article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
larmor s formula of electromagnetism @xcite states that an electric charge in uniform motion does not radiate . however , there are two ways to have radiation from a charge moving with constant velocity .
the first way is to have a particle moving in a medium with velocity exceeding the phase velocity of light in that medium .
this gives rise to the well - known vavilov - cherenkov radiation @xcite .
the second way is to consider charge motion in inhomogeneous media .
ginzburg and frank @xcite first discussed this effect by investigating a particle in uniform motion which crosses a planar interface between two media with dissimilar refractive index .
this kind of radiation is known as transition radiation @xcite .
more generally , any motion near finite - size objects also induces radiation , in a process called diffraction radiation @xcite .
one of the first experimental verifications on this effect was provided by smith and purcell @xcite .
the smith - purcell experimental set - up consisted of an electron moving close to the surface of a metal diffraction grating at right angles to the rulings .
the theory of the smith - purcell effect has been discussed by several authors @xcite .
the aim of this paper is to show the existence of gravitational diffraction radiation ( gdr ) and discuss its physical effects .
the model under consideration is a braneworld scenario @xcite such as , for example , the randall - sundrum i model @xcite .
braneworld models have attracted a lot of interest in recent years , revolutionizing our view of how our universe may be described .
the central idea of braneworld scenarios is that the visible universe is restricted to a four - dimensional brane inside a higher - dimensional space , called the bulk .
the additional dimensions are taken to be compact and other branes may be moving through the bulk .
interactions of the visible brane with the bulk and hidden branes introduce effects not seen in standard physics . in this set - up , a particle in uniform motion on the visible brane radiates gravitational waves due to the presence of a second ( hidden ) brane at finite distance , which plays the role of the metal diffraction grating of the smith - purcell experiment .
gdr on the visible brane is generated by inhomogeneities on the hidden brane due , for example , to bulk - brane interactions and brane fluctuations @xcite .
we will show that the amount of gdr depends on the size of extra dimensions and the length scale of brane perturbations .
this result is general and independent on the fine details of the model . without loss of generality , in our computations we will consider a flat five - dimensional spacetime with the extra dimension taking values within the interval
@xmath0 $ ] .
the distance between the particle and the diffraction grating , @xmath1 , is the distance between the two branes located at the orbifold fixed points .
the model is illustrated pictorially in fig .
[ braneworld1 ] .
the generalization to higher - dimensions is trivial and is discussed at the end of this letter .
is identified with the brane at @xmath2 .
the visible brane is located at @xmath3 .
the particle moves on the visible brane along @xmath4 with constant velocity @xmath5 .
only one transverse coordinate , @xmath6 , is shown .
the right panel shows the wrapping around the extra dimension @xmath7 .
if the two halves of the cylinder are identified , the randall - sundrum model i @xcite is obtained.,title="fig:",scaledwidth=45.0% ] is identified with the brane at @xmath2 .
the visible brane is located at @xmath3 .
the particle moves on the visible brane along @xmath4 with constant velocity @xmath5 .
only one transverse coordinate , @xmath6 , is shown .
the right panel shows the wrapping around the extra dimension @xmath7 .
if the two halves of the cylinder are identified , the randall - sundrum model i @xcite is obtained.,title="fig:",scaledwidth=30.0% ]
in the harmonic gauge , the five - dimensional linearized einstein equations are @xmath8 where @xmath9 represents small corrections to the flat background and @xmath10 is the modified energy - momentum tensor of the source @xcite . for sake of computational simplicity ,
we replace the the gravitational field components with a single scalar degree of freedom , i.e. rewrite the above equation as @xmath11 .
this is a common procedure @xcite .
if we carefully select @xmath12 , the scalar field @xmath13 can mimic all aspects of gravitational waves ( except polarization ) @xcite .
the source of the field is a minimally coupled pointlike particle with nonzero mass @xmath14 .
if we represent its worldline by @xmath15 , where @xmath16 is the proper time , @xmath17 satisfies the inhomogeneous wave equation @xmath18 where @xmath19 is the coupling constant .
we consider a particle in uniform motion on the visible brane .
denoting with @xmath7 the coordinate transverse to the brane and with @xmath20 the longitudinal coordinates perpendicular to the particle direction of motion @xmath4 , the particle worldline in the brane reference frame is @xmath21 , where @xmath22 is the lorentz factor of the particle .
the time - averaged energy radiated per unit time in the direction @xmath23 is @xmath24\ , , \label{powern}\ ] ] where @xmath25 is the surface element with normal @xmath23 . if the tensor structure of the gravitational perturbations is taken into consideration , the source term in eq .
( [ eqmotion2 ] ) must be changed into @xmath26 the results for the scalar field can be translated to the gravitational case by setting @xmath27 @xcite . if the branes are smooth , a particle with constant velocity does not radiate .
this can be checked by deriving the larmor s formula for the field . in order to avoid complications due to the higher - dimensional nature of the spacetime @xcite
, we temporarily assume that there is only one transverse spatial dimension .
the total power emitted per unit of solid angle in the direction @xmath23 is @xmath28 ^ 2}{\gamma^2 ( 1-{\bf v}\cdot{\bf n})^3}{\bf n}\ , .
\label{powerangle}\ ] ] thus a particle in uniform motion in empty space ( on the brane ) does not radiate .
if a hidden brane is present , the green s function has to be modified to take into account its effects .
the hidden brane can be thought as a wall parallel to the particle direction of propagation at distance @xmath29 . by repeating the steps leading to eq.([powerangle ] ) it is straightforward to show that there is no radiation for uniform motion parallel to the wall .
this result can be understood by boosting the solution to the rest frame of the particle .
the problem is reduced to a static problem with one image particle on the other side of the wall .
clearly , the reduction to the static configuration is only possible if the brane is smooth and infinite in the @xmath4 direction .
if the brane is inhomogeneous in the particle direction of propagation , diffraction radiation is generated . in that case , the image configuration is time dependent and the system is equivalent to a set of oscillating charges in the particle reference frame .
diffraction radiation can be understood as being generated by the reflection of the boosted static field on the nearby wall .
we now compute the radiated power in the presence of a brane with typical longitudinal perturbations of length scale @xmath30 and transverse perturbations of length scale @xmath31 .
these perturbations are modeled with a @xmath30-periodic lamellar grating with rulings of width @xmath32 perpendicular to the particle direction of motion as in fig.[braneworld3 ] .
( the smith - purcell effect for a single grating with these characteristics has been discussed in ref .
although this model is clearly an oversimplification , its main features do not depend on the choice of the perturbation structure : the existence of diffraction radiation is due to the excitation of propagating modes by evanescent waves and is independent of the particular mechanism by which the propagating modes are excited . with length @xmath30 .
the transverse perturbations have length @xmath31 .
only one transverse coordinate , @xmath6 , is shown.,scaledwidth=50.0% ] the calculation with a perturbed brane can be conveniently solved by considering the fourier transform of the field @xmath13 : @xmath33 the field equation in the fourier space is @xmath34\phi = g/(\gamma v)\,e^{i\alpha_0 x}\delta(z - h)\ , , \label{fourier2}\ ] ] where @xmath35 .
the general solution for the field @xmath17 in the bulk is the sum of the inhomogeneous solution of the above equation plus a superposition of plane wave solutions of the two - dimensional helmholtz equation : @xmath36e^{i\alpha_n x}\right\}\ , , \label{phi}\ ] ] where @xmath37 , @xmath38 and @xmath39 . since the parameter @xmath40^{1/2}$ ] is pure imaginary , the first term in eq .
( [ phi ] ) describes an evanescent wave that decays exponentially with increasing distance from the particle trajectory .
this evanescent wave is the non - propagating part of the spectrum due to the particle in uniform motion .
the second term in eq .
( [ phi ] ) represents a superposition of propagating plane waves .
the physical interpretation of these propagating modes is that of radiation arising from the modes of the brane perturbations , which are excited by the evanescent waves . the amplitude @xmath41 of the gdr is thus expected to be of the order of the amplitude of the non - propagating modes , i.e. , @xmath42 where we have taken the relativistic limit @xmath43
this can be shown by computing the amplitudes of the radiating modes as function of the evanescent wave amplitude @xmath44 .
the solution in the bulk must match the solution on the brane @xmath45\sin\left[\frac{\pi mx_j}{\delta_j}\right]\ , , \label{phibrane}\ ] ] where @xmath46 and @xmath47 , @xmath48 and @xmath49 in the intervals @xmath50 $ ] ( @xmath51 ) and @xmath52 $ ] ( @xmath53 ) , respectively .
choosing dirichlet boundary conditions , the amplitudes of the propagating waves satisfy the system of linear inhomogeneous algebraic equations : @xmath54 + e^{i\gamma_0l/2}(l\delta_{0k}-v_{1,0k})=0\,,\\ & & \sum_{n=-\infty}^{+\infty}\left[b_ne^{i\gamma_n(l - b)}(l\delta_{nk}+v_{2,nk})+ c_ne^{-i\gamma_n(l - b)}(l\delta_{nk}-v_{2,nk})\right]+ e^{i\gamma_0(l/2-b)}(l\delta_{0k}+v_{2,0k})=0\ , , \label{coeff1}\end{aligned}\ ] ] where @xmath55 and @xmath56 .
the power emitted in the bulk follows from eq .
( [ powern ] ) after solving eqs .
( [ coeff1 ] ) for the propagating wave amplitudes @xmath57 and @xmath58 .
generalizing to @xmath59 spacetime dimensions and taking the relativistic limit , we have @xmath60 where we have set @xmath61 and assumed small @xmath62 . since we assumed the source to be pointlike , eq.([fim ] ) is valid for particle size @xmath63 @xmath64 , @xmath31 and @xmath30 , otherwise it describes the diffraction radiation of a relativistic particle under very general assumptions .
the gdr energy loss per unit distance is @xmath65 where @xmath66 is the @xmath59-dimensional newton constant .
as an example of the physical consequences of gdr , let us consider the scenario with @xmath67 large extra dimensions of size @xmath64 @xcite . in this model @xmath66
is related to the four - dimensional newton constant @xmath68 by @xmath69 .
there are two natural length scales , the planck length , @xmath70 cm , and the size of the large extra dimensions , @xmath71 cm .
the perturbation length scales @xmath30 and @xmath31 are expected to be larger than the planck length .
this implies @xmath72 and @xmath73 .
the largest @xmath22 factors are observed in ultra high - energy cosmic rays , reaching @xmath74 @xcite . in this case , the exponential in eq .
( [ dedx ] ) can be set to unity for all @xmath75 . using the above constraints
, we obtain @xmath76 equation ( [ inuhecr ] ) imposes stringent limits on the smoothness of the brane .
upper bounds on @xmath77 range from @xmath78 for @xmath79 to @xmath80 for @xmath81 .
therefore , gdr implies that the brane must be smooth on scales of order of the size of the large extra dimensions , i.e. , on scales much larger than the planck length .
this fact has important consequences for the problem of initial conditions in models of brane cosmology .
( see , e.g , ref .
in higher - dimensional models of our universe , particles radiate gravitational waves due to the structure of the extra dimensions , specifically the presence of an inhomogeneous hidden brane .
we have called this phenomenon gravitational diffraction radiation " because of its similarity with the electromagnetic diffraction radiation which is emitted by a charge moving near a metallic grating .
gdr is a general phenomenon of all higher - dimensional braneworld scenarios .
it applies to particles propagating both on the brane and in the bulk . as an example of its physical effects
, we have discussed the gdr emitted by a particle on the visible brane and derived lower bounds on the scale of brane perturbations .
the above results can be extended in several directions .
for instance , this paper considers only inhomogeneities on the hidden brane .
the visible brane is also inhomogeneous .
therefore , free motion on the visible brane is not uniform from the five - dimensional perspective .
this effect could be comparable to the gdr effect studied here .
the derivation of gdr presented above is purely classical .
although the scale of brane inhomogeneities suggests that the system can be treated as classical , the importance of quantum effects should be assessed .
it would also be interesting to consider gdr emitted by photons or bulk particles . in the latter case ,
gdr energy loss would manifest itself as a background of stochastic gravitational waves on the visible brane .
gdr is originated by the lumpiness of the spacetime at small scales .
thus gdr effects may show up in a variety of physical situations besides braneworld models .
emission of radiation by free particles in a cosmic string background was discussed in ref .
gdr could also play an important role in the very early universe or near the horizon of black holes .
direct signatures are unlikely to be observed in these cases .
however , gdr emission could lead to interesting predictions at theoretical level and possibly indirect observable effects .
we thank alikram aliev , luca bombelli and roy maartens for discussions and useful suggestions .
vc acknowledges financial support from fundao calouste gulbenkian through the programa gulbenkian de estmulo investigao cientfica .
r. maartens , living rev .
* 7 * , 7 ( 2004 ) .
l. randall and r. sundrum , phys .
lett . * 83 * , 3370 ( 1999 ) .
m. bando , t. kugo , t. noguchi and k. yoshioka , phys .
lett . * 83 * , 3601 ( 1999 ) [ arxiv : hep - ph/9906549 ] .
v. cardoso , o. j. c. dias and j. p. s. lemos , phys .
d * 67 * , 064026 ( 2003 ) .
a. o. barvinsky and s. n. solodukhin , nucl .
b * 675 * , 159 ( 2003 ) . c. w. misner _
et al . _ ,
* 28 * ( 1972 ) 998 .
h. soodak and m.s .
tiersten am .
j. phys .
* 61 * , 395 ( 1993 ) ; a. a. ungar , am .
j. phys .
* 62 * , 279 ( 1994 ) .
y. shibata _ et al .
_ , phys .
e * 57 * , 1061 ( 1998 ) ; o. haeberl , `` electromagnetic radiation generated by relativistic electrons interacting with a diffraction grating '' , phd thesis , universit louis pasteur , 1994 ( unpublished ) . n. arkani - hamed , s. dimopoulos and g. r. dvali , phys .
b * 429 * , 263 ( 1998 ) ; i. antoniadis , n. arkani - hamed , s. dimopoulos and g. r. dvali , phys .
b * 436 * , 257 ( 1998 ) .
a. v. olinto , phys .
rept . * 333 * , 329 ( 2000 ) ; m. nagano and a. a. watson , rev .
phys . * 72 * , 689 ( 2000 ) .
j. khoury , b. a. ovrut , p. j. steinhardt and n. turok , phys .
d * 64 * , 123522 ( 2001 ) ; p. j. steinhardt and n. turok , phys .
d * 65 * , 126003 ( 2002 ) ; r. kallosh , l. kofman and a. d. linde , phys .
d * 64 * , 123523 ( 2001 ) .
aliev and d. v. galtsov , annals of physics * 193 * , 142 ( 1989 ) . | we show that if the visible universe is a membrane embedded in a higher - dimensional space , particles in uniform motion radiate gravitational waves because of spacetime lumpiness .
this phenomenon is analogous to the electromagnetic diffraction radiation of a charge moving near to a metallic grating . in the gravitational case ,
the role of the metallic grating is played by the inhomogeneities of the extra - dimensional space , such as a hidden brane .
we derive a general formula for gravitational diffraction radiation and apply it to a higher - dimensional scenario with flat compact extra dimensions .
gravitational diffraction radiation may carry away a significant portion of the particle s initial energy .
this allows to set stringent limits on the scale of brane perturbations .
physical effects of gravitational diffraction radiation are briefly discussed . |
results from egret revealed a very strong link between @xmath0-ray emission in agn and core - dominated radio emission from relativistic jets , which may originate from super - massive black hole and accretion disk systems @xcite .
recent _ fermi _ data have confirmed this @xcite .
observations at radio wavelengths were critically important for the understanding of the @xmath0-ray emission mechanism in these agn .
in particular , the relative timing of @xmath0-ray flares and radio outbursts provides constraints on the mechanisms and location of the flare origin @xcite .
the main models for @xmath0-ray emission entail inverse compton scattering of low energy photons to high energies by the relativistic electron population in the agn jets .
the low energy photons could have their origin external to the jet , perhaps in the broad line region of the agn @xcite or could originate internal to the jet , being the synchrotron radio / ir / optical emission from the electrons themselves interacting with the jet magnetic field @xcite .
radio flux density monitoring observations , which allow for the evolution of the radio emission to be followed as a function of time , may indirectly reveal when relativistic electrons are being injected into the agn jet .
observations at high frequencies , in the optically - thin part of the radio spectrum , may allow us to probe deep into the structure of the radio jet , to where the @xmath0-ray emission is likely generated @xcite .
since late 2007 , the atca has undertaken a project ( designated c1730 ) to regularly observe a number of southern agn that have been detected in @xmath0-rays . for many sources , particularly those south of @xmath1 declination ,
this is the only project providing flux density monitoring data .
it measures flux densities for these agn at frequencies between 4.5 to 41 ghz , making it ideal for investigating the physics involved in @xmath0-ray emission .
the atca is ideally suited for a program such as c1730 .
it is a six - element interferometer located near the town of narrabri in new south wales .
each antenna has a 22-m diameter , with a cassegrain focus .
each antenna has a suite of cryogenically - cooled receivers , which cover 67% of the frequency range between 1.1 - 50 ghz .
the telescope can quickly change between these receivers to make observations at a range of different frequencies within a short timespan .
the c1730 project uses this ability to measure the spectral index of each source it monitors , by obtaining flux densities at 5.5 , 9 , 17 , 19 , 38 and 40 ghz almost simultaneously
. a typical c1730 epoch will observe 5 closely - spaced sources at two simultaneous frequencies in 30 minutes , before switching frequencies and reobserving them again , resulting in full spectral coverage within 2 hours . between the start of the c1730 project in november 2007 and february 2009
, observations were made with the original atca correlator , which provided 256 mhz of bandwidth , with 128 mhz at each of two simultaneously measured frequencies . in march 2009
, the atca correlator was upgraded to the compact array broadband backend ( cabb , @xcite ) , which provides 4 ghz of bandwidth , with 2 ghz at each of two simultaneous frequencies .
the atca changes the locations of its antenna several times each year .
five of the six antenna are set on a 3 km long east - west track , while the sixth is held stationary 3 km west of the western end of this track .
this allows for configurations with baseline lengths between 31 m and 6 km .
antenna can also be placed on a short ( 214 m ) north - south track placed roughly half - way along the east - west track , and this is done when high brightness - sensitivity is required .
the c1730 project has been scheduled in various array configurations , as the sources that it observes are usually point - like .
details of the c1730 observing epochs to - date are listed in table [ epochlist ] .
the project commenced with a pilot study before _
fermi _ was launched , to trial monitoring of 16 southern agn detected by egret .
this successful pilot study paved the way for continued observations , and with the emergence of the first results from _ fermi _ in early 2009 , the source list was quadrupled . in 2011 ,
the source list was revised based on the first and second _ fermi _ catalogs of detected agn , and the expanded list of tanami targets .
the original declination limit of @xmath2 was changed to better match the tanami range ( @xmath3 ) and reflect the larger number of southern sources now identified by _
the most recent epochs included 150 2lac sources with @xmath4 , including @xmath5 proposed counterparts ( primarily x - ray associations ) that had not previously been detected at radio frequencies .
