in the gas from @xmath43 , which is at the same time made by the primordial concentration of electrons from the recombination epoch ( see tegmark et al .
we use equation ( 16 ) of tegmark et al . , although simplified and modified to consider h@xmath44 dissociation in the following way : @xmath45 where @xmath46 is the dissociation rate ( which we discuss later in [ sdiss ] ) and @xmath47 is the electron fraction , given by equation ( [ xr : fe ] ) .
the constant @xmath48 determines the formation rate of @xmath43 ( which determines the formation rate of molecules ) , and is given by @xmath49 @xmath50 s@xmath51 .
the initial , primordial value of @xmath41 is taken as @xmath52 ( galli & palla 1998 ) . note that the fraction of hydrogen that forms molecules from @xmath43 within collapsed halos is much larger than this primordial value , and therefore the long - term evolution of the molecular hydrogen fraction does not strongly depend on its initial value . as mentioned earlier ,
the main assumption in this work is that only one metal - free star is formed in each halo that collapses from progenitors that have never formed a star previously .
the mass of the star , @xmath53 , is much lower than the baryonic mass of the halo , and is in fact lower than the gas mass in the core of the gas distribution .
consequently , we can evaluate the cooling time at the halo center and consider that the star will form when the age of the halo is equal to this cooling time .
the star forms in a halo of mass @xmath54 with formation time @xmath55 at a time @xmath56 that obeys @xmath57 the star will then live up to a time @xmath58 where @xmath59 is the main - sequence lifetime of a metal - free star with mass @xmath60 .
the probability that a halo of mass @xmath54 at time @xmath61 has already formed a star can be evaluated by computing first numerically the formation time @xmath62 that obeys equation ( [ cool_c1 ] ) with @xmath63 , and then calculating the probability @xmath64 that the halo formation time is earlier than @xmath62 ( see eq . [ 2.26 ] of lacey & cole 1993 ) .
similarly , the probability that the halo contains a metal - free star on the main - sequence at time @xmath61 is evaluated as the probability @xmath65 , where @xmath66 is the solution of @xmath55 of equation ( [ cool_c2 ] ) when @xmath67 .
this condition , however , would be valid only if the halo gas were always metal - free . in reality , if a star had already formed in one of the halo progenitors , the supernova explosion from that star would already have polluted the gas with metals and blown it away . later on , as the sequence of halo mergers continues , the final halo present at time @xmath61 will accrete this gas , but since the gas is metal enriched it will form metal - rich stars . the emission from metal - rich stars will be treated differently in our model , as we describe later in
[ semis_met_rich ] . here
, our main assumption is that the very massive stars we consider that are highly efficient at emitting ionizing radiation are formed only when the gas is totally free of heavy elements . to take this into account
, we evaluate at each time @xmath61 the average number of metal - free stars that have formed in a halo of mass @xmath54 , as ( see the appendix ) @xmath68 where @xmath69 is the number of halos of mass between @xmath70 and @xmath71 at time @xmath72 that have been incorporated into a halo of mass @xmath54 at time @xmath61 ( see eq . [ 2.25 ] in lacey & cole 1993 ) , and @xmath73 is the probability per unit time that a star forms at @xmath72 in a halo of mass @xmath70 . the way the probability @xmath74 is calculated is described in the appendix .
we then assume that there is a poisson distribution of the number of stars that have formed within any halo , so that the probability that none has formed is @xmath75 .
therefore , the probability that a halo of mass @xmath54 harbours a main - sequence star at time @xmath61 is @xmath76 where @xmath77 is the differential probability that a halo at time @xmath61 was formed at time @xmath55 and is given by equation ( 2.19 ) in lacey & cole ( 1994 ) . in practice , the lifetime of the main - sequence star ( @xmath78 years ) is short enough that the above integral can be approximated as @xmath79 figure [ fnstarm ] shows the probability that a halo with mass @xmath54 harbours a metal - free main - sequence star .
different plots correspond to different redshifts ( @xmath80 , 15 , 10 , and 6 ) , and different lines to different models that will be described later .
the model a1 corresponds to no feedback effects on the formation of metal - free stars , and the other models incorporate feedback effects that are described in 2.5 .
the probability to contain a star tends to peak at halo masses of several times @xmath81 . at lower masses
the molecular cooling rate is too slow and stars have not yet formed , while in halos of higher mass a metal - free star has typically already formed in one of their lower - mass progenitors ( i.e. , the second term in eq.is very small ) . as a byproduct
, we can also calculate the global formation rate of metal - free stars ( see the appendix ) as @xmath82}{{{\rm d}}t}\ , .
\label{dndt}\ ] ] where @xmath83 is the metal - free star formation rate and @xmath84 is the sheth - tormen halo mass function .
this last formula is useful to calculate the preheating due to x - rays ( see section [ xr ] ) .
we consider three feedback processes that affect metal - free star formation .
the first is halo photoevaporation due to the ionizing background ( section [ sevap ] ) .
the second , photodissociation of molecules due to a soft ultraviolet background ( section [ sdiss ] ) .
finally , the third is reheating of the igm caused by x - rays ( section [ xr ] ) .
the ionizing background reheats the gas in halos and causes a large increase in the entropy floor .
gas trapped in minihalos escapes ( e.g. , shapiro et al .
2004 ) , quenching star formation . cooling and star formation in reionized regions can only occur in halos that are massive enough for the gas to undergo dissipation by atomic cooling .
we neglect any possible formation of metal - free stars in these halos . in practice ,
once these halos collapse they should almost always have a progenitor that already formed a metal - free star in the past , because the number of halo progenitors increases with halo mass ; moreover , even if one metal - free star may form in one of these more massive halos cooling mainly by atomic processes , their contribution to the total number of metal - free stars formed should be negligible because the abundance of high - mass halos is greatly reduced compared to halos of low mass .
we therefore neglect any possible formation of metal - free stars in reionized regions . to take this into account , equation ( [ prob_sh_star ] )
is multiplied by a factor @xmath85 , equal to the fraction of the igm that is neutral at every redshift .
the calculation of the ionized fraction of the igm as a function of time using the global rate of star formation that we compute is described below in 2.6 . actually , the factor @xmath86 represents a lower limit to the suppression of the star formation by reionization , because in practice halos are correlated , or biased relative to the mass , so that more halos should be present in the reionized regions where other high - mass halos have already formed .
is easily photodissociated by soft ultra - violet photons ( haiman et al .
the rate at which dissociation is produced can be approximated by ( abel et al .
1997 ) @xmath87 being @xmath88 the flux at frequency @xmath89 in units of @xmath90 and @xmath91 the self - shielding factor given by draine & bertoldi ( 1996 ) .
the mean free path of dissociating photons in the igm is typically larger than the mean distance between minihalos .
therefore , this flux can be approximated as homogeneous and isotropic . to calculate this background
we apply equation ( 7 ) of haiman et al .
the emissivity of dissociating photons ( @xmath92 in their formula ) is computed similarly to that of ionizing ones ( see [ semis_met_free ] ) , although we take into account only the emission by the metal - free stars .
emission from the metal - rich stars has a small effect at the high redshifts of interest , which we have neglected .
as proposed by oh & haiman ( 2003 ) , soft x - ray photons raise the entropy floor , greatly reducing cooling and star formation in low - mass halos .
x - rays can have a long mean free path through the atomic igm and can therefore heat this medium in an approximately uniform way , whereas the more abundant ultraviolet ionizing photons heat the medium only near the boundaries of hii regions .
the mean free path in the atomic igm is roughly @xmath93 when this mean free path is in the range that goes from the mean distance between neighboring halos containing first - generation stars ( which is about @xmath94 kpc at @xmath95 ; see yoshida et al .
2003 ) to the horizon radius ( @xmath96 mpc at @xmath95 ) , the photons are effective at heating the atomic igm .
this corresponds to a range of frequencies from 200 ev to 2 kev , or the soft x - ray band .
soft x - ray photons can be produced by metal - free stars during two phases .
first , these stars have a high enough effective temperature during the main sequence lifetime to emit substantially in soft x - rays ( schaerer 2002 ) .
second , x - rays can be emitted by the hot gas in a supernova remnant after the death of the star .
for simplicity we consider that all the emission is produced at the end of the metal - free star lifetime ( a good approximation because of the short lifetime ) .
therefore , the rate of increase of the entropy floor is @xmath97 where @xmath98 is the rate of energy increase per particle , @xmath99 and @xmath100 is the rate of sne per unit of volume , equal to the star formation rate @xmath101 ( eq . [ [ dndt ] ] ) at the time @xmath102 .
the constant @xmath103 is the total energy per star that is used to reheat the medium .
note that this is not the same as the total energy radiated in x - rays by the star since some of that energy goes into collisional ionizations or emission of recombination photons .
we leave the value of @xmath103 as a free parameter in our model ( the range of possible values is discussed in [ sres_xr ] ) . even though the reheating by x - rays is roughly homogeneous , the increase of the entropy is of course lower in regions of high gas density . to compute the entropy floor for a specific halo with formation at @xmath55 , we follow the evolution of the gas density during the collapse assuming the spherical top - hat model , and we compute the entropy increase from the previous history of the intensity of the x - ray background generated by metal - free stars . the evolution of @xmath100 is computed self - consistently with the model from the star formation rate .
heating by x - rays is no longer taken into account after virialization ( this would be negligible in any case once the gas has reached a high density ) .
x - rays have yet another effect : owing to secondary ionizations , they can produce a higher ionized fraction than the residual value after recombination ( which is @xmath104 , see galli & palla 1998 ) .
this increases the formation rate of molecular hydrogen , enhancing the cooling rate ( haiman , rees , & loeb 1996 ) . according to shull & van steenberg ( 1985 ) , around 30 % of the total energy in x - rays
is used in ionizations when the neutral fraction is high .
the rate of ionizations is then found by dividing expression ( [ e : reh ] ) by 13.6 ev ( the ionization energy of hydrogen ) . considering this effect ,
the evolution of @xmath105 is determined by @xmath106 where @xmath107 is the recombination rate , which can be approximated by @xmath108 s@xmath51 ( hutchins 1976 ) and @xmath30 is the gas number density in the halo center after collapse . before collapse , the first term is negligible because the density is very low .
the initial value of @xmath105 is given by the residual ionized fraction after recombination .
to summarize , for each halo mass at a fixed time , stars can form only when the formation time satisfies equation ( [ cool_c1 ] ) . when considering feedback effects , cooling is delayed and therefore halo formation has to take place earlier .
consequently , the probability that a halo hosts a main sequence star in equation ( [ prob_sh_star ] ) decreases , reflecting the feedback effects . to calculate the reionization history of the igm we must first calculate the emissivity @xmath109 of ionizing photons per igm baryon .
we consider two contributions to the emissivity , one from metal - free stars which we discuss in [ semis_met_free ] , and another from stars with metals , discussed in [ semis_met_rich ] .
finally , we explain the procedure to calculate the reionization history in [ semis_met_free ] .
metal - free stars emit a higher quantity of ionizing photons than enriched stars ( see schaerer ( 2002 ) ) .
we use here the fitting formula in table 6 of schaerer ( 2002 ) to compute the ionizing luminosity of each star as a function of its mass , @xmath110 . for the values of @xmath53
we consider here , this yields @xmath111 s@xmath51 and @xmath112 s@xmath51 .
in addition , we take into account that some of the stellar photons may be absorbed locally in the halo gas surrounding the star , and only a fraction @xmath113 of these photons will be able to escape and ionize the igm .
we use the following simple model to compute @xmath113 ( see whalen et al .
2004 for a more complete discussion and several examples of the evolution of the regions formed around a metal - free star ) : we assume that the star is initially surrounded by a distribution of atomic gas following a singular isothermal profile .
an region grows around the star , and the heated gas acquires a sound speed of @xmath114 and starts expanding .
this speed is typically larger than the halo circular velocity , but still smaller than the speed of the ionization front as long as the escape fraction is not very small and the star is capable of ionizing the halo ( the stellar lifetime is @xmath115 3 million years , which is short compared to the crossing time of the halo at this sound speed .
hence , the ionization front advances supersonically .
we use the simple approximation where all the gas moves out by the same distance @xmath116 .
the gas distribution then has a central hole of radius @xmath117 , and at @xmath118 the density is @xmath119\frac{[r - r{_{\rm c}}(t)]^2}{r^2},\ ] ] where @xmath120 is the initial gas number density profile of the halo .
the total rate of recombinations is computed from this density profile assuming the halo gas has been fully ionized by the star , and is subtracted from the photon emission rate of the star .
the escape fraction of ionizing photons as a function of halo mass ( @xmath121 ) is plotted for different redshifts in figure [ ffesc ] .
we consider two possible masses of metal - free stars , @xmath34 m@xmath14 ( left panel ) and @xmath122 m@xmath14 ( right panel ) . naturally , the escape fraction increases with stellar mass and decreases with halo mass , because the number of recombinations in a halo with the gas fully ionized increases with halo mass and does not depend on the luminosity of the ionizing star .
these results agree with the numerical simulations of kitayama et al .
( 2004 ) , suggesting that our simple model is able to capture the main physics of the problem .
the escape fraction increases when reheating is considered , being more significant in low mass halos , which have their density profiles flattened .
note that there is a halo mass at which the escape fraction falls to zero ; for halos above this mass , the single central star is not able to ionize all the halo gas and the region reaches a strmgren radius ( the assumption that all the halo gas is ionized is then obviously incorrect , but this affects the calculation of @xmath113 only when @xmath113 is already very small in any case ) . the global emissivity from metal - free stars , @xmath123 , is finally calculated by integrating over all halo masses the number of photons emitted per igm particle corrected by absorption , i.e. , @xmath124 where @xmath125 is the baryonic number density of the igm at @xmath61 , and @xmath126 is the probability that a halo harbors a main sequence metal - free star ( see eq .
[ prob_sh_star ] ) . for the emissivity of dissociating photons
, we also use equation ( [ eq : emis : mf ] ) with @xmath127 .
the emission of dissociating photons from metal - free stars , @xmath128 , is also computed from the fit given in table 6 in schaerer ( 2002 ) which yields @xmath129 s@xmath51 and @xmath130 s@xmath51 .
after the formation of the first metal - free star in any halo , the surrounding gas is polluted with the metals released by this star .
the metal - enriched gas can recombine and fall back into the halo .
we assume the metallicity of this infalling gas is then sufficiently high to cool by the usual mechanisms that are prevalent in present galaxies ( bromm et al .
however , halos with metal - enriched gas below a critical virial temperature may still not form stars because the cooling rates are too small ( e.g. , dijkstra et al .
2004b ) , and gas may remain close to @xmath131 k in a spherical or thick disk distribution where the gas is unable to cool further and fragment .
in this paper , we adopt the simple assumption that in halos were metal - rich star formation takes place , the emissivity per unit baryonic mass is constant at all redshifts .
then , the global emissivity due to enriched stars , @xmath132 , is calibrated to the observations of the ionizing background intensity and mean free path of ionizing photons at redshift @xmath0 ( see onken & miralda - escud 2004 ) : @xmath133 where @xmath134 is the emissivity at redshift 4 in units of ionising photons per particle and per hubble time , @xmath135 is the fraction of mass in halos where metal - rich star formation takes place , and @xmath136 is the same fraction of mass at @xmath0 .
we assume that halos with circular velocity above @xmath137 km / s are the ones forming stars at @xmath0 ( as in onken & miralda - escud 2004 ; this affects only the normalization of the emissivity ) , but consider different possibilities for the emissivity at higher redshift , as explained below .
in order to study the effect of varying the emission from normal , metal - rich galaxies , we consider several models listed in table [ tcases ] .
these models are labeled by a number , and in general any model will be referred to by the letter specifying the feedback processes assumed ( see table 1 ) and this number . in models 1 and 2 ,
we compute @xmath135 as the fraction of mass in metal - enriched halos with a circular velocity greater than @xmath138 km / s , and in models 3 to 6 we assume that all halos that have been polluted by metals form enriched stars , with no minimum halo circular velocity .
note that even by assuming a minimum halo circular velocity that is lower than at the calibration redshift @xmath0 , we are maximizing the emission at high redshift under the constraint that the emission per unit mass is not greater than in halos at @xmath0 .
the value of the normalization constant @xmath134 is also subject to uncertainties arising from the process by which the intensity of the ionizing background is inferred from modelling the ly@xmath5 forest transmitted flux , and the measurement of the photon mean free path from the abundance of lyman limit systems .
we adopt the values @xmath139 for models 1 to 4 ( as in miralda - escud 2003 ) , and twice this value , @xmath140 ( which would be favored by the measurement of the ionizing background intensity by bolton et al .
2005 ) , in models 5 and 6 . finally ,
the emissivity from the metal - free stars is varied by considering all stars to have a mass @xmath141 in models with odd number , and @xmath142 in models with even number .
the evolution of the ionized fraction @xmath143 is calculated according to @xmath144 where @xmath145 is the total emissivity of ionizing photons , and @xmath146 is the recombination rate which we evaluate at @xmath147 k. the cosmic history of @xmath143 then yields the electron scattering optical depth @xmath148 .
in this section , we present the results on metal - free star formation and the reionization history of the igm obtained from the approach given in [ smethod ] . in this model
, there are four free parameters : the stellar mass , @xmath53 , the reheating of igm per metal - free star , @xmath103 , and the emissivity of metal - rich stars , @xmath134 .
in addition , there is some uncertainty in the physical processes governing feedback effects .
for this reason , we play with the possibility to switch on or off the effect of photodissociating photons in order to check how important it is . finally , to compute @xmath135 in equation ( [ emisp12 ] ) we may consider or not halos with circular velocity @xmath149 km / s ( minihalos ) to see whether the value of @xmath7 experiences a big change .
the models including all the different possibilities are labelled xy , whith x running from a to d and y running from 1 to 6 ( see tables [ tcases ] and [ tsub_cases ] ) .
we now present the results on the rate of metal - free star formation and the reionization history of the igm for the models described in 2 .
the models and their nomenclature are described in tables 1 and 2 .
the model number ( 1 to 6 ) refers to properties of the radiation sources : the mass of metal - free stars , and the normalization and low halo mass cutoff for the emission from metal - enriched stars .
the model letter ( a to d ) refers to the presence or absence of the molecular photodissociation and x - ray heating feedback effects .
results are first presented for the case of no feedback effects ( models a , [ sresultsnofeed ] ) , then including molecular photodissociation ( models b , [ sresultsj21 ] ) , and finally including x - ray photoheating ( models c and d , [ sres_xr ] ) .
then , the reionization histories are reviewed ( [ sresreion ] ) .
all models are run with the following set of cosmological parameters : @xmath150 , @xmath151 , @xmath152 , @xmath153 , @xmath154 and @xmath155 .
cccc 1 & 7 & no & 100 + 2 & 7 & no & 300 + 3 & 7 & yes & 100 + 4 & 7 & yes & 300 + 5 & 14 & yes & 100 + 6 & 14 & yes & 300 + [ tcases ] ccc a & no & 0 + b & yes & 0 + c & yes & 10@xmath156 + d & yes & 10@xmath157 + [ tsub_cases ] the evolution of the emissivity for model a1 is shown in figure [ femis ] , both for metal - free and metal - rich stars .
metal - free stars dominate at high redshift , and enriched stars start dominating below @xmath158 .
this behaviour is easily understood .
initially there are only metal - free stars and their contribution grows with time as more halos form .
this results in an increasing fraction of halos being polluted with metals , which at some point causes a decline in the rate of new metal - free stars that are formed .
the halos collapsing with enriched gas start forming enriched stars which eventually dominate the emissivity .
the formation of metal - free stars is also reduced as redshift decreases due to the increased fraction of the igm that is ionized . in ionized regions ,
the gas is less able to cool and low - mass halos would continue merging into more massive ones before forming any stars , implying a great reduction in the number of metal - free stars that can be formed .
this effect is less important because the igm ionized fraction remains relatively low at the epoch where most metal - free stars form .
comparing models 1 ( with metal - free stellar mass of @xmath159 ) with models 2 ( @xmath142 ) , figure 4 shows that the emissivity of metal - free stars increases by a factor of @xmath160 as expected at @xmath161 , but at lower redshift it increases by a smaller factor because the fraction of ionized igm increases faster for @xmath162 , reducing the formation of metal free stars to a greater extent than for models 1 .
we note here that a high bias in the distribution of metal - free stars relative to the mass could further increase the effect of ionization in suppressing the number of metal - free stars that can be formed .
we now examine how the inclusion of a photodissociating background , computed self - consistently from the star formation history , alters the results .
figure 5 shows the evolution of the photodissociating background intensity in models b1 and b2 .
the maximum is reached around @xmath8 as the effects of increased metal pollution and ionized igm fraction suppress new metal - free star formation ( note that emission from enriched stars is not included in this figure ) .
figure 4 shows that the maximum of formation rate of metal - free stars occurs at higher redshift when the photodissociating feedback is included , because of the additional suppression of star formation .
we also see in figure 2 how the formation of metal - free stars is shifted to higher mass halos in models a compared to models b at @xmath163 , because the negative feedback delays the formation of the stars to a later stage at which halos have grown to a higher mass .
we consider two models for the soft x - ray energy produced by a metal - free star that is converted into igm heat , @xmath164 ( table 2 , models c and d ) . during the main - sequence , and at effective surface temperatures near @xmath165 k , the fraction of radiative energy emitted above @xmath166 is @xmath115 0.1% of the total energy output of the star , or @xmath167 erg .
the supernova explosion at the end of the life of a metal - free star can produce as much as @xmath168 erg of kinetic energy for some stellar masses ( see , e.g. , bromm et al .
2003 ) , and the fraction of these kinetic energy that is eventually reprocessed into soft x - rays may be @xmath169 ( oh & haiman 2003 ) , of which a fraction of 50% may be converted into igm heat when the soft x - rays are absorbed in a mostly neutral medium ( shull & van steenberg 1985 ) . given the substantial degree of uncertainty in the soft x - ray energy emitted per star , we consider the two values in models c and d bracketing our estimate .
the increase in igm temperature caused by the absorption of soft x - rays is shown in figure 6 for models c1 and d1 .
the onset of formation of metal - free stars causes a rise in temperature of the neutral medium starting at @xmath170 , and at lower redshift the temperature ceases to increase when the formation of metal - free stars is reduced ( in practice there would probably be an important contribution to the x - ray emission from the enriched population , which has been considered here ) .
the establishment of an entropy floor causes a large reduction of the central density in low mass halos ( see fig . 1 ) , which greatly reduces the cooling rate .
this implies that metal - free stars tend to be formed in models c and d in much more massive halos than in models a and b : the formation of these stars is suppressed in low - mass halos by the feedback effect , so metal - free stars are not formed until later when a more massive halo has formed from the mergers of several small ones ( see figure 2 ) . as long as our assumption that only one metal - free star forms in each halo continues to hold
, the metal - free stars will be much less abundant .
this is seen clearly in figure 4 , which shows the large reduction in the emissivity of metal - free stars caused by the soft x - ray heating .
cccccc 1 & & 0.084 & 0.077 & 0.066 & 0.057 + 2 & & 0.116 & 0.103 & 0.087 & 0.068 + 3 & & 0.094 & 0.089 & 0.079 & 0.066 + 4 & & 0.121 & 0.110 & 0.096 & 0.076 + 5 & & 0.105 & 0.102 & 0.094 & 0.082 + 6 & & 0.127 & 0.119 & 0.108 & 0.090 + finally , we examine the reionization history and electron scattering optical depth to the cmb of electron scattering in all the models . in figure [ fionf ]
we see that , as expected , feedback delays the ionization of the igm .
reionization always occurs gradually over a relatively wide redshift range .
naturally , models with the highest stellar mass and emission from enriched stars achieve a higher ionized fraction at every redshift , and hence a higher optical depth . even though the total emissivity history may be double - peaked owing to the presence of the metal - free and the enriched populations of stars ( see fig . 4 )
, there is never a double reionization because the recombination rate is slow enough that the ionized fraction continues to increase when the emission from metal - free stars starts decreasing .
the redshift at which reionization is completed stays in the range of @xmath171 .
this prediction of the redshift of the end of reionization is a natural consequence of the value of the emissivity assumed for the enriched stars that is derived from observations of ly@xmath5 absorption systems at @xmath172 , and of assuming that the clumping factor of ionized gas is not much larger than unity ( miralda - escud 2003 ) .
the exact redshift of the end of reionization predicted by each model is not an independent test of our models , because this redshift can be modified and adjusted to agree with observations of the gunn - peterson trough in the highest redshift quasars ( white et al .
2003 ) by introducing a moderate clumping factor with an adequate dependence on redshift ( note that we have assumed a clumping factor equal to one ) . the values of the electron scattering optical depth for each model are presented in table 3 .
the presence of metal - free stars has an important effect in creating a small ionized fraction at very high redshifts ( @xmath173 ) .
however , the increase of the optical depth that metal - free stars can cause is modest and , even for @xmath174 and in the absence of any negative feedback effects , they do not increase the optical depth beyond @xmath175 . models d1 , d3 and d5 represent cases where the contribution from metal - free stars is minimal ( due to the very strong x - ray heating assumed ; see fig . 4 )
. comparing them with models a2 , a4 and a6 where the emission from metal - free stars is maximum , we see that the early ionization caused by metal - free stars can cause only a relatively modest increase of the optical depth , under our central assumption that only one metal - free star is made per halo , and no metal - free stars are made in halos that have already merged with a halo in which star had previously formed .
we have considered in this paper the maximum emissivity and the contribution to reionization that metal - free stars could make .
our central assumption is that only one metal - free star is made in every halo where pristine gas is able to cool via molecular hydrogen rotovibrational lines .
this is a reasonable assumption in view of the simulations that have been made of the formation of the first star and the effects of the subsequent ionizing radiation emitted and supernova explosion : the gas cools hydrostatically towards the center , avoiding fragmentation and forming a central star which can then ionize and heat all the gas in its host halo , and probably expel it in the supernova explosion ( e.g. , abel et al . 2002 ; bromm et al . 2002 , 2003 ) .
once a star has exploded , it pollutes its host halo as well as every larger halo into which its host halo merges in the future .
any halo that has already formed a metal - free star in any of its halo progenitors in the past is assumed to form a metal - enriched population of stars that has an emissivity of ionizing photons similar to the emissivity that is observationally determined at @xmath176 . under this basic assumption
, the contribution of the metal - free stars to reionization is relatively modest .
metal - free stars may dominate the emissivity at very high redshifts ( @xmath177 ) , but before they can manage to ionize much of the universe , the metal pollution of halos rapidly reduces their formation rate
. moreover , the effects of the ionization itself prevent cooling in the remaining low - mass halos with pristine gas , delaying the formation of a central metal - free star to the time of formation of more massive halos that collapse later , and thereby reducing the number of metal - free stars that are formed per unit of mass in the universe .
reionization was likely completed by a population of stars and quasars in galaxies forming from pre - enriched gas .
we have analyzed the effects of molecular photodissociation and of increasing the entropy floor by x - ray heating .
these can further reduce the amount of ionizing radiation that is emitted by the population of metal - free stars . in agreement with oh &
haiman ( 2003 ) , we have found that the negative feedback induced by x - ray heating can in principle be very large .
however , as pointed out by kuhlen & madau ( 2005 ) , this negative feedback might be much less important than suggested by the simple calculation of oh & haiman ( 2003 ) and the one we have presented here . in our calculation , we consider only the cooling time of the gas at the halo center in the hydrostatic equilibrium configuration obtained at the end , by assuming an increased initial entropy produced by the x - ray heating .
in reality , the gas should be able to gradually cool and lose entropy as it collapses in dense regions of halos , and these halos later merge to form the final one where a star is formed . the same x - rays that heat the gas would also increase the ionization , allowing faster molecular cooling as the gas density increases , which might greatly reduce the x - ray negative feedback effect .
nevertheless , our main conclusion is that even without feedback effects the emission from metal - free stars is strongly limited by the metal - pollution and the ionization itself .
our models are generally in agreement with previous work .
the value of @xmath148 we infer is always lower than @xmath3 , even when we increase the emissivity of the enriched stellar population and we minimize any negative feedback effects on the metal - free population . for a smaller contribution from metal - free stars , values of @xmath148
are closer to @xmath178 ( see table [ ttaue ] ) , in agreement with the results of ricotti & ostriker ( 2004 ) .
our results are also compatible with those of sokasian et al .
( 2004 ) , who find that high cmb optical depths to electron scattering can be achieved only when many stars are assumed to be formed within halos that cool by molecular hydrogen .
the model m9 of sokasian et al . is the only one that assumes that only one star is formed per halo , and in this case their results for the maximum optical depth are in good agreement with ours .
similarly , wise & abel ( 2005 ) can also produce high optical depths only by assuming that many metal - free stars form in each halo with a total mass much larger than in our model .
naturally , if one is willing to assume a mass for metal - free stars even larger than @xmath179 , their contribution to the cmb optical depth , @xmath148 , can be further increased , but their total emission increases by a factor smaller than the increase in the mass because of the enhanced negative feedbacks . at the same time , if miniquasars emitting copious amounts of ionizing radiation where produced by the black holes created by the metal - free stars ( which could occur when a halo containing a central clump of cooling gas merges with another halo that has already formed a star and contains a black hole in its center ) , much more ionizing radiation could be produced , although the x - rays emitted by these mini - quasars might then be in conflict with limits on any unresolved component of the present soft x - ray background ( dijkstra et al .
2004a ) .
our work can also be used to estimate the contribution of metal - free stars to the cosmic infrared background , which might include the ultraviolet light of these stars from high redshift .
the total number of ionizing photons that are ever emitted by metal - free stars can be computed straightforwardly by integrating the curves shown in figure 4 .
the results are shown in table [ temis ] for each one of our models . even in the absence of negative feedbacks , and for stellar masses
@xmath180 , metal - free stars do not emit more than @xmath115 one photon per baryon in the universe .
this conclusion is not surprising , since in the absence of many recombinations one does not need a large number of ionizing photons to be emitted per baryon to complete the reionization . because of the very high effective temperatures of metal - free stars ( @xmath181 k ) , the number of photons emitted in ultraviolet light at wavelengths longer than the lyman limit is similar .
hence , the contribution of metal - free stars to the present cosmic infrared background is not more than @xmath115 one photon per baryon under the assumptions we have made .
recently , kashlinsky et al .
( 2005 ) detected brightness fluctuations in the cosmic infrared background after subtracting all contributions from known galaxies .
the measured fluctuations are at a level of @xmath182 .
this background intensity at a wavelength of @xmath183 corresponds to @xmath184 photons per baryon in the universe ; as discussed by kashlinsky et al . , in order to account for the observed fluctuations the absolute brightness of the infrared background contributed by metal - free stars would have to be as high as @xmath185 ( or @xmath115 5000 photons emitted per baryon ) , after taking into account the expected high bias in the spatial distribution of these stars .
it is clear that very extreme assumptions about the number of metal - free stars that were formed need to be made if the unaccounted fluctuations in the cosmic infrared background are related in any way to these metal - free stars . as discussed by santos et al .
( 2002 ) , if metal - free stars are to make an important contribution to the cosmic infrared background , one needs to assume that a large fraction of all the gas in each halo where pristine gas cools by molecular hydrogen ( typically more than @xmath81 ) forms metal - free stars , and that most of the radiation from the stars is somehow internally absorbed in the halos in order to prevent an excessively early reionization of the igm .
both of these requirements are not realistic : once a central star is formed in a @xmath186 halo , its ionizing radiation will ionize and push all the halo gas out over the short main - sequence lifetime of the star , and the escape fraction of the ionizing photons is high except in rare cases where a single star forms in a very massive halo ( see figure [ ffesc ] ) .
the detected cosmic infrared background fluctuations have other more likely possible sources , such as an incomplete accounting of the faint - end of the galaxy luminosity function , or a normal population of low - luminosity galaxies with metal - enriched gas and stars at high redshift .
jm acknowledges helpful conversations with mark kuhnen , piero madau , and daniel wang .
jm would also like to thank the institute for advanced study for their hospitality , where part of this work was completed .
this work was supported in part by the direccin general de investigacin cientfica y tcnica of spain , under contract aya2003 - 07468-c03 - 01 .
jmr was supported by a fellowship of the ministerio de educacin , cultura y deporte of spain .
abel , t. , anninos , p. , zhang , y. , & norman , m. l. 1997 , new astronomy , 2 , 181 abel , t. , bryan , g. l. , & norman , m. l. 2002 , science , 295 , 93 becker , r. h. , et al .
2001 , , 122 , 2850 bennett , c. l. , et al .
2003 , , 148 , 1 bolton , j. s. , haehnelt , m. g. , viel , m. , & springel , v. 2005 , , 357 , 1178 bromm , v. , coppi , p. s. , & larson , r. b. 2002 , , 564 , 23 bromm , v. , ferrara , a. , coppi , p. s. , & larson , r. b. 2001 , , 328 , 969 bromm , v. , yoshida , n. , & hernquist , l. 2003 , , 596 , l135 bullock , j. s. , kolatt , t. s. , sigad , y. , somerville , r. s. , kravtsov , a. v. , klypin , a. a. , primack , j. r. , & dekel , a. 2001 , , 321 , 559 chiu , w. a. , fan , x. , & ostriker , j. p. 2003
, , 599 , 759 cuby , j .-
, le f ` evre , o. , mccracken , h. , cuillandre , j .- c . , magnier , e. , & meneux , b. 2003 , , 405 , l19 dijkstra , m. , haiman , z. , & loeb , a. 2004a , , 613 , 646 dijkstra , m. , haiman , z. , rees , m. j. , & weinberg , d. h. 2004b , , 601 , 666 draine , b. t. & bertoldi , f. 1996 , , 468 , 269 eke , v. r. , navarro , j. f. , & steinmetz , m. 2001 , , 554 , 114 fan , x. , narayanan , v. k. , strauss , m. a. , white , r. l. , becker , r. h. , pentericci , l. , & rix , h. 2002 , , 123 , 1247 galli , d. & palla , f. 1998 , , 335 , 403 gnedin , n. y. 2000 , , 535 , 530 haiman , z. , abel , t. , & rees , m. j. 2000 , , 534 , 11 haiman , z. , & holder , g. p. 2003
, , 595 , 1 haiman , z. , rees , m. j. , & loeb , a. 1996 , , 467 , 522 hu , e. m. , cowie , l. l. , mcmahon , r. g. , capak , p. , iwamuro , f. , kneib , j .-
p . , maihara , t. , & motohara , k. 2002 , , 568 , l75 hutchins , j. b. 1976 , , 205 , 103 kashlinsky , a. , arendt , r. g. , mather , j. , & moseley , s. h. 2005 , , 438 , 45 kitayama , t. , yoshida , n. , susa , h. , & umemura , m. 2004 , , 613 , 631 kneib , j. , ellis , r. s. , santos , m. r. , & richard , j. 2004 , , 607 , 697 kodaira , k. , et al .
2003 , , 55 , l17 kogut , a. , et al .
2003 , , 148 , 161 kuhlen , m. , & madau , p. 2005
, , 860 lacey , c. & cole , s. 1993 , , 262 , 627 lacey , c. & cole , s. 1993 , , 271 , 676 madau , p. , & rees , m. j. 2000 , , 542 , l69 miralda - escud ' e , j. 2003 , , 597 , 66 miralda - escude , j. , & rees , m. j. 1998 , , 497 , 21 navarro , j. f. , frenk , c. s. , & white , s. d. m. 1997 , , 490 , 493 oh , s. p. & haiman , z. 2003 , , 346 , 456 onken , c. a. & miralda - escud ' e ,
j. 2004 , , 610 , 1 oh , s. p. , nollett , k. m. , madau , p. , & wasserburg , g. j. 2001 , apj , 562 , l10 raig , a. , gonzlez - casado , g. , & salvador - sol , e. 2001 , , 327 , 939 rhoads , j. e. , et al .
2004 , , 611 , 59 ricotti , m. , & ostriker , j. p. 2004
, , 350 , 539 santos , m. r. , bromm , v. , & kamionkowski , m. 2002 , , 336 , 1082 schaerer , d. 2002 , , 382 , 28 shapiro , p. r. , iliev , i. t. , & raga , a. c. 2004 , , 348 , 753 sheth , r. k. & tormen , g. 2002 , , 329 , 61 shull , j. m. , & van steenberg , m. e. 1985 , , 298 , 268 sokasian , a. , yoshida , n. , abel , t. , hernquist , l. , & springel , v. 2004 , , 350 , 47 spergel , d. n. , et al . 2003 , , 148 , 175 tegmark , m. , silk , j. , rees , m. j. , blanchard , a. , abel , t. , & palla , f. 1997 , , 474 , 1 whalen , d. , abel , t. , & norman , m. l. 2004 , apj , 610 , 14 white , r. l. , becker , r. h. , fan , x. , & strauss , m. a. 2003 , , 126 , 1 wise , j. h. , & abel , t. 2005 , , 629 , 615 yoshida , n. , abel , t. , hernquist , l. , & sugiyama , n. 2003 , , 592 , 645
consider a halo of mass @xmath54 at time @xmath61 .
the average number of smaller halos with mass between @xmath70 and @xmath71 at @xmath187 that are located within the halo of mass @xmath54 at @xmath61 is given by the extended ps formalism , @xmath188 where @xmath189 is the conditional probability for a mass element to be part of a halo of mass @xmath70 at @xmath72 , given that it is part of a larger halo of mass @xmath190 at a later time @xmath191 ( eq . [ 2.16 ] in lacey & cole 1994 ) .
the rate at which stars form at @xmath72 in halos of mass between @xmath70 and @xmath71 , which will be found within the halo of mass @xmath54 at @xmath61 , is @xmath192 where @xmath193 is the probability per unit of time that a halo with mass @xmath70 forms a star at @xmath72 .
this latter probability is obtained by noting that the probability for a star to form in a halo with mass @xmath70 between @xmath72 and @xmath194 is equal to the probability for the corresponding halo to form between @xmath195 and @xmath196 , being @xmath197 the required cooling time in this halo to form a star at @xmath72 , times the probability @xmath198 that no halo ancestor has previoulsy formed any star , @xmath199 in equation ( [ i ] ) , @xmath200 is the probability distribution function of halo formation times @xmath55 , and the derivative @xmath201 is to be evaluated by requiring the condition @xmath202 .
although a better behaved definition of halo formation time with known analytical expressions of the corresponding probability distribution function can be found in the literature ( raig et al .
2001 ) , for simplicity we use here that provided by lacey & cole ( 1993 ) .
these authors define the formation time of a halo with mass @xmath54 at @xmath61 as the earliest time @xmath203 at which some progenitor reaches a mass @xmath204 . at the same time
, the average number of stars that have previously formed in ancestors of the halo with mass @xmath54 at @xmath61 , @xmath205 , is obtained by integration of equation ( [ nstar ] ) , @xmath206\ , .
\label{nave}\ ] ] assuming a poisson distribution of the number of metal - free stars formed in ancestors of a halo of mass @xmath54 at @xmath61 , with mean expectation value provided by equation ( [ nave ] ) , the probability that no ancestor has previously formed any metal - free star is @xmath207 notice that in order to calculate the previous probability we must proceed in an iterative way since the term @xmath208 in equation ( [ nave ] ) includes the factor @xmath198 affecting its ancestors , implying that we need to save at each step the information of all previous halos along the merger tree .
cccccc 1 & & 0.46 & 0.33 & 0.18 & 0.07 + 2 & & 0.80 & 0.61 & 0.40 & 0.19 + 3 & & 0.39 & 0.29 & 0.16 & 0.06 + 4 & & 0.73 & 0.56 & 0.38 & 0.18 + 5 & & 0.31 & 0.24 & 0.14 & 0.06 + 6 & & 0.62 & 0.49 & 0.34 & 0.16 + | we estimate the maximum contribution to reionization from the first generation of massive stars , with zero metallicity , under the assumption that one of these stars forms with a fixed mass in every collapsed halo in which metal - free gas is able to cool .
we assume that any halo that has already had stars previously formed in one of their halo progenitors will form only stars with metals , which are assigned an emissivity of ionizing radiation equal to that determined at @xmath0 from the measured intensity of the ionizing background .
we examine the impact of molecular hydrogen photodissociation ( which tends to reduce cooling when a photodissociating background is produced by the first stars ) and x - ray photoheating ( which heats the atomic medium , raising the entropy of the gas before it collapses into halos ) .
we find that in the cdm@xmath1 model supported by present observations , and even assuming no negative feedbacks for the formation of metal - free stars , a reionized mass fraction of 50% is not reached until redshift @xmath2 .
the combination of ordinary metal - enriched stars and early metal - free stars can yield a cmb optical depth to electron scattering of at most @xmath3 .
the contribution of metal - free stars to the present cosmic infrared background is negligibly small . |
we propose a seesaw model built upon the hypercharge zero fermionic weak quintuplet which , in isolation , provides a viable dark matter ( dm ) particle within the so - called minimal dark matter ( mdm ) model @xcite .
we explore the conditions under which the quintuplets @xmath10 could simultaneously generate the masses of the known neutrinos and provide a stable dm candidate .
the proposed model goes beyond the three tree - level realizations of weinberg s effective dimension - five operator @xmath11 @xcite : type i @xcite , type ii @xcite , and type iii @xcite seesaw mechanisms , mediated by a heavy fermion singlet , a scalar triplet , and a fermion triplet , respectively .
a seesaw mediator of isospin larger than one has to be accompanied by a scalar multiplet larger than that of the standard model ( sm ) higgs doublet .
our previous model @xcite based on an exotic nonzero hypercharge weak quintuplet dirac fermion @xmath12 generates the neutrino masses both at the tree level and at the loop level . at the tree level it corresponds to dimension - nine seesaw operator which reproduces the empirical neutrino masses @xmath2 ev with @xmath13 few 100 gev states testable at the lhc , while for heavier masses , which are outside of the lhc reach , these exotic leptons generate radiative neutrino masses
. however , if one expects from new states to accomplish simultaneously the dm mission , they should have zero hypercharge . a recent `` r@xmath14mdm model '' @xcite of purely radiative neutrino ( @xmath15 ) masses employs hypercharge zero quintuplets @xmath10 which , taken in isolation , provide a viable mdm candidate @xcite . in a separate paper @xcite
we identify an omitted term in @xcite which spoils the claimed dm stability .
therefore , we adopt here a more general approach in which we analyze in detail both the tree - level and the radiative neutrino masses generated by hypercharge zero quintuplets @xmath10 .
thereby we address also the phenomenology of these states at the lhc and recall that by imposing a @xmath9 symmetry , one can reconcile the seesaw mission with the viability as a dm .
as announced above , the model we propose here is based on the symmetry of the sm gauge group @xmath16 .
in addition to usual sm fermions , we introduce three generations of hypercharge zero quintuplets @xmath17 , transforming as @xmath18 under the gauge group .
also , in addition to the sm higgs doublet @xmath19 there is a scalar quadruplet @xmath20 transforming as @xmath21 .
this is in contrast to a recent model for radiative neutrino masses @xcite adopted subsequently in @xcite , where a use of the scalar sextuplet @xmath22 avoids the tree - level contribution . in a tensor notation suitable to cope with higher @xmath23 multiplets @xcite our additional fields are totally symmetric tensors @xmath24 and @xmath25 with the following components : @xmath26 @xmath27 the gauge invariant and renormalizable lagrangian involving these new fields reads @xmath28 here , @xmath29 is the gauge covariant derivative , @xmath30 is the yukawa - coupling matrix and @xmath31 is the mass matrix of the heavy leptons , which we choose to be real and diagonal .
for simplicity we drop the flavor indices altogether .
the scalar potential has the gauge invariant form @xmath32 in the adopted tensor notation the terms in eqs .
( [ lagrangian ] ) and ( [ scalarpot ] ) read @xmath33 accordingly , the majorana mass term for the quintuplet @xmath34 is expanded in component fields to give @xmath35 the terms containing two charged dirac fermions and one neutral majorana fermion @xmath36 the electroweak symmetry breaking proceeds in usual way from the vacuum expectation value ( vev ) @xmath37 of the higgs doublet , corresponding to the negative sign in front of the @xmath38 term in eq .
( [ scalarpot ] ) . on the other hand
, the electroweak @xmath39 parameter dictates a small value for vev @xmath40 of the scalar quadruplet , implying the positive sign in front of the @xmath41 term .
however , the presence of the @xmath42 term in eq .
( [ scalarpot ] ) , given explicitly by @xmath43 leads to the induced vev for the @xmath44 field , @xmath45 since it changes the electroweak @xmath39 parameter from the unit value to @xmath46 , a comparison to the experimental value @xmath47 @xcite gives us a constraint @xmath48 gev .
the yukawa interaction terms from eq .
( [ lagrangian ] ) , when expressed like that in eq .
( [ tensor ] ) , read explicitly as @xmath49 the vev @xmath40 generates a dirac mass term connecting @xmath50 and @xmath51 , a nondiagonal entry in the mass matrix for neutral leptons given by @xmath52 a similar term connecting light and heavy charged leptons gives the mass matrix for charged leptons @xmath53 after diagonalizing the mass matrix for the neutral leptons , the light neutrinos acquire the majorana mass matrix given by @xmath54 in the basis where the matrix of heavy leptons is real and diagonal , @xmath55 , we can write @xmath56 as @xmath57 together with the induced vev in eq .
( [ phivev ] ) , this gives @xmath58 which reflects the fact that the light neutrino mass is generated from the dimension - nine operator corresponding to the tree - level seesaw mechanism displayed on fig .
[ dim9op ] . ) .
the fermion line flow indicates a majorana nature of the seesaw mediator . ] besides at the tree level , the light neutrino masses arise also through one - loop diagrams displayed on fig .
[ oneloop ] . the crucial quartic @xmath59 term in eq .
( [ scalarpot ] ) , when expanded in component fields , gives @xmath60 there are two different contributions on fig .
[ oneloop ] , one with heavy neutral fields ( @xmath61 ) and the other with heavy charged fields ( @xmath62 ) running in the loop .
if we neglect the mass splitting within the @xmath63 and @xmath64 multiplets , the contribution to the light neutrino mass matrix is given by @xmath65 \,.\ ] ] this expression can be further specified @xcite , depending on relative values of the masses for additional fermion and scalar multiplets . for heavy fermions significantly heavier than the new scalars , @xmath66 , @xmath67 \,.\ ] ] in the opposite case , for @xmath68 , @xmath69 finally , if @xmath70 , then @xmath71 the tree - level and the loop contributions added together give the light neutrino mass matrix @xmath72 \ , .
\nonumber\\\end{aligned}\ ] ] in the case of comparable masses the light neutrino mass matrix reads @xmath73 \sum_k { y_{ik } y_{jk } \over m_k } \,.\ ] ] from this expression we can estimate the high energy scale of our model and the corresponding value for @xmath40 in eq .
( [ phivev ] ) . for illustrative purposes
we take the same values for the mass parameters ( @xmath74 ) and the empirical input values @xmath75 gev and @xmath76 ev . for the moderate values
@xmath77 and @xmath78 we get @xmath79 gev and @xmath80 mev . corresponding masses of the new states could be tested at the lhc . on fig .
[ tree_dom - ce ] we display the part of the parameter space for which the tree - level ( loop - level ) contribution dominates . .
right from the equality - division line the one - loop contribution to neutrino masses dominates , and we plot @xmath81 . ]
the production channels of the heavy quintuplet leptons in proton - proton collisions are dominated by the quark - antiquark annihilation via neutral and charged gauge bosons @xmath82 where the gauge lagrangian relevant for the production is given by @xmath83 the cross section for the partonic process is @xmath84 where @xmath85 is the mandelstam variable @xmath86 for the quark - antiquark system , the parameter @xmath87 denotes the heavy lepton velocity , and the left- and right - handed couplings are given by @xmath88 here , @xmath89 and @xmath90 are the sm chiral quark couplings to the @xmath91 boson .
the vector couplings of heavy leptons to gauge bosons are @xmath92 where @xmath93 can be read of the last raw in eq .
( [ gauge - short ] ) , relevant for the production of @xmath94 and @xmath95 pairs , respectively . in evaluating the cross sections for a hadron collider
, the partonic cross section ( [ eq : xspartonic ] ) has to be convoluted with the appropriate parton distribution functions ( pdfs ) . to evaluate the cross sections we have used cteq6.6 pdfs @xcite via lhapdf software library @xcite .
the cross sections for proton - proton collisions are presented on fig .
[ quiy0 ] for @xmath96 appropriate for the 2011 lhc run , and for designed @xmath97 on fig .
[ quiy014 ] .
thereby we distinguish separately the production via neutral currents shown on the left panel , and via charged currents shown on the right panel of figs .
[ quiy0 ] and [ quiy014 ] . via neutral @xmath98 ( a ) and
charged @xmath99 currents ( b ) , in dependence on the heavy quintuplet mass @xmath100 . ] , but for designed @xmath101 at the lhc . ]
.production cross sections for @xmath63-@xmath102 pairs for the lhc run at @xmath96 , for three selected values of @xmath103 . [
cols="^,^,^,^ " , ]
the fermionic quintuplet of this paper was selected previously as a viable mdm candidate in @xcite . to employ it as a seesaw mediator requires additional scalar multiplets , what makes the neutral component of the quintuplet unstable .
if we employ a @xmath9 symmetry under which @xmath104 , @xmath105 and all sm fields are unchanged , the lightest component of the @xmath63 and @xmath64 multiplets is again a dm candidate . the quartic @xmath59 term in the scalar potential is still allowed so that neutrinos acquire radiatively generated masses given by eq .
( [ loop ] ) . if @xmath106 is the dm particle , its mass is fixed by the relic abundance to the value @xmath107 in @xcite .
the choice @xmath108 gives enough suppression to have small neutrino masses with large yukawas , @xmath109 . in this part of the parameter space
the model could have interesting lfv effects like in @xcite .
the neutral component of the scalar multiplet @xmath64 could also be a dm particle , despite its tree level coupling to the @xmath91 boson .
similar to the inert doublet model , the @xmath59 term in the scalar potential in eq .
( [ scalarpot ] ) splits the real and imaginary parts of the @xmath44 field .
if the mass splitting is large enough , the inelastic scattering through the @xmath91 boson exchange is kinematically forbidden in direct detection experiments . in this case
the new states could be within reach of the lhc . to ensure a large enough mass splitting
, the coupling @xmath59 can not be too small .
accordingly , from eq .
( [ loop ] ) , it follows that the yukawa couplings have to be smaller , suppressing the lfv effects .
in this account we propose a tev - scale model for neutrino masses based only on the gauge symmetry and the renormalizability of the sm .
the model employs fermion quintuplets with zero hypercharge , which in isolation correspond to cosmological minimal dark matter candidate .
such tev - scale fermions in conjunction with the scalar quadruplet can generate neutrino masses both by the tree - level contributions in fig .
[ dim9op ] and the loop level contributions in fig .
[ oneloop ] .
also , such states are expected to be abundantly produced at the lhc , and the distinctive signatures could come from doubly - charged components of the fermionic quintuplets . for @xmath4 of already accumulated integrated luminosity at the lhc ( @xmath110 tev ) , there could be @xmath5 produced doubly - charged @xmath6 or @xmath7 fermions , with mass @xmath8 .
the decays of these states to sm charged leptons have an interesting multilepton signature .
there are , in addition , same - sign dilepton events displayed in table 2 .
these events have a negligible sm background , as demonstrated @xcite in generic new physics scenario with lepton number violation .
in particular , the detailed studies of fermionic triplets from type iii seesaw scenario @xcite apply to our case of fermionic quintuplets .
the multilepton events listed in our table 2 are in direct correspondence with the items from type iii seesaw given in table 13 of ref .
@xcite , to which also the analysis in @xcite refers . the signals which are good for the discovery correspond to a relatively high signal rate and small sm background , which is calculated by madgraph @xcite .
in order to compare the signal and sm background cross sections , we assume a particular choice of parameters leading to the branching ratios given in table 2 , restricted to @xmath111 leptons in the final states . in this respect we distinguish four classes of events containing @xmath112 decaying to @xmath113 or @xmath114 lepton and @xmath115 resonance to help in @xmath112 identification : @xmath116 which has too small of a cross section ( 0.03 fb with respect to the sm background of 0.6 fb ) at 7 tev lhc ; @xmath117 which has a cross section of 0.7 fb , comparable to the sm background of 0.8 fb ; @xmath118 the lnv event having 0.7 fb with the same - sign dilepton state , which is nonexistent in the sm and thus devoid of the sm background ; @xmath119 having a relatively high signal rate ( 1.1 fb with respect to sm background of 0.8 fb ) . the classes @xmath120 and @xmath121 lead to signals elaborately detailed in @xcite , which can be rescaled to infer on the lhc discovery reach
the model of neutrino mass generation presented here represents a generalization of the fermion triplet seesaw mediator from the type iii seesaw model to a hypercharge zero quintuplet , which we postponed at the time of elaborating first its nonzero - hypercharge variant in @xcite .
the hypercharge zero quintuplet has been proposed subsequently in @xcite .
one can question the possibility to distinguish between the present hypercharge zero quintuplet with a doubly - charged component and previously considered nonzero hypercharge states @xcite , which provide spectacular falsifiable triply - charged signatures .
consequently , a nonobservation of the latter would put in a forefront the hypercharge zero quintuplets , which provide their neutral components as potential mdm candidates .
however , a blow to the stability of such dm comes from the presence of additional fields needed to generate the described small neutrino masses . as usual ,
the stability can be restored by introducing the discrete symmetry under which the new states are odd , in which case neutrinos get mass only from radiative contributions .
m. cirelli , n. fornengo and a. strumia , _ minimal dark matter _ , nucl .
b * 753 * , 178 ( 2006 ) [ hep - ph/0512090 ] ; m. cirelli and a. strumia , _ minimal dark matter : model and results _
, new . j. phys . * 11 * , 105005 ( 2009 ) [ 0903.3381 [ hep - ph ] ] .
p. minkowski , phys .
b * 67 * , 421 ( 1977 ) ; t. yanagida , in _ proceedings of the workshop on unified theory and the baryon number of the universe _ , edited by o. sawada and a. sugamoto ( kek , tsukuba , 1979 ) , p. 95 ; m. gell - mann , p. ramond , and r. slansky , in _
supergravity _ , edited by p. van nieuwenhuizen and d. freedman ( north holland , amsterdam , 1979 ) , p. 315 ; s.l .
glashow , in _ quarks and leptons _
, edited by m. l@xmath122vy _ et al . _
( plenum , new york , 1980 ) , p. 707
mohapatra and g. senjanovic , phys .
* 44 * , 912 ( 1980 ) .
w. konetschny and w. kummer , phys .
b * 70 * , 433 ( 1977 ) ; m. magg and c. wetterich , phys
b * 94 * , 61 ( 1980 ) ; j. schechter and j.w.f .
valle , phys .
d * 22 * , 2227 ( 1980 ) ; t.p .
cheng and l.f .
li , phys .
d * 22 * , 2860 ( 1980 ) ; g. lazarides , q. shafi , and c. wetterich , nucl .
b * 181 * , 287 ( 1981 ) ; r.n .
mohapatra and g. senjanovic , phys .
d * 23 * , 165 ( 1981 ) .
b. mukhopadhyaya and s. mukhopadhyay , _ same - sign trileptons and four leptons as signatures of new physics at the large hadron collider _ , phys .
d * 82 * , 031501 ( 2010 ) ; s. mukhopadhyay and b. mukhopadhyaya , _ same - sign trileptons at the lhc : a window to lepton - number violating supersymmetry _ , phys .
d * 84 * , 095001 ( 2011 ) [ 1108.4921 [ hep - ph ] ] .
f. del aguila and j.a .
aguilar - saavedra , _ electroweak scale seesaw and heavy dirac neutrino signals at lhc _ , phys .
b * 672 * , 158 ( 2009 ) [ 0809.2096 [ hep - ph ] ] ; j.a .
aguilar - saavedra , _ heavy lepton pair production at lhc : model discrimination with multi - lepton signals _
b * 828 * , 289 ( 20010 ) [ 0905.2221 [ hep - ph ] ] . | we propose a new seesaw model based on a fermionic hypercharge zero weak quintuplet in conjunction with an additional scalar quadruplet which attains an induced vacuum expectation value .
the model provides both tree - level seesaw @xmath0 and loop - suppressed radiative @xmath1 contributions to active neutrino masses .
the empirical masses @xmath2 ev can be achieved with @xmath3 tev new states , accessible at the lhc . for @xmath4 of accumulated integrated luminosity at the lhc , there could be @xmath5 doubly - charged @xmath6 or @xmath7 fermions with mass @xmath8 , leading to interesting multilepton signatures .
the neutral component of the fermion quintuplet , previously identified as a minimal dark matter candidate , becomes unstable in the proposed seesaw setup .
the stability can be restored by introducing a @xmath9 symmetry , in which case neutrinos get mass only from radiative contributions .
= 2 * * * * + = 0.5 kreimir kumeriki , ivica picek and branimir radovi + _ department of physics , faculty of science , university of zagreb , p.o.b .
331 , hr-10002 zagreb , croatia + _ + _ pacs _ : 14.60.pq ; 14.60.st ; 95.35.+d + _ keywords _ : neutrino mass ; exotic leptons ; dark matter |
the unified model for active galaxies @xcite is based on the existence of a dusty toroidal structure surrounding the central region of active galactic nuclei ( agn ) . considering this geometry of the obscuring material ,
the central engines of type-1 agn can be seen directly , resulting in typical spectra with both narrow and broad emission lines , whereas in type-2 agn the broad line region ( blr ) is obscured .
the infrared ( ir ) range ( and particularly the mid - infrared ; mir ) is key to characterize the torus , since the dust reprocesses the optical and ultraviolet radiation generated in the accretion process and re - emits it in this range .
however , considering the small torus size , high angular resolution turns to be essential to isolate as much as possible its emission .
pioneering work in modelling the dusty torus @xcite assumed a uniform dust density distribution to simplify the modelling , although from the start , @xcite realized that smooth dust distributions can not survive within the immediate agn vicinity . to solve the various discrepancies between observations and models , an intensive search for an alternative torus geometry
has been carried out in the last decade .
the clumpy torus models @xcite propose that the dust is distributed in clumps , instead of homogeneously filling the torus volume .
these models are making significant progress in accounting for the mir emission of agn @xcite . in previous works we constructed subarcsecond resolution ir spectral energy distributions ( seds ) for about 20 nearby seyfert galaxies and successfully reproduced them with the clumpy torus models of nenkova et al .
( hereafter clumpy ) .
it is worth mentioning , however , that some observational results show that the clumpy models alone can not explain the ir seds of a sample of pg quasars @xcite .
the latter authors needed an additional hot dust component to reproduce the seds .
moreover , recent interferometry results @xcite indicate that a single component torus does not reproduce the observations .
the clumpy database now contains @xmath15@xmath210@xmath3 models , calculated for a fine grid of model parameters .
the inherent degeneracy between these parameters has to be taken into account when fitting the observables .
to this end , we developed the bayesian inference tool bayesclumpy .
details on the interpolation methods and algorithms employed can be found in @xcite .
we point out that , given the specificities of the bayesian inference approach we follow , in the following analysis we will not be using the original set of models described in @xcite , but an interpolated version of them .
in @xcite we fitted ir seds constructed using mir nuclear fluxes obtained with 8 m telescopes and nir measurements of similar resolution from the literature ( see also @xcite ) .
some of the seds were well - sampled ( e.g. the circinus galaxy and centaurus a ) whilst others comprised just three photometric data points ( e.g. ngc 1365 and ngc 1386 ) .
the better the sampling of the sed , the more constrained the torus parameters ( see e.g. @xcite ) . considering the need for 8 - 10 m telescopes to isolate the torus emission ,
it is necessary to determine the minimum number of filters required to constrain the model parameters .
we utilize the output of our code bayesclumpy in a bayesian experiment design framework .
our aim is to design the experiment ( observation of a source using a selected filter ) that introduces more constraints for the parameters of the clumpy models . using our bayesian approach
, we can investigate which and how many optical , ir , and sub - mm filters restrict the most the parameter space , as well as which wavelengths provide more information about each of the parameters .
although here we present results for the models of nenkova et al . , the formalism can be applied to any other set of models , including multi - component ones .
thus , this work can be useful for the community when applying for telescope observations .
the clumpy models of @xcite hold that the dust surrounding the central engine of an agn is distributed in clumps .
these clumps are distributed with a radial extent @xmath4 , where @xmath5 and @xmath6 are the outer and inner radius of the toroidal distribution , respectively ( see figure [ clumpy_scheme ] ) .
the inner radius is defined by the dust sublimation temperature ( @xmath7 k ) , with @xmath8 pc . within this geometry , each clump has the same optical depth ( @xmath9 ) .
the average number of clouds along a radial equatorial ray is @xmath10 .
the radial density profile is a power - law ( @xmath11 ) .
a width parameter , @xmath12 , characterizes the angular distribution of the clouds , which has a smooth edge .
the number of clouds along the los at an inclination angle @xmath13 is @xmath14 . for a detailed description of the clumpy models
see @xcite . ) and the dust sublimation radius ( @xmath15 ) .
all the clouds are supposed to have the same @xmath9 , and @xmath12 characterizes the width of the angular distribution .
the number of cloud encounters is function of the viewing angle , @xmath13.,width=302 ] in every bayesian scheme , one needs to specify a - priori information about the model parameters .
this is done through the prior distribution ( see * ? ? ? * for more details ) .
we consider them to be truncated uniform distributions for the six model parameters in the intervals reported in table [ tab : parametros ] .
therefore , we give the same weight to all the values in each interval . to compare with the observations ,
bayesclumpy simulates the effect of the employed filters on the sed by integrating the product of the synthetic sed and the filter transmission curve .
observational errors are assumed to be gaussian .
.clumpy model parameters and considered intervals[tab : parametros ] [ cols="<,^,<",options="header " , ] fluxes in mjy from @xcite and @xcite . in the case of the sy1 and sy2 templates ,
the fluxes have been normalized to the central wavelength of the t - recs si2 filter .
the previous definitions are somewhat obvious and the core of bayesian decision theory is located on the appropriate definition of the utility function .
we follow here the suggestion of @xcite ( see also * ? ? ?
* ) , who suggested that , since we want to maximize the information about the model parameters @xmath16 , it makes sense to use the information encoded on the posterior for @xmath16 , as measured by the negative differential entropy : @xmath17 , \label{eq : utility}\ ] ] where the notation means that the entropy is computed for the posterior for the distribution @xmath18 .
the differential entropy is , following the standard definition , given by : @xmath19 = -\int d{\mbox{\boldmath$\theta$}}\ , p({\mbox{\boldmath$\theta$}}|o,\mathbf{d},i_f ) \log p({\mbox{\boldmath$\theta$}}|o,\mathbf{d},i_f ) .
\label{eq : entropy_def}\ ] ] consequently , we have to compute the following quantity for all the available filters @xmath20 and choose the filter @xmath21 that maximizes it .
one could plug the expression for the predictive distribution of eq .
( [ eq : predictive_distribution ] ) on the expected utility and end up with a triple multidimensional integral that needs to be computed for each value of @xmath21 .
however , it turns out to be much easier to plug eqs .
( [ eq : utility ] ) and ( [ eq : entropy_def ] ) onto eq .
( [ eq : expected_utility ] ) and apply bayes theorem ( e.g. , * ? ? ? * ) to @xmath22 , so that we have to end up with : @xmath23 using the definition of entropy of eq . ( [ eq : entropy_def ] ) , we can rewrite the previous expression as : @xmath24 \nonumber \\ & - & \int do \ , p(o|{\mbox{\boldmath$\theta$}},\mathbf{d},i_f ) h[{\mbox{\boldmath$\theta$}}|\mathbf{d},i_f ] \nonumber \\ & + & \int d{\mbox{\boldmath$\theta$}}\ , p({\mbox{\boldmath$\theta$}}|\mathbf{d},i_f ) h[o|\mathbf{d},i_f].\end{aligned}\ ] ] note that the entropies @xmath25 $ ] and @xmath26 $ ] can be extracted from the integrals because they do not depend on the integration variable . given that the probability distributions are normalized to unit area , the expression simplifies to : @xmath27 \nonumber \\ & - & h[{\mbox{\boldmath$\theta$}}|\mathbf{d},i_f ] + h[o|\mathbf{d},i_f ] .
\label{eq : eu}\end{aligned}\ ] ] clearly , the second term ( the entropy of the posterior distribution considering the existing data ) is independent of the election of the new filter , so it becomes constant with respect to @xmath21 and can be dropped from the computation .
the term @xmath28 $ ] , which is the entropy of the likelihood for the new observed point at filter @xmath21 , can be assumed to be constant if the expected noise is independent of the measurement .
this is the case when the measurement for the new filter @xmath21 is done under the presence of additive ( gaussian ) noise with a variance that is independent of @xmath21 .
however , this is not usually the case since the noise variance at different wavelengths can vary a lot . assuming that the observation @xmath29 is perturbed with gaussian noise with variance @xmath30 , the likelihood @xmath31 is a gaussian distribution , so its entropy is analytically expressed as : @xmath32 = \frac{1}{2 } \log \left ( 2\pi \mathrm{e } \sigma_f^2 \right).\ ] ] given that the previous entropy is independent of @xmath16 , it can be extracted from the integral in eq .
( [ eq : eu ] ) and use the fact that the posterior is normalized to unit area . therefore , the final expression for the expected utility that we use in this work simplifies to : @xmath33 with @xmath34 given by eq .
( [ eq : predictive_distribution ] ) . note that if the noise variance does not depend on @xmath21 , we end up with the expected utility derived by @xcite and @xcite .
they noted that the previous expression leads to a maximum entropy sampling , so that we select the new filter where we know the least .
in other words , we should sample where the current values of the model parameters allow for a large variability of the sed . the computation of the integral of eq .
( [ eq : eu_final ] ) can be done following @xcite , who used a monte carlo estimation technique .
first , we can compute a monte carlo estimation of @xmath34 using : @xmath35 where the @xmath36 are @xmath37 samples from the posterior distribution @xmath38 that have been previously computed with bayesclumpy for the current set of observations . then , the integral of eq .
( [ eq : eu_final ] ) , the entropy of the predictive distribution , is computed using another monte carlo estimation : @xmath39 \approx \frac{1}{n_\mathrm{pred } } \sum_{j=1}^{n_\mathrm{pred } } \log p(o_j|\mathbf{d},i_f),\ ] ] where the @xmath40 are samples from @xmath34 .
however , we have tested that better results are obtained if a simple trapezoidal quadrature is used to compute the integral of eq .
( [ eq : eu_final ] ) . to this end , we discretize the posible range of variation of @xmath29 in bins . for each bin , we compute @xmath34 using the monte carlo estimation of eq .
( [ eq : montecarlo_predictive ] ) . at the end , we compute the entropy of the predictive distribution as : @xmath39 \approx \sum_j w_j p(o_j|\mathbf{d},i_f ) \log p(o_j|\mathbf{d},i_f),\ ] ] with @xmath41 the standard trapezoidal weights ( e.g. , * ? ? ? * ) .
the bayesian adaptive exploration theory , as summarized in the previous section , is applied to the case of seds generated by the emission of a clumpy torus around an agn . under the assumption that the parametric models of @xcite are representative of the underlying physics , our analysis allows us to search for the filter that introduces more restrictions on the model parameters . our set of potential filters
is listed in table [ tab : filters ] .
our aim is to consider a sufficiently complete list of photometric filters available in many medium- to large - aperture telescopes .
in particular , we are considering all the optical , nir , mir and far - ir ( fir ) filters that we have employed in our previous work , plus four new mir filters from the canaricam instrument , which has recently started to operate at the 10 m gran telescopio canarias ( gtc ) , and another four filters from the long wavelength camera ( lwc ) on the forcast instrument .
forcast is a mir / fir camera for the stratospheric observatory for infrared astronomy ( sofia ; @xcite ) . finally , to inspect whether or not sub - mm data can set any constraints on the model parameters , we have also added channels 7 and 9 of alma , although they can not be considered filters per se .
they are simulated as a square transmission efficiency with a peak of 80% in the ranges @xmath42 $ ] @xmath0 m for channel 7 and @xmath43 $ ] @xmath0 m for channel 9 .
table [ tab : filters ] presents the filter identification , together with the instrument and telescope in which they are mounted .
additionally , we have tabulated the central wavelength of the filter , computed as @xmath44 , where @xmath45 is the transmission efficiency of the filter , and the full - width at half maximum ( fwhm ) .
the transmission efficiency of the filters have been obtained from the literature .
it is important to note that , for the sake of simplicity , in this work we are not taking into account the angular resolution provided by the different instruments considered here . in other words
, we are assuming that all instruments listed in table [ tab : filters ] are equivalent except in the wavelength coverage .
however , the reader must keep in mind that high spatial resolution is mandatory when trying to isolate the torus emission .
thus , aside from the expected utilities derived here for the different instruments , it is important to consider the resolution when deciding which should be the next observation .
black dots correspond to filters with @xmath46 @xmath0 m , red dots to @xmath47 @xmath0 m , blue dots to @xmath48 @xmath0 m and green dots to @xmath49 @xmath0 m . [ fig : expected_utility ] to demonstrate our approach , we have used four seds , two representative of seyfert 1 ( sy1 ) and another two representative of seyfert 2 ( sy2 ) galaxies .
these seds have been discussed by @xcite and @xcite using a deep bayesian analysis .
additionally , here we use the sy1 and sy2 average templates presented in @xcite .
these mean templates were constructed using individual sy1 and sy2 seds of angular resolution @xmath500.55 arcsec , that were first interpolated onto the circinus wavelength grid ( 1.265 , 1.60 , 2.18 , 3.80 , 4.80 , 8.74 , and 18.3 @xmath0 m ) , and then normalized to the central wavelength of the t - recs si2 filter ( 8.74 @xmath0 m ) . these flux densities and their associated errors
are shown in table [ tab : fluxes ] and are displayed in graphical form in fig .
[ fig : seds ] , normalized to the filter that is closer to 10 @xmath0 m .
note that the seds that we have selected have quite dense samplings , as opposed to other sparsely sampled ones @xcite .
for each sed we have carried out a full bayesian analysis using bayesclumpy .
this analysis relies on sampling the posterior distribution function as a function of the model parameters ( shown in table [ tab : parametros ] ) .
we have used flat priors for all model parameters . for the sy1 galaxies
, we have additionally included an agn spectrum , which is added to the clumpy sed .
the shape of the agn spectrum is the same that is considered for the infrared pumping of the dust in the torus @xcite .
likewise , we also take into account a certain amount of extinction following the law of @xcite .
this introduces the free parameter @xmath51 , which is included into the inference for sy1 galaxies with a flat prior in the range @xmath52 $ ] mag .
the inclusion of these two ingredients turns out to be essential for obtaining reliable results for sy1 sources .
bayesclumpy generates samples of model parameters distributed according to the posterior distribution .
figure [ fig : seds ] shows , apart from the observed points , a representation of the seds reconstructed from the model parameters sampled from the posterior .
note that the sampled seds have a small variability close to the observed points with the smaller error bars . in principle , and according to the maximum entropy sampling derived by @xcite and @xcite , the filter that we should select has to be located close to the wavelength where the seds contain larger variability with the observational information we have acquired .
an evaluation of eq .
( [ eq : eu_final ] ) , following the ideas of sec .
[ sec : technical ] , gives the results displayed in the left column of fig .
[ fig : expected_utility ] .
this figure represents the value of @xmath53 for filter @xmath21 versus the central wavelength of the filter .
the noise associated to any new observation , @xmath30 , depends on the wavelength : we consider 10% of the value of the flux at @xmath54 , 20% at @xmath55 and 30% at @xmath56 .
these are the typical uncertainties that we measured for individual galaxies in our previous work ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the expected utility is usually measured in terms of `` nats '' ( or bits ) , i.e. , units of information . however , since we are not interested in their actual value but in relative measures , we normalize the plots to the maximum .
several properties are evident .
first , the lack of a vertical spread on the plots for a given wavelength means that the expected utility of the filters depends essentially on the wavelength . using different filters with the same central wavelength but different transmission properties
produce the same increase in the information acquired .
in other words , the results are not dependent on the exact shape of the filter transmission curve .
obviously , other reasons ( atmospheric transmission , telescope diameter , availability , etc . )
would make one particular filter preferable among a set of equivalent filters .
however , under the same conditions , they introduce roughly the same information for constraining clumpy models . in general ,
the region between 10 and 200 @xmath0 m produces the largest increase in the expected utility . out of the filters considered here , those from sofia provide the largest constraining power for the clumpy models .
however , the poor spatial resolution of sofia ( @xmath134 arcsec ) represents a clear drawback and it might be preferable to choose other filters with slightly smaller expected utility but better spatial resolution .
the filters in the q - band , especially the canaricam q8 filter ( @xmath5724.5 @xmath0 m ) , are the next most useful ones for constraining the clumpy torus model parameters , specially if we consider their good spatial resolution ( @xmath10.55 arcsec ) .
they are closely followed by filters in the n - band ( @xmath18 - 12 @xmath0 m ) and by the herschel space observatory pacs filters , especially pacs70 and pacs100 .
the previous results are not surprising considering that the bulk of torus emission is observed in the mir , peaking at @xmath120 @xmath0 m
. the pacs observations , which cover the spectral range from 70 to 160 @xmath0 m , are sampling cooler dust within the torus , and thus helping to constrain parameters such as the torus radial extent ( @xmath58 ) and the radial distribution of the clouds ( @xmath59 ) , as discussed in @xcite .
according to our results , data points obtained with the spire instrument on herschel despite its low spatial resolution , which we are neglecting here still have a relatively large expected utility , and the same happens with sub - mm data from alma .
we have considered two alma channels ( 7 and 9 ) and the expected utility of channel 9 ( @xmath60 @xmath0 m ) is comparable to that provided by the nir data in the case of the sy2 galaxies .
this result is encouraging for potential alma users looking for constraints on torus properties .
alma will observe in the sub - mm regime with unprecedented angular resolution , and according to the analysis presented here , data obtained with channel 9 can provide useful constraints on torus model parameters .
the case of herschel spire and alma clearly illustrates what we discussed at the beginning of this section : although the expected utilities measured for the spire filters are slightly higher than those for alma channel 9 , the angular resolution provided by alma makes it a better choice than spire to sample the cool dust within the torus . on the other hand ,
the expected utility of channel 7 ( @xmath61 @xmath0 m ) is negligible in all the cases considered .
it is interesting to note that there is no apparent difference between sy1 and sy2 galaxies regarding the proposed new filters at wavelengths @xmath624 @xmath0 m .
therefore , it seems that spectral features that are supposed to serve to constrain the model parameters because they are generally different in sy1 and sy2 ( such as the silicate feature at 10 @xmath0 m , generally in emission in sy1 and in absorption in sy2 ) do not make any difference in suggesting new observations .
it is worth noting , though , that here we are using photometry only , which can provide just an insight of spectral features such as the silicate band .
the use of high angular resolution spectroscopy in this spectral region ( provided by ground - based instruments such as visir , michelle and canaricam ) can set constraints on the clumpy torus models that the photometry alone can not @xcite .
thus , the use of spectroscopy would possibly change the suggested observations .
unfortunately , the expression for the expected utility of eq .
( [ eq : eu_final ] ) in the spectroscopic case requires the computation of a multidimensional integral because the expected observation @xmath29 transforms into the vector @xmath63 .
tailor - made algorithms which are computationally heavy have to be applied to compute this integral . for this reason
, we defer the spectroscopic case to a later study . at wavelengths
@xmath504 @xmath0 m we start to see a difference between the expected utilities of sy1 and sy2 .
optical and nir observations help constraining the torus parameters for sy1 ( expected utilities between 0.4 and 0.8 ) more than for sy2 ( expected utilities @xmath10.2 around 1 @xmath0 m ) .
the hot dust emission from the directly illuminated faces of the clumps close to the central engine and the direct agn emission strongly flatten the sy1 ir seds ( see figure [ fig : seds ] ) .
thus , in the case of sy1 , more observations at short wavelengths are required to constrain the shape of the sed in this region , that in turn constrains parameters such as the inclination angle of the torus ( @xmath13 ) . as an experiment , we have simulated which is the influence of having a reduced set of observations . to this aim
, we have considered the observed points below and above 4 @xmath0 m from the complete seds .
this separation was chosen to roughly separate the nir and the mir keeping a similar number of points in both spectral domains .
the resulting seds are shown in the left and right six panels of fig .
[ fig : seds_above_below ] respectively , using the same normalization as in fig .
[ fig : seds ] .
again , the grey curves are samples of seds using the model parameters from the bayesian inference . despite the sparser sampling of the seds in the two cases , the results remain essentially similar ( see central and right panels of fig [ fig : expected_utility ] ) , with the expected utility peaking in the 10 - 200 @xmath0 m region .
the only difference in the results is a small increase in the expected utitility of the optical and nir data when only @xmath644 @xmath65 data is considered in the sy2 seds ( bottom right panels in figure [ fig : expected_utility ] ) .
the process we have described until this point is obviously not complete . as already described in sec . [
sec : bae ] , the bayesian adaptive exploration scheme starts from observations , and then applies bayesian inference tools to design a new potentially informative observation .
since the typical timescale of this process can be very large given the necessity to apply for observing time in large - aperture telescopes , we have carried out a simulated full process in fig . [
fig : experiment ] . we have used the sed of the sy2 galaxy ngc 3081 presented in @xcite .
the reason for choosing this galaxy is the availability of herschel pacs nuclear fluxes . as a first step
, we have assumed that we only have the observation in the nicmos f160w filter and let the bayesian adaptive exploration scheme select the following filters .
bayesclumpy then uses this point to sample the posterior distribution and obtain marginal posteriors for the clumpy model parameters .
the observed point , together with the seds sampled from the posterior are shown in the first column of fig .
[ fig : experiment ] .
the second column shows the expected utility , while the remaining panels display the marginal posterior for @xmath12 , @xmath58 , @xmath10 and @xmath13 , respectively . in order to simulate the next experiment ,
we have selected the next observation from the set of observed points of the full sed that has the largest expected utility , which turns out to be the t - recs qa data point ( @xmath66 @xmath0 m ) .
the full process is then repeated until having all the observed points of the sed in the lowest panel .
the order in which the remaining filters have been selected is : pacs70 , pacs100 , pacs160 , visir neii@xmath672 ( @xmath68 @xmath0 m ) and finally , t - recs si2 ( @xmath69 @xmath0 m ) . given that the differences in expected utility are quite small and can change a little because of the inherently random character of the montecarlo sampling done with bayesclumpy the specific order might change from run to run .
in fact , this means that one could choose a filter giving a large expected utility even though it is not the one giving the largest value if observing with this filter is easier for some specific reason .
it is important to note how the marginal posteriors are constrained when augmenting the number of observed points .
once there are two points in the nir and mir , the marginal posterior of @xmath12 is very similar to what we find when considering the full sed .
this is not the case for @xmath58 and @xmath10 , that considerably change after introducing the fir data and the n - band data points ( neii@xmath702 and si2 filters ) .
finally , the fir does not seem to make any difference to the inclination angle of the torus , @xmath13 , but the addition of the n - band data points especially the si2 filter result in a different posterior shape , favouring intermediate torus inclinations .
the clumpy models have recently gained much success on explaining the observed ir sed of the inner parsecs of nearby seyfert galaxies . given the observational difficulty ( one needs to obtain good signal - to - noise ir photometry of sources at subarcsecond resolution , and ideally , spectroscopy ) ,
the sampling of the seds is generally quite scarce . assuming that the clumpy models are successful in explaining the observed sed , we have applied a bayesian adaptive exploration scheme to propose new observations given the presently available information about a source .
the results quantitatively indicate that the region between 10 and 200 @xmath0 m is crucial , almost independently of the presently available observed points .
we also find that optical and nir data have higher expected utilities in the case of sy1 than in sy2 .
finally , data from 200 to 500 @xmath0 m are found to be useful , even if not as much as data in the 10 - 200 @xmath0 m range , for constraining the torus model parameters .
it is important to note that , for simplicity , here we are not considering the angular resolution provided by the different instruments .
thus , aside from the measured expected utility of a given filter , it is important to consider the resolution of the data when deciding the next observation .
we performed a bae experiment for the galaxy ngc 3081 and found that having two data points only one in the nir and another in the mir significantly constrains the the torus width ( @xmath12 ) , which does not change when adding further ir data points .
the addition of fir data helps constraining the torus radial extent ( @xmath58 ) and the number of clouds ( n@xmath71 ) .
finally , the combination of the j , si2 and q filters appears to be the most suitable to constrain the inclination angle of the torus ( @xmath13 ) .
note , however , that these are the results found for a particular galaxy , and they may not be applicable to all seyferts . given the large pressure on large - aperture telescopes , our approach when combined with spatial resolution considerations is very promising for constraining clumpy models , and possibly torus models in general , in sparsely sampled seds with the smallest number of observed points .
the bayesian adaptive exploration scheme can be applied to any parametric model ( e.g. , smooth torus , light curves of classical eclipsing binaries , x - ray binaries ) once the tools to carry out a fully bayesian inference are available .
the application to the clumpy models is straightforward thanks to the bayesclumpy tool .
we note that the computation of the expected utility is already included in the public version of bayesclumpy , which is available for the community .
aar and cra acknowledge almudena alonso herrero for very useful comments .
aar acknowledges financial support by the spanish ministry of economy and competitiveness through projects aya2010 - 18029 ( solar magnetism and astrophysical spectropolarimetry ) and consolider - ingenio 2010 csd2009 - 00038 .
cra acknowledges the spanish ministry of science and innovation ( micinn ) through project consolider - ingenio 2010 program grant csd2006 - 00070 : first science with the gtc ( http://www.iac.es/consolider-ingenio-gtc/ ) and the spanish plan nacional grant estallidos aya2010 - 21887-c04.04 .
we finally acknowledge useful comments from the anonymous referee . t. j. , 2004 , in 23rd international workshop on bayesian inference and maximum entropy methods in science and engineering , vol . 707 , bayesian adaptive exploration , zhai g. j. e. .
y. , ed . , | recent success in explaining several properties of the dusty torus around the central engine of active galactic nuclei has been gathered with the assumption of clumpiness .
the properties of such clumpy dusty tori can be inferred by analyzing spectral energy distributions ( seds ) , sometimes with scarce sampling given that large aperture telescopes and long integration times are needed to get good spatial resolution and signal .
we aim at using the information already present in the data and the assumption of clumpy dusty torus , in particular , the clumpy models of nenkova et al .
, to evaluate the optimum next observation such that we maximize the constraining power of the new observed photometric point .
to this end , we use the existing and barely applied idea of bayesian adaptive exploration , a mixture of bayesian inference , prediction and decision theories .
the result is that the new photometric filter to use is the one that maximizes the expected utility , which we approximate with the entropy of the predictive distribution . in other words
, we have to sample where there is larger variability in the seds compatible with the data with what we know of the model parameters .
we show that bayesian adaptive exploration can be used to suggest new observations , and ultimately optimal filter sets , to better constrain the parameters of the clumpy dusty torus models . in general , we find that the region between 10 and 200 @xmath0 m produces the largest increase in the expected utility , although sub - mm data from alma also prove to be useful .
it is important to note that here we are not considering the angular resolution of the data , which is key when constraining torus parameters .
therefore , the expected utilities derived from this methodology must be weighted with the spatial resolution of the data . [ firstpage ] methods : data analysis , statistical galaxies : active , nuclei , seyfert |
black hole candidates ( bhcs ) are characterized by rapid x - ray variability ( see recent reviews by van der klis 1995 and cui 1999a ) .
it is also common for bhcs that the variability at high energies lags behind that at low energies ( cui 1999a and references therein ) , which is often referred to as hard lag .
the hard lag is often attributed to thermal inverse - comptonization processes ( e.g. , miyatomo et al .
1988 ; hua & titarchuk 1996 ; kazanas et al . 1997 ; bttcher & liang 1998 ; hua et al .
1999 ) , which are generally thought to be responsible for producing the characteristic hard tail in the x - ray spectra of bhcs ( tanaka & lewin 1995 ) . in these models ,
the hard lag arises simply because a greater number of scatterings are required for seed photons to reach higher energies .
therefore , the lag is directly related to the diffusion timescale through the comptonizing region , which scales logarithmically with photon energy ( e.g. , payne 1980 ; hua & titarchuk 1996 ) .
the expected logarithmic energy - dependence of the hard lag is in rough agreement with the observations ( cui et al .
1997 ; crary et al . 1998
; nowak et al . 1999 ) . however , the measured lag is often large ( e.g. , a few tenths of a second ) at low frequencies , which would require a very extended comptonizing region ( kazanas et al .
1997 ; bttcher & liang 1998 ; hua et al .
it is not clear whether such a region can be physically maintained ( nowak et al .
1999 ; bttcher & liang 1999 ; poutanen & fabian 1999 ) .
suggestions have also be made to link the hard lag either to the propagation or drift time scale of waves or blobs of matter through an increasingly hotter region toward the central black hole where hard x - rays are emitted ( miyamoto et al .
1988 ; kato 1989 ; bttcher & liang 1999 ) or to the evolution time scale of magnetic flares ( poutanen & fabian 1999 ) . regardless which scenario turns out to be close to reality , it is clear that the hard lag is an important property of bhcs which we can use to gain insight into the geometry and dynamics of accretion flows in these systems .
recently , however , it was discovered that a strong qpo in grs 1915 + 105 , a well - known microquasar , had a rather complex pattern of phase lag ( cui 1999b ) : while the hard lag was measured for the odd harmonics of the signal , the even harmonics displayed _ soft _ lag .
the pattern is puzzling because it does not fit naturally into any of the models suggested for bhcs .
speculation was made that the complicated qpo lag in this case might be caused by a change in the form of the wave that produced the qpo ( cui 1999b ) .
it is , however , not clear what physical mechanisms could be responsible for such evolution of the wave form .
similar behavior was subsequently observed for some of the qpos in xte j1550 - 564 ( wijnands et al .
therefore , the phenomenon may actually be common for bhcs .
a related timing property to the phase lag is the coherence function between two different energy bands . only recently ,
however , enough attention is paid to the importance of this property ( vaughan & nowak 1997 ) and efforts are made to compute it along with the phase lag .
consequently , the results are very limited .
it is nevertheless interesting to note that for bhcs the coherence function often appears to be around unity over a wide frequency range the x - ray variabilities in different energy bands are almost perfectly linearly correlated on those timescales in fourier domain ( vaughan & nowak 1997 ; cui et al .
1997 ; nowak et al .
this puts additional constraints on the models for x - ray production mechanisms in bhcs .
lower coherence was observed of cyg x-1 when the source was in the transitional periods between the two spectral states ( cui et al .
this could be attributed to the variation of the comptonizing region during those episodes on timescales less than an hour ( cui et al .
1997 ) , in the context of comptonization models ( hua et al .
1997 ) . however , more data is required to verify such a scenario . in this letter , we present the results from measuring the phase lag and coherence function of x - ray variability for xte j1550 - 564 during the initial rising phase of the 1998 outburst ( cui et al .
1999 , paper 1 hereafter ) .
in addition to the intense aperiodic variability , a strong qpo was detected , along with its first and sometimes second harmonics , and the frequency of the qpo increased by almost 2 orders of magnitude during this period ( paper 1 ) .
we examine the timing properties of both the qpo and broad - band variability .
paper 1 should be consulted for the details of the observations . very briefly
, there were 14 rxte observations , covering the rising phase of the outburst . in the first observation , however , the overflow of the on - board data buffers ( due to the inappropriate data modes adopted ) produced gaps in the data . for this work , we chose to ignore this observation entirely .
for the remaining 13 observations , we rebinned the data with @xmath0 s time bins and combined the _ event _ and _ binned _ data into the six energy bands as defined in paper 1 .
we chose the 2 4.5 kev band as the reference band . a cross - power spectrum ( cps )
was computed for each 256-second data segment between the reference band and each of the higher energy bands .
the results from all segments were then properly weighed and averaged to obtain the cpss for the observation .
the phase of a cps represents a phase shift of the light curve in a selected energy band with respect to that of the reference band .
we followed the convention that a positive phase indicates that the x - ray variability in the high energy band lags behind that in the low energy band , i.e. , a hard lag .
the uncertainty of the phase lag was estimated from the standard deviations of the real and imaginary parts of the cps . for the phase lag associated with a qpo ,
the magnitude was derived from fitting the cps in a narrow frequency range around the qpo with a linear function ( for the continuum ) and two lorentzian functions whose centroid frequencies and widths were fixed at those of the qpo ( paper 1 ) .
acceptable fits ( i.e. , the reduced @xmath1 around unity ) were obtained for all cases .
the corresponding errors were derived by varying the parameters until @xmath2 ( i.e. , representing roughly @xmath3 confidence intervals ; lampton et al .
in all observations , significant hard lag was found to be associated with the broad - band variability .
the phase lag shows significant evolution during the rising phase , which can be divided into two distinct periods .
1 shows an example for each period fourier spectra of power density , phase lag , and coherence function . at the early stage of the rising phase
, the broad - band hard lag is significantly measured only in the middle range of frequencies .
the lag increases with frequency first , peaks at some characteristic frequency , and then decreases .
the peak frequency does not appear to correspond to any characteristic features in the power density spectrum ( pds ) . at the frequency of the qpo , _ soft _ lag is clearly detected , on top of the broad - band hard lag .
however , there does not appear to be significant phase lag ( hard or soft ) associated with the first harmonic . at the later stage of the rising phase , while the hard lag is still apparent there also seems to be significant _ soft lag _ associated with the broad - band variability at frequencies around the qpo and its harmonics . the soft lag associated with the fundamental component of
the qpo remains significant , but now _ hard _ lag is also measured for the first harmonic . as for the coherence function , it is high and nearly constant at low frequencies and drops off sharply almost right after the first harmonic of the qpo for both periods of the rising phase .
it is interesting to notice that at the frequencies of the qpo and its harmonics the coherence function increases significantly .
both the broad - band and qpo lags show strong energy dependence , as illustrated in fig .
they become larger at higher energies ( with respect to the reference band ) , which is typical of bhcs ( cui 1999a ) .
the qpo lag also shows correlation with x - ray flux ( and thus with the qpo frequencies ; paper 1 ) for the second period of the rising phase , as is apparent in fig .
both the soft and hard qpo lags increase as the source brightens , reaching as high as 0.1 and 0.3 radians , respectively .
the coherence function is near unity at the beginning of the rising phase .
subsequently , it decreases rapidly and seems to level off in the end . to quantify the evolution
, we averaged the observed values over a frequency range 0.004 - 0.1 hz where the coherence function is roughly constant . as an example , the results for the 8.1 - 13.3 kev band are plotted in fig . 4 .
the loss of coherence is apparent during the initial rising phase .
moreover , the coherence function also decreases as the separation between the two energy bands widens , as shown in fig .
although the broad - band hard lag is significantly detected only at intermediate frequencies , large uncertainties prevent us from drawing any definitive conclusions on the results at low frequencies ( where the lag is expected to be small ) .
it appears , from fig .
2 , that at the early stage of the rising phase the corresponding _ time lag _ ( @xmath4 ) saturates below a characteristic frequency where the _ phase lag _ peak .
this is unusual because for other bhcs the time lag seems to monotonically increase toward low frequencies ( e.g. , miyatomo et al .
1988 ; cui et al . 1997 ; grove et al . 1998 ; nowak et al . 1999 ) . in the context of comptonization models ,
such a saturation in time lag might be the manifestation of the finiteness of the comptonizing region ( hua et al .
if so , the characteristic frequency ( a few hz ) would provide a direct measure of the outer radius of the region ( i.e. , tens of lt - ms ) , which seems to be in rough agreement with that estimated from the measured time lag .
it is interesting to notice that the characteristic frequency appears to increase as the source brightens , perhaps indicating that the size of the comptonizing region decreases , as was suggested for cyg x-1 when the source goes from the hard ( or low ) state to the soft ( or high ) state .
however , it is puzzling why the lag spectrum appears so different at the later stage of the rising phase , or more specifically what causes the observed broad - band _ soft _ lag around the qpo and its harmonic ( see fig . 1 ) . the complex pattern of the phase lag associated with the qpo bear remarkable resemblance to that observed of grs 1915 + 105 ( cui 1999b ) .
perhaps , the phenomenon is common for certain types of qpos in bhcs . combined with the published results ( wijnands et al .
1999 ) , our results show that the phenomenon seems generic for xte j1550 - 564 , as opposed to being limited to certain spectral states . more importantly , we are now seeing the evolution of the phenomenon as the qpo evolves during the rising phase of the outburst .
if we accept the comptonization scenario for the broad - band lag , we would have to rule out this scenario for the qpo because the qpo time lag _ increases _ with x - ray flux during the rising phase ( see fig .
this would not be surprising , since , as discussed by cui ( 1999b ) , it is problematic to attribute the observed soft and hard lags of the qpo ( and its harmonic ) entirely to inverse - comptonization processes in the first place . at present , no models can naturally account for this type of phase lag phenomenon . in the context of comptonization models , the loss of coherence during the rising phase may be due to the variation in the physical conditions of the comptonizing region ( hua et al .
1997 ) , as was suggested for cyg x-1 during spectral state transitions ( cui et al .
this scenario might also explain why more widely separated energy bands are less coherent ( see fig . 5 ) , since the difference in the number of scatterings that seed photons experience is greater .
the observed fourier spectra of the coherence function are , however , somewhat unusual . for a number of bhcs ,
the coherence function is typically close to unity over the entire frequency range where it can be reliably determined ( e.g. , vaughan & nowak 1997 ; cui et al .
1997 ; nowak et al .
1999 ) . here
, the coherence function drops precipitously above a `` break frequency '' ( which appears to be near the first harmonic of the qpo ; see fig . 1 ) .
we speculate that the break frequency might be indicative of the timescale on which the comptonizing region varies .
the break frequency appears to evolve in unison with the frequency of the qpo , which could imply that ( 1 ) the qpo originates in the comptonizing region ( instead of the accretion disk , as often thought ) and ( 2 ) the comptonizing region varies on an increasingly short timescale during the initial rising phase of the outburst .
the former is also supported by the fact that the qpo becomes stronger at higher energies ( paper 1 ; also see discussion in cui 1999b ) .
for bhcs in general , such energy dependence of qpos is observed in most cases ( cui 1999a ) . associating the qpo with the comptonizing region directly
might also account for the intriguing increase of the coherence function at the frequencies of the qpo and its harmonics , if the qpo is more localized , given that the broad - band variability is probably a disk phenomenon and compton upscattering of disk photons in an extended non - static `` corona '' can cause the loss of coherence .
bttcher , m. , & liang , e. p. 1998
, , 506 , 281 bttcher , m. , & liang , e. p. 1999
, , 511 , l37 crary , d. j. , et al .
1998 , , 493 , 71 cui , w. , zhang , s. n. , focke , w. , & swank , j. 1997 , , 484 , 383 cui , w. 1999a , proc .
`` high - energy processes in accreting black holes '' , eds .
j. poutanen & r. svensson ( asp : san francisco ) , asp conf .
161 , p. 97
( astro - ph/9809408 ) cui , w. , chen , w. , zhang , s. n. , & morgan , e. h. 1999 , , 512 , l43 ( paper 1 ) cui , w. 1999b , , 524 , l59 grove , j. e. , et al .
1998 , , 502 , 45 hua , x .- m . , &
titarchuk , l. 1996 , , 496 , 280 hua , x .-
m . , kazanas , d. & titarchuk , l. 1997 , , 482 , l57 hua , x .- m . , kazanas , d. , & cui , w. 1999 , , 512 , 793 kato , s. 1989 , , 41 , 745 kazanas , d. , hua , x .-
& titarchuk , l. 1997 , , 480 , 735 lampton , m. , margon , b. , & bowyer , s. 1976 , , 208 , 177 miyamoto , s. , et al .
1988 , nature , 336 , 450 nowak , m. a. , vaughan , b. a. , wilms , j. , dove , j. b. , & begelman , m. c. 1999 , , 510 , 874 payne , d. g. 1980 , , 441 , 770 poutanen , j. , & fabian , a. c. 1999 , , 306 , l31 tanaka , y. , & lewin , w. h. g. 1995 , in `` x ray binaries '' , eds .
w. h. g. lewin , j. van paradijs , & e. p. j. van den heuvel ( cambridge u. press , cambridge ) p. 126
van der klis , m. 1995 , in `` x - ray binaries '' , eds . w. h. g. lewin , j. van paradijs , & e. p. j. van den heuvel ( cambridge u. press , cambridge ) p. 252
vaughan , b. a. & nowak , m. a. 1997 , , 474 , l43 wijnands , r. , homan , j. , & van der klis , m. 1999 , , 526 , l33 | we report the results from measuring the phase lag and coherence function of x - ray emission from black hole candidate ( bhc ) xte j1550 - 564 .
these x - ray temporal properties have been recognized to be increasingly important in providing important diagnostics of the dynamics of accretion flows around black holes . for xte j1550 - 564
, we found significant hard lag the x - ray variability in high energy bands _ lags _ behind that in low energy bands associated both with broad - band variability and quasi - periodic oscillation ( qpo ) . however , the situation is more complicated for the qpo : while hard lag was measured for the first harmonic of the signal , the fundamental component showed significant _
soft _ lag .
such behavior is remarkably similar to what was observed of microquasar grs 1915 + 105 .
the phase lag evolved during the initial rising phase of the 1998 outburst .
the magnitude of both the soft and hard lags of the qpo increases with x - ray flux , while the fourier spectrum of the broad - band lag varies significantly in shape .
the coherence function is relatively high and roughly constant at low frequencies , and begins to drop almost right after the first harmonic of the qpo .
it is near unity at the beginning and decreases rapidly during the rising phase .
also observed is that the more widely separated the two energy bands are the less the coherence function between the two .
it is interesting that the coherence function increases significantly at the frequencies of the qpo and its harmonics . we discuss the implications of the results on the models proposed for bhcs |
optical networks have traditionally employed three main switching paradigms , namely circuit switching , burst switching , and packet switching , which have extensively studied respective benefits and limitations @xcite . in order to achieve the predictable network service of circuit switching while enjoying some of the flexibilities of burst and packet switching ,
_ dynamic circuit switching _ has been introduced @xcite .
dynamic circuit switching can be traced back to research toward differentiated levels of blocking rates of calls @xcite .
today , a plethora of network applications ranging from the migration of data and computing work loads to cloud storage and computing @xcite as well as high - bit rate e - science applications , e.g. , for remote scientific collaborations , to big data applications of governments , private organizations , and households are well supported by dynamic circuit switching @xcite
. moreover , gaming applications benefit from predictable low - delay service @xcite provided by circuits , as do emerging virtual reality applications @xcite .
also , circuits can aid in the timely transmission of data from continuous media applications , such as live or streaming video .
video traffic is often highly variable and may require smoothing before transmission over a circuit @xcite or require a combination of circuit transport for a constant base bit stream and packet switched transport for the traffic burst exceeding the base bit stream rate .
both commercial and research / education network providers have recently started to offer optical dynamic circuit switching services @xcite . while dynamic circuit switching has received growing research attention in core and metro networks @xcite , mechanisms for supporting dynamic circuit switching in passive optical networks ( pons ) , which are a promising technology for network access @xcite , are largely an open research area . as reviewed in section [ lit : sec ] , pon research on the upstream transmission direction from the distributed optical network units ( onus ) to the central optical line terminal ( olt )
has mainly focused on mechanisms supporting packet - switched transport @xcite . while some of these packet - switched transport mechanisms support quality of service akin to circuits through service differentiation mechanisms , to the best of our knowledge
there has been no prior study of circuit - level performance in pons , e.g. , the blocking probability of circuit requests for a given circuit request rate and circuit holding time . in this article , we present the first circuit - level performance study of a pon with polling - based medium access control .
we make three main original contributions towards the concept of efficiently supporting both * * dy**namic * * c**ircuit * * a**nd * * p**acket traffic in the upstream direction on a * pon * , which we refer to as * dycappon * : * we propose a novel dycappon polling cycle structure that exploits the dynamic circuit transmissions to mask the round - trip propagation delay for dynamic bandwidth allocation to packet traffic . * we develop a stochastic knapsack - based model of dycappon to evaluate the circuit - level performance , including the blocking probabilities for different classes of circuit requests .
* we analyze the bandwidth sharing between circuit and packet traffic in dycappon and evaluate packet - level performance , such as mean packet delay , as a function of the circuit traffic .
this article is organized as follows .
we first review related work in section [ lit : sec ] . in section [ sec :
model ] , we describe the considered access network structure and define both the circuit and packet traffic models as well as the corresponding circuit- and packet - level performance metrics . in section [ dycappon : sec ] , we introduce the dycappon polling cycle structure and outline the steps for admission control of dynamic circuit requests and dynamic bandwidth allocation to packet traffic . in section
[ sec : analysis ] we analyze the performance metrics relating to the dynamic circuit traffic , namely the blocking probabilities for the different circuit classes .
we also analyze the bandwidth portion of a cycle consumed by active circuits , which in turn determines the bandwidth portion available for packet traffic , and analyze the resulting mean delay for packet traffic . in section [ eval : sec ]
we validate numerical results from our analysis with simulations and present illustrative circuit- and packet - level performance results for dycappon .
we summarize our conclusions in section [ sec : conclusion ] and outline future research directions towards the dycappon concept .
the existing research on upstream transmission in passive optical access networks has mainly focused on packet traffic and related packet - level performance metrics .
a number of studies has primarily focused on differentiating the packet - level qos for different classes of packet traffic , e.g. , @xcite .
in contrast to these studies , we consider only best effort service for the packet traffic in this article . in future work
, mechanisms for differentiation of packet - level qos could be integrated into the packet partition ( see section [ dycappon : sec ] ) of the dycappon polling cycle .
the needs of applications for transmission with predictable quality of service has led to various enhancements of packet - switched transport for providing quality of service ( qos ) .
a few studies , e.g. , @xcite , have specifically focused on providing deterministic qos , i.e. , absolute guarantees for packet - level performance metrics , such as packet delay or jitter .
several studies have had a focus on the efficient integration of deterministic qos mechanisms with one or several lower - priority packet traffic classes in polling - based pons , e.g.,@xcite .
the resulting packet scheduling problems have received particular attention @xcite . generally , these prior studies have found that fixed - duration polling cycles are well suited for supporting consistent qos service .
similar to prior studies , we employ fixed - duration polling cycles in dycappon , specifically on a pon with a single - wavelength upstream channel .
the prior studies commonly considered traffic flows characterized through leaky - bucket parameters that bound the long - term average bit rate as well as the size of sudden traffic bursts .
most of these studies include admission control , i.e. , admit a new traffic flow only when the packet - level performance guarantees can still be met with the new traffic flow added to the existing flows . however , the circuit - level performance , i.e. , the probability of blocking ( i.e. , denial of admission ) of a new request has not been considered .
in contrast , the circuits in dycappon provide absolute qos to constant bit rate traffic flows without bursts and we analyze the probability of new traffic flows ( circuits ) being admitted or blocked .
this flow ( circuit ) level performance is important for network dimensioning and providing qos at the level of traffic flows . for completeness
, we briefly note that a pon architecture that can provide circuits to onus through orthogonal frequency division multiplexing techniques on the physical layer has been proposed in @xcite .
our study , in contrast , focuses on efficient medium access control techniques for supporting circuit traffic .
a qos approach based on burst switching in a pon has been proposed in @xcite . to the best of our knowledge ,
circuit level performance in pons has so far only been examined in @xcite for the specific context of optical code division multiplexing @xcite .
we also note for completeness that large file transmissions in optical networks have been examined in @xcite , where scheduling of large data file transfers on the optical grid network is studied , in @xcite , where parallel transfer over multiple network paths are examined , and in @xcite , where files are transmitted in a burst mode , i.e. , sequentially .
sharing of a general time - division multiplexing ( tdm ) link by circuit and packet traffic has been analyzed in several studies , e.g. @xcite .
these queueing theoretic analyses typically employed detailed markov models and become computationally quite demanding for high - speed links . also , these complex existing models considered a given node with local control of all link transmissions . in contrast , we develop a simple performance model for the distributed transmissions of the onus that are coordinated through polling - based medium access control in dycappon .
our dycappon model is accurate for the circuits and approximate for the packet service .
more specifically , we model the dynamics of the circuit traffic , which is given priority over packet traffic up to an aggregate circuit bandwidth of @xmath2 in dycappon , with accurate stochastic knapsack modeling techniques in section [ percir : sec ] . in section [ pkt_perf : sec ] , we present an approximate delay model for the packet traffic , which in dycappon can consume the bandwidth left unused by circuit traffic .
@xmath3 & transmission rate [ bit / s ] of upstream channel + @xmath2 & transm .
rate limit for circuit service , @xmath4 + @xmath5 & number of onus + @xmath6 & one - way propagation delay [ s ] + + @xmath7 & bit rates [ bit / s ] for circuit classes @xmath8 + @xmath9 & aggregate circuit requests arrival rate [ circuits / s ] + @xmath10 & prob . that a request is for circuit type @xmath11 + @xmath12 & mean circuit bit rate [ bit / s ] of offered circuit traf .
+ @xmath13 & mean circuit holding time [ s / circuit ] + @xmath14 & offered circuit traffic intensity ( load ) + @xmath15 , @xmath16 & mean [ bit ] and variance of packet size + @xmath17 & packet traffic intensity ( load ) ; @xmath18 is agg .
packet + & generation rate [ packets / s ] at all @xmath5 onus + + @xmath19 & total cycle duration [ s ] , constant + @xmath20 & cycle duration ( rand . var . ) occupied by circuit traf .
+ @xmath21 & mean per - cycle overhead time [ s ] for upstream + & transmissions ( report transm . times , guard times ) + + @xmath22 & state vector of numbers of circuits of class @xmath11 + @xmath23 & aggregate bandwidth of active circuits + @xmath24 & equilibrium probability for active circuits having + & aggregate bandwidth @xmath25 + + @xmath26 & blocking probability for circuit class @xmath11 + @xmath27 & mean packet delay [ s ] + we consider a pon with @xmath5 onus attached to the olt with a single downstream wavelength channel and a single upstream wavelength channel @xcite .
we denote @xmath3 for the transmission bit rate ( bandwidth ) of a channel [ bits / s ] .
we denote @xmath6 [ s ] for the one - way propagation delay between the olt and the equidistant onus .
we denote @xmath19 [ s ] for the fixed duration of a polling cycle .
the model notations are summarized in table [ not : tab ] .
( 160,32 ) ( -10,0)(1,0)190 ( -10,20)(1,0)190 ( -15,21)(0,0)[b]olt ( -15,-1)(0,0)[t]onu ( -5,20 ) ( 1,-3)6.6 ( -10,0)(1,3)6.6 ( 18,0)(1,3)6.6 ( 52,20)(1,-3)6.6 ( 46,0)(1,3)6.6 ( 59,0)(1,3)6.6 ( 107,20)(1,-3)6.6 ( 101,0)(1,3)6.6 ( 112,0)(1,3)6.6 ( 162,20)(1,-3)6.6 ( 156,0)(1,3)6.6 ( 169,0)(1,3)6.6 ( 11,0.75)(0,0)[b]@xmath28 ( 2.5,9.75)(0,0)[b]@xmath29 ( 33,0.75)(0,0)[b]@xmath30 ( 52.5,0.75)(0,0)[b]@xmath31 ( 57,9.75)(0,0)[b]@xmath29 ( 81,0.75)(0,0)[b]@xmath32 ( 107.25,0.75)(0,0)[b ] ( 112.5,9.75)(0,0)[b]@xmath29 ( 137,0.75)(0,0)[b]@xmath33 ( 167.5,9.75)(0,0)[b]@xmath29 ( -3,20)(0,1)10 ( 52,20)(0,1)10 ( 107,20)(0,1)10 ( 162,20)(0,1)10 ( -10,11)(0,0 ) [ ] @xmath34 ( 46,11)(0,0 ) [ ] @xmath34 ( 102,11)(0,0 ) [ ] @xmath34 ( 157,11)(0,0 ) [ ] @xmath34 ( 7,25)(-1,0)10 ( 42,25)(1,0)10 ( 25,25)(0,0)[]cycle @xmath35 , dur .
@xmath19 ( 62,25)(-1,0)10 ( 97,25)(1,0)10 ( 77.5,25)(0,0)[]cycle @xmath36 , dur .
@xmath19 ( 117,25)(-1,0)10 ( 152,25)(1,0)10 ( 135.25,25)(0,0)[]cycle @xmath37 , dur .
@xmath19 for circuit traffic , we consider @xmath38 classes of circuits with bandwidths @xmath39 .
we denote @xmath9 [ requests / s ] for the aggregate poisson process arrival rate of circuit requests .
a given circuit request is for a circuit of class @xmath40 , with probability @xmath10 .
we denote the mean circuit bit rate of the offered circuit traffic by @xmath41 .
we model the circuit holding time ( duration ) as an exponential random variable with mean @xmath13 .
we denote the resulting offered circuit traffic intensity ( load ) by @xmath42 . for packet traffic ,
we denote @xmath15 and @xmath16 for the mean and the variance of the packet size [ in bit ] , respectively .
we denote @xmath18 for the aggregate poisson process arrival rate [ packets / s ] of packet traffic across the @xmath5 onus and denote @xmath43 for the packet traffic intensity ( load ) . throughout
, we define the packet sizes and circuit bit rates to include the per - packet overheads , such as the preamble for ethernet frames and the interpacket gap , as well as the packet overheads when packetizing circuit traffic for transmission . for circuit traffic , we consider the blocking probability @xmath44 , i.e. , the probability that a request for a class @xmath11 circuit is blocked , i.e. , can not be accommodated within the transmission rate limit for circuit service @xmath2 .
we define the average circuit blocking probability as @xmath45 . for packet traffic ,
we consider the mean packet delay @xmath27 defined as the time period from the instant of packet arrival at the onu to the instant of complete delivery of the packet to the olt .
in order to provide circuit traffic with consistent upstream transmission service with a fixed circuit bandwidth , dycappon employs a polling cycle with a fixed duration @xmath19 [ s ] . an active circuit with bandwidth @xmath46 is allocated an upstream transmission window of duration @xmath47 in every cycle .
thus , by transmitting at the full upstream channel bit rate @xmath3 for duration @xmath48 once per cycle of duration @xmath19 , the circuit experiences a transmission bit rate ( averaged over the cycle duration ) of @xmath46 .
we let @xmath31 denote the aggregate of the upstream transmission windows of all active circuits in the pon in cycle @xmath36 , and refer to @xmath31 as the circuit partition duration .
we refer to the remaining duration @xmath49 as the packet partition of cycle @xmath36 . as illustrated in fig .
[ fig : cycle ] , a given cycle @xmath36 consists of the circuit partition followed by the packet partition . during the packet partition of each cycle
, each onu sends a report message to the olt .
the report message signals new circuit requests as well as the occupancy level ( queue depth ) of the packet service queue in the onu to the olt .
the signaling information for the circuit requests , i.e. , requested circuit bandwidth and duration , can be carried in the report message of the mpcp protocol in epons with similar modifications as used for signaling information for operation on multiple wavelength channels @xcite .
specifically , for signaling dynamic circuit requests , an onu report in the packet partition of cycle @xmath35 carries circuit requests generated since the onu s preceding report in cycle @xmath50 .
the report reaches the olt by the end of cycle @xmath35 and the olt executes circuit admission control as described in section [ ac : sec ] .
the onu is informed about the outcome of the admission control ( circuit is admitted or blocked ) in the gate message that is transmitted on the downstream wavelength channel at the beginning of cycle @xmath36 . in the dycappon design ,
the gate message propagates downstream while the upstream circuit transmissions of cycle @xmath36 are propagating upstream .
thus , if the circuit was admitted , the onu commences the circuit transmission with the circuit partition of cycle @xmath37 . for signaling packet traffic , the onu report in the packet partition of cycle @xmath35 carries the current queue depth as of the report generation instant .
based on this queue depth , the olt determines the effective bandwidth request and bandwidth allocation as described in section [ dba : sec ] .
the gate message transmitted downstream at the beginning of cycle @xmath36 informs the onu about its upstream transmission window in the packet partition of cycle @xmath36 . as illustrated in fig .
[ fig : cycle ] , in the dycappon design , the circuit partition is positioned at the beginning of the cycle , in an effort to mitigate the idle time between the end of the packet transmissions in the preceding cycle and the beginning of the packet transmissions of the current cycle .
in particular , when the last packet transmission of cycle @xmath35 arrives at the olt at the end of cycle @xmath35 , the first packet transmission of cycle @xmath36 can arrive at the olt at the very earliest one roundtrip propagation delay ( plus typically negligible processing time and gate transmission time ) after the beginning of cycle @xmath36 .
if the circuit partition duration @xmath31 is longer than the roundtrip propagation delay @xmath51 , then idle time between packet partitions is avoided . on the other hand , if @xmath52 , then there an idle channel period of duration @xmath53 between the end of the circuit partition and the beginning of the packet partition in cycle @xmath36 .
( 160,32 ) ( 0,20)(1,0)160 ( 0,30)(1,0)160 ( 0,5)(0,1)25 ( 160,5)(0,1)25 ( 80,10)(0,1)20 ( 8,20)(0,1)10 ( 12,20)(0,1)10 ( 30,20)(0,1)10 ( 34,20)(0,1)10 ( 76,20)(0,1)10 ( 80,20)(0,1)10 ( 92,20)(0,1)10 ( 96,20)(0,1)10 ( 110,20)(0,1)10 ( 114,20)(0,1)10 ( 142,20)(0,1)10 ( 156,20)(0,1)10 ( 4,25)(0,0 ) [ ] @xmath54 ( 9.5,25)(0,0 ) [ ] @xmath55 ( 21,25)(0,0 ) [ ] @xmath56 ( 31.5,25)(0,0 ) [ ] @xmath55 ( 50,25)(0,0 ) [ ] @xmath57 ( 77.5,25)(0,0 ) [ ] @xmath55 ( 85,25)(0,0 ) [ ] @xmath58 ( 93.5,25)(0,0 ) [ ] @xmath55 ( 102,25)(0,0 ) [ ] @xmath59 ( 111.5,25)(0,0 ) [ ] @xmath55 ( 126,25)(0,0 ) [ ] @xmath60 ( 148,25)(0,0 ) [ ] @xmath61 ( 157.5,25)(0,0 ) [ ] @xmath55 ( 20,15)(-1,0)20 ( 60,15)(1,0)20 ( 40,16)(0,0)[]@xmath62 ( 40,11)(0,0)[]circuit partition ( 100,15)(-1,0)20 ( 140,15)(1,0)20 ( 120,15)(0,0)[]packet partition ( 60,5)(-1,0)60 ( 100,5)(1,0)60 ( 80,5)(0,0)[]cycle duration @xmath19 note that this dycappon design trades off lower responsiveness to circuit requests for the masking of the roundtrip propagation delay
. specifically , when an onu signals a dynamic circuit request in the report message in cycle @xmath35 , it can at the earliest transmit circuit traffic in cycle @xmath37 .
on the other hand , packet traffic signaled in the report message in cycle @xmath35 can be transmitted in the next cycle , i.e. , cycle @xmath36 .
[ fig : cycle_det ] illustrates the structure of a given cycle in more detail , including the overheads for the upstream transmissions .
each onu that has an active circuit in the cycle requires one guard time of duration @xmath55 in the circuit partition .
thus , with @xmath63 denoting the number of onus with active circuits in the cycle , the duration of the circuit partition is @xmath64 . in the packet partition ,
each of the @xmath5 onus transmits at least a report message plus possibly some data upstream , resulting in an overhead of @xmath65 .
thus , the overhead per cycle is @xmath66 the resulting aggregate limit of the transmission windows for packets in cycle @xmath36 is @xmath67 if there is little packet traffic , the circuit partition @xmath31 and the immediately following packet transmission phase denoted p1 in fig .
[ fig : cyclell ] may leave significant portions of the fixed - duration cycle idle .
in such low - packet - traffic cycles , the olt can launch additional polling rounds denoted p2 , p3 , and p4 in fig .
[ fig : cyclell ] to serve newly arrived packets with low delay . specifically ,
if all granted packet upstream transmissions have arrived at the olt and there is more than @xmath68 time remaining until the end of the cycle ( i.e. , the beginning of the arrival of the next circuit partition @xmath69 ) at the olt , then the olt can launch another polling round . ( 160,32 ) ( 0,0)(1,0)140 ( 0,20)(1,0)140 ( 0,21)(0,0)[b]olt ( 0,-1)(0,0)[t]onu ( 8,20 ) ( 1,-3)6.6 ( 3,0)(1,3)6.6 ( 17,0)(1,3)6.6 ( 38,20)(1,-3)6.6 ( 31,0)(1,3)6.6 ( 45,0)(1,3)6.6 ( 67,20)(1,-3)6.6 ( 60,0)(1,3)6.6 ( 74,0)(1,3)6.6 ( 97,20)(1,-3)6.6 ( 90,0)(1,3)6.6 ( 104,0)(1,3)6.6 ( 119,0)(1,3)6.6 ( 133,0)(1,3)6.6 ( 10,0.75)(0,0)[b]@xmath31 ( 143,0.75)(0,0)[b]@xmath70 ( 26,0.75)(0,0)[b]@xmath71 ( 53,0.75)(0,0)[b]@xmath72 ( 82,0.75)(0,0)[b]@xmath73 ( 112,0.75)(0,0)[b]@xmath74 ( 126,0.75)(0,0)[b]idle ( 10,20)(0,1)10 ( 140,20)(0,1)10 ( 3,11)(0,0 ) [ ] @xmath34 ( 31,11)(0,0 ) [ ] @xmath34 ( 60,11)(0,0 ) [ ] @xmath34 ( 90,11)(0,0 ) [ ] @xmath34 ( 119,11)(0,0 ) [ ] @xmath34 ( 14,11)(0,0 ) [ ] @xmath29 ( 43,11)(0,0 ) [ ] @xmath29 ( 72,11)(0,0 ) [ ] @xmath29 ( 102,11)(0,0 ) [ ] @xmath29 ( 50,25)(-1,0)40 ( 100,25)(1,0)40 ( 76,25)(0,0)[]cycle @xmath36 , fixed duration @xmath19 for each circuit class @xmath40 , the olt tracks the number @xmath75 of currently active circuits , i.e. , the olt tracks the state vector @xmath76 representing the numbers of active circuits . taking the inner product of @xmath77 with the vector @xmath78 representing the bit rates of the circuit classes gives the currently required aggregate circuit bandwidth @xmath79 which corresponds to the circuit partition duration @xmath80 for a given limit @xmath81 , of bandwidth available for circuit service , we let @xmath82 denote the state space of the stochastic knapsack model @xcite of the dynamic circuits , i.e. , @xmath83 where @xmath84 is the set of non - negative integers . for an incoming onu request for a circuit of class @xmath11 ,
we let @xmath85 denote the subset of the state space @xmath82 that can accommodate the circuit request , i.e. , has at least spare bandwidth @xmath86 before reaching the circuit bandwidth limit @xmath2 .
formally , @xmath87 thus , if presently @xmath88 , then the new class @xmath11 circuit can be admitted ; otherwise , the class @xmath11 circuit request must be rejected ( blocked ) . with the offline scheduling approach @xcite of dycappon
, the reported packet queue occupancy corresponds to the duration of the upstream packet transmission windows @xmath89 , requested by onu @xmath90 . based on these requests , and the available aggregate packet upstream transmission window @xmath91 ( [ gp : eqn ] )
, the olt allocates upstream packet transmission windows with durations @xmath92 , to the individual onus .
the problem of fairly allocating bandwidth so as to enforce a maximum cycle duration has been extensively studied for the limited grant sizing approach @xcite , which we adapt as follows .
we set the packet grant limit for cycle @xmath36 to @xmath93 if an onu requests less than the maximum packet grant duration @xmath94 , it is granted its full request and the excess bandwidth ( i.e. , difference between @xmath94 and allocated grant ) is collected by an excess bandwidth distribution mechanism . if an onu requests a grant duration longer than @xmath94 , it is allocated this maximum grant duration , plus a portion of the excess bandwidth according to the equitable distribution approach with a controlled excess allocation bound @xcite . with the limited grant sizing approach , there is commonly an unused slot remainder of the grant allocation to onus @xcite due to the next queued packet not fitting into the remaining granted transmission window .
we model this unused slot remainder by half of the average packet size @xmath15 for each of the @xmath5 onus .
thus , the total mean unused transmission window duration in a given cycle is @xmath95
in this section , we employ techniques from the analysis of stochastic knapsacks @xcite to evaluate the blocking probabilities @xmath26 of the circuit class .
we also evaluate the mean duration of the circuit partition @xmath20 , which governs the mean available packet partition duration @xmath91 , which in turn is a key parameter for the evaluation of the mean packet delay in section [ delan : sec ] .
the stochastic knapsack model @xcite is a generalization of the well - known erlang loss system model to circuits with heterogeneous bandwidths . in brief , in the stochastic knapsack model , objects of different classes ( sizes ) arrive to a knapsack of fixed capacity ( size ) according to a stochastic arrival process .
if a newly arriving object fits into the currently vacant knapsack space , it is admitted to the knapsack and remains in the knapsack for some random holding time .
after the expiration of the holding time , the object leaves the knapsack and frees up the knapsack space that it occupied . if the size of a newly arriving object exceeds the currently vacant knapsack space , the object is blocked from entering the knapsack , and is considered dropped ( lost ) .
we model the prescribed limit @xmath2 on the bandwidth available for circuit service as the knapsack capacity .
the requests for circuits of bandwidth @xmath96 , arriving according to a poisson process with rate @xmath97 are modeled as the objects seeking entry into the knapsack .
an admitted circuit of class @xmath11 occupies the bandwidth ( knapsack space ) @xmath86 for an exponentially distributed holding time with mean @xmath13 .
we denote @xmath98 for the set of states @xmath77 that occupy an aggregate bandwidth @xmath99 , i.e. , @xmath100 let @xmath24 denote the equilibrium probability of the currently active circuits occupying an aggregate bandwidth of @xmath25 . through the recursive kaufman - roberts algorithm @xcite , which is given in the appendix ,
the equilibrium probabilities @xmath24 can be computed with a time complexity of @xmath101 and a memory complexity of @xmath102 . the blocking probability @xmath44 is obtained by summing the equilibrium probabilities @xmath24 of the sets of states that have less than @xmath86 available circuit bandwidth , i.e. , @xmath103 we define the average circuit blocking probability @xmath104 the performance evaluation for packet delay in section [ pkt_perf : sec ] requires taking expectations over the distribution @xmath24 of the aggregate bandwidth @xmath25 occupied by circuits . in preparation for these packet evaluations
, we define @xmath105 $ ] to denote the expectation of a function @xmath106 of the random variable @xmath25 over the distribution @xmath24 , i.e. , we define @xmath107 = \sum_{\beta = 0}^{c_c } f(\beta ) q(\beta).\end{aligned}\ ] ] with this definition , the mean aggregate bandwidth of the active circuits is obtained as @xmath108 = \sum_{\beta = 0}^{c_c } \beta q(\beta).\end{aligned}\ ] ] note that by taking the expectation of ( [ xin : eqn ] ) , the corresponding mean duration of the circuit partition is @xmath109 = \bar{\beta } \gamma /c$ ] . in this section
we analyze the delay and delay variations experienced by circuit traffic as it traverses a dycappon network from onu to olt .
initially we ignore delay variations , i.e. , we consider that a given circuit with bit rate @xmath46 has a fixed position for the transmission of its @xmath110 bits in each cycle .
three delay components arise : the `` accumulation / dispersal '' delay of @xmath19 for the @xmath110 bits of circuit traffic that are transmitted per cycle .
note that the first bit arriving to form a `` chunk '' of @xmath110 bits experiences the delay @xmath19 at the onu , waiting for subsequent bits to `` fill up ( accumulate ) '' the chunk . the last bit of a chunk experiences essentially no delay at the onu , but has to wait for a duration of @xmath19 at the olt to `` send out ( disperse ) '' the chunk at the circuit bit rate @xmath46 .
the other delay components are the transmission delay of @xmath47 and the propagation delay @xmath6 .
thus , the total delay is @xmath111 circuit traffic does not experience delay variations ( jitter ) in dycappon as long as the positions ( in time ) of the circuit transmissions in the cycle are held fixed .
when an ongoing circuit is closing down or a new circuit is established , it may become necessary to rearrange the transmission positions of the circuits in the cycle in order to keep all circuit transmissions within the circuit partition at the beginning of the cycle and avoid idle times during the circuit partition .
adaptations of packing algorithms @xcite could be employed to minimize the shifts in transmission positions .
note that for a given circuit service limit @xmath2 , the worst - case delay variation for a given circuit with rate @xmath46 is less than @xmath112 as the circuit could at the most shift from the beginning to the end of the circuit partition of maximum duration @xmath113 . inserting the circuit partition duration @xmath20 from ( [ xin : eqn ] ) into the expression for the aggregate limit @xmath91 on the transmission window for packets in a cycle from ( [ gp : eqn ] ) and taking the expectation @xmath114
$ ] with respect to the distribution of the aggregate circuit bandwidth @xmath25 , we obtain @xmath115 - \omega_o.\end{aligned}\ ] ] considering the unused slot remainder @xmath116 ( [ omegau : eqn ] ) , the mean portion of a cycle available for upstream packet traffic transmissions is limited to @xmath117 - \frac{\omega_o + \omega_u}{\gamma}.\end{aligned}\ ] ] that is , the packet traffic intensity @xmath118 must be less than @xmath119 for stability of the packet service , i.e. , for finite packet delays . in this section ,
we present for stable packet service an approximate analysis of the mean delay @xmath27 of packets transmitted during the packet partition . in dycappon ,
packets are transmitted on the bandwidth that is presently not occupied by admitted circuits .
thus , fluctuations in the aggregate occupied circuit bandwidth @xmath25 affect the packet delays . if the circuit bandwidth @xmath25 is presently high , packets experience longer delays than for presently low circuit bandwidth @xmath25 .
the aggregated occupied circuit bandwidth @xmath25 fluctuates as circuits are newly admitted and occupy bandwidth and as existing circuits reach the end of their holding time and release their occupied bandwidth .
the time scale of these fluctuations of @xmath25 increases as the average circuit holding time @xmath13 increases , i.e. , as the circuit departure rate @xmath120 decreases ( and correspondingly , the circuit request arrival rate @xmath121 decreases for a given fixed circuit traffic load @xmath122 ) @xcite . for circuit holding times that are orders of magnitude larger than the typically packet delays ( service times ) in the system , the fluctuations of the circuit bandwidth @xmath25 occur at a significantly longer ( slower ) time scale than the packet service time scale .
that is , the bandwidth @xmath25 occupied by circuits exhibits significant correlations over time which in turn give rise to complex correlations with the packet queueing delay @xcite .
for instance , packets arriving during a long period of high circuit bandwidth may experience very long queueing delays and are possibly only served after some circuits release their bandwidth . as illustrated in section [ mu_impact : sec ] , the effects of these complex correlations become significant for scenarios with moderate to long circuit holding times @xmath13 when the circuit traffic load is low to moderate relative to the circuit bandwidth limit @xmath2 ( so that pronounced circuit bandwidth fluctuations are possible ) , and the packet traffic load on the remaining bandwidth of approximately @xmath123 is relatively high , so that substantial packet queue build - up can occur .
we leave a detailed mathematical analysis of the complex correlations occurring in these scenarios in the context of dycappon for future research . in the present study ,
we focus on an approximate packet delay analysis that neglects the outlined correlations .
we base our approximate packet delay analysis on the expectation @xmath124 $ ] ( [ ebeta : eqn ] ) , i.e. , we linearly weigh packet delay metrics @xmath125 with the probability masses @xmath24 for the aggregate circuit bandwidth @xmath25 .
we also neglect the `` low - load '' operating mode of section [ lowload : sec ] in the analysis .
in the proposed dycappon cycle structure , a packet experiences five main components , namely @xmath0 the reporting delay from the generation instant of the packet to the transmission of the report message informing the olt about the packet , which for the fixed cycle duration of dycappon equals half the cycle duration , i.e. , @xmath126 , @xmath1 the report - to - packet partition delay @xmath127 from the instant of report transmission to the beginning of the packet partition in the next cycle , @xmath128 the queuing delay @xmath129 from the reception instant of the grant message to the beginning of the transmission of the packet , as well as @xmath130 the packet transmission delay with mean @xmath131 , and @xmath132 the upstream propagation delay @xmath6 . in the report - to - packet partition delay
we include a delay component of half the mean duration of the packet partition @xmath133 to account for the delay of the reporting of a particular onu to the end of the packet partition .
the delay from the end of the packet partition in one cycle to the beginning of the packet partition of the next cycle is the maximum of the roundtrip propagation delay @xmath51 and the mean duration of the circuit partition @xmath20 .
thus , we obtain overall for the report - to - packet partition delay @xmath134\\ & = & \frac{1}{2 } \left ( \gamma + \e_{\beta } \left[\max \left\ { 2 \tau,\ \frac{\beta \gamma}{c } \right\ } \right ] - \omega_o \right).\end{aligned}\ ] ] we model the queueing delay with an m / g/1 queue . generally , for messages with mean service time @xmath135 , normalized message size variance @xmath136 , and traffic intensity @xmath137 , the m / g/1 queue has expected queueing delay @xcite @xmath138 for dycappon , we model the aggregate packet traffic from all @xmath5 onus as feeding into one m / g/1 queue with mean packet size @xmath15 and packet size variance @xmath139 .
we model the circuit partitions , when the upstream channel is not serving packet traffic , through scaling of the packet traffic intensity .
in particular , the upstream channel is available for serving packet traffic only for the mean fraction @xmath140 of a cycle .
thus , for large backlogs served across several cycles , the packet traffic intensity during the packet partition is effectively @xmath141 hence , the mean queueing delay is approximately @xmath142 thus , the overall mean packet delay is approximately @xmath143
we consider an epon with @xmath144 onus , a channel bit rate @xmath145 gb / s , and a cycle duration @xmath146 ms .
each onu has abundant buffer space and a one - way propagation delay of @xmath147s to the olt .
the guard time is @xmath148s and the report message has 64 bytes .
we consider @xmath149 classes of circuits as specified in table [ cir : tab ] .
.circuit bandwidths @xmath86 and request probabilities @xmath10 for @xmath150 classes of circuits in performance evaluations . [ cols= " < , > , > , > " , ] for the packet traffic , we observe from table [ fig : chi ] a very slight increase in the mean packet delays @xmath27 as the circuit traffic load @xmath122 increases .
this is mainly because the transmission rate limit @xmath2 for circuit service bounds the upstream transmission bandwidth the circuits can occupy to no more than @xmath2 in each cycle .
as the circuit traffic load @xmath122 increases , the circuit traffic utilizes this transmission rate limit @xmath2 more and more fully . however , the packet traffic is guaranteed a portion @xmath151 of the upstream transmission bandwidth . formally , as the circuit traffic load @xmath122 grows large ( @xmath152 )
, the mean aggregate circuit bandwidth @xmath153 approaches the limit @xmath2 , resulting in a lower bound for the packet traffic load limit ( [ pimax : eqn ] ) of @xmath154 and corresponding upper bounds for the effective packet traffic intensity @xmath155 and the mean packet delay @xmath27 . in fig .
[ fig : cc ] we examine the impact of the transmission rate limit @xmath2 for circuit traffic .
we consider different compositions @xmath156 of the total traffic load @xmath157 .
we observe from fig .
[ fig : cc](a ) that the average circuit blocking probability @xmath158 steadily decreases for increasing @xmath2 . in the example in fig .
[ fig : cc ] , the average circuit blocking probability @xmath158 drops to negligible values below 1 % for @xmath2 values corresponding to roughly twice the offered circuit traffic load @xmath122 .
for instance , for circuit load @xmath159 , @xmath158 drops to 0.9 % for @xmath160 gb / s .
the limit @xmath2 thus provides an effective parameter for controlling the circuit blocking probability experienced by customers . from fig .
[ fig : cc](b ) , we observe that the mean packet delay abruptly increases when the @xmath2 limit reduces the packet traffic portion @xmath151 of the upstream transmission bandwidth to values near the packet traffic intensity @xmath118 .
we also observe from fig .
[ fig : cc](b ) that the approximate packet delay analysis is quite accurate for small to moderate @xmath2 values ( the slight delay overestimation is due to neglecting the low packet traffic polling ) , but underestimates the packet delays for large @xmath2 .
large circuit traffic limits @xmath2 give the circuit traffic more flexibility for causing fluctuations of the occupied circuit bandwidth , which deteriorate the packet service . summarizing , we see from fig . [
fig : cc](b ) that as the effective packet traffic intensity @xmath161 approaches one , the mean packet delay increases sharply .
thus , for ensuring low - delay packet service , the limit @xmath2 should be kept sufficiently below @xmath162 .
when offering circuit and packet service over shared pon upstream transmission bandwidth , network service providers need to trade off the circuit blocking probabilities and packet delays . as we observe from fig .
[ fig : cc ] , the circuit bandwidth limit @xmath2 provides an effective tuning knob for controlling this trade - off . as a function of packet traffic load @xmath118 . ] the fig .
[ fig : lm ] we examine the impact of low - packet - traffic mode polling from section [ lowload : sec ] on the mean packet delay @xmath27 .
we observe from fig .
[ fig : lm ] that low - packet - traffic mode polling substantially reduces the mean packet delay compared to conventional polling for low packet traffic loads .
this delay reduction is achieved by the the more frequent polling which serves packets quicker in cycles with low load due to circuit traffic .
we have proposed and evaluated dycappon , a passive optical network that provides dynamic circuit and packet service .
dycappon is based on fixed duration cycles , ensuring consistent circuit service , that is completely unaffected by the packet traffic load .
dycappon masks the round - trip propagation delay for polling of the packet traffic queues in the onus with the upstream circuit traffic transmissions , providing for efficient usage of the upstream bandwidth .
we have analyzed the circuit level performance , including the circuit blocking probability and delay experienced by circuit traffic in dycappon , as well as the bandwidth available for packet traffic after serving the circuit traffic .
we have also conducted an approximate analysis of the packet level performance . through extensive numerical investigations based on the analytical performance characterization of dycappon as well as verifying simulations
, we have demonstrated the circuit and packet traffic performance and trade - offs in dycappon .
the provided analytical performance characterizations as well as the identified performance trade - offs provide tools and guidance for dimensioning and operating pon access networks that provide a mix of circuit and packet oriented service .
there are several promising directions for future research on access networks that flexibly provide both circuit and packet service .
one important future research direction is to broadly examine cycle - time structures and wavelength assignments in pons providing circuit and packet service . in particular , the present study focused on a single upstream wavelength channel operated with a fixed polling cycle duration .
future research should examine the trade - offs arising from operating multiple upstream wavelength channels and combinations of fixed- or variable - duration polling cycles .
an exciting future research direction is to extend the pon service further toward the individual user , e.g. , by providing circuit and packet service on integrated pon and wireless access networks , such as @xcite , that reach individual mobile users or wireless sensor networks @xcite .
further , exploring combined circuit and packet service in long - reach pons with very long round trip propagation delays , which may require special protocol mechanisms , see e.g. , @xcite , is an open research direction .
another direction is to examine the integration and interoperation of circuit and packet service in the pon access network with metropolitan area networks @xcite and wide area networks to provide circuit and packet service @xcite . in this appendix , we present the recursive kaufman - roberts algorithm @xcite for computing the equilibrium probabilities @xmath163 that the currently active circuit occupy an aggregated bandwidth @xmath25 . for
the execution of the algorithm , the given circuit bandwidths @xmath164 and limit @xmath2 are suitably normalized so that incrementing @xmath25 in integer steps covers all possible combinations of the circuit bandwidth .
for instance , in the evaluation scenario considered in section [ eval_setup : sec ] , all circuit bandwidth are integer multiples of 52 mb / s .
thus , we normalize all bandwidths by 52 mb / s and for e.g. , @xmath160 gb / s execute the following algorithm for @xmath165 .
( the variables @xmath166 , and @xmath25 refer to their normalized values , e.g. , @xmath167 for the @xmath160 gb / s example , in the algorithm below ) .
the algorithm first evaluates unnormalized occupancy probabilities @xmath168 that relate to a product - form solution of the stochastic knapsack @xcite .
subsequently the normalization term @xmath29 for the occupancy probabilities is evaluated , allowing then the evaluation of the actual occupancy probabilities @xmath24 .
m. batayneh , d. schupke , m. hoffmann , a. kirstaedter , and b. mukherjee , `` link - rate assignment in a wdm optical mesh network with differential link capacities : a network - engineering approach , '' in _ proc .
honet _ , nov .
2008 , pp . 216219 .
l. qiao and p. koutsakis , `` adaptive bandwidth reservation and scheduling for efficient wireless telemedicine traffic transmission , '' _ ieee trans .
vehicular technology _
60 , no . 2 ,
632643 , feb .
m. reisslein , j. lassetter , s. ratnam , o. lotfallah , f. fitzek , and s. panchanathan , `` traffic and quality characterization of scalable encoded video : a large - scale trace - based study , part 1 : overview and definitions , '' arizona state univ . ,
tech . rep . , 2002 .
k. shuaib , f. sallabi , and l. zhang , `` smoothing and modeling of video transmission rates over a qos network with limited bandwidth connections , '' _ int .
journal of computer networks and communications _
, vol . 3 , no . 3 , pp .
148162 , may 2011 .
g. van der auwera and m. reisslein , `` implications of smoothing on statistical multiplexing of h. 264/avc and svc video streams , '' _ ieee trans . on broadcasting _ , vol .
55 , no . 3 , pp .
541558 , sep .
2009 .
n. charbonneau , a. gadkar , b. h. ramaprasad , and v. vokkarane , `` dynamic circuit provisioning in all - optical wdm networks using lightpath switching , '' _ opt .
_ , vol . 9 , no . 2 , pp . 179 190 , 2012
a. munir , s. tanwir , and s. zaidi , `` requests provisioning algorithms for dynamic optical circuit switched ( docs ) networks : a survey , '' in _ proc .
multitopic conference ( inmic ) _ , dec .
2009 , pp .
r. skoog , g. clapp , j. gannett , a. neidhardt , a. von lehman , and b. wilson , `` architectures , protocols and design for highly dynamic optical networks , '' _ opt .
_ , vol . 9 , no . 3 , pp . 240251 , 2012 .
e. van breusegem , j. cheyns , d. de winter , d. colle , m. pickavet , p. demeester , and j. moreau , `` a broad view on overspill routing in optical networks : a real synthesis of packet and circuit switching ? '' _ optical switching and networking _ , vol . 1 , no . 1 , pp .
5164 , 2005 .
m. mcgarry , m. reisslein , f. aurzada , and m. scheutzow , `` shortest propagation delay ( spd ) first scheduling for epons with heterogeneous propagation delays , '' _
ieee j. on selected areas in commun .
_ , vol .
28 , no . 6 , pp .
849862 , aug .
a. sivakumar , g. sankaran , and k. sivalingam , `` performance analysis of onu - wavelength grouping schemes for efficient scheduling in long reach - pons , '' _ opt .
switching netw .
_ , vol . 10 , no . 4 ,
pp . 465474 , 2013 .
i. tomkos , l. kazovsky , and k .-
kitayama , `` next - generation optical access networks : dynamic bandwidth allocation , resource use optimization , and qos improvements , '' _ ieee netw . _ ,
26 , no . 2 ,
pp . 46 , 2012 .
f. zanini , l. valcarenghi , d. p. van , m. chincoli , and p. castoldi , `` introducing cognition in tdm pons with cooperative cyclic sleep through runtime sleep time determination , '' _ opt .
switching netw .
, in print _ , 2013 .
f. aurzada , m. scheutzow , m. herzog , m. maier , and m. reisslein , `` delay analysis of ethernet passive optical networks with gated service , '' _ osa journal of optical networking _ , vol . 7 , no . 1 , pp .
2541 , jan . 2008 .
f. aurzada , m. scheutzow , m. reisslein , n. ghazisaidi , and m. maier , `` capacity and delay analysis of next - generation passive optical networks ( ng - pons ) , '' _ ieee trans . on communications _ ,
59 , no . 5 , pp .
13781388 , may 2011 .
j. angelopoulos , h .- c .
leligou , t. argyriou , s. zontos , e. ringoot , and t. van caenegem , `` efficient transport of packets with qos in an fsan - aligned gpon , '' _ ieee comm . mag .
_ , vol .
42 , no . 2 ,
9298 , feb .
c. assi , y. ye , s. dixit , and m. ali , `` dynamic bandwidth allocation for quality - of - service over ethernet pons , '' _ ieee journal on selected areas in communications _
21 , no . 9 ,
14671477 , nov .
2003 .
y. luo and n. ansari , `` limited sharing with traffic prediction for dynamic bandwidth allocation and qos provioning over epons , '' _ osa journal of optical networking _ , vol . 4 , no . 9 , pp . 561572 , sep . 2005 .
m. radivojevic and p. matavulj , `` implementation of intra - onu scheduling for quality of service support in ethernet passive optical networks , '' _ ieee / osa j. lightw .
_ , vol . 27 , no . 18 , pp . 40554062 , sep . 2009
s. sherif , a. hadjiantonis , g. ellinas , c. assi , and m. ali , `` a novel decentralized ethernet - based pon access architecture for provisioning differentiated qos , '' _ ieee / osa j. of lightwave technology _ , vol .
22 , no . 11 , pp . 24832497 , nov .
a. shami , x. bai , n. ghani , c. assi , and h. mouftah , `` qos control schemes for two - stage ethernet passive optical access networks , '' _ ieee j. on sel .
areas in commun .
_ , vol .
23 , no . 8 , pp . 14671478 , nov .
m. vahabzadeh and a. ghaffarpour rahbar , `` modified smallest available report first : new dynamic bandwidth allocation schemes in qos - capable epons , '' _ optical fiber technolgy _ ,
17 , no . 1 , pp .
716 , jan .
t. berisa , a. bazant , and v. mikac , `` bandwidth and delay guaranteed polling with adaptive cycle time ( bdgpact ) : a scheme for providing bandwidth and delay guarantees in passive optical networks , '' _
j. of opt .
_ , vol . 8 , no . 4 , pp .
337345 , 2009 .
y. qin , d. xue , l. zhao , c. k. siew , and h. he , `` a novel approach for supporting deterministic quality - of - service in wdm epon networks , '' _ optical switching and networking _ , vol .
10 , no . 4 , pp .
378392 , 2013 .
l. zhang , e .- s . an , h .-
yeo , and s. yang , `` dual deb - gps scheduler for delay - constraint applications in ethernet passive optical networks , '' _ ieice trans .
_ , vol .
86 , no . 5 , pp . 15751584 , 2003 . f. an , y. hsueh , k. kim , i. white , and l. kazovsky , `` a new dynamic bandwidth allocation protocol with quality of service in ethernet - based passive optical networks , '' in _ proc .
iasted woc _ ,
vol . 3 , jul .
2003 , pp . 165169 .
i. hwang , j. lee , k. lai , and a. liem , `` generic qos - aware interleaved dynamic bandwidth allocation in scalable epons , '' _ ieee / osa j. of optical commun . and
_ , vol . 4 , no . 2 ,
99107 , feb .
2012 .
n. merayo , t. jimenez , p. fernandez , r. duran , r. lorenzo , i. de miguel , and e. abril , `` a bandwidth assignment polling algorithm to enhance the efficiency in qos long - reach epons , '' _ eu .
_ , vol . 22 , no . 1 ,
3544 , jan . 2011 .
s. de , v. singh , h. m. gupta , n. saxena , and a. roy , `` a new predictive dynamic priority scheduling in ethernet passive optical networks ( epons ) , '' _ opt . switch .
_ , vol . 7 , no . 4 , pp .
215223 , 2010 .
w. kwong , g .- c .
yang , and j .-
g . zhang , `` @xmath177 prime - sequence codes and coding architecture for optical code - division multiple - access , '' _ ieee transactions on communications _ , vol .
44 , no . 9 , pp . 11521162 , 1996 .
m. mcgarry , m. reisslein , and m. maier , `` ethernet passive optical network architectures and dynamic bandwidth allocation algorithms , '' _ ieee communication surveys and tutorials _
10 , no . 3 , pp . 4660 ,
2008 . c. assi , y. ye , s. dixit , and m. ali , `` dynamic bandwidth allocation for quality - of - service over ethernet pons , '' _ ieee journal on selected areas in communications _ , vol .
21 , no . 9 ,
14671477 , nov .
g. kramer , b. mukherjee , s. dixit , y. ye , and r. hirth , `` supporting differentiated classes of service in ethernet passive optical networks , '' _ osa j. of optical netw .
_ , vol . 1 , no . 9 , pp .
280298 , aug .
2002 .
f. aurzada , m. levesque , m. maier , and m. reisslein , `` fiwi access networks based on next - generation pon and gigabit - class wlan technologies : a capacity and delay analysis , '' _ ieee / acm trans .
, in print _ , 2014 .
j. coimbra , g. schtz , and n. correia , `` a game - based algorithm for fair bandwidth allocation in fibre - wireless access networks , '' _ optical switching and networking _ , vol . 10 , no . 2 , pp . 149 162 , 2013 . a. dhaini , p .- h . ho , and x. jiang , `` qos control for guaranteed service bundles over fiber - wireless ( fiwi ) broadband access networks , '' _ ieee / osa j. lightwave techn
_ , vol . 29 , no . 10 , pp . 15001513 , 2011 .
m. maier , n. ghazisaidi , and m. reisslein , `` the audacity of fiber - wireless ( fiwi ) networks , '' in _ proc . of accessnets
lecture notes of the institute for computer sciences , social informatics and telecommunications engineering.1em plus 0.5em minus 0.4emspringer , 2009 , vol . 6 , pp .
n. moradpoor , g. parr , s. mcclean , and b. scotney , `` iidwba algorithm for integrated hybrid pon with wireless technologies for next generation broadband access networks , '' _ opt
. switching .
_ , vol . 10 , no . 4 , pp . 439457 , 2013 . a. seema and m. reisslein , `` towards efficient wireless video sensor networks : a survey of existing node architectures and proposal for a flexi - wvsnp design , '' _ ieee comm . surv . & tut . _ , vol . 13 , no . 3 , pp . 462486 , third quarter 2011 .
a. mercian , m. mcgarry , and m. reisslein , `` offline and online multi - thread polling in long - reach pons : a critical evaluation , '' _ ieee / osa j. lightwave techn .
_ , vol . 31 , no . 12 , pp . 20182228 , jun
. 2013 .
h. song , b. w. kim , and b. mukherjee , `` long - reach optical access networks : a survey of research challenges , demonstrations , and bandwidth assignment mechanisms , '' _ ieee commun .
_ , vol . 12 , no . 1 ,
pp . 112123 , 1st quarter 2010 .
m. maier , m. reisslein , and a. wolisz , `` a hybrid mac protocol for a metro wdm network using multiple free spectral ranges of an arrayed - waveguide grating , '' _ computer networks _
41 , no . 4 , pp . 407433 , mar .
m. scheutzow , m. maier , m. reisslein , and a. wolisz , `` wavelength reuse for efficient packet - switched transport in an awg - based metro wdm network , '' _ ieee / osa journal of lightwave technology _ , vol .
21 , no . 6 , pp . 14351455 , jun .
yang , m. maier , m. reisslein , and m. carlyle , `` a genetic algorithm - based methodology for optimizing multiservice convergence in a metro wdm network , '' _ ieee / osa journal of lightwave technology _ , vol .
21 , no . 5 , pp .
11141133 , may 2003 .
m. yuang , i .-
f . chao , and b. lo , `` hopsman : an experimental optical packet - switched metro wdm ring network with high - performance medium access control , '' _ ieee / osa journal of optical communications and networking _ , vol . 2 , no . 2 ,
91101 , feb . | dynamic circuits are well suited for applications that require predictable service with a constant bit rate for a prescribed period of time , such as cloud computing and e - science applications .
past research on upstream transmission in passive optical networks ( pons ) has mainly considered packet - switched traffic and has focused on optimizing packet - level performance metrics , such as reducing mean delay .
this study proposes and evaluates a dynamic circuit and packet pon ( dycappon ) that provides dynamic circuits along with packet - switched service .
dycappon provides @xmath0 flexible packet - switched service through dynamic bandwidth allocation in periodic polling cycles , and @xmath1 consistent circuit service by allocating each active circuit a fixed - duration upstream transmission window during each fixed - duration polling cycle .
we analyze circuit - level performance metrics , including the blocking probability of dynamic circuit requests in dycappon through a stochastic knapsack - based analysis . through this analysis
we also determine the bandwidth occupied by admitted circuits .
the remaining bandwidth is available for packet traffic and we conduct an approximate analysis of the resulting mean delay of packet traffic . through extensive numerical evaluations and verifying simulations we demonstrate the circuit blocking and packet delay trade - offs in dycappon . dynamic circuit switching ; ethernet passive optical network ; grant scheduling ; grant sizing ; packet delay ; stochastic knapsack . |
in the arena of cosmology and stellar structure , the importance of the perfect fluid source can not be understated .
the assumption of large - scale isotropy will demand , via the field equations , that the stress - energy tensor supporting the universe possess the algebraic structure of a perfect fluid .
as well , cosmologies containing scalar fields have long been considered .
the motivation being that the scalar field would represent some exotic component of matter , which could explain various puzzling phenomena presented by observational cosmology .
more recently , viable alternatives to the standard big bang theory have been put forward , the most promising of which involve an inflationary era at some early time . most of these theories invoke a scalar field to play the role of the inflaton in the early universe . at late times , the scalar field may again play an important role as a `` dark energy '' field .
it has been observed that the universe has recently entered an acceleration phase and some exotic dark energy must presently dominate @xcite , @xcite .
it is well known that a scalar field provides a simple model to explain this acceleration .
the scalar field may take many forms . aside from the traditional quintessence field , which arises from a lagrangian motivated by relativistic continuum mechanics
, there is also the recently popular `` tachyonic '' scalar field from string theory , which possesses a born - infeld type action .
both of these fields have been used extensively in cosmology as the inflaton , dark matter and dark - energy ( see @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite @xcite and references therein ) .
usual cosmological models are spherically symmetric
. however , more general cosmological models , admitting a three - parameter group of motions ( bianchi type i - ix ) have also been studied . for the above reasons
, we believe that it is important to study the combined system of einstein - scalar field - perfect fluid equations in detail .
specifically , in section 2 we present the system in its full generality for both the quintessence field and tachyon field .
we discuss the system of equations and the number of functions which can be apriori prescribed .
this is dictated by the number of functions present as well as the number of equations and identities .
the eigenvector and eigenvalue structure is briefly discussed as well as its relation to physical quantities such as energy density and pressure .
the cauchy problem has long been studied in general relativity and is of great use due to the difficulty involved in obtaining exact solutions to physical problems within the theory ( see @xcite , @xcite , as well as @xcite where the gravitating complex scalar - maxwell field system has been studied ) . in solving einstein s equations ,
one often considers physically acceptable initial data ( for example : data from plasma physics for stellar collapse , a specific form for the scalar potential inspired by string theory or particle physics in cosmology , etc . )
the future evolution of this prescribed data is to be determined , ideally by analytically solving the einstein equations and matter field equations of motion .
however , this is often not possible and either some simplifying assumptions must be made or else approximation methods must be utilized .
it is useful therefore to numerically evolve the initial data of the full , unsimplified model to determine the behaviour of the system at later times and , in some cases , to aid in determining what form the exact solution should take . a thorough treatment of the system consisting of a massless self - gravitating scalar field in the context of gravitational collapse ( in the retarded time gauge )
may be found in @xcite .
the cauchy problem in general relativity is non - trivial .
initial data is not freely prescribable due to the existence of four constraining equations for the initial data .
( these arise due to the differential identities among the system of field equations . )
the constraint equations for both the quintessence plus perfect fluid and tachyon plus perfect fluid systems are derived in this paper along with the functions which one , in principle , may prescribe . in section 3 the cauchy problem for the einstein - perfect fluid - scalar field
is properly posed using geodesic normal coordinates ( which are particularly well suited to cosmological studies ) .
a scheme is presented which will evolve the this data from the initial @xmath0 hyper - surface analytically into the bulk ( @xmath1 ) .
finally , we end with a simple example to illustrate the use of the scheme and its convergence properties . for an excellent review of the cauchy problem in general relativity ,
the reader is referred to @xcite .
the system considered consists of gravity coupled to a perfect fluid and quintessence scalar field , @xmath2 .
the fluid energy density , pressure and scalar potential are denoted by @xmath3 , @xmath4 and @xmath5 respectively .
derivatives with respect to the scalar field are denoted by a prime .
the quintessence lagrangian density has the form : @xmath6.\ ] ] the einstein and supplemental equations for this system are given by : @xmath7 \delta^{\mu}_{\;\nu } + \left(\rho + p \right ) u^{\mu}u_{\nu } + p\delta^{\mu}_{\;\nu } , \label{eq : tq } \\ & \mathcal{t}_{\nu}:=t^{\mu}_{\;\nu ; \mu}=0 , \label{eq : consq } \\ & \sigma:=\phi^{;\mu}_{\ ; ; \mu}-{v^{\prime}(\phi)}=0 , \label{eq : siq } \\ & \mathcal{u}:=u^{\alpha}u_{\alpha}+1=0 , \label{eq : uq } \\ & \mathcal{k}:=\left[(\rho + p)u^{\alpha}\right]_{;\alpha } -u^{\alpha}p_{,\alpha}=0 , \label{eq : kq } \\ & \mathcal{f}_{\nu}:=(\rho + p)u^{\alpha}u_{\nu;\alpha } + \left(\delta^{\alpha}_{\;\nu } + u^{\alpha}u_{\nu}\right)p_{,\alpha}=0 , \label{eq : fq } \\ & \mathcal{c}^{\nu}(g_{\alpha\beta } , g_{\alpha\beta,\gamma})=0 .
\label{eq : cq}\end{aligned}\ ] ] ( the @xmath8 denote four possible coordinate conditions . ) the algebraic and differential identities are : @xmath9 in this system of equations the number of unknown functions is @xmath10 whereas , from ( [ eq : einstq]-[eq : cq ] ) , there exist twenty - five equations
. however , there are nine identities ( [ eq : id1q]-[eq : id3q ] ) and therefore only sixteen independent equations yielding an underdetermined system .
we can therefore prescribe two functions out of the eighteen . in case
an equation of state is imposed , @xmath11^{2}+\left[\frac{\partial s(p,\rho)}{\partial p}\right]^{2}&>&0 , \nonumber\end{aligned}\ ] ] we can still prescribe one function .
we can now explore eigenvalues of @xmath12 .
assuming that @xmath13 we obtain from ( [ eq : tq]-[eq : uq ] ) and ( [ eq : eigassumpq ] ) that @xmath14u^{\alpha } , \label{eq : evec1q } \\
t^{\alpha}_{\;\beta}\phi^{;\beta}= & \left[p+ \frac{1}{2 } \phi^{;\beta}\phi_{,\beta } -{v(\phi)}\right]\phi^{;\alpha } + \left[\left(\rho + p \right)u_{\beta}\phi^{;\beta}\right ] u^{\alpha } , \label{eq : evec2q } \\ t^{\alpha}_{\;\beta}s^{\beta}=&\left[\phi_{,\beta}s^{\beta}\right]\phi^{;\alpha } -\left[\frac{1}{2 } \phi^{;\beta}\phi_{,\beta } + { v(\phi)}-p\right]s^{\alpha}. \label{eq : evec3q}\end{aligned}\ ] ] it is evident from the above equations that , in general , none of the vectors @xmath15 , @xmath16 , @xmath17 are eigenvectors .
however , consider two special cases . in case - i
, we take the scalar gradient perpendicular to the fluid velocity and @xmath17 : @xmath18 by ( [ eq : evec1q]-[eq : evec2q ] ) and ( [ eq : perpgradq ] ) , we get : @xmath19u^{\alpha } , \label{eq : class1vecsa } \\ t^{\alpha}_{\;\beta}\phi^{;\beta}= & \left[p + \frac{1}{2 } \phi^{;\beta}\phi_{,\beta } -{v(\phi)}\right]\phi^{;\alpha } , \label{eq : class1vecsb } \\ t^{\alpha}_{\;\beta}s^{\beta}=&-\left[-p + \frac{1}{2 } \phi^{;\beta}\phi_{,\beta } + { v(\phi)}\right]s^{\alpha}. \label{eq : class1vecsc}\end{aligned}\ ] ] in this case , @xmath15 , @xmath20 are eigenvectors and @xmath17 is a two - fold degenerate eigenvector .
the corresponding eigenvalues in ( [ eq : class1vecsa]-[eq : class1vecsc ] ) represent the proper mass density and two principal pressures , respectively , of the matter . a positive value of the potential , @xmath5 , make a positive contribution to the energy density and negative contribution to the pressures .
the structure here is algebraically similar to the _
anisotropic fluid_. in case - ii , assume that the scalar gradient is colinear with the fluid velocity : @xmath21 by equations ( [ eq : evec1q]-[eq : evec3q ] ) and ( [ eq : eigassumpq2 ] ) , we arrive at : @xmath22u^{\alpha } , \label{eq : class2vecsa } \\
t^{\alpha}_{\;\beta}\phi^{;\beta}= & -\left[\rho -\frac{1}{2 } \phi^{;\beta}\phi_{,\beta } + { v(\phi)}\right]\phi^{;\alpha } , \label{eq : class2vecsb } \\ t^{\alpha}_{\;\beta}s^{\beta}=&\left[p-\frac{1}{2 } \phi^{;\beta}\phi_{,\beta } -{v(\phi)}\right]s^{\alpha}. \label{eq : class2vecsc}\end{aligned}\ ] ] in this case , @xmath15 , @xmath20 and @xmath17 are all eigenvectors .
the structure here is algebraically similar to a perfect fluid with proper mass density @xmath23 and pressure @xmath24 .
the tachyon field lagrangian density is given by : @xmath25^{1/2},\ ] ] with @xmath26 representing the tachyon field .
in the case of the tachyonic scalar field , the governing equations read : @xmath27 + \left(\rho+p\right ) u^{\mu}u_{\nu } + p\delta^{\mu}_{\;\nu } , \label{eq : tt } \\ & \mathcal{t}_{\nu}:=t^{\mu}_{\;\nu;\mu}=0 , \label{eq : const } \\ & \sigma:=\left[\frac{v({\phi})\ , { \phi}_{,\mu}}{{\sqrt{1+{\phi}^{;\kappa}{\phi}_{,\kappa}}}}\right]_{;\mu } -v^{\prime}({\phi}){\sqrt{1+{\phi}^{;\kappa}{\phi}_{,\kappa}}}=0 , \label{eq : sit } \\ & \mathcal{u}:=u^{\alpha}u_{\alpha}+1=0 , \label{eq : ut } \\ & \mathcal{k}=\left[(\rho + p)u^{\alpha}\right]_{;\alpha}-u^{\alpha } p_{,\alpha}= 0 , \label{eq : kt } \\ & \mathcal{f}_{\nu}=(\rho+p)u^{\alpha}u_{\nu;\alpha}+ \left(\delta^{\alpha}_{\;\nu } + u^{\alpha}u_{\nu } \right ) p_{,\alpha } = 0 , \label{eq : ft } \\ & \mathcal{c}^{\nu}(g_{\alpha\beta } , g_{\alpha\beta,\gamma})=0 .
\label{eq : ct}\end{aligned}\ ] ] again the system is underdetermined by two which allows us to prescribe an equation of state and one quantity ( usually @xmath3 or @xmath4 ) .
defining vectors as in ( [ eq : eigassumpq ] ) we find : @xmath28u^{\alpha } , \label{eq : evaleqta } \\
t^{\alpha}_{\;\beta}{\phi}^{;\beta}= & \left[p-\frac{v({\phi})}{{\sqrt{1+{\phi}^{;\kappa}{\phi}_{,\kappa}}}}\right]{\phi}^{;\alpha } + \left[(\rho+p)u^{\beta } { \phi}_{,\beta}\right ] u^{\alpha } ,
\label{eq : evaleqtb}\\ t^{\alpha}_{\;\beta}s^{\beta}= & \frac{v({\phi})}{{\sqrt{1+{\phi}^{;\kappa}{\phi}_{,\kappa } } } } s^{\beta}{\phi}_{,\beta } { \phi}^{;\alpha } + \left[p - v({\phi}){\sqrt{1+{\phi}^{;\kappa}{\phi}_{,\kappa}}}\right]s^{\alpha}. \label{eq : evaleqtc}\end{aligned}\ ] ] as with the quintessence field , none of the vectors @xmath15 , @xmath16 , @xmath17 are eigenvectors . in case
the vectors are mutually orthogonal the following equations hold : @xmath29u^{\alpha},\label{eq : evaltaorth } \\ t^{\alpha}_{\;\beta}{\phi}^{;\beta}=&\left[p-\frac{v({\phi})}{{\sqrt{1+{\phi}^{;\kappa}{\phi}_{,\kappa}}}}\right]{\phi}^{;\alpha } , \label{eq : evaltborth } \\
t^{\alpha}_{\;\beta}s^{\beta}=&\left[p - v({\phi}){\sqrt{1+{\phi}^{;\kappa}{\phi}_{,\kappa}}}\right]s^{\alpha}. \label{eq : evaltcorth}\end{aligned}\ ] ] note that in this case , not only are @xmath15 , @xmath16 and @xmath17 eigenvectors , but the stress - energy tensor possesses the algebraic structure of an _
anisotropic fluid_. in case the fluid velocity and scalar gradient are colinear , the stress - energy tensor has following structure : @xmath30u^{\alpha } , \label{eq : evalta } \\
t^{\alpha}_{\;\beta}{\phi}^{;\beta}=&-\left[\rho+\frac{v({\phi})}{{\sqrt{1+{\phi}^{;\kappa}{\phi}_{,\kappa}}}}\right]{\phi}^{;\alpha } , \label{eq : evaltb } \\
t^{\alpha}_{\;\beta}s^{\beta}=&\left[p - v({\phi}){\sqrt{1+{\phi}^{;\kappa}{\phi}_{,\kappa}}}\right]s^{\alpha } , \label{eq : evaltc}\end{aligned}\ ] ] which is similar to that of a _
here we present the cauchy problem for the self - gravitating system of a perfect fluid and scalar field .
let us consider a contractible space - time domain ( see figure [ fig : cauchfig ] ) @xmath31 and is evolved to the hyper - surface domain @xmath32 . ] let a differentiable symmetric tensor field @xmath33 exist in @xmath34 of the space - time .
we cite synge s lemma @xcite + * synge s lemma : * let @xmath33 be a symmetric , differentiable tensor field in the domain @xmath34 of space - time . then
, the following two statements are mathematically equivalent : @xmath35 the above lemma can be applied to einstein s field equations . in that case
, the following two statements are mathematically equaivalent : @xmath36 we shall investigate equations ( [ eq : einstlem ] ) involving a quintessence scalar field and perfect fluid in the following geodesic normal coordinates : @xmath37_{|t_{0}},\;\ ; \overline{g}_{ij,0,0}(\mathbf{x}):=\left[g^{\sharp}_{ij,0,0}(\mathbf{x},t)\right]_{|t_{0}}\end{aligned}\ ] ] to clarify , @xmath38 will subsequently be used to denote the _ bulk spatial metric _ and @xmath39 will be used to denote the _ spatial metric on the initial hyper - surface_. computing the ricci and einstein tensor components from ( [ eq : einstlem ] ) we obtain : @xmath40 , \label{eq : iconst2 } \\
g_{00}(x)= & \frac{1}{2 } r^{\sharp}(\mathbf{x},t ) + \frac{1}{8 } \left[g^{\sharp kl}g^{\sharp}_{kl,0}\right]^{2 } -\frac{1}{8 } g^{\sharp km}g^{\sharp ln}g^{\sharp}_{kl,0}g^{\sharp}_{mn,0}. \label{eq : iconst3}\end{aligned}\ ] ] here , sharp symbols ( @xmath41 ) on indices denotes covariant differentiation with respect to the three dimensional metric @xmath38 .
the ricci and einstein tensor components can be expressed as ( from ( [ eq : cauchdefs ] ) and ( [ eq : iconst1 ] -[eq : iconst3 ] ) ) : @xmath42 , \label{eq : newiconst2 } \\
g_{00}(\mathbf{x},t_{0})= & \frac{1}{2}\overline{r}(\mathbf{x } ) + \frac{1}{8 } \left[\overline{g}^{kl}\overline{g}_{kl,0}\right]^{2 } -\frac{1}{8 }
\overline{g}^{km}\overline{g}^{ln}\overline{g}_{kl,0}\overline{g}_{mn,0}. \label{eq : newiconst3}.\end{aligned}\ ] ] note that the _ extrinsic curvature components _ of the initial hyper - surface are given by @xmath43 and the above equations are also derivable from the gauss - codazzi equations @xcite , @xcite , @xcite .
we shall now explore the field equations and subsequent evolution with quintessence scalar field and perfect fluid sources .
using synge s lemma and consequent equations ( [ eq : einstlem ] ) and ( [ eq : iconst1 ] - [ eq : iconst3 ] ) we arrive at : @xmath44\right\}=0 , \label{eq : cauchqe } \\ & \sigma=\frac{1}{\sqrt{g^{\sharp } } } \left[\sqrt{g^{\sharp}}g^{\sharp kl}\phi _ { , k}\right]_{,l } -\phi_{,0,0 } -\left(\ln \sqrt{g^{\sharp}}\right)_{,0}\phi_{,0 } -{v^{\prime}(\phi)}=0 , \label{eq : cauchqs } \\ & u^{0}=u(u):=\sqrt{1+g^{\sharp mn}u_{m}u_{n}}\ , \geq 1 , \label{eq : cauchqu } \\ & \mathcal{k}=u(u)\rho_{,0 } + u^{l}\rho_{,l } + \frac{(\rho+p)}{\sqrt{g^{\sharp } } } \left\{\left ( \sqrt{g^{\sharp}}u^{k}\right)_{,k } + u(u)\left(\sqrt{g^{\sharp}}\right)_{,0 } + \sqrt{g^{\sharp}}u_{,0}\right\}=0 , \label{eq : cauchqk } \\ & \mathcal{f}_{0}=\left[1-\left(u(u)\right)^{2 } \right]p_{,0 } -u^{k}u(u)p_{,k } -\left(\rho+p\right ) \left\{u^{l}u_{,l } + \frac{1}{2}g^{\sharp bm}g^{\sharp}_{ma,0}u^{a}u_{b}\right\}-uu_{,0}=0 , \label{eq : cauchqf } \\ & \mathcal{f}_{j}=(\rho+p ) u(u ) u_{j,0 }
-\frac{1}{2}(\rho + p)u g^{\sharp bm}g^{\sharp}_{mj,0}u_{b } + \left[p_{,j}+u^{b}u_{j}p_{,b } + uu_{j}p_{,0 } \right]=0 .
\label{eq : cauchqf2}\end{aligned}\ ] ] moreover , on the initial hyper - surface ( @xmath45 ) , the initial data must satisfy : @xmath46 -8\pi \left[\phi_{,i}\phi_{,0}-(\rho + p)u u_{i}\right]_{|t_{0 } } , \label{eq : qinit1 } \\ \mathcal{e}_{00}(\mathbf{x},t_{0})=&\frac{1}{2 } \overline{r}(\mathbf{x } ) + \frac{1}{8 } \left[{\overline}{g}^{mn}(\mathbf{x } ) \overline{g}_{mn,0}\right]^{2 } -\frac{1}{8 } \overline{g}^{ma}\overline{g}^{nb } \overline{g}_{mn,0 } \overline{g}_{ab,0 } \nonumber \\ & -8\pi \left\{\frac{1}{2 } \left[(\phi_{,0})^{2 } + g^{\sharp ab}\phi_{,a}\phi_{,b}\right]+ v(\phi ) + ( \rho+p ) u^{2 } -p\right\}_{|t_{0 } } = 0 .
\label{eq : qinit2}\end{aligned}\ ] ] the initial data is to be prescribed as : @xmath47 the given functions @xmath48 , @xmath49 , @xmath50 and @xmath51 are of class @xmath52 in @xmath53 , the functions @xmath54 are of class @xmath55 and the functions @xmath56 are of class @xmath57 .
moreover , the consistency conditions ( [ eq : qinit1 ] - [ eq : qinit2 ] ) must be satisfied as : @xmath58 + 8\pi \left\ { \xi\chi_{,a } -\left[\mu+\eta\right]u\,w_{a}\right]=0 , \label{eq : qconsist1 } \\ & \frac{1}{2}\overline{r}+\frac{1}{8 } \left[\gamma^{mn } \psi_{mn}\right]^{2 } -\frac{1}{8 } \gamma^{ma } \gamma^{nb } \psi_{mn}\psi_{ab } \nonumber \\ & \;\;\ ; -8\pi \left[\frac{1}{2 } \left(\xi^{2 } + \overline{\gamma}^{ab } \chi_{,a}\chi_{,b } \right ) + v(\chi ) + \left(\mu+\eta\right ) u^{2 } -\eta \right]=0 .
\label{eq : qconsist2}\end{aligned}\ ] ] ( here , barred quantities are derived from the prescribed metric @xmath59 ) .
the system of three dimensional partial differential equations is underdetermined so that infinitely many solutions should exist locally . restricting the field equations to the initial hyper - surface , @xmath45
, we can rearrange the equations symbolically as : @xmath60p_{,0}\right\}_{|t_{0}}&= & \;\;\ ; " \;\;\;\;\;\;\;\ ; " \;\;\;\;\;\ ; " \;\;\;\;\;\ ; " \;\;\;\;\;\;\;\;\ ; " \;\;\;\;\;\;\;\;\ ; " \;\;\;\;\;\;\;\;\;\;,\nonumber \\ \left[\mu+\eta\right]u\,u_{a,0\;|t_{0}}&=&\;\;\ ; " \;\;\;\;\;\;\;\ ; " \;\;\;\;\;\ ; " \;\;\;\;\;\ ; " \;\;\;\;\;\;\;\;\ ; " \;\;\;\;\;\;\;\;\ ; " \;\;\;\;\;\;\;\;\;\ ;
.\nonumber\end{aligned}\ ] ] thus , the higher derivatives of @xmath61 , @xmath2 , @xmath3 , @xmath4 , and @xmath62 at the initial hyper - surface are determined by allowable cauchy data and their spatial derivatives . assuming the functions @xmath63 , @xmath64 , @xmath65 , @xmath66 , and @xmath67 are _ real - analytic functions _ in @xmath68
, we can determine arbitrary higher order derivatives with respect to @xmath69 by differentiating the field equations in ( [ eq : cauchqe ] - [ eq : cauchqf2 ] ) and restricting subsequently those equations to the initial hyper - surface .
thus , the power series : @xmath70 can be generated . by the real analyticity conditions ,
there exist a @xmath71 such that all the power series in ( [ eq : gevolve]-[eq : uevolve ] ) converge absolutely for all @xmath72 $ ] and converge uniformly for all @xmath73 $ ] for a sufficiently small @xmath74 .
in the case of the tachyonic scalar field , the field and supplemental equations may be expressed as : @xmath75 + ( \rho+p)u_{i}u_{j } + \frac{1}{2}g^{\sharp}_{ij}(\rho -p ) \right\}=0 , \label{eq : etach } \\ & \sigma=\frac{1}{\sqrt{g^{\sharp } } } \left[\sqrt{g^{\sharp } } g^{\sharp ab}{\phi}_{,a } \right]_{,b } -{\phi}_{,0,0 } -\left(\ln \sqrt{g^{\sharp}}\right)_{,0 } { \phi}_{,0 } -\left[\ln |{v(\phi)}| \right]^{\prime } \nonumber \\ & \;\;\;\;\;\ ; -\left[1+{\phi}^{;\mu}{\phi}_{,\mu } \right]^{-1 } \left[{\phi}_{,i } { \phi}_{,j } g^{\sharp \,nj}g^{\sharp\ , bi } \left({\phi}_{,n;b^{\sharp } } + \frac{1}{2}g^{\sharp}_{bn,0}{\phi}_{,0}\right ) \left({\phi}_{,0}\right)^{2}{\phi}_{,0,0 } \right .
\nonumber \\ & \;\;\;\;\;\ ; \left .
-2{\phi}_{,0}{\phi}_{,j}g^{\sharp\,nj}{\phi}_{,n;0}\right ] = 0 , \label{eq : sigtach } \\ & u^{0}=:u:=\sqrt{1+g^{\sharp ab}u_{a}u_{b } } \geq 1 , \label{eq : utach } \\ & \mathcal{k}:=u\,\rho_{,0 } + u^{a } \rho_{,a } + \frac{(\rho+p)}{\sqrt{g^{\sharp } } } \left[\left(\sqrt{g^{\sharp}}u^{a}\right)_{,a}+ u\sqrt{g^{\sharp}}_{,0 } + \sqrt{g^{\sharp } } u_{,0 } \right]=0 , \label{eq : ktach } \\ & \mathcal{f}_{0}=\left[1-\left(u\right)^{2 } \right]p_{,0 } -u^{k}u(u)p_{,k } -\left(\rho+p\right ) \left\{u^{l}u_{,l } + \frac{1}{2}g^{\sharp bm}g^{\sharp}_{ma,0}u_{b}\right\}-uu_{,0}=0 , \label{eq : f0tach } \\ & \mathcal{f}_{j}=(\rho+p ) u(u ) u_{j,0 } -\frac{1}{2}(\rho + p)u g^{\sharp bm}g^{\sharp}_{mj,0}u_{b } + \left[p_{,j}+u^{b}u_{j}p_{,b } + uu_{j}p_{,0 } \right]=0 .
\label{eq : f1tach}\end{aligned}\ ] ] as before , there exist constraints on the initial data on the @xmath45 hyper - surface : @xmath76 \nonumber \\ & - 8\pi \left\{{v(\phi)}\left[\frac{{\phi}_{,a}{\phi}_{,0}}{\sqrt{1 + g^{\sharp rs}{\phi}_{,r}{\phi}_{,s}-({\phi}_{,0})^{2}}}\right ] -(\rho + p)u u_{a } \right\}_{|t_{0}}=0 , \label{eq : tachcons1 } \\
\mathcal{e}_{00}(\mathbf{x},t_{0})= & \frac{1}{2 } \overline{r}+\frac{1}{8 } \left[\overline{g}^{mn } \overline{g}_{mn,0 } \right]^{2 } - \frac{1}{8 } \overline{g}^{ma } \overline{g}^{nb } \overline{g}_{mn,0 } \overline{g}_{ab,0 } \nonumber \\ & -8\pi \left\{{v(\phi)}\left[\frac{1+g^{\sharp ab}{\phi}_{,a } { \phi}_{,b}}{\sqrt{1+g^{\sharp mn } { \phi}_{,m}{\phi}_{,n } -({\phi}_{,0})^{2 } } } \right]+ ( \rho+p ) u^{2 } - p\right\}_{|t_{0 } } = 0 .
\label{eq : tachcons2}\end{aligned}\ ] ] the initial data is prescribed as in the quintessence case ( [ eq : quintindat1])-([eq : quintindat5 ] ) with @xmath2 replaced by @xmath26 ( differentiability requirements remain the same ) . in this case , the consistency equations ( [ eq : tachcons1])-([eq : tachcons2 ] ) become : @xmath77 + 8\pi \left\{v(\chi ) \left[\frac{\xi\chi_{,a}}{\sqrt{1+\gamma^{rs}\chi_{,r}\chi_{,s } -\xi^{2 } } } \right ] -\left(\mu+ \eta\right)u\,w_{a } \right\}=0 , \label{eq : inittacha } \\ & -\frac{1}{2}\overline{r } -\frac{1}{8 } \left[\gamma^{mn } \psi_{mn}\right]^{2 } + \frac{1}{8 } \gamma^{ma } \gamma^{nb } \psi_{mn } \psi_{ab } \nonumber \\ & \;\;\ ; + 8\pi \left\{v(\chi ) \left[\frac{1+\gamma^{ab}\chi_{,a}\chi_{,b}}{\sqrt{1 + \gamma^{mn}\chi_{,m}\chi_{,n } -\xi^{2}}}\right ] + ( \mu+\eta ) u^{2 } -\eta\right\ } = 0 .
\label{eq : inittachb}\end{aligned}\ ] ] at this point , the evolution is governed by the equations ( [ eq : gevolve ] ) - ( [ eq : uevolve ] ) as before .
it is instructive to demonstrate how the above scheme works with an explicit example .
we shall study the numerical evolution of a system whose analytic properties are known .
comparison with a known solution will show whether or not the method works as well as provide a benchmark on its convergence properties . specifically , we consider the evolution of a constant scalar field in an otherwise empty , flat friedmann - lematre - robertson - walker ( flrw ) space - time .
although extremely simple , this example is pedagogically useful as it serves well to elucidate the employment of the scheme without unnecessary complications which arise from more complex systems . as well , the constant field evolution is identical for both the quintessence and tachyonic scenario . we show how to extract various quantities for the cauchy evolution and compare with the known analytic result .
since the source consists of only the constant scalar field , we can immediately set the following initial data : @xmath78 with this prescription , the equation pairs ( [ eq : qconsist1 ] - [ eq : qconsist2 ] ) and ( [ eq : inittacha ] - [ eq : inittachb ] ) both take the form : @xmath79 = 0 , \label{eq : exampconst1 } \\ & -\frac{1}{2}\overline{r } -\frac{1}{8}\left[\gamma^{mn}\psi_{mn}\right]^{2 } + \frac{1}{8 } \gamma^{ma } \gamma^{nb } \psi_{mn}\psi_{ab } + 8\pi v(\chi_{0 } ) = 0 .
\label{eq : exampconst2}\end{aligned}\ ] ] it should be noted that in general , the consistency equations , although underdetermined , are difficult to solve even in vacuum and one must appeal to numerical techniques for solutions . as well , we consider the spatially flat flrw metric : @xmath80=a^{2}(t_{0})\left [ \begin { array}{ccc } 1&0&0\\ \noalign{\medskip } 0&r^{2}&0\\ \noalign{\medskip } 0&0&r^{2}\sin^{2}\theta \end { array } \right ] , \nonumber\ ] ] with @xmath81 the value of the scale factor on the initial data hyper - surface .
we consider an inflationary scenario and prescribe : @xmath82_{,0|t = t_{0}}=&2e^{2t_{0}}. \nonumber\end{aligned}\ ] ] at this point , all the allowable initial data has been prescribed .
the einstein - scalar field equations ( [ eq : etach]-[eq : sigtach ] ) dictate that @xmath83 at this point it is useful to reiterate that the above prescription is only valid _ if _ it satisfies the constraint equations ( [ eq : tachcons1 ] - [ eq : tachcons2 ] ) . in this example , the relevant values for the constraint equations are : @xmath84 it is a simple matter to check that these quantities indeed satisfy the constraint equations ( [ eq : exampconst1 ] ) and ( [ eq : exampconst2 ] ) . we show both the cauchy evolved and the analytic @xmath85 square of the scale factor in figure [ fig : evolved ] . as expected , there is excellent agreement at small values of @xmath86 .
deviations at larger @xmath86 may be minimized by retaining more terms in the taylor series ( [ eq : gevolve ] ) as can be seen from the vatious dotted lines in the figure ( see figure caption ) .
it may readily be verified that the other parameters to be evolved will display the proper evolution by a simple inspection of ( [ eq : gevolve ] - [ eq : uevolve ] ) .
that is , in this simple example , all other parameters will retain their initial values as prescribed by ( [ eq : exinitdat ] ) .
( quadratic , cubic , quartic ) .
the solid line represents the analytic result @xmath85 . ]
both the quintessence and tachyonic scalar field , supplemented with a perfect fluid , were considered in the context of general relativity .
the general mathematical properties of the system , including the eigenvalue structure were studied .
it is seen that the system may behave either as a two component perfect fluid or an anisotopic fluid , the anisotropy being due to the properties of the scalar field .
finally , in the geodesic coordinates , the cauchy problem as well as the initial constraint equations have been derived .
assuming analyticity , the cauchy scheme presented here is convergent .
the scheme is iterative and may easily be executed by computer . | the system consisting of a self gravitating perfect fluid and scalar field is considered in detail .
the scalar fields considered are the quintessence and `` tachyonic '' forms which have important application in cosmology .
mathematical properties of the general system of equations are studied including the algebraic and differential identities as well as the eigenvalue structure .
the cauchy problem for both quintessence and the tachyon is presented .
we discuss the initial constraint equations which must be satisfied by the initial data .
a cauchy evolution scheme is presented in the form of a taylor series about the cauchy surface .
finally , a simple numerical example is provided to illustrate this scheme .
pacs numbers : 04.20.ex , 04.40.-b , 98.80jk + msc : 83f05 , 83c05 + key words : scalar fields , cauchy problem , relativity + |
the graal beam line facility @xcite currently in operation at the esrf in grenoble , is the first source of high intensity and completely polarized @xmath0 rays in the energy range [email protected] gev .
this project has been realized , with the prevailing support of the istituto nazionale di fisica nucleare ( infn ) , to study polarization observables in photoproduction reactions including strangeness . the graal apparatus ( see fig .
1 ) , consists of a high resolution and large solid angle bgo electromagnetic calorimeter combined with multiwire proportional chambers ( mwpc ) that covers a solid angle range of almost 4@xmath2 . particles emitted at small angles are also detected by a scintillator wall , that is installed three meters from the target and permits particle identification by means of their time of flight and their energy loss in the scintillators .
the particle identification in the central region is accomplished with a plastic scintillator barrel through the measurement of de / dx . in this paper
we describe the photon - detection system , which has been designed to measure the energy of @xmath0 and neutral mesons decaying in two or more photons ( @xmath3 , @xmath4 , @xmath5 , k@xmath6 , @xmath7 ) , with a good angular resolution .
the calorimeter covers 90% of the entire solid angle , detecting particles emitted at angles from 25^@xmath8to 155^@xmath8 .
the accuracy and reliability of the energy calibration is a basic requirement for this detector in which a large number of bgo sectors , comprising about 500 pms , are involved and high resolution is expected . the problem is to keep under control the variations in the gain and temperature of the different sectors as a function of time , thus ensuring uniformity of response during data taking and keeping to a minimum the time spent calibrating the calorimeter .
we shall , therefore , give particular emphasis to the description of our led - based monitoring system , which plays a key - role in this respect . in sect .
2 we describe briefly the characteristics of the apparatus . in sect .
3 the principles of the electronics and data acquisition . sect .
4 is devoted to the calibration procedure .
the gain monitoring system is described in sect .
5 . in sect . 6 we report on the linearity of the calorimeter energy response . in sect .
7 we report on the performances of the bgo calorimeter and the monitoring system , with special emphasis on the energy resolution and time stability .
the bgo detector is shown in detail in fig .
the mechanical support structure consists of 24 baskets of carbon fiber composite material supported by an external steel frame .
each basket is divided into 20 cells with very thin walls , 0.38 mm for the inner and 0.54 mm for the outer walls , to keep the crystals optically and mechanically separated .
the carbon fiber has been preferred to other materials like aluminum for its higher rigidity and lower gamma ray attenuation due to its low z number .
the support frame is divided into two halves which can be taken apart by 1.5 meters to allow access to the target and central detector region .
when closed the structure leaves a 20 cm diameter hole along the beam - line for the insertion of the target , the cylindrical wire chambers and the plastic scintillator barrel .
the crystals are of 8 different dimensions and are shaped like pyramidal sectors with trapezoidal basis @xcite ( see fig.3 ) .
they define 15 angular regions ( @xmath9 ) in the plane containing the symmetry axis of the calorimeter , coincident with the beam axis , and 32 ( @xmath10 ) in the plane orthogonal to the beam axis ( see tab.1 of @xcite and @xcite for details ) .
the 480 crystals have all the same length of 24 cm ( @xmath1121 radiation lengths ) , for a good confinement of photon showers in the gev region , and are arranged in such a way that the reaction products emitted in all directions from the target center encounter a constant thickness of bgo .
each crystal is wrapped up in a thin ( 30@xmath12 m ) aluminized mylar reflector , and its back side is optically coupled to a photomultiplier ( pm ) @xcite .
two holes in the back side of the crystal support are used for monitoring the temperature and for the input , through optical fiber , of light pulses which are used for the measurements of the linearity and gain stability of the photomultipliers .
+ during the production phase each one of the crystals has been accurately tested to check the accomplishment of the requirements imposed for acceptance : longitudinal uniformity @xmath13 95% and resolution at the 0.661 mev cesium @xmath0-peak @xmath14 20% fwhm .
the quality tests gave results better than the design specifications .
two thirds of the crystals have an average resolution at cesium better than 18% fwhm and two thirds have a longitudinal uniformity greater than 97% .
since the bgo calorimeter is operating in a region without magnetic field and we need to measure with a good resolution electromagnetic showers with energy less than few hundreds mev , we choose for the readout of the signal standard photomultipliers due to their noise much smaller than that of other devices such as photodiodes .
the anode signals from the pms enter in 15 adders ( mixer ) , each having 32 input channels with programmable attenuators @xcite .
the outputs from each module consist of : a linearly summed prompt output , with a fan - out of 6 , used for trigger purposes and to build up the calorimeter total energy hardware sum . a 300 ns delayed and , if necessary , attenuated output that is sent for digitization to two fera modules ( fast encoding and readout adc , charge - sensitive , 11-bit , 16 channels per unit ) @xcite .
the linearly summed output of each mixer ( @xmath15 ) corresponds to the sum of the signals coming from the 32 bgo crystals having the same @xmath9 angle .
the 15 @xmath15 outputs are sent to another mixer ( mixer-@xmath16 ) .
the individual linear output signals from the mixer-@xmath16 are sent to another fera adc for the acquisition of the partial sums . the summed output signal from the mixer-@xmath16 , that corresponds to the total energy of the bgo calorimeter ,
is sent to the different trigger circuits and to another fera channel ( see fig .
4 ) . the linear part of the acquisition electronics is thus formed by 16 mixer and 31 fera adc modules . + for the adc read - out of the bgo calorimeter and other detectors of the graal apparatus we have found an original solution that allow a fast and sequential acquisition of all the camac crates using the fera - bus . in fig .
5 is shown the scheme of the handshake logic for the read - out of the adc .
the fera adc of each crate are connected in parallel to the fera - driver @xcite through the command - bus and the data - bus .
we have specifically designed a system that needs only one bus for the data coming from the drivers of each crate . for this purpose
we have used three bus - switch modules @xcite as interfaces between the fera - drivers and the single data - bus .
the bus - switch module was originally designed to accept as input the signals from two different data - bus giving at the output one of the two depending on the status of the strobe signals of the two data - bus . in our scheme
we exploit the three state port characteristic of the bus - switch output .
the open circuit state of the output allows to disconnect the module from the bus , in such a way that the transmission on the same bus of signals from other modules is not disturbed . using one bus - switch module for each crate and connecting the driver data output with one of the two bus - switch data inputs , with a suitable signal timing logic , we achieve the `` mixing '' of the data on a single bus .
the fera - adc are sequentially read starting from the third crate and continuing with the adc of the second and first crate pertaining to the bgo calorimeter .
+ during the calibration phase the data are temporarily stored in an auxiliary memory area formed by two high speed memory modules @xcite hsm-1 and hsm-2 , as shown in fig.5 .
these memories are configured to work in a flip - flop to combine acquisition and data readout without loss of speed .
when hsm-2 is filling , hsm-1 transfers the recorded data to the memory of the calibration processor @xcite , according to an interrupt scheme , and will be ready to accept new data when hsm-2 is full . during the experimental data acquisition
the gate signal for the fera - adc is given by the experimental trigger and the fera - bus is automatically connected to the general data - bus for the global acquisition of all the detectors of the graal apparatus .
the absolute calibration of the crystals is obtained using the 1.27 mev photons ( @xmath17 ) from a @xmath18na source .
the response of the 480 bgo sectors is equalized using an automatic procedure , that sets all the pm s gains , varying their high voltage .
if we call @xmath19 the calibration channel we would like to obtain , we can say that the equalization is achieved when all the adc channels @xmath20 ( @xmath21 ) are within the following range : @xmath22 @xmath23 indicates the precision of the procedure . to maintain the time spent for the equalization at the reasonable value of about one hour , we fix @xmath24 .
this value affects only the precision on the hardware energy sum used for the experimental trigger , but does not affect the absolute calibration resolution because in the data analysis the exact value of the calibration constant for each crystal is used .
the choice of the equalization channel is determined by the maximum energy value that has to be measured and by the range of the detector linearity response .
the maximum @xmath0 energy of the graal beam is 1.5 gev , but the monte carlo simulations and the experimental data @xcite indicate that no more than 80% of the incident @xmath0 energy can be contained in a single crystal ( @xmath25 gev ) , due to the transverse spread of the electromagnetic shower .
the saturation effects of the crystals pms become evident for peak pulse amplitudes of the order of 5 v , so that the signals corresponding to the maximum energy released should not exceed this value . for the signals coming from bgo scintillation events
, we have measured the following relation between the peak amplitude of the signal @xmath26(mv ) and the adc channels : @xmath27 the end - scale value of our fera - adc is 1920 channels , corresponding to a signal equivalent charge of 480 pc . to this value
we must subtract the channels corresponding to the pedestal signal , that have been adjusted in the range 100 - 120 ; for this reason we can assume that the maximum number of adc channels available is : @xmath28 .
in the hypothesis of a linear response , the following relation holds : @xmath29 where @xmath30 is the attenuation factor on the analog signal introduced by the mixer . according to ( [ emax ] )
, we obtain in our case : @xmath31 the best value for @xmath30 is determined knowing that the @xmath26 values of the signal corresponding to @xmath32 depend on the attenuation factor , according to the following relation : @xmath33 from which we decided to use @xmath34 , that represents the best approximation between the available attenuation factors . from eq.([ceq ] ) it follows that @xmath19=64 is a reasonable compromise between calibration precision and saturation effects . this way , without attenuation
, we obtain a calibration constant @xmath350.02 mev / channel .
the maximum energy that can be measured with attenuation is thus equal to @xmath361.15 gev compatible with the expected @xmath37 value .
the calibration constant of each crystal measured at the equalization time @xmath38 may change as a function of time , due to two principal reasons : variation in the crystal light output due to temperature effects and variation in the pm gain . to control the temperature effects we have set up a monitoring system , that is discussed in details in @xcite and
will not be described further here .
gain variations of the photomultipliers may occur as a consequence of the following effects : * photocathode temperature variation : at the bgo peak emission wavelength ( 480 nm ) the efficiency variation of a bialkali photocathode is of the order of -0.4%/@xmath39c ; * instability in the high voltage power supply @xcite : these are of the order of @xmath401 v and assuming a hv typical value v=1500 v , the induced gain variations are 0.6% for the r580 pm model and 0.8% for the r329 - 02 model ; during the acquisition of the calibration peaks , this effect is however strongly reduced as it is averaged over a long time interval ; * pm drifts and shifts ; for both source and experimental events , we have an anodic mean current of the order of @xmath41 @xmath12a , much lower than the maximum value of 1 @xmath12a recommended in order to avoid drift effects ; also shift effects can be neglected as previous tests made on a basket of 20 bgo crystals @xcite have shown that for rates lower than 1 khz , as in our case , we have a gain variation due to shift effects of the order of 0.01% ; * aging of cathode and dynodes materials : a typical time scale for these effects is of the order of few years and can be kept under control with a regular monitoring of the pms gains ; * voltage dividers instability : this effect does not cause a time variation of the pm gain but a loss of linearity in the pm response ; the linearity curves have been measured as reported in section 6 . in order to keep under control all time variations of the calibration constants previously discussed we have realized a photomultiplier gain monitoring system as described below .
+ the monitoring system , as shown in fig . 6 ,
is made of : * high luminosity , blue emitting , pulsed leds as light source ; * a light distribution system consisting of an integrating sphere and optical fibers ; * a reference nai scintillation detector . a matrix of 7 leds @xcite ,
is directly coupled to the input port of the integrating sphere .
the following aspects were taken into account with particular care in the final choice of the light source : * emission spectrum ; * time stability ; * time shape of the light pulse ; * intensity .
the emission spectrum of the chosen leds has a maximum at 450 nm with a fwhm of 140 nm , that is quite similar to the bgo one .
the led pulse - to - pulse stability is of the order of 1% , depending on the light intensity . in fig .
7 is shown the led pulse distribution corresponding to the maximum light intensity , in a generic bgo crystal .
measurements done in a time interval of several days have shown that the led time stability is of the order of a few percent .
we decided not to use the pulsed laser source @xcite , in spite of its high intensity , because it had a pulse to pulse instability greater than 20% ( fwhm ) and showed a time drift in the emission intensity as high as 50% .
the time shape of the light pulse is important for two fundamental reasons .
the first is that the pm anode pulse from the light source should reach a charge equivalent to @xmath421 gev energy without having a current peak that gives rise to saturation in the pm .
the second reason is that the pm response may depend on the time shape of the input pulse also without saturation effects . especially for these reasons we realized a driver circuit for the leds that produces an anode pulse ,
see fig . 8 , very similar to that produced by scintillation in bgo .
the intensity of the light source should be high enough to simulate the maximum energy deposited in each detector .
the actual configuration of the leds produce an equivalent energy of almost 1 gev for each of the 480 bgo crystals .
we have chosen an integrating sphere of 6 " internal diameter @xcite to match the leds light to the 600 optical fibers , 6 meters long , carrying the light to each detector .
this solves in a simple way the challenging problem of having a uniform light distribution over the fibers that are connected to the exit port of the sphere .
thus the ratios of the incident light intensities over any couple of fibers are constants independently of the variations of the light source .
the inner surface of the sphere consists of spectralon reflectance material that gives the highest diffuse reflectance of any known material over the uv - vis - nir region of the spectrum .
the reflectance is generally @xmath1199% over the range 400 - 1500 nm .
radiation introduced into the sphere undergoes multiple scattering , which results in a uniform , spatially integrated radiation distribution within the sphere .
the exit port serves as a source of radiation characterized by a radiance which is constant over the plane of the exit port and independent of viewing angle .
the light carried by each fiber is then optically coupled to each crystal .
the reference detector is a crucial part of the linearity and gain monitoring system because it measures the intensity of the light pulses from the leds and the outputs of all the other detectors are compared to this value .
thus its linearity and gain variation should always be under control .
this detector is a 2``x2 '' nai(tl ) scintillator @xcite coupled to the same type of pm used for the bigger bgo crystals .
five optical fibers carrying the leds light are coupled to the detector through a quartz window .
a @xmath43cs radioactive source was used to monitor the gain drift of its response .
we have realized a system to remotely remove the source when the leds light is flashing to avoid that random coincidences between the leds and the source signal alter the measure of the leds intensity .
the electronic chain and trigger circuit for the acquisition of the reference detector are different from those of the calorimeter .
the anode signal is formed by a preamplifier and a spectroscopy amplifier and is then digitized by a high performance buffered spectroscopy adc @xcite with 8000 channels .
the data from the adc are sent via an external bus to an histogramming memory @xcite and at the end of the acquisition are transfered to the calibration processor memory for analysis .
our method of calibration of the bgo calorimeter , as described in section 4 , is simple , fast and very effective , allowing to determine the absolute calibration constants of all the cristals in less than ten minutes during a pause of the experimental data taking , tipically a new injection of the electron beam in the esrf .
nevertheless the extrapolation of the calibration constant from the 1 mev region of the gamma source to the 1 gev region of the maximum energy deposited in a single cristal requires careful account of the possible non - linear behaviour of the electronics chain ( pms , mixer and fera adc ) in this large energy range .
we have thus measured the linearity curve of each crystal of the calorimeter and applied these correction factors to the measured energy .
the variations of the light intensity of the leds over almost 3 orders of magnitude was obtained attenuating the driving leds input pulse using a camac programmable attenuator @xcite with steps of 0.25 db .
the relative intensities of the light pulses for each crystal were normalized to the response of the reference detector , the excellent linearity of which has been repeatedly verified .
the measure of the nai reference detector linearity is crucial and preliminary to that of the bgo detectors .
we set up a system that allows to send to the reference the light of any combination of five fibers .
the five fibers coming from the integrating sphere are in fact coupled to other five fibers carrying the light to the reference in such a way that it is possible to adjust the distance between each pair of coupled fibers and thus adjust , for each fiber , the light reaching the detector . moreover a remotely controlled shutter placed between each couple of fibers enables or disables the optical contact .
two out of five couples of fibers are adjusted in such a way that they send the same amount of light to the reference .
the output of the reference detector is measured three times : when one shutter is open , when the other one is open and when both of them are open .
if the system is linear the sum of the individual measurements should equal , for any light intensity , the response when both shutters are open .
the variations of the led light intensity during the three measurements , of the order of few tenths of percent , are monitored and corrected using the light measured by one bgo detector . in fig .
9 it is shown the ratio between the sum of the adc channels corresponding to the two single fibers and the channel obtained when the shutters of the two fibers were both opened .
each point represents the linearity of that adc channel respect to its half .
these points are then fitted with a polynomial as shown in the same figure . from this curve ,
moving in steps of a factor of two , we can deduce the integral non linearity points of fig .
this figure indicates that the non linearity does not exceed @xmath400.3% in a range larger than a factor of 30 .
these data are fitted again with a polynomial function to obtain the curve shown in fig .
10 . applying this correction to the original points of fig .
9 we obtain the distribution of fig . 11 which indicates that , after correction , the distribution of the individual measurements is better than @xmath400.1% . with this procedure we can measure the linearity of the bgo over a factor of 30 extending approximately from 1 to 30 mev . to extend our measurements of a further factor of 30 to arrive , in the bgo , at a light pulse equivalent to around 1 gev , we shut off four of the five optical fibers delivering the light pulse to the reference detector .
the only fiber left open had been previously adjusted to carry about 1/30 of the total light .
this way the pulse in the reference detector returns to around channel 200 .
increasing further the light of another factor of 30 , up to its maximum value , we can extend our measurements over three orders of magnitude . this way the absolute calibration done with the @xmath18na source at 1.27 mev is extended up to 1 gev .
the result is a calibrated light source which covers a range of 1000:1 . in order to correct the non linear response of the bgo detectors
, we developed an automatic procedure to build the linearity curves for each crystal .
the leds light intensity has been initially adjusted in such a way to correspond to an energy of @xmath421.27 mev around the adc calibration channel 64 , with no attenuation .
successively a 30 db attenuation on the analog signal is inserted and the leds light is progressively increased until the maximum intensity corresponds to an equivalent energy of almost 1 gev . this way we obtain the response of each bgo detector to light pulses of known amplitude corresponding to energies up to 1 gev .
two typical response curves from two photomultipliers of different type are indicated in fig .
12 . they have been taken with an attenuation of 30 db .
each distribution was separated in three regions and each region fitted with a polynomial function up to the eighth degree .
the energy correction factor for each adc channel of each crystal has been extracted from these functions .
the absolute energy calibration of the adc scale for each bgo detector with the 30 db attenuation is obtained calculating the energy corresponding to the first point of the linearity distribution at 30 db .
this is done , for each bgo crystal , multiplying the energy calibration by the ratio between the light measured in the reference , in correspondence to this point at 30 db , and that corresponding to the calibration channel at 0 db .
the validity of the calibration method and the linearity correction procedure for the bgo calorimeter have been recently tested using the first experimental data obtained with the polarized and tagged graal @xmath0 ray beam incident on a liquid hydrogen target placed at the center of the calorimeter .
in particular we looked at the reaction @xmath44 in order to identify the @xmath4 mesons we selected the events with two neutral clusters in the calorimeter , corresponding to the decay @xmath45 .
we required also no charged cluster in the calorimeter and one charged particle in the forward scintillator wall , corresponding to the recoil proton . in fig .
13a and 13b are shown the reconstructed @xmath4 mass distribution with and without the bgo linearity corrections .
the value of the @xmath4 mass has been deduced using the energy of the two neutral clusters , the incident @xmath0 energy from the tagging system and the angle of the proton in the forward scintillator wall .
the two distributions in fig .
13 have been fitted with two gaussian functions and their parameters are reported in the figures , where p2 corresponds to the mean value and p3 to the @xmath46 of each gaussian . from these figures
we can see that the linearity corrections increase the @xmath4 mass by @xmath42 1% and improve the resolution of the mass distribution by almost 20% . this way the @xmath4 mass distribution has a mean value of 542.8 mev that differs by less than 1% from the true value ( 547.45 mev ) , due to uncertainties on the incident @xmath0 ray energy and on the corrections introduced by the monte carlo simulation .
the value of the @xmath4 mass and its resolution , comparable to that measured for the bgo at these energies @xcite , proves the validity of this method of calibration .
a typical drift distribution of the calibration peaks for all the bgo detectors , corresponding to two consecutive calibration runs , is showed in fig.14 , where we can see that in the time interval of 12 hours the peak positions are stable at a few tenths of a percent level .
we have designed and installed a calibration and monitoring system for the bgo electromagnetic calorimeter in operation at the graal beam of the esrf facility .
we have developed a method for the correction of the non - linearity that allows the measure of the energy in a range of three orders of magnitude with an accuracy that is a small fraction of the intrinsic detector resolution .
photomultiplier tubes r580 ( 38 mm diameter ) for the 192 smaller crystals and r329 - 02 ( 52 mm diameter ) for the 288 bigger ones , both manufactured by hamamatsu photonics .
these type of pms have been chosen after a detailed comparative test with other models ; see @xcite for details . | we describe the electromagnetic calorimeter built for the graal apparatus at the esrf .
its monitoring system is presented in detail .
results from tests and the performance obtained during the first graal experiments are given .
the energy calibration accuracy and stability reached is a small fraction of the intrinsic detector resolution . |
the year 1999 has already seen impressive progress in flavor physics .
cdf reported the first measurement of @xmath6 from @xmath7 decay @xcite which strongly hints to cp violation in a system other than the neutral kaon system .
the situation of direct cp violation @xmath0 in the neutral kaon system used to be somewhat unclear , but ktev result @xcite and the na48 result reported at this meeting @xcite made the experimental situation basically settled .
the numbers on @xmath2 reported are . ]
@xmath8 .various standard model estimates of @xmath2 in the literature .
two bosch estimates use different renormalization schemes ( ndr and hv ) . [
cols="<,^",options="header " , ] on the other hand , the theoretical situation is rather unclear .
the calculation of @xmath0 in the standard model is difficult partly because of a cancellation between gluon and electroweak penguins which makes the result sensitive to the precise values of the hadronic matrix elements .
a ( not complete ) list of theoretical calculations is given in table [ tab : estimates ] .
see @xcite and @xcite for more details on this issue .
the experimental values are compared to the `` probability density distributions for @xmath2 '' @xcite in fig .
[ fig : pdf ] .
there is a feeling in the community that the data came out rather high , even though one can not draw a definite conclusion if the standard model accounts for the observed high value because of theoretical uncertainties .
= 0.7 of course the correct strategy to resolve this issue is to improve theoretical calculations , probably relying on the progress in lattice calculations .
unfortunately , this is a challenging program and we can not expect an immediate resolution .
what i instead attempt in this talk is the alternative approach : think about new physics candidates which are `` reasonable '' and at the same time account for the observed value of @xmath2 .
since any new physics explanation would probably give rise to other consequences , the cross - check can start eliminating such possibilities .
if all `` reasonable '' candidates get excluded , one can draw a conclusion that the standard model should account for the observed high value of @xmath0 . or such a cross - check might confirm other consequences which would be truly exciting .
naturally , i turned my attention to supersymmetry , most widely discussed candidate of physics beyond the standard model , and asked the question if supersymmetric contribution to @xmath0 can be interesting at all .
there are many studies of @xmath0 in supersymmetric models , most notably @xcite .
their detailed study found the possible range of supersymmetric contribution , @xmath9@xmath10 .
then clearly supersymmetry can not account for the observed high value , and indeed i do not have anything new to add to their beautiful analysis within the framework they used : the minimal supergravity model . the reason why supersymmetric contribution to @xmath0 is small
can be quite easily understood . within the minimal supergravity framework ,
the main contribution to flavor - changing effects between the first and second generations originate in the left - handed squark mass - squared matrix . in general , the superpartners of the left - handed quarks have a mass - squared matrix @xmath11 if there is a non - vanishing off - diagonal element @xmath12 in the above mass - squared matrix , it would contribute to flavor - changing processes from loop diagrams .
the easiest way to study such contributions is to use the mass insertion formalism where one treats the off - diagonal element perturbatively as an `` interaction '' in the squark propagator , because the existent constraints require the off - diagonal element to be small anyway .
the size of such a perturbation can be nicely parameterized by @xmath13 where @xmath14 is the average of the two diagonal elements .
= 0.45 = 0.45 contribution of the mass insertion to @xmath0 is given by for simplicity .
see @xcite for more details . ]
@xcite : @xmath15 a representative feynman diagram is shown in fig . [
fig : ll ] . on the other hand , the insertion of the same parameter @xmath16 induces kaon mixing parameters @xmath17 a representative feynman diagram
is shown in fig .
[ fig : ll ] .
therefore , once the constraints from @xmath18 and @xmath19 are satisfied , it does not allow a large contribution to @xmath0 unless some fine - tuning is done .
this simple comparison already gives a typical order of magnitude @xmath20 , with some wiggle room by varying @xmath21 , allowing chargino contributions , etc . beyond this estimate by assuming that @xmath16 is nearly pure imaginary because @xmath22 is then nearly pure real and hence @xmath19 constraint is satisfied .
] this leads to the basic conclusion in @xcite that the supersymmetric contribution to @xmath0 is never important .
amplitude from supersymmetry . a generally missed contribution to @xmath23 amplitude due to the isospin breaking
@xmath24 leads to a too - large @xmath0 if the splitting is @xmath25 .
a more natural size of splitting from the renormalization - group point of view is of order 10% , and with a quarter of the maximum possible value of @xmath26 , the resulting size of @xmath0 is in the ballpark of the observed value @xcite . ]
antonio masiero and myself found that there is a loophole in previous discussions why supersymmetric contribution to @xmath0 is small @xcite . from an enhanced @xmath27 vertex
was suggested @xcite which is subject to tighter constraints phenomenologically @xcite .
] we pointed out that a broad class of supersymmetric models actually gives an important contribution , @xmath28 where the uncertainties come from model dependence and hadronic matrix elements .
i discuss the angle @xmath29 and the origin of this new contribution below .
the masses of quarks arise from the coupling of left - handed and right - handed quarks to the higgs expectation value , @xmath30 here , @xmath31 is higgs field which couples to the down - type quarks in the minimal supersymmetric standard model .
similarly , in addition to the mass - squared mass matrix ( [ eq : ll ] ) for the left - handed squarks and the analogous one for the right - handed squarks , there is another contribution to the masses of squarks which mix left - handed and right - handed particles , @xmath32 the important point here is , without going into detailed discussions of models , we expect the matrix @xmath33 to have similar structure as the yukawa matrix @xmath34 because they have exactly the same property under the flavor symmetries .
i will describe this point more clearly in the context of models later on , but based on this simple fact , i expect @xmath35 where @xmath36 is the typical mass scale of supersymmetry breaking ( _ i.e. _ , superparticle masses ) .
it is possible that @xmath33 and @xmath37 are exactly proportional to each other as matrices , which was assumed in the minimal supergravity and hence in previous analyses , but not necessarily .
generically , one expects the following patterns for yukawa and @xmath38 matrices , @xmath39 where @xmath40 , @xmath41 , @xmath42 are @xmath25 unknown constants and @xmath43 .
the ( 2,1 ) elements are intentionally left blank because we do not know the mixing angles of right - handed quarks ( read : model - dependent ) . in order to discuss flavor - changing effects from squarks ,
one first needs to go to the mass basis for quarks , which is achieved by rotating the yukawa matrix by the cabibbo angle @xmath44 .
the same rotation acts on the @xmath38 matrix , but unless @xmath45 , it still leaves the off - diagonal element after rotation @xmath46 .
we can again describe its effect in terms of a mass insertion parameter @xmath47 it contributes to the flavor - changing chromo - electric dipole moment @xmath48 @xmath49 ( fig .
[ fig : lr ] ) , where @xmath50 is the gluon field strength .
this operator generates a supersymmetric contribution to @xmath0 as @xmath51 , which gives eq .
( [ eq : epsilonp ] ) .
the model uncertainty ( 0.52 ) comes from the lack of knowledge on the @xmath25 constant @xmath52 , and also that we do not know the ratio of @xmath36 in the lr mass matrix relative to the average squark mass .
the phase factor is @xmath53 .
finally the matrix element of this operator between kaon and two - pion states vanishes at the lowest order in chiral perturbation theory , and it was estimated in @xcite using the chiral quark model .
i have arbitrarily assigned a factor of two uncertainty ( 0.52 ) as a guess.@xmath54 in their notation and have emphasized that the sign of the matrix element is model - dependent . ]
note that the exact proportionality between yukawa and @xmath38 matrices assumed in the minimal supergravity automatically makes the off - diagonal element vanish after diagonalizing the yukawa matrix and hence no flavor - changing effects .
this is one of the reasons why the lr mass insertion has been regarded unimportant in the literature .
but such an exact proportionality is a strong assumption with no justification .
the constraint from @xmath19 on lr mass insertion is an order of magnitude weaker : @xmath55 @xcite .
this phenomenological fact that the constraint from @xmath19 still allows an interesting lr contribution to @xmath0 has been known ( see , _
e.g. _ , @xcite ) .
but the natural size of @xmath56 has not been discussed and it was always assumed to be smaller than the size estimated here due to various prejudices partly because of the constraint from the neutron electric dipole moment ( see discussions later ) .
= 0.45 = 0.45 the reason why this supersymmetric contribution is competitive ( or larger ) than the standard model contribution is basically because of the larger km factor as seen in the following back - of - envelope estimate .
the estimate of the supersymmetry contribution is given by @xmath57 and is suppressed only by the cabibbo angle in the lr mass insertion . on the other hand the standard model qcd penguin
is suppressed because all three generations should participate for cp - violation : @xmath58 @xmath59 . to compare them
, @xmath60 in the supersymmetric contribution can be regarded as an @xmath25 parameter in kaon decays .
the mass scale is higher @xmath61 , but the km factor is larger by 100 .
however the matrix element of the chromo - electric dipole moment operator vanishes at the leading order in chiral perturbation theory @xcite , resulting in about a factor of ten suppression , which brings the supersymmetry contribution roughly on par with the qcd penguin . finally , the qcd model penguin is partially canceled by the electroweak penguin .
this leaves the supersymmetry contribution important in @xmath0 .
now that we have found the new supersymmetric contribution to @xmath0 based on general grounds , it is important to ask what class of models actually gives such a contribution . because we are talking about flavor physics issues in supersymmetry here ( see @xcite for a good summary in this context )
, we need to look at models which naturally explain the hierarchy in fermion masses . the easiest example to explain is the model based on u(2 ) flavor symmetry @xcite . in this model ,
the first and second generations form a doublet under the u(2 ) flavor symmetry , and the third generation a singlet .
this makes the ll squark mass - squared matrix ( [ eq : ll ] ) an identity matrix and helps to keep the first and second generation scalars to avoid too - large flavor - changing effects in @xmath18 and @xmath19 .
the flavor symmetry allows the top yukawa coupling , but forbids down and strange yukawa couplings and hence needs to be broken by some vacuum expectation values ( vevs ) .
they introduce two such vevs , one in the symmetric tensor of u(2 ) @xmath62 and the other in the anti - symmetric tensor @xmath63 , which can generate the yukawa matrix @xmath64 where @xmath65 is the energy scale of the flavor physics .
the point here is that ( 2,2 ) and ( 1,2 ) elements originate from couplings to different fields @xmath66 and @xmath67 and hence their supersymmetry breaking counterparts can have different coefficients , giving in general @xmath68 in the lr mass matrix ( [ eq : a ] ) .
a similar situation occurs in the string theory , where the yukawa couplings @xmath34 are in general functions of so - called moduli fields @xmath69 : @xmath70 .
the moduli fields also acquire supersymmetry breaking vev ( @xmath71 ) , and the lr mass matrix is given by @xmath72 . generally , these two matrices are not proportional to each other , even though their structures are likely to be similar if the yukawa hierarchy is natural , _
i.e. _ , stable under small changes of the moduli
. a particular realization of this idea is an attempt to generate yukawa hierarchy as a consequence of a somewhat large compactification radius of an orbifold and fields of different generations belonging to twisted sectors of different fixed points .
then the modular invariance requires yukawa couplings to be proportional to powers of dedekind eta function @xmath73 : @xmath74 ( @xmath75 is the modular weight ) @xcite .
this mechanism can generate a yukawa hierarchy because @xmath76 for a large @xmath69 . then the derivative picks the power @xmath77 which is generation dependent and hence @xmath78 again .
the flavor - changing effects from the ll mass insertion , however , is a problem in general string models .
i have to emphasize that not all supersymmetric models give this type of contributions to @xmath0 .
for instance , the models with gauge - mediated supersymmetry breaking ( see @xcite for a review ) give either vanishing @xmath38-matrix or one proportional to the yukawa matrix and hence does not give interesting @xmath0 .
if the cabibbo angle originates in the up sector , the naive analysis here also fails .
also , the @xmath38-matrix may be real .
all i can argue here is that we generically expect supersymmetric contribution to @xmath0 as in ( [ eq : epsilonp ] ) , but there are exceptions .
see @xcite for more discussions on model - dependence .
since supersymmetry can give a contribution to @xmath0 or an interesting size , we should now ask if there are other consequences of this . the immediate question is the neutron electric dipole moment . again using the mass insertion technique , the current constraint is @xmath79 @xcite , while my estimate is @xmath80 .
if we are interested in a large supersymmetric contribution to @xmath0 , it is natural to expect a phase of order unity in @xmath81 as well .
this basically saturates the current upper limit for a phase of order a third , and an improve experimental bound would be extremely interesting .
the phase in @xmath81 , however , is in principle unrelated to that in @xmath82 , and in fact some models give a real @xmath81 despite a complex @xmath82 @xcite .
a more exciting aspect is that hyperon cp violation arise from the same chromo - electric dipole moment operator which generates supersymmetric @xmath0 and in principle there is a perfect correlation .
unfortunately the matrix element uncertainties obscures the correlation but pakvasa argued that the hyperon cp violation should appear within the sensitivity of the current hypercp experiment if @xmath0 is saturated by supersymmetry in the way i described @xcite . it would the most direct test of this possibility .
another important question is what the corresponding flavor - changing effects in the lepton sector would be .
in fact , we found a new contribution to @xmath83 which was also generally missed in the literature .
unfortunately the estimate is uncertain because we do not know the mixing angle in the lepton sector , but a reasonable guess is @xmath840.05 suggested by the small angle msw solution .
the new limit from mega collaboration @xcite @xmath85 ( 90% cl ) leads to the bound on the mass insertion parameter @xmath86 .
this is to be compared to the estimate @xmath87 .
note that this bound is on the _ magnitude _ of the parameter , not its imaginary part , and is hence more difficult to avoid .
one can regard this estimate as a lower bound on @xmath88 of about 500 gev , which is much stronger than previously discussed limits .
the electron electric dipole moment @xmath89 ( 90% cl ) translates to the limit @xmath90 , compared to the estimate @xmath91 .
the supersymmetric contribution again nearly saturates the constraint and a future improvement of the bound would be useful .
more recently , it has been suggested that there may also be enhancements in @xmath92 @xcite and @xmath93 @xcite .
see also discussions in @xcite .
overall , the naive estimate of the lr mass insertion i ve described gives interesting and important contributions to @xmath0 , @xmath1 , @xmath94 , and @xmath95 and those on the first two had been generally missed in the literature .
we reexamined the supersymmetric contribution to @xmath0 , and found that an important contribution had been missed in the literature .
the assumptions that went into the estimate are : ( a ) hierarchical yukawa matrix is probably controlled by certain flavor symmetries , ( b ) @xmath38-matrix is subject to the same control which makes its structure similar to the yukawa matrix but not necessarily exactly proportional to each other , ( c ) cabibbo rotation arises in the down sector , and ( d ) @xmath25-ish phase in the @xmath38-matrix . given these assumptions , we came up with the estimate eq .
( [ eq : epsilonp ] ) which surprisingly is in the ballpark of reported experimental values without any fine - tuning .
if @xmath0 is dominated by the supersymmetric contribution described here , it should give important contribution to the hyperon cp violation and possibly also to the neutron electric dipole moment .
correspondingly , there is an important contribution to @xmath96 and possibly also to the electron electric dipole moment .
i thank antonio masiero for collaboration . this work was supported in part by the u.s .
department of energy under contracts de - ac03 - 76sf00098 , in part by the national science foundation under grant phy-95 - 14797 , and in part by alfred p. sloan foundation . | i first motivate why we may want to look at possible new physics contributions to @xmath0 given relatively clear experimental but unclear theoretical situations .
i reexamine the supersymmetric contribution to @xmath0 and find an important one generally missed in the literature .
based on rather model - independent arguments based on flavor symmetries , an estimate of the possible supersymmetric @xmath0 is given , which interestingly come around the reported values without fine - tuning . if the observed values are dominated by supersymmetry , it is likely to give interesting consequences on hyperon cp violation , @xmath1 , and neutron and electron electric dipole moments .
august 23 , 1999 lbnl-44172 + ucb - pth-99/32 + hep - ph/9908442 + .1 in * can @xmath2 be supersymmetric ? * 0.3 in hitoshi murayama@xmath3 0.05 in @xmath4__department of physics , university of california + berkeley , ca 94720 - 7300 _ _ 0.05 in and 0.05 in @xmath5__theoretical physics group , ms 50a-5101 + lawrence berkeley national laboratory , berkeley , ca 94720 _ _ |
the late time acceleration of the universe can be explained by modifying general relativity ( gr ) on cosmological scales , avoiding the need of invoking a cosmological constant @xmath11 or an exotic repulsive fluid ( a.k.a . , dark energy ) .
many popular modified gravity ( mg ) theories rely on an extra scalar field @xmath12 to mediate a fifth force , making the distributions and motions of galaxies different from those predicted by gr ( see * ? ? ?
* and references therein ) .
in particular , the coherent infall of galaxies onto massive clusters will exhibit systematic deviations due to the enhanced gravitational forces .
@xcite ( 2013 , hereafter zw13 ) demonstrated that in @xmath13+gr simulations the velocity distribution of galaxies in the virial and infall regions of clusters ( hereafter abbreviated gik for galaxy infall kinematics ) is well described by a 2component velocity distribution model , which can be reconstructed from measurements of the redshift space cluster
galaxy cross correlation function , @xmath10 . in this paper
, we apply gik modelling to two suites of different mg simulations and investigate the possible signals of mg imprinted on the redshift space distribution of galaxies around clusters , using dark matter particles and halos as proxies for galaxies .
( for more general discussions of clusters as tests of cosmic acceleration theories we refer readers to @xcite , and for a succinct discussion of distinguishing mg from dark energy to @xcite 2009 . )
while deviations from gr may be welcomed on cosmological scales , a `` screening '' mechanism must be invoked in mg theories to recover gr in high density regions like the solar system , where gr has passed numerous stringent tests ( e.g. , * ? ? ?
* ; * ? ? ?
current viable screening mechanisms generally fall into two classes : * the chameleon like mechanism , in which the self interactions of the scalar field are regulated by a potential @xmath14 @xcite .
objects are screened when their gravitational potential @xmath15 is larger than @xmath16 , where @xmath17 is the cosmic mean of @xmath12 and @xmath18 is the coupling between matter and @xmath12 .
in other words , the effective scalar charge @xmath19 that responds to @xmath12 is reduced by the ambient gravitational potential .
this type of screening operates in @xmath0 @xcite , symmetron @xcite ) , and dilaton @xcite theories . in this paper
we will focus on the chameleon mechanism within @xmath0 , where the ricci scalar , @xmath20 , in the einstein
hilbert action is replaced by @xmath21 where @xmath0 is an arbitrary function of @xmath20 .
is tightly constrained by observations @xcite .
* the vainshtein mechanism , in which the self interactions of the scalar field are determined by the derivatives of @xmath12 , which suppress the scalar field and fifth force in high density regions @xcite .
scalar fields that exhibit vainshtein screening are generally called ` galileons ' because of an internal galilean symmetry @xcite . for an isolated spherical source
, the force transition happens at a characteristic radius @xmath22 ( called the vainshtein radius ) , where @xmath23 is the schwarzschild radius of the source and @xmath24 in models of interest is on the order of the hubble radius @xmath25 . within @xmath26 the scalar field
is suppressed ( @xmath27 ) , forming a `` sphere '' of screened region around the source .
this mechanism is at play in the dvali
porrati ( dgp , * ? ? ?
* ) and massive gravity @xcite theories . for our purpose , we simplify this class of model as a theory with a @xmath13 background cosmology and an extra galileon type scalar field that manifests vainshtein screening @xcite . in both the @xmath0 and galileon models ,
the maximum force enhancement is @xmath28 times the normal gravity , but the `` fifth force '' that produces this enhancement has different ranges in the two models .
since the chameleon scalar field becomes yukawa screened , the fifth force does not have infinite range , i.e. , it can not reach to cosmological scales .
galileons , however , are never massive , so their force has an infinite range , thus having a much larger impact on linear perturbation theory than chameleons do . in the local universe , however , the chameleon screening predicts a richer set of observational signatures that are detectable with astrophysical tests , because it is possible to have order unity violation of the macroscopic weak equivalence principle ( wep ) , i.e. , extended objects do not fall at the same rate as in gr @xcite . in environments of low background @xmath15 , objects with deep gravitational potential can self
screen , while those with shallow potential remain unscreened .
for example , @xcite estimated that there could be up to @xmath29 separation of the stellar disk ( composed of self screened objects ) from the dark matter and gas ( both unscreened ) inside unscreened dwarf galaxies , using orbital simulations under @xmath0 .
in contrast , there is no analogous order one violation in the vainshtein case , but the vainshtein `` spheres '' of individual objects interfere with each other .
for example , in a two body system where the separation is @xmath30 , the interference reduces the infall acceleration , and this reduction becomes most significant for two objects with equal masses @xcite .
the infall zone around clusters lies at the transition between the linear scale , where gravity is universally enhanced , and the local universe where gr is frequently recovered , providing a unique avenue for distinguishing mg from gr .
however , in both screening mechanisms the scalar @xmath12 is coupled to density fluctuations via a nonlinear field equation , which can only be solved jointly with the matter field using numerical simulations . @xcite
proposed a halo model - based approach to model the line - of - sight ( los ) velocity dispersions of galaxies in the infall region under both modified and normal gravities . in this study
we hope to provide a complete picture of the coherent motions and distributions of galaxies around clusters in the two mg theories and their systematic deviations from gr , fully taking into account the nonlinearities that are intrinsic to the chameleon and vainshtein mechanisms .
the gik model of zw13 describes the average galaxy infall in cluster
centric coordinates in terms of a 2d velocity distribution at each radius , comprising a virialized component with an isotropic gaussian velocity distribution and an infall component described by a skewed 2d _
t_-distribution with a characteristic infall velocity @xmath31 and separate radial and tangential dispersions .
the virialized component is confined within a `` shock radius '' that is close to the virial radius of the cluster , so in the infall region ( several to @xmath32 ) ] the gik model reduces to a single infall component .
zw13 demonstrated that gik profiles can be robustly recovered from the measurement of @xmath10 within the millennium simulation @xcite , and they applied the method to rich galaxy groups in the sloan digital sky survey ( sdss , * ? ? ?
in particular , the inferred @xmath31 profile provides a promising way of estimating the average dynamical mass of clusters and is likely insensitive to baryonic physics and galaxy bias ( zu et al . in prep ) .
although the zw13 method is calibrated against gr simulations , we will show that the gik model is also an excellent description of the infall behavior in mg simulations .
studies have shown that the peculiar velocities are more distinctively affected by modifications to gravity than the matter density field alone @xcite , and we expect the gik to be a particularly acute test of mg theories and their screening mechanisms .
a virtue of using galaxy dynamics in the outskirts of clusters is that independent information on the average cluster mass profiles can be robustly extracted from stacked weak lensing ( wl ) experiments @xcite .
since photons do _ not _ respond to the extra scalar field , lensing mass estimates will be different from dynamical mass estimates if gravity is modified from gr on relevant scales @xcite .
however , implementing this test requires measuring mass profiles on scales where screening is inefficient , i.e. , the cluster infall region rather than within the virial radius . for any given cluster sample detected by imaging , x - ray , or sunyaev-zeldovich ( sz ) experiments , one can
either directly compare the average lensing mass and the gik estimated dynamical mass or search for inconsistency between the measured @xmath10 and the gik predicted @xmath10 using lensing mass estimates and assuming @xmath13+gr .
alternatively , a simplified test can be performed when wl measurements are unavailable . since any volume
limited cluster sample is thresholded by some mass observable ( i.e. , galaxy richness , x - ray luminosity , or sz decrement ) that correlates with the true mass ( with some scatter ) , we can estimate the corresponding threshold in true mass using the abundance matching ( am ) technique .
however , the uncertainties of am may be large when the scatter in the mass observable true mass relation is large or / and the completeness of the cluster sample is low .
for the sake of simplicity , in this paper we concentrate on the am - based approach , and focus on cluster samples selected to have the same rank order in mass in the gr and mg simulations .
we present the results of gik modelling of the chameleon and galileon simulations separately in the paper , as the two simulation sets were run with different initial conditions , cosmic expansion histories , and box sizes and resolution .
we first introduce the chameleon simulations in [ sec : frsim ] , including the specific @xmath0 model , and halo statistics and kinematics .
[ sec : gik ] presents the results of gik modelling and the measurements of @xmath10 for the chameleon clusters and compares that to gr . in
[ sec : vain ] , we presents a similar set of results for the galileon clusters .
we summarize and discuss the future prospects of our method in [ sec : con ] .
for the chameleon modified gravity , we use a suite of large volume @xmath0 simulations ( @xmath33 box and @xmath34 particles ) evolved with the n - body code ecosmog @xcite .
the same set of simulations has been used to study the non - linear matter and velocity power spectra @xcite , redshift space distortions @xcite , and halo and void properties @xcite in @xmath0 gravity .
the large volume of the simulations allows us to derive robust statistics from a large number of massive clusters ( @xmath35 ) .
we do not make distinctions between the main halos and sub halos , but include all the bound groups of particles identified by the spherical density halo finder ahf ( amiga s halo finder , * ? ? ?
the halo mass is defined by @xmath36 , where @xmath37 is the overdensity for virialized halos ( @xmath38 at @xmath5 for typical @xmath13 cosmology ; * ? ? ?
* ) and @xmath39 is the mean density of the universe
. we will briefly describe the @xmath0 models here and refer the readers to @xcite for more details on the simulations .
the simulations adopt a specific @xmath0 gravity model introduced by @xcite , with the functional form of @xmath0 as @xmath40 where the mass scale @xmath41 , @xmath42 and @xmath43 are dimensionless parameters , and @xmath44 controls the sharpness of the transition of @xmath0 from @xmath45 in high curvature limit ( @xmath20@xmath46@xmath45 ) to @xmath47 in low
curvature limit ( @xmath20@xmath46@xmath48 ) .
the corresponding scalar field in @xmath0 theory is @xmath49 , commonly denoted as @xmath50 ( i.e. , the scalaron ) .
matching to the expansion history of flat @xmath13 universes ( e.g. , @xmath51 and @xmath52 ) requires a @xmath53 and a field value @xmath54 , where the subscript @xmath55 represents the present day values .
therefore , the hu & sawicki model is effectively described by two parameters : @xmath44 and @xmath56 .
models with @xmath57 are capable of evading solar system tests . for cosmological tests of chameleon theories , models with @xmath58 are ruled out by cluster abundance constraints from @xcite , and models with @xmath59 below @xmath60 are nearly indistinguishable from @xmath13 universes .
we shall study two cosmologically interesting @xmath0 models with @xmath61 and @xmath62 , @xmath63 , which will hereafter be referred to as f5 and f4 , respectively .
@xcite reported constraints of @xmath64 from other astrophysical tests , so these models may no longer be observationally viable , but they nonetheless provide useful illustrations of mg effects and a natural comparison for the vainshtein screening model discussed later .
the expansion history of the two @xmath0 simulations is matched to one flat @xmath13 simulation that evolves from the same initial conditions under normal gravity with @xmath51 , @xmath52 , @xmath65 , and @xmath66 ( referred to as the `` gr '' simulation ) .
the evolution of structure is the same in the three universes up to epochs around @xmath67 , which is the starting time of the simulations , since the fifth force in @xmath0 gravity is vastly suppressed until then . by studying the time evolution of the matter and velocity divergence power spectra in @xmath0 , @xcite showed that different @xmath0 models are in different stages of the same evolutionary path at any given time , and that varying the model parameter @xmath56 mainly varies the epoch marking the onset of the fifth force .
the exact epoch of onset in each model depends on the scale of interest , i.e. , at higher redshift for smaller scales ( see figure 8 in * ? ? ?
we choose to focus on the @xmath5 output of the simulations , mimicking the portion of the universe most observed by existing and near
term future redshift surveys that contain large samples of clusters .
figure [ fig : mf ] compares the halo mass functions from the gr , f5 , and f4 simulations .
the bottom panel shows the fractional enhancement of the halo mass function in the @xmath0 simulations relative to the gr one .
the shaded region ( @xmath68 ) indicates the mass range where the halo catalogs are incomplete and the bumps in the shaded area of the bottom panel are likely due to numerical effects . although sub halos were included in the halo catalogs , at @xmath69 the halo mass functions are mostly contributed by main halos .
the two @xmath0 models predict very similar halo abundances below @xmath70 , enhanced by @xmath71 over the gr abundances . for halos in this mass range ,
the fifth force was activated early enough that the structure formation has somehow converged in the two @xmath0 models . however , on the high mass end , the halo mass function in the f4 model shows even stronger enhancement over gr , while in the f5 model the number of halos becomes closer to the @xmath13 prediction with increasing mass .
the divergence of the halo mass functions predicted by the two @xmath0 models beyond @xmath72 indicates that the fifth force only started affecting the formation of massive clusters recently in the f5 model , producing smaller enhancement in the halo abundance compared to f4 .
for some very massive clusters in the f5 model ( e.g. , see figure 3 in * ? ? ?
* ) , the gravitational potential has begun to dominate the background scalaron , activating the chameleon mechanism to recover gr in the infall region , while in the f4 model the chameleon screening likely never activates anywhere in the universe .
figure [ fig : mf ] is in good agreement with the study of @xcite , where a series of smaller but higher resolution @xmath0 simulations are employed .
@xcite modeled the halo mass functions measured from the same sets of @xmath0 simulations using environment and mass dependent spherical collapse model in combination with excursion set theory . with neither
well resolved sub halos nor simulated galaxies in the simulations , the common prescription for constructing mock galaxy catalogs is through halo occupation distributions ( hods , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . however , the minimum required mass threshold for a complete halo sample in the chameleon simulations , @xmath73 , is overly high for any meaningful hod galaxies similar to the milky way and m31 would be absent .
therefore , we simply use particles and halos as our proxies for galaxies in the paper , and the behavior of hod galaxies should be intermediate between that of the `` particle '' galaxies and `` halo '' galaxies . since we are focused on the relative behavior of mg and gr simulations , the impact of choices for galaxy proxy on our conclusions should be small . we will usually refer simply to `` galaxies '' , when it is clear from context whether we are using particles or halos as our proxies . before
going into the statistical properties of galaxy infall , we hope to gain some intuitive understanding of the differences in infall between mg and gr by looking at figure [ fig : vflow ] .
taking advantage of the same initial condition shared by the @xmath0 and gr simulations , we locate at the frame center two primary clusters ( @xmath74 ) at @xmath5 that formed from the same seed in the initial density fluctuation field in the f4 ( red , for which the fifth force is stronger ) and gr ( gray ) simulations , respectively , and plot all halos with mass above @xmath75 as circles , with their relative velocities to the primary cluster indicated by the arrows .
the dark matter density field in the gr simulation is illustrated by the grayscale background , highlighting the three filamentary structures that funnel the infalling halos .
the radius of each circle is proportional to the halo mass , with the thick ones representing halos more massive than @xmath76 .
the length of each arrow is @xmath77 , corresponding to the redshift space displacement ( in units of @xmath78 ) that would be seen by a distant observer aligned with the velocity vector .
figure [ fig : vflow ] shows that the gik around _ individual _
clusters is highly anisotropic , and while the difference in the spatial distribution between gr and f4 halos is irregular and fairly mild , it is somewhat enhanced in redshift space . to avoid clutter ,
we do not show halos from the f5 model , which should display smaller differences from gr because of its smaller impact from the fifth force . in the next section
, we will show that by stacking individual frames like figure [ fig : vflow ] for clusters of similar mass , the anisotropy goes away and the average infall kinematics can be well described by the gik model proposed in zw13 , for both the @xmath0 and @xmath13 models .
more importantly , the enhanced difference in the redshift space between mg and gr models can be captured by systematic differences in the parameters of the gik model , namely , the characteristic infall velocity @xmath31 , the radial velocity dispersion @xmath79 , and the tangential velocity dispersion @xmath80 .
the halo mass functions in figure [ fig : mf ] suggest that the fifth force became unscreened earlier in f4 than in f5 , which gives it more time to affect the velocity field in the former . as a result , at @xmath5 clusters in both the f4 and f5 models should exhibit more enhanced galaxy infall compared to gr , but we expect the enhancement to be more substantial in the f4 model . for
the small number of screened clusters in the f5 model , the infalling galaxies around chameleon clusters should feel similar instantaneous accelerations as their counterparts around similar clusters in gr . however , the peculiar velocities of galaxies were enhanced by the fifth force when they were further away from the clusters . by the time they reached to the screened region ,
the peculiar velocities were already higher , so the infall stays stronger even though the underlying gravity recovers to gr .
our goal here is to quantify this modification to galaxy infall induced by @xmath0 gravities as function of @xmath56 within the framework of gik modelling . to compare the average gik among the three simulations , we select dark matter halos with @xmath81 in the @xmath13 simulation as our fiducial gr cluster sample , and those with the same mass range in the @xmath0 simulations as the `` equal mass '' ( em ) cluster samples .
as mentioned in the introduction , we also select specific halo samples in the @xmath0 simulations to have the same rank order in mass as the fiducial gr clusters , i.e. , the `` equal rank '' ( er ) cluster samples . the er clusters generally have slightly larger masses than the em ones , with @xmath82 and @xmath83 in f5 and f4 models , respectively .
the er sample thus resembles the set of clusters that formed from the same initial density peaks as the fiducial gr ones . in the limit of very rare , highly biased peaks , the large scale cluster bias @xmath84 , defined by the ratio between the cluster matter correlation function and the matter auto
correlation @xmath85 , is @xmath86 , yielding @xmath87 on large scales .
operationally , the em comparison would be most relevant to an observational study of clusters whose virial masses are calibrated by wl ( and thus accurate in both gr and mg ) . alternatively ,
if one ranks clusters by a mass proxy such as galaxy richness , x ray luminosity , or sz signal , then selects clusters above a threshold ( i.e. , am method ) , the er comparison is more relevant .
hereafter we simply denote the fiducial gr sample as `` gr '' while comparing it to the `` em '' and `` er '' samples in mg simulations .
as mentioned in the introduction , the two component gik model is an excellent analytic description of the joint 2d distribution of radial and tangential velocities of galaxies in the cluster
centric frame , @xmath88 . here
we will present the results of gik modelling for the cluster samples defined above , and refer the readers to zw13 for details on the gik parameterization and fitting procedures .
figure [ fig : vblob ] shows the best fitting @xmath88 for the gr ( top ) , f5em ( middle ) , and f4em ( bottom ) cluster samples at five different radial bins , using dark matter particles as proxy for galaxies .
negative @xmath89 indicates falling toward clusters . following zw13 ,
we define @xmath90 as the tangential velocity component that is projected in the plane of los axis and galaxy position vector in the cluster
centric frame ( see the 3d diagram in the figure 2 of zw13 ) .
since the _ average _ galaxy motion around the cluster center is isotropic , the probability distribution of @xmath90 is symmetric about zero .
hereafter we refer to @xmath90 simply as the `` tangential velocity '' .
note that the hubble flow is subtracted when defining @xmath89 , but it will be incorporated when modelling @xmath10 .
the gik of the three cluster samples in figure [ fig : vblob ] show some generic trends with radius : 1 ) the distribution has two distinct components on very small scales ( leftmost column ) but only shows a single infall component on large scales ; 2 ) the infall component is symmetric about the mean @xmath89 near the turn around radius , where infall velocity is comparable to hubble flow ( middle column ) ; 3 ) the @xmath89 distribution of the infall component is skewed toward positive velocities beyond the turn around radius ( two right columns ) , but is negatively skewed below that radius ( two left columns ) .
the impacts of modified gravity on @xmath88 are subtle but nonetheless visually apparent in figure [ fig : vblob ] when comparing the three models at the same radial bin ( i.e. , within the same column ) .
the solid vertical line in each panel indicates the most probable radial velocity , which shifts to more negative @xmath89 with increasing @xmath59 ( i.e. , from top to bottom ) at each radius .
simultaneously , the dispersions of the infall component in the radial and tangential directions also increase as function of @xmath56 the joint distributions in the f4 model are more extended than in gr .
this increased width results in the decreased peak amplitude of the distributions ( which are normalized to unity by definition ) . to quantify the differences in gik among the three simulations ,
we will focus on the impacts of modified gravity on three of the gik parameters ( @xmath31 , @xmath79 , and @xmath80 ) .
we ignore the other gik parameters defined by zw13 ( seven in total ) , including two parameters describing the virialized component that are irrelevant to this paper , and two others that describe the skewness and the kurtosis of the infall component , which we found to be insensitive to modified gravity .
note that the characteristic radial velocity @xmath31 is not the mean , median , or mode of the radial velocity distribution , but is a characteristic velocity naturally associated with the definition of the skewed _
t_-distribution used for describing the gik infall component ( see zw13 , equation 6 ) .
figure [ fig : v6p ] presents the best
fit gik parameters as functions of radius for the f5 ( left ) and the f4 ( right ) models , respectively . in each panel
we show the parameter profiles for all three types of cluster samples ( gr , em , and er ) .
the shaded region in each panel indicates the radial bins where gik is dominated by the virialized component and the fit to the infall component is less robust . for the f5 model , the halo mass function is close to that of @xmath13 at high masses , so the em and er samples have little difference in mass and their gik profiles look very similar
. however , as expected from figure [ fig : vblob ] , the two f5 cluster samples in the @xmath0 simulation show stronger infall ( top left ) and larger velocity dispersions ( center and bottom left ) compared to the gr sample . for the f4 model , the difference between the em and er samples is larger , especially in the @xmath31 profiles ( top right ) , where the difference between the er and gr profiles is almost double that between the em and gr ones . for both samples , the f4 results are more easily distinguished from gr than the f5 results , as expected .
since we anticipate that er comparisons will be more observationally relevant in most cases , we will focus henceforth on the gik of er cluster samples in the mg simulations .
we highlight the differences in gik profiles between er and gr clusters for both chameleon models in figure [ fig : vdiff ] .
the characteristic infall velocity profiles exhibit significant effects from the chameleon gravity on scales below @xmath91 , showing @xmath92 and @xmath93 enhancement at @xmath94 for the f5 and f4 models , respectively .
beyond @xmath91 the @xmath31 profiles converge to the gr prediction .
the dispersion profiles in @xmath0 models deviate from gr on all distance scales , with the differences almost constant and decreasing with radius for @xmath79 and @xmath80 , respectively .
the magnitude of the deviations we see here is very encouraging @xmath92 difference in both the @xmath31 and the dispersions is already detectable within @xmath95 in zw13 , where a preliminary gik constraint is obtained using two samples of sdss rich groups ( with group number @xmath96 and @xmath97 , respectively ) and the sdss dr7 main galaxy sample .
using dark matter particles as proxy for galaxies effectively assumes that galaxies have the same density profile and velocity distribution as dark matter particles within halos . in reality
, we expect central galaxies to have low peculiar velocities relative to the halo center of mass , and the spatial distribution may be less concentrated than the matter ( see , e.g. , * ? ? ?