.c1730 epoch list . for each epoch , the number of sources observed is shown , along with the atca antenna configuration at the time . details of each configuration can be found from the atca website . [ cols="<,^,^,<",options="header " , ] our c1730 observations will contribute significantly toward the identification of the physics of extragalactic _ fermi _ sources , along with monitoring programs at other frequencies , for example : * `` fermi large area telescope observations of the active galaxy 4c + 55.17 : steady , hard gamma - ray emission and its implications '' @xcite * `` fermi large area telescope observations of markarian 421 : the missing piece of its spectral energy distribution '' @xcite * `` the first fermi multifrequency campaign on bl lacertae : characterizing the low - activity state of the eponymous blazar '' @xcite * `` insights into the high - energy gamma - ray emission of markarian 501 from extensive multifrequency observations in the fermi era '' @xcite c1730 observations are also occassionally made to rapidly follow - up sources that are reported to be flaring by _ fermi _
, for example atels 3788 , 3734 , 3713 , 3703 , 3579 and 3394 . in this way
, we have a higher probability of observing `` interesting '' agn , and thus examining their temporal behaviour during flare events . with the sensitivity that cabb provides with its 4 ghz of bandwidth ,
accurate flux density measurements are possible even for objects that are not normally very bright .
for example , in 20 minutes of on - source integration time , the atca with cabb reaches an rms noise level of 30 ujy at 5.5 ghz , 40 ujy at 9 , 17 and 19 ghz , and 70 ujy at 38 and 40 ghz .
it is therefore possible to follow - up interesting flare events during periods of unscheduled time as small as 2 hours .
all data from c1730 is reduced through the same custom pipeline , ensuring consistency over multiple epochs .
flux density calibration is achieved via the atca primary flux density calibrators pks b1934 - 638 for frequencies lower than 25 ghz and the planet uranus for higher frequencies .
early analysis has focussed on determining which of the agn monitored by this program have varied , and how . using the method described in @xcite ,
we have found 107 sources that have varied between 2007 and now .
the spectral curves for these sources are shown in figure [ allspecvar ] .
the variability is characterised by a change in the spectral turnover frequency .
for sources that vary in brightness achromatically , the spectral turnover frequency will remain constant , but if , for example , the source becomes disproportionately brighter at higher frequencies , the spectral turnover may shift toward the higher frequency end .
we used the roma - bzcat @xcite , cgrabs @xcite and 1lac @xcite catalogues of blazars to separate the agn in this survey into bl lacs , flat - spectrum radio quasars ( fsrq ) and unclassified blazars .
sources that do not appear in any of these catalogues are listed here as `` unclassified '' .
fifty - nine percent of our sources are blazars , while most others are flat - spectrum radio sources ( fsrs , @xcite ) . from our survey ,
72% of the bl lacs have shown spectral - index variability , along with 87% of the fsrqs and 80% of the unclassified blazars , while 89% of the unclassified sources also show spectral variation . of these variable sources ,
97% are seen to have a spectral turnover frequency at the very highest observed frequencies in at least one epoch .
this indicates that the vast majority of our variable sources become periodically brighter at high frequencies , but do not always stay that way .
further work is required to determine whether these high - frequency flux increases are associated with @xmath0-ray activity .
observations at the atca continue on a regular basis , and analysis of the bulk of the data has yet to be completed .
this section describes some of the analysis we plan to do in the near future .
the atca receivers are capable of measuring polarisation accurately at all the frequencies observed by this project
. it will therefore be possible to examine the polarisation position angle variability for these blazars , and determine if the magnetic field direction is changing when the spectral index varies .
the atca monitoring data provides information crucial in constructing the spectral energy distributions ( seds ) of agn both in their flaring and quiescent states .
quasi - simultaneous broadband seds are necessary to distinguish between competing models for the high - energy emission from agn @xcite .
the atca data are also being used to look for changes in spectral index associated with the ejection of vlbi components from the agn cores .
the atca has monitored a set of 133 agn over multiple epochs for four years , and will continue to monitor even more in the future . through a combination of regular monitoring and target - of - opportunity observations of flaring agn , we will investigate the mechanisms responsible for both radio and @xmath0-ray emission in their jets . | as a critical part of the tracking active galactic nuclei with austral milliarcsecond interferometry ( tanami ) program @xcite , in november 2007 the australia telescope compact array ( atca ) started monitoring the radio spectra of a sample of southern hemisphere active galactic nuclei ( agn ) that were selected as likely candidates for detection ( as well as a control sample ) by the large area telescope ( lat ) aboard the _ fermi gamma ray space observatory_. the initial sample was chosen based on properties determined from agn detections by the energetic gamma ray experiment telescope ( egret ) .
most of the initial sample has been detected by _
fermi_/lat and with the addition of new detections the sample has grown to include 226 agn , 133 of which have data for more than one epoch . for the majority of these agn , our monitoring program provides the only dynamic radio spectra available .
the atca receiver suite makes it possible to observe several sources at frequencies between 4.5 and 41 ghz in a few hours , resulting in an excellent measure of spectral index at each epoch . by examining how the spectral index changes over time , we aim to investigate the mechanics of radio and @xmath0-ray emission from agn jets . |
as presented in ref .
@xcite , a periodically - dressed bose - einstein condensate ( pdbec ) is formed of atoms of mass @xmath3 with two internal states , @xmath4 and @xmath5 , separated by an energy difference @xmath6 . these atoms are exposed continuously to two laser beams of wavevectors @xmath7 and @xmath8 and frequencies @xmath9 and @xmath10 , which couple the states @xmath4 and @xmath5 via a raman transition ( see figure [ fig : scheme ] ) .
in such a raman transition , an atom in internal state @xmath4 with wavevector @xmath11 absorbs a photon from beam @xmath12 and emits a photon into beam @xmath13 , arriving in internal state @xmath5 with wavevector @xmath14 gaining a momentum @xmath15 and a kinetic energy @xmath16 . and
@xmath10 induce raman transitions between internal states @xmath17 and @xmath18 . the beams share a common large detuning @xmath19 from the excited atomic state , and this state is adiabatically eliminated in the theoretical treatment .
the lasers are detuned by an amount @xmath20 from the two - photon raman transition .
( b ) such a raman transition imparts a momentum transfer of @xmath21 , where @xmath7 and @xmath8 are the wavevectors of the raman coupling lasers .
, width=288 ] keeping in mind this coupling between atoms of different momenta , we let @xmath22 ( @xmath23 ) be the creation ( annihilation ) operator for an atom in internal state @xmath4 and wavevector @xmath11 , and similarly let @xmath24 ( @xmath25 ) be the creation ( annihilation ) operator for an atom in internal state @xmath5 and wavevector @xmath14 . we refer to these as the bare - state creation and annihilation operators .
the justification for this particular notation lies in the dressed - state picture @xcite , in which one considers the driving optical fields to be part of the quantum system . in this picture , the raman transition connects the states @xmath26 and @xmath27 , where @xmath28 represents an atom in state @xmath4 and wavevector @xmath11 and an optical field with @xmath29 photons in beam @xmath12 and @xmath30 photons in beam @xmath13 , and similarly for @xmath31 . in the dressed - state picture ,
@xmath32 and @xmath31 are states for which the total momentum of the atom + photon system is equal by a proper choice of intertial frame , this total momentum is taken to be @xmath33 ) .
the states are separated by an energy difference @xmath34 , and coupled by a static and ( in the dressed - state picture ) spatially uniform raman coupling .
the system is then represented by a many - body hamiltonian of the form @xmath35\ ] ] here @xmath36 is the two - photon rabi frequency characterizing the strength of the raman transition , determined by the polarizations and intensities of the driving laser beams and by dipole matrix elements of the atoms .
this hamiltonian is diagonalized by transforming to the dressed - state creation and annihilation operators @xmath37 , @xmath38 , @xmath39 , @xmath40 which are defined as linear combinations of the bare - state operators @xmath41 the mixing angle @xmath42 is defined by the relation @xmath43 .
the dressed states have energies @xmath44 here , the minus sign refers to the lower dressed - state dispersion curve and to the operators @xmath37 and @xmath38 , while the plus sign refers to the upper dressed - state dispersion curve and to the operators @xmath39 and @xmath40 .
a bose - einstein condensate formed from this periodically - dressed atomic gas has a macroscopic population of @xmath45 atoms in the lowest energy momentum state : this state lies on the lower dressed - state dispersion curve and is taken to have momentum @xmath46 @xcite . in ref.@xcite , a bogoliubov - approximation theory was developed to account for the effects of weak interatomic interactions on a pdbec .
an interaction hamiltonian of the form @xmath47 was considered , where @xmath48 is the total number of atoms in the system and @xmath49 is the fourier transform of the density operator .
quasi - particle energies and operators were found by diagonalizing a @xmath50 matrix @xmath51 where @xmath52 is a diagonal matrix with entries @xmath53 with @xmath54 , @xmath55 is the chemical potential , and @xmath56 , where @xmath57 .
diagonalization yields @xmath58 where @xmath59 is a diagonal matrix with entries @xmath60 by which the lower ( minus sign ) and upper ( plus sign ) quasi - particle energies @xmath61 are defined .
one also obtains the creation and annihilation operators for the lower ( @xmath62 , @xmath63 ) and upper ( @xmath64 , @xmath65 ) branch quasi - particles through the following relation : @xmath66 here @xmath67 contains trigonometric functions of @xmath68 and @xmath69 in accordance with eq.[eq : rotation ] .
the bose commutation relations take the form @xmath70 = a_{i j } = \left\ { \begin{array}{c l } 0 & i \neq j \\ 1 & i = j = 1 \mbox { or } i = j = 3 \\ -1 & i = j = 2 \mbox { or } i = j = 4 \end{array } \right.\ ] ] where @xmath71 represents any of the four - component vectors in eq.[eq : baretoquasi ] .
these relations are clearly preserved by the rotation matrix , i.e. @xmath72 , while the imposition of bose commutation relations for the quasi - particle operators enforces a normalization criterion @xmath73 .
experimentally , values for the raman detuning @xmath20 , the rabi frequency @xmath36 , the chemical potential @xmath74 and the raman momentum transfer @xmath75 can be chosen over a wide range of parameters .
for example , let us consider a particular realization of a pdbec using two hyperfine states of @xmath1rb .
one may consider the case of raman excitation using two lasers both tuned near the @xmath76 optical transition at a wavelength @xmath77 nm .
thus , the raman recoil energy @xmath78 can be readily tuned in the range @xmath79 khz .
properties of a pdbec are governed by the ratios @xmath80 , @xmath81 and @xmath82 .
the raman detuning @xmath20 and the rabi frequency @xmath36 can be chosen nearly arbitrarily , reaching maximum values which are orders of magnitude larger than @xmath83 .
the chemical potential @xmath84 has a typical value of 1 khz for a density of @xmath85 and a scattering length @xmath86 bohr thus the ratio @xmath87 can be as small as @xmath88 for counter - propagating raman excitation lasers , and greatly increased by using raman beams intersecting at small angles . for illustration , we show in figure [ fig : quasidisprelations ] the quasiparticle dispersion relations for a pdbec with a chemical potential @xmath89 , a raman detuning of of @xmath90 and raman coupling strength @xmath91 varying between 0 and 2 . for this case of @xmath92 ,
a condensate is formed in the @xmath93 state of the lower dispersion relation . in the absence of raman coupling , this equality is exact , and the condensate is formed in the @xmath94 internal state .
quasiparticle excitations in the @xmath94 state follow the standard bogoliubov result , while quasiparticles in state @xmath95 have a free - particle dispersion relation this can be though of as due to the lack of a spatial interference term between the condensate and excitations in the @xmath95 state at any wavevector .
for small rabi coupling strengths , one speaks more properly of excitations in the lower and the upper quasiparticle dispersion relations , although the energies and internal - state compositions of these states remain relatively unchanged for wavevectors @xmath96 far from the doppler - shifted raman resonance condition @xmath97 .
the lower quasiparticle dispersion relation has a local minimum near @xmath98 caused by the minimum of the dispersion relation for the bare @xmath95 state atoms . by analogy to the roton minimum in the excitation spectrum of superfluid @xmath0he @xcite
, we denote this feature as an `` artificial roton , '' although we mean to suggest no similarity between the specific structure of a roton and of the plane - wave excitations of a pdbec . as the rabi frequency
is increased further , a greater mixing of the internal state composition of the bose condensate and of the quasiparticle excitations occurs , and energy levels are shifted . in particular ,
the energy of the `` artificial roton '' is shifted upwards due to interactions as the internal state composition of the bose condensate and of these excitations becomes more similar . ) and upper ( created by @xmath99 ) dispersion curves with momentum @xmath100 with respect to the condensate .
excitations parallel to the raman momentum transfer are considered , with the definition @xmath101 .
black curves represent pdbec s with @xmath102 , @xmath103 and @xmath104 ( dashed ) , 1 ( dotted ) and 2 ( solid ) .
the gray dot - dashed curve shows the bogoliubov excitation spectrum for a single component condensate with @xmath103 .
, width=288 ]
we now consider the scattering of impurities passing through a uniform pdbec .
we consider impurities of mass @xmath105 which have a free - particle dispersion relation @xmath106 where @xmath107 is the impurity wavevector .
such impurities interact with the two components of the periodically - dressed condensate through an interaction of the form @xmath108 here @xmath22 , @xmath23 , @xmath24 , and @xmath25 are defined as above , and @xmath109 and @xmath110 are the creation and annihilation operators , respectively , for the impurity particles .
the coupling strength for the impurity scattering is @xmath111 , which we parameterize as @xmath112 .
this parameterization correctly describes low - energy @xmath113-wave scattering with a scattering length of @xmath114 and the reduced mass @xmath115 .
note that we have made the assumption that the scattering strengths between the impurity and atoms in states @xmath4 and @xmath5 are equal , and assume further that state changing collisions do not occur ( the impurity is considered `` non - magnetic '' ) .
we may consider also the limit as the impurity mass becomes infinite while the impurity velocity @xmath116 remains constant ( @xmath117 is the impurity wavevector ) .
we thereby obtain a description of the pdbec flowing past a rigid obstacle with a velocity @xmath118 , and the concept of impurity scattering is replaced by the dissipation of superfluid flow .
the scattering rate of these impurities can be calculated using a perturbative approach as follows @xcite .
we consider the system to be initially in the state latexmath:[$ |0\rangle_{ab } impurity particle of wavevector @xmath117 .
the scattering hamiltonian of eq .
[ eq : hscat ] may couple this state to the final state @xmath120 in which the periodically - dressed gas is in the state @xmath121 and the impurity wavevector changes from @xmath117 to @xmath122 .
the matrix element for this transition is @xmath123 , where @xmath124 is described above , and where the @xmath125 subscript labeling the quantum states of the periodically - dressed gas has been dropped .
we then obtain the rate of impurity scattering by summing over all final states @xmath121 and momentum transfers @xmath100 using fermi s golden rule .
we now make use of the bogoliubov approximation .
if we consider only weak scattering , we may restrict the final states to those containing a single quasiparticle , i.e. @xmath126 and @xmath127 .
the rate of impurity scattering is then @xmath128 \label{eq : gammaeq}\ ] ] where @xmath129 is the impurity velocity .
we have introduced the quantities @xmath130 and @xmath131 which relate to the static structure factor describing density - density correlations in the bose condensate . for a single - component bose - einstein condensate ,
the static structure factor @xmath132 has been evaluated and measured @xcite .
for the pdbec , we separate the contributions of the lower and upper excitation spectra as @xmath133 physically , these quantities describe two effects . first , the magnitude of @xmath134 and @xmath135 describes the contributions of quasiparticles in the lower ( @xmath74 ) or upper ( @xmath136 ) quasiparticle state to density fluctuations ( as opposed to phase fluctuations , which dominate , for example , for phonon excitations of a scalar bose condensate @xcite ) .
second , these quantities describe the degree to which the internal state composition of the condensate matches that of the quasiparticle . for instance
, a condensate which is predominantly in the @xmath94 internal state will not scatter strongly into quasiparticle states in which the @xmath95 internal state is dominant , and thus the structure factor describing this scattering will be small . before proceeding to a calculation of @xmath134 and @xmath135 ,
it is useful to consider the scattering rate @xmath137 for an ideal - gas , single - component bose - einstein condensate as the target gas . in this case
, we may conveniently take @xmath138 , @xmath139 and @xmath140 .
replacing @xmath141 , and substituting for @xmath111 we evaluate eq .
[ eq : gammaeq ] and obtain simply @xmath142 i.e. the conventional expression for the scattering rate where @xmath143 is the density of the target gas and @xmath144 is the collision cross section .
we may then determine the impurity scattering rate in a pdbec relative to that in an ideal gas , @xmath145 : @xmath146
to calculate the structure factors @xmath134 and @xmath135 we start by expressing the density operator @xmath124 in terms of the quasiparticle operators @xmath147 acting on a pdbec , the only important terms in the sum are those which include the condensate operators @xmath148 and @xmath149 . applying the relations @xmath150 for @xmath151 , and the relation @xmath152 , we define matrices @xmath153 and @xmath154 and obtain @xmath155 so @xmath156 the static structure factors @xmath157 and @xmath131 for a pdbec are shown in figures [ fig : smuandspivaryrabi ] and [ fig : smuandspivarymu ] . the specific values for the raman detuning @xmath20 , the rabi frequency @xmath36 and the chemical potential @xmath74 are relevant for a particular realization of a pdbec using two hyperfine states of a @xmath1rb , as discussed above .
the division of the scattering weight between the lower ( @xmath134 ) and upper ( @xmath135 ) quasiparticle branches can be understood by considering the composition of the dressed - state quasiparticles , which is related by eq .
[ eq : rotation ] to the mixing angle @xmath158 .
this angle ranges from @xmath159 for @xmath160 to @xmath161 for @xmath162 , reaching @xmath163 at the doppler - shifted raman resonance @xmath164 . for small rabi frequency , @xmath158 , and hence the internal state composition of the dressed states varies rapidly about the raman resonance condition , while for large rabi frequency , this variation occurs over a broader range of @xmath165 .
this behaviour is reflected in the structure factors , as illustrated by figure [ fig : smuandspivaryrabi ] . by continuity ,
the internal state composition of the condensate in the @xmath166 state is nearly identical to that of excitations in the lower quasiparticle branch ( created by @xmath167 ) , and nearly orthogonal to that of excitations in the upper quasiparticle branch ( created by @xmath168 ) .
thus , since elastic scattering from a `` non - magnetic '' impurity does not affect the internal state distribution , @xmath169 generally for small @xmath165 . away from @xmath170 ,
the division of the scattering weight between the upper and lower quasiparticle branches depends on the strength of the rabi coupling . for small rabi frequencies , the scattering strength shifts abruptly from the lower to the upper quasiparticle branch at the doppler - shifted raman resonance . for large rabi frequency , the changes in the scattering strengths are more gradual , with scattering into both branches showing a significant weight for a wide range about the raman resonance condition .
this variation is particularly relevant for the scattering of impurities near the landau critical velocity , as discussed below . and ( b ) @xmath131 with the raman coupling strength @xmath36
. the wavevector @xmath100 is colinear with the raman momentum transfer @xmath117 with @xmath171 .
three black curves describe a pdbec with @xmath172 , @xmath89 , and @xmath173 ( solid ) , @xmath174 ( dashed ) and @xmath12 ( dotted ) .
scattering for @xmath175 and @xmath176 occurs only into the lower quasiparticle branch and for @xmath177 into the upper quasiparticle branch .
in other regions , stronger raman coupling leads to eigenstates with mixed populations in the internal states @xmath95 and @xmath94 over a broader range of @xmath165 , and thus impurity scattering occurs into both the lower and upper quasiparticle branches . also shown is the structure factor @xmath178 for a scalar bose - einstein condensate ( gray dot - dashed curve ) with @xmath89 . ,
width=480 ] and ( b ) @xmath131 for a pdbec . the wavevector @xmath100 is colinear with the raman momentum transfer @xmath117 with @xmath171 .
three black curves describe a pdbec with @xmath172 , @xmath179 , and @xmath180 ( solid ) , @xmath174 ( dashed ) and @xmath12 ( dotted ) .
the condensate in the @xmath166 state is predominantly in the internal state @xmath5 ; thus scattering predominantly occurs into the lower quasiparticle branch for @xmath176 and into the upper quasi - particle branch for large @xmath181 . for @xmath182 both @xmath157 and @xmath131 are significant since quasiparticles contain large fractions of both @xmath95 and @xmath94 .
interactions suppress scattering at small momentum transfer , an effect seen also in the structure factor @xmath178 for a scalar bose - einstein condensate ( gray dot - dashed curve ) with @xmath89 .
, width=480 ] as illustrated in figure [ fig : smuandspivarymu ] , interatomic interactions suppress scattering at small momentum transfers . for a single - component homogeneous bose - einstein condensate ,
the structure factor @xmath183 is small for wavevectors of magnitude @xmath184 where @xmath185 is the healing length .
similarly , for the pdbec , we find a suppression of the structure factors @xmath157 and @xmath131 for @xmath186 , as shown in the figure .
this suppression of the structure factor in a pdbec could be probed experimentally as it was for a scalar bose - einstein condensate by studying shallow - angle optical bragg scattering @xcite .
finally , let us use the formalism developed above to describe superfluid properties of a pdbec . in ref .
@xcite , we applied the landau criterion to determine the critical velocity for superfluidity in a pdbec given the quasiparticle dispersion relations .
this criterion gives the superfluid critical velocity along a direction @xmath187 as having a magnitude @xmath188 where @xmath189 are the energies of possible excitations with wavevector @xmath96 ; using the momentum definitions for a pdbec , these would be the quasiparticle energies @xmath190 . in an unperturbed , single component weakly - interacting condensate ,
this critical velocity @xmath191 is equal to the bogoliubov speed of sound @xmath192 in all directions . in a pdbec
, the anisotropic quasiparticle dispersion relation leads directly to the prediction of an anisotropic superfluid critical velocity , and to the possibility of a critical velocity lower than that of the unperturbed condensate .
one may determine this critical velocity graphically by finding the line of minimum absolute slope which connects the origin and the quasiparticle dispersion curve ( see figure [ fig : quasidisprelations ] ) .
for the pdbec , the superfluid critical velocity may be suppressed by the presence of the `` artificial roton '' feature in the lower dispersion curve .
this suppression occurs in the direction of the raman momentum transfer @xmath193 for the case of @xmath92 ( as shown in the figure ) , or in the opposite direction for @xmath194
. however , while the landau criterion determines the minimum velocity for the dissipation of superfluid flow , it does not quantify the onset of such dissipation as the velocity is increased beyond the critical value . indeed , for extremely weak raman coupling , even though the superfluid critical velocity is strictly different
, one should expect the fluid properties of a pdbec to be nearly identical to that of a two - component bec in the complete absence of raman coupling .
we can quantitatively assess this description using the formalism developed above for the determination of the matrix elements @xmath134 and @xmath134 , and therefrom the normalized scattering rates @xmath195 and @xmath196 .
the results of such calculations for infinitely massive impurities ( i.e. superfluid flow about rigid obstacles ) are presented in figures [ fig : allpdplot ] and [ fig : onepdplot ] .
values of @xmath195 and @xmath196 were determined by numerical integration in which the delta function in eqs .