* ) . to bracket the expectations for constraining gik using realistic galaxy samples
, we repeat the above experiment using halos instead of particles as proxy for galaxies .
this mimics the scenario where a luminous red galaxy ( lrg , * ? ? ?
* ) galaxy sample is employed , implying approximately one galaxy per halo as one extreme of the hod @xcite .
figure [ fig : vdiff_h ] summarizes the result of this experiment .
we denote the curves correspondingly as `` ` gr_h ` '' , `` ` f5_h ` '' , and `` ` f4_h ` '' in the figure .
the gik profiles are much noisier because of the rarity of halos , but the differences among the three samples are similar to those seen in figure [ fig : vdiff ] , but smaller in magnitude by about a factor of two .
we infer that the difference in gik seen in figure [ fig : vdiff ] using `` particle '' galaxies has approximately equal contributions from two sources , chameleon modifications to the random motions within halos ( i.e. , ` 1-halo' ) and the impact of chameleon gravity on the bulk inflow of halos ( i.e. ,
` 2-halo' ) .
quantitative predictions for a particular galaxy sample will require simulations that resolve the host halos and thus allow a full hod model of the population , incorporating both 1-halo and 2-halo effects with appropriate weight .
the redshift space cluster
galaxy cross
correlation function , @xmath10 , is a comprehensive characterization of the statistical relation between clusters and galaxies , influenced by both the real space cross
correlation @xmath98 and the peculiar velocities induced by the cluster gravitational potential .
mathematically , @xmath99 , a function of projected cluster galaxy separation @xmath100 and line - of - sight redshift separation @xmath101 , can be derived by convolving the real space @xmath102 with the hubble flow corrected los velocity distribution , which can be straightforwardly predicted from the gik model ( see equation 11 in zw13 ) .
zw13 demonstrated that the gik can be extracted from the observed @xmath10 , by taking advantage of the non degenerate imprint of each gik element on the 2d pattern of @xmath10 . in this section
we will examine the impact of chameleon gravity on the real and redshift space cluster
galaxy correlation functions in the @xmath0 simulations , using particles as proxy for galaxies . for measuring @xmath98 , we count the numbers of particles around clusters in spherical shells of successive radii , ranging from @xmath103 to @xmath104 with logarithmic intervals , then average over all clusters in each bin and normalize by the particle numbers expected in a randomly located shell of equal volume .
we measure @xmath10 in a similar way , counting galaxies in cylindrical rings of successive los distance @xmath101 for each projected separation @xmath100 ( assuming a distant observer approximation so that the los is an axis of the box ) .
uncertainties in both measurements are estimated by jackknife re sampling the octants of each simulation box .
we start by showing the real space cluster
galaxy correlation function @xmath98 for the gr , f4er , and f5er cluster samples in the left panels of figure [ fig : xi3d_ucurve ] . in the top left panel
, all three correlation functions exhibit a break at @xmath105 , marking the transition from the nfw like density distribution @xcite within halos to a biased version of the matter auto
correlation function on large scales @xcite .
the bottom left panel shows the ratio of @xmath98 between the er samples in @xmath0 simulations and the gr sample . on scales below the break radius , @xmath98 of
the er clusters show enhancement of @xmath106 and @xmath107 in the f5 and f4 models , respectively , because they are intrinsically more massive than their counterparts in the @xmath13 simulation . on scales larger than @xmath108 ,
the er clusters have nearly the same large
scale clustering as the gr sample .
since we are measuring the cross correlation with dark matter particles , these @xmath98 profiles also determine the cluster
galaxy wl profile ( see , e.g. , eq . 13 of @xcite 2012 ) . on intermediate scales
, there is a bump at @xmath109 in the ratio between @xmath98 of the f4er and the gr samples , but not in the ratio curve for f5er . in the f4 model ,
the fifth force became unscreened from quite early time , so that the velocity field has been enhanced for a long time by @xmath5 .
this means that the peculiar velocities are enhanced by roughly the same factor @xmath110 as the gravitational force , where @xmath110 is between unity and @xmath28 , and the kinetic energy of particles in f4 is thus @xmath111 times that in gr ; meanwhile , the gravitational potential in f4 is @xmath112 times as deep as in gr .
the net result is that the mg effect on the particle kinetic energy dominates over its effect on the cluster potential , so that as a compromise the particles tend to move toward the outer parts of clusters .
the situation is different in the f5 model , where the fifth force became unscreened quite late .
by @xmath5 , the fifth force has become unscreened but only for a short period , so that particle velocities have not been significantly affected by it .
on the other hand , due to the disappearance of the screening , the potential of the cluster suddenly became deeper .
the result is a stronger mg effect on the potential than on the kinetic energy of particles , and particles tend to move toward the inner parts of clusters .
these features have been observed in the halo density profiles of @xmath0 simulations before ( e.g. , * ? ? ?
* ) , and similar ones have been found in coupled quintessence simulations . using cluster wl measurements , @xcite exploited this small scale enhancement of cluster density profiles in chameleon gravity and obtained a constraint of @xmath113 at @xmath114 confidence level .
figure [ fig : xirs ] presents the redshift space cluster
galaxy correlation functions for the three cluster samples , showing a stronger small scale fingers - of - god ( fog , * ? ? ?
* ) effect with increasing @xmath59 , but with similar los squashing effect on large scales ( a.k.a . , kaiser effect , * ? ? ?
however , the most easily visible features of @xmath10 in figure [ fig : xirs ] are driven by the radial gradient of cluster density profiles , which are fairly insensitive to the influence of chameleon modifications to gravity according to the left panel of figure [ fig : xi3d_ucurve ] . modified gravity also changes the shape of @xmath10 at fixed @xmath100 via its effects on the gik , re distributing matter / galaxies along the @xmath101 axis . to reveal this los distortion of @xmath10 by @xmath0 , following zw13 , we compute the characteristic los distance @xmath115 by fitting a powered exponential function to @xmath10 at each fixed @xmath100 , @xmath116 where @xmath8 is the characteristic length scale at which @xmath10 drops to @xmath117 of its maximum value at @xmath118 .
the shape parameter @xmath119 yields a gaussian cutoff for @xmath120 and simple exponential for @xmath121 , though any value is allowed in the fit .
the results of this fitting are shown in the right panel of figure [ fig : xi3d_ucurve ] .
the @xmath8 vs. @xmath100 curves exhibit the characteristic u - shape discovered in zw13 fog stretching at small @xmath100 gives way to kaiser compression at intermediate @xmath100 which gives way to hubble flow expansion at large @xmath100 .
clearly , the @xmath10 distribution along the los is a sensitive probe of @xmath0 models at @xmath9 , becoming more extended with increasing @xmath59 ( e.g. , @xmath122 , @xmath123 , and @xmath124 at @xmath125 for gr , f5 , and f4 models , respectively ) .
the detailed shape of @xmath126 reflects a complex interplay among the four elements of the galaxy kinematics around clusters , including the three gik profiles of the infall component ( @xmath31 , @xmath79 , and @xmath80 ) and the virial component ( see figure 10 of zw13 for an illustrative experiment ) .
for the same reason , the increase of @xmath8 with @xmath56 has different origins at different projected distances .
below @xmath127 , the response of @xmath8 to @xmath56 is uniform with @xmath100 , caused by the uniform increase of dispersion in virial motions . for @xmath128 ,
the @xmath56 dependence of @xmath8 has two contributing sources , one being the increase of tangential velocity dispersions , the other the increase of maximum infall velocities , which transport high speed matter / galaxies from one side of the cluster in real space to the opposite side in redshift space ( i.e. , the portion of the fog effect caused by infall ) . for @xmath129 ,
@xmath56 influences the @xmath8 profile mainly via the increase of tangential velocity dispersions .
the radial velocity dispersions only enter into play at large @xmath100 , where the diagnostic power of @xmath10 is diminishing .
although chameleon theories like the hu & sawicki @xmath0 model include the phenomenology of @xmath13 without a _ true _ cosmological constant , @xcite proved that the theories that invoke a chameleon like scalar to explain cosmic acceleration essentially rely on a form of dark energy rather than a genuine mg effect , even if they are initially described in terms of an altered gravitational action .
conversely , the galileon class of theories is capable of accelerating the cosmic expansion even in the absence of any form of dark energy ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , i.e. , they have so called `` self accelerating '' solutions . here
we study the gik for a simplified version of such kind of galileon theories , where an extra galileon type scalar field that manifests the vainshtein mechanism permeates a universe with the @xmath13 background cosmology .
we employ a suite of galileon simulations with @xmath130 particles on a @xmath130 grid of @xmath131 , evolved using the particle mesh code of @xcite , which was updated by @xcite ( for other galileon / dgp simulations , see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the simulations were first used by @xcite for studying the statistics of matter clustering in real and redshift spaces .
we will briefly introduce the galileon implementation and parameterization here and refer the readers to @xcite for details .
as described in the introduction , the vainshtein mechanism has only one parameter , @xmath24 , which is interpreted as the compton wavelength associated with the graviton mass in massive gravity theories , so that the vainshtein radius of a point mass with mass @xmath132 is @xmath133 for extended objects of the same mass @xmath132 , the vainshtein radii are generally several times larger than @xmath134 e.g. , a nfw halo with @xmath135 has @xmath136 , and galaxies generally have @xmath137 . therefore , whereas the cluster interior below the virial radius belongs to the strongly vainshtein screened regime , the cluster infall region
is weakly screened , displaying complex interference among galileon fields sourced by the primary cluster and the infalling galaxies and galaxy groups .
we make use of three simulations with the same expansion history and initial condition , one of them a flat @xmath13 universe evolved under normal gravity with @xmath51 , @xmath52 , @xmath65 , and @xmath138 , and the other two immersed in galileon scalar fields with @xmath139 and @xmath140 . in the figures we refer to them simply as `` gr '' ( i.e. , @xmath141 ) , @xmath139 , and @xmath142 , respectively , and a smaller @xmath24 implies an earlier onset of the fifth force and an effectively stronger fifth force in the late universe ( see , e.g. , figure 5 of * ? ? ?
* for the change of linear growth rate as function of @xmath24 ) .
note that the `` gr '' simulation is different from what we used for comparing to the @xmath0 simulations , albeit with similar cosmology .
dark matter halos are identified via a spherical overdensity finder with the halo mass defined by @xmath143 , different from the @xmath144 used for the chameleon simulations . since @xmath145
, a halo would have a higher @xmath146 than @xmath144 ( e.g. , @xmath147 for a @xmath76 cluster at @xmath5 in @xmath13 ) .
the halo mass functions , halo bias functions , and matter power spectra of the three simulations can be found in @xcite .
targeting the same redshift range as in [
sec : frsim ] , we use the @xmath148 output of the simulations ( not @xmath5 due to the different sets of recorded epochs in the two suites of simulations ) .
since the volume of the galileon simulations is only @xmath149 of that of the chameleon simulations , we have to select samples with a wider mass bin size for robust gik measurements . for the fiducial cluster sample in the gr simulation
we include clusters with @xmath150 , and similar to [ sec : frsim ] we also select two er cluster samples from respective galileon simulations , with @xmath151 and @xmath152 in the @xmath139 and @xmath142 models , respectively . unlike [ sec : frsim ] , we do not show the results from the em cluster samples for the gr vs. galileon comparisons , as the relative difference between the em and er samples is similar to what we see in the chameleon simulations .
because of the smaller volume , we can only afford to use dark matter particles as proxy for galaxies .
note that the galileon cluster sample here comprises halos that are intrinsically smaller than the one used in [ sec : frsim ] , due to different overdensity thresholds used in mass definitions .
figure [ fig : vdiff_mark ] compares the gik profiles of the two er samples in the galileon simulations to that of the fiducial gr cluster sample .
as intrinsically less massive systems , the fiducial gr clusters show weaker infall velocities and velocity dispersions than the gr clusters used in the chameleon comparison at all distances ( e.g. , comparing the black curves in the top panels of figure [ fig : vdiff_mark ] to the black curves in figure [ fig : v6p ] ) .
however , the relative difference between the mg and gr samples overall looks very similar to what we see in the chameleon comparison ( e.g. , comparing the bottom panels of figure [ fig : vdiff_mark ] to figure [ fig : vdiff ] ) .
as expected , @xmath31 , @xmath79 , and @xmath80 all become stronger with decreasing @xmath24 ( i.e. , stronger fifth force ) .
specifically , @xmath31 shows @xmath153 and @xmath154 enhancement at @xmath2 for the @xmath139 and @xmath142 models , respectively , which are comparable to the fractional enhancements in the f5 and f4 chameleon models , respectively . the only major difference between figure [ fig : vdiff_mark ] and figure [ fig : vdiff ] appears in the bottom middle panel : the deviation of @xmath79 profiles from the gr prediction decreases as function of distance , from @xmath155/@xmath156 at @xmath157 to @xmath158/@xmath159 at @xmath2 , whereas in the chameleon comparison the deviation of @xmath79 stays more or less constant with distance
this difference in @xmath79 may be reflecting the different ranges of fifth force in the two models : the galileon has infinite range , so the force is enhanced further way from clusters , whereas the chameleon force is yukawa suppressed on large scales . however , the pattern of @xmath10 is insensitive to @xmath79 at @xmath9 where the @xmath10 measurement is the most robust , making it an unpromising tool for distinguishing the two mechanisms .
the gik profiles of the em cluster samples in the galileon simulations closely follow those of their corresponding er counterparts , albeit with slightly weaker amplitudes .
figure [ fig : xi3d_ucurve_mark ] compares the real space cluster
galaxy cross correlation function @xmath98 ( left ) and the characteristic los distance @xmath8 ( right ) measured for the three cluster samples .
the shaded region indicates the scales below the force softening length of the simulations , where the correlation function and velocity dispersions are artificially suppressed . on small scales , the ratios between @xmath98 of the galileon er samples and fiducial gr sample ( bottom left ) display similar features to these in figure [ fig : xi3d_ucurve ] , including an enhancement interior to the virial radius because the er clusters are more massive , and a bump around @xmath160 , though this is only @xmath161 away from the force resolution limit
to check whether the bump is a numerical artefact , we repeated the same measurements of @xmath98 using a suite of higher resolution ( but smaller volume ) galileon simulations and verified that the bump is physical . similar to the @xmath0 case in the f4 model , the fast transition from galileon force outside to normal gravity inside causes a sudden change of the depth of the potential well , but not the galaxy velocity ( or kinetic energy ) , which has been experiencing enhancement well before infall .
as the galaxies have excessive kinetic energy , they tend to move to the outer parts of halos , making the density profile lower in the central region of clusters and higher near the edges .
on large scales , the er samples in the galileon simulations exhibit stronger clustering than the gr clusters .
this enhancement in @xmath98 can be understood by starting from the findings of @xcite , which suggest that the halo mass function , the halo bias function , and the matter power spectrum of the galileon simulations are like those of @xmath13 universes with higher @xmath162 . by examining the kaiser effect , where changing @xmath162
can not mimic the large scale boost in the redshift
space clustering due to mg .
see @xcite for details .
] for example , at @xmath163 the @xmath139 and the @xmath142 simulations resemble the @xmath13 universes with @xmath164 and @xmath165 , respectively .
since the large scale @xmath98 of er clusters is almost linearly proportional to the effective @xmath162 , we observe @xmath166 and @xmath167 enhancement in @xmath98 at @xmath168 for clusters in the @xmath139 and @xmath142 models , respectively . for the redshift
space cluster
galaxy cross
correlation function , because the impact of the galileon field on gik is similar to that in the chameleon models , the @xmath8 curves in galileon simulations also exhibit similar deviations from the gr curves , as shown in the right panel of figure [ fig : xi3d_ucurve_mark ] .
the lack of upturn of @xmath8 at small @xmath100 is a consequence of force resolution suppressing velocity dispersions ; we expect that simulations with higher force resolution would show the characteristic u - shape for all three models .
the overall lower amplitude of @xmath8 compared to figure [ fig : xi3d_ucurve ] is again the result of selecting intrinsically less massive halos .
when we select the galileon clusters to have equal mass to the gr clusters , the enhancement in large
scale clustering ( left panel ) disappears because the large scale bias is steep function of mass , but the differences in gik and @xmath8 remain .
we have investigated the impact of modified gravity on the galaxy infall motion around massive clusters by applying the gik model developed in zw13 to two suites of @xmath0 and galileon n - body simulations .
both mg theories seek to explain cosmic acceleration by modifying gr on cosmological scales , but they recover gr in dense regions via two distinct `` screening '' effects : the potential driven chameleon mechanism in @xmath0 and the density driven vainshtein mechanism in galileon .
however , within the range of parameter space probed by our simulations ( i.e. , @xmath169 for @xmath0 and @xmath170 for galileon ) , despite having quite different cosmic growth histories , the two theories exhibit strikingly similar gik deviations from gr , with @xmath1 enhancement in the characteristic infall velocity at @xmath2 , and @xmath171 broadening in the radial and tangential velocity dispersions across the infall region , for clusters with mass @xmath172 at @xmath5 .
these deviations are detectable through gik modelling of the redshift space cluster
galaxy correlation function @xmath10 , especially when combined with cluster wl measurements .
we highlight the imprint of mg on @xmath10 using the characteristic u - shaped curve of @xmath8 , which increases by @xmath7 at @xmath9 from the gr prediction .
we find little difference between the gik profiles predicted by the two screening mechanisms , except for slightly different trends of the radial velocity dispersion with distance .
it is unclear whether the similar signature of these two distinct modified gravity theories on gik is a coincidence , or a generic result for any typical scalar
tensor theory that recovers the observed @xmath13like expansion history and reduces to normal gravity in the solar system and binary pulsars . in either case ,
our findings imply that , in combination with wl , galaxy infall kinematics offer a powerful non parametric cosmological test of modified gravity .
ongoing galaxy redshift surveys will provide large samples of clusters with good statistics for measuring @xmath10 and inferring gik out to large scales .
the main systematic uncertainty arises from the imperfect understanding of the impact of galaxy formation physics on gik . within the context of gik modelling and calibration ,
the infall behavior of realistic galaxies could differ from that of tracers in cosmological simulations ( e.g. , halos / sub halos in n - body simulations , post processed galaxies in semi analytical galaxy formation models , and simulated galaxies in hydrodynamic simulations , etc ; see @xcite 2013 ) .
however , we expect minimal impact on the characteristic infall velocity , which is our main tool of estimating the dynamical mass profiles of clusters , as any physical process that modifies galaxy kinematics within halos likely only adds scatter to the velocity dispersions rather than changing the mean . within the context of testing gravity ,
the effects of galaxy formation physics could be partly degenerate with those of modified gravity .
the observed properties of galaxies , including luminosity , morphology , colour , star formation , and clustering , are known to correlate with the environment ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and mg theories also rely on the environment to mediate the strength of the fifth force . for example , a sample of preferentially blue , star
forming galaxies may show enhanced infall velocities compared to a galaxy sample that is unbiased in colour in @xmath13 universes , and the enhancement could be mistaken as signal of modified gravity were the selection bias not properly accounted for .
we will investigate the potential systematic uncertainties induced by galaxy formation physics in a future paper , using mock galaxy samples constructed from different hod and semi - analytical model prescriptions .
redshift surveys that probe a range of galaxy types are especially valuable for those cosmological tests because one can check that different classes of galaxies lead to the same cosmological conclusions even though the galaxy samples themselves have different clustering and kinematics .
our gik modelling of @xmath10 is complementary to other semi analytical approaches based on the halo model @xcite , both seeking to model the velocity distribution around massive clusters for testing gravity ( also see * ? ? ?
* ; * ? ? ?
* for an alternative method of modelling galaxy redshift space distortion based on hod ) . the semi analytical velocity model adopted in @xcite has three components : the empirical infall velocity from the spherical collapse model , the halo - halo pairwise velocity distribution , and the intra - halo velocities ( assumed maxwellian with constant scatter ) .
while the model itself is highly informative , the accuracy is slightly lacking compared to simulation predictions .
we instead use the simulations as emulators for gik , trading more computer time for better accuracy in the prediction of our model . in terms of observational applications ,
our method differs from @xcite in two significant aspects .
first , they use the stacked redshift differences as the observable , but the model predicts the los velocity dispersion , so they are affected by the systematics in the subtraction of hubble flow in the 2-halo term ; in our method hubble flow is naturally incorporated in the calculation of @xmath10 .
second , they consider the velocity distributions up to the second moment , while we are able to model the entire @xmath88 including all higher moments .
established as one of the most powerful probes of dark energy , stacked wl analysis of clusters requires deep imaging surveys that can simultaneously yield lensed background galaxies and foreground cluster sample .
forecasts for stage iii and stage iv dark energy experiments predict cluster wl constraints that are competitive with supernovae , baryon acoustic oscillations , and cosmic shear ( see * ? ? ?
* and 8.4 ) . to complement wl as a cosmological test of gravity , gik modelling of galaxy clusters requires overlap with a large galaxy redshift survey , such as the ongoing baryon oscillation spectroscopic survey ( boss , *
* ) , its higher redshift successor eboss ( see * ? ? ?
* ) , and the deeper surveys planned for future facilities such as bigboss @xcite , despec @xcite , the subaru prime focus spectrograph @xcite , _ euclid _ @xcite , and _ wfirst _ @xcite .
we expect that , in combination with the stacked cluster wl analysis , the redshift space cluster galaxy cross correlations can reveal an accurate and complete picture of the average galaxy infall around clusters , allowing stringent tests of modified gravity theories for the origin of the accelerating expansion of the universe .
we thank lam hui and bhuvnesh jain for helpful discussions .
y.z . acknowledges the hospitality of the dept . of physics and astronomy at the university of pennsylvania where he enjoyed a fruitful discussion with participants of the `` novel probes of gravity and dark energy '' workshop .
d.h.w . and y.z .
are supported by the nsf grant ast-1009505 .
is also supported by the ohio state university through the distinguished university fellowship .
is supported by the royal astronomical society and durham university . e.j . acknowledges the support of a grant from the simons foundation , award number 184549 .
m.w . and e.j .
were partially supported by , and some of the numerical simulations reported here were performed on a cluster supported in part by , the kavli institute for cosmological physics at the university of chicago through grants through grants nsf phy-0114422 and nsf phy-0551142 and an endowment from the kavli foundation and its founder fred kavli .
we also acknowledge resources provided by the university of chicago research computing center .
was additionally supported by u.s .
dept . of energy contract de - fg02 - 90er-40560 . | infrared modifications of general relativity ( gr ) can be revealed by comparing the mass of galaxy clusters estimated from weak lensing to that from infall kinematics .
we measure the 2d galaxy velocity distribution in the cluster infall region by applying the galaxy infall kinematics ( gik ) model developed by @xcite to two suites of @xmath0 and galileon modified gravity simulations . despite having distinct screening mechanisms , namely , the chameleon and the vainshtein effects , the @xmath0 and galileon clusters exhibit very similar deviations in their gik profiles from gr , with @xmath1 enhancement in the characteristic infall velocity at @xmath2 and @xmath3 broadening in the radial and tangential velocity dispersions across the entire infall region , for clusters with mass @xmath4 at @xmath5 .
these deviations are detectable via the gik reconstruction of the redshift
space cluster
galaxy cross
correlation function , @xmath6 , which shows @xmath7 increase in the characteristic line - of - sight distance @xmath8 at @xmath9 from gr predictions . with overlapping deep imaging and large redshift surveys in the future
, we expect that the gik modelling of @xmath10 , in combination with the stacked weak lensing measurements , will provide powerful diagnostics of modified gravity theories and the origin of cosmic acceleration .
galaxy : clusters : general galaxies : kinematics and dynamics cosmology : large - scale structure of universe |
the proximity effect in a superconductor coupled to a mesoscopic normal conductor leads to a wide range of new quantum phenomena .
these include andreev reflection at normal and superconductor interfaces , the formation of andreev bound states ( abs ) in confined geometries , and proximity - induced supercurrent flow through normal conductors.@xcite the recent experimental detection of individual andreev bound states@xcite as well as the efforts towards demonstrating majorana states in superconductor coupled nanostructures devices with strong spin - orbit interaction@xcite received significant experimental and theoretical interest and opened a new area of research .
quantum dots ( qds ) coupled to superconducting leads provide an ideal system to test the theoretical predictions on the interplay of abs with the kondo effect.@xcite a 0 to @xmath0 junction transition@xcite has been reported in s - qd - s systems by measuring the sign reversal ( positive to negative ) of the josephson supercurrent for even to odd occupation of the qd.@xcite theoretical calculations suggest that this quantum phase transition in the s - qd - s josephson junction devices is signaled also by the crossing of two andreev levels.@xcite depending on the ratio of the kondo temperature @xmath1 at the center of a coulomb valley and the superconducting gap energy @xmath2 , the abs display a crossing ( @xmath3 ) or a non - crossing ( @xmath4 ) dispersion @xmath5 as a function of gate voltage @xmath6 .
these predictions have been confirmed by recent experimental studies using al - contacted semiconductor quantum dot / nanowire devices.@xcite it has also been predicted theoretically that there could be up to four abs for a single level model in the superconducting gap .
however the two outer abs may not be visible in the transport spectrum since they can merge with the continuum.@xcite so far most of the experimental studies of abs formation are limited to al as contacts material , which restricts these experiments to very low temperatures and magnetic fields . despite these earlier reports
, the study of hybrid nanostructures with larger gap superconducting elements is required to allow for a more complete understanding of abs formation in these devices . here
, we report on low temperature tunnelling spectroscopy measurements on an individual carbon nanotube quantum dot device strongly coupled to a nb superconducting loop and weakly coupled to an al tunnel probe .
two types of abs are observed in coulomb valleys with different charging energy . in some gate regimes
two pairs of abs are found within the superconducting gap .
in addition , next to the main abs conductance resonance a weaker conductance peak is present , which is interpreted as quasi - particle tunnelling from the abs to the al tunnel probe . at higher temperatures tunneling from the thermally populated upper abs becomes visible and shows an opposite curvature at the center of the coulomb valleys .
calculations based on the superconducting anderson model are used to describe the experimentally observed subgap features .
single wall cnts are grown on a si / sio@xmath7 substrate by chemical vapor deposition using a fe / mo based catalyst and methane as precursor gas .
the highly doped si substrate with 300 nm sio@xmath7 layer serves as a global back gate . ): a peak in the differential conductance is observed when unoccupied ( occupied ) states of the tunnel probe are aligned to the energy ( @xmath8 ) of the lower ( upper ) state of an abs pair .
, width=340 ] a niobium ( @xmath9 pd / @xmath10 nb ) superconducting loop and a tunnel probe ( @xmath11 ti / @xmath10 al ) are patterned using standard electron beam lithography on an individual single wall carbon nanotube ( figure [ fig : device](a ) ) .
a @xmath12 pd interlayer is used to improve the coupling between the superconducting fork and the nanotube and thereby increase the superconducting proximity effect . for weak coupling of the tunnel probe to the cnt a @xmath11 thin ti adhesion layer is used .
low temperature electrical transport measurements are performed in a @xmath13he/@xmath14he dilution refrigerator with heavily filtered signal lines down to @xmath15 .
the differential conductance @xmath16 of the superconducting tunnel probe weakly connected to the nanotube was measured employing conventional lock - in techniques by adding a low frequency ac excitation voltage ( @xmath17 , @xmath18 ) onto the dc bias voltage @xmath19 .
variation of the back gate voltage gives access to the abs spectrum in the cnt - qd .
the device parameters are extracted by measurements in both the superconducting and the normal conducting state ( by applying a magnetic field ) of the contacts . as function of back gate voltage
@xmath20 and bias voltage @xmath21 , for ( a ) @xmath22 and ( b ) @xmath23 .
while the superconducting energy gap in the contacts is clearly visible in ( a ) , in ( b ) superconductivity is suppressed and normal - state coulomb blockade behaviour emerges . , width=302 ] corresponding overview plots of the differential conductance are shown in fig .
[ fig : overview ] .
while in fig .
[ fig : overview](a ) at zero applied magnetic field features pertaining to superconductivity in the leads dominate transport at low bias , a magnetic field of @xmath23 restores the typical pattern of coulomb blockade in fig .
[ fig : overview](b ) .
the superconducting gap of the nb loop is found to be @xmath24 , and for the al tunnel probe @xmath25 and @xmath26 are estimated .
typical parameter values of the cnt - qd are the charging energy @xmath27 , the tunnel coupling to the leads @xmath28 , and the coefficient @xmath29 which relates the energies @xmath30 of the quantum dot states to @xmath6 .
and back gate voltage @xmath6 for two different gate regimes , ( a ) @xmath31 and ( b ) @xmath32 .
the sharp resonances are the signature of the andreev bound states : a main resonance of high conductance ( @xmath33 ) , a weak conductance resonance ( @xmath34 ) running parallel to the main resonance peak ( @xmath33 ) , and a third additional resonance close to the gap edge ( @xmath35 ) .
( c ) , ( d ) abs spectrum calculated using nrg for a two channel superconducting anderson model with parameters @xmath36 and @xmath37 , respectively .
inset : schematic representation of the theoretical model used to describe the main multilevel features , see text .
two independent transport channels connect the quantum dot with the superconducting reservoir .
, width=302 ] figure [ fig : stability ] shows two details of the stability diagram , where the differential conductance @xmath38 is again plotted as a function of bias @xmath21 and gate voltage @xmath20 , at base temperature and zero magnetic field .
the two panels ( a ) and ( b ) correspond to two different gate voltage ranges exhibiting different charging energies @xmath39 and @xmath40 , respectively . within the superconducting gap range @xmath41
there are three subgap features : two main resonances of high conductance ( @xmath33 , @xmath35 ) , and a weak conductance resonance ( @xmath34 ) running parallel to the resonance ( @xmath33 ) .
strong peaks in the differential conductance measurements are expected when an andreev level at @xmath42 is aligned to the bcs singularity of the density of states of the tunnel probe ( see figure [ fig : device](b ) ) .
this results in pronounced conductance peaks at voltages @xmath43.@xcite the weak conductance peak ( @xmath34 ) running parallel to the lower abs ( @xmath33 ) at lower bias voltages is understood as a replica of the abs at low temperatures ( see below ) . as already mentioned the abs ( @xmath33 ) spectrum for odd charge states can show non - crossing ( fig .
[ fig : stability](a ) ) or crossing behavior of the pairs of bound states ( fig .
[ fig : stability](b ) ) , resulting in a 0-@xmath0 quantum phase transition.@xcite this is controlled by the ratio @xmath44.@xcite in case of non - crossing the system always stays in the 0-phase , i.e. a ( singlet ) ground state whereas for a zero bias crossing of the abs the system changes its ground state from 0 to @xmath0-state ( magnetic doublet ) .
[ fig : stability](c ) and ( d ) show the abs spectrum obtained from the superconducting anderson model using the numerical renormalization group ( nrg ) method.@xcite we have found that the presence of two pairs of abs can be explained by assuming the presence of two nearly degenerate levels , e.g. , resulting from the often lifted kk-symmetry that are coupled to two independent channels in the leads , as schematically drawn in the inset of the figure .
for the case of fig .
[ fig : stability](a ) both channels are in the 0-phase and the abs exhibit a non - crossing behavior as correspondingly shown in fig .
[ fig : stability](c ) . on the other hand , in the case of fig .
[ fig : stability](b ) the inner abs exhibit a loop indicating the transition to the @xmath0-phase , which is accounted for in the theoretical result of fig .
[ fig : stability](d ) by a larger @xmath45 ratio for one of the channels .
the @xmath46 values were taken as 8 ( fig .
[ fig : stability](c ) ) and 9 ( fig . [ fig : stability](d ) ) close to the experimental estimations while the parameters @xmath47 were chosen to get a qualitative fit of the experimental results .
to understand the origin of the weak replicas of the inner pair of abs @xmath48 in our experiment , we investigate their dependence on temperature .
figure [ fig : temperature](a d ) shows a detail of the 2d stability diagram in a different coulomb valley at the indicated temperatures . in this gate regime ,
again a weak conductance resonance ( @xmath34 ) running parallel to a pair of abs ( @xmath33 ) is observed .
the apparent replica of the abs can be understood if we assume that the probe dos is finite at the fermi energy @xmath49 and @xmath49 is aligned with the abs : @xmath50 . ) at @xmath51 is observed when the probe dos is finite at the fermi energy @xmath49 and @xmath49 is aligned with the abs .
( b ) at higher temperatures also the upper abs ( @xmath52 ) at @xmath53 is thermally populated . in this case
a peak at @xmath54 ) is observed in the differential conductance , where an unoccupied ( occupied ) dos of the tunnel probe is aligned to a thermally populated upper ( lower ) abs .
, width=340 ] this level arrangement illustrated in fig .
[ fig : sketch](a ) . for a non - superconducting tunnel probe
one would expect the disappearance of the replicas together with a shift in the main abs due to the suppression of the probe gap @xmath55 .
we have verified this by applying a small magnetic field ( @xmath56 ) that drives the al tunnel probe into the normal state : it suppresses the replica and a clear crossing of the main abs ( @xmath33 ) is then present ( data not shown ) .
the main abs peak ( @xmath33 ) remains unchanged up to the maximum temperature @xmath57 investigated in the present study .
the satellite peaks ( @xmath34 ) also do not show any significant change up to @xmath58 . at @xmath59 , however , the dispersion of the gate voltage dependence of the satellite peaks flattens at the center of cb valley and an additional conductance resonance starts to emanate from the main abs resonance ( see the arrow in fig . [
fig : temperature](b ) ) . at even higher temperatures
@xmath60 , the gate voltage dependence of the satellite peak changes its curvature at the center of the coulomb valley compared to the low temperature ( @xmath61 ) case . as illustrated in the schematic of fig .
[ fig : sketch](b ) , we expect a peak in differential conductance when the maximum of the bcs density of states of the tunnel probe is aligned with the thermally populated upper state .
the position of the secondary peak ( @xmath52 ) should then be inverted , and vary as @xmath62 ) . at intermediate temperatures ( @xmath63 )
we observe that quasi particle tunnelling via both the lower and the thermally populated upper abs ( fig .
[ fig : temperature](b ) ) contributes to the current .
the combined contributions lead to a nearly flat dispersion of the secondary peak . at higher temperatures ( fig .
[ fig : temperature](c , d ) ) transport of thermally excited quasi - particles via the upper abs dominates , which leads to the opposite curvature of the dispersion in the center of the coulomb valley .
the slight shift of both the abs and their replica towards smaller energies results from the reduction of @xmath64 as the temperature is increased . according to our model
, the dispersion of all three types of conductance peak originates from the same dispersion relation @xmath5 .
hence , it should be possible to collapse the dispersion of the abs and its replicas at low and high temperature on top of each other by suitable inversions and shifts . for ( a ) @xmath65 and ( b ) @xmath66 .
right : plots combining the data of the left column graphs to demonstrate that the gate dependence of the peak positions can in each case be reduced to a single dispersion relation @xmath5 ; see the main text for details .
, width=302 ] this is illustrated in fig .
[ fig : gatedep ] , confirming our interpretation : at low temperature @xmath65 , fig .
[ fig : gatedep](a ) left panel , we observe in the differential conductance measurement the main abs at @xmath67 and a satellite peak at @xmath68 . after subtracting the sc gap of the tunnel probe ( @xmath69 ) at @xmath70 ,
the peak position of the main abs peaks show a good overlap with the satellite peak ( fig .
[ fig : gatedep](a ) , right panel ) . at higher temperature @xmath66 , as depicted in fig .
[ fig : gatedep](b ) left panel , the dispersion of the satellite peak is inverted .
its position follows a relation @xmath71 with gate voltage in the coulomb blockade valley .
to verify this gate dependence , we compute and plot @xmath72 = \pm [ 2{\ensuremath{\delta_\text{al}}}-\{{\ensuremath{\delta_\text{al}}}- { \ensuremath{\varepsilon_\text{abs}}}({\ensuremath{v_{\textrm{g}}}})\ } ] = \pm [ { \ensuremath{\delta_\text{al}}}+ { \ensuremath{\varepsilon_\text{abs}}}({\ensuremath{v_{\textrm{g } } } } ) ] = e v_\text{main}({\ensuremath{v_{\textrm{g}}}})$ ] for a reduced @xmath73 , and find a very good match with the observed main peak position @xmath74 , see fig .
[ fig : gatedep](b ) right panel .
a similar analysis for other temperatures also shows an excellent agreement with the proposed mechanism .
in addition to the measurement results , calculated conductance patterns for a quantum dot structure as described here are shown in fig . [
fig : temperature](e)-(h ) .
the conductance calculations were performed using a mean field description of the superconducting anderson model coupled to two superconducting leads with two different gap parameters @xmath75 and @xmath76 . as discussed in ref .
the magnetic phase can be represented within a mean field description by introducing an exchange parameter @xmath77 which produces a splitting of the dot energy levels for different spin orientations . in these calculations
the coupling to the nb superconducting leads is included nonperturbatively ( i.e. to all orders in perturbation theory ) while the coupling to the al probe is introduced to the lowest order in perturbation theory .
this approximation is justified by the weak coupling strength to the probe electrode , manifested by the small conductance values experimentally observed ( of the order of @xmath78 ) .
more precisely the conductance was calculated using @xmath79 where @xmath80 are the local densities of states at the probe electrode and at the dot coupled to the nb leads , respectively , @xmath81 is the fermi distribution function and @xmath82 is the tunneling rate from the dot to the probe electrode . while @xmath83 is calculated using the mean field approximation for the anderson model coupled to the nb superconducting lead as described in ref . ,
for @xmath84 we use a standard bcs - like density of states in which we introduce a phenomenological dynes - parameter@xcite of @xmath85 broadening the bcs density of states for the al probe . as can be observed in fig .
[ fig : temperature ] , the model calculations give a good description of the evolution of the weak subgap features with temperature , and once more confirm our interpretation in terms of different transport mechanisms .
it should be noted that the only change in the model parameters when going from fig .
[ fig : temperature](e ) to fig .
[ fig : temperature](h ) is due to the reduction of @xmath64 with temperature , while we assume @xmath86 in this temperature range .
the comparatively high value of the dynes - parameter is at present not understood .
this may be an effect of the electromagnetic environment,@xcite to which superconducting quantum dot devices are known to be extraordinarily sensitive .
( a ) and displaying a multi - loop pattern . ( b ) abs spectrum calculated by nrg assuming a single channel in the leads and using the parameters @xmath87 .
inset : schematic of the nrg model , see text .
, width=302 ] a peculiar feature of the present experimental results are the multiple loop abs that are observed in a more extended gate range , as depicted in fig .
[ fig : multiloop ] . as opposed to the situation considered in fig .
[ fig : stability](c ) and ( d ) , a theoretical description of such multi - loop features is obtained when more than one dot level couples to _ the same _ channel in the leads , see the inset of fig .
[ fig : multiloop](b ) .
a nrg calculation considering such a situation with two hybridized levels on one dot can actually reproduce qualitatively the observed multi - loop patterns , as shown in fig .
[ fig : multiloop](b ) . in several of the coulomb valleys investigated in more detail the abs - peaks are accompanied by a pronounced negative differential conductance ( ndc ) , shown in blue color in figs .
[ fig : stability ] , [ fig : temperature ] , and [ fig : multiloop ] .
ndc features have been predicted to appear due to the presence of the so - called yu - shiba - rusinov ( ysr ) states for a qd with an odd number of electrons@xcite with highly asymmetric coupling to the leads .
ysr - states can be regarded as a variant of abs appearing for a magnetic impurity coupled to a superconductor,@xcite and their existence has been experimentally confirmed via scanning tunneling microscopy of magnetic atoms on superconducting surfaces@xcite and qds with superconducting leads.@xcite
in conclusion , our transport spectroscopy of an individual carbon nanotube strongly coupled to wide gap nb leads reveals several different types of andreev bound state spectra .
weak satellite peaks appear within the smaller probe superconducting gap which are a result of quasiparticle tunneling into a residual density of states within this gap . at higher temperature
these satellite peaks change their dispersion as a function of gate voltage due to the thermal population of the upper state of an abs pair .
our findings are well reproduced within the superconducting anderson model in terms of combined nrg and mean field calculations .
more efforts in this direction could be helpful also to discriminate majorana bound states in similar hybrid nanostructures from other states at zero energy .
we gratefully acknowledge financial support from the deutsche forschungsgemeinschaft within grk 1570 , sfb 689 , the e. noether program ( hu 1808 - 1 ) , from spanish mineco through grant fis2011 - 26516 , and from the eu fp7 project se2nd .
a. k. thanks the alexander von humboldt foundation for providing financial support during this research .
we thank kicheon kang for enlightening discussions . | tunneling spectroscopy of a nb coupled carbon nanotube quantum dot reveals the formation of pairs of andreev bound states ( abs ) within the superconducting gap .
a weak replica of the lower abs is found , which is generated by quasi - particle tunnelling from the abs to the al tunnel probe .
an inversion of the abs - dispersion is observed at elevated temperatures , which signals the thermal occupation of the upper abs .
our experimental findings are well supported by model calculations based on the superconducting anderson model . |
recently orbital degrees of freedom in the strongly correlated electron systems have attracted much attention @xcite due to their crucial role in the stability of the various magnetic phases in the mott insulators @xcite or in the partial explanation of the mott metal - insulator transition itself @xcite .
typical examples of such systems are the transition - metal oxides with partly filled @xmath3 or @xmath0 degenerate orbitals .
the remarkable feature of these oxides is the large on - site coulomb repulsion @xmath7 , with @xmath8 being the effective electron hopping . in case of the stoichiometric ( undoped ) oxides the coulomb repulsion between electrons supresses charge fluctuations and leads to the effective low - energy @xmath9 superexchange ( se ) interactions between spin _ and _ orbital degrees of freedom .
in addition , the atomic hund s interaction @xmath10 aligns the spins of the electrons occupying the degenerate and almost degenerate orbitals on the same site and should be taken into account in the realistic se - type models .
interestingly , it is the hund s interaction which to large extent stabilizes particular spin ( magnetic ) _ and _ orbital order .
it should be stressed that though the @xmath3 orbitals usually order as in e.g. lamno@xmath6 @xcite , it is rather not the case of the @xmath0 orbitals which can form a disordered quantum orbital liquid ( ol ) as in e.g. lavo@xmath6 @xcite .
the different kinds of the orbital state are concomitant with various anisotropic magnetic phases such as @xmath11-af [ ferromagnetic ( fm ) planes staggered antiferromagnetically ( af ) in the other direction ] in lamno@xmath6 @xcite or @xmath2-af [ fm chains staggered af in the other two directions ] in lavo@xmath6 @xcite .
doping with holes destroys the insulating state , modifies the se interaction and hence can also modify the magnetic order and orbital state stabilized in the undoped case .
in such systems the motion of holes is strongly affected by : ( i ) intersite se interaction @xmath12 , ( ii ) on - site hund s interaction @xmath13 , which is captured by the kondo - lattice model . in the realistic limit in the transition metal oxides @xmath14
this model reduces to the double exchange model ( de ) @xcite .
however , for the above discussed class of the transition - metal oxides , orbital degeneracy needs also to be taken into account in such a model .
this was done in the seminal work of van den brink and khomskii @xcite for electrons with the @xmath3 degeneracy , leading to a naively counterintuitive picture of the de mechanism .
one can thus pose a question how the de mechanism would be modified in the case of the @xmath0 orbital degeneracy , as the se interactions for the undoped oxides differ qualitatively for the @xmath0 and @xmath3 cases .
hence , in this paper we want to answer three questions : ( i ) what is the nature of the magnetic order stabilized by the de mechanism for correlated electrons with @xmath0 orbital degeneracy , ( ii ) what is the nature of the orbital order , and ( iii ) how do these results differ from the @xmath3 and the nondegenerate case .
the paper is organized as follows . in the following chapter
we introduce the de hamiltonian with @xmath0 orbital degrees of freedom .
then we solve the model using two different mean - field approximations and we show that the metallic state coexists with the @xmath2-af order and ol state for the broad range of hole - doping .
next we discuss the results , i.e. : the validity of the approximations , the generic role of the @xmath0 orbitals in the stability of the above phases , the physical relevance of the model by comparison with the experiment .
the paper is concluded by stressing : the distinct features of the de mechanism via @xmath3 and via @xmath0 degenerate orbitals , and the crucial role of the coulomb repulsion @xmath1 .
we start with the realistic semiclassical de hamiltonian with @xmath0 orbital degrees of freedom , relevant for the hole - doped cubic vanadates @xcite : @xmath15 where : the hopping amplitude @xmath16 ; the restricted fermion creation operators @xmath17 , @xmath18 , where @xmath19 creates a spinless electron at site @xmath20 in @xmath21 orbital ; and @xmath22 are core spin @xmath23 operators of @xmath0 electrons in occupied @xmath24 orbitals .
we introduce the @xmath25 and @xmath26 variational parameters to be defined later .
the hamiltonian has the following features : ( i ) the first four terms describe the kinetic energy of the electrons in the degenerate @xmath27 and @xmath28 orbitals which can hop only in the allowed @xmath29 or @xmath30 plane to the nearest neighbour ( nn ) site @xmath20 providing there are no other electrons at site @xmath20 in these orbitals ( @xmath31 assumed implicitly ) , ( ii ) the last term describes the af coupling between core spins at nn sites in the @xmath32 plane due to the se interaction originating from the electrons in the always occupied @xmath24 orbitals ( @xmath33 during doping ) , ( iii ) the se interactions due to the itinerant electrons in @xmath27 and @xmath28 orbitals are neglected .
the variational parameters are defined as @xmath34 and @xmath35 , where @xmath36 @xmath37 is the relative angle between core spins in the @xmath38 @xmath39 direction .
this follows from the hund s rule which aligns the spins of the itinerant electrons with the core spins and which is not explicitly written in eq .
( [ eq:2 ] ) but enters via @xmath25 and @xmath26 parameters .
then the last term of eq .
( [ eq:2 ] ) can be written as : @xmath40 where @xmath41 is the number of sites in the crystal .
eq.([eq:2 ] ) together with eq .
( [ eq:3 ] ) shows the competition between the de mechanism allowing for the hopping in the underlying fm background with the se interaction supporting the af order .
the ground state of eq .
( [ eq:2 ] ) was found using two distinct mean - field approximations : hartree - fock , and mean - field applied to the formulation of the model eq .
( [ eq:2 ] ) in the kotliar - ruckenstein slave boson representation @xcite . to facilitate the treatment of the hole doped electronic systems we enlarge the original fock space , which is relevant for the hamiltonian eq .
( [ eq:2 ] ) , by adding a boson creation ( annhilation ) operator @xmath42 , which creates ( annihilates ) a charged hole .
then , in order to get rid of the nonphysical states in the enlarged fock space we need a _
local _ constraint : @xmath43 note that the boson controlling double occupancies is not introduced since such configurations are forbidden for @xmath31 . in order to solve eq .
( [ eq:2 ] ) with the constraint eq .
( [ constrainthf ] ) in a simple way , we introduce mean - field approximation for the above constraint and replace the following combination of fermion operators by averages : @xmath44 where in eq .
( [ mfb ] ) we use constraint eq .
( [ constrainthf ] ) , where @xmath45 and @xmath46 is the number of doped holes per site . besides , the inequalities in eq .
( [ mfa ] ) and ( [ mfb ] ) follow from the fact that we are discussing the hole - doping regime , i.e. with not more than one itinerant electron per site .
we call this mean - field approximation an hfa approximation since it is equivalent to decoupling the interaction term in an hfa way : @xmath47 and then making the @xmath48 limit [ where @xmath49 is the energy of the @xmath50 high spin state ] . note that the above used hfa does not allow for the alternating orbital ( ao ) order but we do not expect such an order to persist during doping .
this is because the kinetic energy of the electrons in the ao is rigorously zero for @xmath31 _ and _ the vanishing hopping between different orbitals @xcite .
then @xmath51 can be diagonalized and the total energy of the system in the mean - field approximation is : @xmath52 where the kinetic energy renormalizing factors follow from the hfa .
the sums go over the mesh of @xmath53 s belonging to the fermi volume @xmath54 which `` accommodates '' @xmath55 or @xmath56 electrons per site ( respectively ) and whose shape depends also on the band dispersion relations for the tight - binding model renormalized by the underlying magnetic order : @xmath57 since we are calculating separately the energies of the @xmath27 and @xmath28 bands we need an additional constraint for the fermi levels @xmath58 as the two bands fill up gradually : @xmath59 for fixed values of se @xmath60 and hole doping @xmath46 we minimise numerically the total energy eq .
( [ e_mf ] ) ( with the constraint eq .
( [ fermi ] ) ) with respect to the @xmath25 and @xmath26 variational parameters , which determine the magnetic structure .
the resulting magnetic phase diagram at @xmath61 is shown in fig .
[ fig : propermean](a ) . in fig .
[ fig : propermean](b ) we show the competing kinetic and magnetic ( se ) energies as a function of hole doping @xmath46 for the realistic value of se @xmath62 and obtained in the minimum of the total energy .
let us also note that for _ all _ of the discussed values of @xmath60 and @xmath46 it was found that @xmath63 , i.e. the number of electrons in both orbitals is the same and the ferrorbital order ( fo ) is not stabilized .
we rewrite the hamiltonian eq .
( [ eq:2 ] ) using kotliar - ruckenstein slave boson representation @xcite adopted to the orbital case .
so we enlarge the fock space by introducing three auxiliary boson fields and decouple the electron creation operator @xmath64 into a fermion creation operator @xmath65 carrying the orbital degree of freedom ( orbital flavour ) , a boson creation operator @xmath66 , and a boson annihilation operator @xmath67 carrying the charge degree of freedom ( i.e. @xmath68 creates a charged hole ) : @xmath69 where @xmath70 applying this to the hamiltonian @xmath51 we get : @xmath71 where we also need constraints to get rid of the nonphysical states in the enlarged fock space : @xmath72 next we introduce mean - field approximation : @xmath73 where we use eq .
( [ cons ] ) and introduce : @xmath74 the mean number of holes per site in orbital @xmath27 , @xmath75 the mean number of holes per site in orbital @xmath28 , and @xmath45 the mean number of holes per site .
furthermore , introducing parameter @xmath56 defined in the same manner as in eq .
( [ mfa]-[mfb ] ) yields @xmath76 and @xmath77 .
then the total energy of the system in the mean - field approximation is : @xmath78 where the electronic bands are defined in the same manner as in eq .
( [ e ] ) and the constraint eq .
( [ fermi ] ) also holds . as in the hfa
we deduce the resulting magnetic structures by minimising the total energy eq .
( [ e^sb2_mf ] ) with the constraint eq .
( [ fermi ] ) .
the resulting magnetic phases are shown in fig .
[ fig : properslave2](a ) and the competing energies at fixed se @xmath60 as a function of hole doping @xmath46 are shown in fig .
[ fig : properslave2](b ) .
we should also stress that the number of electrons in both orbitals does not change during doping and is equal to @xmath79 .
hence , the kinetic energy renormalizing factors are equal to @xmath80 , the well - known gutzwiller factors .
let us first discuss the validity of the two mean - field approximations .
for the hfa we infer that for small doping ( @xmath81 ) the correlated kinetic energy ( @xmath82 in fig .
[ fig : propermean](b ) ) does not tend to zero and instead equals 1/4 of the uncorrelated kinetic energy for @xmath83 .
this is a remnant of the fact that in the hfa the probability for existence of the hole at each site is equal to 25% .
thus , we conclude that the configurations with double occupancies are not suppressed and that hfa fails , at least in the low hole doping regime .
the reason why we can not use hfa , unlike in e.g. the undoped antiferromagnet , is that the orbital state , which is favoured here by the de mechanism , is not an ordered state .
if we were to get the fo or ao state , hfa would give physically relevant results .
we concentrate on the results obtained within kra which give the ( qualitatively ) right value of the correlated kinetic energy of the hamiltonian eq .
( [ eq:2 ] ) , cf .
[ fig : properslave2](b ) .
the @xmath2-af order and a concomitant one - dimensional ( 1d ) metallic state is stable for a broad range of the phase diagram parameters , in particular for the realistic value of se @xmath62 and @xmath84 hole doping .
this means that the de mechanism does not win in the @xmath32 planes , where the se mechanism takes over .
the first explanation of this phenomenon would be to ascribe the role of the driving force for the stability of the @xmath2-af order to the coulomb repulsion @xmath1 : the correlated kinetic energy ( @xmath82 in fig .
[ fig : properslave2](b ) ) is much smaller than the bare band energy ( @xmath85 in fig .
[ fig : propermean](b ) ) and hence the possible kinetic energy gain in the homogeneous fm state is much smaller in the correlated regime than in the conventional free electron regime .
however , for hole - doped manganites @xmath1 , @xmath86 and @xmath8 are of the same order as for hole - doped vanadates @xcite .
but in the manganites for @xmath87 hole doping the homogeneous fm order , concomitant with the metallic state , is stabilized mainly due to the de mechanism @xcite ! the only difference is that in the manganites the itinerant electrons hop between degenerate @xmath3 orbitals and in the vanadates between degenerate @xmath0 orbitals . besides , the af interaction between core spins is truely three - dimensional in manganites and not two - dimensional ( 2d ) as it is the case of our model .
thus the relative weakness of the de mechanism in the hamiltonian eq .
( [ eq:2 ] ) can be attributed to the specific features of the @xmath0 electrons : the strictly 2d and strictly flavor conserving hopping between these degenerate orbitals @xcite .
this generic role of the @xmath0 orbitals in the stability of the @xmath2-af order is threefold : \(i ) the af se interactions are rigorously 2d because for @xmath24 electrons the virtual hopping processes leading to the se interactions can happen only in the @xmath32 plane .
hence it makes possible : the hopping in the @xmath88 direction without destruction of the magnetic order , and thus the gain of the kinetic energy without any loss in the magnetic energy .
though this is rather a specific feature of our simplified se interactions than of the de mechanism itself it should be stressed that the physical setting in which we can have the de via @xmath0 degenerate orbitals is such that the af core spin interactions are 2d .
\(ii ) the degeneracy of the orbitals reduces the correlated kinetic energy and hence makes it energetically unfavourable to get homogenous fm order .
2(a ) of ref . @xcite where for the system without orbital degeneracy with holes doped _ only _ into @xmath28 orbital , i.e. with the hamiltonian : @xmath89 we obtain that for the realistic value of @xmath62 the fm order is already stable for hole doping @xmath90 .
note that this comparison is valid only if the degenerate orbitals are in the ol state which indeed is the case , cf .
discussion below .
\(iii ) since each of the @xmath0 orbitals has a _ different _ inactive axis [ i.e. the cubic direction in which the electrons can not hop ] this means that if one of these orbitals is `` engaged '' in the af se interactions [ cf .
( i ) ] than the other two orbitals have one _ active _ axis in the direction perpendicular to the plane with af bonds .
hence , hopping in the @xmath88 direction , assumed in point ( i ) , is indeed possible .
the orbital state which is concomitant with the obtained magnetic phases is the ol since : ( i ) we obtained @xmath91 for all of the magnetic phases , ( ii ) @xmath92 where @xmath93 is the @xmath94 component of the pseudospin orbital operator @xcite , ( iii ) @xmath95 for the @xmath46 and @xmath96 components of the pseudospin orbital operator @xcite .
the reason why none of the ordered state are realized in nature is twofold .
firstly , as already described , the ao state prohibits hopping due to the vanishing hopping integrals between different orbitals .
this mimics the confinement of holes in the neel af , but stays in contrast with the behaviour of the @xmath3 orbitals .
thus we can not gain any kinetic energy upon doping and this state is unphysical .
secondly , the fo state is also unphysical .
a priori , due to the pauli principle the electrons in the polarized state can better avoid each other than in the unpolarized one .
hence the correlated kinetic energy of the fo state is enhanced in comparison with the ol , cf .
( with @xmath97 ) of ref .
though , the condition of the equilibrium for the whole system yields that the chemical potentials ( fermi energies ) of the two fermionic subsystems ( one with electrons in @xmath27 orbitals and the other one in @xmath28 orbitals ) should be equal . for the fo state to be realized in nature
it means that the de mechanism should raise the bottom of one band so that the electrons can pour into the other one .
however , the de is too weak in our system and this does not happen .
as already pointed out our model describes the physics present in the hole - doped cubic vanadates such as la@xmath4sr@xmath5vo@xmath6 where the following phases were observed experimentally upon changing the hole doping @xmath46 in @xmath98 : @xmath2-af , ao and insulating one for @xmath99 hole concentration , @xmath2-af and metallic one for @xmath100 , and paramagnetic and metallic one for @xmath101 @xcite .
rather similar phases were observed experimentally in pr@xmath4ca@xmath5vo@xmath6 and nd@xmath4sr@xmath5vo@xmath6 @xcite .
hence our results stay in well agreement with the experiment , as we did not expect to explain by our model eq .
( [ eq:2 ] ) the filling control metal - insulator transition present in the system , but merely the coexistence of the metallic phase , the ol state and the @xmath2-af magnetic order in the above mentioned doping range . however , let us note that including the se interactions between the itinerant electrons would stabilize the 1d ol in the @xmath88 direction and ao in the @xmath32 plane @xcite . thus , if we included the realistic se in the hamiltonian eq .
( [ eq:2 ] ) we would get the insulating behaviour for finite hole doping due to the ao order in the plane which would act as a string potential for holes moving in the @xmath88 direction .
this suggests that further investigation of the _ even more realistic _ de model with @xmath0 orbital degrees of freedom _ and _ se interactions between itinerant electrons would not only uncover some new interesting physics but also could explain the experimentally observed phases in the doped cubic vanadates .
in summary , we have shown the distinct features of the _
double exchange via @xmath0 degenerate _ orbitals .
these features , attributed to the @xmath0 orbital degeneracy , are the following : \(i ) similar to the electron - doped @xmath3 degenerate orbitals @xcite the orbital degeneracy _ naturally allows _ for the coexistence of the anisotropic magnetic phases and the metallic phase .
this is a generic feature of the de mechanism via degenerate orbitals .
\(ii ) in contrast with the hole - doped @xmath3 case @xcite the homogenous fm order is not stable for large range of doping in the discussed here hole - doped regime of @xmath0 case .
this is due to the different geometry of the @xmath0 orbitals causing the hopping being strictly 2d and strictly flavour conserving , hence weakening the de mechanism .
\(iii ) upon hole - doping the orbitals form the ol state due to the weakness of the de mechanism . hence
, they do not order as it is typical for the electron - doped @xmath3 systems @xcite .
however , it should be stressed that the type of the orbital state for hole - doped @xmath3 systems is still under discussion @xcite . besides , the _ correlated _ behaviour of the itinerant electrons plays _ a crucial role _ in the stability of the obtained magnetic and orbital phases .
this is especially dramatic in the orbital sector , where the orbitals form the orbital liquid and the hfa [ suitable either for weakly interacting electrons or for the ordered states ] fails qualitatively in giving the right values of the kinetic energy .
altogether these are striking results which can serve as a basis for further investigation of the correlated phenomena in hole - doped systems with @xmath0 orbital degeneracy . *
acknowledgments * i am particularly grateful to a. m. ole for his invaluable idea and discussions .
i would like to acknowledge support by the polish ministry of science and education under project no . 1 p03b 068 26 . | we formulate the double exchange ( de ) model for systems with @xmath0 orbital degeneracy , relevant for hole - doped cubic vanadates . in the relevant regime of strong on - site
coulomb repulsion @xmath1 we solve the model using two distinct mean - field approximations : hartree - fock , and mean - field applied to the formulation of the model in the kotliar - ruckenstein slave boson represenation .
we show how , due to the relative weakness of the de mechanism via @xmath0 degenerate orbitals , the anisotropic @xmath2-type antiferromagnetic metallic phase and the orbital liquid state can be stabilized .
this is contrasted with the de mechanism via @xmath3 degenerate orbitals which stabilizes the ferromagnetic order in the hole - doped regime of manganites .
the obtained results are in striking agreement with the observed magnetic structures and the collapse of the orbital order in the doped la@xmath4sr@xmath5vo@xmath6 , pr@xmath4ca@xmath5vo@xmath6 and nd@xmath4sr@xmath5vo@xmath6 . |
the _ exponential map _ induces a surjection from the vector space @xmath0 of real _ skew symmetric matrices _ of order @xmath1 and the manifold @xmath2 of real _ special orthogonal matrices _ of the same order .
the study and the explicit computation of the fibers of this map are relevant subjects in matrix theory and in its applications ( see for instance @xcite ) . here
we analyse some of their differential - geometric properties .
the set of all _ real skew symmetric logarithms _ of @xmath3 ( i.e. the fiber over a matrix @xmath4 ) can be described in terms of the set @xmath5 of its _ real skew symmetric principal logarithms _ ( i.e. the real skew symmetric logarithms of @xmath4 with eigenvalues having absolute value in @xmath6 $ ] ) and of linear combinations with integer coefficients of suitable skew symmetric matrices ( theorem [ all_skew_symmetric_logarithms ] ) .
@xmath5 is ( implicitly ) studied in @xcite , where the so - called _ rodrigues exponential formula _ for skew symmetric matrices of order @xmath7 is extended to any order @xmath1 ( proposition [ rgx_exp_form ] ) ; we point out the role of the _ singular value decomposition _ of a skew symmetric matrix ( proposition [ svd ] and definition [ svd - system ] ) . also
the problems of the existence and of uniqueness of real skew symmetric principal logarithms of @xmath3 are approached in @xcite , while many differential - geometric properties of @xmath5 , discussed here , could be new : @xmath5 has a differential - geometric structure , depending on the presence of @xmath8 among the eigenvalues of @xmath4 ( proposition [ class_with_pfaff ] , theorem [ characterize_plog ] and corollary [ cor_characterize_plog ] ) . in particular in case of matrices having @xmath8 as an eigenvalue ,
it is diffeomorphic to the manifold of real skew symmetric orthogonal matrices of order equal to the multiplicity of the eigenvalue @xmath8 and has two connected components .
we are also able to describe all real skew symmetric logarithms of a matrix @xmath3 in some particular , but relevant , cases , where they form a discrete lattice of rank @xmath9 in @xmath0 ( theorem [ partskewsymmlogthm ] and remark [ rempartskewsymmlogthm ] ) .
the exponential map is involved in the description of the _ geodesic curves _ on the manifold @xmath10 of real _ orthogonal matrices _ equipped with the metric induced by the _ trace metric _ @xmath11 or by the _
frobenius metric _
@xmath12 ( recalls [ trace_metric ] ) . in @xcite
we have studied the trace metric on the whole manifold of real nonsingular matrices @xmath13 , where it defines a structure of semi - riemannian manifold .
this metric is often considered also in the setting of positive definite real matrices ( see for instance @xcite chapt.xii , @xcite 2 , @xcite chapt.6 , @xcite 3 ) , where it defines a structure of riemannian manifold .
on @xmath14 the trace metric @xmath11 is the opposite of the frobenius metric @xmath12 ( lemma [ g =- g - segnato ] ) .
@xmath10 is an einstein riemannian manifold whose main properties are listed in [ conseguenze_bi - invarianza ] .
moreover we get a suitable _ foliation _ on @xmath13 with leaves isometric to @xmath14 ( proposition [ foliations ] ) .
we describe the geodesic curves on @xmath14 with respect to @xmath12 ( and to @xmath11 ) ( proposition [ conseguenze_bi - invarianza ] ( b ) , remark [ princ_geod ] and proposition [ rat_numb_so_n ] ) and in particular the minimal geodesics joining @xmath15 and @xmath16 , which turn out to be in bijection with the skew symmetric principal logarithms of @xmath17 , furthermore we express the distance @xmath18 in terms of the eigenvalues of @xmath17 ( theorem [ minimalgeodesics ] and remark [ minimal_length ] ) . after computing the _ diameter _ of @xmath10 as @xmath19 ( corollary [ misura - diametro ] ) ,
we introduce the notions of _ weakly diametral pair of points _ and of _ diametral pair of points _ among the pairs of real orthogonal matrices ( definition [ weakly_diametral ] ) and characterize them in terms of the manifolds of real symmetric orthogonal matrices of order @xmath1 with the eigenvalues @xmath20 of multiplicity @xmath21 and @xmath8 of multiplicity @xmath22 , each one of them is diffeomorphic to the grassmannian of @xmath21-dimensional vector subspaces of @xmath23 ( propositions [ grassmannian ] , [ weakly_diametral_pairs ] and [ diametral_pairs ] ) . * acknowledgement .
* we want to thank the anonymous referee for many useful and precious suggestions about the matter and the writing of this paper .
[ initial_recalls ] in this paper all matrices are supposed to be square of order @xmath1 .
we denote by @xmath24 , @xmath0 , @xmath25 , @xmath14 and @xmath2 respectively the vector space of real matrices of order @xmath1 , its subspace of skew symmetric matrices , the multiplicative group of nondegenerate matrices of @xmath24 , the group of real orthogonal matrices and its special subgroup .
@xmath14 is a differentiable submanifold of @xmath13 of dimension @xmath26 with two connected components @xmath2 and @xmath27 ( the orthogonal matrices with determinant @xmath8 ) . as usual @xmath28 is the identity matrix of order @xmath1 and we put @xmath29 .
note that @xmath30 and @xmath31 ( @xmath32 is the transpose of the matrix @xmath33 ) are the unique real skew symmetric orthogonal matrices of order @xmath34 and that @xmath35 , where @xmath36 is the permutation ( orthogonal ) matrix @xmath37 .
analogously : @xmath38 , where @xmath39 denotes the _ block diagonal matrix _ with blocks @xmath40 .
next proposition collects some facts about the _ singular value decomposition _ of a skew symmetric matrix ; for general information on this subject we refer for instance to @xcite and to @xcite for a more geometric point of view .
[ svd ] every matrix @xmath41 has a unique singular value decomposition @xmath42 where @xmath43 are the distinct nonzero singular values of @xmath33 , while the matrices @xmath44 are in @xmath45 and are uniquely determined in this set by @xmath46 and by the conditions : @xmath47 for every @xmath48 and @xmath49 as soon as @xmath50 ( when @xmath51 , the second condition is vacuous ) .
it is well - known that every @xmath52 admits a unique singular value decomposition @xmath53 where @xmath43 are the distinct nonzero singular values of @xmath33 , and the matrices @xmath54 are uniquely determined in this set by @xmath46 and by the conditions : @xmath55 for every @xmath56 and @xmath57 as soon as @xmath58 .
we refer for instance to @xcite theorem 3.4 , for a proof of this statement . by the skew symmetry of @xmath33 and the uniqueness of the singular value decomposition
we get easily the skew symmetry of every @xmath59 s .
so the previous conditions about the @xmath59 s become equivalent to the conditions of our statement .
[ svd - system ] from now on , we will say that a set of nonzero skew symmetric matrices @xmath44 is an _ svd system _ if @xmath47 for every @xmath48 and @xmath49 as soon as @xmath50 .
the previous proposition justifies this definition .
the exponential map of a linear combination of an svd system has an interesting expression , we refer as rodrigues exponential formula : [ rgx_exp_form ] @xmath60
let @xmath61 be an svd system
. then @xmath62.\ ] ] note that @xmath63 are respectively the symmetric part and the skew symmetric part of the exponential of the skew symmetric matrix @xmath64 .
[ eigenvalues_b ] from standard facts on singular values we get that , if @xmath65 is a skew symmetric matrix with its singular value decomposition , then @xmath66 are simultaneously diagonalizable over @xmath67 , the eigenvalues of @xmath33 are @xmath68 both with the same multiplicity @xmath69 , for every k = 1, ... ,s , and possibly @xmath70 with multiplicity @xmath71 . moreover @xmath72 and @xmath73 , where @xmath74 are all the ( possibly repeated ) eigenvalues of @xmath33 .
[ eigenvalues_r ] in this and in the following section @xmath4 is a fixed matrix in @xmath2 .
since the eigenvalues of @xmath4 have absolute value @xmath20 , we can express them in this way : @xmath75 , both with multiplicity @xmath76 , @xmath77 , both with multiplicity @xmath78 , until @xmath79 , both with multiplicity @xmath80 , where @xmath81 $ ] are distinct and , in addition , possibly @xmath82 , with multiplicity @xmath83 , where @xmath84 , with the convention that , if @xmath85 is an eigenvalue of @xmath4 , then @xmath86 and in this case @xmath8 has multiplicity @xmath87 .
a _ real skew symmetric logarithm _ of @xmath4 is any real skew symmetric matrix @xmath88 such that @xmath89 .
a _ real skew symmetric principal logarithm _ of @xmath4 is any real skew symmetric logarithm of @xmath4 such that its eigenvalues have absolute value in @xmath6 $ ] .
we denote by @xmath5 the set of real skew symmetric principal logarithms of @xmath4 .
in particular @xmath90 consists only of the null matrix .
[ inequality ] assume the notations [ eigenvalues_r ] and suppose @xmath91 .
\1 ) @xmath92 is a skew symmetric principal logarithm of @xmath3 if and only if @xmath93 is its singular value decomposition and @xmath94 .
\2 ) for every skew symmetric principal logarithm @xmath88 of @xmath3 we have @xmath95 so @xmath96 is constant on @xmath5 .
\3 ) if @xmath33 is a real skew symmetric logarithm of @xmath3 , then @xmath97 with equality if and only if @xmath33 is a skew symmetric principal logarithm of @xmath4 .
\1 ) if @xmath98 , the eigenvalues of @xmath33 are complex logarithms of the eigenvalues of @xmath4 . therefore @xmath99 if and only if it is a real skew symmetric logarithm of @xmath4 and its eigenvalues are @xmath100 , both with multiplicity @xmath76 , @xmath101 , both with multiplicity @xmath78 , until @xmath102 , both with multiplicity @xmath80 , and @xmath70 with multiplicity @xmath83 .
this fact together with proposition [ svd ] gives the statement .
\2 ) it follows directly from [ eigenvalues_b ] .
\3 ) assume @xmath103 with @xmath104 .
we pose @xmath105 , next @xmath106 , until @xmath107 .
the eigenvalues of @xmath33 are necessarily of the form @xmath108 where @xmath109 @xmath110 for every @xmath111 .
we can assume that @xmath112 with @xmath113 and @xmath114 for every @xmath115 and , if @xmath86 , then @xmath116 for every @xmath117 .
standard computations show that @xmath118 and @xmath119 for every @xmath120 .
hence , for every @xmath121 , we have @xmath122 with equality if and only if @xmath123 . by [ eigenvalues_b ] we have @xmath124 where @xmath125 and the equality holds if and only if @xmath123 for every @xmath121 and @xmath126 for every @xmath127 , i.e. if and only if @xmath99 .
[ all_skew_symmetric_logarithms ] a real skew symmetric matrix @xmath33 of order @xmath1 is a real skew symmetric logarithm of @xmath3 if and only if there exist a skew symmetric principal logarithm @xmath88 of @xmath4 , an svd system of skew symmetric matrices @xmath128 all commuting with @xmath88 , and @xmath129 such that @xmath130 if @xmath131 as above , then by rodrigues exponential formula @xmath132 . for the converse , let @xmath33 be any real skew symmetric logarithm of @xmath4 and let @xmath133 , @xmath134 , @xmath135 , be the distinct nonzero eigenvalues of @xmath33 . by theorem [ svd ] , @xmath33 has a unique singular value decomposition : @xmath136 . for every @xmath121
, there is a unique pair @xmath137 with @xmath138 $ ] and @xmath139 such that @xmath140 , so , for every @xmath121 , we pose @xmath141 hence the nonzero distinct eigenvalues of @xmath33 are @xmath142 with @xmath143 where @xmath144 $ ] and @xmath145 and we get @xmath146 .
the eigenvalues of @xmath4 are the exponentials of the eigenvalues of @xmath33 , i.e. @xmath147 with @xmath148 $ ] and possibly @xmath149 . remembering the notations [ eigenvalues_r ] , up to reorder the indices
, we can assume that the first @xmath76 , among the @xmath150 s , are equal to @xmath151 , the next @xmath78 are equal to @xmath152 , until the last @xmath80 nonzero equal to @xmath153 and that the remaining @xmath154 are zero .
we set @xmath155 , @xmath156 , until + @xmath157 . then + @xmath158 . to conclude , we check that @xmath159 is a skew symmetric principal logarithm of @xmath4 .
we remark that the @xmath160 s form an svd system , indeed so are the @xmath161 s . furthermore @xmath88 commutes with any @xmath161 , hence @xmath162 ,
by rodrigues exponential formula and so @xmath88 is a real skew symmetric principal logarithm of @xmath4 , by proposition [ inequality ] ( 1 ) .
[ skew_symmetric_and_principal_logarithm ] if @xmath3 , then @xmath163 .
it follows from the previous theorem [ all_skew_symmetric_logarithms ] via the existence of real skew symmetric logarithms of @xmath4 ( see for instance @xcite iv , 2.2 ) .
[ explicitlogprin ] the existence of a real skew symmetric principal logarithm of @xmath4 is constructively obtained in @xcite lemma 2.4 . we resume here that construction .
first of all we observe that , for every @xmath164 , @xmath165 . as in @xcite cor.2.5.11 ( c ) , there is a real orthogonal matrix @xmath166 such that @xmath167 @xmath168 @xmath169 where @xmath170 + @xmath171 , until @xmath172 .
then @xmath173 is the real skew symmetric principal logarithm constructed in @xcite with its singular value decomposition . also its uniqueness ,
exactly when @xmath8 is not an eigenvalue of @xmath4 , is proved in @xcite thm .
precisely the following result holds [ system_r ] let @xmath174 with @xmath81 $ ] be any real skew symmetric principal logarithm of @xmath3 together with its singular value decomposition .
\1 ) if @xmath8 is not an eigenvalue of @xmath4 , then @xmath175 are uniquely determined by @xmath4 , so there is a unique real skew symmetric principal logarithm of @xmath4 .
\2 ) if @xmath8 is an eigenvalue of @xmath4 ( i.e. if @xmath176 ) , then @xmath177 are uniquely determined by @xmath4 and moreover also @xmath178 is uniquely determined by @xmath4 because we have @xmath179 see @xcite thm .
4.1 and its proof .
[ ossmatrices ] in order to characterize @xmath5 with @xmath3 , we must study @xmath180 : the compact subset of @xmath181 of skew symmetric orthogonal matrices of order @xmath182 . the lie group @xmath183 acts on @xmath181 on the left as @xmath184 and @xmath180 is the orbit of the matrix @xmath185 .
the isotropy group of @xmath185 is @xmath186 , where @xmath187 is the symplectic group of real matrices of order @xmath188 .
it is a standard fact that this intersection is isomorphic to the unitary group @xmath189 of complex matrices of order @xmath190 , then @xmath180 is a compact submanifold of @xmath181 diffeomorphic to @xmath191 ( see for instance @xcite p.1 ) .
its dimension is equal to @xmath192 [ class_with_pfaff ] with the same notations as in remark [ ossmatrices ] , the manifold @xmath180 is compact of dimension @xmath193 and its connected components are @xmath194 and @xmath195 : the sets of real skew symmetric orthogonal matrices of order @xmath188 with pfaffian equal to @xmath20 and @xmath8 respectively .
these components are simply connected manifolds both diffeomorphic to the homogeneous space @xmath196 . in particular if @xmath197 , each of them is a single point and if @xmath198 , they are both diffeomorphic to a @xmath34-sphere .
we refer for instance to @xcite for the _ pfaffian _ , @xmath199 , of a skew symmetric matrix @xmath200 .
every real skew symmetric orthogonal matrix of order @xmath188 can be written as + @xmath201 with @xmath202 .
now @xmath203 , so @xmath204 . hence the map @xmath205 defines a transitive action of @xmath206 on @xmath194 ( which is diffeomorphic to @xmath195 by means of a congruence with an orthogonal matrix of determinant @xmath8 ) .
this implies that @xmath207 , hence @xmath194 is connected , being @xmath206 connected . from the continuity of the pfaffian map , @xmath194 and
@xmath195 are the connected components of @xmath180 .
the case @xmath208 is trivial and already noted in recalls [ initial_recalls ] .
it is known that bott has computed the homotopy groups of order @xmath209 of such quotients ( see @xcite and also @xcite ) .
in particular the fundamental group is trivial and so @xmath194 is simply connected .
moreover , if @xmath210 , @xmath211 is homeomorphic ( therefore diffeomorphic ) to a @xmath34-sphere ; this case can be also deduced as a consequence of @xcite thm.1 .
[ characterize_plog ] assume the same notations and conventions introduced in [ eigenvalues_r ] .
\1 ) if @xmath8 is not an eigenvalue of @xmath3 , then @xmath4 has a unique skew symmetric principal logarithm .
\2 ) if @xmath8 is an eigenvalue of @xmath3 ( i.e. if @xmath176 ) with multiplicity @xmath212 and @xmath173 is any real skew symmetric principal logarithm of @xmath4 together with its singular value decomposition , then @xmath5 is the embedded submanifold of @xmath0 diffeomorphic to @xmath213 defined by @xmath214 part ( 1 ) is proposition [ system_r ] ( 1 ) .
when @xmath8 is an eigenvalue of @xmath4 , by proposition [ system_r ] ( 2 ) , @xmath215 are uniquely determined by @xmath4 . from this observation and from proposition [ inequality ]
we get that every real skew symmetric principal logarithm of @xmath4 has the form @xmath216 where @xmath217 and @xmath218 is an svd system with @xmath219 .
we remark that if a matrix has the form @xmath216 , with @xmath220 as above , then it is a real skew symmetric principal logarithm of @xmath4 .
indeed @xmath221 + 2 w^2 = \\
i_n + \sum_{j=2}^p[\sin\theta_j b_j + ( 1-\cos\theta_j)b_j^2 ] + 2 b_1 ^ 2 = r.\end{gathered}\ ] ] now let @xmath220 be a skew symmetric matrix with @xmath219 and such that @xmath218 form an svd system . as in @xcite 2 ( and as in remark [ explicitlogprin ] )
we can write @xmath222 ^ 2 = - k \ diag(i_{2m_1 } , \underbrace{0 , \cdots , 0}_{n-2m_1 } ) \
k^t $ ] with @xmath223 , so @xmath224 has @xmath8 as eigenvalue of multiplicity @xmath225 and @xmath70 as eigenvalue of multiplicity @xmath226 .
hence the eigenvalues of the real matrix @xmath220 must be @xmath227 both with multiplicity @xmath76 and @xmath70 with multiplicity @xmath228 .
since @xmath220 is skew symmetric , there exists a real orthogonal matrix @xmath229 such that @xmath230 ( see @xcite cor .
2.5.11 ( b ) ) . from @xmath219 , we get @xmath231
. therefore @xmath232 is an orthogonal matrix which commutes with the block diagonal matrix @xmath233 .
this is equivalent to say that @xmath232 is block diagonal too , say : @xmath234 . but
@xmath232 is orthogonal , so the blocks must be orthogonal matrices too of orders @xmath87 and @xmath228 respectively . hence @xmath235 now ( again by @xcite cor . 2.5.11 ( b ) ) @xmath236 is the form of a generic real skew symmetric orthogonal matrix of order @xmath87 , hence @xmath220 can be written as @xmath237 , where @xmath200 is a skew symmetric orthogonal matrix of order @xmath87 .
vice versa if @xmath238 , with @xmath200 as before , it is easy to check that @xmath239 , @xmath240 for any @xmath241 , @xmath219 and @xmath217
. moreover @xmath242 + 2w^2 = exp(\pi w + \theta_2 b_2 + \cdots + \theta_p b_p ) $ ] .
this implies that @xmath243 , defined by @xmath244 is a bijection .
we can conclude , because @xmath245 is clearly a @xmath246-embedding of @xmath213 into @xmath0 .
[ cor_characterize_plog ] assume that @xmath8 has multiplicity @xmath188 as eigenvalue of @xmath3 ( @xmath247 means that @xmath8 is not eigenvalue ) .
if @xmath247 , @xmath5 consists of a single point . if @xmath197 , @xmath5 consists of two distinct points
. if @xmath198 , @xmath5 is diffeomorphic to the disjoint union of two @xmath34-spheres .
if @xmath248 , @xmath5 has two mutually diffeomorphic connected components each of them has the following homotopy groups of order @xmath209 : @xmath249 it follows from theorem [ characterize_plog ] and from proposition [ class_with_pfaff ] , while the above homotopy groups of @xmath250 , said to be _ stable _ , are computed in @xcite ( see also @xcite ) . in the statement of the previous corollary
we have mentioned only the stable homotopy groups of @xmath250 .
some other results , about _
unstable _ homotopy groups ( i.e. of order @xmath251 ) , could be stated , by considering further studies since @xcite and @xcite .
[ partskewsymmlogrem ] we conclude this section by giving an explicit representation of the set of all real skew symmetric logarithms in some particular , but relevant , cases . from now on and until the end of this section , @xmath4 is a fixed matrix in @xmath2 such that all the following conditions hold : - its nonreal complex eigenvalues are all of multiplicity @xmath20 ; - either @xmath8 is not an eigenvalue or it has multiplicity @xmath34 ; - either @xmath20 is not an eigenvalue or it has multiplicity at most @xmath34 .
these are equivalent to @xmath252 and @xmath253 ( remember notations [ eigenvalues_r ] ) .
note that @xmath252 is equivalent to @xmath254 .
let @xmath255 be the real skew symmetric principal logarithm of @xmath4 , with its singular value decomposition , whose construction is recalled in remark [ explicitlogprin ] .
remember that there exists a matrix @xmath223 such that @xmath256 , @xmath257 .
when @xmath20 is an eigenvalue of @xmath4 with multiplicity @xmath34 ( i.e. if @xmath258 ) , we pose @xmath259 , @xmath260 . hence , if @xmath261 ( the integer part of @xmath262 ) , in any case we can write + @xmath263 . if @xmath264 the previous one is the unique real skew symmetric principal logarithm of @xmath4 , while if @xmath86 , the only two real skew symmetric principal logarithms of @xmath4 are @xmath265 and @xmath266 ( remember theorem [ characterize_plog ] and corollary [ cor_characterize_plog ] ) .
finally for @xmath267 , we have @xmath268 for @xmath269 .
[ partskewsymmloglemma ] with the same notations as in remark [ partskewsymmlogrem ] , assume @xmath270 and let @xmath271 defined by @xmath272 , where the @xmath273 s are real numbers such that @xmath274 as soon as @xmath50 when @xmath1 is even , and @xmath274 as soon as @xmath50 and @xmath275 for every @xmath121 when @xmath1 is odd .
then @xmath276 commutes with @xmath277 if and only if @xmath278 for some @xmath279 .
one implication is trivial .
assume first that @xmath280 .
we write the matrix @xmath281 as a block matrix with @xmath282 blocks : @xmath283 .
then the condition @xmath284 is equivalent to @xmath285 fixed @xmath286 , we denote @xmath287 , so the previous conditions give @xmath288 and therefore @xmath289 hence , for @xmath50 , we have @xmath290 and so easy computations give @xmath291 as soon as @xmath50 .
therefore @xmath292 . but @xmath281 is skew symmetric , so @xmath293 and we can conclude for @xmath1 even . when @xmath294 we pose @xmath295 where , for every @xmath296 , @xmath297 , while @xmath298 and @xmath299 are respectively column and row @xmath34-vectors and @xmath300 .
from @xmath301 , we get @xmath291 , for every @xmath302 , as above , and moreover , by looking at the entries @xmath303 and @xmath304 , we get also that @xmath305 and @xmath306 are both zero for every @xmath307 , taking into account that all @xmath308 s are nonzero . since @xmath281 is skew symmetric , we have @xmath309 , for every @xmath121 and @xmath310 .
this concludes the proof .
[ partskewsymmlogthm ] with the same notations as in [ eigenvalues_r ] and in [ partskewsymmlogrem ] , let @xmath3 and suppose that all the following conditions on @xmath4 hold : - its nonreal complex eigenvalues are all of multiplicity @xmath20 ; - either @xmath8 is not an eigenvalue or it has multiplicity @xmath34 ; - either @xmath20 is not an eigenvalue or it has multiplicity at most @xmath34 .
fix @xmath311 @xmath312 @xmath5 ( as in [ partskewsymmlogrem ] ) . if @xmath8 is not an eigenvalue of @xmath4 ( i.e. @xmath264 ) , then @xmath313 , while if @xmath8 is an eigenvalue of @xmath4 ( i.e. @xmath86 ) , then @xmath5 consists of @xmath314 and @xmath315 .
moreover @xmath104 is a real skew symmetric logarithm of @xmath4 if and only if @xmath316 for some @xmath317 .
in particular the set of all real skew symmetric logarithms of @xmath4 is a discrete lattice of rank @xmath9 in @xmath0 .
the first sentence follows directly from remark [ partskewsymmlogrem ] .
one implication of the equivalence is trivial by rodrigues exponential formula . for the other implication assume that @xmath33 is a real skew symmetric logarithm of @xmath4
then , by theorem [ all_skew_symmetric_logarithms ] , @xmath318 where @xmath319 , @xmath320 form an svd system of skew symmetric matrices all commuting with @xmath321 and @xmath129 . according to remark [ partskewsymmlogrem ]
, we can express the matrix @xmath321 as @xmath322 , where @xmath261 , @xmath323 for every @xmath324 and @xmath325 or @xmath326 .
standard computations show that , fixed @xmath327 , the condition @xmath328 is equivalent to @xmath329 .
hence , by lemma [ partskewsymmloglemma ] , we get @xmath330 , so @xmath331 , where @xmath332 . since @xmath333
, we get @xmath334 for every @xmath286 .
hence we have @xmath335 , where @xmath336 .
if we pose @xmath337 for @xmath338 , @xmath339 when @xmath340 and @xmath341 when @xmath342 , we obtain the formula @xmath343 , for some @xmath344 .
[ rempartskewsymmlogthm ] when @xmath345 , the hypotheses of the previous theorem are satisfied by every @xmath346 .
this implies that the set of real skew symmetric logarithms of every such @xmath4 is countable , discrete and closed in @xmath347 .
it is also possible to prove that the set of real skew symmetric logarithms of @xmath348 is a countable disjoint union of submanifolds of @xmath347 , all diffeomorphic to @xmath34-spheres , and of the point consisting in the null matrix .
we propose to investigate this subject in the context of a more general study about logarithms of matrices .
[ trace_metric ] this last metric induces a structure of symmetric semi - riemannian manifold on @xmath13 .
in particular we have described its _ levi - civita connection _
@xmath352 , its _ curvature tensors _ and characterized its _ geodesics _ as the curves of the type : @xmath353 for any @xmath354 and any @xmath355 ( @xcite thm .
the metric @xmath11 is an important object of study on the submanifold of @xmath13 of real positive definite matrices , where it defines a riemannian structure ( see for instance @xcite chapt.xii , @xcite 2 , @xcite chapt.6 , @xcite 3 ) . for ( c ) , again by the bi - invariance
, we have : @xmath366 where @xmath88 is the cartan - killing form of @xmath14 ( see for instance @xcite p. 324 ) .
now it is known that , if @xmath367 , then @xmath368 ( see for instance @xcite chapt.6 , 2 ) , so @xmath369 . therefore @xmath370 .
finally denoted by @xmath364 the scalar curvature of @xmath10 , we get that @xmath371 .
[ rivestimento_universale ] it is known that a @xmath7-dimensional einstein manifold has constant sectional curvature equal to @xmath372 ( see for instance @xcite p.293 prop.2 ) .
then , from proposition [ conseguenze_bi - invarianza ] ( c ) , we get that @xmath373 has constant sectional curvature @xmath374 , hence @xmath375 is a connected riemannian manifold with constant sectional curvature equal to @xmath20 and therefore it is isometric to @xmath376 with the fubini - study metric ( see @xcite chapt .
vi , thm . 7.10 ) .
( where the unions are taken in the set of all symmetric positive definite matrices ) are two foliations of @xmath13 , whose leaves are totally geodesic anti - riemannian submanifolds of @xmath377 , all isometric to @xmath361 .
indeed let @xmath356 and @xmath382 .
the geodesic on @xmath377 through @xmath15 with velocity @xmath383 is @xmath384 . but
@xmath385 for every @xmath386 , hence @xmath387 for every @xmath386 and this is equivalent to say that @xmath361 is totally geodesic in @xmath377 ( see for instance @xcite chapt.xiv cor.1.4 ) . for every fixed symmetric positive definite matrix @xmath378
, we have @xmath388 .
hence if @xmath389 , then standard computations show that the left translation @xmath390 is an isometry of @xmath377 onto itself , mapping @xmath14 onto @xmath391 .
@xmath26 , it follows that @xmath391 is a totally geodesic submanifold of @xmath377 of the same dimension .
moreover @xmath392 is also an isometry .
furthermore every @xmath393 belongs to the submanifold @xmath394 , because @xmath395 is a symmetric positive definite matrix .
therefore @xmath396 is a foliation of @xmath397 as predicted .
we can conclude the proof , because the transposition is an isometry of @xmath377 onto itself mapping @xmath391 onto @xmath398 for every positive definite matrix @xmath378 ( see @xcite prop .
let @xmath399 be in the same connected component of @xmath14 + ( so @xmath400 ) .
we call _ geodesic arc joining _ @xmath15 _ and _ @xmath16 _ in _ @xmath14 any geodesic curve + @xmath401 \to ( o_n , g)$ ] with @xmath402 , @xmath403 . as in @xcite cor
. 2.2 , from proposition [ conseguenze_bi - invarianza ] ( b ) we can get that there is a bijection between the geodesic arcs connecting @xmath399 and the real skew symmetric solutions of the matrix equation @xmath404 , i.e. the real skew symmetric logarithms of the matrix @xmath405 . [ princ_geod ] let @xmath407 ( with @xmath408 and @xmath409 for @xmath50 ) be the distinct nonzero eigenvalues of a given real skew symmetric matrix @xmath33 .
we know that there exists an svd system of skew symmetric matrices @xmath410 such that @xmath411 , and , moreover , @xmath412 $ ] ( remember theorem [ svd ] ) .
if @xmath414 is a geodesic arc joining @xmath418 , then @xmath33 is a skew symmetric logarithm of @xmath406 .
hence , by theorem [ all_skew_symmetric_logarithms ] , there exist a real skew symmetric principal logarithm @xmath88 of @xmath17 , an svd system of skew symmetric matrices @xmath128 all commuting with @xmath88 and @xmath129 such that @xmath419 thus the geodesic arc @xmath414 is @xmath420 where @xmath421 is still a geodesic arc joining @xmath399 corresponding to the real skew symmetric principal logarithm b of @xmath406 , we call _ principal geodesic arc _ , while the _ variation _ , @xmath422 , is a loop in @xmath24 at the null matrix . [ minimalgeodesics ]
let @xmath399 be in the same connected component of @xmath14 .
the curves of minimal length among the curves in @xmath10 , joining @xmath15 and @xmath16 , are precisely the principal geodesic arcs i.e. the curves
@xmath445 \to o_n$ ] of type @xmath446 , where @xmath88 is any skew symmetric principal logarithm of @xmath406 .
moreover , denoted by @xmath447 the distance associated to the metric @xmath12 , we have @xmath448 , where @xmath449 are the ( possibly repeated ) @xmath1 eigenvalues of @xmath406 and @xmath450 denotes the principal complex logarithm of the complex number @xmath190
. since @xmath451 and @xmath452 are geodesically complete , by hopf - rinow theorem ( see for instance @xcite chapt.5 , 2122 ) there are curves of minimal length among the curves joining @xmath399 ( belonging to the same connected component of @xmath14 ) ; they are geodesic arcs joining @xmath399 and have length equal to the riemannian distance @xmath18 .
a geodesic arc @xmath414 joining @xmath399 is a curve of type @xmath453 , with @xmath33 real skew symmetric logarithm of @xmath405 . finally , by [ eigenvalues_b ] , @xmath458 , where @xmath459 are the ( possibly repeated ) eigenvalues of any real skew symmetric principal logarithm of @xmath406 , hence we can conclude because @xmath460 for every @xmath121 . by theorem [ characterize_plog ]
, we get that if @xmath8 is not an eigenvalue of @xmath17 , then the geodesic of minimal length joining @xmath15 and @xmath16 ( belonging to the same connected component of @xmath14 ) is unique , while if @xmath8 is an eigenvalue of @xmath17 with ( even ) multiplicity @xmath87 , then the geodesics of minimal length joining @xmath15 and @xmath16 ( both in @xmath2 or in @xmath27 as above ) can be parametrized by means of the manifold @xmath213 diffeomorphic to @xmath463 .
[ grassmannian ] let @xmath464 be the set of real symmetric orthogonal matrices of order @xmath1 with the eigenvalues @xmath20 of multiplicity @xmath21 and @xmath8 of multiplicity @xmath22 .
then @xmath464 is a submanifold of the vector space @xmath465 of real symmetric matrices of order @xmath1 , diffeomorphic to the grassmannian @xmath466 of @xmath21-dimensional vector subspaces of @xmath23 .
in particular @xmath467 is diffeomorphic to @xmath468 .
analogously to remark [ ossmatrices ] and proposition [ class_with_pfaff ] , we denote by @xmath469 the left action of @xmath14 on @xmath465 . by spectral theorem , @xmath464 is the orbit of @xmath470 .
the isotropy group of @xmath471 consists of orthogonal matrices commuting with this matrix and , therefore , of block diagonal matrices belonging to @xmath472 .
again @xmath464 is a differentiable submanifold of @xmath465 diffeomorphic to @xmath473 ( see for instance @xcite p.1 ) .
we conclude , since this quotient is the expected grassmannian ( see for instance @xcite i , 4.12 ) .
[ diameter ] if @xmath104 , then @xmath478^t = exp(-a)$ ] .
so it is easy to show that the matrices @xmath399 ( as above ) form a weakly diametral pair if and only if @xmath406 is symmetric . hence @xmath399 form a weakly diametral pair if and only if @xmath479 for some @xmath480 , i.e. if and only if @xmath481 ( @xmath482 denotes the left translation ) or equivalently , by theorem [ characterize_plog ] , if and only if @xmath483 is symmetric with respect to the null matrix
. proposition [ grassmannian ] allows to obtain the following [ weakly_diametral_pairs ] let @xmath484 be as in proposition [ grassmannian ] and @xmath356 . then the set of @xmath485 such that @xmath399 form a weakly diametral pair , is the disjoint union @xmath486 , where every @xmath487 is a submanifold of @xmath14 , diffeomorphic to @xmath488 .
in particular @xmath399 is a diametral pair if and only if @xmath490 in case of @xmath1 even and if and only if @xmath491 in case of @xmath1 odd . in this last case
the set of points @xmath16 such that @xmath492 is a diametral pair is a submanifold of @xmath14 diffeomorphic to @xmath468 .
moakher maher , zra mourad , `` the riemannian geometry of the space of positive - definite matrices and its application to the regularization of positive - definite matrix - valued data '' , in _
imaging vis .
_ , 40 , no .
2 , 171187 . | we investigate the connections between the differential - geometric properties of the exponential map from the space of real skew symmetric matrices onto the group of real special orthogonal matrices and the manifold of real orthogonal matrices equipped with the riemannian structure induced by the frobenius metric . keywords .
skew symmetric and orthogonal matrices , singular value decomposition of skew symmetric matrices , pfaffian , exponential map , skew symmetric and principal logarithms , trace and frobenius metrics , riemannian manifolds , geodesic curves , diameter , pairs of ( weakly ) diametral orthogonal matrices .
mathematics subject classification ( 2010 ) : 15a16 , 53c22 . |
traversable wormholes are solutions to the einstein field equation that violate the classical energy conditions and are primarily useful as `` gedanken - experiments '' and as a theoretician s probe of the foundations of general relativity ( gr ) .
they are obtained by solving the einstein field equation in the reverse direction , namely , one first considers an interesting and exotic spacetime metric , then finds the matter source responsible for the respective geometry .
it is interesting to note that they allow `` effective '' superluminal travel , although the speed of light is not surpassed _ locally _ , and generate closed timelike curves , with the associated causality violations . in this rapporteur article , we consider a brief historical review ranging from the flamm - einstein - rosen bridge,@xcite to the geon wormhole - like structure obtained by wheeler in 1955,@xcite to the modern renaissance of wormhole physics.@xcite wormhole physics can originally be traced back to flamm in 1916,@xcite when he analyzed the then recently discovered schwarzschild solution .
one finds next that wormhole - type solutions were considered , in 1935 , by einstein and rosen ( er),@xcite where they constructed an elementary particle model represented by a `` bridge '' connecting two identical sheets .
this mathematical representation of physical space being connected by a wormhole - type solution was denoted an `` einstein - rosen bridge '' .
indeed , er were attempting to build a geometrical model of a physical elementary `` particle '' that is finite and singularity - free .
they based their discussion in terms of neutral and quasicharged `` bridges '' across a double - sheeted physical space . however , these can can easily be generalized.@xcite it is important to emphasize that at the time er were writing , the notions of `` coordinate singularity '' and `` physical singularity '' were not cleanly separated : it was supposed that the event horizon was the singularity . in this section
we follow ref . closely .
the neutral einstein - rosen bridge is an observation that a suitable coordinate change seems to make the schwarzschild ( coordinate ) singularity disappear , at @xmath0 .
more specifically , er discovered that certain coordinate systems naturally cover only two asymptotically flat regions of the maximally extended schwarzschild spacetime . to see this ,
consider the ordinary schwarzschild geometry : @xmath1 followed by a coordinate change @xmath2 , so that the line element is represented by the er form : @xmath3 with @xmath4 .
note that this coordinate change discards the region containing the curvature singularity @xmath5 .
the region near @xmath6 is interpreted as a `` bridge '' connecting the asymptotically flat region near @xmath7 with the asymptotically flat region near @xmath8.@xcite to justify the `` bridge '' appellation , consider a spherical surface , with constant @xmath9 , so that the area of the surface is given by @xmath10 .
the latter possesses a minimum at @xmath6 , with @xmath11 , which in modern terminology is defined as the `` throat '' , while the nearby region is denoted the bridge , or the `` wormhole '' .
thus , the neutral `` einstein - rosen '' bridge , or the `` schwarzschild wormhole '' , is identical to a part of the maximally extended schwarzschild geometry .
however , this wormhole is non - traversable , as the throat will pinch off before an observer may traverse the throat.@xcite relative to the quasi - charged einstein - rosen bridge , one starts off with the reissner - nordstrom metric : @xmath12 in order to obtain the bridge construction , er reversed the sign of the electromagnetic stress - energy tensor , which implies a negative energy - density .
@xcite considering @xmath13 , with a coordinate change @xmath14 , results in : @xmath15 which is indeed a very peculiar geometry , as it represents a massless , quasicharged object , with a negative energy density , possessing an horizon at @xmath16 or @xmath6 .
in fact , this was the object that er wished to interpret as an `` electron '' .
thus , the key ingredient of the bridge construction is the existence of an event horizon .
the er bridge is a coordinate artifact arising from choosing a coordinate patch , which is defined to double - cover the asymptotically flat region exterior to the black hole event horizon .
one may easily consider generalizations of these contructions ( see ref .
for more details ) . after the pioneering work by einstein and rosen , in 1935 , the field lay dormant for approximately two decades . in 1955
, john wheeler was beginning to be interested in topological issues in gr.@xcite it is interesting to note that wheeler considered a multiply - connected spacetime , where two widely separated regions were connected by a tunnel , which was denoted by a `` geon '' .
the `` geon '' concept denoted a `` gravitational - electromagnetic entity '' , which are hypothesized solutions to the coupled einstein - maxwell field equations . in modern language
the geon may be considered as a hypothetical `` unstable gravitational - electromagnetic quasisoliton''.@xcite furthermore , wheeler s concept can be used as the basis for building a model of nonzero charge with an everywhere `` zero charge density '' , where one of the tunnel mouths will `` send forth lines of force into the space , and appear to have a charge '' . while the other mouth `` must manifest an equal and opposite charge''.@xcite essentially , two routes of research were then available .
first , intensive research was dedicated to the classical dynamics of these tunnel configurations , assuming their possible existence .
second , the investigation of the quantum gravitational processes that might give rise to such configurations were explored . this led wheeler to propose the concept of `` spacetime foam''.@xcite in 1957 , misner and wheeler presented an extensive analysis , where riemannian geometry of manifolds of nontrivial topology was investigated with an ambitious view to explaining all of physics.@xcite indeed , this work was one of the first uses of abstract topology , homology , cohomology , and differential geometry in physics,@xcite and their point of view is best summarised by their phrase : `` physics is geometry '' .
this is also the first paper@xcite that introduces the term `` wormhole '' .
it is interesting to note that misner and wheeler considered that the existing well - established `` already unified classical theory '' allows one to describe in terms of empty curved space,@xcite the following concepts : _ gravitation without gravitation _ ; _ electromagnetism without electromagnetism _ ; _ charge without charge _ ; and _ mass without mass _ ( where around the mouth of the `` wormhole '' lies a concentration of electromagnetic energy that gives mass to this region of space ) .
the objective of misner and wheeler was essentially to use the source - free maxwell equations , coupled to einstein gravity , in the context of nontrivial topology , to build models for classical electrical charges and all other particle - like entities in classical physics.@xcite it is now known that this geon solutions are unstable , so that classically the tunnels will collapse to form black holes , which inevitably hides the interesting multiply - connected topology behind event horizons .
wheeler also noted the overwhelming importance of the planck scale in gravitational physics , where at distances below the planck length quantum fluctuations are considered to be extremely large , so that linearized theory breaks down and the quantum physics of full nonlinear einstein gravity must be faced.@xcite once the metric fluctuations become nonlinear and strongly interacting , one may expect that spacetime is flooded with a `` foamlike '' structure . in this context , the phrase `` spacetime foam '' is often referred to wheeler s suggestion that the geometry and topology of space might be constantly fluctuating .
an outstanding question is if these fluctuation may induce a change in topology , in order to form microscopic wormholes .
after the geon solutions devised by wheeler and misner , there is a thirty year gap between their original work and the 1988 morris - thorne renaissance of wormhole physics.@xcite in fact , despite the fact that considerable effort was invested in attempting to understand the `` spacetime foam '' picture and the `` geon '' concept , during this period the geonlike - wormhole structures seem to have been considered a mere curiosity and were inevitably relegated to a backstage . however , isolated pieces of work did appear , such as the homer ellis drainhole@xcite concept and bronnikov s tunnel - like solutions,@xcite in the 1970s .
it is only in 1988 that a full - fledged renaissance of wormhole physics took place , through the seminal paper by morris and thorne.@xcite as effective technology , traversable wormholes are greatly lacking , and the fundamental approach was to ask the question : `` what do the laws of physics permit ? '' in finding wormhole solutions , one adopts the reverse philosophy in solving the einstein field equation , namely , one first considers an interesting and exotic spacetime metric , then finds the matter source responsible for the respective geometry . in this manner
, it was found that some of these solutions possess a peculiar property , namely `` exotic matter '' , involving a stress - energy tensor that violates the null energy condition ( nec ) .
these geometries also allow closed timelike curves , with the respective causality violations.@xcite another interesting feature of these spacetimes is that they allow `` effective '' superluminal travel , although , locally , the speed of light is not surpassed .
these solutions are primarily useful as `` gedanken - experiments '' and as a theoretician s probe of the foundations of gr , and will be extensively reviewed throughout this review paper.@xcite
consider the following static and spherically symmetric wormhole solution @xmath17 where @xmath18 and @xmath19 are arbitrary functions of the radial coordinate @xmath20 .
@xmath18 is denoted the redshift function , for it is related to the gravitational redshift , and @xmath19 is denoted the shape function , as can be shown by embedding diagrams , it determines the shape of the wormhole.@xcite for the wormhole to be traversable it must have no horizons , which implies that @xmath21 , so that @xmath18 must be finite everywhere .
the coordinate @xmath20 is non - monotonic in that it decreases from @xmath22 to a minimum value @xmath23 , representing the location of the throat of the wormhole , where @xmath24 , and then it increases from @xmath23 to @xmath22 .
although the metric coefficient @xmath25 becomes divergent at the throat , which is signalled by the coordinate singularity , the proper radial distance @xmath26^{-1/2}}\,dr$ ] is required to be finite everywhere .
the proper distance decreases from @xmath27 , in the upper universe , to @xmath28 at the throat , and then from zero to @xmath29 in the lower universe . using the einstein field equation , @xmath30 ( with @xmath31 ) , we obtain the following stress - energy scenario @xmath32 \label{prwh}\,,\\ p_t(r)&=&\frac{1}{8\pi } \left(1-\frac{b}{r}\right)\left[\phi ' ' + ( \phi')^2- \frac{b'r - b}{2r(r - b)}\phi ' -\frac{b'r - b}{2r^2(r - b)}+\frac{\phi'}{r } \right ] \label{ptwh}\,,\end{aligned}\ ] ] where @xmath33 is the energy density , @xmath34 is the radial pressure , and @xmath35 is the lateral pressure measured in the orthogonal direction to the radial direction . using the conservation of the stress - energy tensor , @xmath36 , we obtain the following equation @xmath37 which can be interpreted as the relativistic euler equation , or the hydrostatic equation of equilibrium for the material threading the wormhole .
note that one now has three equations with five unknown functions of the radial coordinate .
several strategies to solve these equations are available , for instance , one can impose an equation of state,@xcite and consider a specific choice of the shape function or of the redshift function . to be a solution of a wormhole
, one needs to impose that the throat flares out .
using embedding diagrams,@xcite this flaring - out condition entails the following condition @xmath38 at the throat , we verify that the shape function satisfies the condition @xmath39 .
note that the above treatment has the drawback of being coordinate dependent . for a covariant treatment ,
we refer to the analysis outlined in refs . [ ]
. we will see below that this condition plays a fundamental role in the analysis of the violation of the energy conditions .
the nec asserts that for _ any _ null vector @xmath40 , we have @xmath41 . for a diagonal stress - energy tensor ,
i.e. , @xmath42 $ ] , this implies @xmath43 and @xmath44 . using the einstein field equations ( [ rhowh ] ) and ( [ prwh ] ) , evaluated at the throat @xmath23 , and taking into account the finite character of the redshift function so that @xmath45 , the flaring - out condition ( [ flareout ] ) imposes the condition @xmath46 .
this violates the nec .
in fact , it implies the violation of all the pointwise energy conditions.@xcite although classical forms of matter are believed to obey the energy conditions , it is a well - known fact that they are violated by certain quantum fields , amongst which we may refer to the casimir effect .
thus , the flaring - out condition ( [ flareout ] ) entails the violation of the nec , at the throat .
note that negative energy densities are not essential , but negative pressures at the throat , @xmath47 , are necessary to sustain the wormhole throat . as the violation of the energy conditions is a problematic issue , it is useful to minimize this violation .
several approaches have been explored extensively in the literature , namely , rotating solutions,@xcite evolving wormhole spacetimes,@xcite thin - shell wormholes using the cut - and - paste procedure,@xcite and modified theories of gravity.@xcite it is interesting to note that the violations of the pointwise energy conditions led to the averaging of the energy conditions over timelike or null geodesics.@xcite the averaged energy conditions permit localized violations of the energy conditions , as long as on average the energy conditions hold when integrated along timelike or null geodesics .
now , as the averaged energy conditions involve averaging over a line integral , with dimensions ( mass)/(area ) , not a volume integral , they do not provide useful information regarding the `` total amount '' of energy - condition violating matter . in order to overcome this shortcoming ,
the `` volume integral quantifier '' was proposed.@xcite thus , the amount of energy condition violations is then the extent that these integrals become negative .
a new set of energy constraints was introduced by ford and roman in 1995 , @xcite denoted by the quantum inequalities ( qi ) .
contrary to the averaged energy conditions , one does not average over the entire wordline of the observer , but weights the integral with a sampling function of characteristic width .
essentially , the inequality limits the magnitude of the negative energy violations and the time for which they are allowed to exist .
ford and roman applied the analysis to static and spherically symmetric morris - thorne wormholes and concluded that either the wormhole possesses a throat size which is only slightly larger than the planck length , or there are large discrepancies in the length scales which characterize the geometry of the wormhole.@xcite due to these results , ford and roman argued that the existence of macroscopic traversable wormholes is very improbable .
but , there are a series of considerations that can be applied to the qi .
firstly , the qi is only of interest if one is relying on quantum field theory to provide the exotic matter to support the wormhole throat .
but there are classical systems ( non - minimally coupled scalar fields ) that violate the null and the weak energy conditions,@xcite while presenting plausible results when applying the qi .
secondly , even if one relies on quantum field theory to provide exotic matter , the qi does not rule out the existence of wormholes , although they do place serious constraints on the geometry .
building on the ford - roman qis , martin - moruno and visser propsed classical and quantum versions of a `` flux energy condition '' ( fec and qfec),@xcite based on the notion of constraining the possible fluxes measured by timelike observers .
the naive classical fec were shown to be satisfied in certain situations , and even for some quantum vacuum states , while its quantum analogue ( the qfec ) was satisfied under a rather wide range of conditions . furthermore , martin - moruno and visser presented and developed several nonlinear energy conditions suitable for usage in the semiclassical regime .
more specifically , they considered the fec , and novel concepts such the `` trace - of - square '' ( tosec ) and `` determinant '' ( detec ) energy conditions , and showed that these nonlinear energy conditions behave much better than the classical linear energy conditions in the presence of semiclassical quantum effects.@xcite moreover , whereas the quantum extensions of these nonlinear energy conditions seem to be quite widely satisfied as one enters the quantum realm , analogous quantum extensions are generally not useful for the linear classical energy conditions.@xcite an interesting application of the energy conditions is to the buchert averaging.@xcite note that a key feature of buchert averaging is the realization that , either with spatial averaging or averaging over a suitably defined ensemble of spacetimes , the average of the einstein tensor is typically not equal to the einstein tensor of the average spacetime
. the discrepancy can be viewed as an `` effective '' stress - energy , one that often violates the classical energy conditions .
a particularly attractive example of this phenomenon arises when one considers spatial averages in a conformal - flrw ( cflrw ) cosmology or the ensemble average over conformal deformations of a specific flrw geometry .
these cflrw - based models are particularly tractable , and attractive for observational reasons , as the cmb is not distorted .
furthermore , it is possible to prove some rigorous theorems regarding the interplay between buchert averaging and the classical energy conditions .
we refer the reader to ref . for more details .
the null energy condition has sweeping consequences in gr .
however , it has been argued that it has been misunderstood as a property of matter,@xcite when in fact it is better viewed as a constraint on spacetime geometry .
in fact , the geometric formulation of the nec was derived from worldsheet string theory , where it is simply the virasoro condition for a closed string moving in a curved background .
furthermore , it was shown that there is an entirely different thermodynamic origin of the nec , if gravity emerges from some holographic theory .
thus , rather than being an incidental property of matter , it was argued in ref . that the validity of the nec appears to hint at the origins of gravity .
it is interesting that wormholes are not the only geometries that violate the energy conditions , and it was shown that superluminal spacetimes , such as the _ warp drive _ , also violate the weak energy condition ( wec).@xcite indeed , while gr ranks undoubtedly among the best theories of physics ever developed , it is also among those with the most striking implications . in particular , gr admits solutions which allow faster than light motion , and consequently effective time travel .
it was shown that a `` pre - emptive '' chronology protection mechanism destabilises superluminal warp drives via quantum matter back - reaction , and therefore forbids even the conceptual possibility to use these solutions for building a time machine.@xcite this result was considered both in standard quantum field theory in curved spacetime as well as in the case of a quantum field theory with lorentz invariance breakdown at high energies .
its interesting to note that evolving wormhole spacetimes may involve `` flashes '' of wec violation , where the matter threading the wormhole violates the energy conditions for small intervals of time .
one can consider specific cases , in which the intervals of wec violation can be chosen to be very small .
for instance , consider the following line element of a wormhole in a cosmological background given by @xmath48\,,\ ] ] where @xmath49 is the conformal factor , which is finite and positive definite throughout the domain of @xmath50 .
it is also possible to write the metric ( [ evolvingwhmetric ] ) using `` physical time '' instead of `` conformal time '' , by replacing @xmath50 by @xmath51 and therefore @xmath52 by @xmath53 , where the latter is the functional form of the metric in the @xmath54 coordinate.@xcite when the shape function and the redshift function vanish , @xmath55 and @xmath56 , respectively , the metric ( [ evolvingwhmetric ] ) becomes the frw metric .
as @xmath57 and @xmath58 , it approaches the static wormhole metric ( [ metricwormhole ] ) .
one can easily use the raychaudhuri equation to verify the non - violation of the wec for dynamic wormholes.@xcite recall that the raychaudhuri equation for a congruence of null rays is given by @xmath59 where @xmath60 is the expansion of the congruence of null rays ; @xmath61 and @xmath62 are the shear and vorticity of the geodesic bundle , respectively , which are zero for this case . from the einstein field equation we have @xmath63 for all null @xmath40 .
so if @xmath64 , we have @xmath65 by ( [ nullraychaud ] ) .
thus if @xmath60 is negative anywhere it tends to @xmath29 at a finite value of the affine parameter , @xmath66 , i.e. , the bundle must necessarily come to a focus . for the case of a static wormhole , the expansion @xmath60 is given by @xmath67 where @xmath68 is a positive quantity . for @xmath69 ,
@xmath70 is negative then so is @xmath60 negative .
@xmath71 only if @xmath72 , since @xmath70 is always finite .
therefore , either the wormhole has a vanishing throat radius , which renders it as non - traversable , or the wec is violated . for the evolving case ,
the expansion is given by@xcite @xmath73 real time @xmath54 has now been used , and as long as @xmath74 , i.e. , the wormhole is opening out fast enough , so that @xmath60 is never negative .
thus , the fact that the bundle does not focus no longer implies that the wec is satisfied .
we refer the reader to for specific solutions .
consider the stationary and axially symmetric @xmath75dimensional spacetime .
the latter possesses a time - like killing vector field , which generates invariant time translations , and a spacelike killing vector field , which generates invariant rotations with respect to the angular coordinate @xmath76 .
consider the following metric@xcite @xmath77\,,\ ] ] where @xmath78 , @xmath79 , @xmath80 and @xmath81 are functions of @xmath20 and @xmath82 .
@xmath83 may be interpreted as the angular velocity @xmath84 of a particle that falls freely from infinity to the point @xmath85 .
consider the definition @xmath86 , which can be used to describe a traversable wormhole .
assume that @xmath87 is a positive , nondecreasing function of @xmath20 that determines the proper radial distance @xmath88 , i.e. , @xmath89 and @xmath90
. we shall adopt the notation that the subscripts @xmath91 and @xmath92 denote the derivatives in order of @xmath20 and @xmath93 , respectively.@xcite note that an event horizon appears whenever @xmath94.@xcite the regularity of the functions @xmath78 , @xmath95 and @xmath79 are imposed , which implies that their @xmath82 derivatives vanish on the rotation axis , @xmath96 , to ensure a non - singular behavior of the metric on the rotation axis .
the metric ( [ 3rwh ] ) reduces to the morris - thorne spacetime metric ( [ metricwormhole ] ) in the limit of zero rotation and spherical symmetry , @xmath97 , @xmath98 , @xmath99 and @xmath100 . in analogy with the morris - thorne case ,
@xmath24 is identified as the wormhole throat , and the factors @xmath78 , @xmath79 and @xmath80 are assumed to be well - behaved at the throat .
thus , one may conclude that the metric ( [ 3rwh ] ) describes a rotating wormhole geometry , with an angular velocity @xmath80 .
the factor @xmath79 determines the proper radial distance .
@xmath78 is the analog of the redshift function in the morris - thorne wormhole and is finite and nonzero to ensure that there are no event horizons or curvature singularities .
@xmath95 is the shape function which satisfies @xmath101 ; it is independent of @xmath82 at the throat , i.e. , @xmath102 ; and obeys the flaring out condition @xmath103 .
the nec at the throat is given by @xmath104 rather than reproduce the analysis here , we refer the reader to ref . , where it was shown that the nec is violated in certain regions , and is satisfied in others .
thus , it is possible for an infalling observer to move around the throat , and avoid the exotic matter supporting the wormhole .
however , it is important to emphasize that one can not avoid the use of exotic matter altogether .
recently , rotating wormhole solutions in gr , supported by phantom scalar fields , were further presented.@xcite it was shown that in four space time dimensions this family of solutions evolves from the static ellis wormhole , when a finite angular velocity is imposed at the throat . at a maximal value of the rotational velocity , the family of wormhole solutions ends in an extremal kerr solution .
the properites of these rotating wormhole solutions including their mass , angular momentum , quadrupole moment , and ergosphere were extensively discussed . in five space
time dimensions rotating wormholes with equal magnitude angular momenta were also presented . applying stability analysis the altter
, it was shown that the unstable mode of the ellis solutions disappears , when the angular momentum of the wormhole is sufficiently high.@xcite it is also interesting to note that for the static and spherically symmetric case , wormholes were also considered by two scalar fields.@xcite these provide the possibility to obtain topologically trivial solutions in contrast with wormholes created by one scalar field .
wormholes with one scalar field are topologically non - trivial because the scalar field possesses a kink - like behaviour .
the solutions with different potentials were considered , and furthermore , the profile of mass vs parameters of scalar fields was obtained.@xcite the kerr - like wormhole with phantom matter as a source , was also analyzed.@xcite it was show to possess three parameters , namely , mass , angular momentum and the scalar field charge .
the wormhole presented was shown to have a naked ring singularity , otherwise it is regular everywhere .
the main feature of the solution is that the throat lies on a sphere of the same radius as the ring singularity , and impedes an observer to see or to reach the singularity , so that it essentially behaves as an anti - horizon .
the geodesics of the wormhole were also analyzed , and it was found that an observer can traverse without consequences , however , the equator presents an infinite potential barrier which prevents the observer of reaching the throat .
furthermore , this wormhole contains a ring singularity only on the south hemisphere , without an horizon , but as in the kerr solution , there is no manner to see the singularity .
thus , it was argued that this solution is a new kind of cosmic censorship.@xcite consider two distinct spacetime manifolds , @xmath105 and @xmath106 , with metrics given by @xmath107 and @xmath108 , in terms of independently defined coordinate systems @xmath109 and @xmath110 , respectively .
the manifolds are bounded by hypersurfaces @xmath111 and @xmath112 , respectively , with induced metrics @xmath113 and @xmath114 .
a single manifold @xmath115 is obtained by gluing together @xmath116 and @xmath117 at their boundaries , i.e. , @xmath118 , with the natural identification of the boundaries @xmath119 .
now , taking into account that the interior wormhole spacetime is given by metric ( [ metricwormhole ] ) and the exterior geometry is the schwarzschild solution , the surface stresses are given by the following quantities @xmath120 \ , , \label{surfpressure}\end{aligned}\ ] ] where @xmath121 and @xmath122 are the surface energy density and the tangential surface pressure , respectively .
the conservation equation provides us with @xmath123 where @xmath124 , defined for notational convenience , is given by @xmath125 \sqrt{1-\frac{b}{a}+\dot{a}^2 } \ , .