[ eq : fmuexpression ] and [ eq : fpiexpression ] was approximated by a narrow gaussian distribution whose width was adjusted to give reasonable numerical convergence . as we expect , for small rabi frequencies @xmath36 the properties of a periodically - dressed bec are quite similar to those of an uncoupled two component bose - einstein condensate .
the scattering probability for impurity atoms ( related to the viscosity of the fluid ) is indeed non - zero at velocities lower than the superfluid critical velocity for the unperturbed gas , demonstrating that the critical velocity is suppressed along the direction of raman scattering .
this scattering at low velocities occurs through the excitation of quasiparticles near the local minimum ( the `` artificial roton '' ) in the lower dispersion relation .
however , this scattering channel is weak for small values of the rabi frequency @xmath36 , and thus the scattering strength @xmath197 remains weak until the curve joins that of the uncoupled condensate and the scattering strength is increased by phonon scattering .
the effects of increasing the rabi frequency are twofold .
first , as @xmath36 increases , the internal state composition of the `` artificial roton '' excitation becomes more similar to the internal state composition of the bose - einstein condensate .
this causes an increase in the scattering strength ( as indicated also in the calculations of @xmath134 shown in figures [ fig : smuandspivarymu ] and [ fig : smuandspivaryrabi ] ) .
simultaneously , because of this overlap of the internal state composition , the energy of these excitations is increased , approaching the energy given by the one - component bogoliubov spectrum , albeit with an appropriately modified effective mass .
thus , the minimum velocity at which scattering occurs increases as the rabi frequency increases . at moderate values of the rabi frequency ( where @xmath198 )
, the scattering rate vs. velocity has a two - step shape as scattering is first enhanced by strong scattering into the `` artificial roton '' state , and then later enhanced by phonon scattering when the velocity is near @xmath191 .
as shown in figure [ fig : onepdplot ] for one typical parameter setting , only @xmath195 contributes to the weak scattering at velocities smaller than @xmath191 . at velocities greater than @xmath191 , scattering into the upper dispersion curve
is allowed and @xmath199 . for infinitely massive impurities traveling through the condensate , normalized to the scattering in an ideal gas bose - condensate at similar densities , is plotted vs. the velocity of the impurity .
here we consider velocities parallel to the raman momentum transfer , with @xmath200 .
black curves show results for pdbec s with varying strengths of the raman coupling ; for all curves , @xmath201 and @xmath89 while @xmath202 ( dashed ) , 1 ( solid ) or 2 ( dotted ) . the gray dot - dashed curve shows results for a one - component condensate with @xmath89 . the superfluid critical velocity of the pdbec in the direction of the raman momentum transfer is suppressed with respect to an uncoupled condensate . for weakly raman coupling ,
the dissipation at low velocities is weak .
this dissipation grows stronger as the raman coupling strength is increased . , width=288 ] , dashed line ) and upper ( @xmath196 , dotted line ) quasiparticle excitations due to scattering off infinitely massive impurities moving at velocities @xmath203 in a pdbec .
velocities parallel to the raman momentum transfer are considered , with @xmath200 .
black curves show results for a pdbec with @xmath102 , @xmath89 , and @xmath204 , with the solid curve showing the total normalized scattering rate @xmath197 .
the gray dot - dashed curve shows the normalized scattering rate for a one - component condensate with @xmath89 . for @xmath205 ( for this case where @xmath206 ) , impurity scattering at low velocities is enhanced as
compared to the one - component condensate .
this scattering leads to the creation of excitations in the lower dispersion curve ( @xmath207 ) around the `` artificial roton '' feature in the quasiparticle dispersion relations . at velocities near the critical velocity for the unperturbed condensate , @xmath191
, scattering is allowed into the upper dispersion relation @xmath208 , and the normalized scattering rate approaches that of a one - component condensate . for @xmath209 , energy and momentum conservation imply that only scattering into the lower quasiparticle excitations is allowed at the velocities considered .
, width=288 ] this two - step curve gives a clear signature which should be visible in experimental probes of superfluid flow .
indeed , experiments such as those performed by chikkatur _ _ e__t al @xcite , in which atoms of another ground state hyperfine level were used as scattering impurities in a dilute atomic bose condensate , could be carried out in a straightforward manner .
it would be interesting to consider further how other aspects of superfluidity , such as the onset of turbulent flow @xcite or the metastability of quantized vortices @xcite , are affected by the anisotropic nature of a periodically - dressed bose - einstein condensate . in conclusion , dissipation of superfluid motion in a periodically - dressed bose - einstein condensate above the superfluid critical velocity
was treated .
a formalism was developed for the calculation of the static structure factors @xmath210 and @xmath211 describing density - density correlations from quasiparticles in the lower ( @xmath74 ) and upper ( @xmath136 ) excitation branches of the periodically coupled gas . following a perturbation theory approach , the scattering rate of massive impurities , or equivalently the damping rate of superfluid motion past microscopic obstacles , was calculated .
raman coupling of a two - component condensed gas leads to a suppression of the superfluid critical velocity .
the impurity scattering rate evolves smoothly as the raman coupling strength is increased .
the the superfluid damping is weak for weak raman coupling , and becomes more significant as the raman coupling strength is increased .
cornell , j.r .
ensher , and c.e .
wieman , in _ bose - einstein condensation in atomic gases _ , _ proceedings of the international school of physics `` enrico fermi , '' course cxl _ , edited by m. inguscio , s. stringari , and c.e .
wieman ( ios press , amsterdam , 1999 ) .
we do not consider here the possibility of a degenerate or near - degenerate ground state .
proper treatment of a periodically - dressed bose - einstein condensate for such a situation requires a many - body treatment beyond the bogoliubov approximation , as will be discussed elsewhere . | the introduction of a steady - state spatially - periodic raman coupling between two components of an ultracold atomic gas produces a dressed - state gas with an anisotropic and tunable dispersion relation . a bose - einstein condensate formed in such a gas is consequently characterized by an anisotropic superfluid critical velocity . the anisotropic dissipation of superfluid flow is quantified by considering the scattering of impurities flowing through this superfluid
a gradual transition from the isotropic nature of an uncoupled bose - einstein condensate to the anisotropic periodically - dressed condensate is obtained as the strength of the raman coupling is varied .
these results present a clear signature for future experiemental realizations of this novel superfluid .
the experimental attainment of quantum degenerate dilute gases , composed of both bosonic and fermionic atoms , has created many new opportunities in the study of quantum fluids . in particular , recent years have seen a flurry of experimental studies of superfluidity in dilute , scalar bose - einstein condensates .
these include the observations of critical velocities for superfluid flow about microscopic @xcite and macroscopic @xcite obstacles , the onset of turbulent flow above this critical velocity @xcite , beautiful studies of quantized vortices and vortex lattices in rotating bose - einstein condensates @xcite , and other manifestations of irrotational superfluid flow @xcite .
these experiments are closely analogous to those performed on another scalar superfluid , liquid @xmath0he , but apply new experimental probes and allow new insights due to the vastly different parameter regime of the dilute atomic gases , and their amenability to a new set of tools for manipulation and probing .
dilute atomic gases also offer the opportunity to create novel quantum fluids by using these various tools to manipulate the internal and external states of the atoms comprising the superfluid .
for example , different types of multi - component condensates have been studied : externally - coupled two - component condensates of @xmath1rb at jila ( reviewed in @xcite ) , and @xmath2 spinor condensates of sodium at mit ( reviewed in @xcite ) .
one experiment at jila explored how a spatially - selective coupling between two trapped hyperfine levels of a rubidium condensate leads to a `` winding '' of the order parameter , akin to the phase winding of the order parameter which occurs due to rotation about a vortex core , and how this winding is recurrently `` undone '' due to the motional dynamics of a two - component condensate @xcite .
this observation illustrates that vortices in such a novel system should not be metastable ( similarly predicted for spinor condensates @xcite ) , a major modification from the superfluid behaviour of a single component gas . in a similar vein
, we have previously considered the novel superfluid properties introduced to a two - component bose condensed atomic gas which is coupled by a spatially periodic coupling field @xcite .
this spatially - periodic coupling results from intersecting , non - collinear laser beams which induce a raman coupling between two internal states of the ultracold gas .
the raman coupling introduces an anisotropy to this periodically - dressed fluid which should manifest itself in an anisotropic superfluid critical velocity .
controlling various parameters of the raman coupling laser beams allows one to dramatically alter this critical velocity , leading to the dissipation of superfluid flow at velocities much lower than the critical velocity for the uncoupled bose condensate . in this paper
, we develop further a theory describing the dissipation of superfluid flow in a periodically - dressed bose - einstein condensate . using the bogolibov approximation theory which was introduced in ref .
@xcite , the possible disspation of superfluid flow is treated by analyzing the scattering of massive impurities which flow through the superfluid . in analogy with similar calculations for a scalar bose - einstein condensate @xcite
, impurity scattering at low impurity velocities is found to be suppressed by the energetics of the quasi - particle spectrum ( related to the landau criterion for superfluidity @xcite ) , by the suppression of density fluctuations in the low - momentum phonon regime @xcite , and by variations of the internal - state compositions of quasiparticles at different wavevectors .
these effects combine to give a smooth evolution in the disspation rates for superfluid flow as the raman coupling strength is increased , gradually converting an uncoupled two component bose - einstein condensate ( essentially a scalar superfluid ) to a periodically - dressed condensate with dramatically different superfluid properties .
this superfluid dissipation rate gives an appealingly clear signature for future experimental studies . |
several theories of particle physics as well as cosmology predict the existence of at least one sub
ev scalar , that is , spin - zero , boson @xcite .
correspondingly , many theories of physics beyond the standard model ( sm ) can accommodate scalars with very small masses and feeble couplings to sm fields @xcite .
an intriguing possibility in astrophysics and cosmology is that these weakly interacting , sub ev particles ( wisps ) may constitute at least some component of the cold dark matter in the universe @xcite .
it has been shown that these arguments apply for both pseudoscalar , namely axion @xcite , and scalar @xcite wisps , e.g. axion like particles ( alps ) that couple to two photons .
the current experimental programs that explore the parameter space of weakly interacting , light , spin - zero bosons include sensitive searches that make use of resonant cavities @xcite .
scalar alps may be detected by stimulation of the conversion to photons in a magnetic field , similar to the way that axion conversion may take place @xcite .
this conversion can occur in a suitable cavity that is properly tuned to the mass of the alp .
while it should be noted that scalar couplings to two photons are strongly excluded by fifth force experiments @xcite , it is possible that a low energy form factor could relax these constraints in some models @xcite .
the lagrangian that describes the coupling of a light scalar particle to two photons in the presence of an external magnetic field takes the form : @xmath3 where @xmath4 is the electromagnetic field strength tensor , @xmath5 is the strength of the coupling between the scalar particle @xmath6 and two photons , @xmath7 is the magnetic field of the incident photon , and @xmath8 is the external magnetic field . from this lagrangian ,
the equations of motion that result may be used to write the power @xmath9 that results from axion to photon conversion on resonance in the microwave cavity as @xcite @xmath10 here , @xmath11 is the cavity volume , @xmath12 is the form factor associated with the cavity mode , @xmath13 is the quality factor of the cavity , and @xmath14 is the assumed scalar alp density .
the results presented here are the first from a search for primordial scalar particles coupling to a strong external magnetic field using resonant cavities in the mass range of 0.14 mev corresponding to 34 ghz frequency .
the apparatus @xcite consists of a tunable , 34 ghz resonant cavity ( q@xmath1510@xmath16 ) made from oxygen - free copper and a high electron mobility transistor ( hemt ) cryogenic amplifier @xcite located at the bottom of a cold gas cryostat that is oriented vertically and cooled to approximately 4 k. the cavity and cryogenic amplifier are coupled by approximately 10 cm of wr28 waveguide .
the cryostat rests inside the vertical bore of a 7 t cryomagnet .
the temperature inside the cryostat is monitored at multiple locations with cryogenic thin film resistance temperature sensors .
the resonant cavity has one critically coupled and one weakly coupled port , each connected to wr28 waveguide .
the critically coupled waveguide terminates at the input to the hemt amplifier , after which both waveguides feed out of the cryostat .
the air in the cavity is pumped out through the weakly coupled waveguide .
the q of the cavity is measured prior to each data run using a network analyzer connected to both waveguides . the cavity is tuned with an adjustable plunger that is vacuum tight at 4 k. a fiberglass g-10 rod is threaded through the top of the cryostat and fastened to the tuning plunger with a horizontal lever .
when the g-10 rod is turned , the plunger moves vertically .
the range of tuning in the cavity is @xmath175 mm which corresponds to approximately @xmath170.8 ghz .
the signal from the cryogenic amplifier terminates outside the cryostat at a waveguide to coaxial adapter and triple heterodyne microwave receiver @xcite . after the receiver the voltages from the in phase and quadrature components
are digitized for the complex fast fourier transform ( fft ) and further analysis offline .
the sensitivity of the experiment is limited by the system noise temperature @xmath18 according to the dicke radiometer equation @xcite . @xmath19
typical power spectra measured with the cryogenic amplifier held at 6 k are shown in the left panel of figure [ fig : results ] .
the resonance of the cavity appears as a dip in the spectrum .
this dip is qualitatively consistent with the combination of two effects : a frequency - dependent reflection of noise power coming from the input of the hemt , and a change in the gain of the hemt as the impedance of the source decreases ( e.g. @xcite ) .
the width of the dip is approximately 3 - 4 mhz which is in accordance with the cavity s q of 10@xmath16 .
coupling from data similar to those shown in the left panel . in this
initial result the measurements are centered at 34.29 ghz ( 0.142 mev ) and span across the width of the last bandpass filter ( 30 mhz ) .
the gap in the center of the plot corresponds to frequencies at or near 0 hz in the baseband and have been excluded from this analysis.[fig : results],title="fig:",width=288 ] coupling from data similar to those shown in the left panel . in this initial
result the measurements are centered at 34.29 ghz ( 0.142 mev ) and span across the width of the last bandpass filter ( 30 mhz ) .
the gap in the center of the plot corresponds to frequencies at or near 0 hz in the baseband and have been excluded from this analysis.[fig : results],title="fig:",width=288 ] the expected power from couplings between a scalar alp and 2 photons is given in equation [ eq : power ] @xcite .
the cavity form factor @xmath12 , adapted from @xcite for the case of a scalar alp , is @xmath20 where @xmath21 is the oscillating magnetic field in the cavity , @xmath22 is the static magnetic field , and @xmath23 is the magnetic permeability . for the configuration used in these measurements
@xmath12=10@xmath24 is lower than would be desirable , but the cavity geometry was dictated by the constraints of a different experiment @xcite . using a limited set of measurements similar to those in the left panel of figure [ fig : results ] and assuming a primordial alp density @xmath14 of 10@xmath25/@xmath26 @xcite
, the right panel shows that we exclude couplings with g@xmath010@xmath1/gev between 0.14 mev scalar particles and two photons with 5@xmath2 confidence .
although the sensitivity of the present measurement does not exceed model - dependent limits on g@xmath27 set in previous searches for solar alps @xcite , by astrophysical observations @xcite , and by fifth force experiments @xcite , it is the first glimpse into this energy regime with a technique that has mass resolution .
it is also the first direct search for alps as cold dark matter at 0.1 mev .
immediate plans for the experiment include a cavity in a transverse magnetic mode which will allow coupling to pseudoscalar alps and will be 3 orders of magnitude more sensitive than the present measurement .
a wider mass range will also be searched by tuning the receiver . in conclusion
although this measurement is primarily a first step toward the goal of a more sensitive experiment , it is still an unprecedented , narrow band test of @xmath28 coupling limits that are otherwise model - dependent .
this work was supported by the office of naval research award n000140910481 .
99 m. ahlers et .
al . , phys . rev .
d77 ( 2008 ) 095001 . j. jaeckel and a. ringwald , phys . lett .
b659 ( 2008 ) 509 .
b. holdom , phys .
b166 ( 1986 ) 196 .
r. foot and x.g .
he , phys .
b267 ( 1991 ) 509 .
j. jaeckel , j. redondo , and a. ringwald , phys .
( 2008 ) 131801 .
k. choi , phys .
d 62 ( 2000 ) 043509 .
i. waga and j. frieman , phys .
d 62 ( 2000 ) 043521 .
p. sikivie , invited talk at the idm2010 international conference on the identification of dark matter , montpellier , france , july 26 - 30 , ( 2010 ) .
l. visinelli and p. gondolo , phys .
d81 ( 2010 ) 063508 .
p. arias , d. cadamuro , m. goodsell , et al .
, desy-11 - 226 , mpp-2011 - 140 , cern - phth- 2011 - 323 , ippp-11 - 80 , dcpt-11 - 160 ( 2012 ) .
p. sikivie , phys . rev .
d32 ( 1985 ) 2988 .
r. bradley et .
al . , rev .
phys . 75 ( 2003 ) 777 .
f. caspers , j. jaeckel , and a. ringwald jinst 4 ( 2009 ) p11013 . j. jaeckel and a. ringwald , phys .
b659 ( 2008 ) 509 .
p. slocum et .
al . , proceedings of the 6th patras workshop on axions , wimps and wisps , zurich , switzerland ( 2010 ) . e. g. adelberger , b. r. heckel , s. hoedl , et al . ,
98 ( 2007 ) 131104 .
s. j. asztalos , g. carosi , c. hagmann , et al .
104 ( 2010 ) 104 .
p. slocum et .
, proceedings of the 5th patras workshop on axions , wimps and wisps , durham , u.k .
a. martin et .
al . , proceedings of the 7th patras workshop on axions , wimps and wisps , mykonos , greece ( 2011 ) .
s. weinreb , m. w. pospieszalski , r. norrod , proc .
1988 ieee mit - s international microwave symp .
( 1988 ) 945 .
r. h. dicke , rev .
sci instrum .
17 ( 1946 ) 268 .
s. w. wedge and d. b. rutledge , ieee trans . on microwave theory and tech .
40:11 ( 1992 ) 2004 .
r. p. meys , ieee trans .
on microwave theory and tech . 26 ( 1978 ) 34 .
e. i. gates , g. gyuk , m. s. turner , astrophys . j. lett .
l123 ( 1995 ) 449 .
k. zioutas , phys .
94 ( 2005 ) 121301 .
g. raffelt , stars as laboratories for fundamental physics : the astrophysics of neutrinos , axions , and other weakly interacting particles , chicago university press ( 1996 ) . | light axion like particles ( alps ) that couple to two photons are allowed in a number of proposed extensions to the standard model of elementary particles . of particular interest from a theoretical and observational standpoint
is the energy regime near 0.1 mev .
we present results from a pilot experiment to search for a signal from a 0.14 mev scalar alp by way of its coupling to two photons . using a copper resonant cavity cooled to four degrees kelvin while immersed in a seven tesla magnetic field , and coupled to a low noise cryogenic amplifier and room temperature receiver
, we exclude an alp driven excess of 34 ghz photons with g @xmath010@xmath1/gev with 5@xmath2 confidence .
we discuss the ramifications of this initial measurement as well as planned modifications to the experiment for increased sensitivity . |
magnetohydrodynamic ( mhd ) turbulence is a key element for understanding star formation ( see * ? ? ?
* ; * ? ? ?
* and references therein ) .
in a new emerging paradigm , density perturbations produced by compressible mhd turbulence create molecular clouds and clumps within the clouds .
this makes studies of density statistics very timely .
in addition , recent advances in the understanding of the properties of compressible magnetic turbulence ( see review by * ? ? ? * and references therein ) call for observational testing . while velocity statistics are currently emerging as an essential tool for turbulence studies ( see review by * ? ? ?
* and references therein ) , density statistics are more readily available from observations . variations of density have been extensively studied for a long time ( see * ? ? ?
* ; * ? ? ?
electron density fluctuations have become widely used to gain insight into small - scale properties of interstellar turbulence ( see * ? ? ? * and reference therein ) .
fluctuations of column density , at the same time , became a useful source of information about turbulence at larger scales @xcite .
those studies of column densities are not limited to studies of spectra .
higher order structure functions have been employed recently ( see * ? ? ?
* ; * ? ? ?
in fact , it was claimed in @xcite , that studies of higher order correlations could determine the mach number of the interstellar flows .
this motivates further studies of this type of statistics .
in addition , the topology of density in the interstellar medium ( ism ) can provide an important insight into ism physics .
for instance , the topology of the ism is expected to be different for models of the ism in which most of the gas is in the hot phase with islands of other phases embedded within ( see * ? ? ? * ) and in models where a substantial part of the volume is filled with cold and warm gas ( see * ? ? ? * ) .
it was pointed out in @xcite , that the topology of density can also be studied using genus statistics ( see * ? ? ?
however , no detailed study has been performed as far as we know .
the issues above call for a comprehensive study of density statistics , which we undertake in this paper .
we use an extended set of mhd numerical simulations , obtained for different sonic @xmath2 and alfvnic @xmath3 mach numbers , to study spectra , higher order statistics , and the genus of the density .
we follow turbulence for many dynamical time steps ( e.g. 25 for 128@xmath4 simulations ) to provide good sampling . in all cases
we provide results of synthetic observations to compare the statistics of density with the statistics of the column density that are available through observations . in this article
we investigate the scaling properties of the structure functions of density and the logarithm of density for compressible mhd turbulence with different sonic and alfvnic mach numbers . in
[ sec : models ] we describe the numerical models of compressible mhd turbulence . in
[ sec : pdfs ] we discuss one - point probability distribution functions ( pdfs ) . in
[ sec : spectra ] we study spectra of density and the logarithm of density and the anisotropy of density structures , and we show their dependence on mach numbers . in [ sec : self - similarity ] we perform tests of the property of self - similarity in the case of density structures . in [ sec : intermittency ] we present the intermittency of density and the logarithm of density and compare the intermittency of density and column density . in [ sec : moments ] we analyze the statistical moments , skewness and kurtosis , as possible measures of the @xmath2 of turbulence . in [ sec : considerations ] we study the effects of cloud boundaries , resolution and time averaging . in [ sec : topology ] we analyze density structure using genus statistics .
we compare the genus lines to those of projected density . in
[ sec : discussion ] we discuss our results and their relation to the previous studies .
we draw our conclusions in [ sec : summary ] .
we use a third - order - accurate hybrid essentially non oscillatory ( eno ) scheme ( see * ? ? ? * cl02 ) to solve the ideal mhd equations in a periodic box , @xmath5 = { \bf f } , \\
\frac{\partial { \bf b}}{\partial t } - \nabla \times ( { \bf v } \times{\bf b } ) = 0,\end{aligned}\ ] ] with zero - divergence condition @xmath6 , and an isothermal equation of state @xmath7 , where @xmath8 is density , @xmath9 is velocity , @xmath10 is magnetic field , @xmath11 is the gas pressure , and @xmath12 is the isothermal speed of sound . on the right - hand side ,
the source term @xmath13 is a random large - scale driving force .
the rms velocity @xmath14 is maintained to be approximately unity , so that @xmath9 can be viewed as the velocity measured in units of the rms velocity of the system and @xmath15 as the alfvn velocity in the same units .
the time @xmath16 is in units of the large eddy turnover time ( @xmath17 ) and the length in units of @xmath18 , the scale of the energy injection .
the magnetic field consists of the uniform background field and a fluctuating field : @xmath19 .
initially @xmath20 .
we drive turbulence solenoidally in fourier space to minimize the influence of the forcing on the generation of density structures .
density structures in turbulence can be associated with the slow and fast modes ( see * ? ? ?
? * ; * ? ? ?
* blc05 ) .
we use units in which @xmath21 and @xmath22 .
the values of @xmath23 have been chosen to be similar to those observed in the ism turbulence .
the average rms velocity @xmath14 in a statistically stationary state is around @xmath24 . for our calculations ,
similar to our earlier studies ( blc05 ) , we assume that @xmath25 . in this case , the sound speed is the controlling parameter , and basically two regimes can exist : supersonic and subsonic .
note that within our model , supersonic means low @xmath26 , i.e. the magnetic pressure dominates , and subsonic means high @xmath26 , i.e. the gas pressure dominates .
we present 3d numerical experiments of compressible ( mhd ) turbulence for a broad range of mach numbers ( @xmath27 and @xmath28 or @xmath29 ; see table [ tab : models ] ) .
the model name contains two letters : `` p '' and `` b '' followed by a number . the letters b and p mean the external magnetic field and the initial gas pressure , respectively , and the numbers designate the value of the corresponding quantity .
for example , a name `` b.1p.01 '' points to an experiment with @xmath30 and @xmath31 .
we understand the mach number to be defined as the mean value of the ratio of the absolute value of the local velocity @xmath32 to the local value of the characteristic speed @xmath12 or @xmath33 ( for the sonic and alfvnic mach number , respectively ) .
we analyzed models with three different resolutions : 128@xmath4 , 256@xmath4 , and 512@xmath4 on a cartesian grid with periodic boundary conditions .
we drove the turbulence solenoidally at wave scale @xmath34 equal to about 2.5 ( 2.5 times smaller than the size of the box ) .
this scale defines the injection scale in our models .
we do not set the viscosity and diffusion explicitly in our models .
the scale at which the dissipation starts to act is defined by the numerical diffusivity of the scheme .
the eno - type schemes are considered to be relatively low diffusion ones ( see * ? ? ?
* ; * ? ? ?
the numerical diffusion depends not only on the adopted numerical scheme but also on the `` smoothness '' of the solution , so it changes locally in the system .
in addition , it is also a time - varying quantity .
all these problems make its estimation very difficult and incomparable between different applications .
however , the dissipation scales can be estimated approximately from the velocity spectra . in the case of our models
we estimated the dissipation scale @xmath35 at 15 , 22 , and 30 for the low , medium , and high resolution , respectively .
we divided our models into two groups corresponding to sub - alfvnic ( @xmath36 ) and super - alfvnic ( @xmath30 ) turbulence . for each group
we computed several models with different values of pressure ( see table [ tab : models ] ) .
observations usually provide 2d maps of column density in the direction of the observer along the line of sight ( los ) .
thus , it is very important in computational astrophysics to study the relation between the actual 3d properties of the simulated turbulence and the synthetic observations prepared on the basis of these simulations .
first , we compare the probability distribution functions ( pdfs ) of density and column density .
pdfs give us information about the fraction of the total volume occupied by a given value of a measured quantity .
for the case of compressible turbulence in which evolution is described by the navier - stokes equation and the isothermal equation of state , the pdf of density obeys a lognormal form @xcite .
if we select a point in space and assume that the density at this point results from subsequent events perturbing the previous density , then the final density is a product @xmath37 , where @xmath38 is the initial density at the selected point and @xmath39 is a small compression / rarefaction factor . by the power of the central limit theorem @xcite , the logarithm of the resulting density , @xmath40 , should obey a gaussian distribution .
the process of perturbation of the column density @xmath41 is similar to the perturbations of density described above , but here a perturbation occurs only in the fraction of the los smaller than unity ; thus , each individual event induces a perturbation of column density with a smaller relative amplitude .
this implies a smaller dispersion of the column density distribution ( see * ? ? ?
@xcite proposed the parameter @xmath42 for determining the form of the column density pdf in molecular clouds , which is the ratio of the cloud size to the decorrelation length of the density field .
they defined the decorrelation length as the lag at which the density autocorrelation function ( acf ) has decayed to its 10% level .
if the density perturbation events are uncorrelated for @xmath43 , large values of the ratio @xmath42 imply that the central limit theorem can be applied to those events .
the decorrelation length estimated from the acf of the density for our medium - resolution models ranges from about 20 cell sizes of the computational mesh for supersonic models to about 50 cell sizes for subsonic models , which corresponds to @xmath44 if we take the size of the computational box as the column length ( in the case of medium resolution , it is equal to 256 cell sizes ) .
found values of @xmath42 signify that in our models at least partial convergence to a gaussian pdf should occur . in the top row of figure
[ fig : dens_pdfs ] we show pdfs of density normalized by its mean value for all models with medium resolution ( 256@xmath4 ) .
the plot on the left shows results obtained from sub - alfvnic experiments ( @xmath45 ) , and the plot on the right results from super - alfvnic models ( @xmath46 ) .
the pdf is additionally divided by the total number of mesh cells , which allows us to compare models with different resolutions . in the bottom row of figure
[ fig : dens_pdfs ] we also see pdfs for column densities for the same models .
the widths of pdf lines for @xmath47 are smaller then the widths of the corresponding lines for @xmath48 .
this confirms that a contribution from density perturbations to the perturbations of column density comes only from a fraction of the distance along the los giving smaller variations of the resulting column density .
our plots confirm the strong dependence of pdfs of density on the sonic mach number , an already known and well - studied property of density fluctuations in compressible turbulence ( see * ? ? ?
* ; * ? ? ?
* and references therein ) . for most of the models pdfs are lognormal functions . for super - alfvnic turbulence pdfs
are very symmetric about a vertical line crossing their maxima .
however , for models with a strong external magnetic field ( sub - alfvnic turbulence , @xmath45 ) and very low pressure ( supersonic turbulence , @xmath49 ) , the shape of the density pdfs is significantly deformed , and its lower value arm ends in higher densities than in the case with a weak magnetic field ( compare models for @xmath50 drawn with solid lines in the left and right plots of figure [ fig : dens_pdfs ] ) .
this supports a hypothesis that the gauge symmetry for @xmath51 that exists in navier - stokes equations is broken in mhd equations because of the magnetic tension term ( blc05 ) , which physically manifests itself by preventing the formation of highly underdense regions . in the higher density part of the distribution we do not see a similar effect , because the highly dense structures are created mainly due to shocks , and in this case , the magnetic tension is too weak to prevent the condensation .
we show the pdfs of column density normalized to the mean value of @xmath41 and divided by the number of pixels in the 2d map of density projected along the direction of @xmath23 .
we see that the maximum and dispersion of pdfs still strongly depend on the sonic mach number .
potentially this means that direct measurements of one - point statistics taken from the maps of molecular clouds can give us information about the speed of sound and the regime of turbulence taking place in these objects .
see , however , discussion of the boundary effect in [ sec : boundary ] . in figure
[ fig : dens_pdfs ] we included the degree of variation of pdfs in time as grey error bars .
we note that the departure of pdfs from their mean profiles is very small in the middle part around the mean value .
the strongest time changes are observed in the low- and high - density tails , but the pdfs for different models are still separable .
the spectrum of density fluctuations is an important property of a compressible flow . in some cases ,
the spectrum of density can be derived analytically . in nearly incompressible motions with a relatively strong magnetic field ,
the spectrum of density scales similarly to the pressure as @xmath52 if we consider the polytropic equation of state @xmath53 @xcite . in weakly magnetized mhd turbulence , however , the density spectrum is predicted to scale as @xmath54 @xcite . in supersonic flows , these relations might not be valid anymore because of shocks accumulating matter into the local and highly dense structures . due to the high contrast of density , the linear relation @xmath55 is no longer valid , and the spectrum of density can not be related to pressure so straightforwardly . in addition , the strong asymmetry of density fluctuations suggests the need to analyze the logarithm of density instead of density itself . in figure
[ fig : spectra ] we present the spectra of fluctuations of density and the logarithm of density ( _ top and middle rows , respectively _ ) for models with different @xmath2 . for comparison
we also show the spectra of velocity fluctuations ( fig . [
fig : spectra ] , _ bottom row _ ) .
the two columns of plots in figure [ fig : spectra ] correspond to experiments with @xmath45 ( _ left column _ ) and @xmath46 ( _ right column _ ) .
as expected , we note a strong growth of the amplitude of density fluctuations with the sonic mach number at all scales . this behavior is observed both in sub - alfvnic as well as in super - alfvnic turbulence ( see fig .
[ fig : spectra ] , _ top left and top right , respectively _ ) .
although the spectra significantly deviate from the power law , in table [ tab : slopes ] we calculate the spectral index of density @xmath56 and the logarithm of density @xmath57 within the inertial ranges estimated from the power spectrum of velocity .
the width of the inertial range is shown by the range of solid lines with slopes @xmath58 and @xmath59 in all spectra plots .
it is estimated to be within @xmath60 . in table
[ tab : slopes ] we also show the errors of estimation which combine the error of the fitting of the spectral index at each time snapshot and the standard deviation of variance of @xmath61 in time .
the slopes of the density spectra do not change significantly with @xmath2 for subsonic experiments and correspond to analytical estimations ( about @xmath62 , which is slightly less than @xmath63 , for turbulence with @xmath45 and about @xmath64 , which is slightly more than @xmath58 , for weakly magnetized turbulence with @xmath46 ) .
such an agreement confirms the validity of the theoretical approximations .
nevertheless , these theoretical considerations do not cover the entire regime of compressible turbulence .
while the fluid motions become supersonic , they strongly influence the density structure , making the small - scale structures more pronounced , which implies flattening of the spectra of density fluctuations ( see values for @xmath65 in table [ tab : slopes ] ) .
although , the spectral indices are clearly different for sub- and super - alfvnic turbulence , their errors are relatively large ( up to @xmath66 ) .
these errors have been calculated by summing the maximum value of the uncertainty of fits for individual spectra at each time snapshot and the standard deviation of time variation of spectral indices .
the uncertainty of fit contributes the most to the total error bar ( about 60%-70% ) .
nevertheless , there is still a small probability that both spectra , for sub- and super - alfvnic models , could overlap , resulting in less of a difference between these cases .
note that in the case of a large contrast , a more proper quantity to analyze is the logarithm of density ; we also show spectra of the logarithm of density fluctuations ( see fig .
[ fig : spectra ] , _ middle row _ ) .
as we can see , the amplitude of fluctuations of @xmath51 , like the case of @xmath8 , grows with @xmath2 at all scales .
however , the logarithmic operation significantly `` filters '' the extreme values of the density , and they do not distort the spectra anymore .
the slopes of spectra of @xmath51 are much less sensitive to the sonic mach number , which is confirmed by small changes of the slope with the value of @xmath2 ( see respective columns in table [ tab : slopes ] ) .
in addition , comparing sub- and super - alfvnic models in table [ tab : slopes ] , we see that , consistent with the theoretical considerations , the density gets a more shallow spectrum as the external magnetic field decreases ( or @xmath3 increases ) . for subsonic turbulence ,
the energy of the density fluctuations ( and the energy of fluctuations of the logarithm of density ) scales as @xmath52 in the case of strong @xmath23 and as @xmath54 in the case of weak @xmath23 . for comparison
we also show the power spectra of velocity ( figure [ fig : spectra ] , _ bottom row _ ) . while the spectra of density and the logarithm of density change both their amplitude and slope with the sonic mach number , for velocity fluctuations we see almost the same - shaped spectra within the inertial range , insensitive to the strength of sonic motions .
the only visible differences are present in the dissipation range , where the motions are strongly influenced by small - scale dissipation .
this dissipation depends on the pressure of the ambient medium , and thus on @xmath2 , as this is the case for shocks .
another question is the anisotropy of density and the logarithm of density structures . in figure
[ fig : anisotropy ] we show lines that mark the corresponding separation lengths for the second - order structure functions parallel and perpendicular to the local mean magnetic field . , the direction of the local mean magnetic field might change with @xmath67 at an arbitrary point .
this is an extension of the procedures employed in @xcite and @xcite .
] in the case of sub - alfvnic turbulence , the degree of anisotropy for density is very difficult to estimate due to the high disperion of points .
however , rough estimates suggest more isotropic density structures , because the points extend along the line @xmath68 . for models with @xmath69 the points in figure [ fig : anisotropy ] have lower dispersion , and
the anisotropy is more like the type from ( * ? ? ?
* hereafter gs95 ) , i.e. @xmath70 . in both alfvnic regimes ,
the anisotropy of density does not change significantly with @xmath2 .
plots for the logarithm of density show more smooth relations between parallel and perpendicular structure functions .
the dispersion of points is very small .
moreover , we note the change of anisotropy with the scale .
lower values of structure functions correspond to lower values of the separation length ( small - scale structures ) , so we might note that the logarithm of density structures are more isotropic than the gs95 model at small scales , but the anisotropy grows a bit larger than the gs95 prediction at larger scales . this difference is somewhat larger in the case of models with stronger external magnetic field ( compare plots in the left and right columns of figure [ fig : anisotropy ] ) , which may signify their dependence on the strength of @xmath23 .
the anisotropy of @xmath51 structures is marginally dependent on the sonic mach number , similar to the density structures .
all these observations allow us to confirm the previous studies ( see blc05 ) that suggest that the anisotropy depends not only on the scale but on @xmath3 . in comparison to the discussed anisotropy of density and the logarithm of density
, we also show the anisotropy of velocity ( see fig .
[ fig : anisotropy ] , _ bottom row _ ) .
the anisotropy of velocity is independent of @xmath2 too , but it slightly changes with @xmath3 ( compare the bottom left and right plots in figure [ fig : anisotropy ] ) , especially at small scales .
for both the sub - alfvnic and super - alfvnic turbulence we see the correspondence of the anisotropy to gs95 model . from these observations
we see that the anisotropy of the logarithm of density structures is very similar to the anisotropy of velocity .
the statistical distribution of velocity eddies of size @xmath67 is described by the moments of the velocity increments called structure functions @xmath71 where @xmath72 is a longitudinal increment @xmath73\cdot\hat{l } , \label{eqn : vel_incr}\ ] ] and @xmath74 denotes an ensemble averaging .
note that the structure function is defined here with the absolute value of the argument .
this ensures the positivity of the structure functions also for the odd orders .
an interesting and not yet fully understood property of these functions is their scaling behavior . a structure function of an arbitrary order scales with the separation length @xmath67 within the intertial range as @xmath75 . from such a scaling relation ,
one can deduce a self - similarity between structure functions of two arbitrary orders .
taking two different orders @xmath11 and @xmath76 , we need only a few transformations to derive a relation ( see * ? ? ?
* ) , @xmath77^{\zeta(p)/\zeta(p')}. \label{eqn : sf_relation}\ ] ] although simple , this relation implies two important properties of the function @xmath78 .
first , it shows how self - similarity connects the structure functions , and second , it lets us introduce a normalization using a structure function of an arbitrary order @xmath76 . in kolmogorov phenomenology ,
the relation @xmath79 has been derived analytically , implying @xmath80 ; thus , the third moment has been taken as the normalization .
let us note again that the above considerations concern only the part of the structure function within the inertial range .
@xcite reported that for hydrodynamic turbulence , functions @xmath81 exhibit a much broader power - law range compared to @xmath78 .
while for the inertial range a similarity in the scaling of two functions stems from the kolmogorov scaling @xmath82 , the power - law scaling of @xmath81 protrudes well beyond the inertial , into the dissipation range . as a function of @xmath67 ,
one gets @xmath83 as a function of @xmath84 , which is nonlinear in a way that corrects for the distortions of @xmath83 . ]
this observation shows that the dissipation `` spoils '' different orders of the structure function in the same manner .
therefore , there is no particular need to use the third moment , but one can use any other moment @xmath85 and obtain a good power law of the form analogous to eq .
( [ eqn : sf_relation ] ) ( see * ? ? ? * ) .
though not understood , the above property of @xmath83 allows insight into the scaling of realistically high re flows with limited re number simulations or laboratory experiments . in this way
, the work of @xcite allowed much progress in studies of the properties of self - similar flows .
all the above considerations originally concerned the properties of velocity and magnetic field structure .
relations between the fluctuations of density and the fluctuations of the other quantities like pressure , velocity and magnetic fields in the nonlinear regime of mhd turbulence are nontrivial to define ; therefore , the scaling function and its self - similarity are still not well understood .
although , there is no clear analytical reason for a self - similarity to be valid for density structures , we check this property by calculating analogous structure functions @xmath86 where @xmath87 might be an arbitrary scalar function of position , in particular , density or the logarithm of density .
note that we do not assume any property of scaling so far .
we calculate structure functions for density and the logarithm of density according to eq .
( [ eqn : sf_scalar ] ) .
we point out that in general , for a normalization function one can take an arbitrary order of the structure function , not necessarily that resulting from the kolmogorov law @xmath88 ( where the structure function is defined without the absolute value before averaging ; see * ? ? ?
the latter is only valid in the case of incompressible hydrodynamic turbulence for the velocity field within the inertial range where the viscous term is negligible .
the difference in definition of the structure function requires additional comment .
the averaging in both definitions is taken over the ensemble .
it means that if we get the negative value of the increment of the quantity @xmath89 , we can switch positions and take the increment @xmath90 into the averaging . this operation is permitted , because the kolmogorov law is valid only for the absolute values of @xmath67 . to test if a similar scaling law exists also for density and the logarithm of density in the magnetized compressible turbulence , we present figure [ fig : scaling_law ] , which shows third - order structure functions for density , the logarithm of density , and velocity as functions of the separation length .
the solid lines show the hypothetical kolmogorov scaling law @xmath91 , where @xmath92 is a constant . in the case of density ,
the models with a strong magnetic field show relatively large dispersion of points .
in addition , the line flattens very quickly ( @xmath93 ) .
in this situation , the region where the structure function scales linearly with the separation length is difficult to find .
the situation is somewhat better in the case of the logarithm of density .
although the relations are far from linear , we can estimate the region of linear scaling ( @xmath94 for sub - alfvnic models and @xmath95 for super - alfvnic models ) . in figure
[ fig : scaling_law ] ( _ bottom row _ ) we show the third - order structure functions for velocity .
we can clearly see the best linear scaling withing the range @xmath96 for subalfvnic models and @xmath97 for superalfvnic models .
the linear scaling regions for the logarithm of density and velocity do not correspond .
the dependence of third - order structure functions on the sonic mach number is marginal .
the only dependence can be observed in the case of velocity in models with a weak magnetic field .
however , the sonic mach number does not change the slope of the structure function . in figure [
fig : dens_logd_ess ] we show a test of an extended self - similarity .
we plot the structure functions of second , fourth , and sixth order as a function of the structure function of third order .
plots contain functions for sub - alfvnic models only with different sonic mach numbers for density and the logarithm of density ( top left and right plots in figure [ fig : dens_logd_ess ] , respectively ) .
the alignment of points along straight lines in figure [ fig : dens_logd_ess ] clearly suggests the existence of a self - similarity both for density and the logarithm of density . in the case of density , this nice picture is slightly disturbed for supersonic models , but only for @xmath83 with order @xmath98 .
structures with large density contrasts significantly distort structure functions of higher orders .
nevertheless , the linear relation is still well seen . applying the logarithmic operation on density before computing the structure function significantly improves a linear alignment of points and introduces some ordering in the placement of lines ( see fig . [
fig : dens_logd_ess ] , _ top right _ ) .
we see that lines corresponding to the structure functions of the same order are placed on the same line .
this is especially well seen for functions of second order .
this draws a hypothesis that the `` perfect '' low - order structure functions of @xmath51 have the same slope independent of the sonic mach number and the structure functions are only shifted along the line of a fixed inclination according to the strength of the sonic motions . for higher orders the strong asymmetries in the distribution of @xmath51 dominate , giving the reason for the flattening of the relation between structure functions of different orders .
the existence of a self - similarity of structure functions of density and the logarithm of density is very useful .
the question of to what degree these relations are linear induces us to perform a fitting , which is shown in figure [ fig : dens_logd_ess ] ( _ bottom row _ ) .
we have chosen one particular model with @xmath99 and @xmath45 .
for this model we perform linear fitting in the log - log domain of second , fourth and sixth order structure functions of density ( fig . [
fig : dens_logd_ess ] _ bottom left _ ) and the logarithm of density ( fig .