\label{h(a)}\end{aligned}\ ] ] the construction of dynamic shells in wormholes have been extensively analyzed in the literature @xcite , where the stability of generic spherically symmetric thin shells to linearized perturbations around static solutions were considered , and applying the analysis to traversable wormhole geometries , by considering specific choices for the shape function , the stability regions were deduced .
negative tension branes as stable thin - shell wormholes in reissner - nordstroem-(anti ) de sitter spacetimes in @xmath126 dimensional einstein gravity , were also investigated.@xcite imposing @xmath127 symmetry , traversable static thin - shell wormholes were constructed and classified in spherical , planar and hyperbolic symmetries . in spherical geometry , it was found that the higher - dimensional counterpart of barcelo and visser s wormholes@xcite are stable against spherically symmetric perturbations .
classes of thin - shell wormholes in planar and hyperbolic symmetries with a negative cosmological constant were also found , which are stable against perturbations . in most cases ,
stable wormholes were found with the combination of an electric charge and a negative cosmological constant .
however , as special cases , stable wormholes were found with a vanishing cosmological constant in spherical symmetry and with vanishing electric charge in hyperbolic symmetry .
an interesting application of thin - shell collapse was also investigated.@xcite more specifically , the gravitational collapse of ( massive or null ) thin shells of dust in the adm hamiltonian formalism was studied , in a particular foliation , namely , for a constant - mean - extrinsic - curvature ( cmc ) in the case of a spatially compact universe , and its analogue in the case of an asymptotically flat space .
exact solutions to einstein s equations at the nonlinear level were obtained , which take into account the backreaction of matter on geometry .
the result is interesting because , in addition to providing an exact solution of gr , it also represents a solution of the newly discovered theory of shape dynamics.@xcite this theory is classically equivalent to gr , but it highlights a different ( dual ) symmetry to refoliation invariance , namely , spatial weyl invariance .
for this reason shape dynamics is expected to differ at the quantum level from the standard covariant quantization schemes of gr , and suggests that the fundamental degrees of freedom of gr are spatial conformal and diffeorphism invariant .
wormholes have also been extensively studied in modified theories of gravity . without a significant loss of generality ,
consider the generalized gravitational field equations for a large class of modified theories of gravity , given by the following field equation@xcite @xmath128 where @xmath129 ( @xmath130 ) are multiplicative factors that modify the geometrical sector of the field equations , and @xmath131 denote generically curvature invariants or gravitational fields such as scalar fields ; the term @xmath132 covers the coupling of the curvature invariants or the scalar fields with the matter stress - energy tensor , @xmath133 .
the additional geometric term @xmath134 includes the geometrical modifications inherent in the modified gravitational theory under consideration . in order to analyse a generalized form of the energy condition ,
it is rather useful to rewrite this field equation as an effective einstein field equation , @xmath135 , where the effective stress - energy tensor is given by @xmath136 where @xmath137 and @xmath138 are defined for notational convenience . for this case
, the violation of the generalized nec , @xmath139 , implies the following restriction @xmath140 note that imposing @xmath141 , @xmath142 , and @xmath143 , we recover gr and the standard violation of the nec for the matter threading the wormhole , i.e. , @xmath144 .
if one imposes an additional condition given by @xmath145/g_1(\psi^i)>0 $ ] , one obtains a general bound for the normal matter threading the wormhole , given by @xmath146 one may demand that the latter condition is fulfilled even if the matter stress - energy tensor satisfies the usual nec.@xcite one may also impose the wec , i.e. , @xmath147 , where @xmath148 is the four - velocity of an observer . in order for normal matter to satisfy the wec , in order to have normal matter threading the wormhole , one also needs to impose the following condition @xmath149 thus , imposing @xmath147 entails a restriction on the geometry arising from the modified gravity under consideration . considering that normal matter
is given by a diagonal stress - energy tensor , one can physically interpret @xmath150 as the energy density measured by any timelike observer with four - velocity @xmath148 .
this definition is useful as using local lorentz transformations it is possible to show that @xmath151 implies that the energy density is positive in all local frames of reference . in a scalar - tensor theory of gravity with a scalar field possessing the nonminimal kinetic coupling to the curvature
, static and spherically symmetric solutions were considered.@xcite the lagrangian of the theory contains the term @xmath152 and represents a particular case of the general horndeski lagrangian , which leads to second - order equations of motion .
the rinaldi approach was used to construct analytical solutions describing wormholes with nonminimal kinetic coupling .
it was shown that wormholes exist only if @xmath153 ( phantom case ) and @xmath154 .
furthermore , the wormhole throat connects two anti - de sitter spacetimes , and the metric possesses a coordinate singularity at the throat .
however , since all curvature invariants are regular , there is no curvature singularity there.@xcite as shown above , it is possible that wormholes arise as solutions of extensions of gr without violations of the energy conditions .
working in a metric - affine framework , which is experimentally supported by the physics of crystalline structures with defects,@xcite explicit models supporting such solutions in four and higher dimensions were found.@xcite it is shown that they actually represent explicit realizations of the concept of the geon introduced by wheeler , whcih were interpreted as topologically non - trivial self - consistent bodies generated by an electromagnetic field without sources .
several of their properties were discussed , and we refer the reader to refs . for more details .
furthermore , using exactly solvable models , it was shown that black hole singularities in different electrically charged configurations can be cured.@xcite the solutions obtained describe black hole space - times with a wormhole giving structure to the otherwise point - like singularity .
furthermore , it was shown that geodesic completeness was satisfied despite the existence of curvature divergences at the wormhole throat . in some cases ,
physical observers can go through the wormhole and in other cases the throat lies at an infinite affine distance .
the removal of singularities occurs in a non - perturbative way.@xcite in the context of gravity s rainbow , the graviton one - loop contribution to a classical energy was computed in a traversable wormhole background.@xcite the form of the shape function considered was obtained by a linear equation of state @xmath155 .
the approach was evaluated by means of a variational approach with gaussian trial wave functionals . instead of using a regularization / renormalization process , and to to handle the divergences ,
the distortion induced by gravity s rainbow was used .
the energy density of the graviton one - loop contribution , or equivalently the background spacetime , was let to evolve , and consequently the classical energy was determined .
more specifically , the background metric was fixed to be minkowskian in the equation governing the quantum fluctuations , which behaves essentially as a backreaction equation , and the quantum fluctuations were let to evolve .
the classical energy , which depends on the evolved metric functions , was then evaluated .
analysing this procedure , a natural ultraviolet ( uv ) cutoff was obtained , which forbids the presence of an interior spacetime region , and it was argued that this may result in a multipy - connected spacetime .
thus , in the context of gravity s rainbow , this process may be interpreted as a change in topology , and in principle results in the presence of a planckian wormhole .
the process of quantum tunneling was studied in a self - interacting scalar field theory with non - minimal coupling to gravity.@xcite it was demonstrated that in these theories gravitational instantons can develop a neck , which is a feature prohibited in theories with minimal coupling .
furthermore , it was show that such instantons with necks lead to the materialization of bubble geometries containing a wormhole region .
the relationship of neck geometries to violations of the nec was also explored , and the bound on the size of the neck relative to that of the instanton was derived .
general parameterizations for generic off - diagonal spacetime metrics and matter sources in gr and modified gravity were found,@xcite when the field equations decouple with respect to nonholonomic frames of reference .
this allows one to construct various classes of exact solutions when the coefficients of the fundamental geometric / physical objects depend on all spacetime coordinates via corresponding classes of generating and integration functions and/or constants .
such ( modified ) spacetimes display killing and non - killing symmetries , describe nonlinear vacuum configurations and effective polarizations of cosmological and interaction constants .
certain examples of exact locally anisotropic wormholes and generic off - diagonal cosmological solutions were analysed in modified gravity , such as , in @xmath156 gravity.@xcite it was concluded that considering generic off - diagonal nonlinear parametric interactions in gr it is possible to mimic various effects in massive and/or modified gravity , or to distinguish certain classes of `` generic '' modified gravity solutions which can not be encoded in gr .
in this paper , we have considered a brief review of wormhole physics .
it is important to emphasize that these solutions are primarily useful as `` gedanken - experiments '' and as a theoretician s probe of the foundations of general relativity .
they have been extremely important in stimulating research in the issues of the energy condition violations , closed timelike curves and the associated causality violations and `` effective '' superluminal travel .
we have outlined a review dating from the `` ( flamm)-einstein - rosen '' bridge , the revival of the topic by wheeler with the introduction of the `` geon '' concept in the 1960s , the full renaissance of the subject by thorne and collaborators in the late 1980s , culminating in the monograph by visser , and detailed the issues that branched therefrom to the present date .
more specifically , we have presented a mathematical overview of the morris - thorne wormhole , paying close attention to the pointwise and averaged energy condition violations , the quantum inequality and modern generalizations of semi - classical nonlinear energy condition .
we then , treated rotating wormholes and evolving wormholes , focussing on the energy condition violations .
indeed , a fundamental ingredient in wormhole physics is the flaring - out condition at the throat which , in classical general relativity , entails the violation of the null energy condition .
we have also presented the most general conditions in the context of modified gravity , in which the matter threading the wormhole throat satisfies all of the energy conditions , and it is the higher order curvature terms , which may be interpreted as a gravitational fluid , that support these nonstandard wormhole geometries .
thus , we explicitly show that wormhole geometries can be theoretically constructed without the presence of exotic matter , but are sustained in the context of modified gravity .
specific models were also briefly outlined .
fsnl was supported by the fundao para a cincia e tecnologia ( fct ) through the grants expl / fis - ast/1608/2013 , uid / fis/04434/2013 and by a fct research contract , with reference if/00859/2012 .
f. s. n. lobo , `` exotic solutions in general relativity : traversable wormholes and ` warp drive ' spacetimes , '' _ classical and quantum gravity research _ , 1 - 78 , nova sci . pub .
isbn 978 - 1 - 60456 - 366 - 5 ( 2008 ) [ arxiv:0710.4474 [ gr - qc ] ] .
n. m. garcia , f. s. n. lobo and m. visser , `` generic spherically symmetric dynamic thin - shell traversable wormholes in standard general relativity , '' _ phys .
d _ * 86 * , 044026 ( 2012 ) [ arxiv:1112.2057 [ gr - qc ] ] .
n. montelongo garcia and f. s. n. lobo , `` nonminimal curvature - matter coupled wormholes with matter satisfying the null energy condition , '' _ class .
_ * 28 * , 085018 ( 2011 ) [ arxiv:1012.2443 [ gr - qc ] ] .
l. h. ford and t. a. roman , `` quantum field theory constrains traversable wormhole geometries , '' _ phys .
d _ * 53 * , 5496 ( 1996 ) [ arxiv : gr - qc/9510071 ] . c. barcelo and m. visser , `` scalar fields , energy conditions , and traversable wormholes , '' _ class .
* 17 * , 3843 ( 2000 ) [ arxiv : gr - qc/0003025 ] .
f. s. n. lobo , j. martinez - asencio , g. j. olmo and d. rubiera - garcia , `` dynamical generation of wormholes with charged fluids in quadratic palatini gravity , '' _ phys .
d _ * 90 * , 024033 ( 2014 ) [ arxiv:1403.0105 [ hep - th ] ] . | we consider the possibility of multiply - connected spacetimes , ranging from the flamm - einstein - rosen bridge , geons , and the modern renaissance of traversable wormholes .
a fundamental property in wormhole physics is the flaring - out condition of the throat , which through the einstein field equation entails the violation of the null energy condition . in the context of modified theories of gravity
, it has also been shown that the normal matter can be imposed to satisfy the energy conditions , and it is the higher order curvature terms , interpreted as a gravitational fluid , that sustain these non - standard wormhole geometries , fundamentally different from their counterparts in general relativity .
we explore interesting features of these geometries , in particular , the physical properties and characteristics of these ` exotic spacetimes ' . |
the pressure dependence of the curie temperature provides important information on a ferromagnetic system and is an object of intensive studies both experimental @xcite and theoretical .
@xcite the key question here is the character of the variation of various magnetic properties with decreasing distances between magnetic atoms . in an early work ,
castellitz @xcite proposed an empirical rule ( interaction curve ) that describes the dependence of the curie temperature of the mn - containing ferromagnetic alloys with 4 - 5 valence electrons per molecule on the ratio @xmath2 where @xmath3 is the nearest - neighbor mn - mn distance and @xmath4 is the radius of the atomic mn 3d shell .
the curve is supposed to represent the curie temperatures of various systems at ambient pressure as well as the pressure dependence of t@xmath1 of a given system .
the function is not monotonous and has a maximum at the @xmath2 value of about 3.6 ( see fig .
[ fig_interactioncurve ] ) . according to the interaction curve
, one can expect @xmath5 for alloys with @xmath6 ( e.g. , ni@xmath0mnsn and cu@xmath0mnin ) . on the other hand , the systems with @xmath7 ( e.g. ,
nias - type mnas , mnsb and mnbi ) are expected to have negative pressure dependence of the curie temperature .
these predictions are in agreement with experiment .
@xcite recently kanomata _ et al .
_ suggested a generalization of the interaction curve to the case of 6 - 7 valence electrons per chemical formula .
@xcite these systems form a new branch of the dependence of the curie temperature on the mn - mn distance ( fig .
[ fig_interactioncurve ] ) .
the available experimental values of the pressure derivative of the curie temperature , @xmath8 , for heusler alloys are consistent with those expected from the interaction curve.@xcite early experiments on the pressure dependence of the curie temperature of heusler alloys have been performed in a low pressure region ( less than 0.5 gpa ) .
recently gavriliuk _ _ et al.__@xcite have studied structural and magnetic properties of ni@xmath0mnsn in the pressure interval up to 10.8 gpa .
the authors have found an increasing linear dependence of the curie temperature on applied pressure .
the mssbaurer spectroscopy revealed partial interchange of the mn and ni atoms .
the purpose of the present work is a first - principles study of the electronic structure , exchange interactions and curie temperature in ni@xmath0mnsn as a function of pressure .
the main attention is devoted to the interval of the interatomic mn
mn distances from @xmath9 to @xmath10 that corresponds to the available experimental variation of this parameter .
these values of the mn mn distance are far above the value of @xmath11 that , according to interaction curve , separates the regions of positive and negative pressure gradients of the curie temperature for this group of systems . to verify the appearance of the non - monotonous behavior we extended the calculation to smaller values of the lattice constant corresponding to larger applied pressures .
we compare empirical and calculated dependencies .
the influence of the mn - ni atomic interchange on the magnetism of the system is also studied .
the paper is organized as follows . in sec .
ii we present the calculational approach .
section iii contains the results of the calculations and discussion .
section iv gives the conclusions .
the calculations are carried out with the augmented spherical waves method @xcite within the atomic
sphere approximation.@xcite the exchange correlation potential is chosen in the generalized gradient approximation .
@xcite a dense brillouin zone ( bz ) sampling @xmath12 is used . to establish the relation between the lattice parameters and applied pressure we use the following expression obtained experimentally in ref .
@xmath13 where @xmath14 , @xmath15 and @xmath16 is the volume of the unit cell at the ambient pressure .
the radii of all atomic spheres are chosen equal .
we describe the interatomic exchange interactions in terms of the classical heisenberg hamiltonian @xmath17 in eq.([eq : hamiltonian2 ] ) , the indices @xmath18 and @xmath19 number different sublattices and @xmath20 and @xmath21 are the lattice vectors specifying the atoms within sublattices , @xmath22 is the unit vector pointing in the direction of the magnetic moment at site @xmath23 .
the systems considered contain three 3d atoms in the unit cell with positions shown in fig.[fig_lattice ] .
we employ the frozen
magnon approach @xcite to calculate interatomic heisenberg exchange parameters .
the calculations involve few steps . in the first step ,
the exchange parameters between the atoms of a given sublattice @xmath18 are computed .
the calculation is based on the evaluation of the energy of the frozen magnon configurations defined by the following atomic polar and azimuthal angles @xmath24 the constant phase @xmath25 is always chosen equal to zero .
the magnetic moments of all other sublattices are kept parallel to the z axis . within the heisenberg model ( [ eq : hamiltonian2 ] )
the energy of such configuration takes the form @xcite @xmath26 where @xmath27 does not depend on * q * and the fourier transform @xmath28 is defined by @xmath29 in the case of @xmath30 the sum in eq .
( [ eq : j_q ] ) does not include @xmath31 .
calculating @xmath32 for a regular @xmath33mesh in the brillouin zone of the crystal and performing back fourier transformation one gets exchange parameters @xmath34 for sublattice @xmath18 .
structure adapted by the full heusler alloys .
the lattice consists of four interpenetrating fcc sublattices with the positions @xmath35 and @xmath36 for the ni and @xmath37 and @xmath38 for the mn and sn , respectively . ]
the determination of the exchange interactions between the atoms of two different sublattices @xmath18 and @xmath19 is discussed in ref . .
the curie temperature is estimated within the mean field approximation for a multi sublattice material by solving the system of coupled equations @xcite @xmath39 where @xmath40 is the average @xmath41 component of @xmath42 and @xmath43 .
eq.([eq_system ] ) can be represented in the form of eigenvalue matrix problem @xmath44 where @xmath45 , @xmath46 is a unit matrix and @xmath47 is the vector of @xmath48 .
the largest eigenvalue of matrix @xmath49 gives the value of curie temperature .
.lattice parameters , magnetic moments and curie temperatures in ni@xmath0mnsn at ambient pressure and applied pressure of 16 gpa . for comparison ,
the magnetic moment obtained with full potential flapw - gga method is presented . [ lattice ] [ cols="<,^,^",options="header " , ] at zero pressure the pattern of exchange parameters ( fig .
[ fig_exchange ] ) and resulting curie temperature ( table [ lattice - swap ] ) are very similar to the case without mn ni interchange .
however , the situation is different at applied pressure of 16 gpa .
both mn - mn and mn - ni@xmath50 nearest - neighbor exchange parameters increase substantially .
the remaining exchange parameters show small pressure dependencies .
the interaction of the mn moment with the moment of the second ni atom is slightly reduced .
the substantial increase of the leading exchange parameters with pressure results in considerable change of the curie temperature from 400 k at ambient pressure to 562 k at 16 gpa . assuming a linear variation of t@xmath1 with pressure we estimate the pressure derivative , @xmath51 , as 12.5 k / gpa .
this value of @xmath51 exceeds strongly the corresponding value for the system without mn ni atomic interchange .
since the number of the swaped mn and ni atoms in our model is much larger than in the samples measured the calculated @xmath51 can not be directly compared with the experimental pressure derivative .
important , however , that the mn ni atomic interchange increases the pressure derivative of t@xmath1 that gives an explanation for the low value of the theoretical pressure derivative in the case of the system without swapping . a detailed study of the influence of the inter
sublattice atomic interchange on the electron properties of the heusler systems is an interesting extension of the present study .
in conclusion , we have systematically studied the pressure dependence of exchange interactions and curie temperature in full heusler alloy ni@xmath0mnsn within the parameter
free density functional theory .
we show that the character of the pressure dependence of the exchange interactions is a consequence of the complex interplay of competing trends in the electronic properties of the system . in agreement with experiment ,
the curie temperature increases with increasing pressure in the pressure region studied . extending our theoretical study onto a larger pressure interval we obtained non
monotonous t@xmath1 dependence and the presence of the metamagnetic transition .
the t@xmath1 behavior in the whole pressure interval is in qualitative correlation with the kanomata s empirical interaction curve . in the low pressure region
there is good quantitative agreement between calculated values and the prediction of the empirical rule . the mn
ni atomic interchange is shown to increase the pressure derivative of the curie temperature that suggests the physical mechanism for the improved agreement between the experimental and theoretical estimations of this parameter .
p. w. anderson , _ theory of magnetic exchange interactions : exchange in insulators and semiconductors _ , in _ solid state physics _ , edited by f. seitz and turnbull ( academic press , new york ) , vol .
99 - 214 .
p. j. webster and k. r. a. ziebeck , in _ alloys and compounds of d - elements with main group elements _ , part 2 , edited by h. r. j. wijn , landolt - brnstein , new series , group iii , vol .
19/c ( springer , berlin , 1988 ) , pp . | the pressure dependence of electronic structure , exchange interactions and curie temperature in ferromagnetic heusler alloy ni@xmath0mnsn has been studied theoretically within the framework of the density - functional theory .
the calculation of the exchange parameters is based on the frozen
magnon approach .
the curie temperature , t@xmath1 , is calculated within the mean field approximation by solving the matrix equation for a multi sublattice system . in agrement with experiment
the curie temperature increased from 362k at ambient pressure to 396 at 12 gpa . extending the variation of the lattice parameter beyond the range studied experimentally we obtained non
monotonous pressure dependence of the curie temperature and metamagnetic transition .
we relate the theoretical dependence of t@xmath1 on the lattice constant to the corresponding dependence predicted by the empirical interaction curve . the mn - ni atomic interchange observed experimentally
is simulated to study its influence on the curie temperature . |
galaxy clusters formed by the collapse of the highest density peaks in primordial density fluctuations ( e.g. , * ? ? ?
* ; * ? ? ?
the cluster environment strongly affects the star - formation history of cluster members .
galaxies in the centers of clusters are observed to be redder and more evolved than galaxies on the outskirts of the cluster , as well as field galaxies at the same redshift @xcite .
thus , the most massive clusters today likely formed in short episodes of vigorous star formation that effectively ceased by @xmath4 @xcite . theoretical simulations of cluster formation and evolution support this scenario @xcite . however , using old stellar populations to reconstruct the star formation conditions and to understand why star formation ceased in clusters is extremely difficult .
recently , several galaxy clusters have been discovered in the `` redshift desert '' ( @xmath5 ) @xcite . with near - infrared multi - object spectroscopy on 8 - 10 m telescopes ,
it is now possible to study the star formation conditions in these clusters , prior to ( or during ) the end of their major star - forming episodes . near infrared spectroscopy probes the optical rest - frame emission - lines at @xmath0 , providing a wealth of information about the ionized gas surrounding the young , hot stars .
the strong lines [ ] , h@xmath1 , [ ] , h@xmath2 , and the [ ] and [ ] doublets diagnose the power source of galaxies , the star formation rate , the gas - phase chemical abundance , the ionization state of the gas , the electron density of the gas , and the amount of dust extinction , among other physical properties ( see * ? ? ? * for an review ) .
knowledge of these properties in clusters across cosmic time helps to build a more comprehensive picture of how star formation progressed in clusters , and how the cluster environment impacted the star formation process .
so far , large near infrared spectroscopic studies have predominantly focused on field galaxies .
high redshift field galaxies appear to have larger [ ] /h@xmath1 and/or [ ] /h@xmath2 optical line ratios ( by @xmath6 dex ) compared with local field galaxies @xcite .
these elevated line ratios may indicate a major change in the ionized gas or star formation properties as a function of time .
many possible causes for this offset have been proposed , including a larger ionization parameter ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , a harder ionizing radiation field @xcite , a higher ism pressure ( density ) @xcite , a contribution from an agn @xcite , and contamination by shocks from galactic outflows @xcite .
sample selection is likely to play a large role in the relative importance of each of these factors @xcite .
if the cluster environment causes more rapid evolution during the star - formation stage than in the field , then we might expect a difference in the ism conditions in clusters at high redshift compared with field galaxies .
the star - formation properties of high redshift cluster galaxies relative to the field are currently under debate . using stacked near - infrared spectroscopy
, @xcite suggested that the relationship between star formation and stellar mass is similar among cluster and field galaxies at @xmath0 .
however , @xcite and @xcite argue that the star formation rate in cluster galaxies is elevated relative to field galaxies at the same stellar mass . in this paper , we test the hypothesis that star - forming conditions in clusters are more evolved than the field using the @xmath0 cosmos cluster @xcite .
the cosmos cluster is an ideal laboratory for studying the formation of massive nearby galaxy clusters .
the cosmos cluster contains four distinct overdensities @xcite .
cosmological simulations suggest that at least two of the overdensities will merge across cosmic time , creating a virgo - like cluster by @xmath7 @xcite .
the cluster has 57 spectroscopically confirmed members , and follows the same mass - metallicity relation as field galaxies at @xmath8 ( kacprzak et al .
2014 , in prep ) .
the cosmos cluster and comparison samples are described in section [ sample ] . in section [ mosfire ] , we describe our near - infrared spectroscopy . in section [ bpt_section ] , we compare the optical line ratios for 10 galaxies in the cosmos cluster with field galaxies at the same redshift , and with local star - forming galaxies .
we investigate the electron density ( or ism pressure ) in the cosmos cluster in section [ pressure ] , and the ionization state of the gas in section [ ionization ] .
we discuss the implications of this work and our conclusions in section [ conclusions ] . throughout this paper , we adopt the flat @xmath9-dominated cosmology as measured by the 7 year wmap experiment ( @xmath10 , @xmath11 ; * ? ? ?
the cosmos cluster was discovered as part of the magellan / fourstar galaxy evolution survey ( zfourge ) photometric redshift survey of the cosmos field @xcite .
the zfourge survey uses five medium bandwidth filters ( @xmath12,@xmath13,@xmath14,@xmath15,@xmath16 ) and the broadband @xmath17 filter on the magellan fourstar instrument ( persson13 , straatman et al . in prep ) .
high - redshift galaxy overdensities were identified using surface density maps within narrow @xmath18 redshift ranges between @xmath19 . with this technique , the cosmos cluster was revealed with photometric redshifts @xmath20 @xcite .
subsequent keck observations revealed 57 spectroscopically confirmed cluster members within a projected radius of @xmath21mpc at @xmath22 @xcite .
we compare the ism properties of the cosmos cluster with two comparison field samples .
we use the sloan digital sky survey ( sdss ) as our local comparison sample @xcite .
although the cosmos cluster galaxies will have evolved into red elliptical galaxies by @xmath23 , the sdss represents the range of properties seen in massive ( @xmath24@xmath25 ) nearby star - forming galaxies .
we select galaxies from the sdss dr4 , as described in @xcite . to avoid aperture effects and incompleteness , we restrict the sdss redshift range to @xmath26 .
the lower redshift limit corresponds to a minimum aperture covering fraction of @xmath27% to minimize aperture effects @xcite .
we use the emission - line fluxes from the jhu / mpa catalog with h@xmath2 , [ ] , h@xmath1 , [ ] , and [ ] emission - lines with signal - to - noise ratios ( s / n ) @xmath28 . these spectra have been corrected by the jhu / mpa team for underlying stellar absorption using stellar population synthesis models .
our sdss sample contains 85,224 galaxies .
we remove agn using the @xcite optical classification scheme , leaving 66,590 star - forming galaxies .
we compare the ism properties of the cosmos cluster with the properties of 10 serendipitous field galaxies between @xmath29 , observed in the same slitmasks as the cosmos cluster candidates .
these galaxies were identified in the zfourge five medium bandwidth filters .
we supplement our small field sample with 39 star - forming galaxies between @xmath30 from the keck baryonic structure survey ( kbss - mosfire ) survey by @xcite .
the kbss - mosfire star - forming galaxy sample contains @xmath31% of galaxies previously selected using rest - frame far - uv spectra , with the remaining galaxies covering a broader range of rest - frame uv and optical colors . the final kbss sample with detected [ ] /h@xmath1 and [ ] /h@xmath2 emission - lines is not biased in star formation rate and stellar mass , relative to the parent ( photometric ) sample
( see figure 4 in * ? ? ?
* ) . to avoid selection effects based on galaxy mass , while ensuring statistically significant samples for comparison , we create two comparison field samples , depending on the suite of emission - lines available
we refer to these two samples as our bpt and [ ] samples , respectively . _ bpt field comparison sample : _ to compare the position of galaxies on the standard optical diagnostic diagrams @xcite , we select galaxies with the full suite of h@xmath2 @xmath32 , [ ] @xmath33 , h@xmath1 @xmath34 , and [ ] @xmath35 lines .
steidel et al .
apply a lower s / n limit of @xmath36 to the [ ] and h@xmath1 emission - lines , and a @xmath37 limit to the [ ] and h@xmath1 emission - lines . for consistency
, we apply the same s / n limits to our field galaxies , but we highlight galaxies with [ ] or h@xmath1 emission - lines with s / n@xmath38 in our figures to indicate that these detections are marginal . the position of a galaxy on the optical diagnostic diagrams is sensitive to its gas - phase metallicity , and hence to its stellar mass through the mass - metallicity relation @xcite . selecting a high redshift sample for the full suite of h@xmath2 , [ ] , h@xmath1 , [ ] emission - lines biases a sample towards galaxies with detectable [ ] and [ ] lines .
this selection can introduce a bias against high mass and low mass galaxies .
metal - rich nebulae produce low [ ] emission - line fluxes and the oxygen lines are sensitive to the electron temperature of the gas , which is cool in metal - rich nebulae . on the other hand , selecting for [ ] detections biases against low metallicity galaxies which have weak [ ] .
the [ ] line is sensitive to metallicity only in the metal - rich regime ( @xmath39@xmath40 ; * ? ? ?
* ) because nitrogen changes from a primary to a secondary nucleosynthetic element at intermediate metallicities
. therefore , we match our field and comparison samples in stellar mass . here
, we use a subsample of the kbss - mosfire galaxies matched in stellar mass to our zfire cluster sample with @xmath41@xmath25@xmath42 .
this stellar mass selection results in 17 galaxies .
our zfire serendipitious field sample contains 4 galaxies with the full suite of emission - lines h@xmath2
@xmath32 , [ ] @xmath33 , h@xmath1 @xmath34 , and [ ] @xmath35 detected to signal - to - noise ratios @xmath28 within the same stellar mass range of @xmath41@xmath25@xmath42 . applying the kbss - mosfire s / n limits yields one additional galaxy . _
[ ] field comparison sample : _ we use the [ ] doublet for calculating the electron density of the ionized gas in the [ ] zone . to compare the electron density of the cluster and field galaxies , we select galaxies with rest - frame red emission - lines h@xmath1 @xmath34 , [ ] @xmath35 , [ ] @xmath43 , and [ ] @xmath44 detected to the 3@xmath45 level .
this selection results in high luminosity , high stellar mass samples because the [ ] lines are only weakly dependent on the metallicity ( see e.g. , * ? ? ?
we therefore match our cluster and field galaxies in stellar mass ( @xmath46@xmath25@xmath47 ) , resulting in 10 field comparison galaxies from our mosfire dataset .
our bpt and [ ] field samples are not mutually exclusive , and they span the same range of sfrs ( @xmath48 )
. for our electron density analysis , we are unable to compare with the kbss - mosfire sample because the kbss - mosfire [ ] line fluxes are not publicly available .
we obtained near - infrared spectroscopy for the cosmos cluster as part of our zfire spectroscopic survey of high - z cluster candidates .
the candidates were selected based on the zfourge photometric redshifts @xcite derived from deep images obtained through the five zfourge near - infrared medium - band filters .
the median uncertainty for the zfourge photometry is @xmath49 dex @xcite .
observations of cluster candidates were made on keck mosfire @xcite on december 24 - 25 , 2013 and february 10 - 13 , 2014 .
eight masks were used in the @xmath50-band ( 1.93 - 2.45 @xmath51 m ) to obtain the h@xmath1 and [ ] emission - lines , as well as two masks in the @xmath52-band ( 1.46 - 1.81@xmath51 m ) , to detect the h@xmath2 and [ ] emission - lines .
total on - source exposure times were 2 hours for the k - band masks , and 5.3 and 3.2 hours for the two h - band masks . seeing varied between 0.4 - 0.7 arcseconds throughout the exposures .
we used a slit width of 0.7 `` , which gives a spectral resolution of @xmath53 in the k - band and @xmath54 in the h - band .
the mosfire field of view of [email protected] arcminutes enabled 224 objects to be targeted in 6 pointings .
we observed an a0v type standard star in both the wide - slit mode and the narrow - science - slit ( @xmath56 '' slit width ) mode before and after our science target exposures for telluric and flux calibration .
the raw mosfire data were reduced using the publicly - available data reduction pipeline ( drp ) developed by the instrument team .
the drp provides background - subtracted , rectified and wavelength calibrated 2d spectra .
all spectra were calibrated to wavelength in a vacuum .
the h - band data were calibrated using night sky lines .
the k - band data were calibrated using both night sky lines and a neon arc lamp to ensure sufficient red emission - lines for wavelength calibration .
the typical residual for the wavelength solution is @xmath57 .
we used our own custom idl routines to correct the 2d spectra for telluric absorption , and for flux calibration using a0v standard stars .
we have visually inspected the 2-dimensional spectra to check for contamination by sky lines at the positions of the emission - lines , including the weak [ ] @xmath586584 line and the [ ] @xmath58@xmath586716,6731 doublet . we have also checked that our error spectra fully account for skyline residuals .
the uncertainty for our flux calibration is @xmath59 8% ( nanayakkara et al . in prep ) .
the 1-d spectra and associated 1@xmath45 error spectra were extracted using an aperture that corresponds to the fwhm of the spatial profile . for objects that are too faint to fit a gaussian to the spatial profile
, we used the fwhm of the stellar profile on the same mask as the extraction aperture .
gaussian profiles were fit to the emission lines in h and k bands separately . for widely separated lines such as [ ] @xmath583727 , h@xmath2@xmath584861 ,
single gaussian functions are fit with 4 free parameters : the centroid ( or the redshift ) , the line width , the line flux , and the continuum .
the doublet [ ] @xmath58@xmath584959,5007 are initially fit as a double gaussian function with 6 free parameters : the centroids 1 and 2 , line widths 1 and 2 , fluxes 1 and 2 , and the continuum .
a triple - gaussian function is simultaneously fit to the three adjacent emission lines : [ ] @xmath58@xmath586548 , 6583 and h@xmath1 .
the centroid and velocity width of the [ ] @xmath58@xmath586548 , 6583 lines are constrained by the velocity width of h@xmath1@xmath586563 , and the ratio of [ ] @xmath586548 and [ ] @xmath586583 is constrained to the theoretical value of 1/3 given in @xcite .
the line profile fitting is conducted using a @xmath60 minimization procedure which uses the inverse of the 1@xmath45 error spectrum of the drp .
this fitting procedure yields the redshift , line flux , line width , continuum and the associated statistical errors .
our typical 3@xmath45 flux limit is @xmath61 ergs / s/@xmath62 .
example spectra of our cosmos cluster members are given in figure 1 of @xcite . from our spectra
, we obtained spectroscopic redshifts for 180 galaxies .
of these , 57 galaxies are confirmed cluster members @xcite .
a total of 8/57 zfire cluster galaxies have the full suite of h@xmath2 @xmath32 , [ ] @xmath33 , h@xmath1 @xmath34 , and [ ] @xmath35 lines detected to signal - to - noise ratios @xmath28 , which we refer to as our bpt cluster sample .
for consistency with the kbss - mosfire field sample , we also show the positions of galaxies with h@xmath2 @xmath32 and [ ] @xmath35 lines detected to signal - to - noise ratios of @xmath37 , and [ ] @xmath33 , h@xmath1 @xmath34 lines detected to @xmath36 .
the kbss - mosfire s / n selection criteria would add 5 additional galaxies to the cluster sample .
we have verified that our results are unchanged regardless of whether we apply our s / n @xmath63 criterion , or whether we apply the kbss - mosfire s / n criteria .
our zfire bpt cluster sample was selected to satisfy @xmath64@xmath25)@xmath42 to match the mass range spanned by the bpt field control sample . of the galaxies with h@xmath1 @xmath34 , and [ ] @xmath35 lines detected to signal - to - noise ratios @xmath28
, 4 galaxies also have both [ ] @xmath43 and [ ] @xmath44 detected to the 3@xmath45 level , and span the stellar mass range @xmath65@xmath25)@xmath47 , where the stellar mass range is matched to the [ ] comparison sample .
we refer to these four [ ] detected cluster galaxies as our [ ] cluster sample . of these
, only 3 galaxies have all of the h@xmath2 , [ ] , h@xmath1 , [ ] , and [ ] , emission - lines detected to the @xmath66 level .
we compute stellar masses using the zfourge @xmath12,@xmath13,@xmath14,@xmath15,@xmath16 ) and @xmath17 photometry .
we fit the @xcite stellar population synthesis models with fast @xcite , assuming a @xcite initial mass function , a @xcite attenuation law , and solar metallicity .
the same models and parameters were used by @xcite for the kbss - mosfire sample .
our bpt cluster sample , the steidel bpt field sample , and our zfire bpt field sample have the same mean stellar masses , within the errors ( @xmath67@xmath25@xmath68 , and @xmath69 , respectively ) .
although limited by small numbers , our [ ] cluster and field comparison samples also have consistent stellar masses , within the errors ( @xmath70 c.f .
@xmath71 , respectively ) .
star formation rates ( sfrs ) are calculated from the extinction - corrected h@xmath1 emission - line and the @xcite calibration , as described in detail in tran et al .
( 2015 , apj , submitted ) .
we correct the h@xmath1 emission - line for extinction using the @xcite extinction curve . the resulting sfrs range between @xmath72 for our bpt cluster sample and @xmath73 for our bpt field sample .
the 2d kolmogorov - smirnov test indicates that our cluster and field samples have statistically consistent ( p(ks)@xmath74 ) extinction and star - formation rates , given the sample sizes and errors .
we do not match the star formation rate range for the cluster and field galaxies because the star formation rate is correlated with the ionization parameter , and the ionization parameter is one of the quantities that we aim to test for differences among field and cluster galaxies .
the [ ] /h@xmath1 versus [ ] /h@xmath2 diagram was originally proposed by @xcite to distinguish between galaxies powered by regions , planetary nebulae , and objects powered by a harder ionizing radiation field , such as galaxies containing active galactic nuclei ( agn ) .
this diagram ( known as the bpt diagram ) is useful at intermediate and high redshift because the [ ] h@xmath1 [ ] and h@xmath2 lines are often the only emission - lines observable in near - infrared spectra of high redshift galaxies , and the wavelength of the emission - lines are sufficiently close that the [ ] /h@xmath1 and [ ] /h@xmath2 ratios do not need to be flux calibrated or corrected for extinction . both regions and star - forming galaxies form a curved locus on the [ ] /h@xmath1 versus [ ] /h@xmath2 optical diagnostic diagram @xcite .
in figure [ bpt](a ) , we show the locus of the star - forming galaxies in the sdss sample .
@xcite showed that theoretical photoionization models can reproduce this curved locus .
this locus is now known as the star - forming abundance sequence because the shape of the locus is dictated primarily by the chemical abundance ( metallicity ) of the star - forming regions within galaxies .
metal - poor galaxies have low [ ] /h@xmath1 and high [ ] /h@xmath2 ratios due to the combination of low chemical abundance and the electron temperature sensitivity of the oxygen collisionally excited lines .
the ism and star formation properties of a galaxy affects its position relative to the star - forming abundance sequence on the bpt diagram ( see figure 1 in * ? ? ?
a rise in the ism pressure causes both the [ ] /h@xmath1 and the [ ] /h@xmath2 line ratios to rise . a hard ionizing radiation field ( such as from a wolf - rayet dominated stellar population ) causes a similar rise in the [ ] /h@xmath1 and [ ] /h@xmath2 line ratios .
however , a rise in the ionization parameter causes the [ ] /h@xmath1 ( and [ ] /h@xmath1 ) to fall , while [ ] /h@xmath2 rises .
the ionization parameter is defined as the number of hydrogen ionizing photons passing through a unit area per second divided by the number density of hydrogen atoms @xmath75 , and is a measure of the amount of ionization that a radiation field is able to produce in an region . for a spherical geometry
, the ionization parameter @xmath76 can be defined using the stromgren radius @xmath77 @xcite : @xmath78 where @xmath79 is the flux of ionizing photons above the lyman limit . in an ionized nebula ,
the number density of hydrogen is approximately the electron density @xmath80 .
the ionization parameter is affected by both the hardness of the ionizing radiation field , as well as the bolometric luminosity of the ionizing source . the combination of theoretical photoionization models , and observations of the [ ] , h@xmath2 , [ ] , h@xmath1 lines , and the [ ] doublet allows one to discriminate between the effect of a differing ionization parameter , the ism pressure , and the radiation hardness . the [ ] @xmath81 [ ] @xmath82 ratio constrains the ism pressure , thanks to the sensitivity of this doublet to the hydrogen density of the gas .
the [ ] /h@xmath1 ratio is particularly sensitive to the hardness of the ionizing radiation field because the [ ] lines are produced in a partially ionized zone that is large when the radiation field contains a significant fraction of high energy ( x - ray and euv ) photons . finally , the combination of the [ ] / [ ] and [ ] /h@xmath2 ratios separate the ionization parameter and the metallicity , as shown in @xcite .
figure [ bpt](a ) gives the [ ] /h@xmath1 versus [ ] /h@xmath2 diagram for the sdss ( grey contours ) and the cosmos cluster ( red symbols ) .
the sdss star - forming sequence is parameterized by @xcite ( their equation 5 ) . using this parameterization
, we calculate the median [ ] /h@xmath2 offset between the cosmos cluster and the sdss star - forming sequence to be @xmath3 dex .
we investigate possible causes for this offset in the following sections .
we compare the position of our cosmos cluster with our field comparison sample in figure [ bpt](b ) . we have matched the stellar mass range of both samples ( i.e. @xmath83 ) to avoid mass biases .
our cosmos cluster has line ratios that are well matched to the field galaxies at the same redshift .
the two - sided kolmogorov - smirnov statistic for the [ ] /h@xmath1 and [ ] /h@xmath2 ratios yields d - values of 0.34 and 0.35 , respectively , with significance levels of 0.36 and 0.32 , indicating that our cluster and the field sample are not significantly different .
we have also verified that the relationships between [ ] /h@xmath1 and stellar mass , and [ ] /h@xmath2 and stellar mass are the same for the cluster and field samples , within the errors .
we conclude that the cluster environment has not produced more rapid evolution in the ism conditions in our star - forming galaxies at @xmath0 .
agn strongly affect the [ ] /h@xmath1 and [ ] /h@xmath2 optical line ratios .
galaxies containing agn form a separate sequence on the bpt diagram , beginning at the high metallicity , high mass end of the star - forming abundance sequence ( figure [ bpt]a ) .
this agn sequence occurs because the hard radiation field from an agn produces more ionizations into the @xmath84 ion , and more collisional excitations of the [ ] line .
the position of an agn host galaxy along the agn sequence depends on the contribution from star - formation to the line ratios , the slope of the euv radiation field , and the ionization parameter of the radiation field @xcite .
we use the redshift - dependent theoretical optical classification scheme of @xcite to identify possible agn in our sample ( figure [ bpt]c ) .
this scheme was based on the predictions of cosmological hydrodynamic simulations , as well as stellar evolutionary synthesis and photoionization models .
we find that three galaxies in our cluster with @xmath85s / n([])@xmath86 may have a small ( @xmath87% ) contribution from an agn based on their optical line ratios ( i.e. [ ] /h@xmath2@xmath88 ) , but these galaxies are within the @xmath89 dex error range of the classification line ( dotted lines in figure [ bpt_niiha]c ) , and may be low metallicity star - forming galaxies with a hard ionizing radiation field , such as from wolf - rayet stars @xcite , or their [ ] line fluxes may simply have too low s / n to reliably classify .
we discuss these galaxies further in section [ ionization ] .
the remaining galaxies in our cluster are likely to be purely star - forming .
shocks can also produce strong [ ] /h@xmath2 , and can mimic agn on the bpt diagram locally ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . at the low metallicities of high redshift galaxies , shocks can also mimic the emission - line ratios of star - forming galaxies @xcite . in figure
[ shocks_bpt ] , we show the position of our cosmos cluster galaxies relative to the slow shock models of @xcite ( cyan ) and the fast shock models of @xcite ( green ) . these models are described in detail in @xcite , and examples of typical mixing sequences are given in rich et al .
( 2010 , 2011 ) .
the shock models occupy different regions of the diagnostic diagram depending on the metallicity of the sample . only the lowest metallicity ( @xmath39@xmath90 ) shock models can account for the line ratios of our cluster galaxies .
however , our cluster galaxies lie along the standard galaxy mass - metallicity relation at @xmath0 @xcite , with metallicities between @xmath91@xmath39@xmath92 in the @xcite metallicity scale . here
, we have converted the metallicities from the @xcite calibration into the @xcite scale using the calibrations provided in @xcite to yield consistent metallicity estimates with our photoionization and shock models .
because the shock models for @xmath91@xmath39@xmath92 produce optical line ratios that are significantly larger than observed ( by 0.1 - 0.4 dex ) , we rule out a dominant contribution from shocks to the optical emission - line ratios of our cluster galaxies .
galaxies containing a mixture of ionizing sources , such as shocks and star formation will lie along mixing sequences between the star - forming sequence and the 100% shock models .
the position and shape of the mixing sequences depend on the metallicity of the galaxy and the shock velocity .
we can not rule out a small ( @xmath5910% ) contribution from shocks to the emission - lines ratios .
high angular resolution and high spectral resolution integral field spectroscopy with adaptive optics can distinguish shocked regions from star - forming regions at high redshift @xcite .
this integral field spectroscopy is one of our future research directions for this sample .
we compare the location of our zfire cluster on the bpt diagram with our predictions from theoretical stellar population synthesis and photoionization models .
our models are described in detail in @xcite and @xcite .
we use the chemical evolution predictions from @xcite to constrain the gas - phase metallicity history for star - forming galaxies with @xmath93 @xmath25 .
the metallicity at a given redshift determines the shape of the euv spectrum produced in the starburst99 or pegase2 evolutionary synthesis models @xcite , which in turn , affects the line intensities predicted by the mappings iv photoionization code @xcite .
we use the instantaneous and continuous burst models from starburst99 and pegase2 with a salpeter initial mass function , and the pauldrach / hillier stellar atmosphere models .
we note that the choice of imf makes negligible difference in the line ratios that we are using in this analysis @xcite .
the resulting stellar population spectra are embedded within a spherical nebula .
our mappings iv photoionization code calculates radiative transfer self - consistently throughout the nebula
. mappings iv uses a kappa temperature distribution which is more suitable for a turbulent ism than a stefan - boltzmann distribution @xcite . for agn
, we use the dusty radiation - pressure dominated models of @xcite . with these models ,
we produce theoretical predictions for how the optical emission - line ratios will appear for galaxy samples at different redshifts , based on the following limiting assumptions for star - forming galaxies and agn : * star forming galaxies at high redshift may have ism conditions and/or an ionizing radiation field that are either ( a ) the same as local galaxies ( _ normal ism conditions _ ) or ( b ) more extreme than local galaxies ( _ extreme ism conditions _ ) .
extreme conditions in star - forming galaxies can be produced by a larger ionization parameter and a denser interstellar medium , and/or an ionizing radiation field that contains a larger fraction of photons able to ionize o@xmath94 into o@xmath95 ( i.e. energy @xmath96 ev ) relative to the number of hydrogen ionizing photons ( i.e. a harder radiation field ) . in @xcite
, we use an ionizing radiation field from the pegase 2 stellar population synthesis models @xcite to mimic the hard ionizing radiation field from the stellar atmospheres from massive stars with the effects of stellar rotation .
the difference between our hard ionizing radiation field and our normal ionizing radiation field varies with wavelength ( see figure 2 in * ? ? ?
* ) , between @xmath97@xmath25 in @xmath98 between @xmath99 .
similar effects on the optical line ratios can be achieved by increasing the ionization parameter by up to an order of magnitude ( i.e. @xmath100 cm / s to @xmath101 cm / s , and/or by raising the electron density by up to two orders of magnitude ( i.e. @xmath102 @xmath103 to @xmath104 @xmath103 ) ( kewley et al . in prep ) . *
the agn narrow line region at high redshift may either ( c ) have already reached the level of enrichment observed in local galaxies ( _ metal - rich _ ) , or ( d ) have the same metallicity as the surrounding star - forming gas . in the latter case ,
the agn narrow - line region at high redshift would be more _ metal - poor _ than local agn narrow - line regions .
our agn narrow - line region models cover a large range of power - law indices ( @xmath105 ) and ionization parameters ( @xmath106 ) .
therefore , the theoretical agn region may be larger than the observed location of agn on the bpt diagram at high redshift .
these two sets of limiting assumptions yield four limiting scenarios for where galaxies might be located on the bpt diagram at different redshifts : 1 .
normal ism conditions , and metal - rich agn narrow - line regions at high - z .
2 . normal ism conditions , and metal - poor agn narrow - line regions at high - z .
3 . extreme ism conditions , and metal - rich agn narrow - line regions at high - z .
4 . extreme ism conditions , and metal - poor agn narrow - line regions at high - z .
our theoretical model predictions for each of these four limiting scenarios are given by the solid lines in figure [ bpt_niiha ] for @xmath23 ( upper panel ) and @xmath0 ( lower panel ) .
the cosmos cluster appears to span a range of ism and star - forming conditions , from nearly local - star - forming conditions ( panels 1 , 2 ) to more extreme star - forming conditions ( panels 3 , 4 ) .
our data do not contain sufficient agn to distinguish between the two limiting agn scenarios .
we conclude that at least 4 of our cluster galaxies have more extreme star forming conditions than local star - forming galaxies , on average , and that their star - forming conditions are similar to field galaxies at the same redshift , as seen in figure 1 .
we investigate whether the [ ] /h@xmath1 ratio correlates with the electron density of the ionized gas . the mean ism pressure , @xmath107 ,
is related to the total gas density , @xmath108 , and mean electron temperature , @xmath109 , through @xmath110 . here , the total density @xmath108 is related to the electron density through @xmath111 . in a fully ionized plasma ,
the electron temperature is @xmath112 k and the ism pressure is directly proportional to the electron density ( see e.g. , * ? ? ? * for a discussion ) .
the mean ism pressure is related to the mean mechanical luminosity flux from the central star clusters within regions . in the standard adiabatic shell model for regions ,
regions expand until the internal pressure from mass - loss and supernova energy - driven bubbles equals the ambient pressure of the ism @xcite .
therefore , in the expanding bubble model , the ism pressure is inversely proportional to the region radius @xcite .
we calculate the electron density using the [ ] doublet ratio , [ ] @xmath43/ [ ] @xmath44 .
we assume a four - level model atom in our mappings iv models .
the relationship between the [ ] @xmath43/ [ ] @xmath44 and electron density given by mappings iv is identical to the [ ] -electron density relationship in @xcite .
when the [ ] ratio approaches @xmath113 , the electron density is in the low density limit ( i.e. , @xmath114 @xmath115 ) .
only one out of five of the cosmos cluster galaxies with measurable [ ] ratios is in the low density limit . in figure
[ niiha_ne ] , we show the electron density versus the [ ] /h@xmath1 ratio for the cosmos cluster and the @xmath0 field galaxies from our [ ] sample . the cluster galaxies span a similar range of electron densities as the field galaxies at the same redshift .
the two - sided kolmogorov - smirnov statistic yields a d - value of 0.4 , with a significance level of 0.54 , indicating that the two samples are not significantly different .
clearly , there is no systematic difference in the electron density among our cluster and field galaxies at @xmath0 , within the limitations of our small sample size and relatively large errors .
the ionization state of the gas ( or `` ionization parameter '' ) is traditionally measured using the ratio of two ionization states of the same atomic species .
the most commonly used ionization parameter diagnostic is the [ ] @xmath33/ [ ] @xmath116 ratio .
this ratio is usually calibrated as a function of ionization parameter and metallicity using stellar population synthesis and photoionization models ( e.g. , * ? ? ?
* ; * ? ? ?
, we use the [ ] / [ ] and [ ] /h@xmath2 ratios to distinguish the ionization parameter from metallicity , because the rest - frame blue [ ] line is not available within the wavelength coverage of our spectra . the [ ] / [ ] ratio is sensitive to ionization parameter because the ionization potentials of the [ ] @xmath117 and [ ] @xmath118 lines differ significantly .
the ratio of [ ] / [ ] is also sensitive to metallicity , particularly for metallicities below @xmath39@xmath119 .
this metallicity sensitivity occurs because the ratio of collisional excitation rates of [ ] and [ ] is a weak function of nebular temperature , which becomes high at low metallicities due to the lack of coolants in the nebula .
we note that although sulphur is a typical alpha element ( i.e. its abundance scales directly proportional to oxygen ) @xcite , nitrogen has both primary and secondary nucleosynthetic origins , and is sensitive to the metallicity in the secondary nucleosynthetic regime ( @xmath39@xmath40 ) . to account for this effect , our models use an empirical function for n / o versus o / h which takes into account both the primary and secondary nucleosynthetic components , as described in @xcite .
the [ ] /h@xmath2 ratio helps to separate the ionization parameter from metallicity , particularly below @xmath39@xmath119 , where the [ ] / [ ] ratio becomes insensitive to ionization parameter . in this regime , the [ ] /h@xmath2 ratio is primarily sensitive to the ionization parameter . at higher metallicities , @xmath39@xmath120 ,
the oxygen lines in the optical and the infrared dominate the cooling of the nebula , leading to less collisional excitations of the @xmath84 ion , and a stronger dependency between [ ] /h@xmath2 and metallicity . in figure [ niisii_oiiihb ]
, we show theoretically how the ionization parameter ( yellow - red curves ) and the metallicity ( green - blue curves ) are related on the [ ] / [ ] versus [ ] /h@xmath2 diagram . for ionization parameters @xmath121 , the [ ] / [ ] and [ ] /h@xmath2 line ratios successfully separate the ionization parameter from metallicity .
we show cluster and @xmath0 field galaxies with @xmath28 detections of [ sii ] ( red circles and black circles respectively ) , as well as cluster galaxies with less reliable @xmath37 detections of [ sii ] ( orange circles ) .
the majority of local star - forming galaxies have ionization parameters between @xmath122 , while our cosmos and field galaxies have larger ionization parameters , i.e. @xmath123 . the two galaxies with the largest [ ] /h@xmath2 ratios lie above the theoretical curves , suggesting a possible small ( i.e. @xmath87% ) agn contribution .
additional information from the x - rays , ir , or radio is required to confirm the presence of an agn in these galaxies . the electron density of our models ( described in * ? ? ?
* ) is fixed by the ism pressure , and corresponds to @xmath124 @xmath103 .
an electron density of @xmath125 @xmath103 is consistent with all of our cluster galaxies and 2/3 of our field galaxies shown on figure [ niisii_oiiihb ] , within the errors .
we note that if the model electron density is increased from @xmath125 @xmath103 to an extreme value of @xmath126 @xmath103 , the theoretical [ ] /h@xmath2 ratio would rise by @xmath127 dex .
the [ ] line fluxes are subject to large uncertainties ( up to 1000 @xmath128 ) . within these large errors , two cluster galaxies and one field galaxy
could potentially have electron densities as large as @xmath126 @xmath103 .
we highlight the position of these three galaxies on figure [ niisii_oiiihb ] with a circular outline .
even with an @xmath126 @xmath103 , two out of the three high electron density galaxies would still have ionization parameters larger than local sdss galaxies at the same stellar mass .
large ionization parameters have been seen in high redshift galaxies in previous work ( e.g. , * ? ? ?
* ) , but the cause of such a large ionization parameter is unknown . @xcite and
@xcite suggest that higher electron densities and a larger escape fraction of h - ionizing photons may be responsible for the larger ionization parameter .
on the other hand , @xcite suggest that a high effective temperature in the massive stellar population , possibly from stellar rotation and/or binaries may be responsible for the large ionization parameters .
our sample is too small to reach statistically significant conclusions on the cause of the high ionization parameter in our cluster and field galaxies , especially given that four of our galaxies lie along the edge of the ionization parameter - metallicity grid in figure [ niisii_oiiihb ] where distinguishing between ionization parameters above @xmath129 is impossible .
we note that the cluster galaxy with the lowest ionization parameter ( @xmath130 ) has a significantly lower specific star formation rate ( @xmath131 ) than the majority of our cluster galaxies ( mean @xmath132 ) .
if there is a relationship between specific star formation rate and ionization parameter , then this may have important implications on the geometry and mass of the star - forming clusters at high redshift .
in this paper , we have investigated the rest - frame optical line ratios of the cosmos galaxy cluster at redshift @xmath133 .
we compare the line ratios of the cosmos cluster with field galaxies at the same redshift .
we incorporate theoretical stellar evolution , photoionization , and shock models into our analysis to investigate any changes in ism conditions , ism pressure , or ionization parameter as a function of redshift or as a function of environment .
we find that the cosmos cluster optical line ratios are indistinguishable from field galaxies at the same redshift .
we find no statistical difference among the ism pressure or ionization parameter of field or cluster galaxies , within the limited size of our sample .
we conclude that for the cosmos cluster , the cluster environment makes no or limited impact on the ism properties of its member galaxies .
we also rule out shocks as a dominant contributor to the optical emission - lines in our cluster galaxies .
we show that the cosmos cluster has larger [ ] /h@xmath1 and/or [ ] /h@xmath2 ratios than local galaxies , similar to the large line ratios seen in field galaxies at high redshift .
this offset is consistent with a theoretical change in ism conditions with redshift .
we investigate possible causes of this change of ism conditions using our theoretical models .
we rule out a change in ism pressure as the dominant cause of the extreme ism conditions in our high redshift sample .
we use the [ ] / [ ] versus [ ] /h@xmath2 diagram to distinguish the ionization parameter from metallicity .
we show that the large line ratios are likely to result from a large ionization parameter ( @xmath123 ) .
the ionization parameters in our cosmos cluster and field galaxies are significantly larger ( by up to 1 dex ) than seen in local galaxies .
we suggest that there may be a relationship between ionization parameter and specific star formation rate . a larger sample of high - redshift galaxies with ionization parameter and specific star formation rate measurements is required to test this idea and to explore possible implications on the region geometry and star cluster properties .
the authors thank the referee for an excellent and comprehensive referee report .
l.j.k . gratefully acknowledges the support of an arc future fellowship and arc discovery project dp130103925 .
l.j.k . and m.a.d
. also acknowledge the support of arc discovery project dp130104879 .
the writing of this paper would not have been possible without the anu chelt 2014 academic women s writing workshop . l.j.k .
thanks the rsaa interstellar plotters and the gears-3d group for providing a supportive and stimulating environment under which to conduct this research .
g.g.k acknowledges the support of the australian research council through the award of a future fellowship ( ft140100933 ) .
this research has made use of nasa s astrophysics data system bibliographic services and the nasa / ipac extragalactic database ( ned ) . data in this paper were obtained at the w.m .
keck observatory , made possible by the generous financial support of the w.m .
keck foundation , and operated as a scientific partnership among caltech , the university of california and nasa .
we recognize and acknowledge the very significant cultural role and reverence that the summit of mauna kea has always had within the indigenous hawaiian community . | we investigate the ism properties of 13 star - forming galaxies within the @xmath0 cosmos cluster .
we show that the cluster members have [ ] /h@xmath1 and [ ] /h@xmath2 emission - line ratios similar to @xmath0 field galaxies , yet systematically different emission - line ratios ( by @xmath3 dex ) from the majority of local star - forming galaxies .
we find no statistically significant difference in the [ ] /h@xmath1 and [ ] /h@xmath2 line ratios or ism pressures among the @xmath0 cluster galaxies and field galaxies at the same redshift . we show that our cluster galaxies have significantly larger ionization parameters ( by up to an order of magnitude ) than local star - forming galaxies .
we hypothesize that these high ionization parameters may be associated with large specific star formation rates ( i.e. a large star formation rate per unit stellar mass ) .
if this hypothesis is correct , then this relationship would have important implications for the geometry and/or the mass of stars contained within individual star clusters as a function of redshift . |
in recent years , there have been continuing interests in the electromagnetic @xmath10 transition .
one of the earlier interests was due to the work of becchi and morpurgo @xcite . in that work
they had shown that in the context of the symmetric , non - relativistic , @xmath11 quark model , the transition @xmath12 is a pure magnetic dipole @xmath13 and the contribution from the electric quadrupole @xmath14 is zero .
one crucial assumption in their derivation of this ` selection rule ' is that the quarks in both the nucleon and the delta are in the zero orbital angular momentum states .
these predictions were to be considered as a check to the validity of the quark model , which was still questionable in those days .
later experimental measurements @xcite showed that , indeed , the magnetic dipole @xmath13 contributes predominantly to the @xmath12 transition , while the contribution from @xmath14 is small but non - vanishing . the non - vanishing value of @xmath14 , and also the coulomb quadrupole @xmath2 in the case of a virtual photon , has generated much theoretical interest .
one way to account for @xmath15 is through the @xmath16-wave mixtures in the @xmath17 and @xmath18 wave functions @xcite .
another way is through the two - body electromagnetic currents from one - gluon and/or one - pion exchange between constituent quarks @xcite . in the latter case
, it is argued that @xmath14 transition is , mainly , due to a two - quark spin - flip operator . on the other hand , in the large @xmath19 limit of quantum chromodynamics
it has been shown @xcite that @xmath20 is of order @xmath21 . to derive this result ,
no assumption about orbital angular momentum of the quarks was necessary .
more recent work in this direction can be found in ref .
@xcite .
another major issue related to the @xmath12 transition , and in general to any hadronic exclusive process , is the applicability of perturbative qcd ( pqcd ) at the range of values of momentum transfer @xmath3 accessible in the current generation of experiments . in terms of the ratios
@xmath22 and @xmath23 , pqcd power counting predicts that @xcite , in the limit @xmath24 , @xmath25 and @xmath26 , up to logarithmic corrections to be discussed in this paper . the former prediction has not yet been observed experimentally . in fact , up to @xmath27 gev@xmath9
, @xmath22 stays negative and very close to zero @xcite .
the comparison between data and pqcd prediction for @xmath7 is the main topic of this paper . to make the pqcd predictions more relevant at finite @xmath3 where
data have been and will be taken , one has to go beyond the asymptotic power counting , make detailed pqcd calculations of hard scattering amplitudes and derive the factorization formula for the experimental observables . in the present case , the relevant quantities are the three independent helicity matrix elements @xcite : @xmath28 where @xmath29 is the mass of the nucleon , @xmath30 , and @xmath31 . a pqcd calculation of the helicity - conserving matrix element @xmath32 has been known for many years @xcite
. however , a pqcd calculation of the helicity non - conserving amplitudes has not been explored in the literature until recently , because _ it necessarily involves the orbital motion of the quarks _
@xcite . in this paper
, we perform a pqcd calculation for the helicity - flip matrix element @xmath1 .
the technical details are basically similar to those in the calculation of the pauli form factor @xmath33 of the proton performed in @xcite . in the calculation , the quarks in the nucleon and the @xmath18 have one unit of relative orbital angular momentum along the direction of motion ( taken as the @xmath34 axis ) _ i.e. _ , @xmath35 . with this
we calculate the `` hard amplitudes '' which , in turn , are to be convoluted with the soft light - cone distribution amplitudes .
the light - cone wave functions of the nucleon and delta have been classified according to their orbital angular momentum dependence in @xcite .
note that we have to add the gluon contribution to maintain color gauge invariance .
we neglect the dynamical gluon effects proportional to the gluon field strength @xmath36 which seems numerically suppressed in general @xcite .
we remark that a detailed calculation of the hard amplitudes for @xmath37 will involve two units of quark relative orbital angular momentum . due to computational complexity
it will not be pursued here .
nevertheless , it can be easily shown that @xmath37 is suppressed , in the high @xmath3 limit by @xmath38 relative to @xmath1 @xcite . assuming our calculation for @xmath1 is relevant at @xmath3 that is currently explored at jefferson lab , our result and a recent experimentally measured values of @xmath7 enable us to make a phenomenological prediction of this ratio for higher values of @xmath3 .
we mention in this regard that the double logarithmic term , @xmath39 , will play an important role in our analysis . the importance of @xmath39 has also been demonstrated recently in the jefferson lab data on the nucleon elastic form factor @xmath40 @xcite .
our plan of representation is as follows : in section ii we give the notations and definitions needed for the coming sections . in section
iii we present in some detail the calculations performed and comment on the obtained results . in section
iv we perform a phenomenological analysis and make a prediction for @xmath7 .
section v will serve as a summary .
in this paper , we treat the bare delta as if it were a bound state of qcd , although in experiment it appears only as a resonance in certain scattering cross sections .
our analysis is relevant for scattering at the resonance peak where the contribution from the background vanishes by definition .
we will not consider the so - called the dressing of the resonance due to the pion cloud which might be important at small @xmath3 @xcite .
consider the following scattering : @xmath41 which appears as a sub - process of the electro - production of pions .
we are interested in the @xmath1 matrix element , in which case the photon is logitudinally - polarized and with a large virtuality @xmath42 .
our analysis will be carried out in a frame in which the virtual photon @xmath43 , the incoming proton and the outgoing @xmath18 are collinear .
the three momenta of the proton and the @xmath18 are in opposite directions and the @xmath43 is moving in the @xmath44 direction .
the photon polarization vector is given by : @xmath45 with @xmath46 .
we will also need the light - cone wave functions of the proton and the @xmath47 ( with the same charge as the proton ) .
these wave functions are given by @xcite : @xmath48[2][3]\varepsilon^{ijk}\left\{\psi_p^{(1)}(1,2,3 ) u_{i\uparrow}^\dagger(1)\left [ u_{j\downarrow}^\dagger ( 2)d_{k\uparrow}^\dagger ( 3)-d_{j\downarrow}^\dagger(2 ) u_{k\uparrow}^\dagger ( 3)\right ]
in this section , we derive a leading - order pqcd factorization formula for the helicity - flip @xmath71 matrix element @xmath1 . as shown in eq .
( 1 ) , this is generated by a virtual photon with longitudinal polarization , corresponding to the coulomb quadrupole amplitude .
the first step in calculating @xmath1 is to obtain the hard scattering amplitudes of partons . for this
we follow the guidelines of ref . @xcite . in principle , one has to consider both three - quark scattering with one unit of orbital angular momentum or three - quark - one - gluon scattering without orbital motion .
here we consider explicitly only the former and take into account the latter through the requirement of color gauge invariance , leaving out the contribution with the gluon field strength tensor .
this practice of leaving out dynamical gluon contribution corresponds to the so - called wendzura - wilczek approximation in the literature @xcite .
a key point in the calculation is that since the quark masses are negligible , the quark helicity is conserved .
this is due to the vectorial nature of the electromagnetic coupling between the struck quark and the virtual photon and of qcd quark - gluon coupling @xcite .
when a fock component of the light - cone wave function of the proton contains two quarks of the same flavor and helicity then , clearly , it contributes only when there are two quarks of the same flavor and helicity in the @xmath18 wave function .
the exchange contribution that results due to the presence of the same two quarks has to be taken into account properly through the second quantized calculation .
transition amplitudes .
the mirror diagrams should be added.,height=9 ] using the light - cone wave functions of the proton and the @xmath18 in eqs.(4 ) and ( 5 ) and keeping terms that are linear in the quark transverse momenta , one obtains : @xmath72[2][3]d[1^\prime ] [ 2^\prime ] [ 3^\prime ] \left\{k'^-_{1\perp}\psi^{(3)*}_\delta(2^\prime , 1^\prime , 3^\prime ) \psi^{(1)}_p ( 1,2,3 ) \right . \nonumber\\ & & \times \left [ t_2(1,2,3,1^\prime , 2^\prime , 3^\prime ) + t_2(1,2,3,3^\prime , 2^\prime , 1^\prime ) - t_3(1,2,3,1^\prime , 2^\prime , 3^\prime ) -t_3(1,2,3,3^\prime , 2^\prime , 1^\prime ) \right ] \nonumber\\ & & + \psi^{(1)*}_\delta ( 1^\prime , 3^\prime , 2^\prime ) ( k_{2\perp}^+ \psi_p^{(3 ) } + k_{1\perp}^+ \psi_p^{(4 ) } ) ( 2,1,3 ) \nonumber\\ & & \left .
\times \left [ t'_2(1,2,3,1^\prime , 2^\prime , 3^\prime ) + t'_2(3,2,1,1^\prime , 2^\prime , 3^\prime ) - t'_3(1,2,3,1^\prime , 2^\prime , 3^\prime ) -t'_3(3,2,1,1^\prime , 2^\prime , 3^\prime ) \right ] \right\}\end{aligned}\ ] ] where @xmath73 and @xmath74 are the three - quark scattering amplitudes introduced in the previous section .
the color wave function of the three quarks is normalized to unity .
the perturbative feynman diagrams for @xmath73 are given in fig .
there are fourteen such diagrams , taking into account the relative position of the photon vertex and the gluon interactions . in each diagram
we have three collinear incoming and outgoing quarks exchanging two gluons
. the photon could interact with any one of the quarks . with the spin - isospin structure of the proton and
the @xmath18 state functions , we assume that the first and third quarks have positive helicities while the second quark has negative one . to find the amplitudes @xmath73 , we let the outgoing quarks carry orbital angular momenta @xmath75 and the incoming quarks have zero orbital angular momenta , _
i.e. _ , @xmath76 . from each one of the feynman diagrams we write the amplitude following the usual qed and qcd feynman rules .
since the calculation is performed in the collinear frame in which the particles are highly relativistic , we can set the masses of the proton and the @xmath18 to zero .
we expand the quark spinors in the final state and the quark propagators to first order in the transverse momenta .
then we collect all such contributions from the given diagram and sum up the results of all the feynman diagrams .
this yields @xmath77 . to find @xmath78 we set @xmath79 and @xmath80 and follow similar steps .
the results for @xmath81 at order of @xmath82 are as follows : @xmath83 , \end{aligned}\ ] ] where @xmath84 and the functions @xmath85 and @xmath86 are : @xmath87 similarly for @xmath88 we have : @xmath89 , \end{aligned}\ ] ] the function @xmath90 is given by : @xmath91 to obtain @xmath92 we interchange @xmath93 and @xmath94 and to obtain @xmath95 we replace @xmath96 with @xmath97 and interchange @xmath58 and @xmath61 .
the above results are consistent with those derived in ref . @xcite . to further simplify the above expressions ,
let us define @xmath98-intergrated wave functions : @xmath99 \psi^{(1)}_p ( k_1,k_2,k_3 ) \nonumber \\
\phi^{(4)}_p(x_1,x_2,x_3 ) & = & 2 \int [ d^2{\bf k_\perp } ] { \bf k_{3 \perp}}\cdot ( { \bf k_{2\perp}}\psi^{(3)}_p + { \bf k_{1\perp}}\psi^{(4)}_p ) ( k_2,k_1,k_3 ) , \nonumber\\ \psi^{(4)}_p(x_1,x_2,x_3 ) & = & 2 \int [ d^2{\bf k_\perp } ] { \bf k_{1 \perp}}\cdot ( { \bf k_{2\perp}}\psi^{(3)}_p + { \bf k_{1\perp}}\psi^{(4)}_p ) ( k_2,k_1,k_3 ) \nonumber \\
\phi^{(3)}_{\delta}(x_1,x_2,x_3 ) & = & 2 \int [ d^2{\bf k_\perp } ] \psi^{(1)}_{\delta } ( k_1,k_2,k_3 ) \nonumber \\
\phi^{(4)}_{\delta}(x_1,x_2,x_3 ) & = & 2 \int [ d^2{\bf k_\perp } ] { \bf k_{3\perp } } \cdot { \bf k_{1\perp } } \psi^{(3)}_\delta(k_2,k_1,k_3),\nonumber \\
\psi^{(4)}_{\delta}(x_1,x_2,x_3 ) & = & 2 \int [ d^2{\bf k_\perp } ] { \bf k_{1\perp } } \cdot { \bf k_{1\perp } } \psi^{(3)}_\delta ( k_2,k_1,k_3 ) \end{aligned}\ ] ] where the integration measure is defined as : @xmath100 & = & \frac { d^2 k_{1\perp } d^2k_{2\perp } d^2k_{3\perp}}{(2\pi)^6 } \delta^{(2)}(\bf k_{1\perp}+ \bf k_{2\perp}+ \bf k_{3\perp } ) . \end{aligned}\ ] ] in terms of these wave functions and by using the obtained expressions for the hard amplitudes we get the following expression for @xmath1 : @xmath101[dy ] \nonumber\\ & & \big\ { \phi^{*(4)}_{\delta}(y_1,y_2,y_3)\phi^{(3)}_p(x_1,x_2,x_3 ) \big [ w_0(y_1,y_2,y_3,x_1,x_2,x_3)+w_0(y_1,y_2,y_3,x_3,x_2,x_1 ) \nonumber\\ & & -w_1(x_3,x_2,x_1,y_1,y_2,y_3 ) -w_2(x_3,x_2,x_1,y_3,y_2,y_1 ) \big ] \nonumber\\ & & + \psi^{*(4)}_{\delta}(y_1,y_2,y_3)\phi^{(3)}_p(x_1,x_2,x_3 ) \big [ w_0(y_3,y_2,y_1,x_3,x_2,x_1 ) + w_0(y_3,y_2,y_1,x_1,x_2,x_3 ) \nonumber\\ & & -w_1(x_3,x_2,x_1,y_3,y_2,y_1 ) -w_2(x_3,x_2,x_1,y_1,y_2,y_3 ) \big ] \nonumber\\ & & + \phi^{(4)}_p(x_1,x_2,x_3)\phi^{*(3)}_\delta ( y_1,y_3,y_2 ) \big [ w_0(x_3,x_2,x_1,y_3,y_2,y_1 ) + w_0(x_3,x_2,x_1,y_1,y_2,y_3 ) \nonumber\\ & & -w_1(y_3,y_2,y_1,x_3,x_2,x_1 ) -w_2(y_3,y_2,y_1,x_1,x_2,x_3)\big ] \nonumber\\ & & + \psi^{(4)}_p(x_1,x_2,x_3)\phi^{*(3)}_\delta ( y_1,y_3,y_2 ) \big [ w_0(x_1,x_2,x_3,y_1,y_2,y_3 ) + w_0(x_1,x_2,x_3,y_3,y_2,y_1 ) \nonumber\\ & & -w_1(y_3,y_2,y_1,x_1,x_2,x_3)-w_2(y_3,y_2,y_1,x_3,x_2,x_1 ) \big ] \big\ } , \end{aligned}\ ] ] where @xmath102= dx_1 dx_2 dx_3 \delta ( 1-x_1-x_2-x_3)$ ] and @xmath103 .
we see from eq.(13 ) that @xmath1 scales like @xmath104 in the high @xmath3 limit , consistent with the general power counting . to find the normalization of @xmath1 we need to know the light - cone distribution amplitudes defined in eq.(12 ) . for the proton , a set of such functions
have been given in ref .
@xcite based on conformal expansion , qcd sum rules and lorentz symmetry . for the @xmath18 no such work has been done explicitly .
however if the @xmath18 is to be treated as a three - quark state , we believe that , at the asymptotically large @xmath3 , the light - cone distribution amplitudes of the @xmath18 will have a @xmath61-dependence which is identical to the @xmath58-dependence of the light - cone distribution amplitudes of the proton @xcite . for the proton we have , for example , the asymptotic form of @xmath105 is @xmath106 and @xmath107 is @xmath108 @xcite .
if we perform the @xmath58 and @xmath61 integration in eq.(13 ) with such functions we get , due to end - point singularities , a double logarithmic divergence results .
this divergence indicates the quarks with small feynman @xmath5 contribute significantly to the hard scattering .
however , for very small @xmath5 for which the parton longitudinal momentum is on the order of @xmath109 , the hard scattering picture breaks down , and the above calculation is invalid .
then one has to add the so - called feynman contribution which is relatively unimportant in the limit of large @xmath3 .
therefore , these divergent integrals are regulated physically by a cut - off from below of order @xmath110 where @xmath111 is some parameter that represents the soft energy scale ( order of few hundreds of mev ) . with this cut - off
, the result of the momentum fraction integration will be a @xmath3-dependent term of the form , @xmath4 .
thus we can write : @xmath112 where @xmath113 is a numerical factor that depends on the explicit expressions of the light - cone distribution amplitudes . with this form of @xmath1 and
with the fact that @xmath37 is of order @xmath114 relative to @xmath32 we will give in the next section a phenomenological prediction of the ratio @xmath7 in the high @xmath3 limit .
in the electro - production of the @xmath18 resonance , the multipoles @xmath13 , @xmath14 and @xmath2 of the exchanged virtual photon are the only ones that contribute @xcite .
one of the experimentally extracted quantities related to this process is the ratio @xmath23 . in order to express this ratio in terms of the helicity matrix elements @xmath115
we introduce the resonance `` helicity amplitudes '' @xmath116 , @xmath117 and @xmath118 @xcite .
these amplitudes are computed in the rest frame of the @xmath18 and the sub - indices refer to the helicity of the @xmath18 .
the scalar amplitude @xmath118 is relevant only for virtual photons .
these amplitudes are related to the helicity matrix elements through the relations @xcite : @xmath119 where @xmath120 is some kinematical factor and @xmath121 is the photon three - momentum in the rest frame of the @xmath18 given by : @xmath122 .
the multipole @xmath2 is proportional to @xmath118 and we also have @xcite : @xmath123 from the results of the previous section we see that @xmath117 is of order @xmath114 relative to @xmath116 in the high @xmath3 limit and thus it will be neglected . with this approximation
the ratio @xmath7 becomes : @xmath124 substituting in the last equation the expression for @xmath125 and using eq.(14 ) we get : @xmath126 the pre - factor @xmath127 is determined by taking @xmath128 gev and by fitting eq .
( 18 ) with the recently obtained experimental values of @xmath7 @xcite : @xmath129 the result is : @xmath130 .,height=9 ] @xmath7 is displayed in figure 2 . from that figure
it is clear that @xmath7 has a slow variation at the presently - accessible momentum transfer @xmath3 at jlab which is about 6 @xmath8 7 gev@xmath9 .
we remark here that for the e2-to - m1 ratio @xmath22 , two units of orbital angular momentum are need to induce the transition , and higher powers of the logarithmic correction may arise when one tries to calculate the matrix element @xmath37 .
a perturbative qcd calculation of the helicity flip matrix element of the electromagnetic @xmath12 transition has been given .
we showed that the transverse momenta of the quarks in the proton and @xmath18 is essential to obtain a finite result .
an explicit calculation of the hard amplitude @xmath131 and @xmath132 has been given and the techniques of the calculation were outlined in some detail .
the essential steps of the calculation are to draw the relevant perturbative feynman diagrams and then to compute the contribution from each such diagram by expanding the quark spinors and propagators to first power of the transverse momenta .
the hard amplitudes are then to be convoluted with light - cone amplitudes .
our pqcd results for the scaling behavior of @xmath32 , @xmath1 and @xmath37 confirm earlier scaling prediction @xcite .
in addition we outlined how one obtains the double logarithmic correction . based on our result for @xmath115 we gave a phenomenological prediction for the ratio @xmath7 . roughly speaking ,
@xmath133 will be of order of @xmath134 at @xmath3 of order 10 gev@xmath9 .
99 c. becchi , g. morpurgo , phys .
17 ( 1965 ) 352 .
dalitz , d.g .
sutherland , phys .
146 ( 1966 ) 1180 .
s. s. gershtein , g. v. dzhikiya , sov . j. nucl
34 ( 1982 ) 870 ; n. isgur , g. karl , r. koniuk , phys . rev .
d 25 ( 1982 ) 2394 .
buchmann , e. hernndez , a. faessler , phys .
c 55 ( 1977 ) 448 ; a. j. buchmann , e. hernndez , k. yazaki , phys . lett .
b 269 ( 1991 ) 35 ; nucl . phys . a 569 ( 1994 ) 661 .
e. jenkins , x. ji , a. manohar , phys .
( 2002 ) 242001 .
buchmann , j.a hester , r.f lebed , phys .
d 66 ( 2002 ) 056002 .
carlson , phys . rev .
d 34 ( 1986 ) 2704 .
et al . _ ,
78 ( 1997 ) 606 .
f. kalleicher _
et al . _ ,
z. phys . a 359 ( 1997 ) 201 . v. v. frolov _ et al .
82 ( 1999 ) 45 .
kamalov , shin nan yang , phys
83 ( 1999 ) 4494 .
t. sato , t.s.h .
lee , phys .
c 54 ( 1996 ) 2660 .
brodsky , g.p .
lepage , phys .
43 ( 1979 ) 545 .
lepage , s.j .
brodsky , phys .
d 22 ( 1980 ) 2157 .
x. ji , j .-
ma , f. yuan , phys .
( 2003 ) 241601 .
belitsky , x. ji , f. yuan , hep - ph/0212351 , to appear in phys .
x. ji , j .-
ma , f. yuan , hep - ph/0304107 . s. wendzura , f. wilczek , phys . lett . 72 ( 1977 ) 195 . v.m .
braun , r.j .
fries , n. mahnke , e. stein , nucl .
b 589 ( 2000 ) 381 .
a. idilbi , x. ji , and j .-
ma , to be published .
jones , m.d .
scadron , ann .
( n.y ) 81 ( 1973 ) 1 .
l. durand , p.c .
decelles , r.b .
marr , phys .
126 ( 1962 ) 1882 .
dufner , y.s .
tsai , phys .
168 ( 1968 ) 1801 .
devenish , t.s .
eisenschitz , j.g .
krner , phys .
d 14 ( 1976 ) 3063 .
l. a. copley , g. karl , e. obryk , phys .
b 299 ( 1969 ) 117 ; nucl . phys .
b 13 ( 1969 ) 303 .
p. stoler , phys .
226 ( 1993 ) 103 .
r. m. davidson , nimai c. mukhopadhyay , r. s. wittman , phys .
d 43 ( 1991 ) 71 . | we present a leading - order pqcd calculation of the helicity - flip @xmath0 matrix element @xmath1 ( coulomb quadrupole amplitude @xmath2 ) , taking into account the transverse momenta of the quarks and the contribution from the gluons . in the large @xmath3 limit ,
its scaling behavior acquires a double - logarithmic correction @xmath4 compared with the standard scaling analysis , due to the contribution from the orbital motion of the small-@xmath5 partons .
based on this and on the latest jlab experimental results of the @xmath6 ratio @xmath7 at @xmath3 = 3 @xmath8 4 gev@xmath9 , we make a phenomenological prediction for the latter at higher values of @xmath3 . |
l dwarfs make up a new spectral class of objects that are believed to have masses near or below the hydrogen - burning limit ( kirkpatrick et al . 1999a ; 1999b ) .
many satisfy currently accepted criteria for identification as _ bona fide _
brown dwarfs ( see tinney 1999 for a review ) .
their local field detection rate in infrared sky surveys suggests they comprise a sizeable population which is well represented by an extension of the field - star mass function , @xmath0 , with @xmath1 ( reid et al . 1999 ) .
the occurrence frequency of multiplicity among these systems is completely unknown ; it is an open question as to whether the distribution of their companions matches that of m dwarfs or bears the stamp of a different , sub - stellar formation mechanism .
stellar companions are detected in approximately 35% of m dwarf systems with a distribution peaking at a radius in the range @xmath2 au ( fischer & marcy 1992 ; henry & mccarthy 1993 ; reid & gizis 1997 ) . efforts to uncover the mass and radial distribution of extra - solar planets around m stars are just beginning to meet with success and have revealed super jovian - mass planets within a few au of their central stars , consistent with results for earlier spectral types ( marcy et al.1998 ) .
the relationship of this population to that of binary companions and planetary systems like our own is a topic of current debate ( black 1997 ) .
the true answer will not be readily apparent until a more complete range of mass and orbital distances has been surveyed .
ground - based imaging of l dwarfs provides a unique piece to this puzzle , since the reduced glare of low - luminosity primaries affords increased sensitivity to very faint companions . to date
, very few multiple l dwarf systems have been identified .
several l dwarf secondaries have been discovered around nearby stars ( becklin & zuckerman 1988 ; rebolo et al .
1998 ; kirkpatrick et al .
1999b ) . among a handful of binary systems believed to be composed of two brown - dwarf components ( e.g. , basri & martn 1997 ) , only two have primary spectral types as late as l : 2massw j0345 is a double - lined spectroscopic l dwarf system ( reid et al .
1999 ) , and denis - p j1228 was shown to be double in hst imaging observations ( martn et al .
the latter is composed of equal - luminosity components with a projected separation of 0.275@xmath3 ( 5 au at the 18 pc distance of denis - p j1228 ) . here
we present the first results of a keck near - infrared imaging survey of a large sample of l dwarfs . at a projected distance of a few au
, our program is capable of detecting companions with luminosity similar to the primary . at further projected distances ,
our survey is sensitive to objects which are several magnitudes fainter than the methane brown dwarf , gl 229b . in this work ,
we report the k - band detection of three l dwarf binaries , including denis - p j1228 .
our target sample was culled from the 2mass and denis near - infrared sky surveys and consisted of objects spectroscopically confirmed to be l dwarfs .
we also included observations of a smaller sample of nearby very late m dwarfs .
imaging was carried out at the keck i telescope with nirc , a cryogenically - cooled near - infrared camera that incorporates a 256@xmath4256 indium - antimonide array at the f/25 focus in an optical framework which yields a 0.15@xmath3 plate scale and 38@xmath3-square field of view ( matthews & soifer 1994 ) .
one - minute integrations were taken in the k - band filter at each of nine dithered positions separated by 5@xmath3 .
individual frames were flat - fielded and differenced , then shifted and combined to create a sensitive composite image suitable for detecting companions to a limiting magnitude of m@xmath5 . at this level of sensitivity ,
several additional sources were typically detected in each frame .
repeat observations in a second epoch were taken one year or more later to determine if any of these share common proper motion with the target ; second - epoch observations are complete for only a subset of the sample which includes 10 l dwarfs at present .
analysis of the completed survey will be presented in a future work .
in addition to the common proper motion analysis of faint sources , we inspected the core of each of the primaries to search for extended emission associated with a marginally resolved binary .
second - epoch observations not only served to provide evidence of common proper motion , but also helped to ensure that any elongation was not due to time - dependent errors in phasing of the segmented primary mirror .
point - like sources observed nearby in the sky and within an hour of the target observations were chosen to serve as psf measurements .
dithered images of candidate binaries and psf stars were not shifted and combined but were treated as independent data sets .
psf stars were fit in duplicate to each of the candidate binary images using a least - squares minimization method .
typically , nine psf frames were fit to each of nine image frames for a total of 81 independent fits .
properties of the psf and target sources used in this work are listed in table 1 ; results of the psf - fitting are given in table 2 .
three objects met our critieria for reliable identification of a true close binary system , i.e. , the availability of suitable psf observations and consistent results in psf - fitting for at least two epochs .
contour plots of denis - p j1228 , denis - p j0205 , and 2massw j1146 are displayed in fig . 1 , together with the psf stars used to decompose them into separate components .
the results of psf - fits in each epoch are listed in table 2 and plotted in fig . 2 and 3 .
parameter estimates are consistent between two epochs ; variations in the uncertainties are largely due to different seeing conditions .
conservatively , we state here the mean of the measurements in all epochs ( rather than simply the best - seeing epoch ) together with an uncertainty calculated as the root mean square difference from the mean . for the separation and pa of the component positions ,
these are plotted as large symbols in fig . 2 and are @xmath6 , @xmath7 , @xmath8 and @xmath9 , @xmath10 , @xmath11 for the component separations and pa s of denis - p j1228 , denis - p j0205 , and 2massw j1146 , respectively .
projected separations correspond to physical separations of 4.9 , 9.2 , and 7.6 au at distances implied by parallaxes listed in table 1 .
histograms of the flux - component ratios are plotted in fig . 3 ; mean values are @xmath12 , @xmath13 , and @xmath14 .
_ denis - p j1228 _ denis - p j1228 was discovered by delfosse et al .
( 1997 ) and shown to have lithium by tinney et al .
( 1997 ) and martn et al .
it is an l5 v dwarf in the classification scheme of kirkpatrick et al .
( 1999a ) with a temperature less than 1700 k implied by the absence of vanadium oxide in its spectrum .
martn et al .
( 1999 ) first discovered the binary nature of denis - p j1228 and reported b / a flux ratios of 0.83 , 0.86 , and 0.87 in the hst / nicmos filters , f110 m , f135 m , and f165 m , respectively .
our k - band flux ratio for the same components , @xmath12 , is consistent with a trend in the hst fluxes which implies that denis - p j1228b is slightly cooler , as is the value obtained on the best night of seeing , @xmath15 .
the separation and pa reported by martn et al .
( 1999 ) are plotted in fig .
2 and were obtained in observations taken mid - way between keck observations spaced one year apart .
the results clearly agree over an observational period during which the system should have moved 0.21@xmath3 along pa 143.8@xmath16 according to measurements by dahn et al .
( 1999 ) that are listed in table 1 .
indeed , we measure the mean proper motion with respect to background stars in our field as @xmath17 = @xmath18 at pa @xmath19 . if either binary component is a background star , this would result in a later - epoch separation that is too large to be consistent with our observations .
thus denis - p j1228 is a true common proper motion binary .
a recently determined value of the trigonometric parallax for denis - p j1228 is listed in table 1 as [email protected]@xmath3 and implies a distance of @xmath21 pc
. this result agrees with the distance estimate of martn et al .
( 1999 ) , based on the spectroscopic parallax of delfosse et al .
( 1997 ) , and from which they derived a projected physical separation of 5 au . assuming circular orbits and masses of 0.05 m@xmath22 , they calculate that orbital motion can be detected by hst in one year .
a precise determination of dynamical mass will require observations over a much larger fraction of an orbit , however .
the orbital period is at least 35 years , so a highly refined mass estimate may not be forthcoming on timescales competitive with analysis of closer binaries discovered in the interim . nevertheless , this binary has the smallest projected physical separation of those presented here , hence , the best opportunity for near - term estimates of the dynamical mass .
the importance of such an observation is highlighted by the fact that , to date , secure dynamical mass determinations have barely begun to extend to masses below the hydrogen - burning limit ( cf .
henry et al . 1999 ) .
_ denis - p j0205 _ denis - p j0205 was first discovered by delfosse et al .
( 1997 ) and is reported in kirkpatrick et al .
( 1999a ) as a prototype of the l7 v spectral class .
although no lithium line is evident in its spectrum , this may indicate a sufficiently cool temperature to foster depletion of atomic li condensing out as licl .
it is , in any case , the coolest dwarf to date for which a similar or less luminous companion has been detected .
the proper motion measured for denis - p j0205 , 0.442@xmath3 at pa 82.6@xmath16 ( table 1 ) , is directed almost entirely along the 92@xmath16 pa derived for the binary and would have produced a total motion of about 0.6@xmath3 in the @xmath23 years between observations reported here .
this is greater than the binary separation itself , which is measured consistently at both epochs to be 0.51@xmath3 .
relative to 3 background stars detected in our field of view , we measure the total proper motion of the system to be @xmath24 along pa @xmath25 between epochs .
consequently , it is securely established that denis - p j0205 comprises a common - proper - motion pair .
the trigonometric parallax , [email protected] , implies a distance of 18 pc , similar to that of denis - p j1228 , and a projected separation of 9.2 au for the binary components . _
2massw j1146 _ the lithium dwarf , 2massw j1146 , is classified l3 v. with a trigonometric parallax @xmath26 = [email protected]@xmath3 , it is the most distant of the three ( d = 26.2 pc ) .
its angular separation , @xmath8 at pa @xmath11 , implies a projected distance of 7.6 au .
it has the smallest proper motion of the three , 0.097@xmath3 at pa 20.2@xmath16 , but in a direction that is nearly anti - parallel to the pa of binary separation .
consequently , differential proper motion would have been easily detected after the nearly 1-year interval separating our observations .
second - epoch observations had relatively poor seeing ; the resulting scatter in fig . 2 is evident .
nevertheless , the results still cluster about a mean coincident with that obtained at the earlier epoch , indicating common proper motion . for 2massw j1146 , a second set of images
was acquired in the first epoch using nirc s `` ch@xmath27 '' filter centered at @xmath28 = 2.269 @xmath17 m with @xmath29 = 0.155@xmath17 m . binary separation and pa estimates from fits to these data
are averaged into the results reported above .
since the ch@xmath27 flux component ratio agrees with the k - band ratios , within the uncertainties , we report here the average of all three : [email protected] .
the absence of any difference precludes a contrasting presence of ch@xmath27 absorption in one of the components , as expected for a luminosity ratio of 1 .
the three binary systems presented here have similar projected separations ( 5 to 9 au ) and luminosity ratios near unity .
they represent the first binary detections in preliminary analysis of a larger dual - epoch survey in which only 10 l - dwarf images have been completely analyzed in two epochs . no companions with wider separation or more highly contrasting luminosities
have been found thus far .
although analysis of our complete sample is required to make unassailable statements about the l dwarf binary distribution , these preliminary results suggest a conjecture for further testing : namely , that multiple systems are not uncommon in the l dwarf population , that their distribution peaks at radial separations like that of both jovian planets in our solar system and binary stars generally ( @xmath30 au ) , and that low - contrast mass ratios are common .
the latter claim is especially in need of testing , since our survey is not very sensitive to companions at the separations reported here if they have high luminosity contrast ratios .
further , the magnitude - limited surveys from which our sample is taken are biased toward the detection of equal - luminosity binaries , since their combined luminosity is greater than for single stars of the same spectral type .
ultimately , techniques with both high resolution and high dynamic range must be applied to a volume - limited sample to reliably identify the distribution of circumstellar bodies that encircle the members of this population of very cool objects .
we thank i.n .
reid for reading an early draft of the manuscript and for useful comments about the mass function of l dwarfs in the field . mass function . for useful discussions , we also thank e. martn with respect to background on denis - p j1228 and t.j .
henry regarding the practicalities of dynamical mass determinations .
we also wish to thank nirc instrument specialists robert goodrich and randy campbell , and operator assistants ron quick , joel aycock , wayne wack , and chuck sorenson .
data presented herein were obtained at wmko , which is operated as a scientific partnership between caltech , university of california , and nasa , and was made possible by the generous financial support of the w.m . keck foundation .
delfosse , x. , tinney , c.g . ,
forveille , t. , epchtein , n. , bertin , e. , borsenberger , j. , copet , e. , de batz , b. , fouque , p. , kimeswenger , s. , le bertre , t. , lacombe , f. , rouan , d. , tiphene , d. , 1997 , a&a , 327 , 25 lccccccc denis - p j0205 & @xmath31 & @xmath32 59@xmath33 29.6@xmath3 & l7 v & [email protected] & [email protected] & 0.442 & 82.6 lp 647 - 13 & @xmath34 & @xmath35 43@xmath33 26.0@xmath3 & m9 v & & & 0.359 & 85 lhs 2351 & @xmath36 & @xmath37 51@xmath33 48.0@xmath3 & m8 v & [email protected] & [email protected] & 0.49 & 318.5 2massw j1145 & @xmath38 & @xmath39 17@xmath33 30.0@xmath3 & l1.5 v & [email protected] & & & 2massw j1146 & @xmath40 & @xmath41 30@xmath33 53.0@xmath3 & l3 v & [email protected] & [email protected] & 0.097 & 20.2 denis - p j1228 & @xmath42 & @xmath43 47@xmath33 11.0@xmath3 & l5 v & [email protected] & [email protected] & 0.21 & 143.8 kelu-1 & @xmath44 & @xmath45 41@xmath33 06.0@xmath3 & l2 v & [email protected] & [email protected] & 0.294 & 270.6 lllcccc denis - p j0205 & 29jul97 & lp 647 - 13 & 0.33@xmath3 & @xmath46 & 106@xmath47 & @xmath48 denis - p j0205 & 24jan99 & lp 647 - 13 & 0.45@xmath3 & @xmath49 & 72@xmath50 & @xmath51 2massw j1146 & 14feb98 & 2massw j1145 & 0.24@xmath3 & @xmath52 & 196@xmath53 & @xmath54 2massw j1146@xmath55 & 16feb98 & 2massw j1145 & 0.36@xmath3 & @xmath56 & 225@xmath50 & @xmath57 2massw j1146 & 24jan99 & lhs 2243 & 0.40@xmath3 & @xmath58 & 197@xmath59 & @xmath60 denis - p j1228 & 14feb98 & kelu-1 & 0.25@xmath3 & @xmath61 & 45@xmath62 & @xmath63 denis - p j1228 & 09feb99 & kelu-1 & 0.28@xmath3 & @xmath64 & 17@xmath65 & @xmath15 | we present keck near - infrared imaging of three binary l dwarf systems , all of which are likely to be sub - stellar .
two are lithium dwarfs , and a third exhibits an l7 spectral type , making it the coolest binary known to date .
all have component flux ratios near 1 and projected physical separations between 5 and 10 au , assuming distances of 18 to 26 pc from recent measurements of trigonometric parallax .
these surprisingly similar binaries represent the sole detections of companions in ten l dwarf systems which were analyzed in the preliminary phase of a much larger dual - epoch imaging survey .
the detection rate prompts us to speculate that binary companions to l dwarfs are common , that similar - mass systems predominate , and that their distribution peaks at radial distances in accord both with m dwarf binaries and with the radial location of jovian planets in our own solar system . to fully establish these conjectures against doubts raised by biases inherent in this small preliminary survey ,
however , will require quantitative analysis of a larger volume - limited sample which has been observed with high resolution and dynamic range .
subject headings : stars : circumstellar matter binaries : spectroscopic planetary systems |
the most important feature of a toric variety @xmath0 is that it is completely described by a fan @xmath1 in the lattice @xmath2 of one parameter subgroups of the big torus @xmath3 .
applying a linear gale transformation to the set of primitive generators @xmath4 of the rays @xmath5 of @xmath1 , gives a new vector configuration in a rational vector space @xmath6 .
this opens an alternative combinatorial approach to the toric variety @xmath0 : the vector space @xmath6 is isomorphic to the rational divisor class group of @xmath0 , and one can shift combinatorial information between the spaces @xmath7 and @xmath6 . in toric geometry , this principle has been used to study the projective case , compare @xcite : roughly speaking , if we consider all fans @xmath8 in @xmath2 having their rays among @xmath9 , then the ( quasi-)projective @xmath8 correspond to so called gelfand - kapranov - zelevinski cones in @xmath6 .
these cones subdivide the cone generated by the gale transform of the vector configuration @xmath10 , and the birational geometry of the associated toric varieties is reflected by the position of their gelfand - kapranov - zelevinski cones . if one leaves the ( quasi-)projective setting , then there are generalizations of gelfand - kapranov - zelevinski cones , compare e.g. @xcite ; but so far there seems to be no concept which is simple enough to serve for practical purposes in toric geometry .
our aim is to fill this gap and to propose a natural combinatorial language which also works in the non quasiprojective case .
the combinatorial data are certain collections which we call _ bunches _ of overlapping cones in the vector space of rational divisor classes .
as we shall see , this approach gives very natural descriptions of geometric phenomena connected with divisors . in order to give a first impression of the language of bunches ,
let us present it here for in a special case , namely for toric varieties with free class group .
consider a sequence @xmath11 of not necessarily pairwise distinct points , called _ weight vectors _ , in a lattice @xmath12 . by a _ weight cone
_ in @xmath13 we then mean a ( convex polyhedral ) cone @xmath14 generated by some of the @xmath15 .
a _ ( free standard ) bunch _ in @xmath13 is a nonempty collection @xmath16 of weight cones in @xmath13 with the following properties : 1 .
a weight cone @xmath17 in @xmath13 belongs to @xmath16 if and only if @xmath18 holds for all @xmath19 with @xmath20 .
2 . for each @xmath21 ,
the set @xmath22 generates @xmath13 as a lattice , and there is a @xmath19 such that @xmath23 holds for the relative interiors . how to construct a toric variety from such a bunch @xmath16 ?
the first step is to _ unpack _ the combinatorial information encoded in @xmath16 . for this , let @xmath24 , and let @xmath25 denote the linear surjection sending the canonical base vector @xmath26 to @xmath15
. denote by @xmath27 the positive orthant .
we define the _ covering collection _ of @xmath16 as @xmath28 the next step is to _ dualize _ the information contained in @xmath29 .
this is done by a procedure close to a linear gale transformation , for the classical setup see e.g. @xcite and @xcite : consider the exact sequence arising from @xmath25 , and the corresponding dualized exact sequence : @xmath30 & { m } \ar[r ] & { e } \ar[r]^{q } & k \ar[r ] & 0 , } \ ] ] @xmath30 & { l } \ar[r ] & { f } \ar[r]^{p } & { n } \ar[r ] & 0 . } \ ] ] note that @xmath31 is not the dual homomorphism of @xmath32 .
let @xmath33 be the dual cone of @xmath27 .
the crucial observation then is that we obtain a fan @xmath34 in the lattice @xmath2 having as its maximal cones the images @xmath35 the _ toric variety associated to the bunch _ @xmath16 is @xmath36 .
the toric variety @xmath37 is _ nondegenerate _ in the sense that it has no torus factors .
moreover , @xmath37 is _ 2-complete _ , that means if @xmath38 is an open toric embedding such that the complement @xmath39 is of codimension at least 2 , then @xmath40 .
consider the sequence @xmath41 of weight vectors in @xmath42 and the bunch @xmath43 in @xmath44 .
then we have @xmath45 , and the associated linear map @xmath46 sends @xmath26 to @xmath21 . the covering collection @xmath29 consists of the following three faces of @xmath47 : @xmath48
if we identify the dual space @xmath49 with @xmath50 , then the cone @xmath51 is again @xmath52 .
moreover , we may identify @xmath2 with @xmath53 and thus realize the map @xmath31 via the matrix @xmath54\ ] ] each @xmath55 equals @xmath56 , where @xmath57 and @xmath58 .
hence the images @xmath59 , which are the maximal cones of the fan @xmath34 , are given in terms of the canonical base vectors @xmath60 as @xmath61 consequently , the toric variety @xmath37 associated to the bunch @xmath16 equals the weighted projective space @xmath62 . introducing a suitable notion of a morphism , we can extend the assignment @xmath63 from bunches to 2-complete nondegenerate toric varieties to a contravariant functor .
in fact , we even obtain a weak antiequivalence , see theorem [ standard2maximal ] : the functor @xmath63 induces a bijection on the level of isomorphism classes of bunches and nondegenerate 2-complete toric varieties . in particular , every complete toric variety arises from a bunch . in order to read off geometric properties of @xmath37 directly from the bunch @xmath16 , one has to translate the respective fan - theoretical formulations via the above gale transformation into the language of bunches .
this gives for example : * @xmath37 is @xmath64-factorial if and only if every cone @xmath19 is of full dimension , see proposition [ qfactorialchar ] .
* @xmath37 is smooth if and only if for every @xmath65 in @xmath29 , the image @xmath66 equals @xmath13 , see proposition [ smoothchar ] . *
the orbits @xmath37 have fixed points in their closures if and only if all cones @xmath67 , where @xmath68 , are simplicial , see proposition [ existenoffixedpointschar ] .
as mentioned , the power of the language of bunches lies in the description of geometric phenomena around divisors , because @xmath6 turns out to be the rational divisor class group of @xmath37 .
for example , we obtain very simple descriptions for the classes of rational cartier divisors , the cone @xmath69 of semiample classes and the cone @xmath70 of ample classes , see theorem [ qcartier ] : for the toric variety @xmath37 arising from a bunch @xmath16 , there are canonical isomorphisms : @xmath71 note that the last isomorphism gives a quasiprojectivity criterion in the spirit of @xcite and @xcite , see corollary [ ewaldcrit ] .
moreover , we can derive from the above theorem a simple fano criterion , see corollary [ fanocrit ] .
finally , we get back reid s toric cone theorem , see @xcite , even with a new description of the mori cone , see corollary [ moricone ] : suppose that @xmath37 is complete and simplicial .
then the cone of numerically effective curve classes in @xmath72 is given by @xmath73 in particular , this cone is convex and polyhedral .
moreover , @xmath37 is projective if and only if @xmath74 is strictly convex .
bunches can also be used for classification problems .
for example , once the machinery is established , kleinschmidt s classification @xcite becomes very simple and can even be slightly improved , see proposition [ kleinschmidt ] ; below we represent a sequence of weight vectors as a set of vectors @xmath75 , each of which carries a multiplicity @xmath76 : the smooth 2-complete toric varieties @xmath0 with @xmath77 correspond to bunches @xmath78 given by * weight vectors @xmath79 , and @xmath80 with @xmath81 , * multiplicities @xmath82 with @xmath83 , @xmath84 and @xmath85 . moreover , the toric variety @xmath0 defined by such a bunch @xmath16 is always projective , and it is fano if and only if we have @xmath86 in general the functor from bunches to toric varieties is neither injective nor surjective on morphisms , see examples [ freegroup ] and [ notfaithfull ] .
but if we restrict to @xmath64-factorial toric varieties , then the language of bunches provides also a tool for the study of toric morphisms , see theorem [ equivcat ] : there is an equivalence from the category of simple bunches to the category of full @xmath64-factorial toric varieties .
we would like to thank the referee for his careful reading and for various helpful comments and corrections .
in this section , we recall some basic facts on the correspondence between fans and toric varieties , and thereby fix our notation used later . for details
, we refer to the books of oda @xcite and fulton @xcite .
we begin with introducing the necessary terminology from convex geometry . by a lattice
we mean a free finitely generated @xmath44-module .
the associated rational vector space of a lattice @xmath2 is @xmath87 .
if @xmath31 is a lattice homomorphism , then we denote the induced linear map @xmath88 of rational vector spaces again by @xmath89 . by a cone in a lattice @xmath2
we always mean a polyhedral ( not necessarily strictly ) convex cone in the associated rational vector space @xmath7 .
let @xmath2 be a lattice , and let @xmath90 denote the dual lattice of @xmath2 .
the orthogonal space and the dual cone of a cone @xmath17 in @xmath2 are @xmath91 the relative interior of a cone @xmath17 is denoted by @xmath92 .
if @xmath93 is a face of @xmath17 , then we write @xmath94 .
the dimension of @xmath17 is the dimension of the linear space @xmath95 generated by @xmath17 .
the set of the @xmath96-dimensional faces of @xmath17 is denoted by @xmath97 , and the one - dimensional faces of a strictly convex cone are called rays . the primitive generators of a strictly convex cone @xmath17 in a lattice @xmath2 are the primitive lattice vectors of its rays . a strictly convex cone in @xmath2
is called simplicial if its primitive generators are linearly independent , and it is called regular if its primitive generators can be complemented to a lattice basis of @xmath2 . [ fandefi ] 1 .
a _ fan _ in a lattice @xmath2 is a finite collection @xmath1 of strictly convex cones in @xmath2 such that for each @xmath98 also all @xmath94 belong to @xmath1 and for any two @xmath99 we have @xmath100 .
2 . a _ map of fans _
@xmath101 in lattices @xmath102 is a lattice homomorphism @xmath103 such that for every @xmath104 there is a @xmath105 containing @xmath106 .
recall that the compatibility condition @xmath100 in the above definition is equivalent to the existence of a _ separating linear form _ for the cones @xmath107 and @xmath108 , i.e. , a linear form @xmath109 on @xmath2 such that @xmath110 if we replace in definition [ fandefi ] `` strictly convex '' with `` convex '' , we obtain the category of _
quasifans_. for a fan @xmath1 in @xmath2 , we denote by @xmath111 its support , that is the union of all its cones .
moreover , @xmath112 is the set of maximal cones of @xmath1 , and @xmath113 is the set of all @xmath96-dimensional cones of @xmath1 . in the sequel
, we shall often make use of a well known universal lifting construction , which makes the set of primitive generators of the rays of a given fan into a lattice basis , compare for example @xcite : [ coxconstr ] let @xmath1 be a fan in a lattice @xmath2 , and let @xmath114 .
let @xmath115 be the map sending the canonical base vector @xmath116 to the primitive generator @xmath117 . for @xmath118 , set @xmath119 then the cones @xmath120 , where @xmath118 , are the maximal cones of a fan @xmath121 consisting of faces of the positive orthant in @xmath122 .
moreover , @xmath115 is a map of the fans @xmath121 and @xmath1 .
now we turn to toric varieties . throughout the entire paper ,
we work over an algebraically closed field @xmath123 of characteristic zero , and the word point refers to a closed point . 1 . a _
toric variety _ is a normal variety @xmath0 containing an algebraic torus @xmath124 as an open subset such that the group structure of @xmath124 extends to a regular action on @xmath0 .
2 . a _ toric morphism _ is a regular map @xmath125 of toric varieties that restricts to a group homomorphism @xmath126 .
the correspondence between fans and toric varieties is obtained as follows : let @xmath1 be a fan in a lattice @xmath2 , and let @xmath127 be the dual lattice of @xmath2 . for every cone @xmath128
one defines an affine toric variety : @xmath129).\ ] ] for any two such @xmath130 , one has canonical open embeddings of @xmath131 into @xmath130 .
patching together all @xmath132 along these open embeddings gives a toric variety @xmath133 .
the assignment @xmath134 is functorial ; it is even a ( covariant ) equivalence of categories . in the sequel
, we shall frequently restrict our investigations to toric varieties that behave reasonably .
for that purpose , we consider the following geometric properties : [ nondegfulldef ] 1 . a toric variety @xmath0 is called _ nondegenerate _ if it admits no toric decomposition @xmath135 .
we call a toric variety @xmath0 _ 2-complete _ if it does not admit a toric open embeding @xmath136 with @xmath137 nonempty of codimension at least two .
we call a toric variety @xmath0 _ full _ if it is 2-complete and every @xmath124-orbit has a fixed point in its closure .
the notion of 2-completeness already occurs in @xcite .
it generalizes completeness in the sense that a toric variety is complete if and only if it is `` 1-complete '' .
for example , the affine space @xmath138 is 2-complete , whereas for a toric variety @xmath0 of dimension at least two and a fixed point @xmath139 , the variety @xmath140 is not 2-complete . in terms of fans ,
the properties introduced in definition [ nondegfulldef ] are characterized as follows : let @xmath0 be the toric variety arising from a fan @xmath1 in a lattice @xmath2 . 1 .
@xmath0 is nondegenerate if and only if the support @xmath141 generates @xmath7 as a vector space .
@xmath0 is 2-complete if and only if the fan @xmath1 can not be enlarged without adding new rays .
@xmath0 is full if and only if @xmath1 is as in ( ii ) and every maximal cone of @xmath1 is of full dimension .
[ section2 ] in this section , we introduce the language of bunches .
intuitively , one should think of a bunch as a collection of pairwise overlapping lattice cones , which satisfies certain irredundancy and maximality properties .
the precise definition of the category of bunches is performed in three steps .
the first one is to introduce the category of projected cones : [ weightedlatticedef ] 1 .
a _ projected cone _ is a pair @xmath142 , where @xmath25 is an epimorphism of lattices and @xmath27 is a simplicial cone of full dimension .
2 . a _ morphism of projected cones _
@xmath143 is a homomorphism @xmath144 such that @xmath145 holds and there is a commutative diagram @xmath146_{q_{1 } } \ar[r]^{\phi } & e_{2 } \ar[d]^{q_{2 } } \\ k_{1 } \ar[r]_{\b{\phi } } & k_{2 } } \ ] ] in the second step , we give the definition of bunches .
such a bunch will live in a projected cone @xmath142 . by a _ projected face _ in @xmath13 we mean the image @xmath67 of a face @xmath147 . [ bunchdef ]
a _ bunch _ in @xmath142 is a nonempty collection @xmath16 of projected faces in @xmath13 with the following property : a projected face @xmath148 belongs to @xmath16 if and only if @xmath149 let us reformulate this definition in a less formal way .
we say that two cones @xmath150 and @xmath151 _ overlap _ , if @xmath152 holds .
now , a nonempty collection @xmath16 of projected faces is a bunch if and only if it has the following properties : * any two members of @xmath16 overlap , * there is no pair @xmath153 with @xmath154 , * if a projected face @xmath155 overlaps each @xmath19 , then @xmath156 for a @xmath157 . [ weightedprojspace ]
let @xmath24 , and let @xmath158 .
moreover , fix a sequence @xmath159 of positive integers having greatest common divisor one .
this gives an epimorphism @xmath160 where @xmath26 denotes the @xmath21-th canonical base vector .
setting @xmath161 , we obtain a projected cone @xmath142 , and @xmath162 is a bunch .
finally , as the third step , we have to fix the notion of a morphism of bunches . for this , we first have to `` unpack '' the combinatorial information contained in a bunch .
this is done by constructing a further collection of cones : let @xmath16 be a bunch in a projected cone @xmath142 .
the _ covering collection _ of @xmath16 is @xmath28 as example [ weightedprojspace ] shows , @xmath29 will in general comprise much more cones than @xmath16 itself
. we can reconstruct the bunch from its covering collection : @xmath163 in general , for an element @xmath164 , the image @xmath67 need not be an element of @xmath16 . for later purposes
, the following observation will be crucial : [ interior ] let @xmath16 be a bunch in @xmath142 .
for any two @xmath165 , we have @xmath166 .
let @xmath167 . by the definition of @xmath29 , there exist cones @xmath168 with @xmath169 .
now assume that the relative interiors of the cones @xmath170 are disjoint .
then there is a proper face @xmath171 such that @xmath172 is contained in @xmath93 .
clearly , @xmath173 is contained in @xmath93 .
moreover , by the condition [ bunchcondition ] , the intersection @xmath174 is not empty
. in particular , @xmath175 meets @xmath93 .
since @xmath93 is a face of @xmath107 , we conclude @xmath176 . thus @xmath177 is a proper face of @xmath178 such that @xmath179 contains an element of @xmath16 .
this contradicts minimality of @xmath178 .
we come back to the definition of a morphism of bunches .
it is formulated in terms of the respective covering collections : [ bunchmorphdef ] let @xmath180 be bunches in projected cones @xmath143 .
a _ morphism _ from @xmath181 to @xmath182 is a morphism @xmath183 of the projected cones such that for every @xmath184 there is an @xmath185 with @xmath186 .
this concludes the definition of the category of bunches .
the notion of an isomorphism is characterized as follows : [ bunchiso ] a morphism @xmath187 of bunches @xmath181 and @xmath182 is an isomorphism if and only if @xmath187 is an isomorphism of the ambient projected cones and the induced map @xmath188 defines a bijection @xmath189 .
let the bunch @xmath180 live in the projected cone @xmath143 .
suppose first that @xmath183 is an isomorphism of the bunches .
then there is a morphism of bunches @xmath190 from @xmath182 to @xmath181 such that @xmath187 and @xmath191 are inverse to each other as lattice homomorphisms .
note that @xmath187 and @xmath191 are as well inverse to each other as morphisms of projected cones . in order to see that @xmath192 defines a bijection @xmath189
, it suffices to show that @xmath187 defines a bijection @xmath193 .
by bijectivity of @xmath187 and @xmath191 , we only have to show that for every @xmath185 the image @xmath194 belongs to @xmath195 .
this is done as follows : given @xmath185 , we apply definition [ bunchmorphdef ] to @xmath191 , and obtain an @xmath184 with @xmath196 .
applying @xmath187 gives @xmath197 .
again by definition [ bunchmorphdef ] , we find an @xmath198 with @xmath199 .
thus we have @xmath200 . by the definition of @xmath201
, we obtain @xmath202 .
consequently , @xmath203 belongs to @xmath195 .
now suppose that @xmath187 is an isomorphism of projected cones and that @xmath188 defines a bijection @xmath189 .
let @xmath190 denote the inverse of @xmath187 as a morphism of projected cones .
the only thing we have to show is that @xmath191 is a morphism from @xmath182 to @xmath181 .
this is done below : let @xmath185 .
then @xmath204 is a face of @xmath205 .
we check @xmath184 . by the definition of @xmath201 ,
the face @xmath206 is minimal with @xmath207 for some @xmath208 .
thus @xmath209 is minimal with @xmath210 for some @xmath208 .
since @xmath188 induces a bijection @xmath189 , we obtain @xmath184 .
the reminder of this section is devoted to the visualization of bunches .
the idea is that one should be able to recover many basic properties of a bunch @xmath16 without knowing the ambient projected cone @xmath142 .
this will work for the following important class of bunches : by a _ free _
we mean a bunch @xmath16 in a projected cone @xmath142 , where @xmath211 is a regular cone in @xmath212 .
the following construction shows that every free bunch arises from a certain collections of data in some lattice @xmath13 : [ bunchfromdownstairs ] let @xmath213 be a sequence in a lattice @xmath13 such that the @xmath214 generate @xmath13 .
we speak of the _ weight vectors _
@xmath15 , and call any cone generated by some of the @xmath15 a _
weight cone_. let @xmath16 be a collection of weight cones in @xmath13 satisfying condition [ bunchcondition ] , i.e. , a weight cone @xmath155 belongs to @xmath16 if and only if @xmath215 then there is an associated projected cone @xmath142 with the lattice @xmath24 , the cone @xmath161 spanned by the canonical base vectors , and the map @xmath25 sending @xmath26 to @xmath15 . by construction ,
the collection @xmath16 is a bunch in the projected cone @xmath142 .
this construction allows us to visualize bunches .
we regard a given sequence of weight vectors as a set @xmath216 , where each @xmath15 has a multiplicity @xmath217 counting the number of its repetitions .
then we may put these data as well as the cones of a given bunch in a picture .
for example , in the setting of [ weightedprojspace ] , the bunch @xmath16 defined by the sequence @xmath218 arises from the picture as one might expect , this picture describes the three - dimensional weighted projective space @xmath219 .
moreover , as we shall see in proposition [ kleinschmidt ] , the smooth complete toric varieties with picard group @xmath53 arise from sequences in @xmath53 and a collection @xmath220 according to the following picture : in order to compare two bunches arising from construction [ bunchfromdownstairs ] , there is no need to determine the covering collection .
namely , using proposition [ bunchiso ] , we obtain : [ downstairsisochar ] two sets of data @xmath221 and @xmath222 as in [ bunchfromdownstairs ] have isomorphic associated free bunches if and only if there is a lattice isomorphism @xmath223 such that 1 .
@xmath224 and @xmath225 differ only by enumeration , 2 . the collections @xmath226 and @xmath227 coincide .
in this section , we provide basic duality statements for translating from the language of bunches into the language of fans .
first we need a concept of a dual of a given projected cone @xmath142 . for this , note that @xmath142 determines two exact sequences of lattice homomorphisms @xmath30 & { m } \ar[r ] & { e } \ar[r]^{q } & k \ar[r ] & 0 , } \ ] ] @xmath30 & { l } \ar[r ] & { f } \ar[r]^{p } & { n } \ar[r ] & 0 , } \ ] ] where @xmath228 is the kernel of @xmath25 , and the second sequence arises from the first one by applying @xmath229 ; note that @xmath89 is not the dual homomorphism of @xmath230 .
let @xmath231 denote the dual cone .
then @xmath232 is again strictly convex , simplicial and of full dimension .
we call @xmath233 the _ dual projected cone _ of @xmath142 . in the sequel ,
fix a projected cone @xmath142 , and denote the associated dual projected cone by @xmath233 . recall that we have the face correspondence ,
see for example ( * ? ? ?
* appendix a ) : the sets of faces of the cones @xmath27 and @xmath234 are in order reversing correspondence via @xmath235 our task is to understand the relations between the projected faces @xmath67 , where @xmath147 , and the images @xmath236 of the corresponding faces .
the following observation is central : [ invariantseparation ] let @xmath237 , let @xmath238 , and let @xmath239
. then the following statements are equivalent : 1 .
there is an @xmath240-invariant separating linear form for @xmath241 and @xmath242 .