[ fig : dens_logd_ess ] _ bottom right _ ) as a function of the third order structure function of the corresponding quantity . in this case , points are regularly spaced and fitting on the log - log axes gives very accurate values of slopes even for high - order structure functions ( fractional errors in the order both for density structure functions and the logarithm of density for the lines are presented in fig . [ fig : dens_logd_ess ] ) .
such a high degree of linearity suggests the existence of highly self - similar structures independent of scale , also self - similar in some way , because at small scales the dissipation starts to play an important role , which we clearly see in spectra of density , velocity , or magnetic fields .
the effect of dissipation is not seen for relations between structure functions of different orders , which suggests that the dissipation acts in a consistent way at different scales .
moreover , the dissipation must undergo a similar scaling relation , otherwise we would get different slopes of the sfs within the inertial and dissipative ranges . going further , we could ask if the extendend self - similarity is valid for structure functions of different quantities . in figure
[ fig : sf3velo ] we draw plots of the structure functions of the second , third , fourth and sixth order of density ( _ left _ ) and the logarithm of density ( _ right _ ) as a function of the structure function of third order of the velocity .
the existence of any similarity should be manifest by the linear relation here also , but as we see , this relation is far from linear .
there are at least two reasons for this .
at first we may suppose that scaling relations of the structure functions of a particular order , which are used as a normalization ( in this paper we use the third moment ) , look different for different quantities , but it might also be the influence of dissipation acting differently on different quantities .
whatever the reason is , the problem of relations , if any exist , between the structures of different quantities involved in turbulence requires further investigation .
an important insight into the turbulence can be achieved by studies of turbulence self - similarity .
this property , which is also called scale invariance , implies that fluid turbulence can be reproduced by the magnification of some part of it .
take the famous model of incompressible kolmogorov turbulence as an example . in this model ,
the energy is injected at a large scale @xmath18 and forms eddies that transfer the energy to smaller and smaller scales . at the scales where the corresponding reynolds number re@xmath100 is much larger than unity , the dissipation over the eddy turnover time @xmath101 is negligible . as a result ,
the energy cascades to smaller and smaller scales without much dissipation , i.e. @xmath102 , which gives the well - known kolmogorov power - law scaling for the eddies of scale @xmath67 , namely , @xmath103 .
the cascade terminates at the dissipation scale , which provides an inertial range from @xmath18 to @xmath104 , where re is the reynolds number corresponding to the flow at the injection scale @xmath18 . at
the dissipation scales , self - similarity is known to fail with turbulence forming non - gaussian dissipation structures as exemplified , e.g. in @xcite .
interestingly enough , present - day research shows that self - similarity is not exactly true even along the inertial range .
instead , fluctuations tend to get increasingly sparse in time and space at smaller scales .
this property is called intermittency .
note that power - law scaling does not guarantee scale invariance or the absence of intermittency .
intermittency is an essential property of astrophysical fluids .
as intermittency violates the self - similarity of motions , it becomes impossible to naively extrapolate the properties of fluids obtained computationally with relatively low resolution to the actual astrophysical situations . in terms of astrophysics , intermittency affects turbulent heating , momentum transfer , interaction with cosmic rays , radio waves , and many more essential processes .
physical interpretation of intermittency started after the work by kolmogorov , but the first successful model was presented by @xcite .
the scaling relations suggested by @xcite related @xmath105 to the scaling of the velocity @xmath106 , the energy cascade rate @xmath107 , and the co - dimension of the dissipative structures @xmath108 , @xmath109 parameter @xmath108 is connected to the dimension of the dissipative structures @xmath110 through relation @xmath111 @xcite . in hydrodynamical turbulence , according to kolmogorov scaling , we have @xmath112 and @xmath113 . vortex filaments , which are 1d structures , correspond to @xmath114 ( @xmath115 ) . for mhd turbulence
we also observe current sheets , which are 2d dissipative structures , corresponding to @xmath116 ( @xmath117 ) .
for these two types of dissipative structures we obtain two different scaling relations ( substituting @xmath112 and @xmath113 ) , @xmath118 \ \mathrm{for } \
c=2 \label{eqn : sl}\ ] ] and @xmath119 relation ( [ eqn : sl ] ) is often called the she & lvque scaling @xcite and equation ( [ eqn : mb ] ) the mller - biskamp scaling @xcite .
there are theoretical arguments against the model ( see * ? ? ?
* ) , but so far the she & lvque scaling is the best for reproducing the intermittency of incompressible hydrodynamic turbulence .
so far there is no similar theory for density scaling in compressible mhd turbulence .
the presence of shocks disturbs the spectra and structure functions of density .
however , as we discussed in [ sec : spectra ] , the logarithm of density exhibits both spectrum and goldreich - sidhar scale - dependent anisotropy ( see also blc05 ) . in figure [
fig : dens_expons ] we show the scaling exponents for density ( _ top row _ ) and the logarithm of density ( _ bottom row _ ) for models with @xmath45 ( _ left column _ ) and @xmath120 ( _ right column _ ) . in the case of density (
_ top row _ ) we show only the subsonic models because of broken self - similarity of density structure functions in supersonic models . to overcome this problem one can use techniques filtering the high - density structures , which deform statistics the most .
the application of such filtering has been discussed by blc05 .
nevertheless , we do not investigate this problem further in this article .
when possible , we use extended self - similarity to obtain the exponents , applying the third - order structure function as a normalization .
we perform such fitting procedures for all available time steps for each model . in this way we obtained the mean values of exponents and their standard deviation . comparing plots for sub- and superalfvnic models ( fig .
[ fig : dens_expons ] , _ top left and top right , respectively _ )
we see that in the presence of a strong external magnetic field , density structures are less intermittent .
although the error bars are relatively large , the difference is undisputed .
moreover , the value of the parameter @xmath108 seems to be between one and two for the sub - alfvnic models with very low sonic mach numbers and approximately equal to two when @xmath121 ( see figure [ fig : dens_expons ] , _ top left _ ) .
when the magnetic field becomes weak , the intermittency of density grows .
in addition , it grows somewhat with the sonic mach number .
the changes are smaller then the errors however . in figure
[ fig : dens_expons ] ( _ bottom row _ ) we present the scaling exponents calculated from the structure functions of the logarithm of density . after this transition
, the scaling exponents can be more precisely determined for all models , even for supersonic ones . in turbulence with a strong magnetic field ( fig .
[ fig : dens_expons ] , _ bottom left _ ) we see that intermittency behaves almost identically for subsonic models
. we can restore the scaling of density even for supersonic motions .
the value of @xmath108 is between one and two for models with @xmath122 and exactly two for the model with @xmath123 .
the errors in the latter case are rather small . our models with a weak magnetic field
reveal more intermittent structure of @xmath51 as in the case of density ( see fig .
[ fig : dens_expons ] , _ bottom right _ ) . in these models ,
the scaling exponents suggest the dimension @xmath114 .
although we see some small variations of exponents with the value of @xmath2 , they are smaller then the error bars .
what we can say about all the above models is that the structure of density is intermittent and the intermittency depends on the strength of the external magnetic field .
one may note some changes in intermittency with the strength of sonic motions , but these changes are weak .
in general , the supersonic motions generate more intermittent structures of density or the logarithm of density .
the only possible exception is the logarithm of density , which has structure more intermittent with stronger sonic motions in the presence of a weak magnetic field , but with a larger uncertainty at the same time .
finally , in comparison to the density , the intermittency decreases when we use the logarithm of density . in figure
[ fig : dens_int_expon ] we show the scaling exponents for column densities for different models .
although the estimation of exponents induce substantial uncertainties , we see a small change in intermittency with the sonic mach number in all plots , especially in plots showing the column density integrated across @xmath23 in the models with a strong magnetic field . to investigate this change deeper
we show how the intermittency of column density changes with sonic mach number . in figure
[ fig : col_dens_scal ] we present the scaling exponents of the structure functions of column densities @xmath41 for different experiments . as a result
we obtain the relation of the scaling exponent and the sonic mach number .
the exponents are normalized to the third - order one , so figure [ fig : col_dens_scal ] shows exponents for the structure functions of the first , second , fourth , fifth , and sixth orders .
we average the exponents for each model over 16 equally spaced moments in time @xmath124 $ ] , and we calculate their standard deviation . in this way we obtain also a measure of uncertainty . figure [ fig : col_dens_scal ] shows the exponents for column densities integrated along and across @xmath23 . in all plots we see a systematic decrease of the exponents for orders larger than 3 and simultaneously an increase of the exponents for orders less than 3 .
this behavior signifies a progressive growth of the intermittency of density structure with sonic mach number .
the most straightforward statistical properties of density and column density distributions are the mean value and the variance . however , on a deeper level we are interested to know the degree of asymmetry of the distribution and its relation to the normal distribution .
these two properties are measured by the skewness and kurtosis . the mean arithmetic value and the variance of the distribution of , for example , density @xmath8 are given by @xmath125 where @xmath126 is the number of samples or points of the mesh in the case of simulation data .
the mean value is a less important property in our studies , so we do not consider it here ; however , it is required to calculate higher moments .
variance measures the width of the distribution and is always positive by definition .
skewness as a measure of the asymmetry is defined through the third - order statistical moment , @xmath127 the skewness for a normal distribution is zero .
negative values of the skewness indicate distribution that is skewed in the left direction , and positive values for the skewness indicate the distribution that is skewed in the right direction . by skewed left , for instance ,
we mean that the left tail is longer compared to the right tail .
kurtosis is a measure of whether the quantity has a distribution that is peaked or flat in comparison to the normal distribution .
that is , data sets with high kurtosis tend to have a distinct peak near the mean , decline rather rapidly , and have heavy tails .
data sets with a low kurtosis tend to have a flat top near the mean rather than a sharp peak
. a uniform distribution would be the extreme case .
kurtosis is defined in a similar way to skewness , but using the fourth - order statistical moment , @xmath128 positive kurtosis indicates a `` peaked '' distribution and negative kurtosis indicates a `` flat '' distribution . in figure
[ fig : dens_moments ] we show the variance , skewness and kurtosis ( _ top , middle , and bottom rows , respectively _ ) for densities and column densities for all low - resolution models ( models with resolution 128@xmath4 , see table [ tab : models ] ) .
from such a set of models we obtain the dependence of the moments on the sonic mach number .
error bars are determined using the standard deviation of the results for each model from all subsequent steps for times @xmath129 $ ] .
the variance of density fluctuations has been studied extensively , so its relation to @xmath2 is well known ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the variance of density grows with the sonic mach number and consequently with the broadening of the pdf .
the same behavior is observed in the variance of column densities .
lines for variances in figure [ fig : dens_moments ] look very similar for models with strong and weak magnetic fields .
however , the comparison of the variances for column densities integrated along and across the external magnetic field reveals a difference in its values , a sort of split , in the case of the sub - alfvnic models ( see fig . [
fig : dens_moments ] , _ top left _ ) . in the case of super - alfvnic turbulence
, there is no split at all .
lines perfectly overlap .
the latter is natural , as the super - alfvnic turbulence easily tangles the external magnetic field .
according to @xcite , the relations between the variance of density fluctuations and the sonic mach number are @xmath130 when the magnetic pressure dominates and @xmath131 when the gas pressure dominates .
the former case is observed when @xmath132 .
indeed , in figure [ fig : var_mach ] we see that the mean standard deviation of the density fluctuations @xmath133 scales with @xmath2 almost linearly when @xmath134 . when @xmath135 , the relation is much steeper .
a similar behavior is observed in the case of super - alfvnic turbulence ( fig .
[ fig : var_mach ] , _ right _ ) although the gas pressure dominates in all models . for models with a very small value of the sonic mach number
, the relation is different .
the value of @xmath133 depends much less on @xmath2 .
studies by @xcite showed the dependence of density fluctuations on the alfvnic mach number , which is different for different modes .
when @xmath136 is small , the slow mode dominates with the relation @xmath137 .
the fast mode starts to contribute more significantly in the generation of density fluctuations with the relation @xmath138 when @xmath136 is large .
a split in the @xmath139@xmath2 relation for column densities parallel and perpendicular to @xmath23 in the sub - alfvnic turbulence ( fig .
[ fig : dens_moments ] , _ top left _ ) can be explained by a correlation of clumps and filaments of the density structure with the direction of the mean magnetic field resulting from the action of the slow mode as a main contributor to the generation of density fluctuations . in the super - alfvnic models ,
the variance of the column density is independent of the direction of the los ( compare to fig . [
fig : dens_moments ] , _ top right _ ) .
plots of variances show that the strength of sonic motions can be directly measured from the variance of intensity in the maps of molecular clouds .
more interesting information is provided by the two other statistical measures : skewness and kurtosis . in figure
[ fig : dens_moments ] ( _ middle row _ ) we see that asymmetry of density and column density distributions grows with the value of @xmath2 , but only if @xmath140 .
when the sonic mach number falls below this value , the asymmetry of the distribution starts to grow again .
the skewness of column density varies in the same way with the minimum placed exactly at the same value of @xmath2 . in all cases ,
the skewness has positive values , which is interpreted as an asymmetry of the distribution into the direction of higher densities .
the skewness is rather insensitive to the strength of the external magnetic field .
the kurtosis , presented in figure [ fig : dens_moments ] ( _ bottom row _ ) , oscillates around zero for models with highly subsonic motions .
this suggests an almost perfect gaussian distribution for density and column density .
it is also in agreement with the fact that if density is perturbed weakly , its distribution is normal .
the growth of kurtosis with @xmath2 and its positive values signify a distribution `` peaked '' more than the standard normal distribution .
all three measures reveal strong dependence on the sonic mach number .
moreover , the utility of the above measures for the column density follows from the similarity to the corresponding measures of 3d density .
to connect our studies with column densities available from observations , we transform our density cube to a cloudlike shape .
this operation has been done by a simple multiplication of the density by a 3d function , which is uniform within a ball of radius @xmath141 and decaying outside @xmath141 with a gaussian profile . in the next step we calculated column densities from the resulting data and performed an analysis in the same way as was previously done . in this simple way we obtain the synthetic map of a molecular cloud .
we should emphasize here that the results obtained from the synthetic observations of this type are mostly applicable to supersonic turbulence .
the restriction is due to the fact that the density fluctuations in subsonic turbulence are small in comparison to the mean cloud profile , so the two - point statistics may be dominated by this large - scale distribution .
therefore , in this section we narrow down our studies only to the supersonic experiments .
examples of maps obtained after applying the above procedure are presented in figure [ fig : synth_maps ] .
the change in structure is visible `` by eye '' . for highly supersonic motions ( fig .
[ fig : synth_maps ] , _ left column _ ) , the column density exposes filamentary structure , which is created by strong shocks .
the overall shape of the cloud is difficult to determine . in the middle column of figure
[ fig : synth_maps ] , for medium sonic mach numbers , the overall shape of the cloud starts to appear .
the maps still show filaments but in a much more compact form with many spongelike elements . for subsonic turbulence
, the cloud looks like a ball of cotton .
we can recognize some whirled streaks . in this case , the mean shape of the cloud clearly dominates . to answer the question concerning the influence of boundaries on the statistics , we start by presenting the pdfs of column densities ( figure [ fig : dens_sph_pdfs ] ) .
all models have distributions very different from those presented in figure [ fig : dens_pdfs ] .
the pdfs are very far from the symmetry with respect to the mean value .
the shape of the low - density part is influenced mainly by the boundaries of the cloud .
the boundary of a cloud is an interface where the density decreases from the mean value of the interior of the cloud to the density of the ambient medium , so this explains why the shape of the low - density part looks very similar in all models independent of the mach number . however , if we take a look at the high - density part of the pdfs , we see that here the shape does depend on the sonic mach number .
because the densities of the ambient medium are too small to disturb this part of the pdf , it is clear the these are formed only by the internal structure of the cloud .
thus , for our cloud model , the pdfs from the higher density part of the cloud are more reliable . in figure
[ fig : dens_sph_moments ] we show three statistical moments of cloudlike - shaped densities as functions of moments for the original full cube data . in the six plots in figure [ fig : dens_sph_moments ]
we show variances ( _ top row _ ) , skewnesses ( _ middle row _ ) and kurtosises ( _ bottom row _ ) for sub - alfvnic ( _ left column _ ) and super - alfvnic ( _ right column _ ) experiments .
we plot the values of moments obtained from models with different sonic mach numbers as a function of moments for the same models but with densities modulated by the gaussian distribution .
the boundary does not influence significantly the moments if the relations presented in the plots of figure [ fig : dens_sph_moments ] are linearly increasing . on the contrary
, the boundary dominates if values of the moments for @xmath142 or @xmath143 do not change with the moments for @xmath144 and @xmath145 . in this way
we can see how the boundary influences the variance , skewness , and kurtosis of density and column density
. taking a closer look into the relation between the moments of 3d density ( lines plotted using stars ) we see that the linearity of the relation grows with the order of the statistical moment , which means that the boundary contributes the most in variance , less in skewness , and in the case of kurtosis , the contribution is almost negligible ( see fig .
[ fig : dens_sph_moments ] , _ top , middle , and bottom rows , respectively _ ) .
it means that if we want to avoid the influence of the boundaries , we should use the kurtosis as a measure of the sonic mach number in turbulence with a proper scaling factor .
however , we are mainly interested in the moments of 2d data , which are available from observations . lines corresponding to the moments of column densities ( plotted using triangles and squares ) reveal much higher dispersion and irregularity . nevertheless , we are still able to determine regions where the relation between statistical moments of column density calculated from cloudlike - shaped density and moments of unmodified column density is linear ( see fig . [
fig : dens_sph_moments ] ) .
the skewness of @xmath41 could also be a measure of the strength of the external magnetic field , because its relations , seen in middle left plot of figure [ fig : dens_sph_moments ] , are different for densities integrated along and across @xmath23 ( lines plotted using squares and triangles , respectively ) .
we do not see such differences in the results of models with weak @xmath23 ( see fig .
[ fig : dens_sph_moments ] , _ middle right _ ) .
the relations for kurtosises in the case of @xmath41 seem to be least helpful .
the points are strongly agglomerated , with only a few points , corresponding to the highest sonic mach numbers , creating a straight line . as a result of these agglomerations
, the relations can only be used in the case of highly supersonic turbulence .
figure [ fig : dens_sph_expon ] shows two plots for a relation between the exponents of the structure functions of column densities for the sub- and super - alfvnic turbulence ( _ left and right plots , respectively _ ) .
these plots suggest a strong dependence of intermittency on @xmath2 .
it seems to be even stronger than in the plots in figure [ fig : col_dens_scal ] due to the presence of a boundary .
as previously described , in models with low values of @xmath2 , the contribution in the structure functions mainly comes from the large - scale structure of the cloud . in this case , the exponents are almost independent of @xmath2 ( see the subsonic part of the plots in figure [ fig : dens_sph_expon ] ) . however , when the sonic motions become stronger , they are able to produce a high contrast in density structures , enough to dominate in the structure functions .
thus , we are able to study the intermittency from the intensity maps only in turbulence with sufficient values of @xmath2 . however , to overcome this problem in the case of subsonic turbulence , we should restrict analogous analysis only to the intensity maps of the interior of the cloud . in both cases
, we should treat the statistics obtained from the intensity maps very carefully , including the boundary effect as much as possible .
one could ask how our results depend on the resolution of our calculations .
undoubtedly , in models with a higher resolution the small - scale structure of density is better resolved .
we have seen in previous sections that these small - scale structures can significantly affect the spectra and structure functions . in this section
we examine to what extent our results depend on numerical resolution . in figure
[ fig : reso_pdfs_spectra ] we show pdfs and spectra of density for experiments with the same physical conditions , but for three different resolutions : 128@xmath4 , 256@xmath4 , and 512@xmath4 . as a reference example
we show only models with @xmath99 and @xmath45 .
plotted functions for the low and middle resolutions are the result of averaging over several time steps to get the results independent of time variations . for the highest resolution
we have only a single snapshot in time .
figure [ fig : reso_pdfs_spectra ] ( _ left _ ) suggests that the distribution of density is almost insensitive to the resolution .
we see very similar shapes and values of lines .
the only visible difference is a tail at higher densities for the high - resolution experiment , which can be a result of time - dependent fluctuations of the highly compressed small - scale structure .
it could also be a result of better resolved shocks due to the smaller numerical diffusivity in the case of the high - resolution model , which smooths the density compression at shocks less , resulting in a higher density contrast .
a progressive growth in the numerical diffusivity is especially well seen in the spectra of density ( see fig . [
fig : reso_pdfs_spectra ] , _ right _ ) . due to a high diffusivity
, the inertial range is very short for the low - resolution model , but its slope is consistent with slopes for higher resolution experiments . almost perfect covering of
the dissipation ranges for all resolutions in the right plot of figure [ fig : reso_pdfs_spectra ] shows that the numerical dissipation acts in the same way on the small - scale density structure independent of the resolution .
the resolution changes only the length of the inertial range .
figure [ fig : reso_expons ] shows the scaling exponents for density for three different resolutions . to estimate exponents and errors we calculate the mean value and standard deviation from results for all available time snapshots at @xmath146 for a given experiment .
figure [ fig : reso_expons ] suggests the growth of the intermittency of density with resolution .
the intermittency is more sensitive to the resolution .
it is especially well seen at higher orders @xmath11 .
simultaneously , we see bigger errors for lower resolutions .
but bigger errors could also indicate that the intermittency of structure is oscillating stronger over time .
an interplay of forces creating small - scale and dense structures through shock formation and diffusion ( also numerical ) and destroying such structures could establish an equilibrium of some sort when the development of turbulence reaches a stationary state . in this kind of analysis , there are also statistical restrictions due to the sampling size effects .
structure functions are reliable to some order depending on the resolution of numerical models . in our studies
we have the size of analyzed samples ranging from about 2@xmath14710@xmath148 to over 10@xmath149 .
in such a case , the maximum order should be for @xmath150 .
this explains the increase of the error bars .
our conclusion is that we should expect a somewhat higher intermittency of density in experiments with higher resolution .
in some sense our study sends a warning to the low - resolution studies of the she - lvque scaling .
we see that while the extended self - similarity may be fulfilled for the low - resolution studies , their higher order scalings can still be different from higher resolution runs .
in our analysis we used 16 snapshots of data uncorrelated in time for the low - resolution models and from eight up to 17 snapshots for the medium - resolution models depending on the model .. ] our results show a relatively significant magnitude for the errors ( see figures [ fig : dens_expons ] , [ fig : col_dens_scal ] , and [ fig : dens_moments ] ) understood here as the measure of the time variance of the analyzed quantities .
continuous interaction between waves results in creation of the structures .
we observe smooth transitions from one structure to another .
the process is stochastic .
it means that as with all stochastic processes , it should not be considered for one particular snapshot , but a relatively large number of snapshots should be taken into account .
for instance , if we had used only one snapshot , we would come to the wrong conclusions about whether the intermittency is higher or lower for a particular case or whether the statistical moments scale differently with @xmath2 .
this is especially important for measures with a relatively high level of uncertainty like scaling exponents or statistical moments for the column densities .
in addition , in numerical simulations the time step is defined by the maximum characteristic speeds in the system .
these speeds depend on the local amplitudes of the velocity and magnetic field , which can be different for different models . in this case , a small shift in time at which we compare snapshots for different models would result in completely different conclusions for measures strongly varying in time . in this way , the analysis of turbulence for only small time averaging produces results that could be very unreliable .
to study the topology of density structures , we applied the genus statistics that are successfully used in the cosmic microwave background studies ( * ? ? ? * for example ) . for a given density threshold
we define an isosurface of density separating higher and lower density regions and evaluate the genus of this contour . roughly speaking
, the genus of an isosurface is the number of holes in this surface minus the number of isolated regions for a given threshold .
this difference changes with the change of the threshold .
we plot the genus of these contours as a function of the density threshold . to perform genus statistics we use a fortran algorithm published by @xcite . from the analysis of the genus of isosurfaces of density
, we can extract information about the density distribution and about the topology of structures that are created by turbulence .
the results and plots in this section refer to the medium - resolution models . in figure
[ fig : genus_statistics ] ( _ left column _ ) we present the distribution of genus for three different sonic mach numbers , one for the subsonic case ( fig .
[ fig : genus_statistics ] , _ top left _ ) and two for the supersonic cases ( fig .
[ fig : genus_statistics ] , _ moddle left and bottom left _ ) . for the subsonic case in which the distribution is almost symmetrical with respect to the mean density value , the shape of the genus line suggests a spongelike density structure with occasional overdense clumps or under - dense holes .
we observe a somewhat higher number of higher amplitude dense regions than rarefactions , because on the high - density side of the distribution , the line decreases to highly negative values ( down to about -80 ) .
correspondingly the low distribution of rarefactions reaches a value of only -50 . for supersonic models with @xmath151 around 2.5 or 7.0
, the genus line looks much different .
although the structure seems to be also spongelike ( the distribution in the vicinity of the mean density value is broad and highly positive ) , the low-@xmath26 turbulence produces an extended tail in the high - density direction .
this suggests the existence of many highly dense singular structures
. it may be evidence for the formation of strong shocks , which perturb and compress density locally in small volumes ( see blc05 ) .
density in these regions grows to a value 25 times larger than the mean value in the case of @xmath152 turbulence . in connection with this
we observe the disappearance of the low - density tail in the distribution , which means that underdense holes are short lived .
we can ask the question about whether the genus statistics of integrated density along one chosen axis can give us adequate information about the 3d topology of the density structure .
with such a relation , we could interpret maps of column densities obtained from observations of the diffuse ism and molecular clouds . in figure
[ fig : genus_statistics ] ( _ right column _ )
we show the genus statistics of density integrated along the @xmath153-axis ( or along the external magnetic field , which is also directed along the @xmath153-axis ) .
we applied the 2d version of the algorithm for genus computations @xcite .
a 2d genus is interpreted as the number of isolated 2d high - density regions minus the number of isolated 2d low - density regions . while the 3d genus is symmetric about the mean value , the 2d version is antisymmetric . for the high-@xmath26 case ( fig .
[ fig : genus_statistics ] , _ top right _ ) the genus distribution is very symmetrical , and we only see one rarefied hole structure with column density lower than 180 .
this corresponds to the gaussian distribution of the integrated density for this model . for the low-@xmath26 cases the structure
is extended more into high - density structures , like in the 3d case of the genus line , and we do not observe any symmetry there . although the number of more dense regions is somewhat lower than the less dense regions , the maximum is shifted to the right .
we can say that all the dense clumps have densities above 300 , while holes have densities below 150 .
it was previously discussed that clumps with density much higher than the mean value can distort the power spectrum and hide the structure of the density created by motions at small scales ( blc05 ) .
following blc05 we applied these techniques to test how the genus statistics would be affected . in figure
[ fig : genus_logarithm ] we present the genus lines for the logarithm of density ( _ left column _ ) and the corresponding 2d genus lines for the logarithm of the integrated density ( _ right column _ ) . from the 3d genus plots ( fig .
[ fig : genus_logarithm ] , _ left column _ ) we see that the lines are almost symmetric about zero .
the most non symmetry reveals itself in the supersonic models , where the maximum shifts to the negative values of @xmath51 . for integrated density plots ( fig .
[ fig : genus_logarithm ] , _ right column _ ) , we do not see any significant differences between the results for different mach numbers . for high mach numbers , the logarithm of density has a higher extremum , which is not unexpected .
we have tested the genus statistics for density with the cutoff of high - density peaks .
we do not show the plots for them , because the procedure of cutting off the density above some level simply does not change the line in plots of figure [ fig : genus_statistics ] below the cutoff value and only removes the high - density tail . the interesting information would be the sonic mach number retrieved from the genus statistics of column density maps .
the sonic mach number can be obtained from the length of the high - density tail of the genus line .
it is just another way to measure the maximum value of density , which is strongly related to the value of @xmath2 , which is explained more in
[ sec : pdfs ] . however , more important information that the genus can give is the type of structure . for low sonic mach number models we have both underdense holes and overdense clumps with approximately the same count of each .
this is supported by the top right plot in figure [ fig : genus_statistics ] , where we observe distribution a roughly symmetric about the mean value . however , with the increase of sonic mach number , turbulence produces more high - density structures , extending the right tail of the genus line significantly .
this means that the structure starts to look rather like a medium with lower density in the most of volume , but with many singular 1d or 2d structures with much higher density .
the reader can compare this picture to figure [ fig : synth_maps ] , which supports our claims .
we can conclude then that the genus statistics can be a valuable tool to study the structure of molecular clouds .
we feel that it is important for studies of turbulent density to be understood in the context of broad mhd turbulence .
there has long been an understanding that mhd turbulence is anisotropic ( e.g. * ? ? ?
a very important insight into incompressible mhd turbulence by @xcite ( gs95 ) has been followed by progress in the understanding of compressible mhd turbulence ( cl02 ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in particular , simulations in cl02 showed that the alfvnic cascade evolves on its own and the exchange of energy between alfvn , slow , and fast modes is marginal .
, where @xmath33 and @xmath12 are the alfvn and sound velocities respectively
. however , the coupling becomes marginal at smaller scales , as turbulence cascades and turbulent velocities get smaller .
] this understanding differs from some earlier claims of a high degree of coupling of compressible and incompressible motions .
incidentally , the notion of high coupling was used to explain the fast damping of compressible mhd turbulence reported in @xcite and in @xcite . in view of the current understanding of mhd turbulence
we claim that the fast decay of mhd turbulence is due to the fast decay of the alfvnic mode ( see * ? ? ?
* ) , the effect that is present even in incompressible mhd ( see * ? ? ?
* ) . in view of the aforementioned picture of mhd turbulence ,
density perturbations arise from the slow and fast modes .
alfvn modes , however , shear and cascade the slow modes ( gs95 ) and therefore affect the density spectrum .
note that for supersonic turbulence the slow modes get modified and produce shocks [ the spectrum @xmath154 . the large - scale shocks produced compress density and produce a shallow spectrum , as we observed in the simulations above ( see also blc05 ) . in terms of the study of pdfs and spectra , our present work goes beyond the one in blc05 , as we provide a wider parameter space survey .
we reach conclusions similar to those in blc05 ( see also * ? ? ?
interestingly enough , our results are also similar to those obtained in non - magnetized fluids by @xcite . indeed , we find that the spectrum of density gets shallower than the kolmogorov one as the sonic mach number increases .
the reason for this is the formation of small high - density compressed regions that dominate the spectrum .
if the logarithm of density is used , these regions do not dominate and the underlying spectrum influenced by anisotropic alfvn mode shearing emerges . as for the pdfs , they are close to lognormal for sub - alfvnic turbulence even if the turbulence is supersonic
. for super - alfvnic turbulence the pdfs can be approximated by a lognormal distribution only for low sonic mach number .
the remarkable regularity of @xmath51 should have a physical explanation . a possible one
is related to the multiplicative symmetry with respect to density in the equations for isothermal hydrodynamics ( see * ? ? ?
this means that if a stochastic process disturbs the density , it results in perturbations of density being multiplied rather than added together .
consequently , the distribution of density for a gaussian driving of turbulence tends to be lognormal , which is consistent with our pdf measurements .
magnetic forces should affect the multiplicative symmetry above .
however , they do not affect the compressions of gas parallel to the magnetic field .
compressions in magnetically dominated fluid will be of the highest intensity and therefore most important
. they are sheared by alfvnic modes , as their own evolution will be slower than that imposed by alfvnic cascade ( see theoretical considerations in gs95 , * ? ? ?
we note that shearing does not affect the pdfs , but it does affect the spectra and the anisotropy of the turbulence .
this explains the close relationship between the properties of velocity in our simulations and those of density .
we confirm the existence for density of a rather mysterious property of turbulence called `` extended self - similarity '' .
while our spectra have a limited inertial range , the moments of density normalized by the third moment show nice power laws .
interestingly enough , we find that the extended self - similarity is not true if one uses the normalization over the third moment of velocity .
while we still may wonder about the origin of this phenomenon , the extended self - similarity is an indispensable tool for studyint turbulence intermittency .
our study of the intermittency of density and the logarithm of density reveals the dependence of the she - lvque exponents on both sonic and alfvn mach numbers .
moreover , we note a rather weak dependence on the intermittency when @xmath155 . for supersonic turbulence
the dependency is much stronger .
the intermittency changes with the strength of the external magnetic field may frequently be more important .
we observe less intermittent structures when @xmath23 is strong ( see figure [ fig : dens_expons ] ) .
the question of anisotropy of structures is in a natural way related to the presence of magnetic field .
when no magnetic field is present , there is a lack of any force that could organize these structures . on the other hand
, the introduction of the external magnetic field to the turbulent system induces the alfvnic mode , which is able to imprint its structure on random densities . in mhd turbulence ,
one direction is chosen , the direction of @xmath23 , so the anisotropy should be observed . in our sub - alfvnic simulations
the anisotropy of density is difficult to distinguish due to large dispersion of the structure functions .
the situation improves in the case of models with a weaker magnetic field . here
the results are in accordance with the gs95 model ( i.e. @xmath70 ) .
the anisotropy of the logarithm of density and velocity structures show similar gs95 behavior in all models with observed dependence on the scale and independence of @xmath2 .
however , the structures of @xmath51 and velocity for sub - alfvnic models seem to be less anisotropic at the small scales .
some basic properties of the density distribution in turbulence have been already studied .
however , most papers were focused mainly on the variance of distribution .
we analyzed here also two higher order statistical moments : skewness and kurtosis .
we observe a strong dependence of these moments on @xmath2 .
for instance , skewness grows with the sonic mach number for mid- and supersonic turbulence . for highly subsonic turbulence
we observe the growth of skewness but with a decreasing @xmath2 .
kurtosis in this case is zero and starts to grow rapidly when the value of @xmath2 exceeds about 0.7 .
combining these two measures can allow one to determine the value of @xmath2 in turbulence .
we do not observe any significant dependency of these moments on @xmath3 .
genus statistics reveal the changes of the structure of density with sonic mach number . for lower values of @xmath2 density
has a normal distribution , signifying roughly the same number of separate rarefactions and condensations . for higher sonic mach numbers ,
overdense structures start to dominate over the underdense ones .
this changes the structure of density significantly . for low sonic mach numbers , the structure of the density is spongelike . in the supersonic turbulence ,
the shocks produce separated high - density 1d and 2d structures embedded in the medium of much lower density . in this sense ,
the genus statistics confirm the topology with the intermittency .
many surveys provide extensive information on column density statistics . those include hi surveys , e.g. the galfa ( http://www.naic.edu/alfa/galfa/ ) survey of galactic hi distribution .
the main goal of another observational project , the wisconsin h - alpha mapper ( wham ) , is to provide a survey of h@xmath156 emission from the ism over the entire northern sky @xcite .
the observations provide the dispersion and emission measures , which can be interpreted as the column density and the integral of the square of the electron density along the entire los .
the survey supported by the numerical studies helps in better understanding the density and distribution of diffuse plasma in a major component of our galaxy , the warm ionized medium . combining data from different surveys and using the techniques in this paper ,
it is possible to get valuable insight into processes taking place in different ism phases .
if the observable column density distributions are concerned , our study confirms the earlier conclusion in @xcite that the pdfs are affected by the los distribution of density , and therefore , the relation between the column density pdfs and the underlying 3d pdfs is rather involved . on the contrary ,
the relation between the 3d and 2d density spectra is very straightforward , as was shown earlier by many authors ( see * ? ? ?
* ; * ? ? ?
this makes the observable column density spectrum a good measure of the underlying 3d density spectrum if the value of the spectral index is concerned .
however , the anisotropies of the density distribution that can be revealed for low sonic mach number by density itself and for high mach number by the distribution of @xmath157 are affected by the los integration that obtaining the column densities involves .
the problem here is that the gs95 anisotropies of the magnetic field and velocities that induce anisotropies of the density are defined in the system of reference aligned with the _ local _ magnetic field ( see discussion in * ? ? ?
los integration averages the local field directions , and therefore , only the anisotropies have to be calculated with respect to the mean magnetic field .
this both decreases the degree of anisotropy and erases the scale dependence of the anisotropy .
the residual anisotropy is the anisotropy of the turbulence at the injection scale .
therefore , for super - alfvnic turbulence for which magnetic fields are dynamically unimportant at the injection scale , the anisotropy of the column density statistics vanishes altogether .
@xcite studied maps of @xmath158co intensity of molecular clouds in taurus and perseus in terms of structure functions .
the integrated intensity maps can be translated into column densities by simple multiplication by a factor ( see * ? ? ?
@xcite show that the scaling exponents of the structure functions , normalized to the third order , follow the velocity scaling of the supersonic turbulence .
this suggests that the density scaling can be described by a similar hierarchical model , like the one proposed by @xcite .
our study shows that the intermittencies of densities and velocities are intrinsically different and one should exercise care in relating the relevant she - lvque exponents . although the intermittencies can reflect the sonic mach number , this dependence that we observe is weak for subsonic models and the changes are below the errors ( see figure [ fig : col_dens_scal ] ) .
the low - resolution models showed the weak dependence of the intermittency of density on the sonic mach number .
however , the intermittency is very sensitive to the resolution , which was stressed in [ sec : resolution ] .
the growth of error bars with the order of exponent @xmath11 is due to the amplification of any irregularities in the structure function by higher order moments .
another difficulty is the change of the intermittency while we switch from density to column density , as seen by the comparison of the scaling exponents in figures [ fig : dens_expons ] and [ fig : dens_int_expon ] .
we note , that the intermittency of the column densities in general is lower than the intermittency of density . in such situations we can only speak about some rough predictions taken from the observational data .
in addition , we see that the effects of cloud boundaries change the intermittency itself depending on @xmath2 . for instance , the highly subsonic and supersonic turbulence reveal more intermittent structure after including the boundary .
the effect is especially strong for turbulence with high @xmath2 and grows with its value .
the strength of the external magnetic field does not influence significantly the effects of the cloud boundary .
we also present a number of different measurements that can be helpful in the determination of the properties of density structures in real cosmic objects , like molecular clouds .
we compared the statistical measures applied for the 3d density and corresponding 2d column densities . by such comparison
, we were able to regain information about the spatial structure from the synthetic maps of the projected density .
for instance , we have shown , that the pdfs of density and column density can be related in particular cases .
we find that the statistical moments skewness and kurtosis , which are easy to obtain , can be very useful .
they can be employed to determine the sonic mach number for supersonic turbulence .
the subsonic regime of turbulence should be treated more carefully , because the cloud boundary effects might be very important in this case , as was shown in [
sec : boundary ] .
boundary effects influence the pdfs and the scaling of density structure ( see figures [ fig : dens_sph_pdfs ] and [ fig : dens_sph_expon ] , respectively ) . in the case of pdfs , however , some information can still be regained , because the boundary of the cloud mostly influences the low - density tail of the distribution .
three - dimensional density spectra can be recovered successfully from observations of column densities .
another promising technique for the topology of density and column density study is genus statistics .
although it is a completely different technique than one- or two - point statistics , we have shown that it can both help us with understanding the topology of density structures in turbulence and relate the results obtained for density and column density ( see [ sec : topology ] ) .
some measures presented in the article , like statistical moments and spectra , should be easily obtained from the molecular maps .
such observation of quantities can be directly compared with our numerical results , allowing quick determination of the conditions in investigated maps .
other measurements , like scaling exponents and genus statistics require more effort to calculate , but can provide more information related to the density structure .
thus , the techniques of the study of turbulence using 2d maps of observed column densities are very useful .
in this paper we investigated the statistics of density structure in compressible mhd turbulence using numerical experiments .
we analyzed pdfs , spectra , structure functions , and the intermittency of density and the logarithm of density for experiments with different sonic @xmath2 and alfvn @xmath3 mach numbers .
we addressed the issue of whether the numerical results can be compared with observations .
we found : * the amplitude of density fluctuations strongly depends on @xmath2 both in weakly and strongly magnetized turbulent plasmas .
1 . pdfs of density fluctuations have a lognormal distribution .
2 . the width of the pdf of @xmath51 grows with @xmath2 .
3 . the flattening observed in density spectra is due to the contribution of the highly dense small - scale structures generated in the supersonic turbulence .
* fluctuations of the logarithm of density are much more regular than those of density .
the logarithm of density exhibits the gs95 scalings and anisotropies .
the pdfs are close to gaussian .
* structure functions of density and the logarithm of density reveal self - similarity .
the density field is not self - similar to the velocity one . *
the intermittency of density structure depends on both @xmath2 and @xmath3 .
we report only weak dependence on @xmath2 for subsonic turbulence . in the case of supersonic models
the dependence is much stronger . *
additional care about resolution is necessary for intermittency studies .
the she - lvque exponents results can be affected by the resolution . *
similar to their 3d counterparts , the 2d measures of column densities , i.e. the variance , skewness and kurtosis , depend on @xmath2 and can be used to determine the value of @xmath2 from observations . *
the cloud boundary effects are important and should be accounted for by the study .
the most sensitive to the effects of cloud boundaries are the pdfs and intermittency studies . *
the genus statistic is sensitive to @xmath2 .
the 2d genus analysis is a promising tool for studying the topology of the ism .
a. l. acknowledges the discussions with annick pouquet .
the research of g. k. and a. l. is supported by nsf grant ast 03 - 07869 , the vilas professorship award , and the center for magnetic self - organization in astrophysical and laboratory plasmas ; a. b. is supported by an ice cube grant .
we thank the referee for valuable input .
armstrong , j. w. , rickett , b. j. & spangler , s. r. , 1995 , , 443 , 209 benzi , r. , ciliberto , s. , tripiccione , r. , baudet , c. , massaioli , f. & succi , s. , 1993 , , 48 , 29 beresnyak , a. , lazarian , a. & cho , j. , 2005 , , 624 , l93 , blc05 biskamp , d. , 2003 , _ magnetohydrodynamic turbulence _
, cambridge university press boldyrev , s. , nordlund , . & padoan , p. , 2002
, , 573 , 678 cho , j. & vishniac , e. t. , 2000 , , 539 , 273 cho , j. & lazarian , a. , 2002 , phys . rev .
letters , 88 , 245001 , cl02 cho , j. , lazarian , a. & vishniac , e. t. , 2002 , , 564 , 291 cho , j. & lazarian , a. , 2003 , , 345 , 325 , cl03 cho , j. , lazarian , a. & vishniac , e. t. , 2003 , lecture notes in physics , 614 , 56 cho , j. & lazarian , a. , 2004 , proc .
summer program 2004 , center for turbulence research , 75 , http://ctr.stanford.edu/sp04.html cho , j. & lazarian , a. , 2005 , theor . and comp .
dynam . , 19 , 127 cox , d. p. & smith , b. w. , 1974 , , 189 , l105 dickman , r. l. , 1985 , _ turbulence in molecular clouds _ , tucson , university of arizona press esquivel , a. & lazarian , a. , 2005 , , 631 , 320 falgarone , e. & perault , m. , 1988 , , 205 , l1 falgarone , e. , pineau des forets , g. & roueff , e , 1995 , , 300 , 870 falgarone , e. , hily - blant , p. , pety , j. & pineau des forts , g. , 2005 , aip conf .
proc . 784 : _ magnetic fields in the universe : from laboratory and stars to primordial structures .