2 . for the relative interiors @xmath243 we have @xmath244 . here , by an _ @xmath240-invariant _ linear form we mean an element @xmath245 with @xmath246 .
moreover , recall from section [ section1 ] that a separating linear form for the cones @xmath241 and @xmath242 is an element @xmath247 with @xmath248 as before , let @xmath249 .
then the @xmath240-invariant linear forms on @xmath250 are precisely the elements of @xmath228 .
thus , since @xmath251 is a face of both @xmath252 , condition ( i ) is equivalent to @xmath253 note that @xmath254 equals @xmath255 .
moreover , @xmath256 equals @xmath257 , because the cone @xmath232 is simplicial .
hence we obtain @xmath258 for the second equality we used simpliciality of @xmath211 . by analogous arguments ,
the expression @xmath259 simplifies to @xmath260 .
thus [ invsepchar ] holds if and only if @xmath261 we claim that the left hand side simplifies to @xmath262 .
indeed , any @xmath263 belonging to the left hand side of [ invsepchar2 ] has a unique representation in terms of the primitive generators of @xmath264 of @xmath211 : @xmath265 now , dividing the middle term into two sums , one with only positive coefficients and the other with only negative ones , gives @xmath266 .
the reverse inclusion is obvious .
consequently [ invsepchar2 ] is equivalent to @xmath267 .
this in turn is obviously equivalent to condition ( ii ) .
let us mention here that simpliciality of the cones @xmath211 and @xmath232 is essential for the invariant separation lemma : consider the `` projected cone '' @xmath142 , where the lattices are @xmath268 and @xmath42 , the map @xmath230 is the projection onto the third coordinate , and the cone @xmath211 is given in terms of canonical base vectors by @xmath269 denote the `` dual projected cone '' by @xmath233 .
then @xmath239 is the sublattice generated by the dual base vector @xmath270 , and the cone @xmath232 is given by @xmath271 the faces @xmath272 and @xmath273 do not satisfy condition [ invariantseparation ] ( ii ) .
nevertheless , the corresponding faces @xmath274 admit @xmath240-invariant separating linear forms . for example we can take the linear form @xmath275 .
next we compare injectivity of @xmath230 with surjectivity of @xmath89 along corresponding faces ( of course , the roles of @xmath230 , @xmath276 etc . and @xmath89 , @xmath65 , etc .
can be interchanged in the statement ) : [ surjectiveandinjective1 ] for a face @xmath147 and @xmath277 , the following statements are equivalent : 1 .
@xmath89 maps @xmath278 onto @xmath7 , 2 .
@xmath230 is injective on @xmath279 .
let @xmath249 and @xmath239 .
using the fact that @xmath278 and @xmath279 are the orthogonal spaces of each other , we obtain the assertion by dualizing : @xmath280 if we take the lattice structure into consideration , then the situation becomes slightly more involved .
the essential observation is : [ surjectiveandinjective2 ] for a face @xmath147 and @xmath277 , the following statements are equivalent : 1 .
@xmath89 maps @xmath281 onto @xmath2 , 2 .
@xmath230 maps @xmath282 isomorphically onto a primitive sublattice of @xmath13 .
set @xmath239 .
assume that ( i ) holds .
then the snake lemma provides an exact sequence @xmath283 & { l \cap \lin(\delta_{0 } ) } \ar[r ] & { l } \ar[r ] & { f/(\lin(\delta_{0 } ) \cap f ) } \ar[r ] & 0 . } \ ] ] the dual lattice of @xmath284 is canonically isomorphic to @xmath285 .
hence , applying @xmath229 gives an exact sequence @xmath286 & e \cap \lin(\gamma_{0 } ) \ar[r]^(.625){q } & k \ar[r ] &
\hom(l \cap \lin(\delta_{0}),\zz ) \ar[r ] & 0.}\ ] ] this implies condition ( ii )
. the reverse direction can be settled by similar arguments .
finally , we consider morphisms @xmath183 of projected cones @xmath143 .
these data define a commutative diagram of lattices with exact rows : @xmath30 & { m_{1 } } \ar[r ] \ar[d ] & { e_{1 } } \ar[r]^{q_{1 } } \ar[d]_{{\phi } } & { k_{1 } } \ar[r ] \ar[d]^{{\bar{\phi } } } & 0 \\ 0 \ar[r ] & { m_{2 } } \ar[r ] & { e_{2 } } \ar[r]_{q_{2 } } & { k_{2 } } \ar[r ] & 0 } \ ] ] applying @xmath229 to this diagram , we obtain the following commutative diagram , again with exact rows : @xmath30 & { l_{1 } } \ar[r ] & { f_{1 } } \ar[r]^{p_{1 } } & { n_{1 } } \ar[r ] & 0 \\ 0 \ar[r ] & { l_{2 } } \ar[r ] \ar[u ] & { f_{2 } } \ar[r]_{p_{2 } } \ar[u]^{{\psi } } & { n_{2 } } \ar[r ] \ar[u]_{{\bar{\psi } } } & 0 } \ ] ] the dual map @xmath287 is a morphism of the dual projected cones @xmath288 of @xmath143 . now , consider faces @xmath289 , and let @xmath290 denote the corresponding faces of the cones @xmath291 .
[ shiftingmorphisms ] we have @xmath186 if and only if @xmath292 holds .
we only have to verify one implication .
the other then is a simple consequence of @xmath293 .
so , suppose @xmath294 .
then we obtain @xmath295
in this section , we compare bunches with fans .
we shall show that the category of bunches is equivalent to the category of `` maximal projectable fans '' , see theorem [ dualitythm ] .
the latter category is defined as follows : [ admissiblefandefi ] 1 .
let @xmath233 be a projected cone , and let @xmath239 .
1 . a _ projectable fan _ in @xmath233 is a fan @xmath8 consisting of faces of @xmath232 such that any two maximal cones of @xmath8 can be separated by an @xmath240-invariant linear form . 2 .
we call a projectable fan @xmath8 in @xmath233 _ maximal _ if any @xmath296 , which can be separated by @xmath240-invariant linear forms from the maximal cones of @xmath8 , belongs to @xmath8 .
2 . a _ morphism _ of projectable fans @xmath297 ( maximal or not ) in projected cones
@xmath288 is a morphism @xmath298 of projected cones which is in addition a map of the fans @xmath297 .
note that a projectable fan is the collection of all faces of the cones belonging to a `` locally coherent costring '' in the sense of ( * ? ? ?
5.1 ) , but the converse does not hold in general .
we shall demonstrate later by means of an example the importance of the maximality condition ( b ) , see [ underlinemaximal ] .
we define now a functor @xmath299 from bunches to maximal projectable fans .
let @xmath16 be a bunch in the projected cone @xmath142 .
consider the associated dual projected cone @xmath233 and the following subfan of the fan of faces of @xmath232 : @xmath300 @xmath8 is a maximal projectable fan in @xmath233 .
let @xmath301 . by the overlapping property [ interior ] of the covering collection @xmath29 and the invariant separation lemma [ invariantseparation ] , any two maximal cones of @xmath8
can be separated by @xmath240-invariant linear forms .
so we only have to verify the maximality condition [ admissiblefandefi ] ( i ) ( b ) for @xmath8 .
suppose that the face @xmath302 can be separated by @xmath240-invariant linear forms from the maximal cones of @xmath8 but does not belong to @xmath8 .
the projected face @xmath303 does not belong to @xmath16 , because otherwise any minimal face @xmath304 projecting onto @xmath155 would belong to @xmath29 , which contradicts the choice of @xmath17 .
the invariant separation lemma [ invariantseparation ] yields @xmath305 for every @xmath306 .
since @xmath155 is not an element of @xmath16 , it has to contain some element of @xmath16 .
but then some face of @xmath307 belongs to the covering collection @xmath29 .
again this contradicts the choice of the face @xmath308 .
the assignment @xmath309 extends canonically to morphisms .
namely , let @xmath180 be bunches in projected cones @xmath143 .
let @xmath297 denote the associated maximal projectable fans in the respective dual projected cones @xmath288 .
lemma [ shiftingmorphisms ] tells us that for every morphism @xmath183 of the bunches @xmath181 and @xmath182 , the dual map @xmath310 is a morphism of the maximal projectable fans @xmath311 and @xmath312 .
thus , we obtain : the assignments @xmath309 and @xmath313 define a contravariant functor @xmath299 from the category of bunches to the category of maximal projectable fans .
now we go the other way round .
consider a maximal projectable fan @xmath8 in a projected cone @xmath233 .
let @xmath142 denote the associated dual projected cone .
define @xmath16 to be the set of the minimal cones among all projected faces in @xmath13 arising from @xmath8 : @xmath314 @xmath16 is a bunch in @xmath142 .
we verify property [ bunchcondition ] for a given @xmath315 . according to the invariant separation lemma [ invariantseparation ]
, we have @xmath305 for any further @xmath19 .
moreover , since @xmath16 consists of minimal cones , @xmath316 is not contained in @xmath317 provided that @xmath19 is different from @xmath155 .
conversely , let the projected face @xmath155 satisfy [ bunchcondition ] .
choose @xmath147 with @xmath318 .
the invariant separation lemma [ invariantseparation ] tells us that @xmath277 belongs to @xmath8 .
let @xmath319 be a maximal cone with @xmath320 , and consider the image @xmath321 .
then we have @xmath156 , because @xmath322 holds . by the definition of the collection @xmath16
, there is a cone @xmath323 with @xmath324 .
in particular , we have @xmath325 .
applying once more the invariant separation lemma [ invariantseparation ] , gives even @xmath326 .
thus property [ bunchcondition ] yields @xmath327 .
this shows @xmath315 .
according to lemma [ shiftingmorphisms ] , associating to a map @xmath191 of maximal projectable fans its dual map @xmath187 makes this construction functorial .
thus we have : [ projectablefan2bunch ] the assignments @xmath328 and @xmath329 define a contravariant functor @xmath330 from the category of maximal projectable fans to the category of bunches . summing up
, we arrive at the main result of this section , namely the following duality statement : [ dualitythm ] the functors @xmath299 and @xmath330 are inverse to each other .
in particular , the categories of bunches and maximal projectable fans are dual to each other .
let us emphasize here the role of the maximality condition [ admissiblefandefi ] ( i ) ( b ) in this result .
the following example shows that there is no hope for a similar statement on ( non maximal ) projectable fans : [ underlinemaximal ] consider the projected cone @xmath233 , where the lattices are @xmath331 and @xmath332 , the cone @xmath232 is the positive orthant in @xmath333 , and the projection map is given by @xmath334 then the fan @xmath8 in @xmath250 having @xmath335 and @xmath336 as its maximal cones is projectable . but
@xmath8 is not maximal , because we may enlarge it to a ( maximal ) projectable fan @xmath337 by adding the cone @xmath338 .
let @xmath142 be the dual projected cone of @xmath233 .
then we have @xmath339 and @xmath340 .
moreover the projection @xmath230 sends each dual base vector @xmath341 to @xmath342 .
in particular , we obtain @xmath343 thus , @xmath8 and @xmath337 determine the same collection @xmath344 of projected faces in @xmath13 .
in other words , there is no way to reconstruct @xmath8 via face duality from a collection of projected faces in @xmath13 . in the rest of this section ,
we associate to any maximal projectable fan its `` quotient fan '' .
so , let @xmath8 be a maximal projectable fan in a weighted lattice @xmath233
. then the images @xmath345 , where @xmath17 runs through the maximal cones of @xmath8 are the maximal cones of a quasifan @xmath337 in @xmath2 .
we reduce @xmath337 to a fan as follows : let @xmath346 be the primitive sublattice generating the minimal cone of @xmath337 , let @xmath347 , and let @xmath348 denote the projection .
the _ quotient fan _ of @xmath8 is @xmath349 .
note that @xmath350 is a map of the fans @xmath8 and @xmath1 .
in fact , this is a special case of a more general construction , see ( * ? ? ?
* theorem 2.3 ) . in our setting , it is easy to see that everything is compatible with morphisms . thus we obtain : [ quotfan ] the assignment @xmath351 defines a covariant functor @xmath352 from the category of maximal projectable fans to the category of fans . the following simple example shows that dividing by @xmath353 in the construction of quotient fan is indeed necessary : consider the lattices @xmath354 and @xmath42 , the map @xmath25 , @xmath355 , and the positive orthant @xmath356 .
let @xmath16 be the bunch consisting just of the trivial cone @xmath357 .
then the quasifan @xmath337 determined by @xmath16 consists of the single cone @xmath358 . we note an observation on the composition @xmath359 .
consider a bunch @xmath16 in @xmath142 and its associated maximal projectable fan @xmath8 in @xmath233 .
let @xmath1 be the quotient fan of @xmath8 , and , as before , let @xmath360 be the projection .
[ conesofquotfan ] there is a canonical order reversing bijection @xmath361 the inverse map is given by @xmath362 .
here we present the first application of the language of bunches .
we consider the action of a subtorus on a @xmath64-factorial nondegenerate affine toric variety and give a combinatorial description of the maximal open subsets admitting a good quotient by this action .
this complements results of @xcite for torus actions on @xmath363 .
let us first recall the basic concepts concerning good quotients . let the reductive group @xmath364 act on a variety @xmath0 by means of a morphism @xmath365 .
a _ good quotient _ for this action is a @xmath364-invariant affine morphism @xmath366 such that the canonical map @xmath367 is an isomorphism .
if it exists , then the good quotient space is usually denoted by @xmath368 . in general ,
a @xmath364-variety @xmath0 need not admit a good quotient @xmath369 , but there frequently exist many invariant open subsets @xmath370 with good quotient @xmath371 .
it is one of the central tasks of geometric invariant theory to describe all these open subsets , see ( * ? ? ?
* section 7.2 ) . in the course of this problem ,
one reasonably looks for maximal @xmath370 in the following sense , see ( * ? ? ?
* section 7.2 ) and @xcite : an open subset @xmath370 is called _
@xmath364-maximal _ , if there is a good quotient @xmath372 and there is no open @xmath373 admitting a good quotient @xmath374 such that @xmath375 is a proper @xmath376-saturated subset of @xmath377 . in the setting of subtorus actions ,
the maximal open subsets with good quotient can be characterized in terms of fans .
this relies on the following observation due to wicicka , see ( * ? ? ?
* proposition 2.5 ) : [ swiecicka ] let @xmath0 be a toric variety , and let @xmath378 be a subtorus of the big torus . if @xmath370 is a @xmath379-maximal subset , then @xmath375 is invariant under @xmath124 .
now , let @xmath0 be the toric variety arising from a fan @xmath1 in a lattice @xmath2 , let @xmath378 be the subtorus corresponding to a primitive sublattice @xmath380 . by the above proposition , the @xmath379-maximal subsets @xmath370 correspond to certain subfans of @xmath1 .
the characterization of these fans is standard , see e.g. ( * ? ? ?
* proposition 1.3 ) : [ goodquotchar ] let @xmath370 be the open @xmath124-invariant subset defined by a subfan @xmath8 of @xmath1 . 1 .
there is a good quotient @xmath381 if and only if any two maximal cones of @xmath8 can be separated by an @xmath240-invariant linear form . 2 .
@xmath375 is @xmath379-maximal if and only if ( i ) holds and every @xmath382 that can be separated by @xmath240-invariant linear forms from the maximal cones of @xmath8 belongs to @xmath8 .
though this is a complete combinatorial description of all @xmath379-maximal subsets , it has two drawbacks in practice : on the one hand , the ambient space of the combinatorial data might be of quite big dimension , and , on the other hand , for the explicit checking of the conditions there may be large numbers of cones to go through .
the language of bunches makes the situation more clear .
let @xmath0 be an affine nondegenerate @xmath64-factorial toric variety arising from a cone @xmath232 in a lattice @xmath250 , and let @xmath378 be the subtorus corresponding to a sublattice @xmath383 .
setting @xmath384 , we obtain a projected cone @xmath233 .
moreover , we have the dual projected cone @xmath142 , where @xmath13 is canonically isomorphic to the lattice of characters of the small torus @xmath378 . in order to describe the @xmath379-maximal subsets of @xmath0
, we use the functor @xmath299 associating to a bunch @xmath16 in @xmath142 a maximal projectable fan @xmath385 in @xmath233 . the resulting statement generalizes and complements the results of @xcite : [ bunches2tmax ] the assignment @xmath386 defines a one - to - one correspondence between the bunches in @xmath142 and the @xmath379-maximal open subsets of @xmath0 . by the definition of a maximal projectable fan and proposition [ goodquotchar ]
, the toric open subvariety @xmath387 is indeed @xmath379-maximal .
hence the assignment is well defined .
moreover , it is of course injective .
surjectivity follows from proposition [ swiecicka ] . in the setting of theorem [ bunches2tmax ] ,
the good quotient of the @xmath379-action on @xmath387 is the toric morphism @xmath388 arising from the projection @xmath389 onto the quotient fan .
some of the good quotients are of special interest : a _ geometric quotient _ for an action of a reductive group @xmath364 on a variety @xmath0 is a good quotient that separates orbits .
geometric quotients are denoted by @xmath390 . again , for subtorus actions on toric varieties , there is a description in terms of fans .
let @xmath0 be the toric variety arising from a fan @xmath1 in a lattice @xmath2 , let @xmath378 be the subtorus corresponding to a primitive sublattice @xmath380 .
existence of a geometric quotient is characterized as follows , see ( * ? ? ?
* theorem 5.1 ) : the action of @xmath379 on @xmath0 admits a geometric quotient if and only if the projection @xmath391 is injective on the support @xmath8 . let us translate this into the language of bunches
. as before , consider a projected cone @xmath233 and its associated dual projected cone @xmath142 .
a bunch @xmath16 in @xmath142 is called _ geometric _ if @xmath392 holds for every @xmath19 .
we consider the affine toric variety @xmath393 and the subtorus @xmath378 corresponding to the sublattice @xmath394 .
the above notion yields what we are looking for : [ geomquotchar ] let @xmath16 be a bunch in @xmath142 .
the open toric subvariety @xmath395 admits a geometric quotient by the action of @xmath379 if and only if @xmath16 is geometric .
@xmath387 admits a geometric quotient by @xmath379 if and only if @xmath31 is injective on @xmath396 . in our situation ,
the latter is equivalent to saying that @xmath31 is injective on the maximal cones of @xmath8 , that means on the cones @xmath397 with @xmath65 an element of @xmath29 .
thus lemma [ surjectiveandinjective1 ] tells us that @xmath387 admits a geometric quotient if and only if every cone @xmath67 , @xmath398 , is of full dimension in @xmath13 .
since the elements of @xmath16 occur among these cones and for any two cones of @xmath29 their relative interiors intersect , we obtain the desired characterization .
we introduce the class of standard bunches .
the main result of this section , theorem [ standard2maximal ] , says that every nondegenerate 2-complete toric variety can be described by such a standard bunch , and , moreover , the isomorphism classes of free standard bunches correspond to the isomorphism classes of nondegenerate 2-complete toric varieties having free class group .
[ standarddefi ] let @xmath16 be a bunch in a projected cone @xmath142 , and let @xmath399 be the facets of @xmath211 .
we say that @xmath16 is a _
standard bunch _
if 1 . for all @xmath400 we have @xmath401 , 2 . for every @xmath400
there is a @xmath19 with @xmath402 .
if @xmath16 is a standard bunch in @xmath142 , and the cone @xmath27 is regular , then we speak of the _ free standard bunch _ @xmath16 .
the constructions of section [ bunchesandadmissiblefans ] provide a functor from standard bunches to toric varieties : let @xmath16 be a standard bunch in @xmath142 , and let @xmath403 be the quotient fan of the maximal projectable fan corresponding to @xmath16 .
toric variety associated to _
@xmath16 is @xmath404 .
recall from section [ section1 ] that a toric variety @xmath0 is nondegenerate , if it does not admit a toric decomposition @xmath135 .
moreover , @xmath0 is 2-complete , if any toric open embedding @xmath136 with @xmath405 is an isomorphism .
[ standard2maximal ] the assignment @xmath406 defines a contravariant functor from the category of standard bunches to the category of nondegenerate 2-complete toric varieties .
moreover , 1 .
every nondegenerate 2-complete toric variety is isomorphic to a toric variety @xmath37 with a standard bunch @xmath16 .
@xmath407 induces a bijection from the isomorphism classes of free standard bunches to the isomorphism classes of nondegenerate 2-complete toric varieties with free class group . for the proof of theorem [ standard2maximal ]
, we have to do some preparation .
we need a torsion free version of cox s construction [ coxconstr ] for nondegenerate fans , i.e. fans @xmath1 in a lattice @xmath2 such that the support @xmath111 generates the vector space @xmath7 : [ reducedcox ] let @xmath233 be a projected cone , @xmath8 a fan in @xmath250 , and @xmath1 a fan in @xmath2 .
we say that these data form a _ reduced cox construction _ for @xmath1 if 1 .
@xmath408 equals @xmath409 , and @xmath89 induces bijections @xmath410 and @xmath411 , 2 .
@xmath89 maps the primitive generators of @xmath232 to primitive lattice vectors .
we show now that every nondegenerate fan admits reduced cox contructions .
in fact , these reduced cox constructions will be even compatible with a certain type of maps of fans .
let @xmath101 be nondegenerate fans in lattices @xmath102 .
moreover , let @xmath412 be any isomorphism of lattices that is a map of the fans @xmath413 and @xmath414 .
suppose that @xmath415 induces a bijection on the sets of rays @xmath416 and @xmath417 .
[ lift2cox ] 1 .
there exist a projected cone @xmath418 and a fan @xmath312 in @xmath419 defining a reduced cox construction for @xmath413 .
2 . for every reduced cox construction as in ( i )
there exist a reduced cox construction for @xmath420 given by @xmath421 and @xmath311 , a lattice isomorphism @xmath298 , and a commutative diagram of maps of fans @xmath422^{\psi } \ar[d]_{p_{1 } } & { f_{2 } } \ar[d]^{p_{2 } } \\ n_{1 } \ar[r]_{\b{\psi } } & n_{2 } } \ ] ] where the map @xmath191 induces a bijection of the sets of rays @xmath423 and @xmath424 .
moreover , if @xmath415 maps @xmath413 isomorphically onto a subfan of @xmath414 , then @xmath191 maps @xmath312 isomorphically onto a subfan of @xmath311 .
first we perform cox s original construction [ coxconstr ] for the fans @xmath101 .
denote by @xmath425 the set of rays of @xmath101 .
for every maximal cone @xmath426 let @xmath427 then these cones @xmath120 are the maximal cones of a fan @xmath428 in @xmath429 .
moreover , we have canonical projections sending the canonical base vectors to the primitive lattice vectors of the corresponding rays : @xmath430 let us verify ( i ) .
since @xmath431 needs not be surjective , it can not serve as a projection of a projected cone .
we have to perform a reduction step : let @xmath432 , choose a section @xmath433 of @xmath431 , and set @xmath434 then we can view @xmath241 and @xmath312 as well as objects in the lattice @xmath419 . note that @xmath241 needs no longer be regular , but remains simplicial .
together with the surjection @xmath435 , the cone @xmath436 and the fan @xmath437 give the desired data .
we turn to ( ii ) .
define a ( invertible ) linear map @xmath438 of rational vector spaces by prescribing its values on the canonical base vectors as follows : @xmath439 similar to the proof of ( i ) , we may reduce the cox construction of @xmath420 by refining the lattice @xmath440 via @xmath441 .
then the resulting @xmath421 and @xmath311 and the lattice isomorphism @xmath442 are as desired .
the above proof shows in particular that we can not expect uniqueness of reduced cox constructions for a given fan .
[ freecl2regular ] let @xmath1 be a nondegenerate fan in a lattice @xmath2 .
then the associated toric variety @xmath0 has free class group @xmath443 if and only if @xmath1 admits a reduced cox construction @xmath8 in a projected cone @xmath233 with a regular cone @xmath232 .
again , we consider cox s construction [ coxconstr ] for @xmath1 .
let @xmath444 denote the set of rays of @xmath1 , and let @xmath445 be the map sending the canonical base vector @xmath116 to the primitive lattice vector @xmath446 .
if the toric variety @xmath0 has free class group then the lattice homomorphism @xmath447 is surjective , see @xcite .
hence it defines the desired reduced cox construction with @xmath448 and @xmath232 the positive orthant in @xmath122 .
conversely , let @xmath233 and @xmath8 be a reduced cox construction of @xmath1 with @xmath232 regular .
then @xmath447 factors as @xmath449 , where @xmath450 maps the canonical base vectors to the primitive generators of the cone @xmath232 .
it follows that @xmath447 is surjective .
hence @xmath443 is free .
[ liftisos ] let @xmath288 and @xmath297 be reduced cox constructions for @xmath101 such that the @xmath291 are regular cones .
then every isomorphism @xmath412 of the fans @xmath101 admits a unique lifting @xmath451 to an isomorphism of the fans @xmath297 .
for @xmath452 , let @xmath453 denote the primitive generator of @xmath291 above the primitive vector of @xmath454 . then define @xmath455 by setting @xmath456 .
let @xmath16 be a bunch in @xmath142 , and let @xmath8 denote the associated maximal projectable fan in the dual projected cone @xmath233 .
let @xmath1 denote the quasifan in @xmath2 obtained by projecting the maximal cones of @xmath8 .
[ standard2cox ] the following statements are equivalent : 1 .
@xmath1 is a fan having @xmath8 and @xmath233 as a reduced cox construction .
2 . @xmath16 is a standard bunch .
recall that @xmath8 being projectable means that any two maximal cones of @xmath8 can be separated by a @xmath457-invariant linear form .
in particular , @xmath89 sets up a bijection @xmath411 .
thus , the first statement holds if and only if we have : * @xmath89 maps every primitive generator of @xmath232 to a primitive lattice vector in @xmath2 . *
@xmath1 is a fan , we have @xmath458 , and @xmath89 induces a bijection @xmath410 . hence the task is to show that the conditions [ standarddefi ] ( i ) and ( ii ) hold if and only if ( a ) and ( b ) do so .
equivalence of ( a ) and [ standarddefi ] ( i ) is a direct application of lemma [ surjectiveandinjective2 ] ; note that for this one has to interchange the roles of @xmath459 and @xmath460 in the lemma such that for the @xmath65 of the lemma one can take a ray of @xmath8 . now , suppose that the conditions ( a ) and ( b ) are valid .
by equivalence of ( a ) and [ standarddefi ] ( i ) , we only have to check that [ standarddefi ] ( ii ) holds .
fix a facet @xmath461 . by assumption
, there is a maximal cone @xmath462 such that @xmath463 is a ray of @xmath464 and @xmath465 is a ray of @xmath466 .
then @xmath467 belongs to @xmath29 , we have @xmath468 , and the invariant separation lemma implies @xmath469 . by the overlapping property [ interior ] ,
any @xmath19 with @xmath470 is as in [ standarddefi ] ( ii ) .
conversely , if [ standarddefi ] ( i ) and ( ii ) are valid , then we have to show that @xmath1 is a fan , @xmath409 equals @xmath408 and that @xmath89 induces a bijection @xmath410 .
consider any @xmath471 .
then @xmath472 for some facet @xmath461 . by [ standarddefi ]
( ii ) , we have @xmath473 for some @xmath474 .
thus we find a @xmath468 being minimal with the property that @xmath475 holds for some @xmath315 .
then @xmath68 , and @xmath476 is nonempty because it contains @xmath477 .
hence @xmath478 is a maximal cone of @xmath8 with @xmath479 , and the invariant separation lemma yields @xmath480 .
so , this consideration gives in particular @xmath458 .
moreover , since @xmath465 is strictly convex , it gives @xmath481 ; in other words , the quasifan @xmath1 is a fan .
furthermore , since we already know that ( a ) holds , the image @xmath465 is in fact one - dimensional .
hence we obtain @xmath482 , and thus @xmath410 is well defined .
surjectivity of the map @xmath410 follows from the fact that @xmath411 is surjective .
injectivity follows from the observation that by [ standarddefi ] ( ii ) we always have @xmath483 , and hence any two rays of @xmath232 can be invariantly separated . by lemma [ standard2cox ] ,
the toric variety @xmath484 associated to a standard bunch @xmath16 is nondegenerate .
it is also 2-complete : otherwise there is an open toric embedding @xmath38 with nonempty complement of codimension at least two . using lemmas [ lift2cox ] and [ standard2cox ] , we can compare reduced cox constructions of @xmath37 and @xmath0 , and we see that the projectable fan associated to @xmath16 is not maximal , i.e. , does not satisfy [ admissiblefandefi ] ( i ) ( b ) . a contradiction .
so the functor @xmath406 is well defined .
the fact that it is surjective on isomorphism classes follows from existence of reduced cox constructions , proposition [ goodquotchar ] , theorem [ dualitythm ] and lemma [ standard2cox ] .
the correspondence of isomorphism classes of free bunches with isomorphism classes of nondegenerate 2-complete toric varieties with free class group is a direct application of lemmas [ freecl2regular ] and [ liftisos ] .
fix a standard bunch @xmath16 in a projected cone @xmath142 , and let @xmath485 denote the associated toric variety . in this section ,
we characterize basic geometric properties of @xmath0 in terms of the bunch @xmath16 .
let @xmath233 be the dual projected cone .
denote by @xmath8 the maximal projectable fan associated to @xmath16 , and let @xmath1 be the quotient fan of @xmath8 .
recall from lemma [ standard2cox ] that these data form a reduced cox construction of @xmath1 .
in particular , @xmath1 lives in @xmath2 , and we have @xmath486 we study now @xmath64-factoriality , smoothness , existence of fixed points and completeness . for this
we need the following observation : [ fullandsimplicial ] consider a face @xmath68 and the corresponding maximal cone @xmath236 of @xmath1 .
then we have : 1 .
@xmath67 is of full dimension if and only if @xmath236 is simplicial .
@xmath67 is simplicial if and only if @xmath236 is of full dimension .
we prove ( i ) .
let @xmath67 be of full dimension .
by lemma [ surjectiveandinjective1 ] , the map @xmath89 is injective along @xmath397 . in particular
, @xmath236 is simplicial .
conversely , let @xmath236 be simplicial . since @xmath89 induces a bijection from the rays of @xmath397 to the rays of @xmath236 , it is injective along @xmath397
. thus lemma [ surjectiveandinjective1 ] yields that @xmath67 is of full dimension .
we turn to ( ii ) .
if @xmath236 is of full dimension , we see as before that @xmath67 is simplicial . for the converse
we show that @xmath230 is injective along @xmath65 : for every ray @xmath454 of @xmath67 , choose a ray @xmath487 of @xmath65 with @xmath488 .
then the cone @xmath489 generated by these rays @xmath487 is mapped bijectively onto @xmath67 . by minimality of @xmath65 as an element of @xmath29 , we conclude @xmath490 . the first statement of this section is the following characterization of @xmath64-factoriality : [ qfactorialchar ] the toric variety @xmath0 is @xmath64-factorial if and only if @xmath16 consists of cones of full dimension in @xmath13 .
the toric variety @xmath0 is @xmath64-factorial if and only if all cones of @xmath1 are simplicial . by lemma [ fullandsimplicial
] this is equivalent to saying that all cones @xmath67 , @xmath68 , are of full dimension .
the latter holds if and only if @xmath16 consists of cones of full dimension , because every @xmath67 contains some @xmath19 . characterizing smoothness of toric varieties in terms of bunches involves the lattice structure : [ smoothchar ] the toric variety @xmath37 is smooth if and only if for every @xmath68 we have 1 .
@xmath397 is a regular cone in @xmath250 .
@xmath230 maps @xmath282 onto @xmath13 .
suppose that ( i ) and ( ii ) hold . to verify smoothness of @xmath0 , we have to show that all cones @xmath236 , @xmath491 , are regular . by the properties [ reducedcox ] of a reduced cox construction and condition ( i ) , the primitive generators of @xmath236 span the sublattice @xmath492 of @xmath2 . by lemma [ surjectiveandinjective2 ] and condition ( ii ) , this sublattice is primitive .
this proves regularity of the cone @xmath236 .
conversely , let @xmath0 be smooth . for @xmath491 , consider the sublattice @xmath493 spanned by the primitive generators of @xmath397 . since @xmath236 is regular , the properties [ reducedcox ] of a reduced cox construction yield that @xmath494 is a primitive sublattice of @xmath2 . since @xmath89 is injective along @xmath397 , we see that also @xmath495 is primitive .
this gives condition ( i ) .
condition ( ii ) then follows from lemma [ surjectiveandinjective2 ] .
existence of global regular functions on a toric variety is characterized as follows : [ nontrivfunctionschar ] we have @xmath496 if and only if @xmath230 contracts no ray of @xmath211 to a point and the image @xmath497 is strictly convex
. @xmath496 holds if and only if the rays of @xmath1 generate @xmath7 as a convex cone .
this is valid if and only if @xmath498 equals @xmath333 . by dualizing ,
this condition is equivalent to @xmath499 .
this in turn holds if and only if @xmath230 contracts no ray of @xmath211 and @xmath497 is strictly convex .
recall that we speak of a full toric variety @xmath0 if @xmath0 is 2-complete and every @xmath124-orbit contains a fixed point in its closure .
[ existenoffixedpointschar ] the toric variety @xmath0 is full if and only if all cones @xmath67 , where @xmath68 , are simplicial .
existence of fixed points in the @xmath124-orbit closures means that all maximal cones of @xmath1 are of full dimension .
thus lemma [ fullandsimplicial ] gives the assertion .
finally , we also can characterize completeness of a toric variety in terms of its bunch : [ completenesschar ] the toric variety @xmath37 is complete if and only if @xmath16 contains a simplicial cone and any face @xmath147 satisfying @xmath500 for some @xmath19 and @xmath489 for only one @xmath501 belongs to @xmath29 .
this is a direct translation of a well known characterization of completeness in terms of fans .
namely , @xmath0 is complete if and only the fan @xmath1 has the following two properties : firstly , at least one of its ( maximal ) cones is of full dimension ; secondly , any cone contained in only one maximal cone is itself maximal . by lemma [ fullandsimplicial ]
, the first property translates to the property that @xmath16 contains a simplicial cone . for the second ,
recall from proposition [ conesofquotfan ] , that the cones @xmath502 correspond to the faces @xmath147 satisfying @xmath500 for some @xmath19 via @xmath503 thereby the maximal cones @xmath107 of @xmath1 correspond to the elements of @xmath29 .
thus the statement that @xmath504 for only one maximal @xmath107 implies @xmath505 directly translates to the second characterizing condition of the assertion .
one may ask if a full toric variety @xmath0 with @xmath496 is already complete .
if @xmath506 holds , then the answer is positive . for @xmath507
, there are counterexamples , compare ( * ? ? ?
* remark 2 ) .
however , we shall see in section [ section10 ] that every smooth 2-complete toric variety with class group @xmath53 is complete .
in this section we introduce the class of simple bunches , and we show that @xmath63 defines an equivalence of categories between the simple bunches and the full @xmath64-factorial toric varieties . by a _ simple bunch _ we mean a standard bunch @xmath16 in a projected cone @xmath142 such that @xmath230 maps @xmath508 isomorphically onto @xmath13 for every @xmath68 .
note that the cones of a simple bunch are of full dimension and simplicial , but they need not be regular . to state the main result of this section , recall from definition [ nondegfulldef ] that a toric variety @xmath0 is full , if it is 2-complete and every @xmath124-orbit has a fixed point in its closure .
[ equivcat ] the assignment @xmath63 defines an equivalence of the category of simple bunches with the category of full @xmath64-factorial toric varieties . the crucial observation for the proof is that full @xmath64-factorial toric varieties admit a universal reduced cox construction in a certain sense .
first we show existence : [ freeexists ] let @xmath1 be a simplicial fan in a lattice @xmath2 such that any maximal cone is of full dimension . then there are @xmath233 and @xmath8 defining a reduced cox construction of @xmath1 such that 1 .
@xmath89 induces for every @xmath509 an isomorphism @xmath510 .
@xmath250 is the sum of the sublattices @xmath511 , where @xmath509 . as usual
, let @xmath444 be the set of rays of @xmath1 , let @xmath512 be the homomorphism sending the canonical base vector @xmath116 to the primitive vector @xmath446 , and let @xmath513 be the cone generated by the @xmath116
. for any cone @xmath128 set @xmath514 then these cones form a fan @xmath8 in the lattice @xmath515 . in order to achieve the desired properties ,
we refine the lattice @xmath515 as follows : think of @xmath89 for the moment as a map of the vector spaces @xmath122 and @xmath7 and consider the set @xmath516 this set generates a lattice @xmath517 , because by simpliciality of @xmath1 the map @xmath89 is injective along the cones @xmath462 , and hence each @xmath518 is discrete . moreover , the restriction @xmath519 of @xmath89 is an isomorphism for every @xmath509 .
here we use that the maximal cones of @xmath1 are of full dimension .
we may view @xmath232 and @xmath8 as well as data in the lattice @xmath250 .
since @xmath31 is surjective , this gives in particular a projected cone @xmath233 .
now it is straightforward to verify the defining properties of a reduced cox construction for these data . by construction
it satisfies ( i ) and ( ii ) .
we shall call a reduced cox construction with the properties of lemma [ freeexists ] a _ universal reduced cox construction_. uniqueness of universal reduced cox constructions is a consequence of the following lifting property : [ liftinglemma ] let @xmath101 be fans in lattices @xmath102 , and let @xmath520 and @xmath297 be universal reduced cox constructions
. then every map @xmath412 of the fans @xmath101 admits a unique lifting to a map @xmath298 of the fans @xmath297 : @xmath521^{\psi } \ar[d]_{{p_{1 } } } & f_{2 } \ar[d]^{{p_{2 } } } \\
{ n_{1 } } \ar[r]_{\b{\psi } } & n_{2 } } \ ] ] for every maximal cone @xmath104 , fix a maximal cone @xmath105 with @xmath522 .
let @xmath523 denote the cones lying over @xmath170 .
then , by the property [ freeexists ] ( i ) of a universal reduced cox construction , we obtain for every @xmath107 a unique commutative diagram @xmath524^{\psi_{\rq{\sigma}_{1 } } } \ar[d]_{{p_{1 } } } & f_{2 } \cap \lin(\rq{\sigma}_{2 } ) \ar[d]^{{p_{2 } } } \\
{ n_{1 } } \ar[u]_{s_{\rq{\sigma}_{1 } } } \ar[r]_{\b{\psi } } & n_{2 } \ar[u]^{s_{\rq{\sigma}_{2 } } } } \ ] ] where the @xmath525 are the sections mapping the primitive generators of @xmath170 to those of @xmath526 .
any two @xmath527 and @xmath528 have the same values on the primitive generators of @xmath232 that lie in @xmath529 .
thus the @xmath527 fit together to a linear map @xmath530 .
lemma [ freeexists ] ( ii ) gives @xmath531 .
hence we found the desired lifting .
first observe that for a simple bunch @xmath16 , the toric variety @xmath37 is indeed full and @xmath64-factorial : since for every @xmath68 the cone @xmath67 is simplicial and of full dimension , lemma [ fullandsimplicial ] ensures that also the maximal cones of the fan @xmath1 defining @xmath0 are of full dimension and simplicial . next we show that , up to isomorphism , every full @xmath64-factorial toric variety @xmath0 is of the form @xmath37 with a simple bunch .
we may assume that @xmath0 arises from a fan @xmath1 in a lattice @xmath2 .
lemma [ freeexists ] provides a universal reduced cox construction @xmath233 and @xmath8 of @xmath1 . by lemma [ surjectiveandinjective2 ] ,
the bunch @xmath16 corresponding to the maximal projectable fan @xmath8 is as wanted .
finally , we have to show bijectivity on the level of morphisms .
due to lemma [ liftinglemma ] the morphisms between full @xmath64-factorial simplicial toric varieties are in one - to - one correspondence with the maps of maximal projectable fans defined by their universal reduced cox constructions . by theorem [ dualitythm ] , the latter morphisms correspond to the morphisms of the associated bunches .
@xmath63 defines an equivalence of categories from free simple bunches to full smooth toric varieties . by propositions [ smoothchar ] and [ existenoffixedpointschar ] ,
every free simple bunch @xmath16 defines a full smooth toric variety @xmath37 .
conversely , for a full smooth toric variety @xmath0 , the usual cox construction [ coxconstr ] is a universal reduced cox construction .
hence @xmath0 arises from a free simple bunch .
we conclude this section with some observations and examples .
the first remark gives the geometric interpretation of the universal reduced cox construction : in terms of toric varieties , the universal reduced cox construction @xmath532 arises from the usual cox construction @xmath533 by dividing @xmath534 by the ( finite ) group generated by all isotropy groups of the action of @xmath535 on @xmath534 .
the following example shows that for nonsimplicial toric varieties the functor @xmath63 associating to a standard bunch its toric variety is not surjective on the level of morphisms : [ freegroup ] we present a toric morphism that can not be lifted to the cox constructions . in @xmath50
consider the vectors @xmath536 let @xmath414 be the fan of faces of the cone generated by @xmath537 , and let @xmath413 be the subdivision of @xmath414 at @xmath538 .
mapping the @xmath21-th canonical base vector to @xmath539 , we obtain projections @xmath540 let @xmath312 and @xmath311 be the fans above @xmath413 and @xmath414 , respectively .
note that these data are in fact reduced cox constructions .
we claim that the identity @xmath541 does not admit a lifting .
in fact , since @xmath538 equals @xmath542 , a possible lifting @xmath543 must satisfy @xmath544 where @xmath545 generates @xmath546 .
an explicit calculation shows that the right hand sice does not contain integral points with nonnegative coefficients .
this excludes existence of a map @xmath547 of the fans @xmath312 and @xmath311 lifting @xmath548 .
the next example shows that in the nonsimplicial case the functor @xmath63 is not injective on the level of morphisms .
[ notfaithfull ] we give a toric morphism that admits two different liftings .
let @xmath414 be the fan of faces of the cone generated by the vectors @xmath549 let @xmath413 be the subdivision of @xmath414 at @xmath550 .
then there are two different liftings of the identity map @xmath551 to the respective cox - constructions , namely the maps @xmath552 defined by the matrices @xmath553
in this section , we come to one of the most powerful parts of the language of bunches : we study geometric properties of divisors and divisor classes .
let us fix the notation . as usual
, @xmath16 is a standard bunch in a projected cone @xmath142 , and @xmath485 is the associated toric variety with its big torus @xmath554 .
let @xmath233 be the dual projected cone , @xmath8 the maximal projectable fan corresponding to @xmath16 , and @xmath1 the quotient fan of @xmath8 . recall that these data define a reduced cox construction of @xmath1 , and we have @xmath555 .
our first task is to relate the lattice @xmath13 to the divisor class group @xmath556 .
let @xmath557 be the primitive generators of the one - dimensional cones of @xmath1 , and let @xmath558 be the primitive generators of the rays of @xmath232 , numbered in such a way that we always have @xmath559 .
every ray @xmath560 determines an invariant prime divisor @xmath561 in @xmath0 .
there is a canonical injection mapping @xmath212 into the lattice @xmath562 of invariant weil divisors on @xmath0 : @xmath563 by construction , an element @xmath263 is mapped to the principal divisor @xmath564 of @xmath0 .
hence , we obtain , compare @xcite : [ divisors ] there is a commutative diagram with exact rows and injective upwards arrows : @xmath565 & \pdiv^{t}(x ) \ar[r ] & { \wdiv^{t } ( x ) } \ar[r ] & \cldiv(x ) \ar[r ] & 0 \\ 0 \ar[r ] & m \ar[r ] \ar[u]^{\cong } & e \ar[r ] \ar[u]^{\mathfrak{d } } & k \ar[r ] \ar[u ] ^{\b{\mathfrak{d } } } & 0 } \ ] ] moreover , tensoring this diagram with @xmath64 provides isomorphisms of rational vector spaces : @xmath566 in the main result of this section , we study the group @xmath567 of rational cartier divisor classes .
we obtain very simple descriptions of the cone @xmath568 of semiample classes and the cone @xmath569 of ample classes : [ qcartier ] the map @xmath570 defines canonical isomorphisms : @xmath571 let @xmath572 , and set @xmath573 .
our task is to characterize the statements that @xmath574 is a @xmath64-cartier , a semiample or an ample divisor in terms of the image @xmath575 in @xmath6 . for the description of @xmath576 , recall from @xcite , that @xmath574 is @xmath64-cartier if and only if it arises from a support function , i.e. , there is a family @xmath577 with @xmath578 such that @xmath579 holds on each affine chart @xmath580 with a postive integral multiple @xmath581 . if @xmath574 is @xmath64-cartier , then the describing linear forms @xmath582 are unique up to elements of @xmath583 . now , suppose that @xmath574 is @xmath64-cartier , and let @xmath584 be a describing support function
. then @xmath585 holds for every @xmath586 .
define @xmath587 and denote the cone above @xmath17 by @xmath464 .
then @xmath588 lies in @xmath589 , and hence @xmath590 lies in @xmath591 .
since this applies for all @xmath118 , we obtain @xmath592 conversely , let @xmath593 belong to the last intersection . since for @xmath118 the image @xmath594 contains an element of @xmath16 , we find for each @xmath118 an @xmath595 with @xmath596
. then @xmath597 maps to zero and can therefore be viewed as an element of @xmath598 . hence , @xmath577 provides a support function describing @xmath574 . for the descriptions of @xmath599 and @xmath600 , let @xmath574 be @xmath64-cartier .
recall that @xmath574 is semiample ( ample ) if and only if it is described by a support function @xmath601 , which is convex ( strictly convex ) in the sense that @xmath602 is nonnegative ( positive ) on @xmath603 for any two @xmath604 .
suppose that @xmath574 is semiample ( ample ) with convex ( strictly convex ) support function @xmath601 . in terms of @xmath587
this means that each @xmath605 is nonnegative ( positive ) on @xmath606 .
since @xmath607 holds , this is equivalent to nonnegativity ( positivity ) of @xmath608 on every @xmath606 . since all rays of the cone @xmath232 occur in the fan @xmath8 , the latter is valid if and only if @xmath609 ( resp .
@xmath610 ) holds for all @xmath17 .
this in turn implies that for every @xmath118 we have @xmath611 now , the @xmath612 , where @xmath118 , are precisely the cones of @xmath29 . since any interior @xmath594 contains the interior of a cone of @xmath16 , we can conclude that @xmath75 lies in the respective intersections of the assertion .
conversely , if @xmath75 belongs to one of the right hand side intersections , then we surely arrive at [ intheimages ] .
thus , for every @xmath118 , we find an @xmath613 ( an @xmath610 ) mapping to @xmath75 .
reversing the above arguments , we see that @xmath597 is a convex ( strictly convex ) support function describing @xmath574 . as an immediate consequence of theorem [ qcartier ]
, we obtain a quasiprojectivity criterion in the spirit of @xcite .
[ ewaldcrit ] the variety @xmath0 is quasiprojective if and only if the intersection of all relative interiors @xmath316 , where @xmath19 , is nonempty .
combining theorem [ qcartier ] with proposition [ qfactorialchar ] shows that a @xmath64-factorial quasiprojective toric variety has an ample cone of full dimension .
hence we get back a result of oda and park ( * ? ? ?
* theorem 3.5 ) : suppose that @xmath0 is quasiprojective and @xmath64-factorial .
then every rational weil divisor @xmath574 of @xmath0 admits a representation @xmath614 with ample rational cartier divisors @xmath615 and @xmath616 .
so far , we considered rational divisors and divisor classes . for toric varieties with free class group ,
we also obtain a simple picture for integral divisors .
we need the following notation : every ray @xmath617 of @xmath232 has a unique `` opposite '' ray , namely the unique ray of @xmath211 that is not contained in @xmath618 .
we denote the primitive generator of this opposite ray by @xmath619 , and its image in @xmath13 by @xmath620 .
[ dualbase ] the bunch @xmath16 is free if and only if @xmath621 holds for all @xmath622 . in the free case
, we have the following integral version of the corresponding statements in proposition [ divisors ] and theorem [ qcartier ] : assume that @xmath16 is a free bunch .
then we have : @xmath623 using remark [ dualbase ] , we infer from freeness of @xmath16 that the map @xmath624 is an isomorphism . using the 5-lemma in the diagram of proposition [ divisors ] , we obtain that also the map @xmath625 is an isomorphism .
this gives the first part of the assertion .
the rest of the proof is similar to that of proposition [ qcartier ] : suppose that @xmath626 lies in all @xmath66 , where @xmath68 .
let @xmath627 with @xmath628 .
then for every @xmath118 we can choose @xmath588 even in @xmath629 with @xmath596 .
but then all @xmath597 lie in @xmath228 .
therefore @xmath630 is a cartier divisor and thus we have @xmath631 .
conversely , given @xmath626 with @xmath631 , choose @xmath627 with @xmath628 .
then @xmath630 is a cartier divisor , and hence it is described by a support function @xmath577 with @xmath632 .
but then each @xmath633 lies in @xmath634 , which proves the assertion .
we present some further applications .
if @xmath16 is free , then we can easily describe the canonical divisor class of @xmath0 : by ( * ? ? ?
4.3 ) , the negative of the sum over all invariant prime divisors is a canonical divisor on @xmath0 ; its class is given by @xmath635 = - \b{\mathfrak{d}}(w_1 + \ldots + w_r).\ ] ] recall that a toric variety is said to be _ @xmath64-gorenstein _ if some multiple of its anticanonical divisor is cartier . in terms of bunches
, we have the following characterization .
suppose that @xmath16 is free .
then @xmath0 is @xmath64-gorenstein if and only if we have @xmath636 similarly we can decide whether a toric variety is a fano variety , i.e. its anticanonical divisor is cartier and ample : [ fanocrit ] 1 .
suppose that @xmath16 is free .
then @xmath0 is fano if and only if we have @xmath637 2 .
suppose that @xmath16 is free and @xmath0 is smooth .
then @xmath0 is fano if and only if we have @xmath638 we now apply our results to the case of simple bunches .
recall that these describe precisely the @xmath64-factorial full 2-complete toric varieties .
the first observation is that in this case we immediately obtain the picard group .
suppose that @xmath16 is a simple bunch .
then @xmath624 induces an isomorphism @xmath639 .
it suffices to show that the image of @xmath640 is the group of invariant cartier divisors @xmath641 . as @xmath16 is simple , lemma [ surjectiveandinjective2 ] tells us that the dual projected cone @xmath233 and the maximal projectable fan @xmath8 associated to @xmath16 are the universal reduced cox construction of the fan @xmath1 of @xmath0 .
now , for an element @xmath627 , consider the restrictions @xmath642 , where @xmath118 . by the properties of a universal reduced cox construction
, @xmath643 is an isomorphism for each @xmath118 .
hence , @xmath642 defines a @xmath44-valued linear function on @xmath2 .
this shows that @xmath630 is in fact a cartier divisor .
on the other hand , if @xmath574 is a cartier divisor , we have to show that @xmath644 is @xmath44-valued on @xmath250 .
as above , we see that the restriction of @xmath645 to any sublattice @xmath511 is @xmath44-valued .
this implies the assertion , because lemma [ freeexists ] ( ii ) yields @xmath646 for a @xmath64-factorial toric variety over @xmath647 , we can identify @xmath648 with @xmath6 .
hence we may identify @xmath649 with @xmath650 .
then the mori cone , i.e. , the cone @xmath651 of numerically effective curve classes is dual to the cone of numerically effective divisor classes @xmath652 .
reid proved that the mori cone of a @xmath64-factorial complete toric variety is generated by the classes of the invariant curves , see ( * ? ? ?
using the fact that on a @xmath64-factorial complete toric variety semiample and numerically effective divisors coincide , theorem [ qcartier ] gives a new description of the mori cone : [ moricone ] suppose that @xmath37 is complete and simplicial
. then the cone of numerically effective curve classes in @xmath72 is given by @xmath73 in particular , this cone is convex and polyhedral .
moreover , @xmath37 is projective if and only if @xmath74 is strictly convex .
in this section , we present some applications and examples .
first , we perform kleinschmidt s classification in the setting of bunches .
we use the visualization techniques introduced at the end of section [ section2 ] .
we first show that every smooth 2-complete toric variety @xmath0 with @xmath77 and @xmath653 arises from a bunch as in the assertion . from theorem [ standard2maximal ] ( ii )
, we infer that the toric variety @xmath0 arises from a free bunch @xmath16 in a projected cone @xmath654 with @xmath655 .
let @xmath656 be the set of weight vectors , and let @xmath217 be the multiplicity of @xmath15 .
smoothness of @xmath0 means that every image @xmath657 , where @xmath68 , generates @xmath53 , see proposition [ smoothchar ] . since we have @xmath496 , proposition [ nontrivfunctionschar ] tells us that the cone @xmath658 generated by the weight vectors is strictly convex .
being two - dimensional , @xmath659 is generated by two vectors , say @xmath660 .
now we make essential use of the fact that @xmath16 lives in a two - dimensional space : consider the intersection @xmath155 of all @xmath19 .
then @xmath155 is a weight cone .
moreover , @xmath155 is of dimension two , because for any two cones of @xmath16 their relative interiors intersect .
thus the defining property of a bunch implies @xmath661 . as a strictly convex two - dimensional weight cone containing @xmath155 , also @xmath659 occurs among the images @xmath67 , where @xmath68 .
consequently , @xmath660 is a regular cone in @xmath53 , and moreover , using remark [ downstairsisochar ] , we may assume that @xmath662 and @xmath663 are the canonical base vectors of @xmath53 .
we claim that the cone @xmath155 has at least one ray in common with @xmath659 . indeed
, otherwise one of the cones generated by @xmath664 or @xmath665 would be non regular . but
both cones occur among @xmath67 , where @xmath68 .
a contradiction .
again by remark [ downstairsisochar ] , we may assume that @xmath155 contains @xmath666 , and hence is generated by vectors @xmath666 and @xmath667 , where @xmath668 .
consider any @xmath669 .
then @xmath670 overlaps @xmath155 , and thus the properties of a bunch give @xmath671 .
consequently @xmath672 for some @xmath68 , and hence @xmath15 and @xmath673 generate @xmath53 . but
this means @xmath674 with @xmath675 .
so we arrive at the desired picture .
note that the conditions @xmath83 and @xmath676 are due to property [ standarddefi ] ( ii ) .
the fact that the toric variety @xmath0 associated to @xmath16 is projective is clear by proposition [ completenesschar ] and corollary [ ewaldcrit ] .
moreover , corollary [ fanocrit ] tells us that @xmath0 is fano if and only if the weighted sum @xmath677 lies in the relative interior of the cone @xmath487 .
but this holds if and only if the condition stated in the assertion is valid .
[ oda ] the simplest nonprojective complete simplicial fans @xmath1 live in @xmath50 and are of the following type : consider a prism @xmath678 over a 2-simplex such that @xmath679 lies in the relative interior of @xmath89 .
here we take @xmath89 with the vertices @xmath680 then subdivide the facets of @xmath89 according to the picture below , and let @xmath1 be the fan generated by the cones over the simplices of this subdivision . the bunch @xmath16 corresponding to @xmath1 is free and lives in a 3-dimensional lattice @xmath681 .
combinatorially , it looks as indicated above .
explicitly it is given by the weight vectors @xmath666 , @xmath682 , @xmath683 , @xmath684 , @xmath685 , and @xmath686 in @xmath50 , and the four cones @xmath687 using the propositions [ fullandsimplicial ] , [ completenesschar ] , and corollary [ ewaldcrit ] it is immediately clear from the picture that @xmath485 is @xmath64-factorial and nonprojective .
moreover , by theorem [ qcartier ] , the cone @xmath688 of semiample divisors is spanned by the class of the anticanonical divisor .
the last example concerns the problem whether or not the betti numbers of a toric variety are determined by the combinatorial type of the defining fan .
a first counterexample was given by mcconnell , see @xcite .
later eikelberg @xcite gave the following simpler one : [ eikelberg ] let @xmath678 be a prism over a 2-simplex such that @xmath679 lies in the relative interior of @xmath89 , and define @xmath1 to be the fan generated the cones over the facets of @xmath89 . as vertices of @xmath89 we take : @xmath689 eikelberg constructs a nonprojective fan @xmath690 from @xmath1 by moving the ray through @xmath691 into the ray through @xmath692 .
the picture is the following ( the dotted diagonals indicate the edges of the convex hull over @xmath693 that do not define cones of @xmath690 ) : what does this mean in terms of bunches ?
first note that the bunches @xmath16 and @xmath226 corresponding to @xmath1 and @xmath690 contain 2-dimensional cones , because @xmath1 as well as @xmath690 have nonsimplicial cones .
the loss of projectivity is reflected in terms of bunches as follows : here we draw the intersection of @xmath16 and @xmath226 with a plane orthogonal to an inner vector of the ( strictly ) convex hulls @xmath694 and @xmath695 of the unions of the respective bunch cones .
so the fat lines correspond to 2-dimensional bunch cones , whereas the shaded simplices represent 3-dimensional bunch cones .
now , both fans @xmath1 and @xmath690 have the same combinatorial type . in terms of bunches , we see immediately that the associated toric varieties @xmath0 and @xmath696 have different betti numbers : the second betti number @xmath697 equals the dimension of @xmath698 .
hence theorem [ qcartier ] gives us @xmath699 .
for @xmath696 , we obtain @xmath700 by the same reasoning .
we conclude with a remark concerning ewald s construction of `` canonical extensions '' of a given toric variety presented in ( * ? ? ?
* section 3 ) , compare also @xcite .
the bunch theoretical analogue is the following : consider a free bunch , represented in the sense of construction [ bunchfromdownstairs ] by a set weight vectors @xmath656 in a lattice @xmath13 , multiplicities @xmath701 and a collection @xmath16 of weight cones .
define a new free bunch by setting @xmath702 and replacing the multiplicities @xmath217 with bigger ones , say @xmath703 . for the toric varieties @xmath0 and @xmath696 associated to these bunches , we have @xmath704 moreover , applying the respective characterizations of section [ section7 ] and [ section9 ] , one immediately verifies that @xmath696 is non quasiprojective ( complete , @xmath64-factorial , smooth ) if @xmath0 was so .
a. acampo - neuen , j. hausen : quotients of toric varieties by the action of a subtorus .
tohoku math .
j. 51 ( 1999 ) , 112 f. berchtold , j. hausen : demushkin s theorem in codimension one .
z. 244 , no.4 , 697703 ( 2003 ) a. biaynicki - birula : algebraic quotients . in : r.v .
gamkrelidze , v.l .
popov ( eds . ) , encyclopedia of mathematical sciences , vol .
, 182 ( 2002 ) a. biaynicki - birula , j. wicicka : a recipe for finding open subsets of vector spaces with a good quotient .
77 , no . 1 ( 1998 )
, 97114 l. bonavero : sur des varits toriques non projectives .
128 , no.3 , 407431 ( 2000 ) d. cox : the homogeneous coordinate ring of a toric variety .
geom . 4 ( 1995 ) , 1750 m. eikelberg : the picard group of a compact toric variety .
22 , no . 1 - 2 ( 1992 )
, 509527 g. ewald : spherical complexes and nonprojective toric varieties .
discrete comput .
geometry 1 ( 1986 ) , 115122 g. ewald : polygons with hidden vertices .
contributions to algebra and geometry , vol .
2 , 439442 ( 2001 ) w. fulton : introduction to toric varieties .
the william h. roever lectures in geometry , princeton university press ( 1993 ) b. grnbaum : convex polytopes .
london , new - york , sydney : interscience publishers , john wiley & sons .
j. hausen : producing good quotients by embedding into toric varieties .
et congr . 6
( 2002 ) , 193212 h. a. hamm : very good quotients of toric varieties . in : bruce , j. w. ( ed . ) et al .
real and complex singularities .
proceedings of the 5th workshop ; so carlos , brazil , july 27 - 31 , 1998 .
chapman / hall / crc res .
notes math .
412 ( 2000 ) , 6175 y. hu : combinatorics and quotients of toric varieties .
discrete comput .
28 , 151174 ( 2002 ) p.
kleinschmidt : a classification of toric varieties with few generators .
aequationes math .
35 , no.2/3 ( 1988 ) , 254266 m. mcconnell : the rational homology of toric varieties is not a combinatorial invariant .
105 , no . 4 ( 1989 ) , 986991 .
mcmullen : transforms , diagrams and representations .
contributions to geometry , proc .
symp . , siegen 1978 , 92 - 130 ( 1979 ) .
t. oda : convex bodies and algebraic geometry .
ergebnisse der mathematik und ihrer grenzgebiete , band 15 .
springer verlag ( 1988 ) t. oda , h.s .
park : linear gale transforms and gelfand - kapranov - zelevinskij decompositions .
thoku math .
j. , 43 ( 1991 ) , 375399 g.c .
shephard : spherical complexes and radial projections of polytopes .
israel j. math .
35 ( 1971 ) , 236238 m. reid : decomposition of toric morphisms . in : arithmetic and geometry , papers dedicated to i.r .
shafarevich on the occasion of his 60th birthday , vol .
ii , geometry . progress . in math .
36 , birkhuser , boston , basel , stuttgart .
1983 ( 395418 ) j. wicicka : quotients of toric varieties by actions of subtori .
colloq . math .
82 , no . 1 ( 1999 ) , 105116 | as an alternative to the description of a toric variety by a fan in the lattice of one parameter subgroups , we present a new language in terms of what we call bunches these are certain collections of cones in the vector space of rational divisor classes .
the correspondence between these bunches and fans is based on classical gale duality .
the new combinatorial language allows a much more natural description of geometric phenomena around divisors of toric varieties than the usual method by fans does .
for example , the numerically effective cone and the ample cone of a toric variety can be read off immediately from its bunch . moreover
, the language of bunches appears to be useful for classification problems . |
apparently very different systems as high - gain free electron laser ( fel ) @xcite and collective atomic recoil laser ( carl ) @xcite exhibit similar behaviors , showing self - bunching and exponential enhancement of the emitted radiation . originally conceived in a semiclassical framework , they can be as well described quantum - mechanically @xcite .
however , it is not explicitly evident how to obtain the classical limit starting from the quantum description .
first attempts to give a quantum description of the fel have been proposed in the 80s , starting from a canonical quantization of the @xmath3-particle hamiltonian in the heisenberg picture , to study photon statistics and quantum initialization from vacuum in the linear regime @xcite . in 1988 , preparata proposed a quantum field theory of fel @xcite , in which he has shown that , for @xmath4 , the fel dynamics is solved by a single - electron schrdinger equation coupled to a self - consistent radiation mode .
the same model has been recently obtained to describe carl from a bose - einstein condensate ( bec ) at zero temperature @xcite .
furthermore , it has been also proved experimentally @xcite that carl in a bec exhibits quantum recoil effects when the average recoil velocity remains less than the photon recoil limit . in this letter
we start from the classical model describing both carls and fels and we extend it to the quantum realm showing the correspondence between the preparata model @xcite and the carl - bec model @xcite . in particular
, it is possible to derive an equation for the wigner function of the @xmath3-particle system .
the wigner function obeys to a finite difference equation which reduces to the classical vlasov equation @xcite in the limit in which the number of photons emitted per particle is much larger than unity . in the opposite limit ,
both carls and fels behave as a two - state system @xcite described by the well - known maxwell - bloch equations @xcite .
apparently the physics of fel and carl appears to be quite different .
the first describes a relativistic high current electron beam with energy @xmath5 , injected in a magnet ( wiggler ) with a transverse , static magnetic field @xmath6 and periodicity @xmath7 , which radiates in the forward direction at the wavelength @xmath8 , where @xmath9 is the wiggler parameter and @xmath10 . instead , carl consists of a collection of two level atoms in a high - q ring cavity driven by a far - detuned laser pump of frequency @xmath11 which radiates at the frequency @xmath12 in the direction opposite to the pump . in both cases
the radiation process arises from a collective instability which originates a symmetry breaking in the spatial distribution , _
i.e. _ a self - bunching of particles which group in regions smaller than the wavelength .
it can be shown that , under suitable conditions and introducing proper dimensionless variables , the dynamics of both fels and carls is described by the following hamiltonian @xcite @xmath13-\frac{\delta}{\bar \rho}\ ; a^{\dag } a , \label{ham}\ ] ] where @xmath14 and @xmath15 are the phase operator of the @xmath16-th particle and its conjugate momentum operator , obeying @xmath17=i\delta_{jj'}$ ] . in eq.([ham ] ) , @xmath18 and @xmath19 are annihilation and creation operators for the forward radiation mode photon , with @xmath20=1 $ ] .
notice that the dynamics described by eq.([ham ] ) depends only on the parameter @xmath21 and on the detuning @xmath22 , properly defined for the two systems : * for fels , @xmath23 and @xmath24 , where @xmath25 , @xmath26 is the resonant energy and @xmath27 is the bpn parameter for a fel @xcite , @xmath28 , @xmath29 and @xmath30 . * for carls , @xmath31 and @xmath32 , where @xmath33 is the recoil frequency , @xmath34 , @xmath35 $ ]
, @xmath36 is the pump rabi frequency , @xmath37 is the pump - atom detuning , @xmath38 is natural decay constant of the atomic transition and @xmath39 is the dipole matrix element @xcite .
finally , @xmath40 and @xmath41 , where @xmath42 is the longitudinal atomic velocity . in these definitions
, @xmath43 is the particle density in the radiation volume @xmath44 and @xmath45 is the particle mass .
notice that in both cases @xmath46 scales as @xmath47 , i.e. as the reciprocal of the inter - particle distance . introducing @xmath48 and @xmath49 , the heisenberg equations associated with eq.([ham ] )
are @xcite : @xmath50 where @xmath51 and @xmath52 for carl whereas @xmath53 for fel . considering the operators @xmath14 , @xmath54 and @xmath55 in eqs.([c1])-([c3 ] ) as c - numbers , one obtain the well - known classical description for fel and carl .
the scaling of eqs.([c1])-([c3 ] ) is called ` universal ' in the sense that , assuming resonance ( _ i.e. _ @xmath56 ) , the equations do not contain any parameter .
hence , the scaling law of the various physical quantities can be obtained from their definition in term of @xmath21 . in particular , from the definition of @xmath55 and @xmath57 it follows that the photon number per particle and the momentum recoil are proportional to @xmath21 , both for fels and carls .
hence , it is expected that when @xmath1 the system behaves classically , whereas for @xmath2 quantum effects becomes relevant .
notice that @xmath58 is a constant of motion in eqs.([c1])-([c3 ] ) , _ i.e. _ the radiated intensity is due to the average recoil .
in ref . @xcite preparata , using quantum field theory , has shown that the collective dynamics of the system of @xmath59 electrons in an fel can be described by means of a single complex scalar quantum field whose behavior is governed by a schrdinger - type equation in the self - consistent radiation field , which originates a pendulum - like potential : @xmath60\psi \label{psi}\\ \frac{da}{d\tau}&=&\int_{0}^{2\pi } d\theta|\psi|^2e^{-i\theta}+\frac{i\delta}{\bar\rho}a , \label{a}\end{aligned}\ ] ] where @xmath61 is normalized to one , i.e. @xmath62 . note that eq.([psi ] ) is the schrdinger equation associated to the hamiltonian ( [ ham ] ) and eq.([a ] ) corresponds to eq.([c3 ] ) when the classical average of @xmath63 is replaced by the quantum ensemble average . quoting ref .
@xcite , eqs.([psi ] ) and ( [ a ] ) are derived if one `` formulate the many - electron problem in the language of quantum field theory and uses the large number @xmath3 of electrons to evaluate the resulting path integral by saddle - point techniques '' .
recently , the same model of eqs.([psi ] ) and ( [ a ] ) has been used to describe carl from a bec * @xcite*. hence , we propose the nonlinear system of eqs.([psi ] ) and ( [ a ] ) as the quantum extension of the carl - fel classical model .
we now show that the classical equations ( [ c1])-([c3 ] ) are recovered in the limit @xmath64 .
let s consider the standard definition of the wigner function for a state with wave function @xmath65 : @xmath66 so that @xmath67 one can show that eq.([psi ] ) is equivalent to the following finite difference equation for the quasi - probability distribution @xmath68 : @xmath69 \left [ w\left(\theta,\bar p+\frac{1}{\bar \rho } , \tau\right ) - w\left(\theta,\bar p - \frac{1}{\bar \rho } , \tau\right ) \right ] = 0.\ ] ] using eq .
( [ w : psi ] ) , eq . ( [ a ] ) becomes : @xmath70 we underline again that eqs.([weqn ] ) and ( [ a : w ] ) are equivalent to eqs.([psi ] ) and ( [ a ] ) using the wigner function representation . in the right hand side of eq .
( [ weqn ] ) , the incremental ratio @xmath71/(2\epsilon)\rightarrow \partial w(\theta,\bar p)/\partial\bar p$ ] when @xmath72 . hence , for @xmath1 eq .
( [ weqn ] ) becomes the vlasov equation : @xmath73 \frac{\partial w(\theta,\bar p,\tau)}{\partial\bar p } = 0.\ ] ] eqs .
( [ a : w ] ) and ( [ vlasov ] ) are equivalent to the classical eqs .
( [ c1])-([c3 ] ) .
this means that the particles behave classically , following a newtonian motion , when @xmath1 i.e. when the average number of photons scattered per particle is much larger than unity . in this limit ,
the quantum recoil effects due to the single photon scattering process is negligible . on the contrary
, a quantum regime of carl or fel occurs when @xmath2 , in which each particle scatters only one photon . ) and ( [ a3 ] ) for @xmath74 ( first row ) , @xmath75 ( second row ) and @xmath76 ( third row ) .
the other parameters are @xmath77 , @xmath78 and @xmath79 . left column : dimensionless radiation intensity @xmath80 vs. @xmath81 ; central and right column : occupation probabilities @xmath82 vs. @xmath83 and density distribution @xmath84 vs. @xmath85 , for @xmath81 near the first maximum of @xmath80.,scaledwidth=90.0% ] in fact , expanding the wave function in fourier series as @xmath86 and inserting this ansatz in eqs.([psi ] ) and ( [ a ] ) , one can easily obtain the following closed set of equations for @xmath87 : @xmath88\label{w3}\\ \frac{d\bar{a}}{d\tau}&=&\sum_{n=-\infty}^{\infty}\varrho_{n-1,n } , \label{a3}\end{aligned}\ ] ] where @xmath89 .
these equations are equivalent to eqs.([psi ] ) and ( [ a ] ) for the density matrix in the momentum representation and have been discussed in ref.@xcite .
fig.[figura1 ] shows the numerical solution of eq.([w3 ] ) and ( [ a3 ] ) for @xmath74 ( first row ) , @xmath75 ( second row ) and @xmath76 ( third row ) .
the other parameters are @xmath78 , @xmath79 and @xmath77 , which corresponds to a single photon scattering recoil . in the first column @xmath80
is plotted as a function of @xmath81 .
the central column shows the occupation probabilities @xmath82 vs. @xmath83 , whereas the right column shows the density distribution @xmath84 vs. @xmath85 , for a value of @xmath81 near the first maximum of @xmath80 .
we note that for @xmath74 the system behaves classically : the momentum states are occupied in a range of the order of @xmath21 and the average momentum is @xmath90 , as can be seen from the first row of fig.[figura1 ] .
furthermore , the particle distribution shows periodic narrow peaks of density .
when @xmath75 , the mainly occupied momentum states are those for @xmath91 and @xmath92 , corresponding to particles in the initial state or in the recoil state , respectively . finally , when @xmath93 , the dynamics is that of a pure two - level system .
in this limit , if the initially occupied state is the @xmath83th momentum state , the only two momentum states involved in the interaction are those for @xmath83 and @xmath94 , so that eq.([w3 ] ) and ( [ a3 ] ) , after defining the ` polarization ' @xmath95 and the ` population difference ' @xmath96 , reduce to the maxwell - bloch equations for a two - state system @xcite : @xmath97 where @xmath98 , @xmath99 and @xmath100 . with this new
scaling and assuming resonance ( i.e. @xmath101 ) , eqs.([s])-([a:4 ] ) do not contain any parameter .
hence , the characteristic timescale is ruled by @xmath102 instead of @xmath47 as in the classical case .
the quantum regime for carls and fels is analog to the coherent spontaneous emission regime predicted quantum - mechanically for a two - level system in ref .
@xcite , where a series of optical `` @xmath103-pulses '' are generated .
in fact , assuming resonance ( i.e. @xmath101 ) , @xmath104 and @xmath105 are real . hence , we can introduce the `` bloch angle '' @xmath106 such that @xmath107 , @xmath108 .
then , eqs.([s])-([a:4 ] ) reduce to a pendulum equation @xmath109 and @xmath110 . hence , in the quantum regime
, the dynamics is that of a pendulum moving away from the unstable equilibrium point ( @xmath111 ) and undergoing periodically a complete revolution ( @xmath103-pulse ) with angular velocity @xmath104 . finally , we note that , adopting the same scaling of eqs.([s])-([a:4 ] ) in eq.([psi ] ) ,
this can be interpreted as a schrdinger equation for a single particle with a `` mass '' @xmath112 in a self - consistent pendulum potential .
this provides an intuitive interpretation of the classical limit , that holds when the particle s mass is large .
the strong differences between the quantum and classical regimes are evident from fig.[figura1 ] .
in this letter we presented a unified quantum model that stands for apparently very different systems as fel and carl .
the dynamics is described by a schrdinger equation in a self - consistent pendulum potential and is ruled by an unique parameter @xmath21 which represents the maximum number of photons scattered per particle and the maximum momentum recoil in units of the photon recoil momentum .
the schrdinger equation can be transformed in an exact equation for the wigner quasi - probability distribution .
the main results are the following : i ) the classical model is recovered in the limit @xmath113 ; this because the finite difference equation for the wigner function reduces to the classical vlasov equation .
ii ) in the limit @xmath114 a completely different dynamical regime occurs ( see fig . [ figura1 ] ) : due to momentum quantization the system reduces to only two momentum states obeying to the maxwell - bloch equations which describe the dynamics of a two - level atomic system coupled to a coherent field .
one of us ( rb ) has to acknowledge a big mistake he made almost 20 years ago when giuliano preparata , a well - known field theorist , presented to him his general `` quantum field theory of a free electron laser '' @xcite .
rb did not understand this work thinking it was incorrect . on the contrary ,
preparata s theory was perfectly correct , as recognized in this letter , which is dedicated to his memory .
99 bonifacio r. , pellegrini c. and narducci l. , opt .
50 ( 1984 ) 373 .
bonifacio r. , de salvo souza l. , nucl .
instrum . and meth .
res . a 341 ( 1994 ) 360 ; bonifacio r. , de salvo souza l. , narducci l. and dangelo e.j . ,
rev.a 50 ( 1994 ) 1716 .
preparata g. , phys .
rev.a 38 ( 1988 ) 233 .
bonifacio r. and casagrande f. , optics comm .
50 ( 1984 ) 251 .
bonifacio r. , optics comm .
146 ( 1998 ) 236 .
moore m.g . , zobay o. and meystre p. , phys .
rev.a 60 ( 1999 ) 1491 . ling h.y .
, pu h. , baksmaty l. and bigelow n.p .
rev.a 63 ( 2001 ) 053810 .
piovella n. , cola m. , bonifacio r. , phys .
rev.a 67 ( 2003 ) 013817 .
becker w. and mciver j.k .
rev.a 28 ( 1983 ) 1838 .
bonifacio r. , cataliotti f.s . ,
cola m. , falani l. , fort c. , piovella n. , inguscio m. , optics comm . 233 ( 2004 ) 155 .
inouye s. et al . , science 285 ( 1999 ) 571 .
kozuma m. et al . , science 286 ( 1999 ) 2309 ; inouye s. et al . ,
nature 402 ( 1999 ) 641 .
bonifacio r. , verkerk p. , optics comm .
124 ( 1996 ) 489 .
bonifacio r. and de salvo souza l. , optics comm . 115 ( 1995 )
inouye s. et al . , phys .
85 ( 2000 ) 4225 .
piovella n. , gatelli m. and bonifacio r. , optics comm .
194 ( 2001 ) 167 .
avetissian h.k . and mkrtchian g.f .
rev.e 65 ( 2002 ) 046505 .
arecchi f.t .
, bonifacio r. , ieee quantum electron . 1 ( 1965 ) 169 .
bonifacio r. and preparata g. , phys .
rev.a 2 ( 1970 ) 336 . | free electron laser ( fel ) and collective atomic recoil laser ( carl ) are described by the same model of classical equations for properly defined scaled variables .
these equations are extended to the quantum domain describing the particle s motion by a schrdinger equation coupled to a self - consistent radiation field .
the model depends on a single collective parameter @xmath0 which represents the maximum number of photons emitted per particle .
we demonstrate that the classical model is recovered in the limit @xmath1 , in which the wigner function associated to the schrdinger equation obeys to the classical vlasov equation . on the contrary , for @xmath2 , a new quantum regime
is obtained in which both fels and carls behave as a two - state system coupled to the self - consistent radiation field and described by maxwell - bloch equations . |
the role of phonons in the mechanism of high - temperature superconductivity in ferropnictides remains poorly understood .
while density function theory ( dft ) calculations indicate that conventional electron - phonon coupling is not responsible for superconductivity @xcite , there is nevertheless a strong relationship between the lattice and electronic properties , including both magnetism and superconductivity .
for example , there is an isotope effect due to fe @xcite , and the superconducting temperature strongly depends upon the fe - as - fe bond angle @xcite ( or alternatively the height of the pnictogen atom above the plane of the transition metal@xcite ) .
in addition , the lattice and magnetism are strongly coupled , which manifests in several ways .
first , in order to correctly refine the internal parameters of the structure , dft calculations must be performed with spin - polarization @xcite .
second , cafe@xmath1as@xmath1 enters a `` collapsed '' tetragonal state at mild isotropic pressures @xcite , in which the c - axis contracts by @xmath010% and the magnetism becomes quenched @xcite ; this is the result of strong coupling between the fe moment and interplanar as - as bonding @xcite .
third , the increase of the static moment with as - fe separation can be described by a landau / stoner theory @xcite , in which a larger volume around each fe atom allows a larger magnetic moment .
this dependence of the moment on as position has been confirmed experimentally @xcite for the static moment in the cefeas@xmath2p@xmath3o series .
this strong magnetostructural coupling can in turn influence the phonon spectrum .
for instance , the frequency of the raman - active mode corresponding to vibration of the as atoms is not predicted correctly unless the fe moment is taken into account @xcite .
in addition , the phonon density of states in lafeaso is appreciably softer than predicted by nonmagnetic dft @xcite , especially for the higher - energy modes corresponding to in - plane motion of the fe and as atoms . while the dispersion of the c - axis modes has been studied elsewhere @xcite ) , we are not aware of any reports regarding the in - plane fe and as phonons . here
we present measurements of the phonon dispersion in srfe@xmath1as@xmath1 obtained through time - of - flight ( tof ) inelastic neutron scattering , focusing on the closely - spaced higher - energy modes which are sensitive to fe - as coupling and compare these results to the dft calculations . in order to isolate the dispersions of different phonons from nearby branches
, we developed a novel approach that utilizes the entire tof spectrum across many brillouin zones ( bz ) , which we call multizone phonon refinement ( mpr ) .
we show that the experimental phonon dispersions agree well with dft calculations performed with the structure constrained to the measured structure .
we also found that there is no significant temperature dependence across the magnetic ordering transition of any of the phonons for which it was measured .
phonon dispersion along ( 2,k,0 ) .
this is the inelastic scattering for the regions shown as red boxes in panels @xmath4 and @xmath5 . near ( 2,0,0 )
the phonons are strictly transverse , but near ( 2,6,0 ) the longitudinal phonons are also evident .
@xmath6 map of elastic scattering in the ( 2,k , l ) plane , showing many bragg peaks .
@xmath7 map of elastic scattering in the ( h , k,0 ) plane , showing 8 bragg peaks . ]
the experiment was performed using the arcs time - of - flight spectrometer at the spallation neutron source , located at the oak ridge national laboratory @xcite .
data were collected using the recently - developed rotating sample technique @xcite , which allows us to collect a 4-dimensional dataset ( h , k , l , energy ) .
the sample was a single crystal of srfe@xmath1as@xmath1 , grown from self - flux in glassy carbon crucibles @xcite , using extra - low cooling rates of @xmath8 - @xmath9 c / h .
the sample mass was @xmath0 1.5 g , and the mosaic was @xmath0 @xmath10 .
at room temperature the material is tetragonal ( space group 139 , _ i4/mmm _ ) , and at @xmath0 210 k undergoes an afm / orthorhombic transition @xcite ( space group 69 , _ fmmm _ ) . because the orthorhombic distortion is quite small ( @xmath11 @xmath0 1% ) , we use tetragonal notation throughout this paper .
the sample was mounted using twisted aluminum foil , and placed in a closed - cycle refrigerator .
no exchange gas was used , so the temperatures are accurate to within 10 k. data were collected over a period of 6 days ; approximately 3 days over an angular range of @xmath12 at t = 10 k , and approximately 1.5 days over a range of @xmath13 at both t = 220 k and t = 320 k. this corresponds to roughly 45 bz in the energy range of interest for 10 k and 20 bz for 220 k ( just slighly above the magnetic ordering temperature ) , and 320 k. the sample was mounted in the hk0 plane , but the out - of - plane detector coverage allows access up to l @xmath0 @xmath14 4 .
the incident energy was 70 mev , with a fwhm of @xmath0 3 mev at the elastic line , and @xmath0 1.5 mev in the range of interest .
the details of the dft calculations have been described previously @xcite .
we used the non - magnetic tetragonal unit cell , with the experimental lattice parameters and atomic positions taken from previously - published work @xcite .
the eigenvectors are closely related to the atomic displacements , which in turn determine the intensities of the phonon peaks in various bz .
these eigenvectors can be calculated quite well at the high - symmetry points , and the calculation allows for the phonon intensity at other points to be calculated using a perturbative method .
these energies and intensities are then used as a starting point for the refinement , as described below .
$ ] ( gamma point ) .
the background ( solid red line ) can be subtracted from the raw data ( red diamonds ) to produce a remarkably clean spectrum ( blue circles ) .
dashed lines show the refined values of the peak centers .
the intensities shown are predicted from dft , which provides reasonable initial conditions for least - squares fitting . ] = [ 0.7 , 0 , 0 ] from selected brillouin zones .
the left panel displays the data ( points with error bars ) together with phonon energies and intensities as predicted by dft ; the right panel displays the same data , along with the energies and intensities after multizone phonon refinement .
although data from only 5 symmetry points are shown here , a total of 154 symmetry points were used for the fitting .
data from different @xmath15 are vertically offset for clarity .
the model consists of 6 phonons each constrained to have the same energy and lineshape in all bz .
the individual phonons broadened by experimental resolution are shown with thin lines , and the summation with solid lines .
the dashed vertical lines mark the phonon energies , and facilitate comparison between the different @xmath15 . ]
optical phonons in many compounds with a large unit cell form dense `` spaghetti '' of branches in large parts of reciprocal space .
in such a case energies of different phonons at the reduced wavevector are separated by much less than the energy - resolution of the spectrometer .
quite often it is easy to mistake two nearby peaks for a single broad peak .
here we outline a novel technique , multizone phonon refinement ( mpr ) that allows accurate determination of phonon dispersions in such cases .
we use this technique to resolve a group of branches between 24 and 36 mev in srfe@xmath1as@xmath1 , measured as described in the previous section . in neutron - scattering experiments the inelastic background arises from a combination of factors , primarily incoherent phonon scattering by the sample , powder - averaged phonon scattering from the mount ( in this case aluminum foil ) , secondary scattering ( a bragg peak and a phonon ) , and multiphonon scattering ( strongest at high temperature ) .
the first factor necessarily produces isotropic scattering with a @xmath16 dependence characteristic of phonons . for this work all contributions to the inelastic background are assumed to follow this form . in order to determine the background , we divided measured reciprocal space into equally - sized volume elements ( known as `` voxels '' ) , normalizing by @xmath16 .
each voxel corresponds to an energy scan at a fixed @xmath15 . for a given energy , the mean of the lowest - intensity voxels
was taken to be the background ( see fig .
[ fig : bg_subt ] ) .
in other words , for each energy we found the lowest intensities over the entire measured dataset ( after dividing by @xmath16 ) and took that to be the background .
figure [ fig : bg_subt ] shows such energy spectra plotted together in thin lines . the thick solid line corresponding to the average intensity of the lowest 1% of the @xmath15-voxels . we experimented with using the lowest 1% to the lowest 5% of the @xmath15-voxels at each energy , and found only minor changes to the resulting background function . since our primary interest here was the higher - energy phonons , we made no attempt to subtract the elastic line . in future experiments
we can subtract the elastic line separately from each voxel of @xmath15 , and then subtract the background .
an additional linear term was included in the fitting to account for any residual intensity , but was generally insignificant . at lower @xmath15
there is some additional background , but these data have little effect on the overall fit . as can be seen in fig .
[ fig : bg_subt ] , the subtraction works quite well in the sense that it results in spectra that consist of phonon peaks on top of a very small linear background . note that because this method measures the background empirically , the source of the background ( sample , aluminum , multiphonon , elastic line , etc . )
is irrelevant ; the only stipulation is that the angular range of data collection be large enough so that at some point of the measured range of s@xmath17 , the phonon structure factor goes to zero .
the precise range of @xmath15 required to meet that condition varies depending on symmetry of the sample and the energy range of interest . in the worst - case scenario ( collecting over the entire energy range ) , the _ i4/mmm _ symmetry of this sample would require a rotation of @xmath18 , plus an additional range of @xmath0 @xmath13 to account for the different region of reciprocal space which is measured at different energy transfers . in practice the actual range covered may be substantially less ; we found no appreciable difference whether gathering the background from a @xmath12 or @xmath13 range .
one potential limit of the background subtraction procedure is in the case of a non - dispersive mode .
if a mode maintained both a constant energy and a nonzero structure factor , then some of its intensity might be incorrectly assigned to the background . however , this situation can only occur for an isolated atom in an isotropic environment ( a rattling mode inside of a rigid cage might approximate this condition ) .
other than this special case , the intensity of a well - behaved mode must vary with @xmath15 , due to either the phase factor , or to the projection of the eigenvector onto @xmath15 .
traditionally , phonon measurements have been made at specific wavevectors in one or two bz .
this works when phonons are well - separated in energy , but in cases of multiple overlapping peaks it is often impossible to distinguish a single broad peak from two or more narrow peaks .
tof inelastic neutron scattering instruments capture data over many bz .
thus phonon peaks belonging to a multiplet of overlapping peaks at a specific reduced wavevector will have different intensities in different bz , but they will have the same energies and intrinsic linewidths . to perform the fitting , we first start with predictions of dft calculations , which give approximate phonon frequencies and structure factors .
some of these are very accurate and agree with the data without any fitting , whereas others are significantly off .
( see left panel in fig .
[ fig : multizone ] ) by simultaneously fitting the experimental phonon spectrum at the same reduced wavevector in multiple bz , we can identify experimental positions of individual phonons them quite well if we constrain their widths and energies to be bz - independent while letting their intensities vary from zone to zone ( see right panel in fig .
[ fig : multizone ] ) .
the energy resolution of arcs is nearly independent of @xmath15 over the entire instrument , so values for resolution widths were calculated as a function of energy , as described elsewhere @xcite . because arcs has an asymetric resolution function , the peaks were modeled as split gaussians , with the lower ( higher ) side of the peak having 60% ( 40% ) of the fwhm .
the peak width and asymmetry were checked at several @xmath15 for which dft predicted only a single phonon was allowed in the 32 - 35 mev range , and this modeling of the energy resolution was found to be satisfactory . while on a triple - axis instrument the momentum and energy resolutions are strongly coupled , for arcs they are largely independent .
the @xmath15-resolution at higher momentum transfer ( where phonons are strongest ) is determined primarily by the sample mosaic . our sample mosaic of @xmath10 sets @xmath19/@xmath15 @xmath03% , on the same order as the @xmath15-resolution set by size of the detector pixels ( @xmath01% ) .
the in - plane voxel size ( q - binning ) was either @xmath14 .025 r.l.u . , or @xmath14 0.05 r.l.u . , which makes @xmath19/@xmath15 between 1 - 2% or 2 - 4% , respectively .
the out - of - plane voxel size was @xmath14 0.5 r.l.u .
in all cases ; this puts @xmath19/@xmath15 @xmath06% , but the dft calculation indicates that the energy should vary by less than 1% over this range ( c - axis modes are generally fairly non - dispersive in layered systems such as this one ) , reducing the effect of @xmath15-resolution on determination of the phonon energies . in order to simulate the combined effect of dispersion and finite bin size on the energy resolution , we increased the calculated energy fwhm by 10% . for weakly ( strongly ) dispersive branches , this will cause the refined separation between overlapping modes to be smaller ( larger ) than the true separation . a more detailed analysis could modify the energy broadening based upon bin size and local dispersion as calculated from dft .
other than this broadening , no attempt was made to account for the effect of @xmath15-resolution on the energy linewidth .
the fitting program was written in matlab , using a standard least - squares algorithm .
each reduced-@xmath20 symmetry point was fit separately .
the heights of the peaks were constrained to be positive , but could vary between bz .
peak centers were constrained to be identical in various bz .
resolution widths were fixed as described above .
for each voxel , the empirical background was subtracted , and an additional linear term was included in the fitting to account for any residual background ( but was generally small ) .
the energy linewidth was fixed to be resolution - limited , but can be made a fitted parameter in the future .
solid lines are from t = 10 k , dashed lines from t = 220 k , and dotted lines from t = 320 k. the data from t = 220 k and 320 k were taken over a shorter time and smaller angular region and are correspondingly noisier .
the endpoints are labelled with the same numbers used in fig .
[ fig : disp ] . ] figure [ fig : disp ] shows extracted dispersions and branch assignments for narrowly - spaced phonon branches in srfe@xmath1as@xmath1 .
note that their separation in energy was often much smaller than the experimental resolution .
our method of determining the phonon peak positions automatically provides bz - dependence of peak intensities for each reduced @xmath20 , which is determined by the phonon eigenvectors .
thus we assign phonon peaks to the same branch if their eigenvectors vary smoothly throughout the zone ( a similar method has been previously used to assign eigenvectors using triple - axis ins @xcite ) .
symmetry constraints and comparison of experimental results with the dft calculations of peak positions and eigenvectors served as additional guides in peak assignments .
because the detector banks extend out - of - plane , we were able to determine the c - axis modes as well ( albeit with less precision ) .
figure [ fig : disp ] shows that dft calculations reproduce the phonon dispersions and intensities quite well .
although some small deviations from the calculated dispersions have been observed , they are on the order of 5% , which is typical for the agreement between experiment and dft calculations in general .
this good agreement has been achieved despite the fact that the calculations were carried out without inclusion of magnetism , although the structure was constrained to the experimental structure .
our results for the raman - active modes at the @xmath21-point are in good agreement with previous raman studies @xcite .
we have not been able to establish the extent to which different branches mix with each other when they cross .
such mixing should occur , but the resulting anti - crossing gaps were too small to be observed in our experiment .
the gaps could likely be determined by using a sample which is larger ( for better statistics ) and has a smaller mosaic ( for better resolution ) .
successful mapping of phonon dispersions allowed us to investigate temperature dependence of these phonon branches in order to look for phonon anomalies associated with magnetic ordering .
[ fig : temp ] shows that temperature - dependence of these phonons is weak and is consistent with small changes in lattice constants due to thermal expansion , i.e. these phonons are not particularly sensitive to the magnetic ordering transition .
we note some temperature variation does appear near points 4 and 8 .
however , the polarization in those regions is primarily along the c - axis , and the weak intensity produces larger errors . along with the narrower range of data aquisition for t = 220 k and 320 k , the uncertainty at these points are large enough to account for the variation seen here .
there is no measurable effect of the onset of long range magnetic order on any phonon branch measured so far as a part of this study ( fig .
[ fig : temp ] ) or in previous studies , with the exception of transverse acoustic phonons whose eigenvectors correspond directly to the associated structural deformation @xcite .
it is also possible that phonons couple to the nematic fluctuations found at higher energies @xcite and do not feel the effects of the phase transition .
these fluctuations may vary as a function of doping @xcite , and may in turn influence the phonon spectrum .
the effect of doping on these high - energy phonons can now be investigated using the multizone approach .
one advantage of mpr can be the identification of enhanced electron - phonon coupling at wavevectors away from high symmetry directions that previously would have remained unobserved .
for example , such anomalies have been found in cr @xcite .
we examined temperature - dependence of phonons away from high - symmetry directions in srfe@xmath1as@xmath1 and found that there are no anomalies there either .
we developed a multizone phonon refinement ( mpr ) technique for using time - of - flight neutron chopper spectrometers to extract phonon dispersions even when the phonon frequencies are spaced much more closely than the experimental resolution .
it is based on fitting scattering intensities in multiple bz .
we used this technique to reliably measure closely - spaced phonons above 23 mev in ferropnictide superconductors .
we assigned eigenvectors to these phonons and showed that they do not exhibit any unconventional temperature - dependence related to the magnetic ordering transition .
our results pave the way for future detailed comparison of experimental and calculated phonon frequencies in iron - based superconductors and for future studies of materials with large unit cells where closely spaced phonon branches are expected , such as thermoelectrics and complex oxides .
d.p and d.r . were supported by the doe , office of basic energy sciences , office of science , under contract no .
de - sc0006939 .
the research at ornl s spallation neutron source was sponsored by the scientific user facilities division , office of basic energy sciences , u.s .
department of energy .
26ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1103/physrevlett.101.026403 [ * * , ( ) ] link:\doibase 10.1038/nature07981 [ * * , ( ) ] link:\doibase 10.1143/jpsj.77.083704 [ * * , ( ) ] link:\doibase 10.1088/0953 - 2048/23/5/054013 [ * * , ( ) ] , link:\doibase 10.1038/nphys1160 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.184517 [ * * , ( ) ] link:\doibase 10.1103/physrevb.79.024513 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.037003 [ * * , ( ) ] link:\doibase 10.1155/2010/164916 [ * * , ( ) ] link:\doibase 10.1016/j.physc.2009.11.167 [ * * , ( ) ] link:\doibase 10.1103/physrevb.80.214534 [ * * , ( ) ] link:\doibase 10.1103/physrevb.79.220511 [ * * , ( ) ] link:\doibase 10.1143/jpsj.77.103715 [ * * , ( ) ] link:\doibase 10.1063/1.3680104 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.109.057001 [ * * , ( ) ] @noop `` , '' link:\doibase 10.1103/physrevb.81.060501 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.212502 [ * * , ( ) ] link:\doibase 10.1103/physrevb.60.r3709 [ * * , ( ) ] link:\doibase 10.1088/0953 - 8984/20/45/452201 [ * * , ( ) ] , link:\doibase 10.1103/physrevb.71.092301 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.060503 [ * * , ( ) ] link:\doibase 10.1107/s0108768112008294 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.054544 [ *
* , ( ) ] link:\doibase 10.1038/nphys1512 [ * * , ( ) ] link:\doibase 10.1103/physrevb.82.195121 [ * * , ( ) ] | the ferropnictide superconductors exhibit a sensitive interplay between the lattice and magnetic degrees of freedom , including a number of phonon modes that are much softer than predicted by nonmagnetic calculations using density functional theory ( dft ) .
however , it is not known what effect , if any , the long - range magnetic order has on phonon frequencies above 23 mev , where several phonon branches are very closely spaced in energy and it is challenging to isolate them from each other .
we measured these phonons using inelastic time - of - flight neutron scattering in @xmath0 40 brillouin zones , and developed a technique to determine their frequencies .
we find this method capable of determining phonon energies to @xmath0 0.1 mev accuracy , and that the dft calculations using the experimental structure yield qualitatively correct energies and eigenvectors .
we do not find any effect of the magnetic transition on these phonons . |
common belief holds that only polynomial interactions up to a certain degree depending on the spacetime dimension are renormalizable , in the sense that interactions of even higher order require an infinite number of subtractions in a perturbative analysis .
this can be attributed to the implicit assumption that the higher - order couplings , which in general are dimensionful , set independent scales .
such nonrenormalizable theories can only be defined with a cutoff scale @xmath1 , while the unknown physics beyond the cutoff is encoded in the ( thereby independent ) values of the couplings .
starting from the viewpoint that the cutoff @xmath1 is the only scale in the theory , halpern and huang @xcite pointed out the existence of theories with higher - order and even nonpolynomial interactions within the conventional setting of quantum field theory .
this happens because the higher - order couplings , by assumption , are proportional to a corresponding power of @xmath2 and therefore die out sufficiently fast in the limit @xmath3 ; the theories remain perturbatively renormalizable in the sense that infinitely many subtractions are not required . perhaps most important
, halpern and huang so discovered nonpolynomial scalar theories which are asymptotically free , offering an escape route to the `` problem of triviality '' of standard scalar theories @xcite . to be more precise , halpern and huang analyzed the renormalization group ( rg ) trajectories for the interaction potential in the vicinity of the gaussian fixed point .
the exact form of the potential was left open by using a taylor series expansion in the field as an ansatz . employing the wegner - houghton @xcite ( sharp - cutoff ) formulation of the wilsonian rg , the eigenpotentials ,
i.e. , tangential directions to the rg trajectories at the gaussian fixed point , were identified in linear approximation .
while the standard polynomial interactions turn out to be irrelevant as expected , some nonpolynomial potentials which increase exponentially for strong fields prove to be relevant perturbations at the fixed point .
for the irrelevant interactions , the gaussian fixed point is infrared ( ir ) stable , whereas the relevant ones approach this fixed point in the ultraviolet ( uv ) .
possible applications of these new relevant directions are discussed in @xcite for the higgs model and in @xcite for quintessence .
further nonpolynomial potentials and their applications in higgs and inflationary models have been investigated in @xcite .
considering the complete rg flow of such asymptotically free theories from the uv cutoff @xmath1 down to the infrared , the halpern - huang result teaches us only something about the very beginning of the flow close to the cutoff and thereby close to the gaussian fixed point .
each rg step in a coarse - graining sense `` tends to take us out of the linear region into unknown territory '' @xcite .
it is the purpose of the present work to perform a first reconnaissance of this territory with the aid of the rg flow equations for the `` effective average action '' @xcite . in this framework ,
the standard effective action @xmath4 is considered as the zero - ir - cutoff limit of the effective average action @xmath5 $ ] which is a type of coarse - grained free energy with a variable infrared cutoff at the mass scale @xmath6 .
@xmath7 satisfies an exact renormalization group equation , and interpolates between the classical action @xmath8 and the standard effective action @xmath9 . in this work , we identify the classical action @xmath10 given at the cutoff @xmath1 with a scalar o(@xmath0 ) symmetric theory defined by a standard kinetic term and a generally nonpolynomial potential of halpern - huang type .
therefore , we have the following scenario in mind : at very high energy , the system is at the uv stable gaussian fixed point . as the energy decreases , the system undergoes an ( unspecified ) perturbation which carries it away from the fixed point initially into some tangential direction to one of all possible rg trajectories .
we assume that this perturbation occurs at some scale @xmath1 which then sets the only dimensionful scale of the system .
any other ( dimensionless ) parameter of the system should also be determined at @xmath1 ; for the halpern - huang potentials , there are two additional parameters : one labels the different rg trajectories ; the other specifies the `` distance '' scale along the trajectory .
finally , the precise form of the potential at @xmath1 serves as the boundary condition for the rg flow equation which governs the behavior of the theory at all scales @xmath11 . since the rg flow equations for @xmath7 are equivalent to an infinite number of coupled differential equations of first order , a number of approximations ( truncations ) are necessary to arrive at explicit solutions .
in the present work , we shall determine the rg trajectory @xmath12 for @xmath13 $ ] explicitly only in the large-@xmath0 limit which simplifies the calculations considerably .
the paper is organized as follows : sec .
[ hh ] , besides introducing the notation , briefly rederives the halpern - huang result in the language of the effective average action , generalizing it to a nonvanishing anomalous dimension .
[ largen ] investigates the rg flow equation for the halpern - huang potentials in the large-@xmath0 limit , concentrating on @xmath14 and @xmath15 spacetime dimensions ; here , we emphasize the differences to ordinary @xmath16 theory particularly in regard to mass renormalization and symmetry - breaking properties .
[ conclusions ] summarizes our conclusions and discusses open questions related to finite values of @xmath0 . as an important caveat
, it should be mentioned that the results of halpern and huang have been questioned ( see @xcite and also @xcite ) , and these questions raised also affect the present work . to be honest ,
we have hidden the problems in the `` scenario '' described above in which an `` unspecified '' perturbation controls the shift of the system from the gaussian fixed point ( the continuum limit ) to the cutoff scale @xmath1 along a _
tangential _ direction .
but since the cutoff scale @xmath1 , though large , is not at all infinitesimally separated from the gaussian fixed point , this tangential approximation is probably not sufficient to stay on the true renormalized trajectory during the shift . not only the tangent but also all ( infinitely many ) curvature moments of the trajectory had to be known in order to find an initial point right on the renormalized trajectory at @xmath1 .
this point would correspond to a so - called `` perfect action '' @xcite .
of course , this requires an infinite number of conditions to be imposed on the initial action at @xmath1 which we can not specify . in conventional field theories ,
this problem is solved by adjusting ( fine - tuning ) the initial action close to the unstable gaussian fixed point , leaving open only one a priori chosen relevant direction to the flow .
but in the present case , there is an infinite number of relevant directions corresponding to the continuum of possible halpern - huang directions , and thus it seems impossible to single out only one relevant direction while frustrating the others by tuning infinitely many parameters . in other words , upon studying the flow from @xmath1 down to zero within our scenario , the continuum limit of our system remains unspecified , and therefore one important ingredient to a complete field theoretic system is missing . with these reservations in mind , we nevertheless believe that there are some lessons to be learned from the application of the rg flow equations to such potentials .
concerning the investigation of the rg flow equation for the euclidean effective average action in @xmath17 dimensions , we closely follow the original work of wetterich @xcite .
polynomial potentials and the large-@xmath0 limit to be discussed later have been explored in @xcite and @xcite in the effective average action approach . a comprehensive review and
an extensive list of references on this subject can be found in @xcite .
the effective average action can be expanded in terms of all possible o(@xmath0 ) invariants , @xmath18 = \int d^dx\left\ { u_k(\rho ) + \frac{1}{2 } z_k(\rho)\ , \partial_\mu \phi^b \partial_\mu \phi^b + \frac{1}{4 } y_k(\rho)\ , \partial_\mu \rho \partial_\mu \rho+\dots\right\ } , \label{1}\ ] ] where @xmath19 , @xmath20 labels the real components of the scalar field , and the dots represent terms involving higher derivatives ; for convenience , we shall always assume that @xmath21 during the calculation .
halpern and huang derived their result in the `` local - potential approximation '' which is constituted by setting the wave function renormalization constant @xmath22 and neglecting @xmath23 and higher - derivative terms . in the present work
, we shall generalize their result to a @xmath6-dependent @xmath24 which is parametrized by the anomalous dimension , @xmath25 denotes the derivative with respect to the rg `` time '' , @xmath26-\infty , 0]$ ] . ]
@xmath27 . here
we neglect @xmath23 and any @xmath28 dependence of @xmath24 . following @xcite , the rg flow equation for the effective average potential @xmath29
can be written as @xmath30 where the prime denotes the derivative with respect to the argument @xmath28 .
the cutoff function @xmath31 is to some extent an arbitrary positive function that interpolates between @xmath32 for @xmath33 and @xmath34 for @xmath35 .
it suppresses the small - momentum modes by a mass term @xmath36 acting as the ir cutoff . in eq .
( [ 3 ] ) the distinction between the @xmath37 `` goldstone modes '' and the `` radial mode '' is visible .
provided that @xmath38 is given ( which we shall always assume in the this work ) , the flow of the effective potential @xmath39 ( and thus of the effective action @xmath7 in the present approximation ) is determined by eq .
( [ 3 ] ) . even if @xmath38 is neglected , eq .
( [ 3 ] ) produces qualitatively good results for polynomial effective potentials in @xmath21 @xcite .
we expect similar behavior for nonpolynomial potentials .
the halpern - huang result can be rederived by assuming that the system is close to the gaussian fixed point so that the effective potential and its derivatives are small . linearizing the right - hand side of eq .
( [ 3 ] ) with respect to the potential and its derivatives gives @xmath40 where we introduced the abbreviation @xmath41 , which is related to the volume of @xmath17 spheres .
it is convenient to remove the explicit @xmath24 and @xmath6 dependence by using dimensionless scaling variables : @xmath42 in the same spirit , we write for the cutoff function @xmath43 where @xmath44 is a dimensionless function of a dimensionless argument , satisfying @xmath45 and @xmath46 . rewriting eq .
( [ 4 ] ) in terms of these variables and taking the rg time derivative @xmath47 on the left - hand side at fixed @xmath48 , we obtain the differential equation @xmath49 where the dot denotes a derivative with respect to the argument @xmath48 , and the complete cutoff dependence is contained in @xmath50 .
\label{8}\ ] ] we are looking for eigenpotentials , i.e. , tangential directions to the rg flow of the scaling form @xmath51 , where @xmath52 classifies the possible directions and distinguishes between irrelevant ( @xmath53 ) , marginal ( @xmath54 ) and relevant ( @xmath55 ) perturbations away from the gaussian fixed point . solutions of this form can be given in terms of the kummer function @xmath56 @xcite @xmath57 .
\label{9}\ ] ] for given dimension , @xmath0 , cutoff specification and anomalous dimension , the halpern - huang potential ( [ 9 ] ) depends on two dimensionless parameters : @xmath52 and @xmath58 .
the latter sets a `` distance '' scale along the rg trajectories ; since it is an overall factor , the position of possible extrema of @xmath59 are independent of @xmath58 . to make contact with the literature , we note that we rediscover the results of periwal @xcite in the limit @xmath60 , where the halpern - huang result was generalized to arbitrary cutoffs within the polchinski rg approach @xcite . the results of halpern and huang are recovered by employing a sharp cutoff , for which @xmath61 is related to the volume of the @xmath62 dimensional sphere ) has to be defined carefully ; details can be found in @xcite .
] : @xmath63 various representations for the kummer function @xmath64 exist in the literature @xcite ; for further discussion , it is useful to replace the parameter @xmath52 by the combination @xmath65 then , eq .
( [ 9 ] ) reduces to standard polynomial potentials of degree @xmath66 in @xmath48 ( @xmath67 in @xmath68 ) if @xmath69 ; for all such polynomial potentials , the gaussian fixed point is ir stable . for @xmath70 , the potential vanishes , and for any other value of @xmath71 ,
the potential is nonpolynomial .
for these cases , the asymptotic behavior for large third argument @xmath72 is given by an exponential increase @xmath73 the gaussian fixed point is uv stable ( @xmath55 ) for @xmath74 ( as long as @xmath75 ) .
a particularly interesting case is given by the parameter set @xmath76 for which the eigenpotential eq . ( [ 9 ] ) is nonpolynomial and develops a minimum , inducing spontaneous symmetry breaking . to conclude our derivation of the halpern - huang results
, we mention that in the particular case of @xmath77 there exist ( physically admissible ) solutions to eq .
( [ 7 ] ) which are odd under @xmath78 @xcite .
the linearized flow equation has also been studied from a different perspective employing its similarity to a fokker - planck form @xcite . according to the scenario outlined in the introduction
, we shall now consider the potentials found in eq .
( [ 9 ] ) taken at @xmath79 ( @xmath80 ) as the boundary condition for the complete flow equation ( [ 3 ] ) .
provided that the anomalous dimension @xmath38 is only weakly dependent on @xmath6 and bounded ( as is the case , e.g. , for polynomial interactions in @xmath21 ) , some features can immediately be read off from eq .
( [ 3 ] ) : for nonpolynomial potentials with exponential asymptotics given by eq .
( [ 10 ] ) , the denominators on the right - hand side of the flow equation ( [ 3 ] ) vanish exponentially for large values of @xmath28 . therefore , @xmath81 for large @xmath28 , and the flow halts , leaving @xmath39 essentially unchanged
. in particular , for symmetry - preserving potentials with a minimum at @xmath82 and @xmath83 , we may expect a rather unspectacular flow : for large @xmath28 , the above argument holds , whereas for small @xmath28 , we may always find a small region where the linearization of the flow equation is a good approximation ; there , the halpern - huang potential will still be an appropriate approximation .
therefore , these potentials are expected to behave stiffly under the flow . for potentials with a minimum at nonvanishing @xmath28 ( spontaneous symmetry breaking ) with @xmath84 , the asymptotics for large @xmath28 will also stop the flow
. however , the flow of @xmath39 near the nontrivial minimum can be more complicated , since @xmath85 and @xmath86 are no longer monotonic functions in this region . to the right of the minimum
, these potentials may also be stiff under the flow , but the region around the origin and the minimum appear as a loose end .
these heuristic arguments will be worked out and confirmed in the following section in the large-@xmath0 limit .
for solving flow equations for the effective average potential of the type of eq .
( [ 3 ] ) , several techniques have been developed .
of course , it is always possible to search numerically for solutions by putting the differential equation on a computer ; in fact , if one is looking for accurate results , this is the most appropriate option .
however , since the potentials under consideration exhibit an exponential increase , straightforward numerics may come to its limits and a clever variable substitution has to be guessed .
another possibility is to expand the potential in terms of a complete set of functions and decompose the flow equation into differential equations for the @xmath6-dependent coefficients ( generalized couplings ) . here
, a choice for a useful set of functions again has to be guessed ; obviously , the polynomials as the standard choice are of no use , because the important information is contained in the nonpolynomial nature of the potential .
therefore , we decide to work in the large-@xmath0 limit which puts no a priori restrictions on the form of the potential and allows for a complete integration of the flow equation .
of course , the validity of the results for finite values of @xmath0 can hardly be controlled at this early stage . in the large-@xmath0 limit , the rg flow equation ( [ 3 ] ) for the potential
simplifies considerably ; here we shall follow the presentation given in @xcite and @xcite . not only does the anomalous dimension @xmath38 vanish @xcite , but so does the influence of higher derivative terms ( @xmath87 ) .
moreover , the goldstone modes dominate the right - hand side of eq .
( [ 3 ] ) and any contribution from the radial mode can be neglected ( this essentially changes the order of the differential equation ) . for technical reasons ,
one finally chooses a sharp cutoff function @xmath88 and decides to consider the flow equation for the _ derivative _ of the potential . in dimensionless variables ,
the large-@xmath0 limit of the flow equation reads @xmath89 of course , this equation can be obtained directly from the sharp - cutoff formulation of the rg and has already been studied by wegner and houghton @xcite ; further investigations of the wegner - houghton approach have been made in @xcite .
following @xcite , this partial differential equation of first order can be solved using the standard method of characteristics and we find that the solution @xmath90 has to satisfy the equation @xmath91 where @xmath92 is defined by the integral @xmath93 this function is studied in app .
a and explicit representations for @xmath14 and @xmath15 are given .
the function @xmath94 is implicitly defined by the equation @xmath95 where @xmath96 represents the boundary condition for the flow equation at @xmath80 ( @xmath79 ) ; here , @xmath97 as a variable parametrizes the boundary condition and corresponds to the @xmath48 axis at @xmath79 in the @xmath98 plane .
it is exactly eq .
( [ 18 ] ) that is to be inserted into eq .
( [ 16 ] ) , where the nonpolynomial potentials enter the investigation
. now the route to an explicit solution is clear : ( i ) we specify the boundary condition via eq .
( [ 18 ] ) , ( ii ) insert this and an explicit representation for @xmath92 into eq .
( [ 16 ] ) , and `` solve '' ( or invert ) the resulting equation for @xmath90 . however , in practise , some complications are encountered : e.g. , inverse functions of such complicated objects as the kummer function @xmath56 are not easily obtainable .
but the large-@xmath0 limit comes to the rescue once more as demonstrated in the next subsection .
let us finally extract the flow of a possible minimum of the potential which is defined by @xmath99 ; from eq .
( [ 16 ] ) , we can easily extract that @xmath100 where @xmath101 denotes the minimum of the potential at @xmath80 ( t=0 ) , i.e. , the minimum of the halpern - huang potential ( in the large-@xmath0 limit ) ; by construction , it is identical to @xmath102 .
the function @xmath103 can be read off from eqs .
( [ a3 ] ) and ( [ a4 ] ) of the appendix ( @xmath104 ) . reinstating dimensionful quantities ( cf .
( [ 5 ] ) ) , we find for the flow of a minimum of the potential @xmath105 here , we introduced a `` critical '' ( dimensionless ) field strength @xmath106 .
of course , eq . (
[ 20 ] ) is well known in the literature @xcite and makes no particular reference to the type of potential under consideration .
the only place where the potential type enters is the position of the initial minimum @xmath107 . if @xmath108 , then the classical as well as the quantum theory exhibit spontaneous symmetry breaking , since @xmath109 ; if @xmath110 the quantum theory will preserve o(@xmath0 ) symmetry .
finally , if @xmath111 the classical potential @xmath112 is `` fine - tuned '' in such a way that the theory shows symmetry breaking for finite values of @xmath6 , but restores o(@xmath0 ) symmetry in the limit @xmath113 ; additionally , the potential has a vanishing mass term : @xmath114 ( by construction ) . in standard @xmath16 theory ,
the position of the minimum @xmath107 of @xmath112 can be chosen at will by an appropriate tuning of the negative mass term and the coupling .
by contrast , for halpern - huang potentials , once the precise type of the potential is chosen by fixing @xmath52 ( or @xmath71 ) , there is no parameter left for any fine - tuning , since a possible minimum ( for theories with @xmath84 ) is independent of the last free parameter @xmath58 in eq .
( [ 9 ] ) .
the question as to whether a symmetry - breaking quantum theory of the halpern - huang potentials exists has to be answered by determining the position of the initial minimum @xmath107 .
this will also be investigated in the next subsection in the large-@xmath0 limit .
in our scenario , the halpern - huang potential enters the flow equation as its boundary condition at the cutoff .
since the flow equation is considered in the large-@xmath0 limit , it is not only useful to insert the large-@xmath0 limit of the halpern - huang potential into eq .
( [ 16 ] ) , but it is mandatory for reasons of consistency .
otherwise , nonleading large-@xmath0 information would be mixed with large-@xmath0 behavior , introducing some arbitrariness into this approximation . from eq .
( [ 9 ] ) , we read off that the parameter @xmath0 occurs only in the second argument of the kummer function . unfortunately
, we could not find any asymptotic expression for the kummer function with large second argument in the literature .
instead of investigating this limit in terms of some appropriate series or integral representation , which might involve awkward interchanges of limiting processes , we shall use a more physically motivated approach : since the kummer function was identified with the tangential direction to the rg flow in the vicinity of the gaussian fixed point , its large-@xmath0 approximation should naturally be deducible from a large-@xmath0 study of the same subject . in other words ,
the desired function has to be a solution to the linearized flow equation in the large-@xmath0 limit . for technical reasons
, we again turn to the derivative of the potential with respect to @xmath48 . then
, the desired differential equation is obtained by linearizing eq .
( [ 15 ] ) .
looking for potentials which satisfy the eigenpotential scaling condition @xmath115 , the large-@xmath0 halpern - huang equation reads @xmath116 where we again traded @xmath52 for the parameter @xmath71 as defined in eq .
( [ 12 ] ) . additionally , we made use of the `` critical '' field strength @xmath106 defined in eq .
( [ 20 ] ) .
equation ( [ 21 ] ) can easily be solved for the various boundary conditions ; let us begin with a symmetry - preserving potential satisfying @xmath117 and @xmath118 : @xmath119 where the initial value @xmath120 also satisfies @xmath121 .
the parameter @xmath120 is , of course , the large-@xmath0 analogue of the distance parameter @xmath58 in eq .
( [ 9 ] ) , @xmath122 . at first sight ,
one may doubt the validity of eq .
( [ 22 ] ) as the large-@xmath0 limit of eq . (
[ 9 ] ) , because it diverges at the critical field strength @xmath123 . nevertheless , this indeed reflects the behavior of the kummer function for large second argument , as can be checked numerically ( see fig .
[ fig1](a ) ) @xcite .
the critical field strength @xmath106 marks the point where the asymptotic exponential increase ( cf .
( [ 10 ] ) ) sets in ; and for larger values of @xmath0 , the slope increases without bound limit of the kummer function in the literature : it can not be defined for arbitrary values of @xmath48 . ] .
of course , a potential that diverges for finite values of its argument is usually considered as inadmissible in field theory ( see , e.g. , @xcite and @xcite ) ; however , in the present case , we take the viewpoint that this potential wall at @xmath124 only symbolizes the exponential increase of the potential for finite values of @xmath0 . ( 145,52 ) ( 0,0 ) ( 0,52)(a ) ( 49,18)@xmath0=10 ( 61,13)100 ( 77,40)1000 ( 64,45)10000 ( 7,49)@xmath90 ( 80,3)@xmath125 ( 82,0 ) ( 82,52)(b ) ( 150,7)10 ( 148,40)100 ( 134,40)1000 ( 110,40)@xmath0=10000 ( 92,46)@xmath90 ( 156,20)@xmath125 let us now turn to the solution of eq .
( [ 21 ] ) for the symmetry - breaking potentials with ( @xmath84 ) . in the inner region of the potential where @xmath126
, we obtain the solution @xmath127 where the initial value this time satisfies @xmath128 .
the derivative of the large-@xmath0 halpern - huang potential has a zero at @xmath124 , so that the potential itself exhibits a minimum at this position .
now let us turn to the large-@xmath0 limit of the symmetry - breaking potential to the right of the minimum @xmath129 where @xmath130 . as a matter of fact ,
the unique solution of eq .
( [ 21 ] ) increases only very slowly , @xmath131 for @xmath132 and @xmath84 .
this does certainly not reflect the expected exponential increase ; hence this solution has to be discarded .
even if a more appropriate solution existed for @xmath129 , we would not be able to match them properly at @xmath124 , because the second derivative of the potential diverges at this point . in view of the results for the symmetry - preserving potential and owing to the fact that the asymptotic exponential increase for both types of potentials is the same ( see eq .
( [ 10 ] ) ) , the only possibility for the large-@xmath0 limit is to continue the potential at @xmath124 by a potential wall at @xmath133 . again , a numerical analysis of the full kummer function for large @xmath0 confirms this conjecture , as is depicted in fig .
[ fig1](b ) .
this concludes our large-@xmath0 analysis of the halpern - huang potential ; note that both types of the potential are formally equivalent , so that we combine them in the notation @xmath134 where @xmath135 stands for @xmath120 or @xmath136 and @xmath137 .
actually , this also covers a @xmath16 potential for @xmath138 .
now we are in a position to study the flow of the halpern - huang potentials from @xmath80 into the infrared regime @xmath113 in the large-@xmath0 limit . the missing piece of information to be inserted into eq .
( [ 16 ] ) is given by the inverse of eq .
( [ 24 ] ) : @xmath139 . \label{25}\ ] ] employing the representation ( [ a3 ] ) for the function @xmath92 , we find that the derivative of the potential has to satisfy the equation @xmath140 where the function @xmath141 is defined in eq .