_ , 299 goldreich , p. , sidhar , s. , 1995 , , 438 , 763 , gs95 gott , j. r. , melott , a. l. & dickinson , m. , 1986 , , 306 , 341 haffner , l. m. , reynolds , r. j. , tufte , s. l. , madsen , g. j. , jaehnig , k. p. , percival , j. w. , 2003 , , 149 , 405 hill , a. s. , benjamin , r. a. , reynolds , r. j. & haffner , l. m. , 2005 , bulletin of the american astronomical society , 1302 kallenberg , o. , 1997 , _ foundations of modern probability _
, new york : springer - verlag kaplan , s. a. & pikelner , s. b. , 1970 , _ the interstellar medium _ , cambridge : harvard university press kim , j. and ryu , d. , 2005 , , 630 , l45 lazarian , a. , 1995 , , 293 , 507 lazarian , a. & pogosyan , d. , 2000 , , 537 , 720 lazarian , a. , pogosyan , d. & esquivel , a. , 2002 , asp conf .
ser . 276 : _ seeing through the dust : the detection of hi and the exploration of the ism in galaxies _ lazarian , a. , 2004 , jkas , 37 , 563 lazarian , a. & beresnyak , a. , proceedings of the conference _ the magnetized plasma in galaxy evolution _ , krakw , jagiellonian university press , 2005 levy , d. , puppo , g. & russo , g. , 1999 , mathematical modelling and numerical analysis , 33 , 547 lithwick , y. & goldreich , p. , 2001
, , 562 , 279 liu , x .- d . &
osher , s. , 1998 , journal of computational physics , 141 , 1 mac low , m .- m . , klessen , r. s. , burkert , a. & smith , m. d. , 1998 , physical review letters , 80 , 2754 mac low , m .-
& klessen , r. s. , 2004 , rev . of mod .
phys . , 76 , 125 mckee , c. f. & ostriker , j. p. , 1977
, , 218 , 148 mckee , c. f. & tan , j. c. , 2002 , , 416 , 59 melott , a. l. , cohen , a. p. , hamilton , a. j. s. , gott , j. r. , weinberg , d. h. , 1989 , , 345 , 618 melott , a. l. , 1990 , , 193 , 1 , 1 montgomery , d. , brown , m. r. , matthaeus , w. h. , 1987 , , 92 , 282 mller , w. c. & biskamp , d. , 2000 , , 84 , 3 nordlund , . k. & padoan , p. , 1999
, _ interstellar turbulence _ , proceedings of the 2nd guillermo haro conference , cambridge university press novikov , e. a. , 1994 , , 50 , 3303 ossenkopf , v. , 2002 , , 391 , 295 ostriker , e. c. , stone , j. m. & gammie , c. f. , 2001 , , 546 , 980 ostriker , e. c. , 2003 , lnp , 614 , 252 padoan , p. , boldyrev , s. , langer , w. & nordlund , .
, 2003 , , 583 , 308 padoan , p. , jimenez , r. , nordlund , . & boldyrev , s. , 2004 , , 92 , 1102 passot , t. and vzquez - semadeni , e. , 1998 , , 58 , 4501 passot , t. and vzquez - semadeni , e. , 2003 , , 398 , 845 porter , d. , pouquet , a. & woodward , p. , 2002
, , 66 , 026301 scalo , j. , vazquez - semadeni , e. , chappell , d. & passot , t. , 1998 , , 504 , 835 she , z. & lvque , e. , 1994 , , 72 , 3 shebalin , j. v. , matthaeus , w. h. & montgomery , d. , 1983 , jplph , 29 , 525 stone , j. m. , ostriker , e. c. & gammie , c. f. , 1998 , , 508 , l99 stutzki , j. , 1999 , _ plasma turbulence and energetic particles in astrophysics _
, proceedings of the international conference , cracow ( poland ) , 5 - 10 september 1999 vzquez - semadeni , e. & garca , n. , 2001 , , 557 , 727 vestuto , j. g. , ostriker , e. c. & stone , j. m. , 2003 , , 590 , 858 weinberg , d. h. , 1988 , , 100 , 1373 cccccccccc b1p10 & 1.0 & 10.0 & 0.20@xmath159 & 0.62@xmath160 & @xmath161 & 25.0 & 16 & @xmath1620.06 & @xmath1620.6 + b1p9 & 1.0 & 9.0 & 0.21@xmath159 & 0.62@xmath160 & @xmath161 & 25.0 & 16 & @xmath1620.06 & @xmath1620.6 + b1p8 & 1.0 & 8.0 & 0.22@xmath159 & 0.62@xmath160 & @xmath161 & 25.0 & 16 & @xmath1620.06 & @xmath1620.6 + b1p7 & 1.0 & 7.0 & 0.24@xmath159 & 0.63@xmath160 & @xmath161 & 25.0 & 16 & @xmath1620.06 & @xmath1620.6 + b1p6 & 1.0 & 6.0 & 0.26@xmath159 & 0.63@xmath160 & @xmath161 & 25.0 & 16 & @xmath1620.06 & @xmath1620.6 + b1p5 & 1.0 & 5.0 & 0.28@xmath159 & 0.63@xmath160 & @xmath161 & 25.0 & 16 & @xmath1620.07 & @xmath1620.6 + b1p4 & 1.0 & 4.0 & 0.32@xmath163 & 0.64@xmath160 & @xmath161 & 25.0 & 16 & @xmath1620.08 & @xmath1620.6 + b1p3 & 1.0 & 3.0 & 0.37@xmath160 & 0.64@xmath164 & @xmath161 & 25.0 & 16 & @xmath1620.1 & @xmath1620.6 + b1p2 & 1.0 & 2.0 & 0.46@xmath164 & 0.65@xmath164 & @xmath161 & 25.0 & 16 & @xmath1620.1 & @xmath1620.6 + b1p1 & 1.0 & 1.0 & 0.66@xmath164 & 0.66@xmath164 & @xmath161 & 25.0 & 16 & @xmath1620.2 & @xmath1620.6 + b1p.9 & 1.0 & 0.9 & 0.70@xmath164 & 0.66@xmath164 & @xmath161 & 25.0 & 16 & @xmath1620.2 & @xmath1620.6 + b1p.8 & 1.0 & 0.8 & 0.74@xmath165 & 0.66@xmath164 & @xmath161 & 25.0 & 16 & @xmath1620.3 & @xmath1620.6 + b1p.7 & 1.0 & 0.7 & 0.80@xmath165 & 0.66@xmath164 & @xmath161 & 25.0 & 16 & @xmath1620.3 & @xmath1620.6 + b1p.6 & 1.0 & 0.6 & 0.86@xmath166 & 0.66@xmath164 & @xmath161 & 25.0 & 16 & @xmath1620.3 & @xmath1620.6 + b1p.5 & 1.0 & 0.5 & 0.94@xmath167 & 0.66@xmath164 & @xmath161 & 25.0 & 16 & @xmath1620.3 & @xmath1620.6 + b1p.4 & 1.0 & 0.4 & 1.05@xmath167 & 0.66@xmath164 & @xmath161 & 25.0 & 16 & @xmath1620.4 & @xmath1620.7 + b1p.3 & 1.0 & 0.3 & 1.21@xmath168 & 0.66@xmath164 & @xmath161 & 25.0 & 16 & @xmath1620.4 & @xmath1620.7 + b1p.2 & 1.0 & 0.2 & 1.47@xmath168 & 0.66@xmath160 & @xmath161 & 25.0 & 16 & @xmath1620.5 & @xmath1620.7 + b1p.1 & 1.0 & 0.1 & 2.1@xmath169 & 0.67@xmath165 & @xmath161 & 25.0 & 16 & @xmath1620.9 & @xmath1620.7 + b.1p10 & 0.1 & 10.0 & 0.23@xmath170 & 7.1@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.06 & @xmath1620.7 + b.1p9 & 0.1 & 9.0 & 0.24@xmath170 & 7.1@xmath172 & @xmath161 & 25.0 & 16 & @xmath1620.06 & @xmath1620.7 + b.1p8 & 0.1 & 8.0 & 0.25@xmath159 & 7.1@xmath172 & @xmath161 & 25.0 & 16 & @xmath1620.06 & @xmath1620.7 + b.1p7 & 0.1 & 7.0 & 0.27@xmath159 & 7.1@xmath172 & @xmath161 & 25.0 & 16 & @xmath1620.06 & @xmath1620.7 + b.1p6 & 0.1 & 6.0 & 0.29@xmath159 & 7.2@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.06 & @xmath1620.7 + b.1p5 & 0.1 & 5.0 & 0.32@xmath159 & 7.2@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.06 & @xmath1620.7 + b.1p4 & 0.1 & 4.0 & 0.36@xmath159 & 7.2@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.07 & @xmath1620.7 + b.1p3 & 0.1 & 3.0 & 0.42@xmath159 & 7.3@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.09 & @xmath1620.7 + b.1p2 & 0.1 & 2.0 & 0.52@xmath163 & 7.3@xmath172 & @xmath161 & 25.0 & 16 & @xmath1620.1 & @xmath1620.6 + b.1p1 & 0.1 & 1.0 & 0.72@xmath163 & 7.2@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.2 & @xmath1620.7 + b.1p.9 & 0.1 & 0.9 & 0.77@xmath163 & 7.3@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.3 & @xmath1620.7 + b.1p.8 & 0.1 & 0.8 & 0.82@xmath163 & 7.3@xmath169 & @xmath161 & 25.0 & 16 & @xmath1620.3 & @xmath1620.7 + b.1p.7 & 0.1 & 0.7 & 0.88@xmath163 & 7.4@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.3 & @xmath1620.7 + b.1p.6 & 0.1 & 0.6 & 0.94@xmath163 & 7.3@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.3 & @xmath1620.7 + b.1p.5 & 0.1 & 0.5 & 1.03@xmath163 & 7.3@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.4 & @xmath1620.7 + b.1p.4 & 0.1 & 0.4 & 1.15@xmath160 & 7.3@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.4 & @xmath1620.7 + b.1p.3 & 0.1 & 0.3 & 1.33@xmath160 & 7.3@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.5 & @xmath1620.7 + b.1p.2 & 0.1 & 0.2 & 1.66@xmath165 & 7.4@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.6 & @xmath1620.8 + b.1p.1 & 0.1 & 0.1 & 2.3@xmath173 & 7.4@xmath171 & @xmath161 & 25.0 & 16 & @xmath1620.8 & @xmath1620.8 + b1p8 & 1.0 & 8.0 & 0.23@xmath170 & 0.64@xmath159 & @xmath174 & 10.0 & 8 & @xmath1620.05 & @xmath1620.6 + b1p4 & 1.0 & 4.0 & 0.33@xmath170 & 0.65@xmath159 & @xmath174 & 15.5 & 14 & @xmath1620.1 & @xmath1620.7 + b1p1 & 1.0 & 1.0 & 0.68@xmath163 & 0.68@xmath163 & @xmath174 & 10.1 & 6 & @xmath1620.3 & @xmath1620.6 + b1p.1 & 1.0 & 0.1 & 2.20@xmath168 & 0.69@xmath163 & @xmath174 & 10.1 & 9 & @xmath1620.9 & @xmath1620.7 + b1p.01 & 1.0 & 0.01 & 7.0@xmath171 & 0.70@xmath160 & @xmath174 & 10.1 & 10 & @xmath1621.9 & @xmath1620.7 + b.1p8 & 0.1 & 8.0 & 0.26@xmath163 & 7.3@xmath175 & @xmath174 & 14.5 & 14 & @xmath1620.03 & @xmath1620.7 + b.1p4 & 0.1 & 4.0 & 0.36@xmath160 & 7.2@xmath175 & @xmath174 & 16.7 & 14 & @xmath1620.07 & @xmath1620.7 + b.1p1 & 0.1 & 1.0 & 0.74@xmath165 & 7.4@xmath176 & @xmath174 & 21.0 & 16 & @xmath1620.2 & @xmath1620.7 + b.1p.1 & 0.1 & 0.1 & 2.34@xmath167 & 7.4@xmath171 & @xmath174 & 22.3 & 17 & @xmath1620.8 & @xmath1620.7 + b.1p.01 & 0.1 & 0.01 & 7.1@xmath171 & 7.1@xmath171 & @xmath174 & 20.9 & 16 & @xmath1621.5 & @xmath1620.7 + b1p1 & 1.0 & 1.0 & @xmath1620.6 & @xmath1620.6 & @xmath177 & 7.2 & 1 & @xmath1620.3 & @xmath1620.6 + ccc|ccc 0.23@xmath178 & -2.3@xmath179 & -2.3@xmath179 & 0.26@xmath180 & -1.7@xmath179 & -1.7@xmath179 + 0.33@xmath178 & -2.2@xmath179 & -2.2@xmath179 & 0.36@xmath181 & -1.7@xmath179 & -1.7@xmath179 + 0.68@xmath180 & -2.0@xmath179 & -2.1@xmath179 & 0.74@xmath182 & -1.6@xmath183 & -1.6@xmath179 + 2.20@xmath180 & -1.3@xmath183 & -2.0@xmath183 & 2.34@xmath184 & -1.2@xmath183 & -1.6@xmath183 + 7.0@xmath179 & -0.5@xmath185 & -1.7@xmath183 & 7.1@xmath179 & -0.6@xmath183 & -1.5@xmath183 | we perform three - dimensional ( 3d ) compressible mhd simulations over many dynamical times for an extended range of sonic and alfvn mach numbers and analyze the statistics of 3d density and 2d column density , which include probability distribution functions , spectra , skewness , kurtosis , she - lvque exponents , and genus . in order to establish the relation between the statistics of the observables , i.e. column densities , and the underlying 3d statistics of density
, we analyze the effects of cloud boundaries .
we define the parameter space for 3d measures to be recovered from column densities .
in addition , we show that for subsonic turbulence the spectra of density fluctuations are consistent with @xmath0 in the case of a strong magnetic field and @xmath1 in the case of a weak magnetic field . for supersonic turbulence
we confirm the earlier findings of the shallow spectra of density and kolmogorov spectra of the logarithm of density .
we find that the intermittencies of the density and velocity are very different . |
the huge amount of active galactic nucleus ( agn ) energy is released through accretion onto super - massive black hole ( bh ) , supposed to exist in the center of agn .
the emission of the accretion disk is not only in the continuum , but also in the emission lines ( e.g. in fe k@xmath1 line ) and in low ionization lines , as e.g. in broad balmer emission lines which are seen as double peaked ( dp ) .
dp balmer lines are found in 20% of radio loud agn at @xmath2 @xcite and 4% of the sloan digital sky survey ( sdss ) quasars at @xmath3 @xcite .
broad , double - peaked emission lines of agn provide dynamical evidence for presence of an accretion disk feeding a supermassive black hole in the center of agn .
but in some cases , the variability of these lines shows certain irregularities which could not be explained just by standard model of accretion disk .
the dp line profiles are often used to extract the disk parameters ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in a series of papers dumont & collin - souffrin
( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) investigated the radial structure and emission of the outer regions of the optically thin accretion disks in agn and calculated detailed grid of photoionisation models in order to predict the relative strengths of low - ionization lines emitted from the disk .
they obtained integrated line intensities and line profiles emitted at each radius of the disk , for its different physical parameters .
they also studied the influence of the external illumination on the structure of the disk , considering the point source model , where a compact source of non - thermal radiation located at a given height illuminates the disk and the diffusion model , where the radiation of a central source is scattered back towards the disk by a hot diffusing medium .
one of the first methods for calculating the profiles of optical emission lines from a relativistic accretion disk was proposed by @xcite .
the limitation to this method is that the accretion disk structure required to explain the variability of the line profiles can not be axi - symmetric , i.e. very often the red peak is higher than blue one and that can not be explained by this model .
it is not possible in a circular disk , in which the blue peak is always doppler boosted to be stronger than the red peak .
therefore , @xcite adapted the circular accretion disk model to elliptical disks in order to fit the profiles of double - peaked emitters with a red peak stronger than the blue one .
this model introduced eccentricity and phase angle parameters to the circular model described above , and the pericenter distance of the elliptical orbits ( see * ? ? ?
spectroscopic monitoring of double - peaked emitters ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) has revealed that a ubiquitous property of the double - peaked broad emission lines is variability of their profile shapes on the timescales of months to years .
dp line profiles are observed to vary on timescales of months to years , i.e. on timescales of the order of the dynamical time or longer ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . this slow , systematic variability of the line profile is on the timescale of dynamical changes in an accretion disk , and has been shown to be unrelated to the shorter timescale variability seen in the overall flux in the line , due to reverberation of the variable ionizing continuum .
patterns in the variability of the broad balmer lines are often a gradual change and reversal of the relative strengths of the blue - shifted and red - shifted peaks ( see e.g. * ? ? ?
periodic variability of the red and blue peak strengths has also been attributed to a precessing elliptical disk , a precessing single - armed spiral ( as e.g. 3c 332 , 3c 390.3 : @xcite ; ngc 1097 : @xcite ) , and a precessing warp in the disk .
for instance , @xcite computed the profiles of balmer emission lines from a relativistic , warped accretion disk in order to explore the certain asymmetries in the double - peaked emission line profiles which can not be explained by a circular keplerian disk .
elliptical disks and spiral waves have been detected in cataclysmic variables @xcite , and a radiation induced warp has been detected in the large - scale disk of the agn ngc 4258 @xcite .
spiral waves are a physically desirable model since they can be produced by instability in the vicinity of a black hole .
they can play an important role in accretion disks , because they provide a mechanism for transporting angular momentum outward in the disk , allowing the gas to flow inwards , towards the central black hole .
long - term profile variability is thus a useful tool for extracting information about the structure and dynamics of the accretion disk most likely producing the double - peaked emission lines . in this paper
, we present an investigation of the disk line variations due to instability in accretion disk .
first we developed a model , assuming that instability in the accretion disk affects disk emissivity .
this model and some simulations of expected line profile variability are presented in 2 . in
3 we compare the model with observations taken from long - term monitoring of 3c 390.3 @xcite in order to obtain parameters of perturbations . in
4 we discuss our results in the light of possible physical mechanisms which could cause such perturbations , and finally , in 5 we outline our conclusions .
here we introduce the model and some approximations used in the simulations of accretion disk perturbation . as we mentioned above
, the dp line profile variability does not appear to correlate with changes in the line and/or continuum flux , and consequently one can assume that the changes in line profile are likely caused by changes in the accretion disk structure .
there are several examples of the long - term variability ( on timescales of several years ) of the dp line profile of some objects which has been successfully modeled by the precession of a non - axisymmetric accretion disk , such as an elliptical disk or a disk with a spiral arm ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) .
these models , however , fail to explain the long - term variability of some objects and the short - term variability ( on timescales from several months to a year ) of all objects @xcite .
for instance , @xcite found that the two simple models , an elliptical accretion disk and a circular disk with a spiral arm , are unable to reproduce all aspects of the observed variability , although both account for some of the observed behaviors .
therefore , these authors suggest that many of the observed variability patterns could be reproduced assuming a disk with one or more fragmented spiral arms . other attempts to explain the dp line profile variability through perturbations of the disk structure introduced bright spots over an axisymmetric accretion disk . as an example @xcite successfully modeled the variation of the h@xmath1 peak intensity ratio of arp 102b with a single spot rotating within the disk , but @xcite were not able to apply the same model to the same object at a different time period . in the case of fe
k@xmath1 variability @xcite used the spot model to explain the variability of the iron line profile of mrk 766 in the x - ray band .
also , @xcite studied variations of the iron line due to an orbiting spot which arise by reflection on the surface of an accretion disk , following its illumination by an x - ray flare in form of an off - axis point - like source just above the accretion disk . besides the spots in accretion disk , the fe k@xmath1 line of some agn could be also significantly affected by highly - ionized fast accretion disk outflows . for instance
, @xcite found that the major features in the observed 2 10 kev spectrum of the bright quasar pg1211 + 143 can be well reproduced by their monte carlo radiative transfer simulations which include a variety of disk wind ( outflow ) models . to explain the short - timescale variability of the dp line profiles @xcite constructed stochastically perturbed accretion disk models and calculated h@xmath1 line profile series as the bright spots rotate , shear and decay .
they ruled out spot production by star / disk collisions and favor a scenario where the radius of marginal self - gravity is within the line emitting region , creating a sharp increase in the radial spot distribution in the outer parts .
we model the emission from accretion disk using numerical simulations based on ray - tracing method in kerr metric ( see e.g. * ? ? ?
* and references therein ) .
although this method was developed for studying the x - ray radiation which originates from the inner parts of the disk close to the central black hole ( see e.g. * ? ? ?