( [ a5 ] ) . finally , eq . ( [ 26 ] ) has to be solved for @xmath90 , which we shall do in the limit @xmath142 ( @xmath143 ) for various cases in order to obtain the complete quantum effective potential .
the following consideration will serve as a guide to the necessary approximations : at the cutoff @xmath80 , the dimensionful potential is of the order of the cutoff @xmath144 .
for small deviations from the cutoff , @xmath145 , the potential scales according to the linearized flow equation ( halpern - huang equation ) : @xmath146 .
then , the dimensionful potential scales as @xmath147 . therefore , if @xmath118 ( symmetry - preserving potentials ) , @xmath85 increases as we approach the infrared , whereas if @xmath148 ( symmetry - breaking potentials ) , @xmath85 decreases towards the infrared
. of course , this argument holds strictly close to @xmath149 only , but it turns out to reproduce the unique consistent approximation schemes for extracting analytical results .
let us first consider the @xmath15 potentials with @xmath118 and @xmath150 that exhibit no symmetry breaking at the cutoff .
employing eq .
( [ a6 ] ) and reinstating dimensionful quantities via eq .
( [ 5 ] ) , eq . ( [ 26 ] ) reads , after neglecting terms of order @xmath151 in the limit @xmath113 ( @xmath152 ) : @xmath153 where @xmath154 denotes the mass of the theory at the cutoff .
let us study eq .
( [ 27 ] ) in two limits : first , at @xmath28 close to @xmath155 , and secondly at the origin @xmath156 . at @xmath28 close to the potential wall at @xmath155 ,
the potential diverges , and we can approximate @xmath157 , leading us to @xmath158 in this limit , the effective potential @xmath159 remains formally identical to the large-@xmath0 halpern - huang potential ( cf .
( [ 22 ] ) ) !
this confirms our heuristic argument that the potential behaves stiffly under the flow in the region where it increases exponentially .
concerning the opposite limit @xmath156 , there would be no mass renormalization at all , if eq .
( [ 28 ] ) were also correct in this limit , @xmath160 .
however , in this limit , the approximation @xmath161 no longer holds , and instead we deduce from eq .
( [ 27 ] ) the transcendental equation @xmath162 therefore , the mass renormalization is governed by the only free parameter of the theory , @xmath120 : for large @xmath120 , there is effectively no renormalization , whereas the renormalized mass @xmath163 exceeds the `` classical '' mass @xmath164 for @xmath165 .
typical values are @xmath166 for @xmath167 and @xmath168 ; for larger values of the rg trajectory parameter @xmath71 , the mass shift even increases : @xmath169 for @xmath167 and @xmath170 .
the @xmath171 relation is plotted against @xmath120 for various @xmath71 in fig . [ fig3](a ) . by reintroducing the cutoff again via @xmath172 ,
( [ 29 ] ) can be interpreted differently by writing @xmath173^a .
\label{30}\ ] ] this equation tells us that the physical mass of the theory in the infrared can easily be much smaller than the cutoff by tenth of orders of magnitude , provided that @xmath120 is correspondingly small .
since @xmath120 sets the distance scale on the rg trajectory , the demand for a small value of @xmath120 is consistent with our scenario : if we leave the gaussian fixed point with a very tiny perturbation @xmath174 at the high - energy scale @xmath1 , it is only _ natural _ to arrive at a low - energy theory with a similarly tiny mass compared to the cutoff .
moreover , consistency of our scenario requires @xmath120 to be small in order to justify the linearization of the flow equation in deriving the halpern - huang result .
to summarize , the symmetry - preserving halpern - huang potential qualitative does not change its form during the flow into the infrared ; in particular , no symmetry breaking occurs .
only the slope of the potential at the origin of the theory increases for @xmath142 , which corresponds to a mass renormalization .
let us begin with a dimension - independent statement referring to the position of the minimum of symmetry - breaking halpern - huang potentials with @xmath84 : in subsec .
[ lnhh ] we learned that the position of the minimum of the halpern - huang potentials in the large-@xmath0 limit is independent of the parameters @xmath71 and @xmath175 : @xmath176 , or in dimensionful quantities : @xmath177 .
according to the discussion following eq .
( [ 20 ] ) , the halpern - huang potentials are `` fine - tuned '' in the sense that the minimum vanishes exactly in the infrared limit @xmath113 : @xmath178 therefore , there is no symmetry breaking in the full quantum theory of halpern - huang potentials in the large-@xmath0 limit .
moreover , since @xmath179 , the potential is flat at the origin and the renormalized quantum theory is massless .
following the line of argument given below eq .
( [ 26 ] ) , the inner region of the potential where @xmath180 decreases towards the infrared ; hence we approximate @xmath181 in eq .
( [ 26 ] ) and obtain the transcendental equation in @xmath15 : @xmath182 here we can read off that @xmath183 is always smaller than @xmath36 .
this reflects the approach to convexity of the inner part of the effective potential . to summarize , we have found , on the one hand , that the originally nontrivial minimum of the potential moves to the origin during the flow ; the inner region of the potential shrinks to a point . on the other hand , we know from the preceding subsection that the potential wall at @xmath184 does not change its position under the flow .
it remains to be investigated what happens in between the minimum and the potential wall .
unfortunately , we can not answer this question by the large-@xmath0 version of the flow equation , because we do not have a boundary condition for this region . at the cutoff @xmath80 , the inner region borders directly at the potential wall ; hence , there is no `` in - between '' that could serve as a boundary condition .
of course , it is plausible to assume that the potential may interpolate smoothly between the origin with zero slope and the potential wall at @xmath185 with infinite slope .
but alternatively , the potential can also remain flat for @xmath186 $ ] , resembling a particle - in - a - box potential .
our ignorance about that part of the potential is unfortunately accompanied by our inability to predict the mass of the radial mode ; but this should not come as a surprise , since the large-@xmath0 limit neglects the radial mode anyway .
( 145,52 ) ( 0,0 ) ( 0,52)(a ) : @xmath15 ( 35,50)@xmath187 ( 61,-2)@xmath188 ( 2,10)@xmath189 ( 5,27)@xmath168 ( 12,40)@xmath170 ( 84,0 ) ( 84,52)(b ) : @xmath14 ( 147,-2)@xmath188 ( 127,50)@xmath187 ( 86,9)@xmath189 ( 87,25)@xmath168 ( 101,38)@xmath170 the investigation of the various types of potentials in @xmath14 proceeds analogously to the @xmath15 case with almost identical results . in particular , the symmetry - breaking potentials offer no new information : the inner region shrinks to a point , while the potential minimum moves to zero for @xmath113 , and the potential wall remains at @xmath190 . in between , no confirmed statement can be made within the large-@xmath0 limit , since no boundary condition governs this part of the potential . for symmetry - preserving potentials with @xmath118 , the potential again remains in the same form as at the cutoff for values of @xmath28 close to the potential wall at @xmath191 ( cf .
( [ 28 ] ) with @xmath192 replaced by @xmath1 ) .
close to the origin @xmath156 , the shape of the potential is modified ; this is reflected by a mass renormalization . employing the same line of argument as given above in @xmath15 , and using eq .
( [ a7 ] ) , we find the @xmath14 analogue of eq .
( [ 29 ] ) : @xmath193 again , we find that there is no mass renormalization for large values of @xmath120 ; corrections for small values of @xmath120 are plotted in fig [ fig3](b ) .
in the present paper , we have investigated the rg flow of particular nonpolynomial potentials for o(@xmath0 ) symmetric scalar theories using the effective - average - action method .
these halpern - huang potentials arise from small relevant perturbations at the gaussian fixed point as tangential directions to the rg flow . apart from serious , unresolved problems with the continuum limit of these potentials
, we were able to follow the flow from a given ultraviolet scale @xmath1 down to the nonperturbative infrared ; for this , a number of approximations have been made which are only under limited control . in a first step ,
we have neglected the influence of possible derivative couplings on the flow of the potential . secondly
, assuming that the anomalous dimension is only weakly dependent on @xmath6 and bounded , the qualitative features of the flow could already be guessed from the form of the flow equation : this is because the exponential increase of the potentials essentially causes the flow to stop for large enough field values .
therefore , the form of the potentials was recognized as stiff under the flow ; only the loose ends of the potential near the origin or possible extrema make room for more diversified behavior .
these considerations have been verified explicitly in the large-@xmath0 limit of the system . in this limit , the exponential increase of the potentials is represented by a potential wall .
the potential close to the wall and the wall itself remain unchanged even in the far infrared .
those potentials with an o(@xmath0 ) symmetric ground state ( @xmath118 ) at the cutoff preserve this symmetry down to @xmath113 .
our main result for such potentials is summarized in eqs .
( [ 29 ] ) , ( [ 30 ] ) and ( [ 33 ] ) , where the particular form of the mass renormalization is stated .
contrary to polynomial scalar interactions where the mass varies @xmath194 during the flow , the halpern - huang potentials exhibit corrections which are governed by the rg distance parameter @xmath120 .
in particular , if one demands that a renormalized ( infrared ) mass differ by several orders of magnitude from the cutoff scale @xmath1 , the bare parameters of a _ polynomial _ theory at the cutoff scale have to be fine - tuned accurately to several decimal places .
by contrast , to achieve such a separation of mass scales with a _ nonpolynomial _ halpern - huang potential , an adjustment of the rg distance parameter at the cutoff to some small value is required with much less precision .
additionally , the smallness of this value arises naturally , if the ( unknown ) perturbation at the gaussian fixed point is tiny .
the ( symmetry - preserving ) halpern - huang potentials thus has no problem of _
naturalness_. owing to the general properties of the complete flow equation mentioned above , we believe that these properties of the symmetry - preserving potentials in the large-@xmath0 limit also hold for finite values of @xmath0 . the status of the large-@xmath0 limit is certainly different for halpern - huang potentials which offer spontaneous symmetry breaking ( @xmath84 ) . these potentials exhibit the remarkable property that the nontrivial minimum persists for any finite value of @xmath6 but vanishes in the complete quantum theory for @xmath142 in the large-@xmath0 limit ; the o(@xmath0 ) symmetry is restored and the potential becomes flat near the origin .
the coincidence between the position of the minimum and the critical value of the field strength may finally be ascribed to the formal resemblance between the large-@xmath0 flow equation and its linearized version determining the halpern - huang potentials .
since the complete flow equation is much more complex , it appears rather improbable that this property continues to hold for finite @xmath0 .
therefore , whether or not spontaneous symmetry breaking occurs in the quantum version of the halpern - huang potential at finite @xmath0 remains an open question .
the present investigation at least observes a tendency of the system to restore o(@xmath0 ) symmetry .
this is in concordance with @xcite , where a one - loop calculation for the effective potential reveals a restoration of o(@xmath0 ) symmetry for potentials with ( @xmath195 ) . in this context , a possible application of the halpern - huang potentials to the higgs sector of the standard model is still questionable . even if a quantum version of the potential with spontaneous symmetry breaking exists , the naturalness of the scalar sector alone is not sufficient to solve the hierarchy problem
this is because the ( standard ) yukawa coupling to the fermions leads to large scalar mass renormalizations by fermion loops .
therefore , some appropriate nonpolynomial interaction has to be chosen , also in this sector .
nevertheless , the price to be paid would not be too high , because not only the hierarchy problem could be circumvented without additional degrees of freedom , but also the problem of `` triviality '' would be evaded . from an intuitive point of view
, the fact that the form of the potential is stable under the rg flow appears to be disappointing : since the potential remains inherently nonpolynomial , it is impossible to make contact with a would - be classical behavior that is determined by only a few ( polynomial ) terms .
the latter is usually expected at large distances .
for example , merely for very weak fields do the first terms in a taylor expansion of the kummer function represent a good approximation . for stronger fields , the application of the halpern - huang potentials
might therefore be limited in this sense . from a technical viewpoint ,
our calculations hold for @xmath21 .
we have given explicit results for @xmath14 and @xmath15 , and generalizations to higher dimensions are straightforward .
the limiting case @xmath196 has to be treated with great care for several reasons .
first of all , finite @xmath0 results may only be trusted if the flow of the anomalous dimension @xmath38 is taken into account ; at least in the case of polynomial potentials , this turned out to be obligatory @xcite in order to obtain a good picture of the kosterlitz - thouless transition .
furthermore , the limit @xmath197 of the halpern - huang potentials offers several possibilities .
it has already been observed variously in the literature ( see , e.g. , @xcite ) , that the sine - gordon as well as the liouville potentials solve the linearized flow equation in @xmath196 .
in fact , as can be easily shown with the aid of some identities of @xcite , both types of potentials arise as limiting cases of the halpern - huang potentials for @xmath77 in combination with the @xmath198 odd solution of the linearized flow equation : to be precise , the sine - gordon potential is recovered in the limit @xmath199 for @xmath200 , whereas the liouville potential is obtained by taking the limit @xmath201 for @xmath202 .
as far as the liouville theory is concerned , further similarities to the present results for the symmetry - preserving potentials are visible . in @xcite ,
the liouville potential has also been found to behave stiffly under the rg flow for similar reasons as in the present case . in particular , quantum liouville theory appears to equal classical liouville theory , except for a flow of the central charge by one unit and a modified mass parameter .
these similarities confirm the viewpoint that the halpern - huang potentials can be regarded as higher - dimensional analogues of liouville theory .
the author wishes to thank w. dittrich for helpful conversations and for carefully reading the manuscript .
useful discussions with r. shaisultanov are also gratefully acknowledged .
in this appendix , we present some details about the function @xmath92 appearing in the solution ( [ 16 ] ) to the flow equation ( [ 15 ] ) ; this function is defined as @xmath203 substituting @xmath204 $ ] , we arrive at the form @xmath205 where @xmath27 is always nonpositive : @xmath206-\infty , 0]$ ] . separating the zeroth - order term of a taylor expansion of the integrand , we find the convenient representation @xmath207 with the auxiliary functions @xmath208 and @xmath141 defined by @xmath209 note that @xmath210 for @xmath211 and @xmath212 .
the explicit form of @xmath213 depends on the spacetime dimension . for @xmath15 , the integral
can easily be evaluated by standard means , yielding @xmath214 in @xmath14 , we take care of the possibility of a nontrivial minimum ( spontaneous symmetry breaking ) and find to the right of a possible minimum @xmath215 in the `` inner '' region to the left of a possible minimum , we obtain @xmath216 where @xmath212 for reasons of consistency .
99 k. halpern and k. huang , phys . rev .
lett . * 74 * , 3526 ( 1995 ) .
k. halpern and k. huang , phys .
d * 53 * , 3252 ( 1996 ) .
a.i . larkin and d.e .
khumelnitskii , sov .
j. nucl .
* 29 * , 1123 ( 1969 ) ; k.g .
wilson , phys .
* 28 * , 248 ( 1972 ) .
f.j . wegner and a. houghton ,
rev . a * 8
* , 401 ( 1973 ) .
v. branchina , hep - ph/0002013 ( 2000 ) .
k. langfeld and h. reinhardt , mod .
a * 13 * , 2495 ( 1998 ) ; r.f .
langbein , k. langfeld , h. reinhardt , l. v. smekal , mod .
a * 11 * , 631 ( 1996 ) . c. wetterich , phys .
b * 301 * , 90 ( 1993 ) .
morris , phys .
* 77 * , 1658 ( 1996 ) ; k. halpern and k. huang , phys . rev .
lett . * 77 * , 1659 ( 1996 ) . c. bagnuls and c. bervillier , hep - th/0002034 ( 2000 ) .
n. tetradis and c. wetterich , nucl .
b * 422 * , 541 ( 1994 ) .
p. hasenfratz , in proc .
_ advanced school of non - perturbative quantum field physics _ , edited by m. asorey and a. dobado , singapore , world scientific , 1998 , hep - lat/9803027 .
n. tetradis and d. f. litim , nucl .
b * 464 * , 492 ( 1996 ) [ hep - th/9512073 ] .
j. berges , n. tetradis and c. wetterich , hd - thep-00 - 26 , hep - ph/0005122 ( 2000 ) .
v. periwal , mod .
a * 11 * , 2915 ( 1996 ) .
j. polchinski , nucl .
b * 231 * , 269 ( 1984 ) .
m. abramowitz and i.a .
stegun , _ handbook of mathematical functions _ , national bureau of standards , washington , ( 1964 ) .
a. bonanno , phys .
d * 62 * , 027701 ( 2000 ) .
j. zinn - justin , _ quantum field theory and critical phenomena _ , oxford university press , ( 1989 ) .
j. comellas and a. travesset , nucl .
b498 , 539 ( 1997 ) .
the kummer functions ( hypergeometric1f1[a , b , z ] ) can numerically as well as partly algebraically be treated by mathematica , version 4.0.1.0 , wolfram research , champaign ( 1999 ) .
a. hasenfratz and p. hasenfratz , nucl .
phys . * b270 * , 687 ( 1986 ) . k. halpern , phys . rev .
d * 57 * , 6337 ( 1998 ) .
m. reuter and c. wetterich , nucl .
b * 506 * , 483 ( 1997 ) . | a class of asymptotically free scalar theories with o(@xmath0 ) symmetry , defined via the eigenpotentials of the gaussian fixed point ( halpern - huang directions ) , are investigated using renormalization group flow equations .
explicit solutions for the form of the potential in the nonperturbative infrared domain are found in the large-@xmath0 limit . in this limit ,
potentials without symmetry breaking essentially preserve their shape and undergo a mass renormalization which is governed only by the renormalization group distance parameter ; as a consequence , these scalar theories do not have a problem of naturalness .
symmetry - breaking potentials are found to be `` fine - tuned '' in the large-@xmath0 limit in the sense that the nontrivial minimum vanishes exactly in the limit of vanishing infrared cutoff : therefore , the o(@xmath0 ) symmetry is restored in the quantum theory and the potential becomes flat near the origin . |
consider an equity market with asset capitalizations @xmath0 at time @xmath1 , and with covariance and relative risk rates @xmath2 and @xmath3 , respectively . at any given time @xmath4 , these rates are _ nonanticipative functionals _ of past - and - present capitalizations @xmath5 , @xmath6 ; they are not specified with precision but are , rather , subject to `` knightian uncertainty . '' to wit , for a given collection @xmath7 of nonempty compact and convex subsets on @xmath8 , where @xmath9 is the space of real , symmetric , positive definite @xmath10 matrices , and @xmath11 is the origin in @xmath12 , they are subject to the constraint @xmath13 in other words , the pair @xmath14 must take values at time @xmath4 inside the compact , convex set @xmath15 which is determined by the current location of the asset capitalization process ; but within this range , the actual value @xmath16 is allowed to depend on past capitalizations as well .
[ to put it a little differently : the constraint ( [ e.3 ] ) is not necessarily `` markovian , '' as long as the sets in ( [ 1.b ] ) are not singletons . ] under these circumstances , what is the highest return on investment relative to the market that can be achieved using nonanticipative investment rules , and with probability one under all possible market model configurations that satisfy the constraints of ( [ e.3 ] ) ?
what are the weights in the various assets of an investment rule that accomplishes this ? answers : subject to appropriate conditions , @xmath17 and @xmath18\\[-8pt ] \eqntext{i = 1 , \ldots , n , 0 \le t \le t,}\end{aligned}\ ] ] respectively . here
the function @xmath19 $ ] is the smallest nonnegative solution , in the class @xmath20 , of the fully nonlinear parabolic partial differential inequality @xmath21 subject to the initial condition @xmath22 , with @xmath23 we use in ( [ fg ] ) , ( [ e.3.d ] ) and throughout this paper , the notation @xmath24 , @xmath25 , @xmath26 , @xmath27 and define @xmath28 we call the function @xmath29 the _ arbitrage function , _ as @xmath30 gives the smallest initial capital starting with which an investor , who uses nonanticipative investment rules , can match or outperform the market portfolio by time @xmath31 , if the initial configuration of asset capitalizations is @xmath32 at @xmath33 , and does so with probability one under any `` admissible '' market configuration that might materialize
. it is perhaps worth noting that this function @xmath29 is characterized almost entirely in terms of the prevalent covariance structure @xmath34 .
the relative risk @xmath35 enters only indirectly , namely , in determining the family of sets ( [ e.3.a ] ) which are admissible for the covariance structure .
put a bit differently , the only role @xmath35 plays is to ensure the asset capitalization process @xmath36 takes values in @xmath37 . under additional regularity conditions ,
@xmath38 is the value of a stochastic control problem : the maximal probability that the diffusion process @xmath39 with initial configuration @xmath40 , values in the punctured nonnegative orthant @xmath41 of ( [ 1.b ] ) , infinitesimal generator @xmath42 and controlled through the choice of covariance function @xmath43 which satisfies @xmath44 for all @xmath45 , does not hit the boundary of the orthant @xmath46 by time @xmath47 . under appropriate conditions
the function @xmath29 satisfies then , in the notation of ( [ e.3.d ] ) , the _ hamilton jacobi bellman _ ( _ hjb _ ) _ equation _ @xmath48 _ relation to extant work _ : stochastic control problems of the `` maximal probability of containment '' type were apparently pioneered by van mellaert and dorato ( @xcite ) ( see also @xcite , pages 157158 ) .
the `` knightian uncertainty '' constraint imposed in ( [ e.3 ] ) is very similar to the formulation of stochastic control and stochastic game problems for one - dimensional diffusions pioneered by william sudderth , that sudderth and his collaborators developed in a series of articles that includes @xcite , heath et al .
( @xcite ) , @xcite , @xcite ; indeed , the developments in sections [ sec6][sec8 ] of our paper can be construed as a multidimensional extension of the sudderth approach .
we rely strongly on krylov s ( @xcite , @xcite ) work , which studies solutions of stochastic differential equations with constraints on the drift and diffusion coefficients in terms of `` supermartingale problems '' and characterizes sets of stochastic integrals via appropriate supermartingales .
the approach we adopt has a lot in common with the effort , started in the mid-1990s , to understand option pricing and hedging in the presence of uncertainty about the underlying volatility structure of assets .
we have been influenced by this strand of work , particularly by the papers of @xcite , @xcite , @xcite , @xcite , @xcite ; other important papers include @xcite , @xcite , @xcite , @xcite , ekstrm and tysk ( @xcite ) , @xcite , @xcite , whereas the recent preprints by @xcite ( @xcite , @xcite ) contain very relevant results .
similar in this spirit is the strand of work by shige peng and his collaborators , surveyed in @xcite , regarding the so - called `` @xmath49-brownian motion '' which exhibits volatility uncertainty [ see also @xcite , as well as @xcite for extensions to settings where the range of uncertainty is stochastic ] .
whereas in both these strands the relevant fully nonlinear parabolic - type partial differential ( so - called `` black scholes barenblatt'')equation has a typically unique solution , here the main interest arises from _ lack of uniqueness _ on the part of the rather similar , fully nonlinear equation ( [ e.3.c ] ) .
let us mention that optimization problems in stochastic control , mathematical economics and finance that involve model uncertainty have also been treated by other authors , among them @xcite , @xcite , shied and wu ( @xcite ) , @xcite , fllmer and gundel ( @xcite ) , @xcite , @xcite , bayraktar and yao ( @xcite ) , bayraktar , karatzas and yao ( @xcite ) and @xcite [ see also the survey by @xcite ] . _ preview _ : sections [ sec2 ] and [ sec3 ] set up the model for an equity market with knightian model uncertainty regarding its volatility and market - price - of - risk characteristics , and for investment rules in its context .
section [ sec4 ] introduces the notion of optimal arbitrage in this context , whereas section [ sec5 ] discusses the relevance of the fully nonlinear parabolic partial differential inequality of hjb type ( [ e.3.in ] ) , ( [ e.3.d ] ) in characterizing the arbitrage function and in finding an investment rule that realizes the best outperformance of the market portfolio .
section [ sec6 ] presents a verification - type result for this equation .
sections [ sec7 ] and [ sec8 ] make the connection with the stochastic control problem of maximizing the probability of containment for an auxiliary it process , controlled in a nonanticipative way and in a manner that affects both its drift and dispersion characteristics .
finally , section [ sec9 ] develops yet another characterization of the arbitrage function , this time as the min - max value of a zero - sum stochastic game ; the investment rule that realizes the best outperformance of ( optimal arbitrage with respect to ) the market , is now seen as the investor s best response to a `` least favorable '' market model configuration .
we shall fix throughout a canonical , filtered measurable space @xmath50 , @xmath51 and assume that @xmath52 contains the space @xmath53 of all continuous functions @xmath54 .
we shall specify this canonical space in more detail in section [ sec7 ] below , when such detail becomes necessary . on this space
, we shall consider a vector of continuous , adapted processes @xmath55 with values in @xmath37 ; its components will represent stock capitalizations in an equity market with @xmath56 assets , and thus the total market capitalization will be the sum @xmath57 we shall also fix throughout a collection @xmath58 of nonempty , compact and convex subsets of @xmath8 as in ( [ 1.b ] ) .
we shall consider @xmath59-valued functionals @xmath60 and @xmath61-valued functionals @xmath62 , all of them defined on @xmath63 and progressively measurable [ see @xcite , definition 3.5.15 ] .
we shall assume that , for every continuous function @xmath64 and @xmath65 , these functionals satisfy the constraint and integrability conditions , respectively , @xmath66\\[-8pt ] \int_0^t \bigl ( \| \vartheta(t , \mathfrak{w } ) \|^2 + \operatorname{tr } ( \alpha(t , \mathfrak{w } ) ) \bigr ) \,\mathrm{d } t & < & \infty.\nonumber\end{aligned}\ ] ] we shall also consider @xmath67-matrix - valued functionals @xmath68 where @xmath69 for a given collection of sets @xmath70 as in ( [ 1.b ] ) and a fixed initial configuration @xmath71 of asset capitalizations , we shall call _ admissible system _ a collection @xmath72 consisting of the underlying filtered space @xmath73 , @xmath74 , of a probability measure @xmath75 on it , and of a pair of processes @xmath76 , with @xmath77 an @xmath78-dimensional @xmath79-brownian motion and @xmath80 taking values in @xmath37 .
these processes have the dynamics @xmath81 , \qquad x_i(0)=x_i>0\ ] ] for some progressively measurable functionals @xmath82 and @xmath83 which satisfy ( [ 1.2 ] ) and ( [ alphasigma ] ) above .
we shall think of this admissible system @xmath72 as a _ model _ subject to `` knightian '' uncertainty ; this is expressed by the requirement @xmath84 in ( [ 1.2 ] ) , ( [ 1.b ] ) about the market price of risk and the covariance structure of the asset capitalization vector process @xmath85 . in order not to lose sight of the underlying probability space , we shall denote by @xmath86 ( resp .
, @xmath87 ) the probability measure ( resp . , the corresponding expectation operator ) on this space . finally , @xmath88 will denote the collection of all such admissible systems or `` models '' with initial configuration @xmath89 .
we shall think of @xmath90 as a _ meta - model _
, a collection of admissible models , and of the collection @xmath91 as a `` family of meta - models . ''
the interpretation is that the components of the driving brownian motion @xmath92 represent the independent factors of the resulting model ; the entries of the matrix @xmath93 are the local volatility rates of the asset capitalization vector process @xmath85 at time @xmath94 ; the entries of the matrix @xmath95 as in ( [ alphasigma ] ) represent the local covariance rates ; whereas the components of the vector @xmath96 are the market price of risk ( also called _ relative risk _ ) rates prevalent at time @xmath4 . in particular
, @xmath97 is , in the notation of ( [ alphasigma ] ) , the vector of mean rates of return for the various assets at time @xmath4 , when the equations of ( [ 1.1 ] ) are cast in the more familiar form @xmath98 the integrability condition of ( [ 1.2 ] ) guarantees that the process @xmath85 takes values in @xmath37 , @xmath99-a.s . ; it implies also that the exponential process @xmath100 is well defined and a strictly positive local martingale , thus also a supermartingale .
this process @xmath101 plays the role of a state - price - density or `` deflator '' in the present context .
just as in our earlier works , fernholz and karatzas ( @xcite , @xcite ) as well as @xcite mostly in a markovian context , and without model uncertainty an important feature of this subject is that @xmath101 has to be allowed to be a _
strict _ local martingale , that is , that @xmath102 be allowed to hold for some , if not all , @xmath103 . constraint ( [ e.3 ] ) brings us in the realm of the krylov ( @xcite , @xcite ) approach , which studies stochastic differential equations with constraints on the drift and diffusion coefficients in terms of `` supermartingale problems . ''
in particular , theorem 2.2 of @xcite shows that , under a suitable regularity condition on the family of sets @xmath70 in ( [ 1.b ] ) , solving stochastic equation ( [ 1.1 ] ) subject to the requirements of ( [ 1.2 ] ) can be cast as a _ supermartingale problem , _ as follows .
consider the nonlinear partial differential operator associated with ( [ b1 ] ) , ( [ b ] ) , namely @xmath104 \\[10pt ] \eqntext{(q , p , \mathbf { z } ) \in \mathbb{s}^n \times\mathbb{r}^n \times(0 , \infty ) ^n .}\end{aligned}\ ] ] the supermartingale problem is to find a probability measure @xmath105 on the filtered measurable space @xmath50 , @xmath51 , under which @xmath85 takes values in @xmath37 a.s .
, and the process @xmath106 is a local supermartingale for every @xmath107 of class @xmath20 with compact support .
the regularity condition on the family @xmath70 in ( [ 1.b ] ) that we alluded to earlier , mandates that the function @xmath108 if , in addition , the family of sets @xmath70 in ( [ 1.b ] ) satisfies the linear growth condition @xmath109 and the upper - semicontinuity condition @xmath110 for every @xmath111 and some real constant @xmath112 , then theorem 3.2 in @xcite shows that the family @xmath113 is convex and sequentially compact in the topology of vague convergence of probability measures .
we shall also consider the subcollection @xmath114 of _ markovian _ admissible systems , for which the functionals @xmath115 and @xmath116 as in ( [ 1.2])([1.1 ] ) are given as @xmath117 with measurable functions @xmath118 , that satisfy @xmath119 under condition ( [ lingro ] ) it follows then from so - called markovian selection results [ @xcite ; @xcite , chapter 12 ; @xcite , section 4.5 ] that the state process @xmath120 of ( [ 1.1 ] ) can be assumed to be ( strongly ) markovian under @xmath121 , @xmath122 .
we shall make this selection whenever admissible systems in @xmath123 are invoked .
consider now an investor who is `` small '' in the sense that his actions do not affect market prices .
he starts with initial fortune @xmath124 and uses a rule that invests a proportion @xmath125 of current wealth in the @xmath126th asset of the equity market , for any given time @xmath127 and all @xmath128 ; the remaining proportion @xmath129 is held in cash ( equivalently , in a zero - interest money market ) . here
@xmath130 is a progressively measurable functional assumed to satisfy the requirement @xmath131\\[-8pt ] \eqntext{\forall t \in(0 , \infty)}\end{aligned}\ ] ] for every continuous function @xmath132 .
( thus , the requirement ( [ 1.3 ] ) will be in force under all admissible systems . )
we shall denote throughout by @xmath133 the collection of all such ( nonanticipative ) _ investment rules _ , and by @xmath134 the sub - collection of all _ markovian _ investment rules , that is , those that can be expressed as @xmath135 , @xmath136 for some measurable function @xmath137 .
an investment rule is called _ bounded _
, if the functional @xmath138 is bounded uniformly on @xmath139 ; for a bounded investment rule , requirement ( [ 1.3 ] ) is satisfied automatically , thanks to ( [ 1.2 ] ) .
an investment rule is called _ portfolio _ if the functional @xmath140 satisfies @xmath141 on @xmath142 , and a portfolio is called _ long - only _ if @xmath143 also hold on this domain .
a long - only portfolio is clearly bounded .
given an initial wealth @xmath144 , an investment rule @xmath145 and an admissible model @xmath146 , the resulting wealth process @xmath147 satisfies the dynamics @xmath148\ ] ] and the initial condition @xmath149 .
in conjunction with ( [ expo ] ) in the differential form @xmath150 and the product rule of the stochastic calculus , this gives @xmath151\\[-8pt ] \eqntext{0 \le t < \infty.}\end{aligned}\ ] ] for any initial configuration @xmath152 , initial wealth @xmath153 , investment rule @xmath154 and admissible model @xmath155 , the product @xmath156 is therefore under @xmath157 a continuous , positive local martingale , thus also a supermartingale .
once again , it is important that this process be allowed to be a strict local martingale .
the choice of markovian investment rule @xmath158 given by @xmath159 leads to the long - only _ market portfolio _
@xmath160\\[-8pt ] \eqntext{\mbox{with weights } \mu_i ( t ) = x_i ( t ) / x(t ) , \
i=1 , \ldots , n , \ 0 \le t < \infty}\end{aligned}\ ] ] in the notation of ( [ 1.a ] ) .
it follows from ( [ 1.5 ] ) that investing according to this portfolio amounts to owning the entire market , in proportion to the initial wealth , @xmath161 .
reading ( [ 1.6 ] ) for the market portfolio of ( [ 1.7.a ] ) , and recalling ( [ 1.a ] ) , leads to @xmath162\\[-8pt ] \eqntext{0 \le t
< \infty}\end{aligned}\ ] ] or equivalently @xmath163 , where @xmath164\\[-8pt ] \eqntext{\displaystyle \mbox{satisfies } \int_0^t \| \widetilde{\vartheta } ( t , \mathfrak{w } ) \|^2 \,\mathrm{d } t < \infty}\end{aligned}\ ] ] for all @xmath165 , thanks to ( [ 1.2 ] ) .
with this notation , it follows from ( [ 1.c ] ) that @xmath166 on the strength of the integrability condition in ( [ 1.d ] ) , the dambis dubins schwartz representation [ e.g. , @xcite , page 174 ] of the @xmath167-local martingale @xmath168 gives @xmath169 where @xmath170 is one - dimensional , standard brownian motion under @xmath171.whereas the equations of ( [ 1.1 ] ) can be written as @xmath172 for @xmath128 , with @xmath173 we then have the representation @xmath174\\[-8pt ] & \qquad\hspace*{3pt}= & \exp\biggl\ { \int_0^\cdot ( \widetilde{\vartheta } ( t , \mathfrak { x } ) ) ' \,\mathrm{d } \widetilde { w}(t ) - { 1 \over2 } \int_0^\cdot \| \widetilde{\vartheta } ( t , \mathfrak { x } ) \|^2 \,\mathrm{d } t \biggr\}\nonumber\end{aligned}\ ] ] for the normalized reciprocal of the deflated total market capitalization , and @xmath175 for the dynamics market weights in ( [ 1.7.a ] ) ; here @xmath176 is the @xmath126th unit vector in @xmath177 .
let us consider now the smallest proportion @xmath178\\[-8pt ] & & \qquad\hspace*{198pt}\forall\mathcal{m } \in\mathfrak{m } ( \mathbf{x } ) \bigr\}\nonumber\end{aligned}\ ] ] of the initial total market capitalization @xmath179 which allows the small investor , starting with initial capital @xmath180 and through judicious choice of investment rule in the class @xmath181 , to match or exceed the performance of the market portfolio over the time - horizon @xmath182 $ ] , and to do this with @xmath167-probability one , under any model @xmath183 that might materialize .
we shall refer to @xmath184 of ( [ 1.7 ] ) as the _ arbitrage function _ for the family of meta - models @xmath185 , and think of it as a version of the arbitrage function studied in fernholz and karatzas ( @xcite ) which is `` robust '' with respect to @xmath186 .
the quantity of ( [ 1.7 ] ) is strictly positive ; see proposition [ proposition_1 ] below and the discussion following it . on the other hand , the set of ( [ 1.7 ] )
contains the number @xmath187 , so clearly @xmath188 if @xmath189 , then for every @xmath190and even for @xmath191 when the infimum in ( [ 1.7 ] ) is attained , as indeed it is in the context of theorem 1 below
there exists an investment rule @xmath192 such that @xmath193 holds for every @xmath194 . in other words ,
the investment rule @xmath195 leads then to _ strong arbitrage relative to the market _ portfolio in the terminology of @xcite here with the extra feature that such arbitrage is now _ robust _ , that is , holds under any possible admissible system or `` model '' that might materialize . if , on the other hand , @xmath196 , then such outperformance of ( equivalently , strong arbitrage relative to ) the market is just not possible over all meta - models @xmath197 . in either case ,
the highest return on investment relative to the market @xmath198 achievable using ( nonanticipative ) investment rules , is given as @xmath199 .
[ remark_1 ] instances of @xmath189 occur , when there exists a constant @xmath200 such that either @xmath201 or @xmath202 holds for every @xmath203 .
see the survey paper @xcite , examples 11.1 and 11.2 [ as well as @xcite , @xcite for additional examples ] .
[ proposition_1 ] the quantity of ( [ 1.7 ] ) satisfies @xmath204\\[-8pt ] \eqntext{\displaystyle \mbox{where } \phi(t,\mathbf{x}):= \sup_{\mathcal{m } \in \mathfrak{m } ( \mathbf{x } ) } \biggl ( \frac { \mathbb{e}^ { \mathbb { p}^\mathcal{m } } [ l(t ) x(t ) ] } { x_1 + \cdots+ x_n } \biggr ) .}\end{aligned}\ ] ] furthermore , under conditions ( [ measur])([uppsemicont ] ) , there exists an admissible system @xmath205 such that @xmath206 } { x_1 + \cdots+ x_n } .\ ] ] take an arbitrary element @xmath207 of the set on the right - hand side of ( [ 1.7 ] ) and an arbitrary admissible system @xmath155 .
there exists then an investment rule @xmath208 with the inequality @xmath209 valid @xmath210-a.s . on the strength of ( [ expo ] ) and ( [ 1.6 ] ) ,
the process @xmath211 is a @xmath210-supermartingale ; thus ( [ repr ] ) and ( [ 1.d ] ) lead to @xmath212 \nonumber\\[-8pt]\\[-8pt ] & \ge&\mathbb{e}^ { \mathbb{p}^\mathcal{m } } [ l ( t ) x(t ) ] > 0 .\nonumber\end{aligned}\ ] ] the inequality @xmath213 in ( [ 1.8 ] ) follows now from the arbitrariness of @xmath207 and @xmath214 .
the existence of an admissible system @xmath215 that satisfies ( [ 1.88 ] ) follows from theorem 3.4 in @xcite , in conjunction with the dynamics of ( [ 1.1 ] ) and ( [ expotoo ] ) .
although strong arbitrage relative to the market may exist within the framework of the models @xmath155 studied here ( cf .
remark [ remark_1 ] ) , the existence of a strictly positive supermartingale deflator process @xmath216 as in ( [ expo ] ) proscribes scalable arbitrage opportunities , also known as _ unbounded profits with bounded risk _ ( _ upbr _ ) ; this is reflected in the inequality @xmath217 of ( [ 1.8 ] ) .
we refer the reader to @xcite for the origin of the resulting _ nupbr _ concept , and to @xcite for an elaboration of this point in a different context , namely , the existence and properties of the numraire portfolio .
finally , let us write ( [ 1.8 ] ) as @xmath218 } { x_1 + \cdots+ x_n } .\hspace*{-10pt}\ ] ] we have for this quantity the interpretation @xmath219 as the `` arbitrage function for the model @xmath220 , '' at least when the matrix @xmath221 in invertible for every @xmath222 and when @xmath223-martingales can be represented as stochastic integrals with respect to the brownian motion @xmath92 in ( [ 1.1 ] ) .
consider now a continuous function @xmath224 with @xmath225 which is of class @xmath20 on @xmath226 and satisfies on this domain the fully nonlinear partial differential inequality ( pdi ) @xmath227\\[-8pt ] \eqntext{\forall a \in\mathcal{a } ( \mathbf { z } ) .}\end{aligned}\ ] ] we shall denote by @xmath228 the collection of all such continuous functions @xmath229 which are of class @xmath20 on @xmath230 and satisfy ( [ 2.1 ] ) and ( [ 2.2 ] ) .
this collection @xmath228 is nonempty , since we can take @xmath231 ; however , @xmath228 need not contain only one element .
let us fix an initial configuration @xmath232 and consider any admissible system @xmath233 .
applying it s rule to the process @xmath234 in conjunction with ( [ 1.c ] ) and ( [ 1.1 ] ) , we obtain its @xmath210-semimartingale decomposition as @xmath235\\[-8pt ] & & { } + \sum_{\nu=1}^n \bigl [ r_\nu(t , \mathfrak{x } ) - u \bigl(t - t , \mathfrak{x}(t ) \bigr ) \widetilde{\vartheta}_\nu(t , \mathfrak{x } ) \bigr ] \,\mathrm{d}w_\nu(t ) .\nonumber\end{aligned}\ ] ] here we have used the notation of ( [ 1.d ] ) , and have set @xmath236 from the inequality of ( [ 2.2 ] ) , coupled with the fact that @xmath237 holds for all @xmath238 , this last expression is clearly not positive . as a result , the positive process @xmath239 of ( [ 2.4 ] ) is a @xmath210-supermartingale , namely , @xmath240 \nonumber\\[-8pt]\\[-8pt ] & = & \mathbb{e}^ { \mathbb{p}^\mathcal{m } } [ l(t)x(t ) | \mathcal{f}(t ) ] \nonumber\end{aligned}\ ] ] holds @xmath167-a.s . ,
@xmath241 and @xmath242 ; in particular , @xmath243 \\ & = & \mathbb{e}^ { \mathbb{p}^\mathcal{m } } [ l(t ) x(t ) ] \qquad \forall\mathcal{m } \in\mathfrak{m } ( \mathbf{x } ) .\end{aligned}\ ] ] with the notation of ( [ 1.8 ] ) , we obtain in this manner the following analog of the inequality in proposition [ proposition_1 ] : @xmath244 digging in this same spot , just a bit deeper , leads to our next result ; this is very much in the spirit of theorem 5 in @xcite and of section ii.2 in lions ( @xcite ) .
[ proposition_2 ] for horizon @xmath103 , initial configuration @xmath245 and function @xmath229 in the collection @xmath246 , we have the inequality @xmath247 furthermore , the markovian investment rule @xmath248 generated by this function @xmath249 through @xmath250\\[-8pt ] & & { } + \frac{z_i } { z_1 + \cdots+ z_n } , \qquad ( t,\mathbf{z } ) \in[0,t ] \times(0 , \infty)^n\nonumber\end{aligned}\ ] ] for each @xmath251 , satisfies for every admissible system @xmath252 the inequality @xmath253 for
a fixed initial configuration @xmath254 , an arbitrary admissible model @xmath155 and any function @xmath255 , let us recall the notation of ( [ 2.4 ] ) and re - cast the dynamics of ( [ 2.5 ] ) as @xmath256 here by virtue of ( [ 2.5 ] ) , ( [ 2.6 ] ) and ( [ 1.d ] ) we have written @xmath257\\[-8pt ] & & { } - \vartheta_\nu(t , \mathfrak{x})\nonumber\end{aligned}\ ] ] for @xmath258 and have introduced in the notation of ( [ 2.7 ] ) the continuous , increasing process @xmath259 the expression of ( [ 2.10 ] ) suggests considering the markovian investment rule @xmath260 as in ( [ 2.11.a ] ) ; then we cast the expression of ( [ 2.10 ] ) as @xmath261 on the strength of ( [ 1.6 ] ) , the value process generated by this investment rule @xmath262 starting with initial wealth @xmath263 , satisfies the equation @xmath264 juxtaposing this to ( [ 2.9 ] ) , and using the positivity of @xmath265 along with the nonnegativity and nondecrease of @xmath266 , we obtain the @xmath167-a.s .
comparison @xmath267 , thus @xmath268 with @xmath269 this leads to ( [ 2.11.b ] ) , in conjunction with ( [ 2.1 ] ) .
we conclude from ( [ 2.11.b ] ) that the number @xmath270 belongs to the set on the right - hand side of ( [ 1.7 ] ) , and the first comparison in ( [ 2.8 ] ) follows ; the second is just a restatement of ( [ 1.8 ] ) .
suppose that the function @xmath271 of ( [ 1.8 ] ) belongs to the collection @xmath272 .
then @xmath273 is the smallest element of @xmath274 ; the infimum in ( [ 1.7 ] ) is attained ; we can take @xmath275 in ( [ 2.11.b ] ) , ( [ 2.11.a ] ) ; and the inequality in ( [ 2.8 ] ) holds as equality , that is , @xmath276 coincides with the arbitrage function @xmath277 _ interpretation : _
imagine that the small investor is a manager who invests for a pension fund and tries to track or exceed the performance of an index ( the market portfolio ) over a finite time - horizon .
he has to do this in the face of uncertainty about the characteristics of the market , including its covariance and price - of - risk structure , so he acts with extreme prudence and tries to protect his clients against the most adverse market configurations imaginable [ the range of such configurations is captured by the constraints ( [ e.3 ] ) , ( [ 1.b ] ) ] .
if such adverse circumstances do not materialize his strategy generates a surplus , captured here by the increasing process @xmath278 of ( [ 2.11 ] ) with @xmath279 , which can then be returned to the ( participants in the ) fund .
we are borrowing and adapting this interpretation from @xcite .
similarly , the markovian investment rule @xmath260 generated by the function @xmath279 in ( [ 2.11.a ] ) , ( [ fg ] ) implements the best possible outperformance of the market portfolio , as in ( [ 2.11.b ] ) .
for the purposes of this section we shall impose the following _ growth _ condition on the family @xmath280 of subsets of @xmath281 in ( [ e.3.a ] ) , ( [ 1.b ] ) : there exists a constant @xmath282 , such that for all @xmath283 we have @xmath284 we shall also need the following _ strong ellipticity _ condition , which mandates that for every nonempty , compact subset @xmath285 of @xmath37 , there exists a real constant @xmath286 such that @xmath287 [ assa ] there exist a continuous function @xmath288 , a @xmath289-function @xmath290 , and a continuous square root @xmath291 of @xmath292 , namely @xmath293 such that , with the vector - valued function @xmath294 defined by @xmath295 condition ( [ d.22 ] ) is satisfied , whereas the system of stochastic differential equations @xmath296 , \nonumber\\[-8pt]\\[-8pt
] \eqntext{x_i(0)=x_i>0 , i=1,\ldots , n}\end{aligned}\ ] ] has a solution in which the state process @xmath120 takes values in @xmath37 .
a bit more precisely , this assumption posits the existence of a markovian admissible system @xmath297 consisting of a filtered probability space @xmath298 , @xmath74 and of two continuous , adapted process @xmath85 and @xmath92 on it , such that under the probability measure @xmath299 the process @xmath92 is @xmath300-dimensional brownian motion , the process @xmath85 takes values in @xmath37 a.s . and ( [ 1.1 ] ) holds with @xmath301 as in ( [ d.2 ] ) , and with @xmath302 , @xmath303 ( @xmath304 ) .
the system of equations ( [ d.1 ] ) can be cast equivalently as @xmath305\\[-8pt ] & & \hphantom{x_i(t ) \biggl [ } { } + \biggl ( \sum_{j=1}^n \mathbf{a}_{i j } ( \mathfrak { x } ( t ) ) x_j ( t ) d_j h ( \mathfrak { x } ( t ) ) \biggr)\ , \mathrm{d } t \biggr].\nonumber\end{aligned}\ ] ] [ assb ] in the notation of the previous paragraph and under the condition @xmath306 we define on @xmath37 the continuous functions @xmath307 and @xmath308 $ ] and assume that the function @xmath309 , \qquad ( \tau , \mathbf{x } ) \in[0 , \infty)^n\ ] ] is continuous on @xmath310 and of class @xmath311 on @xmath312 sufficient conditions for assumptions [ assa ] , [ assb ] to hold are provided in @xcite , sections 8 and 9 .
it is also shown there , that we have the @xmath313-martingale property @xmath314 = x(t ) l(t ) \cdot\gamma\bigl(t - t , x ( t ) \bigr ) , \qquad 0 \le t \le t\ ] ] for the function @xmath315 this function is of class @xmath20 on @xmath316 and satisfies the initial condition @xmath317 on @xmath37 as well as the _ linear _ second - order parabolic equation @xmath318\\[-8pt ] \eqntext{(\tau , \mathbf { z } ) \in(0 , \infty ) \times(0 , \infty)^n .}\end{aligned}\ ] ] [ proposition_3 ] under the assumptions [ assa ] , [ assb ] and the conditions ( [ 3.2.gg ] ) , ( [ 3.2 ] ) , suppose that the functions @xmath319 and @xmath320 satisfy the inequality @xmath321\\[-8pt ] \qquad & & \qquad \ge\sum_{i=1}^n \sum_{j=1}^n z_i z_j a_{ij } \biggl ( \frac{1 } { 2 } d^2_{ij } \gamma(\tau , \mathbf{z})+ \frac{d_{i } \gamma(\tau , \mathbf{z})}{z_1 + \cdots+z_n } \biggr ) \qquad \forall a \in\mathcal{a } ( \mathbf { z})\nonumber\end{aligned}\ ] ] for every @xmath322
. then , in the notation of ( [ 1.7])([1.9 ] ) , we have : @xmath323 for the markovian admissible system @xmath324 posited in assumption [ assa ] ; the conclusions of proposition [ proposition_2 ] and its corollary for @xmath325 ; as well as the @xmath167-a.s . comparison @xmath326,\qquad
0\le t \le t , \ ] ] which holds for every @xmath194 and as equality for @xmath327 . under condition ( [ d.11 ] )
the function @xmath328 belongs to the collection @xmath228 , as ( [ 2.2 ] ) is satisfied with @xmath329 on the strength of ( [ d.10 ] ) and ( [ d.11 ] ) ; thus , we deduce @xmath330 from ( [ 2.7.b ] ) . on the other hand , equality ( [ d.8 ] ) with @xmath33 , and the definition of @xmath331 in ( [ 1.8 ] ) , give @xmath332 } { x_1 + \cdots+ x_n } = \mathfrak{u}_{\mathcal{m}_o } ( t,\mathbf{x } ) \le\phi(t,\mathbf{x } ) , \ ] ] so the equality @xmath333 follows . in other words ,
we identify @xmath334 as a markovian admissible system that satisfies ( [ 1.88 ] ) and attains the supremum in ( [ 1.8 ] ) .
the remaining claims come from proposition [ proposition_2 ] and its corollary [ in particular , from reading ( [ 2.8 ] ) with @xmath335 and from ( [ 2.7.7 ] ) , ( [ d.8 ] ) .
[ remark_2 ] proposition [ proposition_3 ] holds under conditions weaker than those imposed in assumptions [ assa ] and [ assb ] above , at the `` expense '' of a certain localization .
more precisely , one posits the existence of locally bounded and locally lipschitz functions @xmath336 and @xmath337 ( @xmath304 ) for which ( [ d.22 ] ) , ( [ 3.2.gg ] ) , ( [ 3.2 ] ) and ( [ d.4 ] ) are satisfied with @xmath338 , and for which there exists a markovian admissible system @xmath339 whose state process @xmath340 in ( [ d.1 ] ) is , under @xmath341 , a strong markov process with values in @xmath342 a.s . using results from the theory of stochastic flows [ @xcite , @xcite ] and from parabolic partial differential equations [ @xcite ,
ekstrm and tysk ( @xcite ) ] , theorem 2 in ruf ( @xcite ) shows that the function @xmath343 is then of class @xmath311 locally on @xmath344 , and solves there equation ( [ d.10 ] ) .
we have now gone as far as we could without having to specify the nature of our filtered measurable space @xmath345 , @xmath346 . to proceed further , we shall need to choose this space carefully .
we shall take as our sample space the set @xmath52 of right - continuous paths @xmath347 . here
@xmath348 is an additional `` absorbing point '' ; paths stay at @xmath348 once they get there , that is , after @xmath349 , and are continuous on @xmath350 ; we are employing here , and throughout this work , the usual convention @xmath351 .
we also select @xmath352 and @xmath353 , where @xmath354 is the zero matrix . with @xmath355 ,
the filtration @xmath356 is a _
standard system _ in the terminology of @xcite .
this means that each @xmath357 is isomorphic to the borel @xmath358-algebra on some polish space , and that for any decreasing sequence @xmath359 where each @xmath360 is an atom of the corresponding @xmath361 , for some increasing sequence @xmath362 , we have @xmath363 [ see the appendix in fllmer ( @xcite ) , as well as @xcite and fllmer ( @xcite ) ] . with all this in place we take @xmath364 as our filtered measurable space , where @xmath365 an admissible system @xmath366 , @xmath367 defined as in section [ sec2 ] consists of this filtered measurable space @xmath368 , of a probability measure @xmath86 on it , of an @xmath300-dimensional brownian motion @xmath92 on the resulting probability space and of the coordinate mapping process @xmath369 which is assumed to satisfy ( [ 1.1 ] ) , ( [ 1.2 ] ) and to take values in @xmath37 , @xmath370-a.s
. we shall take @xmath371 in the notation of ( [ lambda ] ) , ( [ expo ] ) and ( [ 1.a ] ) , and note @xmath372 . with this setup , there exists a probability measure @xmath373 on @xmath374 such that @xmath375 we express this property ( [ bigtriangleleft ] ) by writing @xmath376 . under the measure @xmath377 ,
the process @xmath378 of ( [ 1.1.b ] ) is brownian motion ; whereas the processes @xmath379 and @xmath380 of ( [ mu ] ) , ( [ lambda ] ) in section [ subsec3.1 ] are nonnegative @xmath381-martingales . the `` absorbing state '' @xmath348 acts here as a proxy for @xmath210-null sets to which the new measure @xmath377 may assign positive mass ; the possible existence of such sets makes it important that the filtration @xmath382 be `` pure , '' that is , _ not _ completed by @xmath167-null sets .
this probability measure @xmath383 satisfies @xmath384 } { x_1 + \cdots + x_n } & = & \mathbb{e}^{\mathbb{p}^\mathcal{m } } \bigl [ ( 1 / \lambda(t ) ) \cdot\mathbf { 1}_{\ { \mathcal{t } > t\ } } \bigr ] \nonumber\\[-8pt]\\[-8pt ] & = & \mathbb{q } ( \mathcal { t } > t ) \qquad
\forall t \in[0 , \infty)\nonumber\end{aligned}\ ] ] and @xmath385 we also have @xmath386-a.e . on @xmath387 @xmath388 and @xmath389 whereas , @xmath390-a.e . on @xmath391 , we have @xmath392 we deduce from ( [ foell ] ) that the arbitrage function of ( [ 1.9 ] ) for the model @xmath194 is given by the _ probability of _ `` _ _ containment _ _ '' _ under the measure _ @xmath383 , namely , the @xmath393-probability that the process @xmath394 , started at @xmath395 , stays in @xmath37 throughout the time - horizon @xmath396 $ ] . at this point we shall impose the following requirements on @xmath397 , the family of compact , convex subsets of @xmath398 in ( [ 1.b ] ) , ( [ e.3 ] ) :
there exists a constant @xmath399 , such that for all @xmath400 we have the strengthening @xmath401 of the growth condition in ( [ 3.2.gg ] ) , as well as the `` shear '' condition @xmath402 \le c .\ ] ] then the following identity holds @xmath390-a.s .
: @xmath403 for justification of the claims made in this subsection , we refer to section 7 in fernholz and karatzas ( @xcite ) , as well as @xcite , @xcite and @xcite in addition , of course , to the seminal work by fllmer ( @xcite , @xcite ) .
the special structure of the filtered measurable space @xmath404 that we selected in this section is indispensable for this construction and for the representation ( [ foell ] ) ; whereas the inequality @xmath405 from condition ( [ thetatrace ] ) is important for establishing the representation of ( [ ttaui ] ) .
let us fix then an initial configuration @xmath406 and denote by @xmath407 the collection of stochastic systems @xmath408 that consist of the filtered measurable space @xmath409 , of a probability measure @xmath410 , of an @xmath59-valued brownian motion @xmath411 under @xmath377 , and of the cordinate mapping process @xmath412 which satisfies @xmath377-a.s .
the system of the stochastic equations in ( [ 1.1.a ] ) @xmath413 here the elements @xmath414 , @xmath415 of the matrix @xmath416 are progressively measurable functionals that satisfy , in the notation of ( [ e.3.a ] ) , @xmath417 as in section [ subsec2.3 ] , we shall denote by @xmath418 the subcollection of @xmath419 that consists of _ markovian _ auxiliary admissible systems , namely , those for which the equations of ( [ aux ] ) are satisfied with @xmath420 and @xmath421 , @xmath422 and with measurable functions @xmath423 and @xmath424 that satisfy the condition @xmath425 , @xmath426 .
we invoke the same markovian selection results as in section [ subsec2.3 ] , to ensure that the process @xmath427 is strongly markovian under any given @xmath428 , @xmath429 . by analogy with ( [ ttaui ] ) , we consider @xmath430 then for every @xmath431 we have @xmath432 whereas @xmath433 , @xmath434 holds for every @xmath435 .
[ remark_7 ] as in section [ subsec2.2 ] , solving the stochastic equation ( [ aux ] ) subject to condition ( [ const1 ] ) amounts to requiring that the process @xmath436 be a local supermartingale , for every continuous @xmath437 which is of class @xmath311 on @xmath438 and has compact support ; here @xmath439 is the nonlinear second - order partial differential operator in ( [ e.3.d ] ) .
[ remark_3 ] the total capitalization process @xmath440 satisfies , by virtue of ( [ aux ] ) , the equation @xmath441 , \\
\widetilde{n}(\cdot ) & : = & \sum_{\nu=1}^n \int_0^\cdot\biggl ( \sum_{i=1}^n ( x_i ( t ) / x(t ) ) \sigma_{i \nu } ( t , \mathfrak { x } ) \biggr)\ , \mathrm{d } \widetilde{w}_\nu(t ) .\end{aligned}\ ] ] under the measure @xmath442 , the process @xmath443 is a continuous local martingale with quadratic variation @xmath444 from ( [ 3.2.g ] ) , so the total capitalization process @xmath445 takes values in @xmath446 , @xmath447-a.e . ; here @xmath448 is a one - dimensional @xmath447-brownian motion .
this is in accordance with our selection of the punctured nonnegative orthant @xmath449 in ( [ 1.b ] ) as the state - space for the process @xmath120 under @xmath377 . under this measure , the relative weights @xmath450
, @xmath128 are nonnegative local martingales and supermartingales , in accordance with ( [ mu ] ) , and since @xmath451 these processes are bounded , so they are actually martingales . once any one of the processes @xmath452 [ i.e. , any one of the processes @xmath453 becomes zero , it stays at zero forever ; of course , not all of them can vanish at the same time .
in section [ subsec7.1 ] we started with an arbitrary admissible system @xmath454 and produced an `` auxiliary '' admissible system @xmath455 , for which the property ( [ foell ] ) holds .
thus , for every @xmath456 we deduce @xmath457\\[-8pt ] & \hspace*{3pt}\ge & \sup_{\mathcal { m } \in\mathfrak { m } ( \mathbf { x } ) } \biggl ( \frac { \mathbb{e}^ { \mathbb{p}^\mathcal{m } } [ l(t )
x(t ) ] } { x_1 + \cdots+ x_n } \biggr)= \phi(t , \mathbf { x } ) .\nonumber\end{aligned}\ ] ] we suppose from now onwards that , for every progressively measurable functional @xmath458 which satisfies @xmath459 we can select a progressively measurable functional @xmath460 with @xmath461 [ see the `` measurable selection '' results in chapter 7 of @xcite ] .
we introduce now the functional @xmath462 as in ( [ 1.d ] ) and also , by analogy with ( [ ttau.a ] ) , the stopping rule @xmath463 [ cf .
@xcite , where stopping rules of this type also play very important roles in the study of arbitrage ] .
we recall now ( [ const2 ] ) ; on the strength of the requirement ( [ thetatrace ] ) , this gives @xmath464 for every @xmath465 , and @xmath466 , @xmath467 for every @xmath468 . in conjunction with ( [ 3.2.g ] )
, we obtain from ( [ ttau111 ] ) that @xmath469 holds for every @xmath465 , and that @xmath470 , @xmath471 holds for every @xmath472 .
we deduce for the stopping rules of ( [ ttau1 ] ) and ( [ ttaui1 ] ) the identification @xmath473 .
let us fix now a stochastic system @xmath474 as in section [ subsec7.2 ] , pick a progressively measurable functional @xmath475 with @xmath476 for all @xmath477 and select a progressively measurable functional @xmath478 as in ( [ thetaalpha ] ) . for _ this
_ @xmath115 and _ this _
@xmath479 , we define @xmath480 by ( [ ttau101 ] ) as well as @xmath481\\[-8pt ] \eqntext{\mbox{for } 0 \le t < \mathcal{t}}\end{aligned}\ ] ] as in ( [ lambda ] ) , and set @xmath482 in the notation of ( [ ttau1 ] ) .
the resulting process @xmath380 is a local martingale and a supermartingale under @xmath377 , and we have @xmath483 , @xmath381-a.e . we introduce also the sequence of @xmath79-stopping rules @xmath484 which satisfy @xmath485 and @xmath486 @xmath487 , for every @xmath488 . from novikov s theorem [ e.g. , @xcite , page 198 ] , @xmath489 is a uniformly integrable @xmath381-martingale ; in particular , @xmath490 holds for every @xmath491 .
thus , the recipe @xmath492 , \qquad a \in\mathcal{f}(s_n)\ ] ] defines a consistent sequence , or `` tower , '' of probability measures @xmath493 on @xmath494 .
appealing to the results in @xcite , pages 140143 [ see also the appendix of fllmer ( @xcite ) ] , we deduce the existence of a probability measure @xmath495 on @xmath496 such that @xmath497 \quad \mbox{holds for every } a \in \mathcal{f}(s_n ) , n \in\mathbb{n}.\hspace*{-25pt}\ ] ] ( here again , the special structure imposed in this section on the filtered measurable space @xmath498 is indispensable . ) therefore , for every @xmath103 we have @xmath499 = \mathbb{e}^\mathbb{q } \bigl [ \lambda ( t ) \cdot \mathbf { 1}_{\{s_n > t\ } } \bigr]\ ] ] by optional sampling , whereas monotone convergence leads to @xmath500 .\ ] ] the following result echoes similar themes in @xcite .
[ lemma_1 ] the process @xmath380 of ( [ lamda ] ) , ( [ lamda_too ] ) is a @xmath393-martingale , if and only if we have @xmath501 [ i.e. , if and only if the process @xmath394 never hits the boundary of the orthant @xmath502 , @xmath503-a.s . ] .
if @xmath504 holds , the nonnegativity of @xmath380 and ( [ wald ] ) give @xmath505 \le\mathbb{e}^\mathbb{q } [ \lambda ( t ) ] \qquad \forall t \in(0 , \infty ) .\vadjust{\goodbreak}\ ] ] but @xmath380 is a @xmath393-supermartingale , so the reverse inequality @xmath506 \le \lambda ( 0 ) = 1 $ ] also holds .
we conclude that @xmath507=1 $ ] holds for all @xmath103 , so @xmath380 is a @xmath393-martingale .
if , on the other hand , @xmath380 is a @xmath508-martingale , then @xmath506=1 $ ] and ( [ wald ] ) give @xmath509 for every @xmath510 , from optional sampling and the fact that @xmath511 holds @xmath373-a.e .
on @xmath512 .
we conclude @xmath513 ; in conjunction with the identification @xmath514 and ( [ ttaui1 ] ) , this means that the cordinate mapping process @xmath515 never reaches the boundary of ( i.e. , takes values in ) the strictly positive orthant @xmath516 , @xmath503-a.e .
when the conditions of lemma [ lemma_1 ] prevail , the process @xmath517 is brownian motion under the probability measure @xmath518 introducedin ( [ tower ] ) .
this measure satisfies the equations of ( [ foell ] ) , whereas the process @xmath394 solves @xmath167-a.s .
the system @xmath519 , \qquad x_i(0)=x_i>0\ ] ] for @xmath520 , as in ( [ 1.1 ] ) .
it is then not hard to check that @xmath101 defined by ( [ expo ] ) satisfies @xmath167-a.s .
the identity @xmath521 in accordance with ( [ lambda ] ) .
we formalize these considerations as follows .
[ assc ] suppose that the collection of sets @xmath522 in ( [ 1.b ] ) satisfies ( [ 3.2.g ] ) , ( [ thetatrace ] ) and that for any given progressively measurable functional @xmath523 which satisfies ( [ alpha ] ) and @xmath524 there exists a progressively measurable functional @xmath460 that satisfies the condition ( [ thetaalpha ] ) , thus also by virtue of ( [ thetatrace ] ) @xmath525 the analysis of this subsection shows that , under assumption [ assc ] and starting with any initial configuration @xmath406 and with an arbitrary `` auxiliary '' admissible stochastic system @xmath526 , @xmath527 in @xmath528 as in section [ subsec7.2 ] , the process @xmath380 of ( [ lamda_too ] ) is a @xmath508-martingale , and we can construct a `` primal '' admissible system @xmath529 as in section [ subsec2.1 ] [ i.e. , with the canonical process @xmath427 taking values in @xmath37 , @xmath167-a.s . ] , for which ( [ foell ] ) holds , and we have @xmath530 as in ( [ bigtriangleleft ] ) .
we deduce @xmath531 } { x_1 + \cdots+ x_n } \biggr)= \phi(t , \mathbf { x } ) .\ ] ] the reverse inequality @xmath532 was established in ( [ 3.7.f ] ) . this way , for every function @xmath533 in the collection @xmath246
, we can strengthen ( [ 2.8 ] ) to @xmath534\\[-8pt ] \eqntext{\forall ( t,\mathbf{x } ) \in(0,\infty ) \times(0,\infty)^n .}\end{aligned}\ ] ] we have established the following result . [ proposition_4 ]
recall the functions @xmath535 , @xmath536 and @xmath537 defined on @xmath538 by ( [ 1.7 ] ) , ( [ 1.8 ] ) and ( [ 3.7.f ] ) , respectively , and impose assumption [ assc ] .
then ( [ 3.8 ] ) holds for every function @xmath539 .
[ remark_4 ] here is a situation where assumption [ assc ] prevails : suppose that ( [ 3.2.g ] ) holds and that , for every @xmath540 and @xmath541 , we have @xmath542 for @xmath543 given by @xmath544 , @xmath545 and @xmath546
. then for any progressively measurable @xmath547 that satisfies ( [ aa1 ] ) we select the progressively measurable functional @xmath548 via @xmath549 , @xmath550 .
this choice induces @xmath551 which obeys @xmath552 as in ( [ aa2 ] ) for all @xmath553 ; the process @xmath380 of ( [ lamda ] ) and ( [ lamda_too ] ) is a @xmath393-martingale , whereas ( [ 1.1 ] ) becomes @xmath554 , \quad x_i(0)=x_i>0\ ] ] for @xmath128 .
the condition ( [ aa1 ] ) guarantees now that @xmath555 takes values in @xmath342 , @xmath556-a.s . in the resulting primal admissible system @xmath529 .
the quantity @xmath557 defined in ( [ 3.7.f ] ) is the value of a stochastic control problem : namely , the _ maximal _ `` _ _ containment _ _ '' _ probability _ , over all measures @xmath558 with @xmath559 , that the process @xmath427 with dynamics ( [ aux ] ) , initial configuration @xmath560 , and controlled through the choice of progressively measurable functional @xmath561 as in ( [ const1 ] ) , ( [ const2 ] ) , does _ not _ hit the boundary of the positive orthant by time @xmath562 . let us suppose that the resulting function @xmath537 is continuous on @xmath438
. then it can be checked [ as in lions ( @xcite ) , lemma ii.1 and @xcite , theorem ii.4 ] that it satisfies as well the following dynamic programming principle : _ for every initial configuration @xmath563 , the process _ @xmath564\\[-8pt ] \eqntext{0\le t \le t \mbox { \textit{is a } } \mathbb{q}^ { \mathcal{n}}\mbox{-\textit{supermartingale } , } \forall\mathcal{n }
\in\mathfrak { n}(\mathbf { x } ) .}\end{aligned}\ ] ] [ see @xcite , @xcite , @xcite and @xcite for results in a similar vein . ]
equivalently , the process @xmath565 , @xmath242 is a @xmath566-supermartingale , for every @xmath567 .
consider now an arbitrary continuous function @xmath568 which satisfies @xmath569 on @xmath516 , and is such that , for every @xmath570 and @xmath571 , the process @xmath572 is a @xmath573-supermartingale .
we shall denote by @xmath574 the collection of all such functions and note that @xmath575 and @xmath576 . from optional sampling
we have then for every @xmath577 the comparisons @xmath578 = \mathbb{q}^ { \mathcal{n } } ( \mathcal{t } > t ) , \ ] ] thus also @xmath579 , @xmath580 . in other words , _ the function @xmath581 defined in _ ( [ 3.7.f ] ) _ is the smallest element of the collection _ @xmath582 .
it is also clear from this line of reasoning that @xmath583 attains the supremum @xmath584 in ( [ 3.7.f ] ) , if and only if the process @xmath585 is a @xmath586-martingale .
[ theorem_1 ] suppose that assumption [ assc ] and conditions ( [ measur])([uppsemicont ] ) hold and that the function @xmath537 of ( [ 3.7.f ] ) is continuous on @xmath438 .
then the infimum in ( [ 1.7 ] ) is attained , and @xmath587 consider an arbitrary function @xmath588 in the collection @xmath589 just defined and fix an arbitrary pair @xmath590 ; then for every @xmath591 , consider a mollification @xmath592 of the function @xmath593 with @xmath594 .
proposition [ proposition_2 ] gives then @xmath595 .
since @xmath596 is arbitrary , this shows that @xmath597 is dominated by @xmath598 , the infimum of @xmath599 over all functions @xmath600 .
but the reverse inequality @xmath601 holds on the strength of ( [ 3.8 ] ) , so ( [ 3.8.a ] ) follows . under the conditions of theorem [ theorem_1 ] , the arbitrage function @xmath602 is equal to the function @xmath603 of ( [ 3.7.f ] ) and is continuous on @xmath604 .
thanks to the dynamic programming principle of ( [ dp ] ) , it is also a viscosity solution of the _ hamilton jacobi bellman _ ( _ hjb _ ) _ equation _
@xmath605 on @xmath438 [ cf .
lions ( @xcite ) , theorem iii.1 or @xcite , theorem i.1 ] .
if in addition to being continuous , as we assumed in theorem [ theorem_1 ] , the function @xmath606 of ( [ 3.7.f ] ) is of class @xmath20 locally on @xmath607 , then the arbitrage function @xmath602 is not only a viscosity solution but actually a classical solution of the hjb equation ( [ 3.9 ] ) .
this is the case , for instance , under the combined conditions of theorem [ theorem_1 ] and proposition [ proposition_3 ] ; then the arbitrage function @xmath608 also satisfies on the domain @xmath438 the _ linear _ parabolic equation @xmath609 with @xmath288 as in assumption [ assa ] or remark [ remark_2 ] , in addition to the initial condition @xmath610 in particular , the arbitrage function @xmath602 satisfies , in this case , the requirement ( [ d.11 ] ) and belongs to the class @xmath246 of section [ sec5 ] .
recalling propositions [ proposition_1][proposition_3 ] and theorem [ theorem_1 ] , we summarize the above discussion as follows .
[ theorem_2 ] suppose that conditions ( [ measur])([uppsemicont ] ) , ( [ 3.2.gg ] ) , ( [ 3.2 ] ) and assumptions [ assa ] , [ assb ] and [ assc ] are in force .
then the arbitrage function @xmath602 is the smallest element of the class @xmath246 , as well as a classical solution of both the hjb equation ( [ 3.9 ] ) and of the linear parabolic equation ( [ 7.25 ] ) , subject to ( [ 3.9.a ] ) .
furthermore ( [ 3.8.a ] ) holds , the infimum in ( [ 1.7 ] ) is attained , and the markovian investment rule @xmath611 in ( [ 2.11.a ] ) with @xmath612 satisfies ( [ 2.11.b ] ) for every admissible system @xmath183 .
[ remark_5 ] we note that theorem [ theorem_2 ] is in agreement with general regularity theory for fully nonlinear parabolic equations , as in lions ( @xcite ) , theorem ii.4 ( see also @xcite , section 6.5 ; krylov ( @xcite ) ; wang ( @xcite , @xcite , @xcite ) , theorems ii.3.2 and iii.2 ; or @xcite , chapter xiv ) .
as we mentioned already , assumptions [ assa ] and [ assb ] can be replaced in theorem [ theorem_2 ] by the conditions of remark [ remark_2 ] .
we conjecture that the conclusions of theorem [ theorem_2 ] should hold under even weaker assumptions but leave this issue for future research .
we also remark that the function @xmath613 , @xmath614 satisfies an hjb - type equation simpler than ( [ 3.9 ] ) , namely , the _ pucci maximal equation _ , @xmath615 along with the initial condition @xmath616 . in the setting of theorem [ theorem_2 ] , equation ( [ 3.10 ] ) reduces to @xmath617 let us go back to the volatility - stabilized model introduced in @xcite , but now with some `` knightian '' uncertainty regarding its volatility structure @xmath618 for some given @xmath619 .
the case @xmath620 corresponds to the variance structure of the model studied in fernholz and karatzas ( @xcite , @xcite ) .
more specifically let us assume that , for any given @xmath621 , the compact , convex subset @xmath622 of @xmath9 in ( [ e.3.a ] ) consists of all matrices @xmath623 with @xmath624 for @xmath625 and @xmath626 the sets of ( [ 1.b ] ) are given as @xmath627 \bigr\}\ ] ] for some given constants @xmath628 $ ] , @xmath629 ; these choices satisfy ( [ 3.2.g ] ) , ( [ thetatrace ] ) . condition ( [ 3.2 ] ) is satisfied in this case automatically ( in fact , with @xmath630 ) , as are ( [ 3.2.gg ] ) and ( [ d.4 ] ) : it suffices to take @xmath631 and @xmath632 , which induces @xmath633 in ( [ d.2 ] ) .
these functions are all locally bounded and locally lipschitz continuous on @xmath502 .
the hjb equation ( [ 3.9 ] ) satisfied by the arbitrage function @xmath634 becomes @xmath635 , \ ] ] and reduces to the _ linear _ parabolic equation @xmath636 of ( [ 7.25 ] ) for the choice of variances in ( [ volstab ] ) .
the reason for this reduction is that the expression on the left - hand side of ( [ 3.11 ] ) is negative , so we have considerable simplification in this case . in this example , the arbitrage function @xmath608 can be represented as @xmath637\ ] ] in terms of the components of the @xmath342-valued capitalization process @xmath638 .
these are now time - changed versions @xmath639 , @xmath128 of the independent squared - bessel processes @xmath640 run with a time change @xmath641 common for all components , and with @xmath642 independent standard brownian motions [ see fernholz and karatzas ( @xcite , @xcite ) , @xcite and pall ( @xcite ) for more details ] .
for any given investment rule @xmath145 and admissible system @xmath183 , let us consider the quantity @xmath643 this measures , as a proportion of the initial total market capitalization , the smallest initial capital that an investor who uses the rule @xmath138 and operates within the market model @xmath72 , needs to set aside at time @xmath33 in order for his wealth to be able to `` catch up with the market portfolio '' by time @xmath269 , with @xmath210-probability one .
our next result exhibits the arbitrage function @xmath644 of ( [ 1.7 ] ) as the min max value of a zero - sum stochastic game between two players : the investor , who tries to select the rule @xmath645 so as to make the quantity of ( [ 9.1 ] ) as small as possible and `` nature , '' or the goddess tyche herself , who tries to thwart him by choosing the admissible system or `` model '' @xmath646 to his detriment .
[ theorem_3 ] under the conditions of theorem [ theorem_1 ] , we have @xmath647 for the quantities of ( [ 9.1 ] ) and ( [ 1.9 ] ) we claim @xmath648 indeed , if the set on the right - hand side of ( [ 9.1 ] ) is empty , we have@xmath649 and nothing to prove ; if , on the other hand , this set is not empty , then for any of its elements @xmath207 the process @xmath650 is a @xmath651-supermartingale , and therefore ( [ 1.89 ] ) , that is , @xmath652 , still holds and ( [ 9.3 ] ) follows again . taking the infimum with respect to @xmath145 on the left - hand side of ( [ 9.3 ] ) , then the supremum of both sides with respect to @xmath183 , we obtain @xmath653 from ( [ 1.9 ] ) .
the quantity @xmath654 is the lower value of the stochastic game under consideration . in order to complete the proof of ( [ 9.2 ] ) it suffices , on the strength of theorem [ theorem_1 ] , to show that the upper value @xmath655 of this game satisfies @xmath656 to see this , we introduce for each given investment rule @xmath657 the quantity @xmath658 that is , the smallest proportion @xmath207 of the initial market capitalization that allows an investor using the rule @xmath138 to be able to `` catch up with the market portfolio '' by time @xmath269 with @xmath659-probability one , no matter which admissible system ( model ) @xmath660 might materialize .
we have clearly @xmath661 which leads to @xmath662 and proves ( [ 9.5 ] ) .
let us place ourselves now in the context of theorem [ theorem_2 ] and observe that proposition [ proposition_1 ] , along with proposition [ proposition_2 ] and its corollary , yields @xmath663 by virtue of ( [ 9.3 ] ) for @xmath664 and of ( [ 2.11.b ] ) for @xmath665 . here
@xmath334 is the `` least favorable admissible system '' that attains the supremum over @xmath666 in ( [ 1.8 ] ) , and @xmath667 denotes the investment rule of ( [ 2.11.a ] ) with @xmath668 . in this setting , the investment rule @xmath669 attains the infimum@xmath670 in ( [ 9.8 ] ) , and we obtain then @xmath671\\[-8pt ] \eqntext{\forall \mathcal{m } \in\mathfrak{m } ( \mathbf { x})}\end{aligned}\ ] ] on the strength of ( [ 9.7 ] ) . putting ( [ 9.9 ] ) and ( [ 9.10 ] ) together we deduce @xmath672 the saddle property of the pair @xmath673 .
in particular , the investment rule @xmath674 of ( [ 2.11.a ] ) with @xmath665 is seen to be the investor s best response to the least favorable admissible system @xmath675 of proposition [ proposition_1 ] , and vice - versa . in this sense
the investor , once he has figured out a least favorable admissible system @xmath334 , can allow himself the luxury to `` forget '' about model uncertainty and concentrate on finding an investment rule @xmath669 that satisfies @xmath676 , @xmath677 as in ( [ 9.9 ] ) , that is , on outperforming the market portfolio with the least initial capital within the context of the least favorable model @xmath678 .
we are grateful to professors jaka cvitani , nicolai krylov , mete soner , nizar touzi , hans fllmer , erhan bayraktar , johan tysk , and to the two reviewers , for helpful advice on the subject matter of this paper and for bringing relevant literature to our attention . numerous helpful discussions with drs .
robert fernholz , adrian banner , vasileios papathanakos , phi - long nguyen - thanh , and especially with johannes ruf , are gratefully acknowledged . | in an equity market model with `` knightian '' uncertainty regarding the relative risk and covariance structure of its assets , we characterize in several ways the highest return relative to the market that can be achieved using nonanticipative investment rules over a given time horizon , and under any admissible configuration of model parameters that might materialize .
one characterization is in terms of the smallest positive supersolution to a fully nonlinear parabolic partial differential equation of the hamilton
jacobi
bellman type . under appropriate conditions ,
this smallest supersolution is the value function of an associated stochastic control problem , namely , the maximal probability with which an auxiliary multidimensional diffusion process , controlled in a manner which affects both its drift and covariance structures , stays in the interior of the positive orthant through the end of the time - horizon .
this value function is also characterized in terms of a stochastic game , and can be used to generate an investment rule that realizes such best possible outperformance of the market . and . . |
the formation of galactic disks is one of the most important unsolved problems in astrophysics today . in the currently favored hierarchical clustering framework , disks form in the potential wells of dark matter halos as the baryonic material cools and collapses dissipatively .
fall & efstathiou ( 1980 ) have shown that disks formed in this way can be expected to possess the observed amount of angular momentum ( and therefore the observed spatial extent for a given mass and profile shape ) , but only under the condition that the infalling gas retain most of its original angular momentum . however , numerical simulations of this collapse scenario in the cold dark matter ( cdm ) cosmological context ( e.g. , navarro & benz 1991 , navarro & white 1994 , navarro , frenk , & white 1995 ) have so far consistently indicated that when only cooling processes are included the infalling gas loses too much angular momentum ( by over an order of magnitude ) and the resulting disks are accordingly much smaller than required by the observations .
this discrepancy is known as the _ angular momentum problem _ of disk galaxy formation .
it arises from the combination of the following two facts : a ) in the cdm scenario the magnitude of linear density fluctuations @xmath2 increases steadily with decreasing mass scale @xmath3 leading to the formation of non - linear , virialized structures at increasingly early epochs with decreasing mass i.e. the hierarchical `` bottom - up '' scenario .
b ) gas cooling is very efficient at early times due to gas densities being generally higher at high redshift as well as the rate of inverse compton cooling also increasing very rapidly with redshift .
a ) and b ) together lead to rapid condensation of small , dense gas clouds , which subsequently lose energy and ( orbital ) angular momentum by dynamical friction against the surrounding dark matter halo before they eventually merge to form the central disk .
a mechanism is therefore needed that prevents , or at least delays , the collapse of protogalactic gas clouds and allows the gas to preserve a larger fraction of its angular momentum as it settles into the disk .
two such possible solutions are discussed in section 2 . in section 3
we present some new results from our wdm disk galaxy formation simulations on the tully - fisher relation and in section 4 we discuss how the magnetic field strengths of a few @xmath4 g observed in galactic disks can be obtained via disk galaxy _ formation _ , as an alternative to disk dynamo amplification .
two ways of possibly solving the angular momentum problem have recently been discussed in the literature : a ) by invoking the effects of stellar feedback processes from either single , more or less uniformly distributed stars or star - bursts and b ) by assuming that the dark matter is `` warm '' rather than cold . both options lead to the suppression of the formation of early , small and dense gas clouds , for a ) because the small gas clouds may be disrupted due to the energetic feedback of primarily type ii super - nova explosions and for b ) simply because fewer of the small and dense gas clouds form in the first place for wdm free - streaming masses @xmath5-@xmath6 .
sommer - larsen et al .
( 1999 ) showed that the feedback caused by a putative , early epoch of more or less uniformly distributed population iii star formation was not sufficient to solve the angular momentum problem .
based on test simulations they showed , however , that effects of feedback from star - bursts in small and dense protogalactic clouds might do that .
preliminary results of more sophisticated simulations incorporating stellar feedback processes in detail indicate that this is at least partly the case .
considerable fine - tuning seems to be required , however : about 2 - 3% of the gas in the proto - galactic region of a forming disk galaxy should be turned into stars . if less stars are formed the feedback is not strong enough to cure the angular momentum problem and , vice versa , if more stars are formed during this fairly early phase of star - formation , the energetic feedback causes the formation of the main disks and thereby the bulk of the stars to be delayed too much compared to the observed star - formation history of the universe .
this requirement of fine - tuning is advantageous , however , in relation to the early chemical evolution of disk galaxies , since the early star - formation histories of the galaxies are then well constrained .
furthermore , as it is possible to track the elements produced and ejected by ( primarily ) type ii supernovae in the star - bursts one can determine the fraction of these elements , which ultimately settle on the forming disk and hence determine the rate and metallicity of the gas falling onto the disk . in figure 1
we show the time evolution of the oxygen abundance in a forming disk as a result of infall of a mixture of enriched and unenriched gas ( neglecting the contribution of ejecta from stars formed subsequently in the disk ) . we have assumed a salpeter imf with @xmath7 and @xmath8 and that a typical type ii supernova ejects @xmath9 of oxygen .
this abundance can be regarded as the initial abundance of the disk , its value depending on when star - formation subsequently commenced in the disk ( note that such two - epoch star - formation models have been advocated by , e.g. , chiappini , matteucci & gratton 1997 ) . as can be seen from the figure this initial disk abundance is of the order @xmath10
\sim -2 $ ] .
this is similar to the lowest abundance of the low - metallicity tail of the galactic thick disk see beers & sommer - larsen ( 1995 ) .
another , more radical way of solving the angular momentum problem is to abandon cdm altogether and assume instead that dark matter is `` warm '' .
such a rather dramatic measure not only proves very helpful in this respect , as will be discussed below , but may also be additionally motivated : recently , various possible shortcomings of the cdm cosmological scenario in relation to structure formation on galactic scales have been discussed in the literature : 1 ) cdm possibly leads to the formation of too many small galaxies relative to what is observed , i.e. the _ missing satellites problem _ ( e.g. , klypin et al .
2 ) even if galactic winds due to star - bursts can significantly reduce the number of visible dwarf galaxies formed , sufficiently many of the small and tightly bound dark matter systems left behind can still survive to the present day in the dark matter halos of larger galaxies like the milky way to possibly destroy the large , central disks via gravitational heating , as discussed by moore et al .
( 1999a ) .
3 ) the dark matter halos produced in cdm cosmological simulations tend to have central cusps with @xmath11 ( dubinski & carlberg 1991 , navarro et al .
1996 , fukushige & makino 1997 , moore et al .
1998 , kravtsov et al .
1998 , gelato & sommer - larsen 1999 ) .
this is in disagreement with the flat , central dark matter density profiles ( cores ) inferred from observations of the kinematics of dwarf and low surface brightness galaxies ( e.g. , burkert 1995 , de blok & mcgaugh 1997 , kravtsov et al .
1998 , moore et al . 1999b ,
but see also van den bosch et al . 1999 ) .
the first two problems may possibly be overcome by invoking warm dark matter ( wdm ) instead of cdm : on mass scales less than the free - streaming mass , @xmath12 , the growth of the initial density fluctuations in the universe is suppressed relative to cdm due to relativistic free - streaming of the warm dark matter particles . in conventional wdm theory
these become non - relativistic at redshifts @xmath13-@xmath14 for @xmath15 1 kev , which is the characteristic wdm particle mass required to give sub - galactic to galactic free - streaming masses . as a consequence of this suppression , fewer low mass galaxies ( or `` satellites '' )
are formed cf . ,
e.g. , moore et al .
( 1999a ) and sommer - larsen & dolgov ( 1999 , sd99 ) .
the central cusps problem may be more generic ( huss et al .
1999 and moore et al .
1999b ) , but wdm deserves further attention also on this point .
sd99 show that the angular momentum problem may be resolved by going from cold to warm dark matter , with characteristic free - streaming mass @xmath16-@xmath6 , and without having to invoke effects of stellar feedback processes at all .
the reason why this kind of warm dark matter leads to a solution of the angular momentum problem is that because of the suppression of density fluctuations on sub - galactic scales relative to cdm the formation of a disk galaxy becomes a much more coherent and gentle process enabling the infalling , disk - forming gas to retain much more of its original angular momentum .
in fact sd99 find it likely that the angular momentum problem can be _ completely _
resolved by going to the wdm structure formation scenario , which is more than can be said for the cdm+feedback approach so far . in figure 2
we show a face - on view of a disk galaxy with characteristic circular velocity ( where the rotation curve is approximately constant ) @xmath17 300 km / s formed in a wdm simulation ( in this simulation gas was not converted into stars ) .
clearly it is no longer a problem to form extended , high angular momentum disks in fully cosmological simulations . in comparison
the extent of typical disks formed in `` passive '' cdm simulations ( i.e. simulations not incorporating the effects of stellar feedback processes ) is less than 1 kpc
see , e.g. , sommer - larsen et al .
( 1999 ) . unlike the cdm+feedback solution
, one does not get a constraint on the early star - formation histories of the proto - galaxies , so no statements about the abundance of the first generation of disk stars can be made without further assumptions .
sd99 discuss possible physical candidates for wdm particles and find that the most promising are neutrinos with weaker or stronger interactions than normal , majorons ( light pseudogoldstone bosons ) , or mirror or shadow world neutrinos .
in figure 3 we show the cooled - out disk mass @xmath18 at redshift @xmath19=0 as a function of the characteristic circular velocity @xmath20 of model galaxies formed in our wdm simulations ( assuming a hubble parameter @xmath21=70 km / s / mpc ) .
also shown is the @xmath0-band tully - fisher ( tf ) relation of giovanelli et al .
( 1997 ) converted to mass assuming @xmath0-band mass - to - light ratios @xmath22=0.25 , 0.5 and 1.0 in solar units and @xmath21=70 km / s / mpc .
finally , the baryonic mass of the milky way , estimated in a completely independent way , is shown ( see sd99 for details ) .
as can be seen from the figure we can match the slope of the tf relation very well assuming a constant @xmath22 . to get the normalization right a @xmath23 is required .
sd99 argue that this is quite a reasonable value in comparison with various dynamical and spectrophotometric estimates .
moreover , it is clearly gratifying that the milky way data point falls right on top of the theoretical as well as observational @xmath18-@xmath20 relations ( for @xmath23 , @xmath21=70 km / s / kpc ) .
steinmetz & navarro ( 1999 ) and navarro & steinmetz ( 1999 ) find a discrepancy between the observed and `` theoretical '' tf on the basis of cdm simulations of disk galaxy formation .
it is hence possible that wdm helps out also on this point , but this has to be checked with more detailed simulations .
rgnvaldsson ( 1999 ) showed how the typical magnetic field strengths observed in galactic disks can be explained as a result of disk galaxy _ formation _ , as an alternative to the usually assumed dynamo amplification of an initially very weak magnetic field in the disk : hot , virialized gas ( @xmath24 k ) in a dark matter halo is assumed to initially follow the dark matter distribution and to be rotating slowly , corresponding to a spin - parameter @xmath25 0.05 , typical of galactic , dark matter halos .
the hot gas is assumed to be threaded by a weak and random magnetic field . as the hot gas cools radiatively ,
gravity forces it to flow inwards and due to the spin it forms a growing , cold , galactic disk in the central parts of the dark matter halo .
the magnetic field follows the cooling gas inwards and is strongly amplified by compression and shear in the forming disk .
rgnvaldsson ( 1999 ) carried out magnetohydrodynamical ( mhd ) simulations of this process using an eulerian mesh mhd code .
the simulations were run with various initial magnetic field strengths in the hot gas , parameterized by the initial ratio between the gas pressure and magnetic pressure @xmath26 .
the temporal evolution of the average magnetic field strength in the disk gas is shown in figure 4 . for weak initial fields
( @xmath27=100 - 400 was taken as a starting point , since these values are typical values in the hot , intergalactic gas in clusters of galaxies ) the average magnetic field strength grows gradually from about @xmath28=1 gyr ( after an initial relaxation phase ) .
the average values of 1 - 2@xmath4 g reached after about 5 gyr are quite reasonable for typical disk galaxies , indicating a route to the explanation of the magnetic field strengths observed in galactic disks alternative to the usual dynamo one .
another aspect of the growth of the field strength is reflected in the radial average in the disk , shown in figure 5 at various times for a simulation with @xmath27=400 .
the field strength is always highest in the outermost part of the growing disk , since the fieldlines brought in with the cooling flow are stacked on top of the already existing field there and the field is further amplified by the disk shear .
i have benefited from comments by rnlfur rgnvaldsson , sasha dolgov and jens schmalzing and thank the organizers for a magnificent conference . this work was supported by danmarks grundforskningsfond through its support for the establishment of the theoretical astrophysics center .
beers , t. c. , & sommer - larsen , j. 1995 , _ astrophys .
_ , * 96 * , 175 van den bosch , f. c. , et al . 1999 , _ astron . j. _ , submitted ( astro - ph/9911372 )
burkert , a. 1995 , _ astrophys .
j. _ , * 488 * , l55 de blok , w. j. g. , & mcgaugh , s. s. 1997 , _ mnras _ , * 290 * , 533 chiappini , c. , matteucci , f , & gratton , r. 1997 , _ astrophys . j. _ , * 477 * , 765 dubinski , j. , & carlberg , r. 1991 , _ astrophys . j. _ , * 378 * , 496 fall , s. m. , & efstathiou , g. 1980 , _ mnras _ , * 193 * , 189 fukushige , t. , & makino , j. 1997 , _ astrophys . j. _ , * 477 * , l9 gelato , s. , & sommer - larsen , j. 1999 , _ mnras _ , * 303 * , 321 giovanelli , r. , et al . 1999 , _
astrophys . j. _ , * 477 * , l1 huss , a. , jain , b. , & steinmetz , m. 1999 , _ astrophys .
j. _ , * 517 * , 64 klypin , a. , kravtsov , a. v. , valenzuela , o. , & prada , f. 1999 , _ astrophys .
j. _ , * 523 * , 32 kravtsov , a. v. , klypin , a. a. , bullock , j. s. , & primack , j. r. 1998 , _ astrophys . j. _ , * 502 * , 48 moore , b. , governato , f. , quinn , t. , stadel , j. , & lake , g. 1998 , _ astrophys . j. _ , * 499 * , l5 moore , b. , ghinga , s. , governato , f. , lake , g. , quinn , t. , stadel , j. , & tozzi , p. 1999a
, _ astrophys .
j. _ , * 524 * , l19 moore , b. , quinn , t. , governato , f. , stadel , j. , & lake , g. 1999b , _ mnras _ , submitted ( astro - ph/9903164 ) navarro , j. f. , & benz , w. 1991 , _ astrophys .
j. _ , * 380 * , 320 navarro , j. f. , & white , s. d. m. 1994 , _ mnras _ , * 267 * , 401 navarro , j. f. , frenk , c. s. , & white , s. d. m. 1995 , _ mnras _ , * 275 * , 56 navarro , j. f. , frenk , c. s. , & white , s. d. m. 1996 , _ mnras _ , * 462 * , 563 navarro , j. f. , & steinmetz , m. 1999 , _ astrophys .
, in press ( astro - ph/9908114 ) rgnvaldsson , . 1999 , phd thesis , university of copenhagen sommer - larsen , j. , gelato , s. , & vedel . h. 1999 , _ astrophys . j. _ , * 519 * , 501 sommer - larsen , j. , & dolgov , a. 1999 , _ astrophys . j. _ , submitted ( astro - ph/9912166 , sd99 ) steinmetz , m. , & navarro , j. f. 1999 , _ astrophys . j. _ , * 513 * , 555 | two ways of possibly solving the angular momentum problem plaguing cold dark matter ( cdm ) _ ab initio _ simulations of disk galaxy formation are discussed : 1 ) stellar feedback processes and 2 ) warm dark matter ( wdm ) rather than cdm . in relation to the chemical evolution of disk galaxies our simulations indicate that in case 1 ) the first generation of _ disk _ stars formed in disk galaxies like the milky way should have an abundance about two dex below solar , in fairly good agreement with the lowest observed abundance of the metal - weak tail of the galactic thick disk . for the second case
no such statements can be made without further assumptions about the star - formation history of the galaxies .
we find that the @xmath0-band tully - fisher relation can be matched by wdm disk galaxy formation simulations provided @xmath1 0.8 for disk galaxies , which sommer - larsen & dolgov ( 1999 ) argue is a reasonable value .
finally it is discussed how the magnetic field strengths observed in galactic disks can be obtained through disk galaxy _ formation _ , as an alternative to the conventional dynamo hypothesis . |
observations over the past decades have clearly shown that satellite galaxies ( sgs ) are not randomly distributed , but rather are anisotropically distributed around centrals . this characteristic is observed from our milky way galaxy ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , the neighboring m31 @xcite , to large samples of local galaxies , and even in the virgo cluster@xcite . in particular , both the results of 2dfgrs and sloan digital sky survey ( sdss ) have shown that satellites are preferentially distributed along the major axes of centrals .
this phenomenon is known as galaxy alignment ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the alignment strength also depends on the properties of both the satellites and centrals such that red satellites show stronger alignment with centrals than blue satellites , and red centrals have stronger alignment with their satellites than blue centrals .
such an alignment is also observed for high - redshift galaxies @xcite .
several groups have used theoretical arguments and numerical work to explain the origin of this alignment .
studies of the milky way and m31 have focused on the nature of the thin disk - like configuration of the satellites ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , but debate exists concerning the rareness of such a distribution in the cdm model due to the limited number of satellites and host systems observed ( e.g. , * ? ? ?
* ; * ? ? ?
more converged conclusions are reached in the studies of galaxy alignment found in the sdss survey ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the preferential distribution of satellites along the major axes of centrals is found to be common in the cold dark matter ( cdm ) model and arises from the non - spherical nature of dark halos ( e.g. , * ? ? ?
* ) . however , most studies so far have used @xmath0-body simulations where one has to make assumptions about how the shapes of central galaxies ( cgs ) are related to the shapes of their host halos . in most cases ,
the predicted alignment signal is larger than observed if the cg shape follows the overall shape of the dark halo .
furthermore , to explain the dependence of alignment on the galaxy properties , one needs to adopt different assumptions for the shape of centrals with blue and red color @xcite .
to directly predict the alignment signal , one should use simulations which can self - consistently model the shapes of the centrals and the distributions of the satellites ( rather than the sub - halos ) .
@xcite used @xmath0-body simulations to study the dependence of alignment strength on halo assembly history .
due to the limited resolution and the lack of gas physics , most sub - halos suffered from strong tidal stripping and rarely survived in the central regions of the host halos . in this paper
, we will study galaxy alignment using a smoothed - particle - hydrodynamics ( sph ) simulation which includes gas cooling , metal recycling , star formation , and supernova feedback .
as the stellar component is included , the shape of the cg is properly determined by the physics of gravity
. the color of galaxies can be easily obtained from the star formation history and metal evolution from the simulation , using a stellar population synthesis model and an assumed stellar initial mass function .
compared to @xmath0-body simulations , sph simulations typically resolve more sgs in the halo center , enabling the study of galaxy alignment on smaller scales .
this is because the gas cooling and subsequent star formation results in a more compact mass distribution ( than dark matter only simulations ) which is better able to survive the strong tidal force in the host halo @xcite .
there are a few studies which used sph simulations to explore the galaxy alignment .
for example , libeskind et al . ( 2007 ) used high - resolution simulations of nine milky way like halos to study the spatial distribution of satellites , but mainly focused on their spatial configuration ( polar disk ) .
deason et al .
( 2011 ) have investigated the alignment between the satellites positions and cg using zoomed - in simulations , however , they focused on low - mass halos and did not explore the dependence of alignment on satellite properties or compare with the data .
recently , tenneti et al . (
2014 ) utilized a high - resolution simulations with active galactic nucleus ( agn ) feedback , but they only studied the shape correlation between dark matter halos and the stellar component .
in this study , we will focus on the galaxy alignment with a dependence on the galaxy properties directly from our simulation , and also compare the model results with observational data ( yang et al .
2006 ) to understand what is the origin of the observed dependence .
the cosmological simulation used in this paper was run using the non - public version ( including gas physics ) of the massive parallel code gadget-2 @xcite .
it is evolved from redshift @xmath1 to the present epoch in a cubic box of @xmath2 with @xmath3 of dark matter and gas particles , assuming a flat @xmath4 `` concordance '' cosmology with @xmath5 , @xmath6 , @xmath7 , and @xmath8 .
a plummer softening length of @xmath9 was adopted .
each dark matter particle has a mass of @xmath10 .
the initial mass of gas particles is @xmath11 and one gas particle can turn into two star particles later on .
the simulation includes the processes of radiative cooling , star formation , supernova feedback , and outflows by galactic winds , as well as a sub - resolution multiphase model for the interstellar medium .
the readers are referred to @xcite for more details about the treatment of gas physics .
dark matter halos were found using the standard friends - of - friends ( fof ) algorithm with a linking length of 0.2 times the mean particle separation , while the ` galaxies ' were defined as the stellar fof groups with a linking length of @xmath12 ( c.f .
the cg was identified as the most massive galaxy ( in stars ) in the fof halo ; we call the other galaxies in the halo sgs . to ensure the accurate determination of the shape of centrals and to exclude spurious satellites , only those centrals and satellites with more than @xmath13 and @xmath14 star particles , respectively , are included in our sample .
for each star particle , we record its formation time and the metallicity ( [ fe / h ] ) of the cold gas at that time .
the luminosity of each star particle at @xmath15 is calculated using the stellar population synthesis model of @xcite assuming the initial mass function from @xcite .
the sum of all the stellar particles gives the total luminosity of each galaxy . in our simulation
, we calculate the magnitudes in the sdss @xmath16 and @xmath17 bands .
figure 1 shows the predicted color - stellar mass diagram from our simulation .
the left and right panels are for the centrals and satellites , respectively . here , @xmath18 is the galaxy color using filters redshifted to @xmath19 and computed from @xmath20 using the method described in @xcite .
here we do not apply any dust extinction to the model galaxies .
the solid line is the fit to the sdss data by @xcite which divides galaxies into red and blue .
compared to observations , the massive centrals in the simulation are too blue .
this is a well - known problem of galaxy formation models ; without additional feedback , the gas cooling and star formation in massive halos is too efficient , resulting in massive centrals that are too blue ( e.g. , * ? ? ?
* ) . with the inclusion of agn feedback
, the galaxy stellar masses and colors in massive halos can be reproduced @xcite .
the right panel shows that the majority of satellites in our simulation are red .
it is known that the main mechanisms for halting star formation in satellites are ram - pressure stripping and the intrinsic starvation by star formation using up the available gas , both of which are included in our simulations .
yang et al .
( 2006 ) studied the dependence of galaxy alignment on color .
they simply adopted @xmath21 to divide galaxies into red and blue .
in fact , it is more reasonable that the division between red and blue galaxies is stellar mass dependent ( e.g. , baldry et al .
2006 ; van den bosch et al.2008 ; kang & van den bosch 2008 ) .
thus , in our work , we divide the galaxies into red and blue using the fitting line of van den bosch et al .
( 2008 ; the solid line in figure 1 ) .
to compare with the data more consistently , we measure the alignment signal in projection .
we project the model galaxies along one axis of the simulation box and obtain the reduced inertia tensor of cgs from their stellar particles with @xmath22 where @xmath23 is the position of the @xmath24 th stellar particle in the halo center coordinates and @xmath25 .
the eigenvectors of @xmath26 define the orientation of the galaxy , and with this form of the inertia tensor the major axis corresponds to the vector with the large eigenvalue .
the angle @xmath27 ( @xmath28 ) is defined as the angle on the projection plane between the off - center vector of the satellite position and the major axis of the cg .
the distribution probability of @xmath27 is defined as @xmath29 where @xmath30 and @xmath31 are the number of central - satellite pairs in the simulation sample and random samples .
an alignment is found if @xmath32 is less than @xmath33 .
the bootstrapping method is used to derive @xmath31 by randomly selecting 100 sub - samples from the overall data .
the predicted alignment and comparisons with data are shown in figure [ compare_yang2006 ] .
the triangles with solid lines are the observational results of yang et al .
( 2006 ) and model results are plotted with circles connected by dashed lines .
the upper left panel shows that the predicted alignment effect for all galaxies agrees well with the data . for satellites ,
the model predictions roughly agree with the data , but slightly overpredict the alignment for blue satellites and underpredict the alignment for red satellites .
however , the predicted alignment for centrals is inconsistent with the data .
the lower left and middle panels show that the blue centrals have stronger alignment than the red ones , contrary to the observed dependence on color .
as seen in figure 1 and the discussion above , there are too many massive blue galaxies in our simulation due to absence of an effective quenching mechanism in massive haloes ( such as agn feedback ) .
it is commonly believed that these massive galaxies should be red and exist in massive halos .
both theoretical and observational works suggest that there should be a critical halo mass below which most of the centrals are blue , and above which most are red @xcite
. however , this critical halo mass is not well determined , ranging from @xmath34 ( e.g. , * ? ? ?
* ) to @xmath35 ( e.g. , * ? ? ?
* ) . to simply illustrate this effect , in the lower right panel ,
we show the predicted alignment for blue centrals that have a halo mass less than @xmath36 , where @xmath37 ( dotted line ) and @xmath38 ( dashed line ) .
it is found that the predicted alignment for blue centrals is close to the data with @xmath39 , and it increases with increasing critical halo mass . to avoid overlapping
, we do not plot the lines for red centrals whose halo mass are larger than @xmath40 .
the results are close to the observational one and the lines for the two critical mass are almost identical , since massive halos have more satellites than low - mass halos and thus dominate the signal .
the exercise presented in the lower right panel of figure 2 suggests that the observed alignment dependence on cg color is mainly determined by the host halo mass .
later , we will see that the halo mass dependence is rooted in the correlation between the shapes of the cg and the host halo . in figure [ dependence ]
, we further show the dependence of alignment on galaxy properties ( metallicity , color , stellar mass , and halo mass ) from the simulation .
the solid and dashed lines are for the centrals and satellites , respectively .
the upper panels show that the satellite alignment depends on the metallicity and color , with a stronger dependence on metallicity for which metal - rich satellites have very strong alignment .
the lower left panel shows that the dependence on stellar mass is weak and fainter satellites have slightly weaker alignment , in broad agreement with the finding in y06 .
note that the error - bar for the point at the greatest mass bin is large due to the small number statistics .
the halo mass dependence in the lower right panel is close to the dependence from the data there is stronger alignment in more massive halos , which is also consistent with the dependence on stellar mass for centrals in the left panel .
in addition to the dependence of galaxy alignment on color , observations have shown that the alignment angle is a strong function of radial distance to the centrals @xcite .
in the left panel of figure [ alignment_bcg_halo ] , we show the radial spatial distribution of satellites in the dark halo , together with the dependence on color and metallicty . it is found that both metal - rich ( top 30% by order ranking ) and red satellites are distributed predominately in the central halo . in general
, this distribution agrees with the observational facts that galaxy properties , such as color , metallicity , or morphology , depend strongly on environment / local density ( e.g. , wang et al .
2014b and reference therein ) .
such trends are expected since metal recycling and star formation quenching are more efficient in the inner halo @xcite .
the middle panel in figure [ alignment_bcg_halo ] shows the alignment of satellites as a function of the radius .
the observational results of y06 are shown with triangles and a solid line , which show that satellites residing in the inner halo have stronger alignment with the cg than their counterparts residing in the outer halo .
our simulation results ( blue circles and dashed line ) are in good agreement with these observations . to understand the origin of satellite alignment with respect to the cg , theoretical work using @xmath0-body simulations often assumed that the shape of the cg follows the shape of the overall dark matter halo .
however , this assumption leads to a stronger alignment than observed ( e.g. , * ? ? ?
to decrease the predicted signal , one has to introduce some degree of mis - alignment between the cg and that of the dark matter halo @xcite .
a more physical solution is that the cg follows the shape of dark matter in the central region .
this model is better at reproducing the alignment @xcite , however , the dependence on galaxy color has not been reproduced .
as our sph simulation includes the stellar component , we can directly predict the shape of the cg and test the assumption .
the right panel of figure [ alignment_bcg_halo ] shows the average alignment angle between the major axis of the cg and the shape of the dark matter halo as a function of the halo mass .
results are shown for halo shapes calculated within three radii : for the `` whole halo '' inside of @xmath41 ( blue solid ) , the `` intermediate halo '' inside of @xmath42 ( green dashed ) , and the `` inner halo '' inside of @xmath43 ( red dotted ) . here , @xmath41 is the spherically averaged virial radius .
it is found that the shape of the cg better traces that of the inner halo , and this alignment increases with halo mass .
the mean mis - alignment angle varies from @xmath44 for halos with a mass of @xmath45 .
similar results are also found in recent work ( e.g , * ? ? ?
the results in figure [ alignment_bcg_halo ] clearly explain the observed galaxy alignment dependence on galaxy properties .
the most metal rich and reddest satellites are distributed in the inner region of the host halo , and their spatial distribution should closely follow that of dark matter . on the other hand ,
the stellar component of centrals is also greatly shaped by the gravitational force of the dark matter in the inner halo .
the combination of these two effects leads to a better alignment for the metal - rich / red satellites than their metal - poor / blue counterparts . as to the dependence on color of centrals ,
this is related to the halo mass of the centrals bluer centrals most likely reside in relatively lower - mass halos where the alignment between the central stellar component and the inner halo shape becomes weaker .
in this letter , we carry out a study of galaxy alignment using a cosmological simulation including gas cooling , star formation , and supernova feedback , which enables a direct prediction for the shape of cgs and the galaxy properties .
we find that the predicted alignment between the cg and the distribution of satellites agrees with the observations .
furthermore , with a simple assumption about the halo mass of blue and red centrals , the dependence on color for both centrals and satellites is also reproduced .
we also identify that the strongest dependence of the alignment is with metallicity of satellites , which should be testable using future data .
the main source of galaxy alignment is the non - spherical nature of cdm halos , as shown by many previous studies ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
however , the predicted strength of the alignment is too strong if the shape of the cg follows the overall shape of the dark matter halo . from our study
, we find that the shape of the cg better follows the halo in the inner region , and the average mis - alignment is about @xmath46 ( see figure [ alignment_bcg_halo ] ) , similar to the expected or inferred values in previous studies ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . as the most red / metal - rich satellites stay in the inner halo , they naturally follow the shape of the dark matter halo in that region .
this leads to a strong alignment between red satellites with centrals .
furthermore , as the alignment between the cg and inner halo increases with halo mass and red centrals predominately populate massive halos , it explained the observed fact that red central shows stronger alignment with satellites than blue centrals .
although the prediction for the alignment of blue centrals using our simulation fails because of the too blue colors of the most massive central galaxies , the exercises for the alignment dependence on halo mass have given hints that simulations with agn feedback ( e.g. , * ? ? ?
* ; * ? ? ?
* ) should be helpful to solve this problem .
the non - spherical nature of dark matter halos is one the most prominent features of structure formation in a cdm universe , as the mass accretion and mergers predominately occur along the cosmic web or the filament ( e.g. , * ? ? ?
it also naturally produces the galaxy alignment on very large scales up to @xmath47 @xcite .
accurate predictions for galaxy alignment on large scales is crucial to cosmological applications , such as estimating the systematic error used in weak lensing measurements . with the proper modeling of galaxy shapes from hydrodynamical simulation
, we will be able to make predictions for galaxy alignment on large scales in a forthcoming paper .
the authors thank the anonymous referee for useful suggestions .
w.p.l . and x.k .
acknowledge supports by the nsfc projects ( no .
11121062 , 11233005 , u1331201 , 11333008 ) and the `` strategic priority research program the emergence of cosmological structures '' of the chinese academy of sciences ( grant no .
xdb09010000 ) .
is also supported by the national basic research program of china ( 2013cb834900 ) and the foundation for distinguished young scholars of jiangsu province ( no .
bk20140050 ) .
the simulations were run in the shanghai supercomputer center and the data analysis was performed on the supercomputing platform of shanghai astronomical observatory .
, a.a.d . and a.v.m .
, acknowledge support from the mpg - cas through the partnership program between the mpia group lead by a. macci and the pmo group lead by x. kang . | observations have shown that the spatial distribution of satellite galaxies is not random , but rather is aligned with the major axes of central galaxies ( cgs ) .
the strength of the alignment is dependent on the properties of both the satellites and centrals .
theoretical studies using dissipationless @xmath0-body simulations are limited by their inability to directly predict the shape of cgs .
using hydrodynamical simulations including gas cooling , star formation , and feedback , we carry out a study of galaxy alignment and its dependence on the galaxy properties predicted directly from the simulations.we found that the observed alignment signal is well produced , as is the color dependence : red satellites and red centrals both show stronger alignments than their blue counterparts .
the reason for the stronger alignment of red satellites is that most of them stay in the inner region of the dark matter halo where the shape of the cg better traces the dark matter distribution .
the dependence of alignment on the color of cgs arises from the halo mass dependence , since the alignment between the shape of the central stellar component and the inner halo increases with halo mass .
we also find that the alignment of satellites is most strongly dependent on their metallicity , suggesting that the metallicity of satellites , rather than color , is a better tracer of galaxy alignment on small scales .
this could be tested in future observational studies . |
in galactic contexts , we venture to formulate a theoretical magnetohydrodynamic ( mhd ) disc problem to explore possible large - scale structures and dynamics of stationary mhd density waves in a composite system of stellar and magnetized interstellar medium ( ism ) gas discs .
the two gravitationally coupled discs are treated as ` fluid ' and ` magnetofluid ' respectively and are both expediently approximated as razor - thin singular isothermal discs ( sids ) with the gaseous sid being embedded with a coplanar azimuthal magnetic field . for the gravitational effect of a massive axisymmetric dark matter halo , we prescribe a background composite system of two coupled partial sids ( syer & tremaine 1996 ; shu et al . 2000 ;
lou 2002 ; lou & shen 2003 ; shen & lou 2003 ) . in our model analysis , we construct stationary aligned and unaligned logarithmic spiral mhd perturbation configurations in a composite system of two sids with flat rotation curves , and attempt to relate various morphologies of disc galaxies , including barred and lopsided , barred and normal spiral structures . for possible observational diagnostics , we derive phase relationships among perturbation patterns of the stellar surface mass density , the gas surface mass density and the azimuthal magnetic field .
this introduction serves two purposes .
the first one is to provide a general background information relevant to the problem at hand and the second one is to give reasons of pursuing this mhd disc problem . in a pioneering study of a composite system of stellar and gas discs coupled by gravity ,
lin & shu ( 1966 , 1968 ) used a stellar distribution function and a gas fluid disc description to derive and analyze the local dispersion relation of galactic spiral density waves .
since then , there have been extensive theoretical studies on perturbation configurations and stability properties of a composite disc system , mainly in galactic contexts .
kato ( 1972 ) investigated oscillations and overstabilities of density waves using a formalism similar to that of lin & shu ( 1966 , 1968 ) . in a two - fluid formalism ,
jog & solomon ( 1984a , b ) examined the growth of local axisymmetric perturbations in a composite disc system .
bertin & romeo ( 1988 ) studied the influence of a gas disc on spiral modes in a two - fluid model framework .
vandervoort ( 1991a , b ) studied the influence of interstellar gas on oscillations and stabilities of spheroidal galaxies .
the two - fluid approach was also adopted in a stability study of a two - component disc system with finite disc thickness by romeo ( 1992 ) .
the analysis for morphologies of disc galaxies was performed by lowe et al .
for the stability of a composite disc system , different effective @xmath13 parameters ( safronov 1960 ; toomre 1964 ) have been suggested using a two - fluid formalism by elmegreen ( 1995 ) and jog ( 1996 ) .
recently , lou & fan ( 1998b ) used the two - fluid formalism to study properties of open and tight - winding spiral density - wave modes in a composite disc system .
lou & shen ( 2003 ) studied stationary global perturbation structures in a two - fluid system of sids and , instead of a redefinition of a different @xmath13 parameter , shen & lou ( 2003 ) offered a more practical @xmath14criterion for the axisymmetric instability in a composite sid system .
a rich class of disc problems involves stability properties of sids .
there have been numerous studies on this subject since the pioneering work of mestel ( 1963 ) ( e.g. zang 1976 ; toomre 1977 ; lemos , kalnajs & lynden - bell 1991 ; lynden - bell & lemos 1999 ; goodman & evans 1999 ; charkrabarti , laughlin & shu 2003 ) .
specifically , syer & tremaine ( 1996 ) made an important breakthrough to derive semi - analytic solutions for stationary perturbation configurations in a class of sids .
shu et al . ( 2000 ) obtained stationary solutions for perturbation configurations in an isopedically magnetized sid with a flat rotation curve . through numerical explorations
, they interpreted these stationary aligned and unaligned logarithmic spiral configurations as onsets of bar - type and barred - spiral instabilities ( see also galli et al .
different from yet complementary to the analysis of shu et al .
( 2000 ) , lou ( 2002 ) performed a coplanar mhd perturbation analysis in a single background sid embedded with an azimuthal magnetic field , from the perspective of stationary fast and slow mhd density waves ( fmdws and smdws ; fan & lou 1996 ; lou & fan 1998a ) .
lou ( 2002 ) also derived a form of magnetic virial theorem for an msid and suggested the ratio of rotation energy to the sum of gravitational and magnetic energies to be crucial for the onset of bar - like instability in an msid system . in galactic contexts , it would be more realistic to consider large - scale structures and dynamics in a composite system of stellar and magnetized ism discs . as a first step ,
lou & shen ( 2003 ) made a foray on this model problem , constructed stationary aligned and unaligned logarithmic spiral configurations in such a composite sid system and further examined axisymmetric instability properties ( shen & lou 2003 ) . in disc galaxies , the ism disc is magnetized with the magnetic energy density being comparable to the energy densities of thermal gas and of relativistic cosmic - ray gas ( e.g. lou & fan 2003 ) .
information of galactic magnetic fields can be estimated by synchrotron radio emissions from spiral galaxies .
for such a magnetized composite system , mhd will play an indispensable role and reveal more realistic aspects of dynamic and diagnostic information . these important problems ( shu et al .
2000 ; lou 2002 ; lou & shen 2003 ) are not only interesting by themselves , but also serve as necessary steps for establishing an even more realistic model . motivated by this prospect ( lou & fan 1998b ; lou 2002 ; lou & shen 2003 )
, we construct here stationary perturbation configurations for aligned and unaligned logarithmic spiral cases in a composite system of a stellar sid and a coplanarly magnetized gaseous msid , and discuss their stability properties .
we adopt a relatively simple formalism for a composite system involving fluid and magnetofluid discs coupled by gravity .
we provide in section 2 an mhd description for the coplanarly magnetized gaseous msid , obtain conditions for background axisymmetric equilibrium state for both stellar sid and gaseous msid , and derive linearized equations for coplanar perturbations . there
exist aligned and unaligned classes of global mhd perturbation solutions ; they are analyzed in section 3 and section 4 , respectively .
the exact solutions of stationary perturbations , their stability properties and their corresponding phase relationships among perturbation variables are examined and summarized in section 5 .
details are included in appendices a@xmath15e .
it would be physically more precise to adopt a distribution function formalism in dealing with a stellar disc especially in terms of singularities and resonances ( e.g. lin & shu 1966 , 1968 ; binney & tremaine 1987 ) . for the present purpose of modelling large - scale stationary perturbation structures and for mathematical simplicity ( lou & shen 2003 ) , it suffices to start with the fluid - magnetofluid formalism , including an mhd treatment for the gaseous msid ( lou 2002 ; lou & fan 2003 ) . in this section
, we present the basic equations for the fluid - magnetofluid system consisting of a stellar sid and a gaseous msid . for flat rotation curves , conditions on the background rotational equilibrium with axisymmetry can be derived .
we then obtain the linearized equations for coplanar mhd perturbations in the composite msid system .
the two sids , located at @xmath16 , are both approximated as infinitesimally thin . in our fluid - magnetofluid treatment ,
the two sids are coupled through mutual gravitational interaction . for large - scale stationary perturbations , diffusive processes such as viscosity , ambipolar diffusion and thermal diffusion etc .
are ignored . for physical variables under consideration , we shall use superscript or subscript @xmath17 to indicate an association with the stellar sid and superscript or subscript @xmath18 to indicate an association with the gaseous msid . in cylindrical coordinates
( @xmath19 , @xmath20 , @xmath21 ) , the basic fluid - magnetofluid equations for a composite msid system can be readily written out . in the fluid approximation for a stellar sid , the mass conservation , the radial component of the momentum equation and the azimuthal component of the momentum equation
are given below in order , namely @xmath22 @xmath23 @xmath24 where @xmath25 is the radial component of the stellar bulk velocity , @xmath26 is the stellar specific angular momentum along the @xmath27 direction ; and @xmath28 is the azimuthal component of the stellar bulk velocity ; @xmath29 is the total gravitational potential , @xmath30 is the vertically integrated ( effective ) pressure ( sometimes referred to as the two - dimensional pressure ) , and @xmath31 is the vertically integrated stellar mass density ( i.e. stellar surface mass density ) . in the magnetofluid approximation for the gaseous msid , the mass conservation , the radial component of the momentum equation and the azimuthal component of the momentum equation are given below in order , namely @xmath32 @xmath33\ , \quad \end{split}\ ] ] @xmath34\ , \end{split}\ ] ] where @xmath35 is the radial component of the gas bulk velocity , @xmath36 is the gas specific angular momentum along the @xmath27 direction , @xmath37 is the azimuthal component of the gas bulk velocity , @xmath38 is the vertically integrated gas pressure ( sometimes referred to as the two - dimensional gas pressure ) , @xmath39 is the vertically integrated gas mass density ( i.e. gas surface mass density ) and @xmath40 and @xmath41 are the radial and azimuthal components of magnetic field @xmath42 .
the last two terms on the right - hand sides of equations ( [ 5 ] ) and ( [ 6 ] ) are the radial and azimuthal components of the lorentz force due to the coplanar magnetic field .
the coupling of the two sets of fluid and magnetofluid equations @xmath43 and @xmath44 is effected by the total gravitational potential @xmath29 through the poisson integral , namely @xmath45^{1/2}}\ , \ ] ] where @xmath46 is a constant ratio with @xmath47 ; @xmath48 is the gravitational potential from the stellar sid and gaseous msid together , and the fraction @xmath49 is attributed to an axisymmetric dark matter halo that is unresponsive to coplanar mhd perturbations in the composite msid system ( e.g. shu et al . 2000 ) . the divergence - free condition for the coplanar magnetic field @xmath50 is @xmath51 and the radial and azimuthal components of magnetic induction equation are @xmath52 @xmath53 equations @xmath54 form the basis of our theoretical analysis . for a flat rotation curve in a stellar sid
, we write the corresponding angular rotation rate in the form of @xmath55 where @xmath56 , mimicking an ` effective isothermal sound speed ' , represents the velocity dispersion of the stellar sid and @xmath57 is a dimensionless parameter for stellar sid rotation .
we invoke the expedient polytropic approximation @xmath58 to relate the two - dimensional pressure and the surface mass density .
the epicyclic frequency @xmath59 of the stellar disk is defined by @xmath60 in parallel , we have the angular rotation rate of the gaseous msid as @xmath61 where @xmath62 is the isothermal sound speed of the gaseous msid and @xmath63 is a dimensionless parameter for msid rotation .
the polytropic relation and the definition of the epicyclic frequency @xmath64 for the gaseous msid are simply @xmath65 @xmath66 to avoid the magnetic field winding dilemma in a rotating disc ( e.g. lou & fan 1998a ) , the background coplanar magnetic field , which is not force - free , is taken to be purely azimuthal about the symmetry
@xmath67axis : @xmath68 where @xmath69 is a constant ( lou 2002 ) proportional to the encircled magnetic flux within @xmath19 , and @xmath70 in galactic model applications , one needs to invoke a central bulge or other processes to avoid the divergence of @xmath71 as @xmath72 . from the radial momemtrum equations ( [ 2 ] ) and ( [ 5 ] ) in a rotational equilibrium and the fact that @xmath73 by poisson integral ( [ 7 ] ) , one derives the following expressions for the background surface mass densities , namely @xmath74 @xmath75 where @xmath76 is the surface mass density ratio of the two coupled background sids and @xmath77 is the alfvn wave speed in the msid defined by @xmath78 from equations ( [ 20 ] ) and ( [ 21 ] ) , it then follows that @xmath79 physically , condition ( [ 86 ] ) results from the basic fact that the same total gravitational force @xmath80 , including the contribution from the dark matter halo , acts on both the stellar sid and magnetized gaseous sid , and is very useful in our analysis below .
it should be noted that the rotation rates of the two sids are different in general ( lou & shen 2003 ; shen & lou 2003 ) . in dimensionless form ,
condition ( [ 86 ] ) can be written in the form of either @xmath81 or @xmath82 where parameter @xmath83 stands for the square of the ratio of the stellar velocity dispersion to the sound speed of the msid , and parameter @xmath84 stands for the square of the ratio of the alfvn speed to the sound speed in the msid . in disc galaxies ,
the stellar velocity dispersion @xmath56 is usually higher than the sound speed @xmath62 , we naturally focus on the case of @xmath85 ( e.g. jog & solomon 1984a , b ; bertin & romeo 1988 ; jog 1996 ; elmegreen 1995 ; lou & fan 1998b ; lou & shen 2003 ; shen & lou 2003 ) . in this @xmath86 regime
, it follows from condition ( [ 90 ] ) that @xmath87 implying that for @xmath85 , @xmath88 inequality ( [ 93 ] ) is very important to identify physically valid mathematical solutions of @xmath89 for stationary mhd perturbations . for specified parameters @xmath3 and @xmath90 , @xmath89 and @xmath91 are related to each other linearly by condition ( [ 87 ] ) .
we emphasize that mathematical solutions of @xmath91 and @xmath89 become unphysical for either @xmath92 or @xmath93 or both .
the key here is that by inequality ( [ 93 ] ) , we only need to consider @xmath94 because @xmath91 must also be positive . in our analysis , we mainly use equation ( [ 87 ] ) to derive a cubic algebraic equation in terms of @xmath89 and examine solution properties . for small coplanar mhd perturbations in a composite msid system , basic nonlinear equations ( [ 1])@xmath15([10 ] )
can be linearized in a straightforward manner , namely @xmath95 @xmath96 @xmath97 for coplanar hydrodynamic perturbations in a stellar sid , and @xmath98 @xmath99 \\
+ \frac{c_a^2\sigma^g_1}{2\sigma^g_0r } -\frac1{\sigma^g_0}\int\frac{dzb_\theta } { 4\pi r}\frac{\partial ( rb_\theta)}{\partial r}\ , \end{split}\ ] ] @xmath100 for coplanar mhd perturbations in a gaseous msid , @xmath101^{1/2}}\ ] ] for the linearized poisson integral , and @xmath102 @xmath103 @xmath104 for the linearized divergence - free condition and the linearized magnetic induction equation .
we do not consider vertical variations along @xmath21 direction across the composite msid system . with a harmonic @xmath105 dependence for all perturbation variables ,
we introduce complex radial variations @xmath106 ,
@xmath107 , @xmath108 , @xmath109 , @xmath110 , @xmath111 , @xmath112 , @xmath113 and @xmath114 for @xmath115 , @xmath116 , @xmath117 , @xmath118 , @xmath119 , @xmath120 , @xmath121 , @xmath122 and @xmath123 , respectively .
thus , hydrodynamic equations ( [ 23])@xmath15([25 ] ) can be reduced to the form of @xmath124 @xmath125 @xmath126 for the stellar sid , where @xmath127 .
similarly , mhd equations ( [ 26])@xmath15([32 ] ) can be cast into the form of @xmath128 @xmath129\ , \end{split}\ ] ] @xmath130 where @xmath131 , @xmath132 @xmath133 @xmath134 for coplanar mhd perturbations in the gaseous msid . by setting angular frequency @xmath135 in coplanar mhd perturbation equations ( [ 29 ] ) and ( [ 33])@xmath15([41 ] )
, we can construct global stationary mhd perturbation configurations in a composite system of msids without invoking the wkbj or tight - winding approximation and analyze their properties .
coplanar perturbations in a composite msid system can be classified as ` aligned ' and ` unaligned ' solutions ( e.g. kalnajs 1973 ; shu et al .
2000 ; lou 2002 ; lou & shen 2003 ) . for aligned configurations , all streamlines and magnetic field lines are aligned in a composite msid system . for unaligned spiral configurations , neighbouring streamlines shift relative to each other in a systematic manner ( kalnajs 1973 )
; the same physical scenario holds true for neighbouring magnetic field lines ( lou & fan 1998a ) . in this section
, we obtain the stationary dispersion relation for aligned coplanar mhd perturbations ( both full and partial msids ) by perturbation equations ( [ 29 ] ) and ( [ 33])@xmath15([41 ] ) in the preceding section .
the solution behaviours and the corresponding phase relationships are analyzed , mainly in the context of a full sid system ( i.e. @xmath136 ) . at the end of this section
, we derive the mhd virial theorem for a composite msid system and suggest the onset criterion for secular bar - like instabilities in a composite msid system ( ostriker & peebles 1973 ; binney & tremaine 1987 ; shu et al .
2000 ; lou 2002 ) . to construct stationary perturbation configurations of msid that are aligned , we set @xmath135 in equations ( [ 33])@xmath15([41 ] ) .
let us first set @xmath137 or @xmath135 in equations ( [ 36])@xmath15([41 ] ) for coplanar mhd perturbations in the msid to obtain @xmath138 @xmath139,\quad \end{split}\ ] ] @xmath140 @xmath141 @xmath142 for aligned perturbations , we take the following potential - density pair ( shu et al . 2000 ; lou 2002 ; lou & shen 2003 ) @xmath143 @xmath144 @xmath145 such that @xmath146 for a constant @xmath147 , as will be shown presently , combinations of equations ( [ 63 ] ) and ( [ 65 ] ) and expressions ( [ 47])@xmath15([49 ] ) give @xmath148 @xmath149 as @xmath150 is constant by equation ( [ 50 ] ) , it is clear that @xmath147 is another constant .
consequently , equations ( [ 68 ] ) and ( [ 64 ] ) give @xmath151iu_g -\frac{2\omega_g^2r^2\mu_g}{\sigma_0^g } + \frac{c_a^2\mu_g}{2\sigma_0^g}=0\ .\ ] ] substitutions of expressions ( [ 49 ] ) , ( [ 68 ] ) and ( [ 69 ] ) into equation ( [ 70 ] ) give a relation between @xmath152 and @xmath153 , namely @xmath154 where two coefficients @xmath155 and @xmath156 are defined by @xmath157 and @xmath158 respectively . in parallel , we set @xmath135 in equations ( [ 33])@xmath15([35 ] ) for coplanar perturbations in the stellar disc and use relations ( [ 47])@xmath15([49 ] ) to obtain ) can also be obtained by simply setting @xmath159 in equation ( [ 43 ] ) and switching subscripts @xmath18 and @xmath17 . ] another relation between @xmath153 and @xmath152 , namely @xmath160 combining equations ( [ 42 ] ) and ( [ 43 ] ) , we derive the stationary dispersion relation @xmath161 \qquad\qquad\qquad\\ \times\bigg\{m^4\omega_g^4-\bigg[2\omega_g^2 + \bigg(\frac{c_a^2+a_g^2}{r^2 } -\frac{2\pi g\sigma^g_0}{|m|r}\bigg)m^2 -\frac{2c_a^2}{r^2}\bigg ] \qquad\\ \times m^2\omega_g^2 + \frac{m^2c_a^2}{r^2}\bigg[\bigg(\frac{a_g^2}{r^2 } -\frac{2\pi g\sigma^g_0}{|m|r } \bigg)\bigg(m^2-\frac12\bigg)-\frac{c_a^2}{4r^2}\bigg]\bigg\ } \qquad\\ = 4\pi^2g^2\sigma_0^s\sigma_0^g\qquad\qquad \end{split}\ ] ] for aligned coplanar mhd perturbations in a composite msid system . in the absence of the mutual gravitational coupling between the msid and the stellar sid , represented by the term on the right - hand side of equation ( [ 46 ] ) , the left - hand side of equation ( [ 46 ] ) would give rise to two separate dispersion relations , one for the stellar sid and one for the gaseous msid
. the first one would be @xmath162 which gives the stationary dispersion relation for aligned coplanar perturbations in a stellar sid alone , that is , a single sid without magnetic field . substituting expressions ( [ 11 ] ) and ( [ 20 ] ) of @xmath163 and @xmath164 ( with @xmath136 ) into equation ( [ 71 ] ) , we obtain @xmath165=0\ , \ ] ] which is simply equation ( 26 ) of shu et al .
( 2000 ) .
the second factor in the curly braces on the left - hand side of equation ( [ 46 ] ) is @xmath166m^2\omega_g^2 \\
+ \frac{m^2c_a^2}{r^2}\bigg[\bigg(\frac{a_g^2}{r^2}-\frac{2\pi g\sigma^g_0}{|m|r } \bigg)(m^2-\frac12)-\frac{c_a^2}{4r^2}\bigg]=0\ , \ \qquad \end{split}\ ] ] which is the stationary dispersion relation for aligned coplanar mhd perturbations in the gaseous msid alone , that is , a single msid with a coplanar magnetic field .
equation ( [ 73 ] ) is simply equation ( 3.2.6 ) of lou ( 2002 ) . by the above results , it is clear that equation ( [ 46 ] ) represents the dispersion relation for stationary mhd density waves in a composite msid system .
as both sid and msid rotate , the stationarity of mhd density wave patterns in an inertial frame of reference imposes conditions on dimensionless rotation parameter @xmath89 ( or equivalently , @xmath91 ) . for the following analysis
, we substitute expressions ( [ 11 ] ) , ( [ 15 ] ) , ( [ 20 ] ) and ( [ 21 ] ) into equation ( [ 46 ] ) to yield another form of stationary dispersion relation in a composite full or partial msid system , namely @xmath167 \times\bigg\{m^4d_g^4-\bigg[2d_g^2 + \\ \bigg(\lambda^2 + 1-f\frac{1+d_g^2-\lambda^2/2}{|m|}\frac{\delta } { 1+\delta}\bigg)m^2 - 2\lambda^2\bigg]\times m^2d_g^2 \qquad\\ + \bigg[\bigg(1-f\frac{1+d_g^2-\lambda^2/2}{|m|}\frac{\delta } { 1+\delta}\bigg)\bigg(m^2-\frac12\bigg)-\frac{\lambda^2}{4}\bigg ]
\qquad\\ \times m^2\lambda^2\bigg\}-f^2\frac{(1+d_s^2)(1+d_g^2-\lambda^2/2)\delta } { ( 1+\delta)^2}=0\ .\qquad\qquad \end{split}\ ] ] a substitution of expression ( [ 87 ] ) for @xmath91 into ( [ 89 ] ) gives a cubic algebraic equation in terms of @xmath89 . or
equivalently , a substitution of expression ( [ 90 ] ) for @xmath89 into ( [ 89 ] ) would yield a cubic algebraic equation in terms of @xmath91 . as noted earlier , we focus on the cubic equation of @xmath89 to identify physical solutions of @xmath168 , because @xmath169 as a result of @xmath170 in typical disc galaxies . for a later examination of spatial phase relationship between azimuthal magnetic field perturbation @xmath171 and the surface mass density perturbation @xmath116 of the gaseous msid , we combine equations ( [ 66 ] ) and ( [ 67 ] ) to obtain @xmath172 relating the @xmath20-component of the magnetic field perturbation @xmath173 and the radial gas flow speed perturbation @xmath147 .
using equations ( [ 70 ] ) and ( [ 99 ] ) , one can eliminate @xmath147 to obtain @xmath174}\ , \ ] ] that can be further reduced to @xmath175}{2d_g^2-\lambda^2/2}\ .\ ] ] for real @xmath176 , the sign of the right - hand side of the equation ( [ 101 ] ) will determine the phase relationship between the azimuthal magnetic field perturbation @xmath171 and the gas surface mass density perturbation @xmath116 in the gaseous msid .
that is , @xmath171 and @xmath116 are in and out of phase for a positive and negative right - hand side of the equation ( [ 101 ] ) , respectively . for aligned axisymmetric disturbances with @xmath11 , it would be inappropriate to directly use relation ( [ 46 ] ) or ( [ 89 ] ) .
one should carefully examine equations ( [ 33])@xmath15([41 ] ) with @xmath177 . with @xmath178 , equations ( [ 33 ] ) , ( [ 35 ] ) , ( [ 36 ] ) , ( [ 38 ] ) , ( [ 39 ] ) and ( [ 41 ] ) can be satisfied , and equation ( [ 40 ] ) is identically zero . by choosing @xmath179 , @xmath180 , @xmath181 , @xmath182 , @xmath183 , @xmath184 , @xmath185 constant and @xmath186 constant , the two remaining equations ( [ 34 ] ) and ( [ 37 ] ) are consistent with a rescaling of the axisymmetric background . in the present context , this rescaling is somewhat trivial .
we turn to cases of @xmath187 below .
we first come to the aligned case of @xmath189 . for a full composite msid system with @xmath136 ,
it is easy to verify that equation ( [ 89 ] ) can be satisfied for arbitrary @xmath89 .
this is quite similar to cases of a single sid studied by shu et al .
( 2000 ) , of a single msid studied by lou ( 2002 ) and a composite sid system studied by lou & shen ( 2003 ) .
however , we note that for a single partial msid ( lou 2002 ) and for a composite system of two coupled partial sids ( lou & shen 2003 ) , such stationary aligned eccentric @xmath189 perturbations are not allowed for arbitrary @xmath89 .
likewise , in our case of a composite partial msid system with @xmath190 , it is easy to see that equation ( [ 89 ] ) can no longer be satisfied for arbitrary @xmath89 by simply setting @xmath189 . in the following ,
we analyze cases of @xmath5 and focus on the special case of @xmath136 for a full composite msid system . as mentioned earlier , we work in the parameter regime of @xmath191 with a typical disc galaxy in mind . by an extensive numerical exploration of aligned cases from @xmath192 to @xmath193 ,
we note empirically that cases of @xmath194 are fairly similar to the case of @xmath192 . for this reason
, we shall mainly consider the case of @xmath192 .
as noted earlier , we can substitute expression ( [ 87 ] ) for @xmath91 into equation ( [ 89 ] ) to derive a cubic algebraic equation of @xmath195 as @xmath196 where @xmath192 and @xmath136 , and the four coefficients @xmath156 , @xmath197 , @xmath198 and @xmath199 are explicitly defined by @xmath200 @xmath201 @xmath202 and @xmath203 respectively .
formally , three mathematical solutions for @xmath195 can be written out analytically from cubic equation ( [ 91 ] ) or from the more general cubic dispersion relation ( [ 89 ] ) with @xmath194 and @xmath190 , although the solution expressions are fairly involved ( see appendix c for details ) .
practically , we explore numerically various parameter regimes for the three @xmath89 solutions . for specified values of parameters
@xmath4 and @xmath90 , we show three mathematical solutions @xmath195 versus @xmath3 in figs .
[ f11]@xmath15[f13 ] .
three solution curves of @xmath89 versus @xmath3 for the aligned case with @xmath192 , @xmath204 and @xmath205 . ] three solution curves of @xmath89 versus @xmath3 for the aligned case with @xmath192 , @xmath204 and @xmath206 . ] three solution curves of @xmath89 versus @xmath3 for the aligned case with @xmath192 , @xmath204 and @xmath207 . ] for typical parameters , the three solution branches of @xmath89 do not intersect with each other . for the convenience of discussion , we use @xmath208 , @xmath209 and @xmath210 to denote the upper , middle and lower solution branches , respectively . generally speaking , the upper two solution branches , @xmath208 and @xmath209 , are qualitatively similar to those in a composite unmagnetized sid system ( lou & shen 2003 ) , that is , @xmath208 branch remains always positive , while @xmath209 decreases monotonically with increasing @xmath3 and becomes negative when @xmath3 exceeds a certain critical value @xmath211 . as expected for small @xmath90 ( i.e. weak magnetic field ; see fig .
[ f11 ] ) , @xmath208 and @xmath209 respectively obey the limits of those in a composite sids without magnetic field [ see equations ( 54 ) and ( 55 ) of lou & shen ( 2003 ) ] , showing that as @xmath212 , @xmath208 and @xmath209 consistently approach the upper and lower branches respectively of lou & shen ( 2003 ) .
one novel feature is the lowest solution branch @xmath210 owing to the presence of magnetic field . in most cases , @xmath210 is negative when @xmath3 varies from @xmath213 to @xmath214 , but for some special parameters , @xmath210 may become positive when @xmath3 becomes smaller than a critical value @xmath215 .
analytical expressions of @xmath211 and @xmath215 for the case of @xmath192 can be derived from a quadratic equation and are given by the following pair @xmath216\quad \end{split}\ ] ] and @xmath217\ , \quad \end{split}\ ] ] respectively , where @xmath211 remains always larger than @xmath215 .
we note the following .
first , there is no essential mathematical difficulty of obtaining more general forms of @xmath211 and @xmath215 for arbitrary @xmath218 values with @xmath219 .
secondly , by setting @xmath220 in expression ( [ 94 ] ) for @xmath211 , we obtain @xmath221 consistent with the case of @xmath192 in expression ( 56 ) for @xmath222 by lou & shen ( 2003 ) .
finally , while @xmath215 is real in many cases [ i.e. a positive determinant in both expressions ( [ 94 ] ) and ( [ 95 ] ) ] , @xmath215 may be too small to be discernible in the parameter regime of @xmath191 .
for some special parameters specified , @xmath215 becomes noticeable .
for example , in fig .
[ f12 ] with @xmath204 and @xmath206 , we have @xmath223 according to expression ( [ 95 ] ) .
in this case , for @xmath3 smaller than 1.1667 ( still quite restrictive ) , there exist three positive solutions of @xmath89 . for @xmath224 , the upper two branches @xmath208 and
@xmath209 are positive , corresponding to two possible stationary perturbation modes . when @xmath3 exceeds @xmath211 , only @xmath208 remains positive , corresponding to one possible stationary perturbation mode . for the usual case of @xmath225 , there are at most two possible stationary modes when @xmath226 ( see fig . [ f13 ] ) .
we now examine phase relationships among the azimuthal magnetic field and the surface mass density perturbations , because they may provide clues for magnetized spiral galaxies through optical and synchrotron radio observations ( e.g. mathewson et al .
1972 ; beck & hoernes 1996 ; fan & lou 1996 ; lou & fan 1998a , 2002 , 2003 ; frick et al .
2000 , 2001 ; lou 2002 ; lou et al . 2002 ) . for the phase relationship between the two surface mass density perturbations @xmath152 and @xmath153 , a combination of expressions ( [ 11 ] ) , ( [ 20 ] ) and equation ( [ 42 ] ) gives @xmath227 \left(1+\delta \right)}{\left|m\right|\left(y+1\right)}}\\ ] ] where @xmath195 .
a substitution of expression ( [ 97 ] ) into equation ( [ 91 ] ) leads to a cubic algebraic equation of @xmath228 .
we show different curves of @xmath228 versus @xmath3 in figs .
[ f16 ] and [ f18 ] by specifying different values of @xmath4 and @xmath90 . in parallel
, we examine the phase relationship between the azimuthal magnetic field perturbation @xmath171 and the surface mass density perturbation @xmath116 in the gaseous msid . by equation ( [ 101 ] )
, we introduce a dimensionless @xmath229 parameter @xmath230}{2d_g^2-\lambda^2/2}\ , \ ] ] whose sign determines the phase relationship between @xmath173 and @xmath152 . using equation ( [ 102 ] ) , we may express @xmath91 as a function of @xmath229 , namely @xmath231 a combination of equation ( [ 89 ] ) and expression ( [ 90 ] ) gives a cubic algebraic equation of @xmath91 ; and a substitution of expression ( [ 103 ] ) into the resulting cubic equation of @xmath91 leads to a cubic equation in terms of @xmath229 ( with @xmath192 ) .
we present different curves of @xmath229 versus @xmath3 for different values of @xmath4 and @xmath90 in figs .
[ f20]@xmath15[f22 ] . from expressions ( [ 97 ] ) and
( [ 102 ] ) , one can show that both larger @xmath228 and @xmath176 correspond to smaller @xmath232 .
the lowest branches of @xmath228 or @xmath176 are then related to the uppermost @xmath208 ; the middle branches of @xmath228 or @xmath176 are related to the middle @xmath209 ; and the uppermost branches of @xmath228 or @xmath176 are related to the lowest @xmath210 as shown in figs .
[ f16]@xmath15[f22 ] .
curves of @xmath228 versus @xmath3 for the aligned @xmath192 case with @xmath204 and @xmath205 . ] curves of @xmath228 versus @xmath3 for the aligned @xmath192 case with @xmath204 and @xmath207 .
the uppermost branch related to @xmath210 is only discernible for @xmath233 and is not shown here .
specifically , this is caused by a negative @xmath210 when @xmath234 ( see fig .
[ f13 ] ) . ]
curves of @xmath176 versus @xmath3 for the aligned @xmath192 case with @xmath204 and @xmath205 . ] curves of @xmath176 versus @xmath3 for the aligned @xmath192 case with @xmath204 and @xmath207 .
the uppermost branch related to @xmath210 appears somewhat flat at about @xmath235 and is not shown here . ]
several interesting features are noted here for these curves .
first , when @xmath90 is small ( see fig .
[ f16 ] ) , the lowest and middle branches of @xmath228 respectively follow those in a composite sid system without magnetic fields [ see equations ( 58 ) and ( 59 ) of lou & shen ( 2003 ) ] , showing consistently that as @xmath212 , the middle and lowest branches of @xmath228 correspond to the lower and upper branches of lou & shen ( 2003 ) , respectively . secondly , @xmath228 and @xmath176 of each solution branch of @xmath232 bear definite signs in most cases .
for the case of @xmath204 and @xmath205 as shown in figs .
[ f11 ] , [ f16 ] and [ f20 ] , the stationary perturbation mode of @xmath208 branch has a phase relationship between the surface mass density of the two sids ( i.e. @xmath228 ) and a phase relationship between the surface mass density of the msid and the azimuthal magnetic field ( i.e. @xmath176 ) being both always out of phase . in most cases , the stationary perturbation mode of @xmath209 branch has a @xmath228 being in phase ( see fig .
[ f16 ] ) and a @xmath176 being out of phase ( see fig.[f20 ] ) . the stationary perturbation mode of @xmath210 branch has @xmath228 and @xmath176 being both always in phase ( see figs .
[ f16 ] and [ f20 ] ) .
thirdly , for the stationary perturbation mode of @xmath209 branch , both @xmath228 and @xmath176 have a common zero point at @xmath236 that increases with increasing @xmath90 . for a sufficiently large @xmath90 ( e.g. @xmath207 ) , @xmath236 becomes greater than @xmath213 .
thus , @xmath228 and @xmath176 may become zero or even carry opposite signs ( see figs .
[ f18 ] and [ f22 ] ) .
our analysis show that the middle @xmath209 branch for @xmath228 and @xmath176 has a common root at @xmath237 for a fixed @xmath4 , @xmath236 increases with increasing @xmath90 . specifically for @xmath204 and @xmath205 in expression ( [ 98 ] ) , we have @xmath238 smaller than 1 .
therefore , the middle @xmath209 branch remains always positive ( see fig .
[ f16 ] ) . on the other hand , for @xmath207 in expression ( [ 98 ] ) , we have @xmath239 larger than 1 ( see fig .
[ f18 ] ) .
it is not surprising that the middle @xmath209 branch of @xmath228 and @xmath176 has a common root as result of @xmath240 .
the special @xmath240 case may be a possible situation in our model . by equation ( [ 68 ] ) , it follows that @xmath241 vanishes .
as @xmath153 and @xmath173 do not vanish , we see that @xmath242 , @xmath150 and consequently @xmath243 do not vanish by expressions ( [ 49 ] ) , ( [ 50 ] ) and ( [ 69 ] ) . finally ,
our analysis show that the uppermost branch of @xmath228 and @xmath176 corresponding to the @xmath210 branch has a root at @xmath244 .
it follows that @xmath210 remains always positive . for the onset of aligned secular barlike instabilities in a composite msid system
, we here derive the mhd virial theorem for a composite msid system ( either full or partial ) from the background radial equilibrium conditions ( [ 2 ] ) and ( [ 5 ] ) @xmath245 and @xmath246 adding equations ( [ 104 ] ) and ( [ 105 ] ) , we readily obtain @xmath247 multiplying equation ( [ 106 ] ) by @xmath248 and integrating from @xmath249 to a finite radius @xmath250 , that is allowed to approach infinity eventually , we have the following mhd virial theorem . @xmath251\ ] ] within radius @xmath250 , where @xmath252 is the total rotational kinetic energy of the two sids , @xmath253 is the gravitational energy in the composite msid system , @xmath254 is the sum of stellar and gas internal energies and @xmath255 is the magnetic energy contained in the gaseous msid component . for a full composite system with @xmath136
, we combine relations ( [ 11 ] ) , ( [ 15 ] ) , ( [ 20 ] ) and ( [ 21 ] ) with expressions ( [ 122])@xmath15([125 ] ) to obtain @xmath256\delta } { 2g(1+\delta)}r\ , \end{split}\ ] ] @xmath257 @xmath258 based on early numerical simulations ( hohl 1971 ; miller et al .
1970 ) , it was known that a thin self - gravitating disc in rotation may rapidly evolve into bar configurations ( e.g. binney & tremaine 1987 ) .
ostriker & peebles ( 1973 ) suggested an approximate criterion @xmath259 , necessary but not sufficient , for stability against bar - type instabilities , on the basis of their @xmath260body numerical explorations involving 300 particles . in the presence of a coplanar azimuthal magnetic field with a radial scaling of @xmath261 , lou ( 2002 ) proposed that the ratio @xmath6 may play the role of @xmath262 in an unmagnetized sid to determine the onset of instability in a single msid . by the above analogy and a natural extension , we use @xmath6 instead of @xmath262 to examine the onset criterion of instability in a full composite msid system in the presence of coplanar non - axisymmetric aligned mhd perturbations . using expressions ( [ 108 ] ) , ( [ 109 ] ) and
( [ 111 ] ) , we arrive at @xmath263 indicating that the value of @xmath6 falls between @xmath249 and @xmath264 as in the unmagnetized case ( lou & shen 2003 ) and increases with increasing @xmath89 .
therefore in a composite msid system , the three possible values of @xmath89 correspond to three different values of @xmath6 ratio ; and larger values of @xmath89 correspond to higher @xmath6 ratios .
we here examine a few cases to illustrate the utility of our proposed onset criterion for instability .
first , we set @xmath192 , @xmath204 , @xmath265 and @xmath90=1 . from equation ( [ 91 ] ) , one obtains three @xmath89 solutions @xmath266 , @xmath267 and @xmath268 .
insertions of these three values of @xmath89 into expression ( [ 112 ] ) would give @xmath269 , @xmath270 and @xmath271 , respectively .
as another example , we set @xmath192 , @xmath204 , @xmath272 and @xmath90=1 . from equation ( [ 91 ] ) , we have @xmath273 , @xmath274 and @xmath275 ; the third branch of solution should be ignored as @xmath93 .
the corresponding values of @xmath6 for the first two @xmath89 solutions are @xmath276 and @xmath277 , respectively . from these numerical estimates
, it is clear that the ratio @xmath6 can be much smaller than 0.14 ( mainly through the perturbation mode of the @xmath209 branch ) , analogous to an unmagnetized composite sid system ( lou & shen 2003 ) .
shu et al . ( 2000 ) suggested a correspondence between @xmath262 ratio and the onset of secular bar - like instabilities in a single fluid sid .
it seems natural to extend this suggestion to a single msid ( lou 2002 ) and to an unmagnetized composite system of two sids ( lou & shen 2003 ) . by this analogy
, we now propose a further extension of this correspondence of ratio @xmath6 and the onset of secular bar - like instabilities in a composite system composed of one sid and one msid .
it then appears , as in the case of lou & shen ( 2003 ) , the threshold of @xmath6 ratio can be considerably lowered in a composite sid or msid system .
qualitatively , this illustrates that the mutual gravitational coupling tends to make a disc system more unstable ( jog & solomon 1984a , b ; bertin & romeo 1988 ; romeo 1992 ; elmegreen 1995 ; jog 1996 ; lou & fan 1998b ; lou & shen 2003 ; shen & lou 2003 ) .
curves of ratio @xmath6 versus @xmath4 for different values of @xmath278 , @xmath279 , @xmath280 for the aligned @xmath192 case with @xmath206 . ]
we now explore trends of variations for the ratio @xmath6 of the @xmath209 branch . setting @xmath192
, we present curves of @xmath6 versus @xmath4 when parameters @xmath3 and @xmath90 are specified ( see fig . [ f24 ] ) .
as there are three @xmath89 solution branches involving three parameters @xmath4 , @xmath3 and @xmath90 , values of ratio @xmath6 can be diversified .
consequently , instability properties of a composite ( m)sids system are more complex than in a single sid .
moreover , possible perturbation modes of the @xmath89 solution of @xmath209 branch can vary with different values of azimuthal wavenumber @xmath218 as shown in figs .
[ f26 ] and [ f27 ] .
the @xmath89 solution of the @xmath209 branch versus @xmath218 with parameters @xmath204 , @xmath278 and @xmath205 .
the smallest @xmath89 occurs at @xmath281 . ]
the @xmath89 solution of the @xmath209 branch versus @xmath218 with parameters @xmath204 , @xmath282 and @xmath207 .
the smallest @xmath89 occurs at @xmath192 in this case . ] for parameters @xmath204 , @xmath278 and @xmath205 , @xmath89 attains its minimum value around @xmath281 , while for parameters @xmath204 , @xmath282 and @xmath207 , the minimum value of @xmath89 occurs at @xmath192 . in association with the smallest value of @xmath89 ,
the most vulnerable stationary perturbation configuration varies with chosen parameters . in other words ,
a slightly different choice of parameters may lead to entirely different mhd perturbation configurations . again
, this example shows the diversity and complexity in a composite msid system and indicates that bar - like instabilities ( i.e. @xmath283 ) may not necessarily be the dominant instability in a composite msid system .
in this section , we analyze unaligned perturbation configurations of logarithmic spiral structures in a composite msid system . we first derive the stationary dispersion relation for both full and partial composite msid system .
we then focus on the full case ( @xmath136 ) and address the problem of axisymmetric marginal instabilities with @xmath11 . to construct stationary mhd perturbation configurations of unaligned logarithmic spirals in a composite msid system ,
we pick the following potential - density pair ( syer & tremaine 1996 ; lynden - bell & lemos 1993 ; shu et al . 2000 ; lou 2002 ; lou & fan 2002 ; lou & shen 2003 ) , namely @xmath284 @xmath285 where @xmath286 and @xmath287 are two constant coefficients and @xmath288 is the radial scaling parameter related effectively to the radial wavenumber , together with @xmath289 where the two constant coefficients @xmath290 and @xmath291 are related to @xmath286 and @xmath287 respectively by @xmath292 @xmath293 with @xmath294 being the kalnajs function ( kalnajs 1971 ) .
consistently , we may write @xmath295 and @xmath296 where @xmath297 and @xmath298 are two constant coefficients . using expressions ( [ 52 ] ) , ( [ 53 ] ) and ( [ 54 ] ) , equations ( [ 63 ] ) , ( [ 64 ] ) , ( [ 66 ] ) and ( [ 67 ] ) lead to @xmath299 @xmath300\bigg\}iu_g = c_a^2\bigg(i\alpha+2-\frac{c_a^2}{2\omega_g^2r^2}\bigg ) \qquad\\ \times\bigg[\frac{m\omega_g\sigma_g}{\sigma_0^g } -\frac{m}{\omega_gr}\bigg(\frac{a_g^2\sigma_g}{r\sigma_0^g } + v_s+v_g\bigg)\bigg ] \quad\\ + m\omega_gr\bigg(i\alpha+\frac32\bigg ) \bigg(\frac{a_g^2\sigma_g}{r\sigma_0^g}+v_s+v_g\bigg ) -\frac{m\omega_gc_a^2\sigma_g}{2\sigma_0^g}\ , \end{split}\ ] ] @xmath301 @xmath302 a combination of equations ( [ 75 ] ) and ( [ 76 ] ) together with relations ( [ 55 ] ) and ( [ 56 ] ) give @xmath303\sigma_g = \bigg(\frac{\mathcal km}{\omega_gr } -\mathcal a\bigg)2\pi g\mathcal n_m(\alpha)\sigma_s\ , \end{split}\ ] ] where the three coefficients @xmath155 , @xmath156 and @xmath198 are defined by @xmath304 @xmath305 @xmath306 for stationary coplanar perturbations in the stellar disk in parallel , we set @xmath135 in equations ( [ 33])@xmath15([35 ] ) and use equations ( [ 51])@xmath15([56 ] ) and ( [ 74 ] ) to obtain @xmath307\sigma_s \quad\\ = \bigg(m^2+\alpha^2+\frac14\bigg ) 2\pi g\mathcal n_m(\alpha)\sigma_g\ . \end{split}\ ] ] equation ( [ 57 ] ) can also be obtained by setting @xmath159 in equation ( [ 58 ] ) and exchange subscripts @xmath17 and @xmath18 . by equations ( [ 57 ] ) and ( [ 58 ] ) , we derive @xmath308 \times\bigg\{m^4\omega_g^4-m^2\omega_g^2\bigg[2\omega_g^2 \quad\qquad\\ + \bigg(\frac{c_a^2+a_g^2}{r^2}-\frac{2\pi g\mathcal n_m(\alpha)\sigma_0^g}r\bigg)\bigg(m^2+\alpha^2 + \frac14\bigg)-\frac{2c_a^2}{r^2}\bigg ] \qquad\\ + \frac{m^2c_a^2}{r^2}\bigg[\bigg(\frac{a_g^2}{r^2}-\frac{2\pi g\mathcal
n_m(\alpha)\sigma_0^g}r\bigg)\bigg(m^2+\alpha^2-\frac14\bigg ) \qquad\quad\\ -\frac{c_a^2}{4r^2}\bigg]\bigg\ } = 4\pi^2g^2\mathcal n_m^2(\alpha)\sigma_0^g\sigma_0^s\bigg(m^2+\alpha^2+\frac14\bigg ) \qquad\qquad\\ \times\bigg[\bigg(m^2+\alpha^2+\frac14\bigg ) -\frac{c_a^2}{\omega_g^2r^2 } \bigg(m^2+\alpha^2-\frac14\bigg)\bigg]\ .\qquad\qquad \end{split}\ ] ] in the absence of the gravitational coupling between the stellar sid and the gaseous msid represented by the right - hand side term of equation ( [ 62 ] ) , the left - hand side of dispersion relation ( [ 62 ] ) would be reduced to two separate dispersion relation factors .
the first dispersion relation would be @xmath309=0\\ ] ] for stationary logarithmic spirals in a single stellar sid without magnetic field .
substituting expressions ( [ 11 ] ) and ( [ 20 ] ) of @xmath163 and @xmath164 with @xmath136 into equation ( [ 79 ] ) , we have @xmath310}{d_s^2(m^2 - 2)}=1\ , \ ] ] which is simply equation ( 37 ) of shu et al .
the second dispersion relation would be @xmath311m^2\omega_g^2 + \frac{m^2c_a^2}{r^2 } \qquad\qquad\\ \times\bigg[\bigg(\frac{a^2}{r^2}-\frac{2\pi g \mathcal n_m(\alpha)\sigma_0^g}r\bigg ) \bigg(m^2+\alpha^2-\frac14\bigg)-\frac{c_a^2}{4r^2}\bigg]=0 \end{split}\ ] ] for stationary logarithmic spirals in a single gaseous msid with coplanar magnetic field .
equation ( [ 81 ] ) is simply equation ( 3.4.15 ) of lou ( 2002 ) . without magnetic field with @xmath159 , equation ( [ 62 ] )
is equivalent to the dispersion relation ( 86 ) for stationary logarithmic spiral configurations in a composite sid system of lou & shen ( 2003 ) as it should be . for numerical computations , there are two useful formulae of @xmath312 , namely , the recursion relation @xmath313^{-1}\ ] ] and the asymptotic expansion @xmath314 when @xmath315 ( kalnajs 1971 ; shu et al .
2000 ) . for the purpose of examining phase relationship between the azimuthal magnetic field perturbation @xmath171 and the surface mass density perturbation @xmath316 of the gaseous msid in the full case of @xmath136 , we derive the following relations . from equations ( [ 75 ] ) and ( [ 57 ] ) with @xmath136
, we readily obtain @xmath317(\omega_gr ) } \qquad\qquad\\ \times\bigg[\frac{(1+\delta)(1-d_g^2)}{\delta(1+d_g^2-\lambda^2/2 ) } -\frac{\mathcal a\ \mathcal n_m(\alpha ) } { ( \mathcal a-\mathcal b)}\bigg]\ , \end{split}\ ] ] where the two coefficients @xmath156 and @xmath197 are defined by @xmath318 and @xmath319 it follows from expressions ( [ 78 ] ) and ( [ 115 ] ) that @xmath320 \quad\qquad\qquad\qquad\\ \times\bigg[\frac{(1+\delta)(1-d_g^2 ) } { \delta(1+d_g^2-\lambda^2/2 ) } -\frac{\mathcal a\ \mathcal n_m(\alpha ) } { ( \mathcal a-\mathcal b)}\bigg]\ , \end{split}\ ] ] that will be discussed later in our analysis .
while the @xmath11 case is somewhat trivial as a rescaling in the aligned case , it is of considerable interest in the special ` spiral ' case with radial oscillations . in this subsection , we analyze this situation for a full composite msid system with @xmath136 .
a substitution of expressions ( [ 11 ] ) , ( [ 15 ] ) , ( [ 20 ] ) and ( [ 21 ] ) into equation ( [ 62 ] ) with @xmath84 leads to @xmath321\bigg\ } \quad\qquad\qquad\\ \times\bigg\{-d_g^2\bigg[2d_g^2 + \bigg(\lambda^2 + 1 -\frac{\mathcal n_0(\alpha)(1+d_g^2-\lambda^2/2)\delta } { 1+\delta}\bigg ) \qquad\\ \times\bigg(\alpha^2+\frac14\bigg)-2\lambda^2\bigg ] + \lambda^2\bigg[\bigg(1-\frac{\mathcal n_0(\alpha ) ( 1+d_g^2-\lambda^2/2)\delta}{1+\delta}\bigg ) \hbox { } \\ \times\bigg(\alpha^2-\frac14\bigg)-\frac{\lambda^2}4\bigg]\bigg\ } -\frac{\mathcal n_0(\alpha)^2d_g^2(1+d_s^2)(1+d_g^2-\lambda^2/2)\delta } { ( 1+\delta)^2 } \\
\times\bigg(\alpha^2+\frac 14\bigg)\bigg[\alpha^2+\frac 14 -\lambda^2\bigg(\alpha^2-\frac14\bigg)/d_g^2\bigg]=0\ .\qquad\qquad \end{split}\ ] ] by recursion formula ( [ 83 ] ) and asymptotic expression ( [ 84 ] ) for the kalnajs functions , we have an approximate expression of @xmath322 , namely @xmath323\ . \end{split}\ ] ] a substitution of approximate expression ( [ 85 ] ) for @xmath322 and expression([87 ] ) for @xmath91 into dispersion relation ( [ 82 ] ) yields a cubic algebraic equation in terms of @xmath89 ; and similarly , a substitution of approximate expression ( [ 85 ] ) for @xmath322 and expression ( [ 90 ] ) for @xmath89 yields another cubic algebraic equation in terms of @xmath91 .
as noted earlier , we consider the cubic equation of @xmath89 because of the fact that @xmath169 when @xmath191 . for different values of parameters @xmath4 , @xmath3 and @xmath90 , we show curves of @xmath89 versus @xmath288 in figs .
[ f1]@xmath15[f9 ] . the marginal stability curve of @xmath89 versus @xmath288 with @xmath324 , @xmath204 , @xmath278 and @xmath325 . ] the marginal stability curve of @xmath89 versus @xmath288 with @xmath324 , @xmath204 , @xmath278 and @xmath207 . ] the marginal stability curve of @xmath89 versus @xmath288 with @xmath324 , @xmath204 , @xmath326 and @xmath325 . ] the marginal stability curve of @xmath89 versus @xmath288 with @xmath324 , @xmath327 , @xmath326 and @xmath206 . ] the marginal stability curve of @xmath89 versus @xmath288 with @xmath324 , @xmath204 , @xmath265 and @xmath205 . for @xmath328 ,
the lower ring fragmentation branch would gradually rise above the horizontal @xmath288 axis . ]
the marginal stability curve of @xmath89 versus @xmath288 with @xmath324 , @xmath204 , @xmath265 and @xmath329 . compared with fig . [ f9 ] , the increase of @xmath2 moves the lower ring fragmentation downward . ]
for the positive portions of solution @xmath89 with typical parameters , it turns out that the basic features of a @xmath89 versus @xmath288 profile are qualitatively similar to those of a single sid [ see fig . 2 of shu et al .
( 2000 ) ] and to those of a composite sid system ( see figs .
@xmath330 of lou & shen 2003 ) , except that now trends of variation for @xmath89 with three parameters become more complicated . for example , the lower ring fragmentation curve first discussed in a single msid ( lou 2002 ) becomes negative or indiscernible in most cases ( see appendix b ) . for cases of @xmath3 around @xmath213 , however , the lower ring fragmentation curve of @xmath89 may become positive as shown in fig .
[ f9 ] . as parameters @xmath4 , @xmath3 and @xmath90 vary , we note several trends of variation in the profiles of @xmath89 . for example ,
when @xmath90 increases for fixed @xmath4 and @xmath3 , the ring fragmentation curves of @xmath89 seem to be raised , and the collapse regime tends to be enlarged by comparing figs .
[ f1 ] and [ f2 ] . in other words ,
the presence of coplanar magnetic field tends to increase the chance of collapse but reduce the danger of ring fragmentation .
this feature is qualitatively similar to the case of a single coplanar magnetized sid studied by lou ( 2002 ) ( see fig . 1 of lou 2002 ) .
as @xmath3 increases for fixed @xmath4 and @xmath90 , the curves of @xmath89 tend to be lowered by comparing figs .
[ f1 ] and [ f3 ] . when @xmath4 increases for fixed @xmath3 and @xmath90 , the ring fragmentation curves of @xmath89 decrease , and the collapse regime shrinks by comparing figs . [ f3 ] and [ f7 ] .
note that as @xmath4 increases ( @xmath331 ) , the influence of gas disc on the stellar disc becomes more important .
therefore , the gravitational coupling with another disc tends to suppress the collapse but enhance the danger of ring fragmentation .
one behaviour for the variation of the lower ring fragmentation curve appears somewhat surprising .
it seems to be an characteristic feature involving magnetic field ( lou 2002 ) . as already noted earlier , when @xmath3 approaches @xmath213 , this curve could rise above the @xmath288 axis to become a possible perturbation mode ( fig .
however , comparing fig .
[ f9 ] with fig .
[ f10 ] , we find that in typical parameter regimes , the increase of @xmath2 would force the lower ring fragmentation curve to go downward .
this is completely different from that of a single msid ( see fig . 1 of lou 2002 ) . through numerical explorations , we realize that the location @xmath288 of the vertical asymptote for the ring fragmentation branch seems to be independent of @xmath4 , @xmath3 and @xmath90 . by letting the coefficient of the cubic term in the cubic algebraic equation of @xmath195 to vanish
, the condition that this critical value @xmath332 must satisfy is once again @xmath333 which is exactly the same as that of shu et al .
( 2000 ) , lou ( 2002 ) and lou & shen ( 2003 ) for three related but different sid systems . from equation (
[ 85 ] ) and asymptotic expansion ( [ 84 ] ) , this critical value @xmath332 is estimated to be 1.759 .
physical interpretations for the marginal instability curves are clear .
as @xmath288 parameter is a measure for the radial wavenumber , a smaller @xmath288 corresponds to a larger radial scale of perturbations .
thus , the collapse regime indicates a rotationally modified jeans instability to which a composite msid system is vulnerable when a radial perturbation scale is sufficiently large ( i.e. a sufficiently small @xmath288 ) .
however , when @xmath89 becomes larger , the conservation of angular momentum works against the inward gravitational force and the jeans instability is suppressed . nevertheless , a composite msid system is also vulnerable to toomre - type instability ( safronov 1960 ; toomre 1964 ) associated with the ring fragmentation branch ( shu et al .
2000 ) . in various galactic contexts ,
there have been numerous studies to identify an effective @xmath334 parameter for a composite disc system ( jog & solomon 1984a , b ; bertin & romeo 1988 ; kennicutt 1989 ; romeo 1992 ; elmegreen 1995 ; jog 1996 ; lou & fan 1998b ) . in the case of a single sid , shu et al .
( 2000 ) noted that the minimum of the ring fragmentation curve is closely related to the toomre @xmath334 parameter ( toomre 1964 ) . in the case of a coplanar magnetized sid , lou ( 2002 ) and
lou & fan ( 2002 ) further found that the minimum of the upper ring fragmentation curve is tightly associated with the generalized mhd @xmath335 parameter ( lou & fan 1998a ) . for a composite system of two sids ,
shen & lou ( 2003 ) recently suggested a straightforward @xmath14criterion to effectively determine the axisymmetric stability against ring fragmentation ( see elmegreen 1995 and jog 1996 for other proposed criteria ) .
we therefore expect that in the present problem , the minimum of the upper ring fragmentation curve should be related a magnetic field modified @xmath14criterion to effectively determine the axisymmetric stability against ring fragmentation in a composite system of one fluid sid and one msid .
parallel to the study of aligned @xmath336 cases , we substitute expressions ( [ 11 ] ) , ( [ 15 ] ) , ( [ 20 ] ) , ( [ 21 ] ) and ( [ 87 ] ) into equation ( [ 62 ] ) with @xmath84 and obtain a cubic algebraic equation of @xmath195 for unaligned spiral cases .
one can formally derive three analytic @xmath89 solutions ( see appendix d ) that would be valuable in numerical mhd simulation studies of large - scale sid dynamics within the overall mhd density - wave scenario . to explore parameter regimes ,
we obtain @xmath89 solution curves numerically for cases of @xmath337 and so forth . as expected , there are three solution curves of @xmath89 versus @xmath288 given parameters @xmath218 , @xmath4 , @xmath90 and @xmath3 .
in certain parameter regimes , portions of the lower two @xmath89 solution curves may become negative and hence unphysical .
we defer our discussion of @xmath189 case later . in figs .
[ f28]@xmath15[f32 ] , we show a set of numerical examples with @xmath192 . through extensive numerical examples of stationary logarithmic spiral cases from @xmath192 to @xmath193 , it is clear that behaviours of @xmath89 solutions and properties of phase relationships among perturbation variables for @xmath194 cases are very similar to those of @xmath192 logarithmic spiral .
we therefore carefully examine the @xmath192 spiral case for a relatively simple yet meaningful analysis .
in addition , we plot curves of @xmath89 versus @xmath3 by specifying @xmath288 , @xmath4 and @xmath90 in fig . [ f33 ] .
curves of @xmath89 versus @xmath288 for a stationary unaligned logarithmic spiral with @xmath192 , @xmath204 , @xmath278 and @xmath205 . ]
curves of @xmath89 versus @xmath288 for a stationary unaligned logarithmic spiral with @xmath192 , @xmath327 , @xmath338 and @xmath205 . ] curves of @xmath89 versus @xmath288 for a stationary unaligned logarithmic spiral with @xmath192 , @xmath204 , @xmath265 and @xmath206 . ] curves of @xmath89 versus @xmath3 for a stationary unaligned logarithmic spiral with @xmath192 , @xmath204 , @xmath339 and @xmath206 . ]
as in the aligned case , we shall use @xmath208 , @xmath209 and @xmath210 to denote the upper , middle and lower solution branches , respectively .
it is apparent that the three @xmath89 solution branches do not intersect with each other .
similar to the case of axisymmetric disturbances , the lowest @xmath210 solution branch does not appear above zero in many cases , but for cases with relatively small @xmath3 value ( e.g. @xmath265 ) , @xmath210 solution branch can become a possible stationary perturbation mode ( see figs .
[ f32 ] and [ f33 ] ) .
we note that @xmath89 always increases with increasing @xmath288 for the upper two branches @xmath208 and @xmath209 ( see figs .
[ f28]@xmath15[f32 ] ) but decreases with increasing @xmath3 ( see fig .
[ f33 ] ) .
the uppermost @xmath208 branch remains always positive , while the middle @xmath209 branch may cross the horizontal @xmath288 axis to become negative in some cases ( see fig .
[ f31 ] ) , similar to the situation without involving coplanar magnetic field ( lou & shen 2003 ) . in fig .
[ f33 ] , there exist a @xmath211 for the middle @xmath209 branch and a @xmath215 for the lowest @xmath210 branch for given @xmath288 , @xmath4 and @xmath90 such that @xmath209 and @xmath210 become zero and negative with increasing value of @xmath3 , respectively .
parallel to the aligned case , we examine behaviours of the solutions and obtain the two critical values of @xmath3 analytically . for the @xmath192 case ,
the analytic forms of @xmath211 and @xmath215 are given by @xmath340\ ] ] and @xmath341\ , \ ] ] where definitions of the three coefficients @xmath156 , @xmath197 and @xmath198 are given in appendix a. when @xmath342 , the middle solution branch @xmath343 is positive , while for @xmath344 , the middle solution branch @xmath343 becomes negative and hence unphysical .
for the case of @xmath204 , @xmath205 and @xmath339 , we obtain @xmath345 from equation ( [ 113 ] ) , consistent with fig .
[ f28 ] ; in fig .
[ f28 ] , we have @xmath346 and the middle branch @xmath209 is positive at @xmath339 . for @xmath347 , expressions ( [ 113 ] ) and ( [ 114 ] ) give @xmath348 and @xmath349 , respectively , consistent with fig .
[ f33 ] . for stationary logarithmic perturbation spirals ,
we now examine spatial phase relationships between perturbations of the surface mass densities of the two sids ( i.e. @xmath152 and @xmath153 ) , and between perturbations of the azimuthal magnetic field and the surface mass density of the gaseous msid ( i.e. @xmath171 and @xmath116 ) .
this information might provide necessary and useful clues for optical and synchrotron radio observational diagnostics of spiral galaxies . from equation ( [ 57 ] ) , one obtains @xmath350 \quad\\ -{\frac { \left({m}^{2}-2 \right ) \left(1+\delta\right)y}{1+y}}\bigg\rbrace\ \end{split}\ ] ] with @xmath195 .
substitution of expression ( [ 119 ] ) into the cubic equation of @xmath89 gives a cubic equation of @xmath228 , and we plot several curves of @xmath228 versus @xmath3 with parameters @xmath351 , @xmath204 , @xmath205 and @xmath352 in figs .
[ f34 ] and [ f35 ] . for the phase relationship between @xmath152 and @xmath173
, we start from equation ( [ 118 ] ) . for a sufficiently small @xmath288 ,
corresponding to a relatively open spiral structure , one may keep linear terms of @xmath288 and ignore the @xmath353 term .
it follows that @xmath173 is either ahead of or lagging behind @xmath152 with a phase difference of about @xmath354 . on the other hand , for a sufficiently large @xmath288 in the tight - winding or wkbj regime , one may drop the imaginary part in equation ( [ 118 ] ) ( see lou 2002 ) .
when accuracy is not that crucial in a quantitative analysis , this expression may also be roughly used in cases of @xmath355 . we can rewrite expression ( [ 118 ] ) approximately as @xmath356\bigg\}^{-1},\ ] ] where @xmath156 and @xmath197 are defined by equations ( [ 116 ] ) and ( [ 117 ] ) , respectively .
when parameters @xmath288 , @xmath3 , @xmath4 and @xmath90 are specified in equation ( [ 121 ] ) , solutions of @xmath195 correspond to values of @xmath176 .
we then present curves of @xmath176 versus @xmath3 with parameters @xmath351 , @xmath204 , @xmath205 and @xmath352 in figs .
[ f41 ] and [ f42 ] . by equation ( [ 119 ] ) , we note again that a smaller @xmath228 corresponds to a larger @xmath195 , reminiscent of the same variation trend of correspondence in the aligned case described in section 3
. such a relatively simple correspondence does not hold for @xmath176 that should be analyzed with extra care . through extensive numerical explorations , we note that the highest branch of @xmath176 always relates to the lowest @xmath210 branch , while the @xmath208 branch of @xmath176 ( being always negative ) does not remain always less than the @xmath209 branch of @xmath176 as shown in figs . [ f41 ] and [ f42 ] .
these phase relationships among branches of @xmath228 and @xmath176 and the corresponding branches of @xmath195 are shown explicitly in figs .
[ f34]@xmath15[f42 ] .
curves of @xmath228 versus @xmath3 for a stationary logarithmic spiral with @xmath192 , @xmath204 , @xmath351 and @xmath205 . ] curves of @xmath228 versus @xmath3 for a stationary logarithmic spiral @xmath192 , @xmath204 , @xmath351 and @xmath207 .
the uppermost branch corresponding to @xmath357 is only discernible for @xmath233 and is not shown here .
the branch corresponding to @xmath343 may cross the horizontal @xmath3 axis . ]
curves of real parts of @xmath176 versus @xmath3 for a stationary logarithmic spiral of @xmath192 , @xmath204 , @xmath351 and @xmath205 . imaginary parts of @xmath176 are not shown here . while one branch of real part of @xmath176 corresponding to @xmath208 remains always negative , this branch does not remain always lower than the real part of @xmath176 branch corresponding to @xmath209 . ] curves of real parts of @xmath176 versus @xmath3 for a stationary logarithmic spiral of @xmath192 , @xmath204 , @xmath351 and @xmath207 .
imaginary parts of @xmath176 are not shown here .
the uppermost branch for real part of @xmath176 is only discernible for @xmath233 and is not shown here .
the branch of real part of @xmath176 corresponding to @xmath343 may cross the horizontal @xmath3 axis . ]
the unaligned spiral cases here are fairly similar to the aligned cases of section 3 , except that the imaginary part of @xmath176 does not vanish in general . for the stationary mode of @xmath208 branch , the phase relationship between the surface mass density of the two sids ( i.e. @xmath228 ) and the phase relationship between the surface mass density of the msid and the azimuthal magnetic field ( i.e. @xmath176 ) are always out of phase ( see figs .
[ f34]@xmath15[f42 ] ) .
being ` out of phase ' means that the real part of @xmath176 is negative . as @xmath288 ranges from sufficiently small to relatively large values , the phase difference between @xmath152 and @xmath173 ranges from @xmath358 to @xmath359 , either ahead or lagging behind .
likewise , @xmath176 being ` in phase ' implies that the phase difference between @xmath152 and @xmath173 ranges from @xmath249 to @xmath358 as @xmath288 varies from relatively large to sufficiently small values . ] for the unaligned perturbation mode of @xmath209 branch in most cases , @xmath228 is in phase , while the real part of @xmath176 is always negative .
taking into account of the imaginary part of @xmath176 , the phase difference between @xmath152 and @xmath173 is @xmath360 ( see figs .
[ f34 ] and [ f41 ] ) .
this @xmath209 branch has a zero point at @xmath236 ( where @xmath240 ) that increases with increasing @xmath90 .
when @xmath90 becomes sufficiently large ( e.g. @xmath207 ) with @xmath234 , the value of @xmath236 can become larger than @xmath213 .
it is then possible for @xmath228 and @xmath176 to become zero at @xmath361 and reverse their respective signs for @xmath362 as shown in figs .
[ f35 ] and [ f42 ] . taking into account of the imaginary part of @xmath176
, this implies that the phase difference of @xmath152 and @xmath173 switch from between @xmath360 to between @xmath363 as @xmath3 varies from @xmath364 to @xmath365 .
meanwhile during the same @xmath3 transition , the phase difference between @xmath153 and @xmath152 switch from in phase to out of phase .
the special situation of @xmath240 at @xmath361 , occurring for stationary perturbation mode of @xmath209 branch , is very similar to that of the aligned case studied earlier . with @xmath366 but @xmath367 , @xmath368 , @xmath369 and @xmath370
, this may correspond to a unique solution .
for the stationary perturbation mode of @xmath210 branch , both @xmath228 and the real part of @xmath176 remain positive and therefore @xmath152 and @xmath153 are always in phase ( see figs .
[ f34]@xmath15[f42 ] ) ; taking into account of the imaginary part of @xmath176 , the phase difference between @xmath152 and @xmath173 ranges from @xmath249 to @xmath358 .
the analytic expression of @xmath236 for @xmath192 and @xmath351 is given by @xmath371\ , \end{split}\ ] ] that increases with increasing @xmath90 .
for @xmath204 and @xmath205 in equation ( [ 120 ] ) , one has @xmath372 that is not shown explicitly in figs .
[ f34 ] and [ f41 ] ( see fig .
e1 in appendix e ) . for @xmath204 and @xmath207 ,
one has @xmath373 , consistent with figs .
[ f35 ] and [ f42 ] .
we now turn to the somewhat special case of @xmath189 .
mathematically , the qualitative difference between cases of @xmath189 and @xmath5 is perhaps due to the fact that @xmath374 for @xmath5 but otherwise for @xmath189 ( lou & shen 2003 ) .
we now examine behaviours of @xmath195 of the @xmath189 case shown in figs .
[ f37 ] and [ f38 ] .
curves of @xmath195 versus @xmath3 for stationary unaligned perturbations of @xmath189 with @xmath327 , @xmath351 and @xmath207 .
solution branch of @xmath375 may go across the horizontal @xmath3 axis and has a root at about @xmath376 . ]
curves of @xmath195 versus @xmath288 for stationary unaligned perturbations of @xmath189 with @xmath204 , @xmath278 and @xmath206 . ]
it turns out that solution curves of @xmath189 case indeed differ from those of @xmath192 case .
for example , the middle and lower branches of @xmath195 remain always negative and hence unphysical ( see figs .
[ f37 ] and [ f38 ] ) .
portions of the upper branch of @xmath377 may become zero or even negative for a sufficiently large @xmath3 ( fig .
[ f37 ] ) . that is ,
when @xmath3 exceeds a critical value @xmath378 , @xmath375 would become negative .
the analytical form of this critical @xmath378 is given expression ( [ 126 ] ) .
these properties are similar to those in the case of a composite unmagnetized sid system studied by lou & shen ( 2003 ) .
the three solution branches of @xmath195 do not intersect with each other .
the increase of @xmath3 moves both @xmath208 and @xmath209 branches downward ( see fig . [ f37 ] ) and slightly lifts the lowest @xmath210 branch .
one can derive an expression for the critical point of @xmath3 where @xmath379 .
for @xmath351 , the analytical form of this @xmath378 is @xmath380\ .
\end{split}\ ] ] for @xmath327 and @xmath207 , expression ( [ 126 ] ) gives @xmath381 , consistent with the upper curve in fig .
there is no essential mathematical difficulty of obtaining a more general analytical form of @xmath378 for arbitrary values of @xmath288 , although the expression may appear somewhat complicated .
solutions of @xmath89 for the @xmath209 branch versus @xmath218 with parameters @xmath339 , @xmath204 , @xmath278 and @xmath205 .
the smallest @xmath89 appears at @xmath382 . ]
solutions of @xmath89 for the @xmath209 branch versus @xmath218 with parameters @xmath339 , @xmath204 , @xmath282 and @xmath207 .
the smallest @xmath89 appears at @xmath192 . ]
physically , one may exchange the role of @xmath218 and @xmath288 as they both contribute to the total dimensionless wavenumber . given a fixed @xmath288 value , a composite msid system may support different values of @xmath218 ( see fig . 3 of shu et al . 2000 and figs . 14 and 15 of lou & shen 2003 ) .
this is shown here in figs .
[ f39 ] and [ f40 ] .
analogous to the aligned case , the smallest value of @xmath343 occurs at an azimuthal periodicity of @xmath382 with parameters @xmath339 , @xmath204 , @xmath278 and @xmath205 . as a comparison , for the same @xmath204 and @xmath339 yet a smaller @xmath282 and a stronger magnetic field @xmath207
, the smallest value of @xmath383 then occurs at @xmath192 .
again , we see that the stability of a composite ( m)sid system appears to be more complicated than those of a single ( m)sid system by the involvement of three parameters . as parameters @xmath4 , @xmath3 and @xmath90 vary , vulnerable configurations with the smallest @xmath89 value change accordingly .
in this final section we outline some potential applications of our analysis in the context of a nearby spiral galaxy ngc 6946 , and then summarize our theoretical results and suggestions . by the hydrodynamic density - wave scenario for a disc galaxy ,
large - scale high gas density spiral arms are vulnerable to molecular cloud and star formation activities on smaller scales and are therefore luminous in optical bands .
as synchrotron radio observations also reveal large - scale spiral structures in association with optical spiral arms ( e.g. mathewson et al . 1972 ; sofue et al . 1986 ; beck et al . 1996 ) , it is necessary to incorporate galactic magnetic field embedded in the interstellar medium ( ism ) and an relativistic cosmic - ray gas ( lou & fan 2003 ) in the mhd density - wave scenario for a deeper physical understanding ( fan & lou 1996 ; lou & fan 1998a ) .
the interlacing of optical and magnetic spiral arms in the nearby spiral galaxy ngc 6946 was first revealed by beck & hoernes ( 1996 ) in the inner disc portion with an almost rigid rotation .
similar features of displaced optical and radio arms were also reported in portions of spiral galaxies ic 342 ( e.g. krause et al . 1989 ) and m83 ( ngc 5236 ; e.g. sukumar & allen 1989 ; neininger 1992 ; neininger et al . 1993 ) .
these observational results led fan & lou ( 1996 ) to introduce theoretical concepts of slow mhd density waves ( smdws ) and fast mhd density waves ( fmdws ) in magnetized spiral galaxies ( lou & fan 1998a ; lou , yuan & fan 2001 ) and to show that perturbation enhancements of azimuthal magnetic field and gas mass density of smdw are phase shifted by @xmath384 relative to each other . however , several wavelet analysis on multi - wavelength data of ngc 6946 ( e.g. frick et al . 2000 , 2001 ) further revealed large - scale spiral arms extended well into the outer disc portion with a more or less flat rotation curve ( e.g. tacconi & young 1989 ; sofue 1996 ; ferguson et al .
1998 ) , that seems to challenge earlier arguments that smdws are largely confined within the inner disc portion of almost rigid rotation . for coplanar mhd perturbation structures of stationary logarithmic spirals in an msid model with a flat rotation curve , lou ( 2002 )
has shown the existence of smdws . for stationary smdws of lou ( 2002 ) ,
logarithmic spiral enhancements of gas surface mass density and azimuthal magnetic field are phase shifted by about @xmath360 . as effects of long - range self - gravity , magnetic field and disc differential rotation are included in the model analysis , smdw patterns in a single msid model can indeed support extended manifestations of interlaced optical and magnetic spiral arms that persist well into the outer disc portion with a flat rotation curve ( lou & fan 2002 ) .
moreover , lou ( 2002 ) further suggested that significant phase shifts between @xmath363 related to stationary logarithmic spiral structures of fmdws within a single msid .
given idealizations of our composite msid model , the three possible stationary logarithmic spiral patterns might be conceptually relevant to magnetized spiral galaxies .
specifically , for the upper two solution branches of @xmath208 and @xmath209 , the phase difference of the stationary logarithmic spiral enhancements between azimuthal magnetic field and surface mass gas density is between @xmath360 , supporting the idea that interlaced optical and magnetic spiral arms may persist through the inner disc portion to the outer portion ( with a phase difference @xmath385 ) . for a sufficiently large @xmath288 in the tight - winding or wkbj regime , this phase difference approaches @xmath386 , while for a sufficiently small @xmath288 in the open regime , this phase difference approaches @xmath358 .
note that in our modelling , the phase difference of @xmath363 occurs mainly through the @xmath210 solution branch that is usually unphysical with @xmath89 being negative .
as sketched earlier ( e.g. lou & fan 1998a , 2000a , b ; lou et al .
2002 ) , a typical disc galaxy involves a stellar disc , a magnetized gas disc and a massive dark matter halo . our theoretical model of a composite msid system ,
idealized and simplified in many aspects , does contain these basic elements .
with necessary qualifications in mind , it is then of considerable interest to reveal possible types of large - scale structures and their characteristic features in the stationary mhd density - wave scenario and relate them to galactic diagnostics and observations . for the @xmath208 solution with largest value of @xmath89 , stationary logarithmic spiral density arms in gaseous msid and in stellar sid are spatially out of phase ; in particular , spiral gas density and magnetic field arms
are siginificantly phase shifted relative to each other with a phase difference between @xmath387 . for the @xmath209 solution with middle value of @xmath89 , stationary logarithmic spiral density arms in gaseous msid and in stellar sid are spatially in phase ; meanwhile , spiral gas density and magnetic field arms are siginificantly phase shifted relative to each other with a phase difference between @xmath387 for @xmath234 and @xmath364 . at the qualitative level
, this case appears comparable to interlaced optical and synchrotron radio arms of ngc 6946 ( beck & hoernes 1996 ; fan & lou 1996 ; frick et al . 2000 , 2001 ; lou & fan 2002 ) . on the other hand , for @xmath362 ,
stationary logarithmic spiral density arms in gaseous msid and in stellar sid are spatially out of phase ; meanwhile , spiral gas density and magnetic field arms are phase shifted relative to each other with a phase difference between @xmath388 . by the above analogy of unaligned spiral configurations , our results for aligned stationary mhd perturbations with @xmath192 may bear possible diagnostics for optical and synchrotron radio observations of galactic bar structures and their spatial phase relationships .
it is straightforward to compare figs .
[ f16]@xmath15[f22 ] of the aligned cases with figs .
[ f34]@xmath15[f42 ] and infer spatial phase relationships of stellar bars , gas bars and synchroton radio bars involving magnetic fields . for example , by figs .
[ f16 ] and [ f20 ] , the @xmath208 solution describes a configuration in which the stellar density bar and the gas density bar are phase shifted relative to each other with a @xmath359 phase difference , while the gas density bar and the synchrotron radio bar ( i.e. magnetic bar ) are phase shifted relative to each other with a @xmath359 phase difference .
the @xmath209 solution describes a configuration in which the stellar density bar and the gas density bar coincide , while the gas density bar and the synchrotron radio bar ( i.e. magnetic bar ) are phase shifted relative to each other with a @xmath359 phase difference . by figs .
[ f18 ] and [ f22 ] , the @xmath208 solution describes a qualitatively similar bar configuration as the @xmath208 configuration of figs .
[ f16 ] and [ f20 ] . for @xmath364
[ see definition ( [ 98 ] ) for @xmath236 ] , the @xmath209 solution describes a qualitatively similar bar configuration as the @xmath209 configuration of figs .
[ f16 ] and [ f20 ] . for @xmath362 ,
the @xmath209 solution describes a configuration in which the stellar density bar and the gas density bar are phase shifted relative to each other with a @xmath359 phase difference , while the gas density bar and the synchrotron radio bar ( i.e. magnetic bar ) coincide . in context of observations , the prominent bar structure of the nearby barred spiral galaxy m83 ( e.g. sukumar & allen 1989 ; neininger 1992 ; neininger et al .
1993 ) is worth being analyzed in more details in both optical and synchrotron radio bands . in terms of theoretical development of msid model and in reference to previous theoretical analyses of shu
et al . ( 2000 ) on zero - frequency ( i.e. stationary ) aligned and unaligned perturbation configurations of an isopedically magnetized sid , of lou ( 2002 ) on stationary aligned and unaligned perturbation configurations of a coplanarly magnetized sid , and of lou & shen ( 2003 ) on stationary aligned and unaligned perturbation structures in a composite system of two - fluid sids , we have constructed analytically in this paper stationary aligned and unaligned perturbation configurations in either full or partial composite system consisting of a stellar sid and a gaseous msid . while this composite model of one sid and one msid is highly idealized in many aspects , it does contain several necessary and more realistic elements that are pertinent to structures and dynamics of disc galaxies .
galactic structures may not be stationary in general , yet the stationary global mhd perturbation structures might represent a subclass , provide insights for a disc galaxy , and serve as benchmarks for numerical mhd simulations . given our model specifications , we have reached the following conclusions and suggestions . for aligned mhd perturbation structures , we derived stationary dispersion relation ( [ 46 ] ) for both full and partial composite system of one sid and one msid .
mathematically , there are three possible @xmath89 values for the square of dimensionless rotation parameter @xmath57 of the stellar sid ( or equivalently , @xmath91 ) .
physically , only those solutions with @xmath94 and @xmath390 may be conceptually applicable . in the aligned case of barred configurations with @xmath192 and @xmath136 for a full composite msid system ,
the uppermost branch remains always physically possible , with spatial phase relationships @xmath391 and @xmath392 ( where @xmath173 is the amplitude of azimuthal magnetic field perturbation ) ; that is , both @xmath152 and @xmath153 pair and @xmath152 and @xmath173 pair are out of phase .
these qualitative features for possible bars may be detectable observationally .
when parameter @xmath3 is smaller than a critical value @xmath211 defined by ( [ 94 ] ) , the middle solution branch of @xmath195 is positive and thus physically possible .
corresponding phase relationships of @xmath228 and of @xmath176 are in phase and out of phase , respectively .
but when @xmath365 defined by expression ( [ 98 ] ) , @xmath228 and of @xmath176 can become zero or even change to opposite sign ( see figs .
[ f18 ] and [ f22 ] ) .
this special @xmath240 case might be an additional possibility .
the lowest branch of @xmath195 is physically possible only when @xmath393 defined by expression ( [ 95 ] ) .
in many cases @xmath215 is smaller than @xmath213 , this branch is thus invalid .
however in some cases ( see fig .
[ f12 ] ) , this branch can rise up to become positive and thus physically possible .
the corresponding phase relationships of @xmath228 and @xmath176 are both always positive and thus in phase . for aligned perturbations , the case of
@xmath11 can be made to effectively rescale the axisymmetric background .
eccentric @xmath189 displacements may occur for arbitrary @xmath89 values in a full composite msid system .
in contrast , for a partial composite msid system , @xmath89 can no longer be arbitrary for @xmath189 ( lou 2002 ; lou & shen 2003 ) . for coplanar mhd perturbation structures of unaligned logarithmic spirals
, we obtained the stationary dispersion relation ( [ 62 ] ) for both full and partial composite msid system .
mathematically , there are three possible sets of solutions for @xmath89 or equivalently @xmath91 . for stationary logarithmic spiral configurations with @xmath192 and @xmath136 in a full composite msid system , the uppermost @xmath89 solution branch is always physically possible , with corresponding @xmath228 being always out of phase and with the phase difference between @xmath152 and @xmath173 in the range @xmath360 .
the middle branch of @xmath89 solution is physically possible when @xmath342 defined by expression ( [ 113 ] ) .
there also exists a special @xmath236 defined by expression ( [ 120 ] ) ; when @xmath3 decreases below @xmath236 , the corresponding @xmath228 changes from being in phase to being out of phase ( see fig .
[ f35 ] ) , and the phase difference between @xmath152 and @xmath173 changes from being within @xmath360 to being within @xmath363 ( see fig . [ f42 ] ) .
the special @xmath240 case may occur as well .
the lowest branch of @xmath89 solution can be physically possible when @xmath393 defined by expression ( [ 114 ] ) .
similar to the aligned case , while this lowest @xmath89 solution branch is often invalid as @xmath215 is usually less than @xmath213 , it may become physically possible for @xmath3 within interval @xmath394 $ ] ( see figs .
[ f32 ] and [ f33 ] ) for cases with @xmath395 .
correspondingly , @xmath228 remains always in phase and the phase difference between @xmath152 and @xmath173 falls in the range @xmath363 . for the marginal case of @xmath11 with radial propagations , the collapse and ring fragmentation regimes exist , while the lower ring fragmentation curve , which is a novel feature in a single msid ( lou 2002 ) , does not often appear to be positive . for cases with @xmath396
, this lower ring fragmentation branch may rise to become positive and will descend with increasing @xmath90 as shown in figs .
[ f9 ] and [ f10 ] .
this qualitative trend differs from that of a single msid studied by lou ( 2002 ) .
we further examined trends of variation as parameters @xmath4 , @xmath3 and @xmath90 vary , and concluded that the mutual gravitational coupling between the two sids appears to reduce collapse regime but to increase the danger of ring fragmentation . physical interpretations on collapse and ring fragmentation regimes
are provided ( shu et al . 2000 ; lou 2002 ; lou & shen 2003 ; shen & lou 2003 ) .
consistent with prior analyses , we realize that the ring fragmentation curve has a fixed vertical asymptote at @xmath397 with @xmath398 approximately ; this condition for the vertical asymptote remains exactly the same as that of shu et al .
( 2000 ) , lou ( 2002 ) and lou & shen ( 2003 ) . in reference to these previous analyses
, it is highly suggestive that the effective @xmath334 parameter in a composite msid system should be closely related to the minimum of the ring fragmentation curve ( shen & lou 2003 ) .
the unaligned spiral case of @xmath189 is somewhat special .
being negative , the two lower solution branches of @xmath89 are both unphysical .
the uppermost @xmath89 solution branch is physically possible when @xmath399 .
the analytic expression of @xmath378 with @xmath351 is given by equation ( [ 126 ] ) .
we have also derived the mhd virial theorem in a composite msid system , following the suggestion of lou ( 2002 ) that ratio @xmath6 may play the role of ratio @xmath262 in determining the onset criterion for bar - like instability when coplanar magnetic field is involved . similar to a composite system of two fluid sids of lou & shen ( 2003 ) , we found that solution @xmath89 of the middle @xmath209 branch may correspond to @xmath6 ratios being considerably lower than the oft - quoted value of @xmath400 ( ostriker & peebles 1973 ; binney & tremaine 1987 ) ) .
this indicates that a composite ( m)sid system tends to be less stable .
this is just one perspective of showing that the stability in a composite msid system is far more complicated than that of a single msid . in particular , additional solution branches allowed by more dynamic freedoms as well as more parameters @xmath4 , @xmath3 and @xmath90 do introduce extra dimensions for the stability problems . for example , as these dimensionless parameters vary , most vulnerable configurations related to the smallest @xmath89 change .
while our model for a composite system of msids is highly idealized , theoretical results obtained here provide a conceptual basis and useful clues for optical and synchrotron radio observational diagnostics of spiral galaxies of stars and gaseous ism with magnetic field coupled by mutual gravity on large scales . in particular , our analytical solutions for stationary mhd perturbation configurations in a composite msid system as well as analytical expressions for critical points are important and valuable for benchmarking numerical codes designed for large - scale dynamics of magnetized gaseous discs .
this research was supported in part by the asci center for astrophysical thermonuclear flashes at the university of chicago under department of energy contract b341495 , by the special funds for major state basic science research projects of china , by the tsinghua center for astrophysics , by the collaborative research fund from the national natural science foundation of china ( nsfc ) for young outstanding overseas chinese scholars ( nsfc 10028306 ) at the national astronomical observatory , chinese academy of sciences , by nsfc grant 10373009 at the tsinghua university , and by the yangtze endowment from the ministry of education through the tsinghua university .
affiliated institutions of y.q.l .
share the contribution .
frick p. , beck r. , shukurov a. , sokoloff d. , ehle m. , kamphuis j. , 2000 , mnras , 318 , 925 frick p. , beck r. , berkhuijsen e.m .
, patrickeyev i. , 2001 , mnras , 327 , 1145 galli d. , shu f.h . , laughlin g. , lizano s. , 2001 , apj , 551 , 367 for the spiral case of @xmath192 , a substitution of expressions ( [ 11 ] ) , ( [ 15 ] ) , ( [ 20 ] ) , ( [ 21 ] ) and ( [ 87 ] ) into equation ( [ 62 ] ) would lead to a cubic algebraic equation in terms of @xmath195 . by setting the constant term of this cubic equation equal to zero , we derive conditions for @xmath401 , namely @xmath402 where the three coefficients @xmath156 , @xmath197 and @xmath198 are defined by @xmath403 @xmath404 and @xmath405 in the presence of gravitational coupling between one sid and one msid
, we expect three sets of independent mathematical solutions in general
. however , for cases of marginal stabilities for spiral configurations with @xmath11 , the solution curves qualitatively resemble those determined for a single sid ( shu et al .
2000 ) in most cases , and the two extra solutions turn out to be negative . moreover , the lower ring fragmentation curve in a single msid ( see fig
. 1 of lou 2002 ) remains often indiscernible ( see figs .
[ f1]@xmath15[f9 ] ) . here , we provide a complete structure for solution curves in figs .
[ f4 ] and [ f5 ] .
we note that the basic positive @xmath91 versus @xmath288 profile is qualitatively similar to fig . 1 of lou ( 2002 ) ( see fig .
[ f4 ] ) . however , the @xmath89 curve corresponding to the lower ring fragmentation curve of @xmath91 is negative in most cases , indicating that this branch is physically invalid ( see figs .
[ f4 ] and [ f5 ] ) . on the other hand ,
when @xmath3 approaches @xmath213 , this branch of @xmath89 can be raised above zero and thus becomes physically valid in this limit ( see fig .
[ f9 ] ) . in this appendix c
, we provide the complete form of the cubic equation of @xmath195 for the aligned case of coplanar mhd perturations in a composite msid system for interested readers or potential users in numerical mhd simulations . from dispersion relation (
[ 46 ] ) and relation ( [ 87 ] ) , we readily derive @xmath406 where the four coefficients @xmath156 , @xmath197 , @xmath198 and @xmath199 are defined by @xmath407 @xmath408 @xmath409 and in this appendix d , we provide the complete form of the cubic equation of @xmath195 for the unaligned spiral case of coplanar mhd perturbations in a composite msid system for interested readers or potential users in numerical mhd simulations . from dispersion relation (
[ 46 ] ) and relation ( [ 87 ] ) , we readily derive @xmath406 where the four coefficients @xmath156 , @xmath197 , @xmath198 and @xmath199 are defined by @xmath412 @xmath413 @xmath414 + 16\,{m}^{4}{\lambda}^{2}-36\,{m}^{2}\delta\,{\lambda}^{2}+20\ , { \lambda}^{4}{m}^{2}\delta \\ % \nonumber\\ + 48\,{m}^{2}{\lambda}^{2}{\alpha}^{2}+8\ , \delta\,{\lambda}^{2}-8\,{m}^{4}{\lambda}^{4}\delta \\ % \nonumber\\ + 16\,{m}^{4}\delta \,{\lambda}^{2}+20\,{m}^{2}{\lambda}^{4}+16\,{m}^{4}\mathcal n_m(\alpha){\lambda}^{2 } \\ % \nonumber\\ + 8\ , \mathcal n_m(\alpha){\lambda}^2{\alpha}^2 - 64\,{\alpha}^2 -96\,{\lambda}^2{\alpha}^2
\\ % \nonumber\\ -40\,{m}^2\mathcal n_m(\alpha)-64\,{\alpha}^2\delta-48\,{\alpha}^2 \mathcal n_m(\alpha ) \\
% \nonumber\\ + 32\,{\alpha}^2{\lambda}^4 + 2\,{\lambda}^4 \mathcal n_m(\alpha)m^2 - 8\,{\lambda}^4\delta \\
% \nonumber\\ + 16\ , m^2\mathcal n_m(\alpha)\delta\ , { \lambda}^2{\alpha}^2\beta -16\,m^2{\lambda}^4\delta\,{\alpha}^2 -24\,{m}^2{\lambda}^4\mathcal n_m(\alpha){\alpha}^2
\\ % \nonumber\\ + 48\,m^2\delta\,{\lambda}^2{\alpha}^2 + 32\,m^2{\alpha}^2 + 32\,m^2 { \alpha}^2\delta \\ % \nonumber\\ -8\,m^4{\lambda}^4 + 64\,m^2\mathcal n_m(\alpha){\lambda}^2{\alpha}^2 - 64\,m^2 \mathcal n_m(\alpha){\lambda}^2{\alpha}^2\beta \\ % \nonumber\\ + 16\,{\alpha}^4\mathcal n_m(\alpha)\delta\,{\lambda}^2\beta+60\,\mathcal n_m(\alpha)\beta-256\,m^2\mathcal n_m(\alpha){\alpha}^2\beta \\ % \nonumber\\ + 80\,\mathcal n_m(\alpha)\delta\,{\alpha}^2\beta+32\,{\alpha}^4\mathcal n_m(\alpha)\delta\,\beta+32\,m ^2\mathcal n_m(\alpha)\delta\,{\alpha}^2\beta \\
% \nonumber\\ + 96\,m^2{\alpha}^2\delta\,\beta+72 \,\mathcal n_m(\alpha)\delta\,m^2\beta-48\,{\alpha}^2\beta \\
% \nonumber\\ -270\,\delta\,\beta+192\ , \mathcal n_m(\alpha)m^2\beta+224\,\mathcal n_m(\alpha){\alpha}^2\beta \\
% \nonumber\\ -48\,{\alpha}^2\delta\,\beta+96\,m^2{\alpha}^2\beta-192\,m^4 \mathcal n_m(\alpha)\beta \\
% \nonumber\\ + 18\,\mathcal n_m(\alpha)\delta\,\beta+32\,{\alpha}^2 { \lambda}^4\delta+18\,{\lambda}^{2}\beta % \\
\nonumber \end{split}\ ] ] @xmath415 @xmath416 @xmath417 in section 4.3.2 , we indicated that there exists a root @xmath236 for the @xmath209 solution branch of the @xmath176 , and in the case of figure [ f41 ] , the value of this @xmath236 is about 0.5589 by formula ( [ 120 ] ) .
however , one might worry that from figure [ f41 ] in which @xmath3 ranges from 1 to 5 , the trend of variation of @xmath209 may not lead to such a @xmath236 where @xmath209 vanishes . here , we provide the @xmath176 versus @xmath3 curve in which @xmath3 ranges from 0.2 to 1.5 ( all other parameters are exactly the same as those of figure [ f41 ] and only the @xmath209 branch is shown here ) to support our statement in the main text . in figure e1 , it is clear that this @xmath209 branch does indeed go across the horizontal @xmath3 axis at about 0.5589 as expected from formula ( [ 120 ] ) . | we construct aligned and unaligned stationary perturbation configurations in a composite system of stellar and coplanarly magnetized gaseous singular isothermal discs ( sids ) coupled by gravity .
this study extends recent analyses on ( magnetized ) sids of shu et al . , lou and lou & shen .
by this model , we intend to provide a conceptual framework to gain insights for multi - wavelength large - scale structural observations of disc galaxies .
both sids are approximated to be razor - thin and are in a self - consistent axisymmetric background equilibrium with power - law surface mass densities and flat rotation curves .
the gaseous sid is embedded with a coplanar azimuthal magnetic field @xmath0 of a radial scaling @xmath1 that is not force - free . in comparison with sid problems studied earlier , there
exist three possible classes of stationary solutions allowed by more dynamic freedoms . to identify physical solutions ,
we explore parameter space involving three dimensionless parameters : ratio @xmath2 of alfvn speed to sound speed in the magnetized gaseous sid , ratio @xmath3 for the square of the stellar velocity dispersion to the gas sound speed and ratio @xmath4 of the surface mass densities of the two sids . for both aligned and unaligned spiral cases with azimuthal periodicities @xmath5 , one of the three solution branches
is always physical , while the other two branches might become invalid when @xmath3 exceeds certain critical values .
for the onset criteria from an axisymmetric equilibrium to aligned secular bar - like instabilities , the corresponding @xmath6 ratio , which varies with @xmath2 , @xmath3 and @xmath4 , may be considerably lower than the oft - quoted value of @xmath7 , where @xmath8 is the total kinetic energy , @xmath9 is the total gravitational potential energy and @xmath10 is the total magnetic energy . for unaligned spiral cases ,
we examine marginal instabilities for axisymmetric ( @xmath11 ) and non - axisymmetric ( @xmath12 ) disturbances .
the resulting marginal stability curves differ from the previous ones .
the case of a composite partial msid system is also investigated to include the gravitational effect of an axisymmetric dark matter halo on the sid equilibrium .
we further examine the phase relationship among the mass densities of the two sids and azimuthal magnetic field perturbation .
our exact global perturbation solutions and critical points are valuable for testing numerical magnetohydrodynamic codes . for galactic applications ,
our model analysis contains more realistic elements and offer useful insights for structures and dynamics of disc galaxies consisting of stars and magnetized gas .
ism : general galaxies : kinematics and dynamics galaxies : spiral galaxies : structure mhd waves . |
that vacuum polarization process _
la _ heisenberg - euler - schwinger ( @xcite , @xcite ) can occur in the field of a kerr newmann embh and that they naturally lead to a model for gamma - ray bursts was pointed out in @xcite .
the basic energetics of the process , governed by the christodoulou - ruffini mass formula , for an embh gives as shown in @xcite , @xmath4 with @xmath5 where @xmath6 and @xmath7 are respectively the mass energy , the angular momentum and the charge of the embh and @xmath8 is the irreducible mass , @xmath9 is the horizon radius , @xmath10 is the quasi - spheroidal cylindrical coordinate of the horizon evaluated at the equatorial plane and @xmath11 is the horizon surface area .
extreme black holes satisfy the equality in eq.([s1 ] ) . the vacuum polarization process being `` reversible '' transformations in the sense of @xcite can extract an energy up to 29% of the mass - energy of an extremal rotating black hole and up to 50% of the mass - energy of an extremely magnetized and charged black hole .
although in general such a process is endowed with axial symmetry , in order to clarify the pure interplay of the gravitational and electrodynamical phenomena and also for simplicity , we have examined ( @xcite and @xcite ) the limiting cases of such a process in the field of a reissner - nordstrom geometry whose spherically symmetric metric is given by @xmath12 where @xmath13 \equiv - \alpha^2(r)$ ] and @xmath14 .
the dyadosphere , defined ( see fig . 1 of @xcite ) as the region where the electric field exceeds the critical value , @xmath15 ( @xcite , @xcite ) , where @xmath16 and @xmath17 are the mass and charge of the electron , extends between the horizon radius @xmath18 where we have introduced the dimensionless mass and charge parameters @xmath19 , @xmath20 , and the outer radius @xmath21 consequently the local number density of electron and positron pairs created in the dyadosphere as a function of radius @xmath22 ~ , \label{nd}\ ] ] and their energy density is given by @xmath23 ( see figs.[2 ] & [ 3 ] of @xcite ) .
the total energy of pairs converted from the static electric energy and deposited within the dyadosphere is then @xmath24 ~ , \label{tee}\ ] ] and the total number of electron and position pairs in the dyadosphere is @xmath25 , \label{tn}\ ] ] where @xmath26 ( see @xcite ) . in the limit of @xmath27 , eq.([tee ] ) leads to @xmath28 , which coincides with the energy extractable from embhs by reversible processes ( @xmath29 ) , namely @xmath30(see fig . 4 of @xcite ) .
due to the very large pair density given by eq.([nd ] ) and to the sizes of the cross - sections for the process @xmath31 , the system is expected to thermalize to a plasma configuration for which @xmath32 where @xmath33 is the number of @xmath0-pairs created in the dyadosphere(see @xcite ) .
these are the initial conditions for the evolution of the dyadosphere . in a previous paper ( @xcite )
we presented the temporal evolution of the dyadosphere in vacuum giving origin to an extremely sharply defined and extremely relativistic expanding pulse of pair and electromagnetical radiation of a constant length in the laboratory frame : the pem pulse . in this paper we present results of the collision of the pem pulse with a remnant of baryonic matter surrounding the just formed black hole .
we assume the pem pulse to collide with a shell of baryonic matter of constant density and at a radius of the order of 100 times the radius of the dyadosphere @xmath34 eq.([rc ] ) .
the shells have this thickness of the order of 10 times @xmath34 .
the mass - energies of the shells are taken to be @xmath35 of the total energy of the dyadosphere .
the work contains the following sections : sect .
[ hydrodynamic ] presents discussions of the hydrodynamical equations of a pem pulse interacting with the baryonic shell ; sect .
[ baryon ] defines the parameters of the baryonic shells , the behaviour of the pem pulse colliding with a baryon shell as well as before and after the collision are presented ; in sect .
[ comparison ] the results of a numerical calculation solving the hydrodynamical equations of sect .
[ hydrodynamic ] are compared to the results of the analytical model of slab approximation for selected parameters .
the evolution of the pem pulse in vacuum was treated in a previous work ( @xcite ) .
we here generalize that treatment to the case in which baryonic matter is present and we outline the relevant relativistic hydrodynamic equations . as in the previous treatment ( @xcite ) , we assume the plasma fluid of @xmath0-pairs , photons and baryonic matter to be described by the stress - energy tensor @xmath36 where @xmath37 and @xmath38 are respectively the total proper energy density and pressure in the comoving frame of the plasma fluid .
the @xmath39 is the four - velocity of the plasma fluid .
we have @xmath40 where @xmath41 and @xmath42 are the radial and temporal contravariant components of the 4-velocity .
the conservation law of baryon number can be expressed as a function of the proper baryon number density @xmath43 @xmath44 the radial component of the energy - momentum conservation law of the plasma fluid reduces to @xmath45 = 0 ~. \label{cmom2}\end{aligned}\ ] ] the component of the energy - momentum conservation law of the plasma fluid equation along a flow line is @xmath46=0 ~. \label{conse1}\end{aligned}\ ] ] defining the total proper internal energy density @xmath47 and the baryonic mass density @xmath48 in the comoving frame of the plasma fluid , @xmath49 and using the law of baryon - number conservation ( [ contin ] ) , from eq .
( [ conse1 ] ) we have @xmath50 recalling that @xmath51 , where @xmath52 is the comoving volume and @xmath53 is the proper time for the plasma fluid , we have along each flow line @xmath54 where @xmath55 is total proper internal energy of the plasma fluid .
we represent the equation of state by the introduction of a thermal index @xmath56 @xmath57 we now turn to the analysis of @xmath0 pairs initially created in the dyadosphere and ionized electrons contained in the baryonic matter .
let @xmath58 and @xmath59 be the proper number densities of electrons and positrons associated to pairs created , and @xmath60 the proper number densities of ionized electrons , we clearly have @xmath61 where @xmath62 is the number of @xmath0 pairs and @xmath63 the average atomic number @xmath64 ( @xmath65 for hydrogen atom and @xmath66 for general baryonic matter ) .
the rate equation for electrons and positrons gives , @xmath67 , \label{e+contin}\\ ( n_{e^-}u^\mu)_{;\mu}&=&(n_{e^-}u^t)_{,t}+{1\over r^2}(r^2 n_{e^-}u^r)_{,r}\nonumber\\ & = & \overline{\sigma v } \left[n_{e^-}(t)n_{e^+}(t ) - n_{e^-}n_{e^+}\right ] , \label{e - contin}\\ ( n^b_{e^-}u^\mu)_{;\mu}&=&(n^b_{e^-}u^t)_{,t}+{1\over r^2}(r^2 n^b_{e^-}u^r)_{,r}\nonumber\\ & = & \overline{\sigma v } \left[n^b_{e^-}(t)n_{e^+}(t ) - n^b_{e^-}n_{e^+}\right ] , \label{tbe - contin}\end{aligned}\ ] ] where @xmath68 is the mean of the product of the annihilation cross - section and the thermal velocity of the electrons and positrons , @xmath69 are the proper number densities of electrons and positrons associated to the pairs , given by appropriate fermi integrals with zero chemical potential , and @xmath70 is the proper number density of ionized electrons , given by appropriate fermi integrals with non - zero chemical potential @xmath71 , at an appropriate equilibrium temperature @xmath72 .
these rate equations can be reduced to @xmath73 , \label{econtin}\\ ( n^b_{e^-}u^\mu)_{;\mu}&=&(n^b_{e^-}u^t)_{,t}+{1\over r^2}(r^2 n^b_{e^-}u^r)_{,r}=0 , \label{becontin}\\ frac&\equiv&{n_{e^\pm}\over n_{e^\pm}(t)}={n^b_{e^-}(t)\over n^b_{e^-}}. \label{be - contin}\end{aligned}\ ] ] equation ( [ becontin ] ) is just the baryon - number conservation law ( [ contin ] ) and ( [ be - contin ] ) is a relationship obeyed @xmath74 and @xmath75 .
the equilibrium temperature @xmath72 is determined by the thermalization processes occurring in the expanding plasma fluid with a total proper energy density @xmath37 , governed by the hydrodynamical equations ( [ contin],[cmom2],[conse1 ] ) .
we have @xmath76 where @xmath77 is the photon energy density , @xmath78 is the baryonic mass density which is considered to be non relativistic in the range of temperature @xmath72 under consideration , and @xmath79 is the proper energy density of electrons and positrons pairs given by , @xmath80 where @xmath81 is obtained by integration of eq.([econtin ] ) and @xmath82 is the proper energy density of electrons(positrons ) obtained from zero chemical potential fermi integrals at the equilibrium temperature @xmath72 . whereas , @xmath83 is the energy density of the ionized electrons , obtained by the ionization of baryonic matter , @xmath84 where @xmath60 is obtained by integration of eq.([becontin ] ) and @xmath85 is the proper energy density of ionized electrons obtained from an appropriate fermi integral of non - zero chemical potential @xmath71 at the equilibrium temperature @xmath72 .
having intrinsically defined the equilibrium temperature @xmath72 in eq.([eeq ] ) , we can also , analogously , evaluate the total pressure @xmath86 where @xmath87 is the photon pressure , @xmath88 and @xmath89 given by , @xmath90 where the pressures @xmath91 are determined by zero chemical potential fermi integrals , and @xmath92 is the pressure of the ionized electrons , evaluated by an appropriate fermi integral of non - zero chemical potential @xmath71 at the equilibrium temperature @xmath72 . in eq.([eep ] ) , the ions pressure @xmath93 is negligible by comparison with the pressures @xmath94 , since baryons and ions are expected to be non - relativistic in the range of temperature @xmath72 under consideration . finally , using eqs.([eeq],[eep ] ) , we compute the thermal factor @xmath95 of the equation of state ( [ state ] ) .
the calculation is continued as the plasma fluid expands , cools and the @xmath0 pairs recombine , until it becomes optically thin : @xmath96 where @xmath97 is the thomson cross - section and the integration is over the radial interval of the pem - pulse in the comoving frame . at this point
the energy is virtually entirely in the form of free - streaming photons and the calculation is stopped .
the pem pulse expansion in the absence of baryonic matter has been discussed in a previous paper ( @xcite ) where the quasi - analytical approach of an expanding shell of constant thickness in the laboratory frame was adopted and validated by comparison with the numerical integration of the general relativistic hydrodynamical equations .
we here generalize these results by examining the collision of the pem pulse with baryonic matter and adopting the constant - thickness approximation both for the description of the collision and the further expansion of the pem pulse by a simplified approach to the system of equations outlined in the previous paragraph .
we first recall the main results of the pem pulse expanding in vacuum : we indicate by @xmath98 the four - velocity of the slab and by @xmath99 the constant width of the slab in the laboratory frame of the plasma fluid , the average bulk relativistic gamma - factor @xmath100 is , @xmath101 the evolution of the slab is governed by the total energy and entropy conservations , which are cast into the following equations as a function of the coordinate volume of the plasma fluid expanding from @xmath102 to @xmath103 , @xmath104 where the proper volume @xmath52 of the plasma fluid @xmath105 and the thermal index @xmath95 eq.([state ] ) , a slowly - varying function of the state with values around 4/3 , has been approximately assumed to be constant .
the coordinate number density of @xmath106-pairs in equilibrium is @xmath107 and the coordinate number density of @xmath106-pairs @xmath108 .
these equations have already been numerically integrated ( @xcite ) .
the baryonic matter remnant is assumed to be distributed well outside the dyadosphere in a shell of thickness @xmath109 between an inner radius @xmath110 and an outer radius @xmath111 at a distance from the embh at which the original pem pulse expanding in vacuum has not yet reached transparency , @xmath112 the total baryonic mass ( @xmath113 ) is assumed to be a fraction of the dyadosphere initial total energy @xmath114 . the total baryon - number ( @xmath115 )
is then given by @xmath116 where @xmath117 is a parameter in the range @xmath35 and where @xmath118 is the proton mass .
the baryon number density @xmath119 is assumed to be a constant , @xmath120 as the pem pulse reaches the region @xmath121 , it interacts with the baryonic matter which is assumed to be at rest . in our simplified quasi - analytic model
we make the following assumptions to describe this interaction : * the pem pulse does not change its geometry during the interaction ; * the collision between the pem pulse and the baryonic matter is assumed to be inelastic , * the baryonic matter reaches thermal equilibrium with the photons and pairs of the pem pulse .
these assumptions are valid if : ( i ) the total energy of the pem pulse is much larger than the total mass - energy of baryonic matter @xmath122 , @xmath123 , and ( ii ) the comoving number density ratio @xmath124 of pairs and baryons at the moment of collision is extremely high ( e.g. , @xmath125 , and ( iii ) the pem pulse has a large value of lorentz factor ( @xmath126 ) . in the collision process between the pem pulse and the baryonic matter at @xmath127 , we impose the total energy and momentum conservations .
we consider the collision process between two radii @xmath128 , satisfying @xmath129 and @xmath130 .
the amount of baryonic mass acquired by the pem pulse is @xmath131 where @xmath132 is the mean - density of baryonic matter at rest .
the conservation of total energy leads to the estimate of the corresponding quantities before ( with @xmath133 " ) and after such a collision @xmath134 where @xmath135 is the corresponding increase of internal energy due to the collision .
similarly the momentum - conservation gives @xmath136 where radial component of the four - velocities of the pem pulse @xmath137 and @xmath95 thermal index .
we then find @xmath138,\label{heat}\\ \bar\gamma & = & { a\over\sqrt{a^2 - 1}},\hskip0.5 cm a\equiv { \bar\gamma_\circ \over u^\circ_r}+ { \delta m\over ( \gamma\bar\epsilon_\circ + \bar\rho^\circ_b)\bar\gamma_\circ u^\circ_r{\cal v}_\circ}. \label{dgamma}\end{aligned}\ ] ] these equations determine the gamma - factor @xmath100 and the internal energy density @xmath139 in the capture process of baryonic matter by the pem pulse .
after collision ( @xmath140 ) , the further adiabatic expansion of pem pulse is described by the total baryon number , energy and entropy conservations , i.e. , the following hydrodynamical equations which generalize those derived in our previous paper ( @xcite ) with @xmath141 , @xmath142 in these equations ( @xmath140 ) the comoving baryonic mass- and number densities are @xmath143 and @xmath144 , where @xmath52 is the comoving volume of the pem pulse .
the integration is continued untill the transparency condition in eq.([thin ] ) is reached .
the numerical integration of the general relativistic equations given in sect . [ hydrodynamic ]
have already been presented in a series of papers ( see @xcite ) .
we can then proceed to compare and contrast the results obtained by the numerical integration of the eqs.([contin])-([conse1 ] ) and the simplified quasi - analytical approach given by eq.([scale])-([paira ] ) , ( [ ecc])-([pcc ] ) and ( [ be])-([paira ] ) .
we select the specific example of an embh of @xmath145 with a charge to mass ratio @xmath146 . from the equations already published
( @xcite ) , the total energy in the dyadosphere is @xmath147ergs .
the pem pulse is assumed to collide with a baryonic shell of thickness @xmath148 at rest at a radius of @xmath149 .
we have considered five different cases : ( 1 ) a shell of baryonic mass @xmath150 ( @xmath151 ergs rest energy ) , from corresponding to a parameter @xmath152 ; ( 2 ) @xmath153 ( @xmath154 ergs , @xmath155 ) ; ( 3 ) @xmath156 ( @xmath157 ergs , @xmath158 ) ; ( 4 ) @xmath159 ( @xmath160 ergs , @xmath161 ) ; ( 5 ) @xmath162 ( @xmath163 ergs , @xmath158 ) . the collision between the expanding slab and the baryonic matter shell
is treated as an inelastic collision in both calculations .
we first proceed to a qualitative analysis of the evolution of the pem pulse in eqs .
( [ contin])-([conse1 ] ) from figs .
[ b6fig],[b6detailfig ] corresponding to @xmath152 and figs .
[ b5.5fig],[b5fig],[b4.5fig ] corresponding to @xmath164 , we see that the pem pulse after collision with the baryonic matter shell continues to expand as a one dimensional slab .
we see , however , from figs .
[ b4fig ] corresponding to @xmath165 that the expansion after collision becomes much more complex and the constant - thickness approximation ceases to be valid . from this qualitative analysis
we now proceed to a quantitative analysis in fig .
[ twocodecompare ] .
we compare and contrast the bulk gamma factor as computed from the constant - thickness approximation and the one from the full set of eqs .
( [ contin])-([conse1 ] ) .
the computation refers to the case @xmath152 where an excellent qualitative agreement with the one - dimensional - slab approximation has been found .
the extremely good agreement validates the constant - thickness approximation .
this agreement has been found up to values of @xmath117 no larger than @xmath3 .
we now proceed to some specific prediction of grb features computed by using the constant - thickness approximation and the eqs .
( [ scale])-([paira ] ) , ( [ ecc])-([pcc ] ) and ( [ be])-([paira ] ) in the range of validation of this approximation just defined in the previous paragraph . as an example for clearly showing the evolution of pem pulses colliding with baryonic matter
, we take the following black hole mass and charge , as well as the mass of baryon remnant as a typical case : @xmath166 where the total energy of dyadosphere @xmath167ergs , so the total number of @xmath0-pairs created in the dyadosphere ( given by eq.([tn ] ) ) @xmath168 , and baryonic mass @xmath169 .
this baryonic mass is close to the limit of validation of the slab model shown in sect .
[ comparison ] . we have assumed the baryonic matter at a distance of @xmath170 , very close to the transparency condition of the pem pulse in vacuum ( see @xcite ) . in fig .
[ 3gamma ] we represent the lorentz factor of the pem pulse as a function of the radius for collision with different amounts of baryonic matter , corresponding respectively to @xmath171 , @xmath172 and @xmath173 .
the diagram extends to values of the radial coordinate at which the transparency condition given by eq.([thin ] ) is reached
. also represented , for each diagram , is the `` asymptotic '' lorentz factor : @xmath174 the closer the @xmath100 value approaches , at transparency , the `` asymptotic '' value ( [ asymp ] ) , the smaller the intensity of the radiation emitted in the burst , and the larger the amount of kinetic energy left in the baryonic matter .
this point is further clarified in fig .
[ fgamma ] , where are plotted the @xmath100-factor at transparency and the `` asymptotic '' one as functions of the baryonic matter .
it is interesting that , for a given embh , there is a maximum value of the @xmath100-factor at transparency .
after that maximum value , the energy available for the grb is smaller in intensity , and at decreasing values of the energy , for increasing values of the baryonic mass .
the temperature in the laboratory frame @xmath175 at the transparency point is plotted as a function of the baryonic mass in fig .
[ energypeak ] : it strongly decreases as the baryonic mass increase .
the @xmath175 is related to the observed energy - peak of the photon - number spectrum ( see e.g. , @xcite ) .
we plot in fig .
[ fintkin ] the energy radiated in the burst and the final kinetic energy of baryonic matter .
we find that , for small values of @xmath117 ( around @xmath2 ) , almost all total energy is radiated as grb ( see also our previous paper @xcite ) , and very little energy is left as kinetic energy of baryonic matter as afterglow . while for @xmath176 roughly only @xmath3 of the total initial energy of the dyadosphere is radiated away as grb , and almost all energy is restored as the kinetic energy of the baryonic matter .
it is also clear that for @xmath177 the intensity of the burst ( see also fig .
[ 3gamma ] ) and the observed radiation frequency drifted to smaller and smaller values and are of little astrophysical interest from the point of view of grbs .
for such values , the energy of the dyadosphere is transferred practically totally to the bulk kinetic energy of the baryonic matter .
@xmath171 is also the limit of the validation of our computations based on the analytical slab model , described in sect .
[ baryon ] . for values of baryonic matter in between these two extremal cases ( @xmath178 and @xmath171 ) the analytical slab model covers the whole range of the observed properties of gamma ray bursts .
we turn now to the thermodynamic parameters relevant in the evolution of the pem pulse . in fig .
[ tem ] the temperature of pem pulse , both in the comoving and in the laboratory frame , are given as a function of the radius for a typical case ( @xmath179 and @xmath171 ) . in fig .
[ intkin ] , we plot the total energy of the non baryonic components of the pem pulse , which includes both thermal and kinetic energy , and the kinetic energy of the baryonic matter as functions of the radius , for the typical case @xmath179 and @xmath171 .
the total energy of the non baryonic components of the pem pulse is equal to @xmath180 before the collision ( see @xcite ) and drops after the collision . while the kinetic energy of baryonic matter @xmath181 increases as function of radius for @xmath182 .
the sum of both them is equal to the total energy @xmath183ergs during the evolution of the pem pulse . the value of the total energy of the non baryonic components of the pem pulse at the transparency point , the ending point of the curve in fig .
[ intkin ] , is the energy released in the burst . we have discussed this energy as function of baryonic masses in fig . [ fintkin ] . before and after the collision , the condition of entropy conservation applies , and we have : @xmath184 @xmath185 where @xmath186 is the thermal energy of baryonic matter , and we can neglect , in eq .
( [ entrdopo ] ) , the pressure of the baryonic matter . a sudden increase of the entropy occurs during the collision both for the addition of the baryonic matter , and for the thermal reheating due to the inelastic collapse of the pem pulse with the baryonic matter at rest . from the energy and momentum conservations , we obtain for the values of the @xmath100 and the temperature during the collision :
the proper internal energy , @xmath187 , of the plasma increases as @xmath188 and the slab is decelerated , in terms of lorentz factor @xmath189 , as @xmath190 as baryon mass , @xmath191 , is incrementally gained . before the collision the pem pulse expands keeping its temperature in the laboratory frame constant , while its temperature in the comoving frame falls ( see @xcite ) . during the collision ,
a heating of the plasma due to the energy and momentum conservation occurs ( see also fig .
[ pair ] , where reheating process leads to an increment of the number density of @xmath0 pairs ) .
as the system expands further , both the comoving temperature and the temperature in the laboratory frame decreases , since the total energy of the @xmath0 pairs and the photons before the collision is constant and equal to @xmath180 , while after the collision @xmath192 where @xmath193 is only the thermal energy of the baryons .
it is also interesting to monitor the change of temperature in the comoving frame before the collision and after the collision ( see also fig .
[ entropy ] ) . before the collision , due to the entropy conservation , in the process of @xmath0 annihilation the factor @xmath194 ( where the subscript @xmath195 " means initial value " ) increases to a value near to @xmath196
( see @xcite ; @xcite ) since the collision occurs at @xmath197 , near the condition of transparency .
the number of @xmath0 pairs has now been reduced drastically .
the further jump in the value of the ratio @xmath194 is principally due to the energy and momentum conservation during the inelastic collision .
after the collision , there is a small reheating due to the annihilation of the @xmath0 pairs created in the collision ( see fig .
[ pair ] ) , which can not be seen on the scale of our plot .
then , the ratio remains constant .
by the direct comparison with the numerical integration of the complete relativistic hydrodynamical equations , the use of the constant - thickness approximation has been validated for values of the parameter @xmath198 . for @xmath199 the amount of energy released at transparency in the burst decreases , and its energy drifts toward low energy values , which are of little interest for the astrophysics of grb .
we conclude that the constant - thickness approximation is valid for all astrophysically relevant situations for the analysis of the grb at transparency .
based on this validation we have studied the evolution of a pem pulse in the presence of selected amounts of baryonic matter .
we have studied for a typical case of @xmath179 embh and @xmath200 baryonic shell , the thermal , the bulk lorentz factor evolution of the pem pulse as well as the kinetic energy left over in the baryonic matter after the decoupling of matter and radiation and the emission of the grb .
additional results corresponding to a larger range of masses and charges of the embh and the correlations between the peak energy and the duration of grbs to be expected in our model will be considered in a future paper , together with the results of analyzing the interaction of the kinetic energy left over in the baryonic matter , after the decoupling of matter and radiation , with the interstellar medium .
work by jds and jrw was performed under the auspices of the u.s .
department of energy by university of california lawrence livermore national laboratory under contract no . w-7405-eng-48 . | the interaction of an expanding pair - electromagnetic pulse ( pem pulse ) with a shell of baryonic matter surrounding a black hole with electromagnetic structure ( embh ) is analyzed for selected values of the baryonic mass at selected distances well outside the dyadosphere of an embh .
the dyadosphere , the region in which a super critical field exists for the creation of @xmath0 pairs , is here considered in the special case of a reissner - nordstrom geometry .
the interaction of the pem pulse with the baryonic matter is described using a simplified model of a slab of constant thickness in the laboratory frame ( constant - thickness approximation ) as well as performing the integration of the general relativistic hydrodynamical equations .
the validation of the constant - thickness approximation , already presented in a previous paper ruffini , et al.(1999 ) for a pem pulse in vacuum , is here generalized to the presence of baryonic matter .
it is found that for a baryonic shell of mass - energy less than 1% of the total energy of the dyadosphere , the constant - thickness approximation is in excellent agreement with full general relativistic computations .
the approximation breaks down for larger values of the baryonic shell mass , however such cases are of less interest for observed gamma ray bursts ( grbs ) . on the basis of numerical computations of the slab model for pem pulses ,
we describe ( i ) the properties of relativistic evolution of a pem pulse colliding with a baryonic shell ; ( ii ) the details of the expected emission energy and observed temperature of the associated grbs for a given value of the embh mass ; @xmath1 , and for baryonic mass - energies in the range @xmath2 to @xmath3 the total energy of the dyadosphere . |
collective departure is a decision - making faced by all social species that travel in groups . in this process
, an individual generally initiates the movement of the group out of a residence site or towards a new direction .
the identity and motivation of this initiator can widely vary according to the social organisation of the considered species @xcite . on the one hand ,
the leadership is often assumed by a unique or a subset of individuals that monopolise the decisions in hierarchical societies .
these individuals can be older @xcite , dominant @xcite or of a specific sex @xcite .
these characteristics are generally long - lasting and result in a consistant leadership over time , generally observed in stable and closed groups . on the other hand , the initiators can also be temporarily more motivated due to their physiological state @xcite , level of information @xcite or position in the group @xcite . in these cases , the initiation can be done by any individuals of the group without consistency over time . this mechanism is often present in social species that live in open groups with no consistent membership like bird flocks or fish schools .
although each individual can initiate collective movement in these more egalitarian societies , some characteristics may enhance the probability of some members to take the leadership .
for example , bold individuals that have a higher tendency to explore new areas will more often lead departures @xcite .
similarly , group members with higher nutritional needs will be more motivated to initiate movements towards foraging spots @xcite .
therefore , even in non - hierarchical species , leadership can be heterogeneously distributed among the group members . in this context
, we studied the distribution of the leadership in the zebrafish _ danio rerio_. in its natural habits , _ danio rerio _ is a gregarious species that live in small groups ( a few to a dozen individuals ) in shallow freshwaters @xcite .
it has become a widely studied and well known model organism in genetics and neuroscience @xcite but also in ethology . in this context ,
our goal is to evaluate the presence of leaders or their emergence during successive collective departures .
to do so , we observe groups of 2 , 3 , 5 , 7 and 10 zebrafish swimming in an experimental arena consisting of two rooms connected by a corridor .
our aim is to measure the number of collective departure from one room to the other that are initiated by each fish .
then , we put in relation the propensity of the individuals to lead departures to the number of attempts that they made as well as their swimming speed .
fish experiments were performed in accordance with the recommendations and guidelines of the buffon ethical committee ( registered to the french national ethical committee for animal experiments # 40 ) after submission to the state ethical board for animal experiments .
the fish were reared in housing facilities zebtec and fed two times a day ( special diets services sds-400 scientific fish food ) .
we kept fish under laboratory conditions , @xmath0 , 500@xmath1s salinity with a 10:14 day : night light cycle .
water ph was maintained at 7 and nitrites ( no@xmath2 ) are below 0.3 mg / l .
all zebrafish observed in this study were 6 - 12 months old at the time of the experiments .
we observed groups of zebrafish swimming in an arena consisting of two square rooms connected by a corridor starting at one corners of each room placed in 100 x 100 x 30 experimental tank ( fig .
[ fig : setup ] ) .
the walls of the arena were made of white opaque pmma .
the water depth was kept at 6 cm during the experiments in order to keep the fish in nearly 2d to facilitate their tracking .
one lamp ( 400w ) was placed on the floor at each edge of the tank which is 60 cm above the floor to provide indirect lightning .
the whole setup is confined behind white sheets to isolate experiments and homogenize luminosity .
a high resolution camera was mounted 1.60 m above the water surface to record the experiment at a resolution of 2048 x 2048 and at 15 frames per second .
we observed 12 groups of two , three , five , seven and ten adult laboratory wild - type zebrafish ( _ danio rerio _ ) ab strain during one hour for a total of 60 experiments . before the trials
, the fish were placed with a hand net in a cylindrical arena ( 20 cm diameter ) in one of the two rooms . following a 5 minutes acclimatisation period , the camera started recording and the fish were released and able to swim in the experimental arena .
after one hour , the fish were caught by a hand net and replaced in the rearing facilities .
the videos were analysed off - line by the idtracker software @xcite .
this multi - tracking software extracts specific characteristics of each individual and uses them to identify each fish without tagging throughout the video .
this method avoids error propagation and is able to successfully solves crossing , superposition and occlusion problems .
however , the tracking system failed to correctly track one experiment with two fish , one experiment with five fish and two experiments with ten fish . therefore , these four experiments were excluded from our analysis . for all other experiments
, we obtained the coordinates @xmath3 of all fish at each time step @xmath4 . with these coordinates
, we built the trajectories of each fish and computed their position in the arena and their instantaneous speed @xmath5 calculated on three positions and computed as the distance between @xmath6 and @xmath7 divided by 2 time steps ..
first , we quantified for all the replicates the total number of collective residence events ( cre ) defined as the whole group resting in one of the two rooms .
the number of cre decreases with the size of the groups with a median number of 233 cre for 2 fish to 131 cre for groups of 10 fish ( fig .
[ fig : ndepartures]a ) .
moreover , we counted the total number of collective departures events ( cde ) defined as the whole group leaving one of the resting sites for the corridor towards the other one .
the number of cde also decreases with the size of the group but with a stronger difference between the groups ( from a median number of 212 for two zebrafish to 16 cde for groups of 10 zebrafish ( fig .
[ fig : ndepartures]b ) .
these observations show that 90% of the cre were followed by a collective departure in groups of 2 fish while only 12% were in groups 10 fish .
thus , larger groups were more likely to split into subgroups during departures while small groups of fish swam together most of the time .
thanks to the individual tracking of the fish , we were able to determine the identity of the first fish that left a room for all collective departures .
thus , for all groups , we computed the proportion of collective departure initiated by each fish and ranked the group members according to the proportion of departure that they led .
to characterise the distribution of the leadership among the group members , we ranked each individual according to the proportion of collective departures that they have initiated and compared the distributions of the ranks with two theoretical ones .
on the one hand , we simulated a situation where all fish have the same probability ( @xmath8 ) to initiate a departure
. on the other hand , we simulated a despotic configuration with a fish that has a 0.9 probability to initiate collective movement while the others have only a @xmath9 chance to start a departure .
the experimental data lay between these two extreme scenarios ( fig .
[ fig : rankedinitiation ] for 5 fish ) . in groups of five fish ,
the 1@xmath10 ranked fish initiated 45% of the collective departures on average .
this value is largely below the 90% observed in the despotic simulation but also higher than the 25% of the uniform repartition .
these results highlight an heterogeneous distribution of leadership among group members , some fish have a higher tendency to start departure than others .
we observed similar results for groups of 2 , 3 , 7 and 10 fish . ) while the despotic distribution assumed that one fish has a higher probability ( @xmath11 ) to initiate collective departure while other have a small probability to do so ( @xmath12 ) .
the results show that the frequency of initiation in groups of zebrafish differs from both simulated distribution highlighting a distributed leadership . ] to determine whether this distributed leadership was related to a different success rate or the number of initiation attempts , we measured the number of time that each fish was the first to exit a resting site independently of its success to be followed by the other group members ( i.e. an _ attempt _ ) .
the intra - group proportion of attempts made by each fish shows a continuum from an egalitarian to a more despotic distribution for all group sizes ( fig .
[ fig : prop_attempt ] ) . for each group , we compared the distribution of the total number of attempts made by each fish with a theoretical homogeneous distribution . for two fish ,
4 dyads out of 11 did not significantly differ from the homogeneous distribution ( @xmath13 test , see table for details ) . in groups of three fish
, the hypothesis of homogeneous distribution was not rejected in only one trio . finally , all groups of 5 , 7 and 10 fish significantly differ from the equal distribution of the number of attempts .
thus , like the distribution of initiations , the attempts are also heterogeneously distributed among the fish of the group .
then , we checked for a potential correlation between the number of initiations and the number of attempts made by each fish .
a linear regression shows that the number of initiation is linearly correlated to the number of attempts perform by the fish and that the coefficient of this correlation depends on the group size ( fig .
[ fig : attempts]a ) . for groups of two fish ,
92% of the attempts made by a fish resulted in a collective departure of the whole group .
moreover , the linear relation between the numbers of attempts and initiations highlights that this high success rate is shared by all fish .
therefore , the fish leading the largest number of departures were the more active fish .
in addition , the proportion of initiation led is proportional to the proportion of attempts performed by each fish ( fig .
[ fig : attempts]b ) .
these results highlight that all fish have the same success rate in starting a collective departure when they attempt it .
thus , the higher status of initiator of some fish is not related to a higher influence on other fish or a better success but on a higher tendency to exit the resting sites .
the results for other group sizes are qualitatively similar but differ quantitatively : while the linear relation persists , the success rate for each attempt decrease when the group size increases ( 90% for 3 fish , 66% for 5 fish , 53% for 7 fish and 26% for 10 fish ) .
next , we studied the temporal distribution of the leading event to highlight a potential succession of temporary leaders . to do so , we computed the probability of a fish to perform two successive initiations and compared this probability to the proportion of cde that the fish has led .
a temporal segregation of the initiators would results in a high probability of successive initiations compared to the proportion of led cde while homogeneously distributed initiations would give similar probabilities of successive initiations and led cde .
for all group sizes , we observed a linear correlation between the two proportions ( fig .
[ fig : samelead ] ) .
thus , the probability of a fish to initiate a cde was not dependent on its status of initiator or follower during the previous cde .
finally , we looked at a potential link between the motion characteristics of the individual and the number of attempts that they have made .
in particular , we measured the average linear speed of all individual as an indicator of their motility . there is a positive correlation between the average speed of the individual and the number of attempts that they performed .
however , this correlation is only significant for groups of 5 , 7 and 10 fish ( fig .
[ figure4]a - f ) .
we also compared the intra - group ranking of the fish for the number of initiation with their intra - group ranking for the linear speed .
we used the kendall s @xmath14 coefficient to measure the association between the two rankings .
the intra - group ranking for the initiation is positively correlated to the intro - group ranking for the linear speed ( fig .
[ figure4]c ) for groups of 3 , 5 , 7 and 10 fish .
thus , the fish with a highest average speed of its group is more likely to also be the fish that has started the largest number of departures , except for dyads .
these results show that the initiation of collective movements is related to the motility of the fish , particularly in large groups .
the initiation of collective movement in fish is often reported as a distributed process in which each fish can potentially lead a departure .
this distribution of the leadership is particularly suited for large schools of hundreds or thousands of fish that have to detect and avoid attacks from predator coming potentially from any direction .
thus , the first fish to spot a predator can start an escaping manoeuvre that will be propagated from neighbour to neighbour in the whole school . in these conditions
, it is difficult to imagine a permanent leader monopolising the initiation of collective movement .
however , the situation can be quite different in smaller groups of fish .
while they still have to avoid predators , the inter - individual differences can be more marked due to the small number of group members .
therefore , a heterogeneous distribution of the leadership due to individual characteristic is more likely to appear . here
, we showed that the initiation of collective departure is a distributed process among the group members in zebrafish .
however , the role of initiator is not homogeneously distributed , with some individuals leading more departures than others . by also measuring the number of attempt made by each fish
, we highlighted that this heterogeneous distribution was not the result of a higher success rate of some individual that could have a higher tendency to be followed .
on the contrary , the success of the fish to trigger a collective departure was linearly correlated to their number of attempt .
therefore , the best initiators do not seem to occupy a particular hierarchical status in the group but are the most motivated individuals to exit the room .
in addition , we showed that the initiation process was not temporarily organised with a fish leading the group during a particular time period before being relayed by another fish , but a distributed status during the whole experimental time .
finally , we showed that the motility of the fish is a predictor of the tendency of the fish to initiate collective departure .
indeed , the intra - group ranking of the fish for average speed of the fish was correlated to its intra - group ranking for departure attempts .
thus , these results provide evidence that the emergence of specific initiator of collective movement in zebrafish is more related to individual characteristics of the group members than to a particular status in the group .
all authors declare having no competing interests .
b.c . performed the data analysis , ran the simulations and wrote the manuscript .
a.s . and y.c .
performed the experiments and collected the experimental data .
l.c . developed the video treatment and tracking system .
designed , conceived , coordinated the study and revised the manuscript .
all authors gave final approval for publication .
this work was supported by european union information and communication technologies project assisibf , ( fp7-ict - fet n. 601074 ) .
the funders had no role in study design , data collection and analysis , decision to publish , or preparation of the manuscript . | for animals living in groups , one of the important questions is to understand what are the decision - making mechanisms that lead to choosing a motion direction or leaving an area while preserving group cohesion . here , we analyse the initiation of collective departure in zebrafish _
danio rerio_. in particular , we observed groups of 2 , 3 , 5 , 7 and 10 zebrafish swimming in a two resting sites arena and quantify the number of collective departure initiated by each fish .
while all fish initiated at least one departure , the probability to be the first one to exit a resting site is not homogeneously distributed with some individuals leading more departures than others .
we show that this number of initiation is linearly proportional to the number of attempts performed and that all fish have the same success rate to lead the group out of a resting sites after an attempt .
in addition , by measuring the average swimming speed of all fish , we highlight that the intra - group ranking of a fish for its proportion of initiation is correlated to its intra - group ranking in average speed .
these results highlight that the initiation of collective departure in zebrafish is a heterogeneously distributed process , even if all individual have the same success rate after attempting a departure . |
experimental areas in particle accelerator facilities are notoriously noisy places due to beamline elements like magnets , slits , pumps , power supplies , and fans as well as other assorted electronics equipment that either runs continuously or is intermittently turning on and off .
all of the above may cause voltage ripples that significantly compromise photomultiplier tube ( pmt ) current pulses digitized by analog - to - digital converters ( adcs ) .
the problem is compounded especially in analysis of data from segmented detectors where total energy and/or direction of each measured particle are derived from smeared adc values of several adjacent detector modules .
an experimenter might try to follow the recipes for reduction of pedestal noise couplings by designing recommended magnetic shielding and applying proper electrical grounding principles @xcite .
but such efforts are in practice usually met with only limited success . in order to minimize the electronic noise arising from `` dirty '' electrical grounds and from cross - talk between adjacent detector adc channels ,
quality coaxial cables are used for connecting pmt anode outputs with inputs of fast electronics units .
this arrangement though does not protect against stray magnetic fields at low frequencies that cause the so - called `` ground loops '' . while a low frequency noise component can be removed by using isolation transformers and capacitor - coupled inputs to the adcs , that approach is not an option in high - rate environments where ac - coupled devices would produce unacceptable signal distortions and rate - dependent baseline shifts .
custom electronic circuits developed to address the problem of low frequency voltage ripples are described in refs .
they are designed to provide a correspondence between a simple saw - tooth waveform and an experimental ac - power cycle , digitizing that information and passing it to the data stream for an offline noise correction .
another method that uses one or more `` blackened pmts '' ( bpmts ) in parallel with the detector active pmts is described and compared with active ac - noise synchronization circuits in ref .
the bpmts are dummy phototubes under high voltage but with no attached detector modules , whose signals are digitized in exactly the same way as those of the active detectors
. a drawback / disadvantage of the method is that for complicated experimental layouts operating in noisy environments and with multiple local grounds prevailing in the areas close to beamlines , more than a handful of the bpmts would be necessary to account properly for the correlated noise .
moreover , because bpmts should be mechanically and electrically a part of a detector and should be physically close the active pmts , they could be affected by , for example , cherenkov radiation caused by relativistic minimum ionizing particles in photocathode glass windows , destroying the noise correlations . in the analysis that follows
we take advantage of the fact that our apparatus , the pibeta detector at psi , consists of individual detectors that are optically isolated from each other . therefore
, one particular physics trigger will not excite more than a handful of active detectors and the role of a `` blackened pmt '' can be played by different active detector lines in different events .
we call our procedure `` the second pedestal noise correction '' .
the pibeta apparatus @xcite is a large acceptance 3@xmath0sr non - magnetic detector optimized for measurements of electrons and photons in the energy range of 10100@xmath1mev .
the heart of the detector is a spherical 240-module pure csi calorimeter that is supplemented with a segmented 9-piece active plastic target , a pair of conventional cylindrical mwpcs , and a cylindrical 20-counter plastic hodoscope for charged particle tracking and identification .
all parts of the detector are mounted at the center of a 2.5@[email protected]@xmath1 m steel platform that also carries high voltage power supplies , a detector cooling system , coiled coaxial delay cables on the one side , and fast trigger electronics on the opposite side of the platform .
therefore , the pibeta detector is a compact self - contained assembly that can be moved easily from its parking place into a chosen experimental area as a single unit and made operational within a day . due to the detector s proximity to elements of the beamline
, our fast electronics are exposed to a significant contaminating electronic hum .
[ fig : waveform ] shows a typical baseline of an analog signal that is to be digitized in an adc device , captured on a tektronix tds 744 digital oscilloscope .
a snapshot displays ground - loop noise with frequencies of @xmath350@xmath1hz and @xmath3300@xmath1hz in a 20@xmath4s interval .
typical peak - to - peak noise amplitude is @xmath55@xmath1mv .
pibeta pmt signal outputs are split either directly at the pmt voltage dividers or at passive custom - made analog signal splitters .
one branch of a calorimeter analog signal is delayed @xmath3380@xmath1ns in coaxial cables and then split again to provide inputs for fastbus adc units and discriminators / tdcs / scalers .
the other branch of the pmt signal is connected to analog summing and discriminator modules of the fast trigger electronics .
the master trigger outputs are delayed and subsequently used to provide adc gates and tdc start / stop signals .
metal frames of the pmt high voltage supplies on one end of the platform and the detector conductive support structure in the center are connected with 10@xmath1mm/20@xmath1 mm thick copper cables to fast trigger electronics grounding points in order to decrease the noise arising from the ground loops .
voltage differences between different parts of the detector are measured with a digital voltmeter and after the grounding connections are put in place are reduced to less than 4@xmath1mv . for digitizing fast pmt pulses
we have used a lecroy adc model 1882f unit , a high - density fastbus converter with 96 current - integrating negative inputs @xcite .
the 1885f unit provided a wide dynamic range with 12-bit data for the resolution of 0.05@xmath1pc / count and less than 1@xmath1% quantization error .
raw pedestals are set at about 200 - 300 channels and in our particular experimental environment had the average root - mean - square ( rms ) widths of @xmath315 - 20 channels .
[ fig : baseline ] presents an analog signal of a single csi calorimeter detector on 2@xmath1mv / division and 100@xmath1ns / division scale taken in 6@xmath1sec persistence mode .
clearly , with an adc gate width set to 100@xmath1ns , the adc unit will integrate variable amounts of the noise current .
a 70@xmath1mev electron or photon will produce a current pulse with an amplitude of @xmath30.25@xmath1v and a @xmath330@xmath1ns fwhm width .
this current pulse into 50@xmath6 will integrate to a charge of about 150@xmath1pc , placing it at 3/4 of the full scale . a common noise with an amplitude 5@xmath1mv present on the baseline will integrate in a 100@xmath1ns gate to a 10@xmath1pc charge , or almost 7% of an actual particle signal . as the energies and/or directions of measured particles are deduced from adcs of a centrally hit calorimeter module and on average
two nearest neighbors hit , the resulting uncertainty in energy ( direction ) could add up to 20% .
a 190@xmath220@xmath28@xmath1mm@xmath7 plastic scintillator counter ( bc-400 ) is placed horizontally above the electronics racks , about 3@xmath1 m away from the main pibeta detector . by virtue of its position
, the counter is shielded from the experimental radiation area by a 5@xmath1 cm thick lead brick wall as well as by a 50@xmath1 cm wide concrete wall .
operating with a high discriminator threshold it is counting only random cosmic muons at about 1 - 2@xmath1hz frequency , having a stable counting rate for both beam - on and beam - off periods .
the discriminated signals from that scintillator counter define our random trigger .
the main pibeta detector trigger is defined dynamically in a programmable logic unit . in practice ,
the master trigger is a logic or of up to 12 individual triggers .
high rate physics triggers in the mix are prescaled , with prescaling factors depending on the trigger type and the beam stopping rate .
the random trigger is connected to the trigger mix logic unit and in production runs it is always included among the enabled triggers .
a single production run is limited to 200,000 events , yielding typically @xmath310@xmath8 random triggers per run interspersed among the physics triggers .
we note parenthetically that in an off - line analysis adc values for all detectors in random events are written out to a text file .
these data sets are used in a geant @xcite detector simulation , enabling us to account rigorously for residual adc noise and accidental event pile - ups in the energy spectra .
in a pre - production calibration run data are first collected in so - called `` pedestal mode '' with the beam off and with only random trigger enabled in the trigger mix .
raw pedestals are automatically calculated at the end of the calibration run .
namely , the pedestal histograms are filled with the random trigger events @xmath9 expressed in adc channel numbers , where a subscript @xmath10 labels a detector and an index @xmath11 represents an event number . at the end of the run these histograms
are analyzed for the mean pedestal positions @xmath12 .
the mean pedestal positions and widths are saved in the data acquisition database residing completely in shared memory to allow for fast access .
`` calibrated adc values '' @xmath13 in the subsequent production runs would then be differences between raw adc readings @xmath9 and the pedestal values @xmath12 retrieved from the database : @xmath14 the analyzer calibration program also calculates the correlation coefficients @xmath15 among all possible detector pairs `` @xmath16 '' that belong to the same detector group , e.g. , for 240 calorimeter detectors , 40 plastic veto hodoscope counters , etc .
the expression used for calculating the correlation coefficient @xmath15 is @xcite : @xmath17 where sums extend over all pedestal events with calibrated adc values @xmath18 less than 20 channels and @xmath19 is the total number of such events .
the 20 channel limit corresponds to the half - width at tenth - maximum of the typical noise distribution and discriminated against ( very rare ) cosmic events in our random data set that would smear up our correlation matrix .
the correlation coefficient matrix is saved in a text file that is analyzed by an offline computer program to identify the detector noise groups .
the program user can select the maximum number of noise groups ( default 10 ) , minimum group size ( default 8) and the minimum correlation coefficient ( default 0.75 ) between the detectors in the same noise group .
the offline code combines detectors into noise groups on a trial - and - error basis with the over - all goal of maximizing the average intergroup correlation coefficients , while keeping the number of groups as small as possible and the minimum number of channels in any group above a given limit . with a minimum group size of eight detectors
the program generated 10 noise groups for the 240 csi channels with an average intergroup correlation coefficient 0.90 and minimum correlation coefficient greater than 0.75 .
the noise groups suggested by the program are subsequently hard - coded in the analyzer calibration program .
an example of six plastic hodoscope ( pv ) noise groups is given below in a section of c code : .... / * adc groups for common noise suppression * / # define max_noise_group 100 int adc_group[][max_noise_group ] = { / * plastic veto detectors * / { 240,241,242,243,244,245,246,247,248 , 249,250,251,252,253,254,255 , -1 } , { 256,257,258,259 , -1 } , { 260,261,262,263,264 , -1 } , { 265,266,267,268,269,270,271 , -1 } , { 272,273,274,275 , -1 } , { 276,277,278,279 , -1 } , { -1 } , } ; .... where individual pv counters are identified by their signal indices .
the value `` @xmath20 '' at the end of each line follows the last detector in every noise group .
each detector is included in one and only one noise group .
detectors belonging to the same noise group are usually physically close and have their signal and high voltage cables bundled together .
the algorithm for calculating of noise - corrected and calibrated adc values resides in the analyzer program with predefined noise groups .
the algorithm loops over all adc channels and for every event : 1 . subtracts the raw pedestal values retrieved from the online database from raw adc variables ( `` first pedestal correction '' ) ; 2 . identifies the common noise group that the current detector belongs to ; 3 . finds the channel with the smallest adc value in that group
; 4 . finds all the channels in the group under consideration which are within 15 counts above the minimum of the group and calculates their average adc value ; 5 .
subtracts that average value ( `` second pedestal correction '' ) from the adc values of all other group members . at the beginning of each
run the pedestal histograms are reset and at the end of the run they are again analyzed by fitting lineshapes with a gaussian function .
the means of the distributions are updated in the database as new pedestal values @xmath12 .
[ fig : correlation ] shows a scatter plot of adc values for two different csi detectors acquired with a random trigger .
uncorrected adc values are plotted in the top panel showing the correlation in an electronic noise : these two detectors are eventually assigned to the same electronics noise group . using the secondary pedestal correction procedure described above
, we obtained a scatter - plot of the corrected adc values presented in the lower panel of fig .
[ fig : correlation ] .
examples of one - dimensional adc histograms for two different detector types , namely a plastic scintillator counter and one active target segment are shown in figs .
[ fig : improve ] and [ fig : target ] , respectively .
top panels represent raw adc spectra , bottom panels show the noise suppression in the corrected spectra . in fig .
[ fig : improve ] a peak at channel number 6 is inserted by the fitting function in the analyzer program .
its position corresponds to 3 standard deviations of the corrected pedestal lineshape . in readout of the adc modules all channels with the values below 3@xmath21
are suppressed by being considered equal to zero and are not written to tape .
this zero suppression criterion compresses the size of our data files by an order of magnitude .
a comparison between the smearing of raw pedestal data and the spectra obtained after the noise suppression for four different detector types is given in table [ tab1 ] .
the table summarizes the maximum and minimum noise group sizes , rms widths of raw and corrected pedestal histograms as well as an important equivalent energy deposition corresponding to the corrected pedestal rms values .
we see that the energy equivalents of the rms widths range from 0.13@xmath1mev in csi detectors to 0.40@xmath1mev in the segmented target .
the target pedestals are more difficult to correct because of ( a ) the small number of target segments ( 9 ) that form a noise group , and ( b ) the high rate of physics triggers in the target , with up to 1@xmath1mhz stopping pions .
we have implemented correction of modular detector adc readings for the correlated common noise using the method of common noise detector groups . the reduction in the pedestal width of up to a factor of five
is achieved with the corrected pedestal rms values as narrow as 1.5 adc channels .
this improvement compares favorably with the active circuit noise suppression methods @xcite where improvement factors are @xmath35 - 6 . in a representative case of pure csi calorimeter
the procedure reduces the correlated noise contribution to the equivalent energy of 0.13@xmath1mev per detector module . for the energy range of interest , 10100@xmath1mev ,
this value is smaller than the photoelectron statistics contribution of @xmath22mev @xcite and is comparable with the pmt dark current noise that amounts to @xmath30.1@xmath1mev .
this work is supported and made possible by grants from the us national science foundation and the paul scherrer institute .
d. poani , d. day , e. frle , r. m. marshall , j. s. mccarthy , r. c. minehart , k. o. h. ziock , m. daum , r. frosch and d. renker , psi r-89.01 experiment proposal , ( paul scherrer institute , villigen , 1988 ) .
.comparison of the pedestal noise reduction for the various elements of the pibeta detector .
the values represent averages in one typical production run with @xmath35000 pedestal events .
the minimum and maximum number of detectors in one noise group is given in the second column .
the energy ( angle ) equivalent of the corrected pedestal rms is listed in the fifth column . [ cols="<,^,^,^,^ " , ] | we describe a simple procedure for reducing adc common noise in modular detectors that does not require additional hardware . a method using detector noise groups should work well for modular particle detectors such as segmented electromagnetic calorimeters , plastic scintillator hodoscopes , cathode strip wire chambers , segmented active targets , and the like .
we demonstrate a `` second pedestal noise correction '' method by comparing representative adc pedestal spectra for various elements of the pibeta detector before and after the applied correction .
pacs numbers : 07.50.h , 07.05.hd , 07.05.kf _ keywords : _ adc pedestals , correlated common noise , daq zero suppression psfig.sty , , |
present - day cluster ellipticals are a remarkably well - behaved class of objects , with structural and chemical properties obeying simple power - law scaling relations . but
this could not always have been the case in a hierarchical universe . while most galaxy formation models can be tuned to reproduce these relations at @xmath10 , a more stringent test lies in reproducing their evolution with redshift .
to this end , it is important to study rich clusters out to the highest redshifts , when fractional age differences among the galaxies were proportionately greater . in recent years
, deep wide - field optical surveys and deep serendipitous x - ray surveys have uncovered significant numbers of rich galaxy clusters to redshift unity and beyond ( see reviews by postman 2002 ; rosati 2003 ) .
these most distant , and most massive , of known gravitationally bound structures can then be studied in detail through targeted , high - resolution , follow - up optical and near - infrared observations .
we have undertaken a survey of rich galaxy clusters in the redshift range @xmath11 using the advanced camera for surveys ( acs ; ford et al .
2002 ) on the _ hubble space telescope _
( _ hst _ ) .
the aim of this survey is to establish new constraints on the cluster formation epoch and the evolution of early - type galaxies .
the first cluster observed , rdcs1252.92927 ( hereafter rdcs1252 ) at @xmath12 ( rosati 2003 ; rosati et al .
2003 ) , was discovered as part of the rosat deep cluster survey ( rosati et al .
1998 ) and is among the highest - redshift galaxy clusters with spectroscopic confirmation .
this letter presents the first results from our acs cluster survey , focusing on the color - magnitude ( cm ) relation of the early - type galaxies in rdcs1252 .
we adopt the best - fit wmap cosmology : @xmath13 ( bennett et al .
2003 ) , giving a scale of 8.4 kpc per arcsec at @xmath12 .
rdcs1252 was observed in the f775w and f850lp bandpasses ( hereafter i_775@xmath14 and z_850@xmath15 , respectively ) with the acs wide field camera as part of the guaranteed time observation program ( proposal 9290 ) during 2002 may and 2002 june .
the observations were done in a @xmath16 mosaic pattern , with 3 and 5 orbits of integration in i_775@xmath14 and z_850@xmath15 , respectively , at each of the four pointings .
there was nearly 1 of overlap between pointings ; thus , the core of cluster was imaged for a total of 12 orbits in i_775@xmath14and 20 orbits in z_850@xmath15 .
the data were processed with the `` apsis '' pipeline described by blakeslee et al .
( 2003 ) , with some recent updates .
in particular , we used a version of the drizzle software ( fruchter & hook 2002 ) supplied by r.hook that implements the `` lanczos3 '' interpolation kernel ( a damped sinc function ) .
this kernel produces a sharper point spread function ( psf ) and greatly reduces the noise correlation of adjacent pixels and the resulting `` moire '' patterns .
apsis also now removes discontinuities in the residual bias level at the amplifier boundaries , producing a more uniform background .
an earlier processing of these images has been used by bouwens et al .
( 2003 ) for a study of the faint i_775@xmath14 dropout population at @xmath17 .
we calibrate our photometry to the ab system using photometric zero points of 25.640 ( i_775@xmath14 ) and 24.843 ( z_850@xmath15 ) .
these are uncertain at the @xmath40.02 mag level , which has no effect on our conclusions .
we adopt a galactic reddening for this field of @xmath18mag based on the schlegel et al .
( 1998 ) dust maps .
figure[fig : cluscenter ] shows the central @xmath19 region of a color composite made from our reduced i_775@xmath14 and z_850@xmath15 images .
a red galaxy population is clearly visible .
the central pair of galaxies are separated by 18 ( 15kpc ) and are each of magnitude @xmath20 .
we used sextractor ( bertin & arnouts 1996 ) in `` dual - image mode '' with low threshold and deblending settings to find objects in the reduced images and perform the initial photometry .
sextractor `` mag_auto '' values were used for the total magnitudes .
the ( i_775-z_850)@xmath1 color effectively separates out evolved galaxies at @xmath21 , and the cluster is obvious as a central concentration of galaxies with @xmath22 .
figure1b [ removed from the apjl version in order to meet the page limit ] shows histograms of isophotal color within 3 different radii of the cluster center ( defined midway between the two central galaxies ) for galaxies with total @xmath2325 mag .
we selected an initial sample of 312 nonstellar objects with @xmath24 , in the broad isophotal color range @xmath25 , and inside a radius of 192 .
our goal is to study the early - type galaxy population in rdcs1252 , for which we have limited spectroscopic data , and these cuts are designed to select the vast majority of our target sample while reasonably limiting foreground / background contamination .
the color selection is roughly 7 times broader than the full - width of the red sequence we find below .
the radial cutoff corresponds to about 1.0 mpc for both our adopted wmap cosmology and an einstein - desitter cosmology with @xmath26 .
our final colors are measured within galaxy effective radii @xmath27 to avoid biasing the cm slope due to color gradients .
we follow the basic approach outlined by van dokkum et al.(1998 , 2000 ) .
we derive the @xmath27 values using the program `` galfit '' ( peng et al .
2002 ) by fitting each galaxy to a sersic model ( convolved with the psf ) , but constraining the @xmath28 parameter such that @xmath29 .
bright neighboring galaxies were fitted simultaneously .
we note that subtraction of the model for the two central galaxies reveals evidence for interaction in the form of an -shaped residual .
next , we deconvolve i_775@xmath14 and z_850@xmath15 postage stamp images of each galaxy using the clean algorithm ( hgborn 1974 ) in order to remove the differential blurring effects of the psf , which is @xmath410% broader in the z_850@xmath15 band . to reduce noise ,
the clean maps are smoothed with a gaussian of @xmath30 pix before adding the residual images to the maps to ensure flux conservation .
we then measure the i_775@xmath14 and z_850@xmath15 magnitudes of each galaxy within a circular aperture of radius @xmath27 , typically about 6 pix , or 03 .
we did not allow the radius to drop below 3 pix .
photometric errors were determined empirically from the pointing overlap regions .
we reprocessed the 4 pointings separately and measured the color differences within @xmath27 for 202 pairs of measurements for 74 different early - type galaxies ( classifications described below ) in the overlap regions .
we then median - filtered to obtain the errors as a smooth function of magnitude .
the error thus determined for single - pointing ( i_775-z_850)@xmath1 measurements at @xmath31 was 0.025mag , rising to @xmath40.05mag at @xmath32 .
each galaxy in our initial sample was examined and morphologically classified , following a procedure similar to that of fabricant et al .
this was done independently of the profile fitting , but the types show a good correlation with sersic @xmath28 index . here , we simply distinguish between e , s0 , and later types , where the intermediate s0 class indicates the apparent presence of a disk without spiral or other structure .
full details on the classifications for a much larger sample , including the spatial distribution of the various types , will be presented by postman et al .
( 2003 , in preparation ) .
about 180 galaxies in this field have measured redshifts , obtained with vlt / fors ( rosati et al .
2003 , in preparation ) , with 31 being cluster members .
all galaxies in our initial sample classified as early - type , and not known to be interlopers from their spectra , are included in our cm analysis in the following section .
of the 31 known members , 22 are classified as early - type , and we include all of these in our analysis even though one ( an s0 ) happened to lie beyond our 1 mpc cutoff .
we fitted the early - type galaxy cm relations using simple linear least squares ; other methods gave very similar results .
we estimate the scatter from both the standard rms and the biweight scale estimator ( beers et al .
no rejection was done in fitting subsamples composed of known members , the faintest of which has @xmath33 , or @xmath34 .
we also performed fits to samples with unconfirmed members , allowing us to go 1mag further down the galaxy luminosity function .
however , here we iterate to reject the 3-@xmath35 outliers , as these are likely to be interlopers : none of the 22 confirmed early - type members is more than 2.3-@xmath35 discordant .
after the iterative rejection process , we find concordant scatters for those samples , and the rms and biweight estimator are the same to within @xmath360.001mag .
figure [ fig : finecmr ] presents the cm relation for the rdcs1252 galaxies .
the fit to the full sample of early - type galaxies with @xmath37 gives @xmath38 other results are listed in table [ tab : cmr ] .
the mean locations of the cm relations for the elliptical and s0 subsamples agree to well within the errors , while the slopes are consistent at the 1.5-@xmath35 level .
eight known late - type members from rosati et al .
( 2003 , in preparation ) for which we have photometry are bluer than the early - type galaxies by 0.25mag , with a scatter of 0.14 mag about this offset , indicating young stellar populations .
we find an intrinsic scatter @xmath39 mag for the 15 confirmed elliptical members . for a limit of @xmath40 , we derive @xmath41 mag for the clipped sample of 31 ellipticals and @xmath42 for the 52 e+s0 galaxies . at this limit ,
the observational errors become dominant and classification is difficult , which could bias our @xmath43 estimates . for @xmath44 , still 3mag down the luminosity function to about 0.3@xmath45 , we find @xmath46 for 25 ellipticals ( with no outliers ) .
we estimate @xmath47 ab for rdcs1252 , based on local surveys ( norberg et al .
2002 ) , wmap cosmology , and @xmath481.4 mag luminosity evolution ( postman et al . 2001
; van dokkum & stanford 2003 ) .
the correction from z_850@xmath15 to rest - frame @xmath49 is @xmath500.3 mag for an early - type spectrum .
thus the brightest cluster member at @xmath51 corresponds to @xmath52 .
the cluster red - sequence is about 0.1mag bluer than predicted for non - evolving elliptical templates ( coleman et al .
1980 ; schneider et al .
1983 ) . for comparison , a bruzual & charlot ( 2003 ;
hereafter bc03 ) solar metallicity model reddens by 0.11 mag and fades by 1.5mag in aging from @xmath43 to @xmath411.5 gyr .
the present - day relation for the coma cluster ( transformed according to the relations given below ) is also shown in figure [ fig : finecmr ] .
the tight cm relation allows us to constrain the scatter in early - type galaxy ages , subject to model uncertainties . the evolution in ( i_775-z_850)@xmath1 at this redshift is complicated at young ages because z_850@xmath15 straddles the 4000 break and has its blue end near the balmer jump at 3700 .
the balmer jump reaches a maximum at 0.51 gyr , when the relative contribution from a stars is greatest , resulting in a red color at these ages .
it is interesting that several of the s0s lie _ above _ the cm relation , possibly indicating ages @xmath53 gyr .
as the a star contribution lessens , the color quickly evolves towards the blue , reaching a local minimum near @xmath41.5 gyr , before commencing a roughly monotonic reddening with age .
we have simulated the ( i_775-z_850)@xmath1 colors of galaxies formed under two simple star formation models , similar to those used by van dokkum et al .
( 1998 ) . in model 1 , the galaxies form in single bursts randomly distributed over the interval @xmath54 , where @xmath55 is set to the recombination epoch and @xmath56 gyr , the age of the universe at @xmath57 . in model 2 ,
galaxies form stars at constant rates between randomly selected times @xmath58 , where @xmath59 .
we vary @xmath60 and at each step calculate colors and luminosities for 10,000 `` galaxies '' by interpolation and integration of the bc03 solar metallicity models . assuming @xmath61 mag for the ellipticals , model 1 imples a minimum age @xmath62 gyr , i.e.
, all galaxies finish forming at redshifts @xmath63 ; the mean luminosity - weighted age is @xmath64 gyr ( corresponding to @xmath65 ) with a scatter of 30% .
model 2 gives @xmath66 gyr , @xmath67 , and @xmath68 gyr ( corresponding to @xmath69 ) with 38% scatter .
thus , although some galaxies in model 2 have formed stars recently , the mean ages are still high .
both models give a mean color @xmath70 , similar to that observed .
for the s0s , we find @xmath71 for model 1 and @xmath72 , indicating recent star formation , and age scatters of 4447% .
finally , we note that the 1996 version of the bc models would have predicted higher formation epochs , e.g. , @xmath73 and @xmath74 for the ellipticals in models 1 and 2 , respectively , and an age scatter of only 20% , although the predicted color is then redder by 0.1 mag .
overall , we conclude that the ellipticals are an evolved population , with a mean age @xmath5 , a minimum age @xmath75 gyr , and an age scatter of ( @xmath76)% , where the error reflects uncertainty in @xmath43 and scatter in the models .
to enable comparison with previous work , we convert observed ( i_775-z_850)@xmath1 quantities to rest - frame ( u - b)_z@xmath77 at @xmath78 .
the models ( bc03 ; kodama et al .
1998 ) and empirical templates indicate @xmath79(i_775-z_850)@xmath1 , where the error bar reflects the scatter in the models at the relevant ages and adds about 20% uncertainty to our transformed slope and scatter .
figure [ fig : scatterz ] uses this conversion , and other transformations from van dokkum et al .
( 2000 ) , to compare our results to some previous studies of the cm relation in intermediate - redshift clusters with _
hst _ , as well as the results of van dokkum et al .
2001 on rxj0848 + 4453 , the only cluster of comparable redshift to have its cm relation measured .
linear fits to the data shown in figure [ fig : scatterz ] yield slopes of @xmath80 and @xmath81 for the evolution in the absolute slope and the scatter , respectively .
thus , the scatter is constant , and there is at best marginal evidence for slope evolution , which indicates that the slope is due to a variation in metallicity , not age .
previous studies of cluster samples out to @xmath82 have come to similar conclusions ( e.g. , stanford et al .
1998 ; kodama et al .
1998 ; van dokkum et al .
van dokkum et al .
( 2001 ) concluded that the slope at @xmath83 was shallower than in the coma cluster .
however , as shown in the figure , our slope measurement is consistent within the errors with both coma and rxj0848 + 4453 .
further studies of a diversity of clusters at similar redshifts are needed to explore this issue .
the lack of evolution in the cm relation scatter can be explained by progenitor bias ( e.g. , van dokkum & franx 2001 ) : galaxies selected as early - type at any epoch will have old stellar populations , while the later - type progenitors of the youngest ellipticals today will not be selected .
the result is an underestimate in the color scatter for the progenitors of modern ellipticals , and thus overestimated ages .
an upper limit on the scatter for elliptical progenitors may be estimated from a fit to all confirmed rdcs1252 members ; the result is 34 times larger than for the early - type galaxies . a detailed study of the morphological fractions in rdcs1252(postman et al . 2003 , in preparation )
should help illuminate the magnitude of this bias .
we also note that the two central ellipticals themselves , based on their proximity and irregular isophotes , appear likely to undergo dissipationless merger to form a single @xmath84 galaxy , similar to local cd galaxies .
we conclude that massive , evolved early - type galaxies were already present in rich clusters at @xmath57 .
our simple models imply mean luminosity - weighted ages of 2.63.3 gyr , corresponding to formation at @xmath853.6 .
however , the ( i_775-z_850)@xmath1 color is not ideal for the redshift of rdcs1252 , being better suited to measuring the 4000 break at @xmath86 . combining our acs data with deep , high - resolution near - ir imaging of this field ( lidman et al . 2003 ) will enable a more robust assessment of early - type galaxy ages .
in addition , further studies of the cm relations and morphologies of galaxies in other @xmath21 clusters are needed to improve the constraints on the formation epoch of cluster galaxies and on the evolution of their stellar populations and structural properties .
acs was developed under nasa contract nas 5 - 32864 , and this research has been supported by nasa grant nag5 - 7697 .
the stsci is operated by aura inc .
, under nasa contract nas5 - 26555 .
we thank our fellow acs team members for their help , taddy kodama for helpful discussions and models , and stephane charlot for the bc03 models .
lcccccc e & & 15 15 & @xmath87 & 0.029 & @xmath88 & @xmath89 + e+s0 & & 22 22 & @xmath90 & 0.038 & @xmath91 & @xmath92 + e & 24.0 & 25 25 & @xmath93 & 0.033 & @xmath94 & @xmath95 + e+s0 & 24.0 & 44 41 & @xmath96 & 0.045 & @xmath97 & @xmath98 + e & 24.5 & 34 31 & @xmath99 & 0.053 & @xmath100 & @xmath101 + e+s0 & 24.5 & 58 52 & @xmath102 & 0.054 & @xmath103 & @xmath98 + s0 & 24.5 & 24 21 & @xmath104 & 0.058 & @xmath105 & @xmath106 + [ tab : cmr ] | we investigate the color - magnitude ( cm ) relation of galaxies in the distant x - ray selected cluster rdcs1252.92927 at @xmath0 using images obtained with the advanced camera for surveys ( acs ) on the _ hubble space telescope _ in the f775w and f850lp bandpasses .
we select galaxies based on morphological classifications extending about 3.5 mag down the galaxy luminosity function , augmented by spectroscopic membership information . at the core of the cluster
is an extensive early - type galaxy population surrounding a central pair of galaxies that show signs of dynamical interaction .
the early - type population defines a tight sequence in the cm diagram , with an intrinsic scatter in observed ( i_775-z_850)@xmath1 of @xmath2 mag based on 52 galaxies , or @xmath3 mag for @xmath430 ellipticals .
simulations using the latest stellar population models indicate an age scatter for the ellipticals of about 34% , with a mean age @xmath5 gyr ( corresponding to @xmath6 ) , and the last star formation occurring at @xmath7 transforming to rest - frame @xmath8 , we conclude that the slope and scatter in the cm relation for morphologically selected early - type galaxies show little or no evidence for evolution out to @xmath9 .
thus , elliptical galaxies were already well established in x - ray luminous clusters when the universe was a third of its present age . |
_ human migration _ , i.e. the movement of large numbers of people out of , or into specific geographical areas ,
is seen as one of the biggest challenges that face the human societies in the 21st century .
on one hand , part of the human population has reasons to _ emigrate _ into countries which provide a `` better '' life on the other hand , industrialized countries can not sustain their current situation without the _ immigration _ of people .
the real problem arises because the `` demand '' and the `` supply '' side can not be matched .
industrialized countries fear that immigrants do not contribute to their further economic growth but , on the contrary , deplete their wealth by taking advantage of a social security , health , and educational system which they did not contribute to .
if we move this problem on the more abstract level of a game - theoretical model , we can distinguish between two types of agents : those _ cooperating _ , i.e. being able to integrate in a society and to contribute to a common good , namely economic growth , and those _ defecting _
, i.e. without the ability to socially integrate and thus depleting a common good at the cost of the cooperating agents . certainly , based on their past experience , agents can adapt , i.e. they can change their strategy from defection to cooperation and vice versa dependent on the payoff they receive in a given environment .
the question for an industrialized country would be then to define an optimal immigration rate that ( a ) does not destroy the common good , and ( b ) allows agents to adapt to the assumed cooperative environment within one or two generations , even if they may have not immigrated with a cooperative strategy .
the problems of cooperation and defection and the payoff - dependent adoption of strategies have been discussed in the framework of the prisoner s dilemma ( pd ) and the iterated pd ( ipd ) game ( see sect . [ 2 ] ) . with our paper , we add to this framework the ability to migrate between different countries ( `` islands '' ) .
our aim is to reveal optimal conditions for the migration of agents such that cooperating strategies can take over even on those islands where they were initially not present .
we note that migration was previously studied in a game - theoretical context by different authors @xcite .
our work differs from these attemts in various respects .
first of all , we do not assume that migration is based on the anticipated success @xcite this shifts the conceptual problem of the `` outbreak of cooperation '' to proposing rules with non - local information such that two cooperators meet at the same place , from which cooperating clusters can grow .
we also do not make migration dependent on local variables such as the number of defectors in the neighborhood @xcite or random choices of `` empty '' places @xcite .
in fact , human migration is rarely targeted at less densely crowded places , it is rather the opposite .
further , we do not assume one - shot games such as the pd , but instead consider the ipd in which the number @xmath0 of repeated interaction as well as the mix of up to 8 different strategies plays a crucial role .
eventually , we do not use an agent - based model in which update and migration rules are freely defined , to study their impact on computer simulations on a lattice .
our approach proposes a population based model in which subpopulations are defined with respect to ( a ) their interaction strategy , and ( b ) their spatial location .
the consideration of separated `` islands '' allows a coarse - grained implementation of spatial structures which is in between a lattice or network approach and a meanfield description .
it is known that spatial structures have an impact on the outbreak of cooperation @xcite , but their influence varies with other degrees of freedom , such as update rules , synchronization , interaction topology , payoff matrix .
therefore , in this paper we adopt mostly standard assumptions about the interaction type ( ipd with @xmath0 encounters ) and interaction topology ( panmictic subpopulations ) , strategy adoption ( replication proportional to payoff ) , and migration ( fixed fraction of the population ) . to understand the basic dynamics , we first investigate `` isolated '' islands ( no migration ) to find out about the conditions for the `` outbreak of cooperation '' without external influence .
this `` outbreak '' is defined as the critical point ( strategy mix , number of encounters @xmath0 ) beyond which a whole island is going to be occupied by cooperating agents , if agents adopt strategies proportional to their average payoff .
then , we add migration between islands to this dynamics to find out under which conditions the outbreak of cooperation can be enhanced .
it is important to note that migration does not distinguish between strategies or islands , i.e. there are no better suited strategies for immigration , or bad places with high emigration rates which we consider as artificial assumptions to explain the `` outbreak of cooperation '' . to determine the robustness of our findings
, we always consider worst case scenarios , i.e. initial settings in which most islands have either an entirely defective subpopulation , or at least a defective majority .
we further control for important parameters such as the pool of available strategies @xmath2 , the number of interactions @xmath0 or the number of islands @xmath1 , for which critical conditions are derived .
our finding that migration is indeed able to boost the outbreak of cooperation is remarkable both because it is based on minimal assumptions about the dynamics and because it follows from a quite systematic investigation of the underlying conditions .
let us investigate a population of @xmath3 agents divided into subpopulations on @xmath1 different islands which imply a coarse - grained spatial structure , i.e. @xmath4 . agents at island @xmath5 are assumed to interact with the @xmath6 other agents on their island ( _ panmictic population _ ) .
but , in general we assume that migration between the islands is possible , with the respective migration rates given by @xmath7 .
these define the fraction of the agent population at island @xmath8 that migrates to island @xmath9 in a given time interval .
figure [ fig : lattice ] shows the case of 3 different locations .
( with different strategies ) are spatially distributed on @xmath5 islands from where they can migrate at rates @xmath7.,width=226 ] our model basically considers two different time scales for interaction and migration . we define a _ generation _ @xmath10 to be the time in which each agent has interacted with all other @xmath6 agents a given number of times , denotes as @xmath0 .
thus the total number of interactions during each generation in the panmictic population is roughly @xmath11 . at
any given time , agent @xmath12 can interact only with one other agent in a `` two - person game '' .
but during one generation @xmath10 , it interacts with all other @xmath6 agents @xmath0 times ( repeated two - person game ) . for the interaction ,
we adopt the well known _
iterative prisoner s dilemma _ ( ipd ) from game theory . at each encounter between agents @xmath12 and @xmath13 , both have two actions to choose from , either to _ collaborate _ ( @xmath14 ) or to _ defect _ ( @xmath15 ) , _ without knowing _ the action chosen by their counterpart .
the outcome of that interaction is described in the following payoff matrix : @xmath16{ccc } & \textbf{c } & \textbf{d}\\ \textbf{c } & \;(r , r)\ ; & \;(s , t)\ ; \\
\textbf{d } & \;(t , s)\ ; & \;(p , p)\ ; \end{array}\ ] ] if both agents have chosen to collaborate , then they both receive the payoff @xmath17 .
if one of the agents chose to defect and the other one to cooperate , then the defecting agent receives the payoff @xmath18 and the cooperating the payoff @xmath19 .
if both defect , they both receive the payoff @xmath20 . in the special class of _ prisoner s dilemma _ ( pd ) games , the payoffs have to fullfil the following two inequations : @xmath21 which are met by the standard values @xmath22 , @xmath23 , @xmath24 , @xmath25 .
this means in a cooperating environment , a defector will get the highest payoff . for @xmath26 ( `` one - shot '' game ) choosing action
@xmath15 is unbeatable , because it rewards the higher payoff for agent @xmath12 no matter if the opponent chooses @xmath14 or @xmath15 . at the same time ,
the payoff for _ both agents @xmath12 and @xmath13 _ is maximized when both cooperate .
a simple analysis shows that defection is a so - called _ evolutionary stable strategy _ ( ess ) in a one - shot pd .
if the number of cooperators and defectors in the population is given by @xmath27 and @xmath28 respectively , then the expected average payoff for cooperators will be @xmath29 .
similarly the expected average payoff for defectors will be @xmath30 .
since @xmath31 and @xmath32 , @xmath33 is always larger than @xmath34 for a given number @xmath27 , and pure defection would be optimal in a one - shot game .
even one defector is sufficient to invade the complete population of @xmath35 cooperators .
but in a _ consecutive _ game with short _ memory _ , both agents , by simply choosing @xmath15 , would end up earning less than they would earn by cooperating .
thus , the number of games @xmath0 two agents play together becomes important .
this makes sense only if the agents can remember the previous choices of their opponents , i.e. if they have a memory of @xmath36 steps .
then , for the _ iterated prisoner s dilemma _ ( ipd ) , they are able to develop different _ strategies _ based on their past experiences with their opponents .
we note that the ipd game was studied both in the context of a panmictic population ( ) and assuming a spatial population structure @xcite . usually , in an ipd only a _ one - step _ memory is taken into accout . based on the known previous choice of its opponent
, either @xmath14 or @xmath15 , agent @xmath12 has then the choice between _ eight _ different strategies . following a notation introduced in @xcite ,
these strategies are coded in a 3-bit binary string @xmath37 $ ] which always refers to _ cooperation_. the first bit represents the _ initial _ choice of agent @xmath12 : it is @xmath38 if agent @xmath12 has _ cooperated _ , and @xmath39 if it has defected initially .
the two other values refer always to the previous choice of agent @xmath13 .
@xmath40 is set to @xmath38 if agent @xmath12 chooses to cooperate given that agent @xmath13 has cooperated before and @xmath39 otherwise .
@xmath41 is similarily set to @xmath38 if agent @xmath12 chooses to cooperate given that agent @xmath13 has defected before and @xmath39 otherwise .
both @xmath40 and @xmath41 can be also interpreted as _ probabilities _ to choose the respective action given the knowledge of the previous choice of the opponent , i.e. @xmath42 in the stochastic case . in the _ deterministic _ case considered in this paper ,
@xmath41 and @xmath40 are either 0 or 1 . out of the eight possible strategies , to keep our results in a tractable manner
, we consider the following four strategies @xmath43 ( @xmath44 ) : @xmath45{rrrcl } z^{(1 ) } : & & [ 1|10 ] & & \mathrm{tft }
\\ z^{(2 ) } : & \quad & [ 0|00 ] & \quad & \mathrm{all - d } \\ z^{(3 ) } : & & [ 1|11 ] & & \mathrm{all - c } \\
z^{(4 ) } : & & [ 0|01 ] & & \mathrm{a - tft } \end{array}\ ] ] these four strategies were also considered originaly in the famous paper by axelrod & hamilton @xcite .
hence , we will extend these investigations by later adding migration to the game .
we note that all eight strategies have been discussed in @xcite . in this paper
we have dropped those variants that differ only in the first move but then continue as described in the following .
the strategies all - d and all - c are obvious : for @xmath46 agent @xmath12 always chooses to defect regardless of the previous choices of agent @xmath13 , @xmath47 follows likewise .
these two strategies represent the limit case of _ no memory _ , i.e. agent @xmath12 is simply `` fixed '' as a cooperator or a defector .
the most interesting strategy @xmath48 , known as `` tit for tat '' ( tft ) , means that agent @xmath12 initially cooperates and continues to do so , given that agent @xmath13 was also cooperative in the previous move .
however if agent @xmath13 was defective in the previous move , agent @xmath12 chooses to be defective , too .
this strategy was shown to be the most successful one in iterated prisoners dilemmas @xcite .
if initially two cooperative agents meet , they stay cooperative ( forever , in a deterministic game ) , this way maximizing their average payoff .
this encourages other agents to adopt tft .
hence , the `` outbreak of cooperation '' in a community of @xmath3 agents occurs as the amplification of a cooperative initial fluctuation .
the fourth possible strategy @xmath49 is just the negation of the most successful one , therefore named anti - tft .
it means that agent @xmath12 starts as a defector and continues to be so , as long as it meets with cooperative agents @xmath13 , this way receiving the highest possible payoff @xmath18 .
but , if it meets an agent that has previously defected , it changes its `` behavior '' to cooperation .
this makes some sense , since the loss in payoff is not that much
@xmath20 would be just slightly higher than @xmath19 . on the other hand , if the colloborative agent meets again with another cooperative agent , the payoff would be @xmath17 , which is a much higher gain .
so , this strategy , while not making sense in the first place , may end up with a higher payoff in those cases where defecting and cooperating strategies are equally present .
we will later show that a - tft in the long run benefits agents playing all - d and therefore will create a more difficult environment for the invasion of cooperation .
i.e. considering all four strategies captures a worst - case scenario for the outbreak of cooperation which we deem interesting to study ( instead of giving rational arguments in favor of a - tft ) . in this paper , we are interested in how the frequencies of the different strategies would evolve in the agent community .
the fraction of agents choosing strategy @xmath43 on island @xmath5 is defined as @xmath50 , where the different strategies are given by eq .
( [ strat ] ) .
similarly , the _ total fraction _ of agents playing strategy @xmath43 in the _ whole _ system , is given as @xmath51 .
for the dynamics we may assume that during each generation @xmath10 every agent behaves according to a fixed strategy picked up from the pool of possible strategies .
this strategy can be changed only after one generation is completed .
the `` update '' rule for the strategy is simply given by the success of different strategies during the past generation . as the evaluation criteria , we choose the average payoff @xmath52 , each strategy has received during the last generation on island @xmath5 .
this will be compared with the total average payoff @xmath53 that gives an estimate of the overall dynamics on that island : @xmath54 for the evolution of the frequencies of the different strategies , the following dynamics is postulated : @xmath55 it means that in the next generation @xmath56 the share of agents choosing a particular strategy @xmath43 has grown / shrunk i.e. , strategy @xmath43 has replicated according to the _ relative _ performance of this strategy on island @xmath5 during the previous generation @xmath10 . in population dynamics , this is known as _ fitness - proportional selection _
@xcite , we adopt it here since the relative performance of a certain strategy can be also interpreted as its _
fitness_. the dynamics of eq .
( [ a - dyn ] ) does not take into account that agents can migrate between the @xmath1 islands . for simplicity , we assume that each island @xmath5 is occupied by the same number of agents @xmath57 and migration can occur between between any two islands with a constant and equal migration rate @xmath58 .
we interpret @xmath59 as the total fraction of the agent population at any given island that migrates to other islands in a given time interval , namely one generation . in a first approximation
, we assume that migration only occurs _
after _ a generation is completed , i.e. at fixed times @xmath10 , @xmath56 .
while the fraction @xmath60 is fixed ( but controlled for , afterwards ) , the _ composition _ of the migrating subpopulation across the existing strategies @xmath43 is not .
we assume this composition to be proportional to the fraction @xmath61 of these strategies at island @xmath5 at the time when generation @xmath10 is completed and migration occurs , to affect the dynamics during the next time interval @xmath56 .
this means migration changes the fraction @xmath62 by an additional amount @xmath63 representing the difference between the influx and the outflux of the agents playing strategy @xmath43 : @xmath64 which defines the complete dynamics by the following set of iterative equations : @xmath65 using @xmath66 and an equal share @xmath67 for each strategy on each island as initial conditions , the dynamics is completely determined if we know the respective average payoffs which are derived in the following section .
we note that , in general , the fixpoints of a difference equation may not characterise its asymptotic behaviour , while it holds for the given simplified case . in a deterministic game , we are able to calculate the average payoff @xmath68 that is received _ by agent @xmath12 _ playing strategy @xmath69 , @xmath70 , @xmath0 times with an agent @xmath13 playing strategy @xmath71 , @xmath72 . for @xmath44 ,
the results are given in the @xmath73 payoff matrix of eq .
( [ strat - payoff ] ) . note that the matrix is not symmetric , since it gives the average payoff of agent @xmath12 .
@xmath74 refers to the row and @xmath75 to the column of eq .
( [ strat - payoff ] ) .
@xmath76\ ] ] where @xmath77 the symbols t , r , p and s refer to the payoff matrix of the 2-person game , eq .
( [ 2payoff ] ) .
further , we have used @xmath78 , with @xmath79 .
@xmath80 means the _ integer _ part of the division @xmath81 , i.e. @xmath82 for @xmath83 ; while @xmath84 means the _ remainder _ , i.e. @xmath85 for @xmath83 .
the constants appearing are defined as follows : @xmath86 to verify how the matrix entries of eq .
( [ strat - payoff ] ) are derived we have provided a number of illustrative examples shown in [ appa ] . with the known average payoff resulting from each possible interaction ,
the average payoff per strategy at island @xmath5 is simply given as : @xmath87 where @xmath88 denotes the fraction of agents playing strategy @xmath69 at island @xmath5 and the @xmath89 are given in eq .
( [ strat - payoff ] ) .
stricly speaking , eq .
( [ a - per ] ) has to consider the fact that an agent does not play against itself .
this would lead to a correction term of the order of @xmath90 , which is small and therefore neglected here . with the above specifications we are now able to solve the dynamic equations with respect to the fraction of agents @xmath91 playing strategy @xmath43 . here , we are mostly interested in the impact of migration on the prevailence of certain strategies , therefore we will discuss first the case _ without _ migration , eq .
( [ a - dyn ] ) , which should be used as a reference for the case _ with _ migration , eq .
( [ eq : migration ] ) . also , we are mostly interested in the _ stationary _ solutions of the respective dynamics , i.e. in the fixpoints @xmath61 reached after a suffiently large number of generations .
our focus is then on three issues which crucially determine the long - term outcome : \(a ) the initial frequencies of strategies @xmath92 , \(b ) the number of repeated interactions , @xmath0 , as this enters the payoff matrix , eq . ( [ strat - payoff ] ) and therefore affects the average payoff per strategy , eq .
( [ a - per ] ) \(c ) the impact of the migration rate @xmath60 on the prevailing strategies .
the stationary values of @xmath61 are reached if the different frequencies do not change anymore in the new generation .
this leads to the stationary condition : @xmath93 where @xmath61 shall denote the stationary ( i.e. asymptotic ) frequency of strategy @xmath43 , in contrast to @xmath94 that denotes the initial frequency . for the full dynamics which includes migration , eq .
( [ eq : migration ] ) , we solve this condition numerically .
but for the case without migration , eq .
( [ a - dyn ] ) , the stationary condition leads to @xmath95 , which reduces to the quite simple expression @xmath96 only if @xmath97 . to further discuss the case of isolated islands
, we drop the index @xmath5 . for @xmath98
( [ stat ] ) with eq .
( [ a - dyn ] ) basically yields two different types of solutions , which could be either stable or unstable : * the exclusive dominance of any _ one _ strategy @xmath71 @xmath99 while all other strategies @xmath69 , @xmath100 , have disappeared , i.e. @xmath101 * the co - existence of some ( or all ) of the strategies @xmath71 , @xmath69 with the _ same average payoff _ @xmath102 , but probably _ different frequencies of agents _ , @xmath103 , @xmath104 , i.e. @xmath105 in the deterministic case considered here it solely depends on the initial conditions @xmath106 , which of the possible stationary solutions will be eventually reached . after we know the _ basins of attraction _ , i.e. the range of possible initial conditions that lead to a given stationary solution and the _ separatrix _ deviding them , we could tell from the outset which strategies will be adopted by the population , and which strategies will disappear . therefore , in the next section we will calculate the basins of attraction for the case of isolated islands , to compare these later on for the case of migration which , as we will see , shifts these basins considerably .
for isolated islands , condition ( [ stat ] ) leads to @xmath107 coupled equations , which follow from eq .
( [ a - bar ] ) with eq .
( [ a - per ] ) .
in addition , we have the boundary condition @xmath108 , which leads to three independent variables @xmath109 , @xmath110 , @xmath111 .
the separatix then appears as a two - dimensional _ plane _ in this three - dimensional space , which will be hard to calculate analytically . in order to elucidate this multi - dimensional problem
, we discuss in [ 2d ] the case of only two possible strategies , where the separatrix appears as a _ point _ in the one - dimensional space , and in [ 3d ] the case of three possible strategies , where the separatrix appears as a _ line _ in the two - dimensional space . here
, we continue with the full problem of four possible strategies . compared to [ 3d ]
, strategy @xmath112 a - tft is added which in the long run benefits agents playing all - d and therefore will decrease the basins of attraction of the cooperating strategies .
i.e. considering a - tft as an additional strategy will create a more difficult environment for the invasion of cooperation , therefore it will be very interesting to see , under what conditions cooperation can survive also in the worst case .
the stationary solutions for the frequencies @xmath113 result from the following set of coupled equations : @xmath114 the matrix elements @xmath115 can be calculated from eq .
( [ strat - payoff ] ) .
with @xmath116 we find the following stationary solutions : @xmath117 we note that the stationary solutions also cover the results of the two - strategy case , [ 2d ] and the three - strategy case , [ 3d ] .
we find solutions of type ( i ) where one strategy is exclusively dominating the whole population .
however , the solution of all agents playing @xmath112 a - tft is also an unstable attractor , because a small perturbation by agents playing either strategy all - d or tft will lead to the invasion of that respective strategy into the whole a - tft population . to see this
, we also refer to the payoffs received in the respective interactions , eqs .
( [ 4 - 2 ] ) , ( [ 1 - 4 ] ) . from the solutions
( ii ) that describe the coexistence of strategies we further see that the addition of strategy @xmath112 a - tft does not lead to new coexisting states compared to the case of two strategies , [ 2d ] and three strategies , [ 3d ] .
there is only a coexistence between strategies @xmath118 tft and @xmath119 all - c , which is stable as long as no spontaneous mutations towards all - d or a - tft occur .
the separatrix that distinguishes between the different attractors is in the given case a two - dimensional plane in the three - dimensional space of independent frequencies with @xmath120 . with respect to the cooperative strategy tft it defines the _ threshold frequency _ @xmath121 that has to be reached initially , in order to allow the system to converge to a fully cooperative state , @xmath122 . because of the @xmath0-dependence of the payoff matrix , eqs .
( [ strat - payoff ] ) , ( [ matrix-2 ] ) @xmath123 strongly decreases with the number of encounters @xmath0 , as shown in fig .
[ fc-3 ] . in agreement with known results from evolutionary game theory @xcite ,
for a sufficiently large number of interactions the whole population will adopt the cooperative strategy , this way increasing the average payoff . the important point to notice here
is the quantitative analysis which allows us to determine , for a given _
finite _ number of @xmath0 , the threshold initial frequency @xmath123 of agents playing tft , in order let cooperation invade the whole population . or else , for a given initial frequency @xmath124 , we are able to determine the _ minimum number of interactions _ , @xmath125 , to let the whole population adopt the tft strategy . of agents playing strategy @xmath118 tft vs. number of interactions @xmath0 between each two agents , in the case of four strategies .
for @xmath126 , the tft strategy will be adopted by the whole population , while for @xmath127 the @xmath128 all - d strategy will be adopted by the whole population .
[ fc-3 ] , width=226 ] to get some further insights , we find it more convenient to present two - dimensional graphical projections of this three - dimensional space , where the separatrix naturally appears as a line .
the most interesting projection is the @xmath129 projection , since it allows us to compare the results with the previous cases .
the results are shown fig .
[ fig : basin_4 ] . they have been obtained by numerically solving eq .
( [ a - dyn ] ) for the full range of initial frequencies , @xmath130 . in the stationary limit
, the average payoff was then evaluated to find out to which basin of attraction the solution belongs .
that lead to a particluar stable solution , eq .
( [ separat-4 ] ) , in the case of four strategies .
@xmath131 : adoption of @xmath128 all - d strategy in the whole population , @xmath132 : adoption of @xmath118 tft strategy in the whole population , @xmath14 : coexistence of both @xmath118 tft and @xmath119 all - c strategies .
@xmath133 , @xmath134 .
( left ) @xmath135 , ( middle ) @xmath136 , ( right ) @xmath137 . [
fig : basin_4],title="fig:",width=188 ] that lead to a particluar stable solution , eq .
( [ separat-4 ] ) , in the case of four strategies .
@xmath131 : adoption of @xmath128 all - d strategy in the whole population , @xmath132 : adoption of @xmath118 tft strategy in the whole population , @xmath14 : coexistence of both @xmath118 tft and @xmath119 all - c strategies .
@xmath133 , @xmath134 .
( left ) @xmath135 , ( middle ) @xmath136 , ( right ) @xmath137 . [
fig : basin_4],title="fig:",width=188 ] that lead to a particluar stable solution , eq .
( [ separat-4 ] ) , in the case of four strategies .
@xmath131 : adoption of @xmath128 all - d strategy in the whole population , @xmath132 : adoption of @xmath118 tft strategy in the whole population , @xmath14 : coexistence of both @xmath118 tft and @xmath119 all - c strategies .
@xmath133 , @xmath134 .
( left ) @xmath135 , ( middle ) @xmath136 , ( right ) @xmath137 . [
fig : basin_4],title="fig:",width=188 ] the calculation of the basins of attraction in fig .
[ fig : basin_4 ] are done for a fixed number of interactions , @xmath116 , but for a varying initial frequency of agents playing strategy @xmath112 a - tft .
of course , this sets limits to the _ maximum _ initial frequencies @xmath138 and @xmath139 , therefore in fig .
[ fig : basin_4 ] we have scaled the @xmath92-axes by the maximum possible value , which should be noticed when comparing the figures . for a small initial frequency of a - tft ( fig .
[ fig : basin_4 ] left ) we find the picture rather similar to the case of three strategies shown in [ 3d ] , however the separation line between the basins of attraction @xmath131 ( defection ) and @xmath140 ( cooperation ) is now a non - linear function , different from eq .
( [ line ] ) . with an increasing initial frequency of a - tft , we find that the basin of attraction of defection is naturally decreasing in absolute size , but it is increasing _ relative _ to the basin of attraction of cooperation .
that means that an increase in the fraction of agents initially playing a - tft always benefits the adoption of the all - d strategy , at the end .
a similar conclusion can be drawn also for the number of agents initially playing all - c .
we notice that with an increasing initial frequency of a - tft region @xmath14 , that describes the coexistence of both strategies tft and all - c , ceases to exist , i.e. strategy all - c has vanished for the benefit of strategy all - d . this can also be understood by looking at the average payoff matrix , eq .
( [ strat - payoff ] ) .
if an agent playing all - d meets with an agent playing all - c , it gets the maximum payoff , @xmath18 , while it gets almost the maximum payoff @xmath141 when it meets an agent playing a - tft .
thus , strategy all - d will eventually invade the population of agents playing either all - c or a - tft , or , in other words , the presence of strategies like all - c and a - tft in the population helps strategy all - d to invade . eventually , we have also calculated numerically the relative size of each basin of attraction as shown in fig .
[ fig : bar_4 ] for two different values of @xmath116 and @xmath142 . here
the relative sizes @xmath143 , @xmath144 , @xmath145 and @xmath146 refer to the three - dimensional space of all possible initial conditions ( which is different from fig .
[ fig : basin_3 ] where they refer to the areas @xmath131 , @xmath132 and @xmath14 shown , thus eq . ( [ area ] ) does not apply here ) . when @xmath134 , the all - d basin a is slightly bigger than the cooperative basin d as shown in contrast to figure [ fig : bar_3 ] . with @xmath147 ,
the size of the all - d basin a becomes pretty small as compared to cooperative basin d. , while the right bars ( black area ) refer to @xmath142 .
thus , the change indicates the influence of @xmath0 on the size of the basins of attraction .
[ fig : bar_4 ] , width=226 ] to conclude this analysis , we have shown for the case of isolated islands ( no migration ) what are the critical initial conditions for cooperative strategies , such as all - c , tft to be adopted by the whole agent population .
dependent on the number of interactions @xmath0 , we can tell the minimum initial frequency of tft , @xmath124 that has to be reached to ensure an cooperative outcome dependent on the initial mix of strategies ( details of the two and three strategy case are provided in [ 2d ] , [ 3d ] ) . in the limit of small @xmath0 , while one may naively assume that an initial increase of the all - c strategy benefits cooperation , at the end , this analysis shows that with the involvement of a - tft it eventually benefits defection in the system , which is an interesting finding by itself . in the following
, we will investigate how this picture changes if we add the possibility of migration .
the previous section has shown that an increase of the number of interaction @xmath0 always supports the prevailence of cooperating strategies , such as all - c and tft .
therefore , in the following we will concentrate the analysis on the critical conditions which occur for small @xmath0 .
we fix @xmath116 ( note that @xmath148 is the threshold number of interaction to allow the replication of cooperating strategies at all , see [ 2d ] and figure [ fc-3 ] . ) using two strategies .
the defective strategy all - d has the majority on all islands .
( ii ) final state ( @xmath149 ) ( a ) without and ( b ) with a small migration rate . in
( a ) only one island has adopted cooperation , in ( b ) all three islands have adopted cooperation , scaledwidth=80.0% ] to understand how migration affects the existing equilibrium states , let us start with the simple example shown in fig . [
fig : island ] .
we take the setup of fig .
[ fig : lattice ] and consider only two strategies , @xmath118tft and @xmath128all - d , but different initial conditions on the 3 islands . as the detailed analysis of [ 2d ] has shown , below an initial frequency @xmath150 the only stable stationary state is @xmath151 , @xmath152 .
hence , if no migration is possible , with the given initial conditions shown in fig .
[ fig : island](i ) the three islands converge into a final state shown in fig .
[ fig : island](iia ) where all - d prevails in two islands , while in one tft dominates because that island started from an initial condition above the threshold .
if we however introduce a rather small migration rate of @xmath153 all three islands are eventually dominated by the cooperating strategy , tft , as shown in fig .
[ fig : island](iib ) .
hence , even with a rather small advantage on _ one _ island ( @xmath154 is just above the threshold of 0.2 ) , a small migration was able to induce a transition toward tft instead of a relaxation into all - d on _ all _ islands .
it is precisely this kind of phenomenon that we would like to understand better regarding the critical conditions involved . in particular , we concentrate on the impact of the migration rate on the threshold value of tft , to still observe the outbreak of cooperation .
the results are found numerically , by iterating the set of eqs .
( [ eq : migration ] ) , for @xmath116 and @xmath1 islands that have the same total number of agents , but different initial strategy distributions . island model for different migration rates @xmath60 : ( left ) @xmath155 , ( middle ) @xmath156 , ( right ) @xmath157 .
parameters : panmictic population in @xmath158 island model consisting of tft and all - d .
initial frequencies of tft : @xmath159 and @xmath160.,title="fig:",width=200 ] island model for different migration rates @xmath60 : ( left ) @xmath155 , ( middle ) @xmath156 , ( right ) @xmath157 .
parameters : panmictic population in @xmath158 island model consisting of tft and all - d .
initial frequencies of tft : @xmath159 and @xmath160.,title="fig:",width=200 ] island model for different migration rates @xmath60 : ( left ) @xmath155 , ( middle ) @xmath156 , ( right ) @xmath157 .
parameters : panmictic population in @xmath158 island model consisting of tft and all - d .
initial frequencies of tft : @xmath159 and @xmath160.,title="fig:",width=200 ] fig .
[ fig : evol_2d ] shows results of the two strategies case , tft and all - d , for @xmath158 and three different migration rates , in terms of the average payoff @xmath161 . as explained in [ 2d ] @xmath162 means that all - d invades the entire population , whereas @xmath163 is achieved when tft takes over .
[ fig : evol_2d](left ) with @xmath98 is used for comparison only , as it shows the result predicted in [ 2d ] .
i.e. , three islands which have a frequency of tft below the threshold value of @xmath164 are taken over by agents playing all - d .
the remaining two islands are invaded by tft .
this result changes in fig .
[ fig : evol_2d](middle ) , where we allow migration of @xmath156 between islands . then all islands become populated solely by tft which is indicated by @xmath163 . however ,
if we further increase the migration rate to @xmath157 , the _ opposite _ result occurs . in this case
all - d invades all islands as shown in fig .
[ fig : evol_2d](right ) where @xmath162 .
these results suggest that there exists _ an optimal range for @xmath60 _ that ensures the outbreak of cooperation , which will be further investigated in the following sections .
we continue our discussion for the case of two strategies , tft and all - d , and now assume that for @xmath165 @xmath166 islands are initially dominated by stategy all - d , i.e. @xmath167 for @xmath168 , while the _ first _ island @xmath169 is initially populated by both tft and all - d , i.e. @xmath170 .
we then vary the migration rate @xmath60 for different initial frequencies @xmath171 to study numerically under which condition the outbreak of cooperation on all islands can be obtained .
the results are shown in fig .
[ fig : howto_find_threshold ] . for very small
@xmath60 , the average payoff @xmath102 just reflects the mix of the initial strategies .
but , as the middle part of fig .
[ fig : howto_find_threshold ] shows , for a certain value of the initial frequency @xmath171 , we find a spike @xmath172 at a particular value of @xmath60 that indicates the _ outbreak of cooperation _ , thanks to an _ optimal migration rate_. one would assume that increasing the initial fraction of tft would further boost cooperation but , as fig . [ fig : howto_find_threshold ] ( right ) clearly shows , the `` bandwidth '' of optimal migration rates @xmath60 is rather limited .
precisely , if @xmath173 the bandwidth of optimal migration rates is @xmath174 which is to be compared to the maximum bandwidth @xmath175 for @xmath176 .
this makes sense because a large migration rate may have a contrary effect : while cooperative agents are `` exported '' to other islands , the same number of defective agents are `` imported '' , which may reduce the fraction of tft below the critical threshold .
hence , there is a nonmonotonous relation between @xmath60 and the outbreak of cooperation .
leads to increasing bandwidth of @xmath60 in which cooperation will emerge in the system . from left to right : ( 1 ) @xmath177 ( 2 ) @xmath178 ( 3 ) @xmath179 ( 4 ) @xmath180 . parameters : @xmath134 , @xmath165,title="fig:",width=192 ] leads to increasing bandwidth of @xmath60 in which cooperation will emerge in the system . from left to right : ( 1 ) @xmath177 ( 2 ) @xmath178 ( 3 ) @xmath179 ( 4 ) @xmath180 .
parameters : @xmath134 , @xmath165,title="fig:",width=192 ] leads to increasing bandwidth of @xmath60 in which cooperation will emerge in the system . from left to right : ( 1 ) @xmath177 ( 2 ) @xmath178 ( 3 ) @xmath179 ( 4 ) @xmath180 . parameters : @xmath134 , @xmath165,title="fig:",width=192 ] we are now interested in finding the _ minimum inital frequency _ @xmath181 that will lead to invasion of cooperation on all @xmath182 islands , when varying the migration rate @xmath60 .
we define this as the _ threshold frequency _
, @xmath183 : @xmath184 for each @xmath185 , there exists a specific value of the migration rate @xmath186 $ ] such that the outbreak of cooperation happens on all islands .
@xmath187 denotes the the smallest value of the bandwidth of optimal migration rates shown in fig .
[ fig : howto_find_threshold ] . i.e. , for fixed @xmath0 and a given number of islands @xmath1 with the initial conditions specified in eq .
( [ eq : threshold ] ) , there exist a tuple of critical values @xmath188 which determine the _ outbreak of cooperation_. the outbreak shown in fig .
[ fig : howto_find_threshold](middle ) for @xmath165 is observed for @xmath189 .
this leads us to the question , how the threshold value and the critical migration rate change if we change the number of islands .
( black line ) and relative effort @xmath190 ( dashed line ) dependent on the number of islands @xmath1 .
symbols represent results of numerical calculations .
fitted values of eq .
( [ eq : nonlin - k ] ) @xmath191 , @xmath192 , @xmath193 .
( right ) optimal migration rate @xmath194 dependent on the number of islands @xmath1 .
[ fig : migra_vs_k_and_thres_vs_k],title="fig:",width=226 ] ( black line ) and relative effort @xmath190 ( dashed line ) dependent on the number of islands @xmath1 .
symbols represent results of numerical calculations .
fitted values of eq .
( [ eq : nonlin - k ] ) @xmath191 , @xmath192 , @xmath193 .
( right ) optimal migration rate @xmath194 dependent on the number of islands @xmath1 .
[ fig : migra_vs_k_and_thres_vs_k],title="fig:",width=226 ] the results of exhaustive numerical calculations which had to search for the outbreak of cooperation on the whole @xmath195 parameter space are shown in fig .
[ fig : migra_vs_k_and_thres_vs_k](left ) . obviously , for @xmath196 the threshold frequency results from the analysis given in [ 2d ] .
if the number of islands increases , @xmath171 has to increase as well , in order to be able to `` export '' cooperating agents while still staying above the theshold .
this increase , however , is nonlinear in @xmath1 .
precisely , as can be verified in fig .
[ fig : migra_vs_k_and_thres_vs_k ] , @xmath197 the values of the constants @xmath198 depend implicitely also on the payoff matrix , eq . ( [ 2payoff ] ) , which defines the comparative advantage of each strategy for reproduction and invasion , subsequently .
plotting @xmath199 we see that for @xmath200 , with every new island added the _ relative effort _ to invade the new one with cooperators _ decreases _ , while for @xmath201 this effort _ increases _ quadratically with @xmath1
. however , the effort of invading other islands with cooperators always stays below the threshold of an isolated island , which is 0.2 .
this is a direct consequence of the combined processes of migration and reproduction .
newly arriving cooperative agents will reproduce on the other islands , this way reaching the threshold value the before the next generation is ready for migration .
hence , _ diversification of the reproduction sites _ for cooperating agents lower the relative effort for the outbreak of cooperation and makes it more favorable . as a second consequence of the nonlinear dependency , the export of cooperation from _
one _ island to @xmath182 other islands is only possible for a limited number of islands , because of the two effects already mentioned above .
i.e. the `` export '' of cooperating agents and the `` import '' of defecting agents both decrease the fraction of cooperating agents below the threshold frequency . hence , for the two strategy case discussed here , @xmath202 already leads to the collapse of exporting cooperation . the critical migration rate @xmath187 dependent on the number of islands is shown in fig .
[ fig : migra_vs_k_and_thres_vs_k](right ) .
we recall that this gives the minimal fraction of the population that has to migrate in order to allow the outbreak of cooperation on any island . if @xmath203 it is at the same the maximum fraction to not let cooperation collapse back home , i.e. it is the optimal migration rate ( see fig . [
fig : howto_find_threshold](middle ) eventually , one can ask how the outbreak of cooperation @xmath204 is affected if instead of the two strategies tft , all - d three or four strategies are considered . in general , the presence of both all - c and a - tft favors the invasion of all - d at the end , whereas the presence of a - tft favors the invasion of tft thus we can expect that the thresholds are slightly higher or lower in such cases . a more involved discussion is presented in [ divers ] .
before summarizing our findings , we wish to comment on a related strand of literature about multi - level selection in populations @xcite . there , quite similar to the setup presented in our paper , a population consisting of several groups ( subpopulations ) was considered where interaction only occurs between agents of the same subpopulation .
the fitness of agents was determined by the payoff obtained from an evolutionary game , simply chosen as a non - iterative prisoner s dilemma , and their reproduction was assumed to be dependent on their payoff according to a moran process @xcite .
further a stochastic dynamics was considered .
in addition to these differences ( pd , moran process , stochastic dynamics ) , also a separate group dynamics was assumed .
groups can split when reaching a certain size , which denotes a amplified replication process at the group level .
precisely , groups that contain fitter agents reach the critical size faster and , therefore , split more often .
this concept leads to selection among groups , although only individuals reproduce .
it allows for the emergence of higher level selection ( group ) from lower level reproduction ( agents ) .
it was shown @xcite that cooperation in such a setting is favored if the size of groups is small whereas the number of groups is large .
migration does _ not _ support the outbreak of cooperation in this model because it favors the invasion of defectors rather than of cooperators .
in contrast to these investigations , we have contributed to the analysis of evolutionary ( deterministic ) ipd games in two different domains : * for the meanfield limit , represented by a panmictic population where each agent interacts with @xmath35 other agents @xmath0 times , a quantitative investigation of the basins of attraction was provided .
dependent on the initial mix of two cooperative ( all - c , tft ) and two defective ( all - d , a - tft ) strategies and the number of interactions @xmath0 , we could derive the critical conditions ( threshold frequencies ) under which the outbreak of cooperation can be expected in a panmictic population . a detailed analysis of the two , three and four strategy cases helped to better understand the contribution of each strategy on the final outcome . * using the `` unpertubed '' ( or meanfield ) limit as a reference point
, the impact of migration on the outbreak of cooperation in a spatially distributed system could be quantified .
we found that there is a bandwidth of optimal migration rates which lead to the induction ( or `` export '' ) of cooperation to a number of islands which otherwise would converge to defection .
we were able to determine the critical conditions @xmath205 which guarantee the outbreak of cooperation in a worst case scenario .
remarkably , the relative effort @xmath206 to export cooperation to defecting islands was decreasing with @xmath1 in some range and always stayed below the critical value for an isolated island .
i.e. effectively it became easier to turn defecting into cooperating islands , provided the optimal migration rate was reached .
it is important to note that the outbreak of cooperation is not enforced by a maximal migration rate but , as stated above , by a small range of optimal migration rates .
this is because migration , different from other approaches , is not seen as an unidirectional dynamics , where agents just move to a `` better '' place . instead
, our model is based on the assumption that it is a _ bidirectional
_ process , which we deem a more realistic assumptions when considering a dynamics over many generations .
in fact , agents , or their offsprings , which have obtained certain skills or wealth at their immigration country , quite often decide to start new buisness back home at their origin country because their new capabilities give them a considerable advantage there , while it is only a marginal advantage in their immigration country .
this assumption is in line with many policies about immigration / education of foreigners in a country , to give indirect support for development and to allow future business with those countries , where some agents with positive experience have resettled . at the end , as we have shown in our paper , it remarkably helps to spread `` cooperation '' globally , if the number of breeding places for such a strategy is increased ( above a minimum threshold ) .
this work on the outbreak of cooperation was supported by the cost - action mp0801 _ physics of competition and conflicts_. f.s .
announces financial support from the swiss state secretariat for education and research ser c09.0055 .
10 urlstyle [ 1]doi:#1 axelrod , r. and hamilton , w. , the evolution of cooperation , _ science _ * 211 * ( 1981 ) 13901396 .
bladon , a. j. , galla , t. , and mckane , a. j. , evolutionary dynamics , intrinsic noise and cycles of co - operation , _ physcal review e _ * 81 * ( 2010 ) 066122(112 ) .
cohen , m. d. , riolo , r. l. , and axelrod , r. , the emergence of social organization in the prisoner s dilemma : how context - preservation and other factors promote cooperation , technical report 99 - 01 - 002 , santa fe institute ( 1999 ) , working paper .
doebeli , m. , blarer , a. , and ackermann , m. , population dynamics , demographic stochasticity , and the evolution of cooperation , _ proc .
natl acad .
sci , usa _ * 94 * ( 1997 ) 51675171 .
fogel , d. b. , on the relationship between the duration of an encounter and evolution of cooperation in the iterated prisoner s dilemma , _ evolutionary computation _ * 3 * ( 1995 ) 349363 .
helbing , d. and yu , w. , migration as a mechanism to promote cooperation , _ adv .
complex syst _ * 11 * ( 2008 ) 641652 .
helbing , d. and yu , w. , the outbreak of cooperation among success - driven individuals under noisy conditions , _ proceedings of the national academy of sciences _ * 106 * ( 2009 ) 3680 .
hofbauer , j. and sigmund , k. , _ the theory of evolution and dynamical systems _
( cambridge university press , 1988 ) .
imhof , l. a. , fudenberg , d. , and nowak , m. a. , evolutionary cycles of cooperation and defection .
, _ proceedings of the national academy of sciences of the united states of america _ * 102 * ( 2005 ) 10797800 .
jiang , l. , wang , w. , lai , y. , and wang , b. , role of adaptive migration in promoting cooperation in spatial games , _ physical review e _ * 81 * ( 2010 ) 036108 .
lozano , s. , arenas , a. , and snchez , a. , mesoscopic structure conditions the emergence of cooperation on social networks , _ plos one _ * 3 * ( 2008 ) e1892 .
michael , m. , natural selection and social learning in prisoner s dilemma , _ sociological methods and research _ * 25 * ( 1996 ) 103138 .
nowak , m. a. and sigmund , k. , tit for tat in heterogeneous populations , _ nature _ * 355 * ( 1992 ) 250253 .
rapoport , a. , prisoner s dilemma : reflections and recollections , _ simulation and gaming _ * 26 * ( 1996 ) 489503 .
roca , c. , cuesta , j. , and snchez , a. , effect of spatial structure on the evolution of cooperation , _ physical review e _ * 80 * ( 2009 ) 046106 .
schweitzer , f. , behera , l. , and mhlenbein , h. , evolution of cooperation in a spatial prisoner s dilemma , _ advances in complex systems _ * 5 * ( 2002 ) 269300 .
schweitzer , f. , mach , r. , and m , h. , agents with heterogeneous strategies interacting in a spatial ipd agent s strategies , in _ lecture notes in economics and mathematical systems _
lux , t. , reitz , s. , and samanidou , e. , i ( springer , 2005 ) , pp .
87102 .
szabo , g. , antal , t. , szabo , p. , and droz , m. , spatial evolutionary prisoner s dilemma game with three strategies and external constraints , _ physical rev .
* 62 * ( 2000 ) 10951103 .
szabo , g. and toke , c. , evolutionary prisoner s dilemma game on a square lattice , _ physical review e _ * 58 * ( 1998 ) 6973 .
traulsen , a. and nowak , m. a. , evolution of cooperation by multilevel selection , _ proceedings of the national academy of sciences of the united states of america _ * 103 * ( 2006 ) 109525 .
traulsen , a. , sengupta , a. m. , and nowak , m. a. , stochastic evolutionary dynamics on two levels , _ journal of theoretical biology _ * 235 * ( 2005 ) 393401 .
traulsen , a. , shoresh , n. , and nowak , m. a. , analytical results for individual and group selection of any intensity .
, _ bulletin of mathematical biology _ * 70 * ( 2008 ) 141024 .
vainstein , m. h. and arenzon , j. j. , disordered environments in spatial games , _ physical review e _ ( 2002 ) .
vega - redondo , f. , _ economics and the theory of games _ , 1st edn .
( cambridge university press , new york , 2003 ) .
in order to illustrate eq .
( [ strat - payoff ] ) , let us look at two agents @xmath12 and @xmath13 .
if both of them play strategy @xmath48 , tft , then , during the @xmath0 different `` iterations '' , their choices are as follows : @xmath207{ccccccccccccccccc } & & 1 & 2 & 3 & 4 & \cdots & n_{g}\\ & \mathrm{payoff}\;\ , a_{i}^{(11 ) } : & r & r & r & r & \cdots & r \\
i & ( tft ) : & c & c & c & c & \cdots & c\\ j & ( tft ) : & c & c & c & c & \cdots & c \\ & \mathrm{payoff}\;\ , a_{j}^{(11 ) } : & r & r & r & r & \cdots & r \end{array}\ ] ] consequently , the average payoff for both agents @xmath12 and @xmath13 is @xmath208 . if agent @xmath12 playing tft meets with an agent @xmath13 playing strategy @xmath46 ( all - d ) the series of choices would be : @xmath209{rrcccccccccccccccc } & & 1 & 2 & 3 & 4 & \cdots & n_{g}\\ & \mathrm{payoff}\;\ , a_{i}^{(12 ) } : & s & p & p & p & \cdots & p \\ i & ( tft ) : & c & d & d & d & \cdots & d \\
j & ( all - d ) : & d & d & d & d & \cdots & d \\ & \mathrm{payoff}\;\ , a_{j}^{(21 ) } : & t & p & p & p & \cdots & p \end{array}\ ] ] which leads to an average payoff for agent @xmath12 of @xmath210/n_{g}$ ] , while the average payoff for agent @xmath13 is @xmath211/n_{g}$ ] .
if agent @xmath12 playing a - tft interacts with agent @xmath13 playing all - d , the series of choices reads : @xmath212{rrcccccc } & & 1 & 2 & 3 & 4 & \cdots & n_{g } \\ & \mathrm{payoff}\;\ , a_{i}^{(42 ) } : & p & s & s & s & \cdots & s \\
i & ( a - tft ) : & d & c & c & c & \cdots & c \\
j & ( all - d ) : & d & d & d & d & \cdots & d \\ & \mathrm{payoff}\;\ , a_{j}^{(24 ) } : & p & t & t & t & \cdots & t \end{array}\ ] ] that means the average payoff for agent @xmath12 is @xmath213/n_{g}$ ] , while the average payoff for agent @xmath13 is @xmath214/n_{g}$ ] .
the remaining cases can be explained similarly . only the three more complex cases are given below .
@xmath215 follows from : @xmath216{rrcccccccccccccccc } & & 1 & 2 & 3 & 4 & 5 & 6 & 7 & 8 & \cdots \\ & \mathrm{payoff}\;\ , a_{i}^{(14 ) } : & s & p & t & r & s & p & t & r & \cdots\\ i & ( tft ) : & c & d & d & c & c & d & d & c & \cdots \\
j & ( a - tft ) : & d & d & c & c & d & d & c & c & \cdots \\ & \mathrm{payoff}\;\ , a_{j}^{(41 ) } : & t & p & s & r & t & p & s & r & \cdots \end{array}\ ] ] here , the average payoff of agent @xmath12 results from a repeated payoff series @xmath217 , and the value of @xmath0 defines when this series is truncated .
similarly , the series for @xmath218 is @xmath219 .
thus the first term of @xmath68 , eq . ( [ a - qr ] ) calculates all _ completed _ payoff series ( @xmath220 ) , while the second term calculates the remaining payoffs ( @xmath221 ) . for @xmath222 , we find eventually : @xmath223{rrcccccccccccccccc } & & 1 & 2 & 3 & 4 & \cdots \\ & \mathrm{payoff}\;\ , a_{i}^{(44 ) } : & p & r & p & r & \cdots \\ i & ( a - tft ) : & d & c & d & c & \cdots \\
j & ( a - tft ) : & d & c & d & c & \cdots \\ & \mathrm{payoff}\;\ , a_{j}^{(44 ) } : & p & r & p & r & \cdots \end{array}\ ] ] here , the period of the payoff series is just 2 instead of 4 , and the expression for @xmath224 , eq .
( [ a - qr ] ) follows accordingly .
we assume that the population consists of only two strategies : @xmath118tft and @xmath128all - d , eq .
( [ strat ] ) .
this is an interesting combination since defection is known to be the only equilibrium in an one - shot pd game , while tft fares well in a repeated pd interaction , given there is a critical number of encounters ( discussed below ) . the average payoff @xmath68 received by agent @xmath12 playing strategy @xmath71 ( @xmath225 ) results from eq .
( [ strat - payoff ] ) .
choosing @xmath116 we find : @xmath226 = \left[\begin{array}{cc } 3.0 & 0.75 \\ 2.0 & 1.0 \end{array}\right]\ ] ] applying eqs .
( [ a - bar ] ) , ( [ a - per ] ) , we find for the total average payoff : @xmath227 on the other hand , it follows from the stationary condition ( [ stat ] ) : @xmath228 the combined eqs .
( [ a - tot-2 ] ) , ( [ eq : simul ] ) have to be solved simultaneously for the possible stationary frequencies @xmath109 , @xmath110 . as the result
we find : @xmath229 solution ( i ) implies that either strategy all - d or tft invades the population completely , which are the two stable attractors of the system in the case of only two strategies .
on the other hand , solution ( ii ) describes the coexistence of the two strategies with different frequencies within the total population . in the given case ,
it is is an unstable point attractor that separates the basins of two stable attractors , and therefore acts as a separatrix here .
i.e. , for an initial frequency @xmath230 of strategy @xmath118tft , the dynamics of the system will converge into a stationary state that is entirely dominated by strategy all - d ( where the payoff per agent is @xmath231 ) , whereas in the opposite case the population will entirely adopt strategy tft ( average payoff @xmath232 ) .
the threshold value @xmath233 strongly decreases with the number of encounters @xmath0 , as shown in fig .
[ fc-3 ] for the case of four strategies .
this raises the question about the critical @xmath0 for which tft could invade the whole population .
let us compare the average payoffs of the strategies @xmath118 tft and @xmath128 all - d : @xmath234 the respective values of @xmath115 can be calculated from eq .
( [ strat - payoff ] ) for different numbers of @xmath0 .
we find that only the elements @xmath235 and @xmath236 change with @xmath0 , as follows : @xmath237 with @xmath238 and @xmath239 , we find for @xmath240 that @xmath241 .
this implies that @xmath242 for all possible initial frequencies of the two strategies in the population .
thus , from the dynamics of eq .
( [ a - dyn ] ) ( fitness - proportional selection ) the extinction of the strategy tft results , in agreement with the known results . for @xmath243
we find @xmath244 , @xmath245 , which again implies that @xmath242 for all possible initial frequencies of the two strategies in the population , i.e. the extinction of strategy tft .
however , for @xmath246 , we find @xmath247 , @xmath245 . thus , there exist some initial frequencies of the two strategies for which @xmath248 results .
i.e. , we can conclude that for the given payoff matrix @xmath249 is the _ threshold value _ for the number of interactions between each two agents that may lead to the invasion of cooperation in the whole population .
when strategy @xmath119 all - c is added to the previous strategies @xmath118 tft and @xmath128 all - d , agents playing all - d will benefit from agents playing all - c .
therefore , if the frequency of agents playing all - c is increased in the initial population , it can be expected that the basin of attraction of the tft strategy will shrink , while the basin of attraction of the all - d strategy will grow , compared to the case of two strategies , discussed in [ 2d ] .
the stationary frequencies have now to be calculated from the following set of coupled equations that result from eqs .
( [ a - bar ] ) , ( [ a - per ] ) and the stationary condition ( [ stat ] ) : @xmath250 the matrix elements @xmath115 can be calculated from eq .
( [ strat - payoff ] ) .
with @xmath116 we find the following stationary solutions : @xmath251 we note that a mean - field analysis of the three - strategy case was also dicussed in @xcite and , assuming extensions such as mutations and fluctuations , analysed further in @xcite .
we consider the deterministc case here . solutions ( i ) imply that either strategy tft or all - d or all - c invades the entire population .
we note that only the first two solutions are stable ones , while the last point - attractor @xmath252 , @xmath253 is an unstable one , because any small pertubation ( i.e. the invasion of one defecting agent ) will transfer the cooperating system into a defecting one .
the first of the solutions ( ii ) describing coexisting strategies is already known from the investigation in [ 2d ] to be an unstable one . in the absense of the third strategy , it defines the separatrix point .
the second solution ( ii ) however is a stable one , indicating that both agents playing tft and all - c can coexist in the panmictic population .
note that there is neither a stable nor an unstable coexistence of all three strategies .
the separatix that divides the different basins of attraction is now a line in a two - dimensional space of the initial frequencies . but
different from [ 2d ] the stationary solutions of eq .
( [ separat-3 ] ) do not give further information about the separatices . in order to calculate the different basins of attraction , we therefore have numerically solved eq .
( [ a - dyn ] ) for the full range of initial frequencies : @xmath254 , @xmath255 , @xmath256 , and have evaluated the average payoff in the stationary limit . if @xmath257 , obviously the whole population has adopted strategy @xmath128 all - d .
similarly , if @xmath258 and @xmath259 , the whole population has adopted strategy @xmath118 tft .
however , if @xmath258 and @xmath260 , then there is a coexistence of agents playing strategy @xmath118 tft and @xmath119 all - c . the basins of attraction are shown in fig .
[ fig : basin_3 ] for two different values of @xmath134 and @xmath261 . in the latter case , @xmath262 results for the payoffs in eq .
( [ strat - payoff ] ) .
for @xmath116 we can distinguish between three different regions in fig .
[ fig : basin_3 ] .
region @xmath131 denotes the range of initial frequencies @xmath92 that lead to the adoption of the @xmath128 all - d strategy in the whole population , region @xmath132 denotes the range of initial frequencies that lead to the adoption of the @xmath118 tft strategy in the whole population , while region @xmath14 denotes the range of initial frequencies that lead to the coexistence of both @xmath118 tft and @xmath119 all - c strategies .
that lead to a particluar stable solution , eq .
( [ separat-3 ] ) .
@xmath131 : adoption of @xmath128 all - d strategy in the whole population , @xmath132 : adoption of @xmath118 tft strategy in the whole population , @xmath14 : coexistence of both @xmath118 tft and @xmath119 all - c strategies .
( left ) @xmath134 , ( right ) @xmath263 . [
fig : basin_3],title="fig:",width=226 ] that lead to a particluar stable solution , eq .
( [ separat-3 ] ) .
@xmath131 : adoption of @xmath128 all - d strategy in the whole population , @xmath132 : adoption of @xmath118 tft strategy in the whole population , @xmath14 : coexistence of both @xmath118 tft and @xmath119 all - c strategies .
( left ) @xmath134 , ( right ) @xmath263 . [
fig : basin_3],title="fig:",width=226 ] since regions @xmath132 and @xmath14 both describe the adoption of cooperating strategies in the population , the most interesting line in fig .
[ fig : basin_3 ] is the separatrix between region @xmath131 ( all defectors ) and region @xmath132 ( all cooperators ) .
we can easily interpret this line based on our previous analysis of the two - strategy case , [ 2d ] .
the diagonal in fig .
[ fig : basin_3 ] represents @xmath264 , i.e. a population with only two strategies , @xmath118 tft and @xmath128 all - d .
thus the separatrix line between regions @xmath131 and @xmath132 starts from the separatrix point @xmath265 , @xmath266 , @xmath267 , eq . ( [ separat-3 ] ) . further below the diagonal ,
the frequency @xmath111 of the strategy @xmath119 all - c increases in the initial population , which in turn increases the threshold frequency @xmath109 necessary for the invasion of the @xmath118 tft strategy . in a certain range of frequencies , the separatrix line between defection and cooperation can be described by a linear relation , found numerically as : @xmath268 however , we notice that the separatrix line between regions @xmath131 and @xmath132 never hits the x - axis . for very low values of @xmath269 , i.e. close to the x - axis , it makes a sharp turn towards the origin .
this means that for a vanishing initial frequency of all - d there will be no route to the respective attraction region @xmath131 , which is obviously correct .
the influence of the parameter @xmath0 is shown by comparing the left ( @xmath116 ) and the right part ( @xmath270 ) of fig .
[ fig : basin_3 ] . in the latter case
the basin of attraction @xmath131 ( exclusive domination of strategy @xmath128 all - d ) becomes very small . in order to further quantify the influence of @xmath0 on the dominating strategies in the stationary limit
, we have also calculated the relative size of each basin of attraction . if @xmath271 , @xmath272 and @xmath273 denote the area of the regions @xmath131 , @xmath132 and @xmath14 in fig .
[ fig : basin_3 ] , the relative sizes are defined as follows : @xmath274 the results are shown in fig .
[ fig : bar_3 ] for the two different values of @xmath0 . ) of the basins of attraction shown in fig .
[ fig : basin_3 ] . the left bars ( shaded area )
refer to @xmath116 ( fig .
[ fig : basin_3 ] left ) , while the right bars ( black area ) refer to @xmath142 ( fig .
[ fig : basin_3 ] right ) .
thus , the change indicates the influence of @xmath0 on the size of the basins of attraction .
[ fig : bar_3],width=226 ] in fig .
[ fig : bar_3 ] , @xmath146 , eq . ( [ area ] ) denotes the relative size of the basin of attraction for cooperation resulting from both solutions , domination of @xmath118 tft and coexistence of @xmath118 tft and @xmath119 all - c .
as we see , for @xmath134 , the cooperative basin @xmath146 only has about the same size as the basin of attraction for defection @xmath143 , that mean that about half of the possible initial conditions will lead to a population of defecting agents , at the end . only for @xmath261
the size of the defection basin becomes insignificant as compared to the cooperative basin .
this again explains the role of @xmath275 in influencing cooperation .
in section [ sec : threshold ] , we computed the critical conditions for the outbreak of cooperation @xmath276 in the presence of only two strategies , all - d and tft . if we consider the two additional strategies all - c and a - tft , eq .
( [ strat ] ) , these critical conditions change dependent on the initial values of the four strategies and their distribution on the different islands .
thus , instead of a complete investigations we only discuss the following sample configurations with @xmath277 : these initial conditions imply that @xmath285 .
the results of extensive calculations of the relative effort @xmath286 and the critical migration rate @xmath187 are shown in fig .
[ fig : thres_and_migr_vs_k_all ] and can be compared to the two strategy case , eq .
( [ eq : threshold ] ) . versus number of islands @xmath1 .
( right ) migration rate @xmath60 versus number of islands @xmath1 .
migration can only promote cooperation if the number of islands is @xmath287,title="fig:",width=226 ] versus number of islands @xmath1 .
( right ) migration rate @xmath60 versus number of islands @xmath1 .
migration can only promote cooperation if the number of islands is @xmath287,title="fig:",width=226 ] again , we notice that the relative effort to invade other islands by agents playing tft is non - monotonously dependent on @xmath1 and always stays below the threshold values observed without migration , as given in sect .
[ 3.1 ] , [ 3d ] . regarding the impact of the different strategies on the outbreak of cooperation
, we see that in the presence of all - c all - d benefits more than tft in terms of payoff which raises the threshold frequency . however , adding a - tft benefits all - d less than adding all - c , which lowers the threshold frequency .
this explains why the curve for the _ four _ strategy case is in between the curves corresponding to _ two _ and _ three _ strategies . from fig .
[ fig : thres_and_migr_vs_k_all](right ) the optimal migration rate is found to be almost constant for three and four strategies in contrast to the two strategy case . | we consider a population of agents that are heterogeneous with respect to ( i ) their strategy when interacting @xmath0 times with other agents in an iterated prisoners dilemma game , ( ii ) their spatial location on @xmath1 different islands .
after each generation , agents adopt strategies proportional to their average payoff received . assuming a mix of two cooperating and two defecting strategies , we first investigate for isolated islands the conditions for an exclusive domination of each of these strategies and their possible coexistence .
this allows to define a threshold frequency for cooperation that , dependent on @xmath0 and the initial mix of strategies , describes the outbreak of cooperation in the absense of migration .
we then allow migration of a fixed fraction of the population after each generation . assuming a worst case scenario where all islands are occupied by defecting strategies , whereas only one island is occupied by cooperators at the threshold frequency , we determine the optimal migration rate that allows the outbreak of cooperation on _ all _ islands .
we further find that the threshold frequency divided by the number of islands , i.e. the relative effort for invading defecting islands with cooperators decreases with the number of islands .
we also show that there is only a small bandwidth of migration rates , to allow the outbreak of cooperation .
larger migration rates destroy cooperation . _
keywords : _ migration , cooperation , iterated prisoners dilemma |
the visible , infrared and radio afterglows of grb970228 and grb970508 have taught us many things .
some of them are obvious : the absorption redshift @xmath0 of grb970508 established the cosmological distance scale of grb beyond any reasonable doubt , confirming the very strong case made on statistical grounds@xcite from batse data .
it is also evident that afterglows are very faint .
this suggests that the simultaneous visible counterparts to grb , as yet unobserved , will also be faint , so that experiments designed to detect them will have to be very sensitive . the loss of the original hete , carrying an insensitive ultraviolet imager , may therefore have been fortunate , for it will be replaced by hete-2 which instead will carry a soft x - ray ccd which may yield important spectral information .
observations of afterglows can answer a number of harder questions too : 1 .
what is the relative importance of internal _ vs. _ external shocks ? 2 . where does a grb end and its afterglow begin ? 3
does gamma - ray activity last as long as the afterglow , and could it be an inseparable part of the afterglow ?
these questions are central to the understanding and interpretation of afterglows .
external shocks have been widely considered for grb since they were suggested by rees and mszros@xcite .
they predict a hard - to - soft evolution of the spectrum . in many models
the duration or elapsed time @xmath1 is related to a characteristic emission frequency @xmath2 by a power law @xmath3 ; the exponent @xmath4 is model dependent@xcite but is usually close to @xmath5 .
this is in remarkably good agreement with x - ray and soft gamma - ray observations@xcite of the single - peaked grb960720 in which @xmath6 .
however , the fact that grb with a wide range of durations ( from tenths to hundreds of seconds ) have comparable @xmath2 argues against external shock models , which generally predict that these two quantities should vary roughly reciprocally ; it is difficult for @xmath2 to be in the soft gamma - ray range if the duration is minutes .
there must be more to grb than external shocks .
fenimore , _ et al._@xcite and sari and piran@xcite showed that an external shock can produce only smoothly varying time - dependent emission , not the spiky multi - peaked structure found in many ( but not all ) grb . such complex variation must reflect variation in the supply of energy ; it can not be explained solely by interaction with a heterogeneous medium , however complex .
a variable outflow may radiate when different fluid elements interact with each other .
this process is generally called an internal shock because it does not involve an external medium , although there need be no shock in the hydrodynamic sense of a discontinuity in pressure , density and temperature between two fluids each in thermodynamic equilibrium .
kinematic constraints require that there be inelastic interaction between streams of matter emitted with widely differing lorentz factors in order that radiation be produced with reasonable efficiency .
afterglows have , so far , been observed to have a smooth single - peaked time dependence in visible light ( their complex time dependence at radio frequencies results from interstellar scintillation@xcite ) , and therefore are naturally explained as a consequence of external shocks .
this model predicted@xcite both the existence of afterglows and the gradual rise to maximum which was observed in both grb970228@xcite and grb970508@xcite . however , there is no evidence that internal shocks could not produce the required time dependence , and they are therefore possible explanations for smooth single - peaked grb and for afterglows , as well as for the multi - peaked grb for which they are required .
is it meaningful to distinguish an afterglow from the grb itself ?
this question first arises for x - rays , whose photon energy range overlaps that traditionally assigned to grb .
it has no good answer , and should be considered a matter of nomenclature rather than of physics .
rather , we should be concerned with deciding which emission is produced by internal shocks and which by external ; the taxonomist may wish to label the former the grb and the latter the afterglow .
there are two possible limits@xcite . in one
all the radiation is the product of an external shock .
this is excluded , at least for multi - peaked grb .
it may describe single - peaked grb and their afterglows .
external shock models successfully predicted@xcite the delayed maximum of visible afterglow brightness observed in grb970228@xcite and grb970508@xcite , and are likely to explain such afterglow emission even if they can not explain gamma - ray emission .
these models also successfully predicted@xcite self - absorption in the radio afterglow , as was observed@xcite from grb970508 . in the other limit all the emission of a grb
is the product of internal shocks , which may continue for days@xcite beyond the nominal gamma - ray duration ; the afterglow is a continuation , at lower intensity , of the gamma - ray emitting phase .
this has been suggested for grb970228 , where it agrees with the observed instantaneous spectrum@xcite , and the predicted continuing x - ray emission ( beyond the decaying afterglow ) has been observed from grb970508@xcite .
this hypothesis is consistent with the hours - long gamma - ray emission of grb940217@xcite , and may explain the `` gang of four '' bursts of october 2729 , 1996@xcite as a single grb .
early work on grb and their afterglows attempted to construct analytic and numerical models of the entire process . unfortunately , several essential quantities are poorly known ( and unlikely to be the same for all grb or uniform within a single grb ) , such as the distribution of energy among the electrons , ions and magnetic field , the surrounding density and magnetic field , the efficiency of radiation and the degree of collimation . as is so often the case , the real world has turned out to be more complex than theorists could imagine , and a more phenomenological approach is required .
theories of radiation mechanisms generally predict instantaneous spectra , but observations of afterglows are rarely simultaneous across the spectrum ; x - ray , visible and radio observers face different observational constraints .
it is possible to fit multiparameter models@xcite to non - simultaneous data , but the phenomenologist would like a more direct comparison of the emission in different frequency bands .
for example , in each band of interest the peak spectral intensity can be measured , or at least estimated , if measurements are obtained close to and straddling the maximum .
the resulting function @xmath7 is not an instantaneous spectrum , but makes it possible to compare emission mechanisms in different spectral bands . in the case of grb970508 @xmath7
can be estimated at ghz , visible ( and near - ir ) , hard x - ray and soft gamma - ray energies ( the maximum of the soft x - ray intensity was not observed ) , and is shown in figure [ jikfig1 ] .
a single mechanism is unlikely to explain emission in all these bands , because @xmath7 is not consistent with a single power law , as would be predicted by a single mechanism which does not have a characteristic frequency to define a spectral break .
the visible data form a `` hinge '' ; on purely phenomenological grounds it is not possible to say whether this emission is produced by the same mechanism as the ghz emission or the hard x - rays and soft gamma - rays .
neither of the limiting cases discussed in the previous section is satisfactory .
a plausible hypothesis suggests that ghz and visible emission result from external shocks and hard x - rays and soft gamma - rays ( the classical grb ) from internal shocks .
this explains why an early model@xcite underestimated the time to maximum and overestimated the brightness of the visible afterglow of grb970508 : it extrapolated from the gamma - ray emission of an assumed bright grb , but this extrapolation was invalid ( even if scaled to the lesser flux of grb970508 ) because the gamma - rays were produced by a different mechanism .
it is also possible to compare@xcite the properties of different grb and their afterglows by comparing the burst - to - burst ratios of @xmath7 in different spectral bands .
the afterglow of grb970508 was roughly 20 times brighter than that of grb970228 in soft x - rays and visible light , when scaled to their soft gamma - ray brightnesses .
the afterglow of grb970508 was at least four times brighter than that of grb970111 at 1.43 ghz , and at least 1000 times brighter than that of grb970828 in visible light when similarly scaled ( in these last two cases the ratios are lower limits because only upper limits exist to the fluxes of grb970111 and grb970828 in these bands ) .
these ratios quantify the conclusion that grb970508 had a remarkably bright afterglow , compared to the the other grb for which useful data exist , at all frequencies at which comparisons are possible .
this leads to the important ( and unexpected ) conclusion that afterglows are `` all different '' . they are not all scaled versions of the same event , and any single simple model must fail .
it argues against models in which all grb and afterglows are related by a single scaling parameter , such as the ambient density , energy or distance , or even some combination of these .
a natural interpretation is that ( at least in these grb ) the hard x - rays and gamma - rays are produced by internal shocks , and the lower frequency afterglow by an external shock .
if so , there need be no close correlation between the brightnesses of grb and their afterglows , partly because the mechanisms are different , but also because internal shock properties depend on the detailed temporal and spatial structure of the outflow and may well be very different in different grb .
a generic afterglow model is based on the assumption of an external shock , which permits specific predictions to be made .
the interaction between a relativistic debris shell and the surrounding medium , or among various elements of an outflowing wind , has been the subject of many papers , but the basic physics is not understood .
even the essential collisionless shock is largely a matter of speculation , although recent calculations@xcite have begun to attack the problem .
still , a few features common to most afterglow models are independent of assumptions as to the mechanism of entropy production .
the asymptotic ( @xmath8 ) instantaneous spectrum@xcite has the form @xmath9 between a self - absorption frequency @xmath10 and a characteristic synchrotron frequency @xmath2 .
below @xmath10 the spectrum @xmath11 , while above @xmath2 the flux falls off with a slope reflecting the high energy `` tail '' to the particle distribution function ; a power law is typically observed in grb , but its slope is unpredictable , and is observed to differ from burst to burst and with time in a given burst .
the instantaneous spectra @xmath12 are bounded from above by the function @xmath7 . in simple analytic models @xmath13 , with the exponent @xmath14
typically@xcite close to 0 ; any function of @xmath15 other than a power law would define a characteristic break frequency and require a more complex model . as time progresses both @xmath10 and @xmath2 decrease . at a given frequency
@xmath16 the flux @xmath17 rises until @xmath2 declines to @xmath18 , after which @xmath17 decreases .
this predicted behavior has been observed , at least qualitatively , in the afterglows of grb970228@xcite and grb970508@xcite .
the predicted rate of the initial rise @xmath19 , where the preceding expressions for @xmath7 and @xmath20 and the instantaneous @xmath21 lead to @xmath22 typically , @xmath23 is estimated to be slightly less than unity ; in one model@xcite @xmath24 4/5 while in another@xcite @xmath24 6/7 .
figure [ jikfig2 ] shows the evolution of the instantaneous spectrum .
@xmath17 rises until @xmath25 and then declines .
the predictions of the generic external shock afterglow model can be tested with data from the two observed afterglows .
the predicted rise to maximum has been observed .
some data@xcite suggested agreement ( @xmath26 ) with the predicted rate of rise for grb970508 , but more complete data for grb970228@xcite and grb970508@xcite suggest a much steeper rise , with @xmath273 .
these large values of @xmath23 appear to disagree with the model .
sari@xcite pointed out that in the early stages of external shock models , when the relativistic debris shell has not yet been significantly slowed by the ambient medium , the luminosity should rise @xmath28 at all frequencies ( @xmath29 ) , a simple consequence of the increasing ( @xmath30 ) area of the shell .
if this is the explanation then the rapid rise should level out after a time @xmath31 when deceleration becomes important ( the earlier stages of the rising flux could be hidden under steady emission by other processes , such as internal shocks ) .
this leads to an estimate of the ambient density @xmath32 : @xmath33 if the model is correct this implies an extraordinarily low ambient density out to a radius @xmath34pc , or a surprisingly low value of @xmath35 .
such a low density would be remarkable , even in the intergalactic medium ( although a pre - coalescence pulsar wind could create a bubble ) , and such a low @xmath36 is inconsistent with gamma - gamma pair production constraints ( although in a long duration internal shock model a low @xmath36 wind could follow a brief high @xmath36 wind which produces the gamma - ray emission ) .
if the afterglow is produced by internal shocks it is probably not possible to predict its time dependence .
however , the instantaneous spectrum is still predictable .
there is some spectral evidence that the first several hours of afterglow in grb970228 were the product of internal shocks@xcite .
internal shocks are a possible explanation of disagreements between the observed time dependence and predictions of the generic external shock model . this early ( 38 hours after the grb ) period of roughly constant visible intensity in grb970508 poses another problem . in either internal or external shock models ,
if the electron synchrotron cooling time is short compared to the duration of observation then the observed spectrum is that integrated over the electrons cooling history@xcite : @xmath37 .
comparison of the visible and soft x - ray@xcite fluxes during this period shows a deficiency of x - rays compared even to this spectrum , suggesting that @xmath2 lies within or below the x - ray band .
this is consistent with emission by an internal shock with a low value of @xmath36 , as suggested by both the near constant visible intensity and the long - delayed onset of the rapid rise .
these conclusions are based on limited data from two afterglows .
the outline of the external shock afterglow model is supported by the data , but the detailed interpretation , especially of the rise of the visible intensity to maximum , must await more data from more afterglows .
the predicted@xcite instantaneous asymptotic spectrum @xmath38 for @xmath39 should be applicable to both internal and external shocks . it is based on several plausible assumptions : incoherent synchrotron radiation , no cooling ( radiative or adiabatic ) and a phase space argument for the electron distribution function produced by a relativistic shock which heats the entire electron distribution function , rather than just a `` tail '' of suprathermal particles .
it should apply to the synchro - compton spectrum too , but with a different coefficient . like all theoretical predictions , it is only speculation until empirically tested .
so far , the data are inconclusive and not completely consistent . some x - ray and soft gamma - ray observations of grb support the prediction@xcite , but others disagree@xcite .
the inconsistency may result from the difficulty of extracting quantitative spectral information from nai scintillator data , which have low intrinsic spectral resolution , especially at photon energies @xmath40 kev where the asymptotic spectrum is expected .
this difficulty may also account for the long - standing controversy over the reality of line features in grb radiation , and may only be resolved when data from detectors of intrinsically higher resolution become available . visible data@xcite lead to an exponent @xmath41 in the pre - maximum phase of the afterglow of grb970508 , in agreement with the predicted 1/3 (
agreement is not expected at and after maximum , because then the frequency of observation exceeds @xmath2 ) .
radio data@xcite from the same afterglow lead to an exponent @xmath42 , also in agreement with prediction .
the results are encouraging , but not yet conclusive .
if a grb were to be detected in visible light during its initial brief phase of gamma - ray emission the spectral exponent could be determined quite accurately by comparing fluxes at frequencies separated by a factor of more than @xmath43 .
such simultaneous detection remains the holy grail of grb visible counterpart research .
self - absorption of grb afterglows at ghz frequencies was predicted@xcite and confirmed@xcite .
this was no great surprise ; the physics is elementary , though a little different ( the spectral exponent below @xmath10 is predicted to be 2 rather than 2.5 ) than in the usual case of synchrotron self - absorption by a power law distribution of electron energies .
self - absorption is important because it may lead to a measurement of the emission radius @xmath44 as a function of time with few uncertain assumptions .
the flux for @xmath45 is @xmath46 where @xmath47 is the distance ,
@xmath48 the cosmological redshift , @xmath49 the proton mass and @xmath50 an equipartition factor defined as @xmath51 ; @xmath52 in the case of complete electron - ion - magnetic equipartition .
the shock lorentz factor ( assumed @xmath53 ) drops out ; this is fortunate , for it is poorly known and likely to remain so .
if @xmath54 were inferred from measurements of the self - absorbed flux then it would be possible to reconstruct the expansion history and slowing down of the relativistic debris in an external shock model .
this would permit determination of the ambient density and the efficiency of radiation as the initial kinetic energy is radiated or shared with swept - up matter .
a preliminary attempt@xcite to construct such an energy budget ( based on one inferred value of @xmath44 ) for the afterglow of grb970508 led to the conclusion that the total kinetic energy after seven days was only @xmath55 erg , where @xmath56 is the ambient particle density .
this is much less than the @xmath57 erg inferred from the gamma - ray radiation , assuming isotropic emission
. there are three possible implications of this result : @xmath58 @xmath59 , as might be found in a dense star - forming region ; strong beaming of the gamma - rays ( and therefore of the initial relativistic outflow ) ; or a radiation efficiency @xmath60% in the first seven days of the event .
any or all of these are possible , and any or all would be important .
we have compared the observed afterglows to theoretical predictions based on simple models , and have generally found qualitative agreement . in order to test the theory ,
particularly that of relativistic shocks , in more detail , more difficult measurements will have to be performed .
here is a list of ambitious goals a theorist might set for his observational colleagues : 1 .
accurate x - ray spectroscopy at moderate resolution ( @xmath61 ) to test the predicted instantaneous @xmath9 ( @xmath62 if synchrotron cooling is important ) .
this may be performed by zncdte or ge detectors , or by x - ray ccd .
simultaneous spectral measurements across several decades of frequency , with @xmath63 throughout ( hence the intensity must not have reached its maximum anywhere in this range ) .
this could be achieved by observation of the visible counterpart to a grb during its strong gamma - ray activity , or by simultaneous measurements of visible and radio afterglow before the visible maximum ( within the first day for grb970228 and grb970508 ) .
3 . measurement of the intensity as a function of time in the self - absorbed regime of radio afterglow .
each of these measurements is likely to be difficult .
locating and measuring a visible counterpart to a grb within tens of seconds is much harder than doing it with several hours of imaging x - ray observations .
the only afterglow observed at radio frequencies was not strong enough to be detected until long after the visible maximum , and self - absorption reduces the strength of the radio emission even more .
the first goal is probably the most feasible , and would test relativistic shock theory ; the last is probably the one which would lead to the greatest understanding of how grb and their afterglows really work .
we thank r. sari for discussions .
jik thanks nsf 94 - 16904 for support , washington university for the grant of sabbatical leave and the hebrew university for hospitality and a forchheimer fellowship .
meegan , c. a. , _ et al .
_ , _ nature _ * 355 * , 143 ( 1992 ) .
rees , m. j. , and mszros , p. , _ mnras _ * 258 * , 41p ( 1992 ) .
katz , j. i. , _ ap . j. _ * 422 * , 248 ( 1994 ) .
katz , j. i. , _ ap . j. lett . _ * 432 * , l107 ( 1994 ) .
katz , j. i. , and piran , t. _ ap . j. _ * 490 * , 772 ( 1997 ) .
piro , l. , _ et al .
_ _ astron .
_ in press ( astro - ph/9707215 ) ( 1997 ) .
fenimore , e. e. , madras , c. d. , and nayakchin , s. _ ap . j. _ * 473 * , 998 ( 1996 ) .
sari , r. , and piran , t. _ ap . j. _ * 485 * , 270 ( 1997 ) .
goodman , j. , _ new astron . _ * 2 * , 449 ( 1997 ) .
guarnieri , a. , _ et al .
_ , _ astron .
* 328 * , l13 ( 1997 ) .
pedersen , h. _ et al .
_ , _ ap . j. _ in press ( astro - ph/9710322 ) ( 1998 ) .
frail , d. a. , _ et al .
_ , _ nature _ * 389 * , 261 ( 1997 ) .
katz , j. i. , _ ap .
j. _ * 490 * , 633 ( 1997 ) .
katz , j. i. , piran , t. , and sari , r. , _ prl _ submitted ( astro - ph/9703133 ) ( 1997 ) .
piro , l. , _ et al .
_ , _ astron .
ap . _ submitted ( astro - ph/9710355 ) ( 1998 ) .
hurley , k. , _ et al .
_ , _ nature _ * 372 * , 652 ( 1994 ) .
connaughton , v. , _ et al .
_ , _ proc .
18th texas symp .
_ in press ( 1997 ) .
mszros , p. , and rees , m. j. , _ ap . j. _ * 476 * , 232 ( 1997 ) .
wijers , r. a. m. j. , rees , m. j. , and mszros , p. , _ mnras _ * 288 * , 51p ( 1997 ) .
usov , v. v. , and smolsky , m. v. , _ pre _ in press ( astro - ph/9704152 ) ( 1997 ) .
djorgovski , s. g. , _ et al .
_ , _ nature _ * 387 * , 876 ( 1997 ) .
sari , r. , _ ap .
j. lett . _ * 489 * , l37 ( 1997 ) .
cohen , e. , _ et al .
_ , _ ap . j. _ * 488 * , 330 ( 1997 ) .
schaefer , b. e. , _ et al .
_ , _ ap . j. _ submitted ( 1997 ) .
preece , r. d. , _ et al .
_ , `` the ` line of death ' for the synchrotron shock model : confronting the data , '' these proceedings ( 1998 ) .
crider , a. , liang , e. , and preece , r. , `` restrictions and predictions from grb spectral evolution , '' these proceedings ( 1998 ) .
groot , p. , _ et al .
_ , `` optical and radio follow - up of grb970508 , '' these proceedings ( 1998 ) . | the discovery by bepposax and coordinated ground - based observations of persistent x - ray , visible and radio counterparts to grb has successfully concluded a search begin in 1973 , and the observed redshifted absorption lines have proved that grb are at cosmological distances .
the problem of explaining the mechanisms of grb and their persistent counterparts remains .
there are two classes of models : 1 ) grb continue weakly for days at all frequencies ; 2 ) grb emission shifts to lower frequencies as relativistic debris sweeps up surrounding gas ( in an `` external shock '' ) and slows .
1 ) predicts that the visible afterglow should be accompanied by continuing gamma - ray emission , as hinted by the high energy emission of grb940217 and the `` gang of four '' bursts of october 2729 , 1996 .
it also suggests that the persistent emission will fluctuate .
behavior of this sort may be found in `` internal shock '' models .
2 ) has been the subject of several theoretical studies which disagree in assumptions and details but which predict that at each frequency the flux should rise and then decline , with the maximum coming later at lower frequencies . some of this behavior has been observed , but data from grb970508 show that its afterglow can not be simply extrapolated from its gamma - ray emission .
it is likely that both classes of processes occur in most grb .
comparisons between grb show that they are not all scaled versions of the same event .
these results suggest that most gamma - ray emission is the result of `` internal shocks '' while most afterglow is the result of `` external shocks '' , and hint at the presence of collimated outflows .
self - absorption in the radio spectrum of grb970508 permitted the size of the radiating surface to be estimated , and in future grb it may be possible to follow the expansion of the shell in detail and to construct an energy budget . |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.