* ) , it can be also successfully applied for modelling the uv / optical emission which originates from the outer regions of the disk line originates , these effects are negligible and kerr metric with zero angular momentum , i.e. schwarzschild metric , is a very good approximation . ] .
surface emissivity of the disk is usually assumed to vary with radius as a power law ( e.g. * ? ? ?
* ) : @xmath4 where @xmath5 is an emissivity constant and @xmath6 emissivity index .
total observed flux is then given by : @xmath7 where @xmath8 is the energy shift due to the relativistic effects : @xmath9 , @xmath10 is the rest energy of the line , @xmath11 is the local turbulent broadening and @xmath12 is the solid angle subtended by the disk in the observer s sky . in this paper
we adopt the following modification of the power - law disk emissivity , in order to introduce a bright spot like perturbing region in the disk @xcite : @xmath13 where @xmath14 is the modified disk emissivity at the given position @xmath15 expressed in gravitational radii @xmath16 , @xmath17 is the ordinary power - law disk emissivity at the same position , @xmath18 is emissivity of the perturbing region ( i.e. amplitude of the bright spot ) , @xmath19 is the position of perturbing region with respect to the disk center ( in @xmath16 ) and @xmath20 are its widths ( also in @xmath16 ) . a 3d plot of above expression for modified emissivity law is given in fig .
[ fig1 ] . ) for @xmath21 , @xmath22 and for the following parameters of perturbing region : @xmath23 , @xmath24 , @xmath25 and @xmath26.,scaledwidth=40.0% ]
this simple model is suitable for our purpose because it allows us to change amplitude , width and location of bright spot in respect to the disk center . in that way we are able to simulate displacement of bright spot along the disk , its widening and amplitude decrease with time ( decay ) .
moreover , the above bright spot model can be successfully applied for studying the variations of accretion disk emission in different spectral bands , from x - rays to optical band ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in order to test how this model of bright spot affects the h@xmath0 line profile , we performed several numerical simulations of perturbed emission of an accretion disk in kerr metric for different positions of bright spot along @xmath27 and @xmath28-axes in both , positive and negative directions . for these simulations we adopted the following parameters of the disk : inclination @xmath29 ,
inner and outer radii @xmath30 and @xmath31 , power law emissivity with index @xmath32 , local turbulent broadening @xmath33 and normalized angular momentum of black hole @xmath34 .
the corresponding results are presented in fig .
[ fig2 ] . as it can be seen from this figure , when the bright spot moves along the positive direction of @xmath27-axis ( receding side of the disk )
it affects only `` red '' wing of the line ( fig .
[ fig2 ] , top right ) , but when it moves along the negative direction of @xmath27-axis ( approaching side of the disk ) it affects only `` blue '' wing of the line ( fig .
[ fig2 ] , top left ) . in both cases ,
the other wing and the line core stay nearly constant , and therefore almost unaffected by bright spot .
the situation is quite opposite when the bright spot moves in both directions along the @xmath28-axis , because then it affects only the line core , while the both of its wings stay almost intact ( see the bottom panels of fig .
[ fig2 ] ) .
+ we also performed the corresponding simulations for different positions of a bright spot which moves from the inner radius of the accretion disk towards its outer parts along the @xmath35 direction , and found similar behavior of the simulated line profiles ( see fig .
[ fig3 ] ) . as one can see from fig .
[ fig3 ] , for certain positions of perturbing region along @xmath35 direction we obtained the line profile with almost symmetrical wings , while in other cases either the `` blue '' peak is brighter than the `` red '' one , or the `` red '' peak is stronger than the `` blue '' one .
the next step in our analysis was to use our numerical simulations for fitting the observed spectra of 3c 390.3 in order to study the variability of its h@xmath0 spectral line due to emissivity perturbations in its accretion disk . to test the model , we used 22 spectra of 3c 390.3 observed from november 1995 until june 1999 ( see fig . 8 in * ? ? ? * ) .
spectra of 3c 390.3 were taken with the 6 m and 1 m telescopes of the sao ras ( russia , 19952001 ) and at inaoe s 2.1 m telescope of the `` guillermo haro observatory '' ( gho ) at cananea , sonora , mxico ( 19981999 ) in monitoring regime in 1995 - 1999 .
they were obtained with long slit spectrographs , equipped with ccd detector arrays .
the typical wavelength interval covered was that from 4000 to 7500 , the spectral resolution varied between 5 and 15 and the s / n ratio was @xmath36 50 in the continuum near h@xmath1 and h@xmath0 .
spectrophotometric standard stars were observed every night .
the spectrophotometric data reduction was carried out either with software developed at sao ras or with the iraf package for the spectra obtained in mxico .
the image reduction process included bias , flat - field corrections , cosmic ray removal , 2d wavelength linearization , sky spectrum subtraction , addition of the spectra for every night , and relative flux calibration based on standard star observations .
spectra were scaled by the [ ] @xmath374959 + 5007 integrated line flux under the assumption that the latter did not change during the time interval covered by our observations ( 19952001 ) .
a value of 1.7@xmath3810@xmath39 ergs s@xmath40@xmath41 @xcite for the integrated [ ] line flux was adopted . in order to calculate a normalization coefficient ,
the continuum was determined in two 30 wide clean line free windows centered at 4800 and 5420 . after continuum subtraction , blend separation of the h@xmath0 and [ ]
components was carried out by means of a gaussian fitting procedure , applied to the following : h@xmath0 broad blue , broad red and central narrow ; [ ] @xmath374959,5007 broad and narrow components .
the forbidden lines are represented by two gaussian curves with an intensity ratio i(5007)/i(4959)=2.96 .
comparisons between mean and rms spectra of the h@xmath1 and h@xmath0 broad line profiles of 3c 390.3 during 1995 - 2007 ( including the h@xmath0 spectra from this paper ) are given in fig .
12 of @xcite .
it can be seen that these profiles are similar , and moreover , the corresponding h@xmath1 profiles of 3c 390.3 from @xcite and @xcite are also similar .
these comparisons indicate that the [ ] @xmath374959,5007 narrow lines subtraction in h@xmath0 region was performed correctly .
the flux of h@xmath0 and the broad component profile was obtained from scaled spectra after continuum subtraction and removal of the [ ] doublet and narrow h@xmath0 component .
then the observed continuum fluxes were corrected for the aperture effects using scheme by @xcite . the mean error ( uncertainty ) in our flux determinations for the continuum and h@xmath0 flux is @xmath423% .
more details can be found in paper @xcite .
radial velocities of the blue and red peaks of the h@xmath0 and h@xmath1 broad lines of 3c 390.3 vary with time ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
@xcite obtained the h@xmath0 difference profiles by subtracting the average spectrum corresponding to the minimum activity state ( september 9 , 1997 ) from the individual spectra ( see their fig .
these authors found ( see their table 7 ) that the radial velocity of the blue peak increased from -3200 km s@xmath40 in 1995 - 1996 to -5200 km s@xmath40 in 1999 . at the same time the radial velocity of the red peak increased from + 4900 km s@xmath40 in 1995 - 1996 to + 7000 km s@xmath40 in 1999 .
here we analyze the possibility that the velocities , corresponding to the peak shifts in the h@xmath0 integral and difference profiles , vary with time due to perturbations in disk emissivity . in order to fit the spectral h@xmath0 line shapes of 3c 390.3 , we first estimated the disk parameters from several profiles .
we found following parameters of the disk : inclination @xmath43 , inner and outer radii @xmath44 and @xmath45 , broken power law emissivity with index @xmath32 for @xmath46 and @xmath47 for @xmath48 , radius at which slope of emissivity changes @xmath49 , emissivity of perturbing region @xmath50 , local turbulent broadening @xmath33 and normalized angular momentum of black hole @xmath34 .
these values of parameters are in accordance with the corresponding parameters for 3c 390.3 obtained by @xcite , who found the following values : @xmath51 , @xmath52 , @xmath53 and @xmath54 . in all our simulations we held emissivity of perturbing region fixed at @xmath50 , but
nevertheless , the brightness of perturbing region varies with time due to its changeable area ( i.e. its widths are taken as free parameters which vary with time ) , as well as due to power law emissivity of the disk which decreases with radius .
we studied stationary and moving perturbing regions , and in the latter case their coordinates are also assumed as free parameters .
the fitting of the observed h@xmath0 line shapes of 3c 390.3 is performed separately in the case of moving and stationary perturbing regions in the following three steps : ( i ) the simulated h@xmath0 line profiles are calculated for some initial set of values of free parameters ; ( ii ) the corresponding root mean square ( rms ) of residuals between the observed and simulated line profiles is calculated ; ( iii ) the values of free parameters are varied and the procedure ( i)-(iii ) is repeated until the rms deviations ( rmsd ) between the observed and fitted h@xmath0 profiles become as small as possible .
+ + + + + + fig .
[ fig4 ] . shows comparisons between all 22 observed spectra ( black solid line ) and the obtained best fits assuming the moving ( red solid line ) and stationary ( blue dashed line ) perturbing regions , while the corresponding positions of the perturbing regions are presented in fig .
[ fig5 ] . as it can be seen from fig .
[ fig5 ] , the obtained positions of the moving perturbing regions are distributed in the form of two spiral arms , indicating that these perturbations originate in the inner regions of the disk and spiral away towards its outer parts , moving faster in the azimuthal direction as they get further away from the center of the disk .
table [ tab1 ] .
contains the obtained values of the fitted parameters , as well as the corresponding rmsd in the case of both , moving and stationary perturbing regions .
we first studied the moving perturbations , but after examining the obtained results we found that , during a period of a few years , such perturbations move only by small distances which are comparable to their widths ( see table [ tab1 ] ) .
therefore , in order to test whether the obtained displacements are reliable , we repeated the fitting , but this time assuming the stationary perturbing regions with variable widths . in this case
we obtained the best fits for two perturbations positioned at @xmath55 , @xmath56 and at @xmath57 , @xmath58 , respectively ( denoted by crosses in fig .
[ fig5 ] ) . as one can see from fig .
[ fig4 ] . and table [ tab1 ] , regardless the significant variations of the h@xmath0 line profile during the analyzed period , both models of perturbing region resulted in similar fits for the most of these spectra , except for the spectra observed during 1999 , where the moving perturbing region achieved better fits .
the latter result should be taken with caution because the observations from 1999 were performed after a large gap of @xmath59 1 year .
therefore , the displacements of the perturbing regions can not be considered as indisputably confirmed , but on the contrary , their widths almost certainly vary with time .
the last conclusion is valid for both , moving and stationary perturbing regions , since neither of them can provide satisfactory fits with fixed widths .
the obtained best fit positions of both , moving and stationary perturbations are located on the approaching side of the accretion disk ( see fig .
[ fig5 ] ) and these perturbations can be most likely attributed to successive occurrences of two different bright spots .
this assumption is in good agreement with observations , since two large amplitude outbursts of h@xmath0 line are observed during the analyzed period @xcite , and therefore each bright spot can be assigned to one of them : the bright spot which positions are denoted by squares corresponds to october 1994 - july 1997 outburst while the other one , which positions are denoted by triangles , corresponds to july 1997 - june 1999 outburst . using the time differences between two successive observed spectra we were able to estimate the speeds of both moving bright spots ( see table [ tab1 ] ) . for an average velocity of the first bright spot we obtained the value of 7298 @xmath60 and for the second one 6575 @xmath60 . as it can be seen from table [ tab1 ]
, widths of bright spots are increasing with time , indicating that they decay until they completely disappear .
it should be noted that , inevitably , there is a certain degree of degeneracy in the parameter space , since similar results could be obtained with somewhat different combinations of perturbing region positions and widths .
lccccccccc 1995 nov 17 & 50039.156 & -30 & 175 & 100 & 100 & & & 0.09730 & 0.09460 + 1996 feb 14 & 50127.602 & -40 & 200 & 105 & 130 & 26.93 & 5203.24&0.09680 & 0.09900 + 1996 mar 20 & 50162.580 & -50 & 208 & 106.5 & 140 & 12.81 & 6257.63&0.08476 & 0.09138 + 1996 jul 12 & 50276.567 & -75 & 230 & 110 & 180 & 33.3 & 4993.36&0.08518 & 0.08178 + 1996 jul 17 & 50281.434 & -76.3 & 231.3 & 110 & 180 & 1.84 & 6456.24&0.08820 & 0.08249 + 1996 aug 10 & 50305.489 & -90 & 235 & 112 & 190 & 14.19&10082.9 & 0.09435 & 0.09201 + 1996 sep 11 & 50338.309 & -115 & 238 & 120.5 & 200 & 25.18&13112.6 & 0.12337 & 0.12426 + 1997 mar 04 & 50511.622 & -165 & 245 & 130 & 260 & 50.49 & 4978.94&0.11448 & 0.12082 + 1997 aug 30 & 50691.463 & -160 & 100 & 100 & 100 & & & 0.13717 & 0.13593 + 1997 sep 10 & 50701.576 & -165 & 105 & 105 & 110 & 7.07 & 11950.55 & 0.12034 & 0.11905 + 1997 dec 30 & 50813.195 & -205 & 120 & 145 & 175 & 42.72 & 6541.48 & 0.09499 & 0.09398 + 1998 jan 22 & 50835.631 & -215 & 122.5 & 155 & 195 & 10.31 & 7852.39 & 0.10690 & 0.10949 + 1998 feb 23 & 50867.560 & -225 & 125 & 165 & 215 & 10.31 & 5517.75 & 0.09825 & 0.09618 + 1998
may 06 & 50940.354 & -245 & 129 & 185 & 265 & 20.40 & 4788.88 & 0.08867 & 0.08965 + 1998 jun 25 & 50990.302 & -265 & 132 & 205 & 285 & 20.22 & 6920.32 & 0.07229 & 0.07387 + 1998 jul 16 & 51010.719 & -275 & 134 & 210 & 295 & 10.20 & 8537.04 & 0.06809 & 0.07141 + 1998 jul 25 & 51019.723 & -278 & 134.5 & 212 & 300 & 3.04 & 5773.22 & 0.06548 & 0.06615 + 1998 aug 30 & 51055.551 & -285 & 135.5 & 215 & 320 & 7.07 & 3373.22 & 0.07590 & 0.07638 + 1998 sep 26 & 51082.429 & -293 & 136 & 218 & 340 & 8.01 & 5097.09 & 0.07948 & 0.07788 + 1999 aug 19 & 51410.309 & -500 & 140 & 250 & 400 & 207.04&10792.43 & 0.10645 & 0.11163 + 1999 sep 04 & 51426.208 & -505 & 140 & 252 & 400 & 5 & 5375.06 & 0.09147 & 0.09232 + 1999 oct 03 & 51455.172 & -510 & 140 & 254 & 400 & 5 & 2950.49 & 0.09977 & 0.10270 +
several physical mechanisms could be responsible for perturbations in accretion disk emissivity , i.e. for bright spot formations .
the most plausible candidates for such mechanisms are : disk self - gravity , baroclinic vorticity , disk - star collisions ( * ? ? ?
* and references therein ) , tidal disruptions of stars by central black hole ( * ? ? ?
* and references therein ) and fragmented spiral arms ( * ? ? ?
* and references therein ) .
the disk self - gravity , driven by jeans instability , could cause production of clumps in the disk which have typical sizes in the range from 10 to 1000 gravitational radii for a @xmath61 central black hole .
such clumps do not shear with differential rotation and they have high brightness that varies very little over time . since the obtained results indicate that bright spots decay by time and spiral along the disk , it is not likely that these bright spots could be identified as clumps created by self - gravity , although their sizes are comparable .
baroclinic vorticity appears in the accretion disk due to its differential rotation in combination with the radial temperature gradient , causing the material in the disk to spiral around the center of vortex .
such a vortex would have higher density , and hence higher brightness , causing the formation of a bright spot .
the typical sizes of such spots , as well as their shearing with differential rotation of the disk , are still unknown since different numerical simulations gave contradictory results ( for more details see e.g. * ? ? ?
* and references therein ) .
therefore , in the case of 3c 390.3 , this mechanism still can not be neither ruled out , nor accepted .
disk - star collisions are assumed to be very frequent events which happen on daily timescales and which could increase disk surface temperature in the region of collision , and thus , create a bright spot .
such bright spots shear with differential rotation of the disk and decay as the material cools down . however , the typical size of a such bright spot immediately after collision is close to the size of the star , which is very small when expressed in gravitational radii . therefore ,
neither this mechanism could be accepted as a potential cause of two bright spots , detected in the case of quasar 3c 390.3 .
the tidal disruption of stars by central black hole ( * ? ? ?
* and references therein ) happens when a star passes the tidal radius of the black hole , i.e. when the black hole s tidal gravity exceeds the star s self - gravity .
gas of a disrupted star falls back to the black hole at a super - eddington rate , releasing a flare of energy which then blows away a significant fraction of the falling gas as an outflow . such super - eddington flares and outflows could induce instabilities in accretion disk in form of bright spots .
however , this mechanism is an unlikely candidate for a potential cause of the bright spots in the case of 3c 390.3 due to the following weaknesses : ( i ) super - eddington outflows are short living ( @xmath62 days ) ; ( ii ) frequency of star disruptions in a typical elliptical galaxy is very low , between @xmath63 and @xmath64 per year , and in the case of the black hole of 3c 390.3 which mass is @xmath65 @xcite , it is near the low end of this range ( see e.g. * ? ? ? * ; * ? ? ?
* ) ; ( iii ) any stellar debris could be hardly released in the case of 3c 390.3 since the main sequence stars are disrupted within the innermost stable orbit around a non - rotating black hole as massive as that in 3c 390.3 ; ( iv ) it is unclear how the bright spots produced by this mechanism should evolve with time .
spiral arms in agn disks could be formed spontaneously due to self - gravity instabilities ( see e.g. * ? ? ?
* and references therein ) or could be triggered by close passage of some massive object such as another supermassive black hole or a star cluster ( see e.g. * ? ? ?
* and references therein ) .
spiral arms increase flux variability of agn on timescales of a year to several years , but as noted before , they are unable to reproduce all aspects of the observed variability . however , they are also subjected to fragmentation , causing small variations in the flux on timescales of several months .
the fragments in spiral arms can be due to sub - structures in a non - uniform accretion disk , such as isolated clumps which could pass through the arm and dominate in its emissivity , causing the discrete `` lumps '' of excess emission @xcite . the observed variability on timescales from few months to several years in the difference spectra of some agn is probably caused by such lumps .
it is quite possible that some of these lumps are long - living and that they do not vary significantly in strength , shape , or position over a period of several years @xcite . as the obtained results show , it is the same case with the large bright spots which are responsible for amplitude outbursts of the 3c 390.3 h@xmath0 line , because they have constant emissivity and they are either stationary or spiralling over small distances during the period of several years .
the only feature which significantly varies with time is their width .
therefore , these bright spots could be most likely explained by the emissivity lumps , caused by fragments in spiral arms of the accretion disk .
we developed a model of the disk perturbing region in the form of a single bright spot ( or flare ) by a modification of the power law disk emissivity and used this model to simulate the disk line profiles .
this model has been used to fit the observed h@xmath0 line of 3c 390.3 observed from 1995 to 1999 . from this investigation
we can point out the following results : 1 .
the model which includes perturbation ( bright spot ) in the accretion disk can successfully explain difference in double peaked line profiles , as e.g. higher red peak even if we have the standard circular disk .
the position of a bright spot has a stronger influence on one particular part of spectral line profiles ( such as e.g. its core if the spot is in the central part of the disk , or `` red '' and `` blue '' wings if the spot is located on receding and approaching part , respectively ) .
using the model for perturbing region we were able to successfully model and reproduce the observed variations of the h@xmath0 line profile in the case of 3c 390.3 , including the two large amplitude outbursts observed during the analyzed period .
therefore , the observed variations of the 3c 390.3 h@xmath0 line could be caused by perturbations in the disk emissivity .
we found that two outbursts referred by @xcite could be explained by successive occurrences of two different bright spots on approaching side of the disk which are either moving , originating in the inner regions of the disk and spiralling outwards , or stationary . both bright spots decay by time until they completely disappear .
our results support hypothesis that the perturbations in accretion disk emissivity are probably caused by fragments in the spiral arms of the disk .
the results presented above show that a circular disk with perturbations ( bright spots ) can be applied to explain different double peaked line profiles , and can be also used to trace perturbations ( as well as their characteristics ) from the broad double peaked line shapes .
this work is supported by the ministry of science of serbia through project ( 146002 ) `` astrophysical spectroscopy of extragalactic objects '' and with studentship for m. stalevski .
it is also supported by intas ( grant n96 - 0328 ) , and rfbr ( grants n97 - 02 - 17625 , 09 - 02 - 01136 ) . the authors would like to thank the anonymous referee for very valuable and helpful comments and suggestions .
baptista , r. & cataln , m. s. 2000 , , 44 , 1p chen , k. , halpern , j. p. & filippenko , a. v. 1989 , , 339 , 742 chen , k. & halpern , j. p. 1990
, , 354 , l1 collin - souffrin , s. 1987 , , 179 , 60 collin - souffrin , s. , & dumont , a. m. 1990 , , 229 , 292 doviak , m. , karas , v. , matt , g. , goosmann , r. w. 2008 , , 384 , 361 dumont , a. m. , & collin - souffrin , s. 1990 , , 229 , 302 dumont , a. m. , & collin - souffrin , s. 1990 , , 229 , 313 dumont , a. m. , & collin - souffrin , s. 1990 , , 83 , 71 eracleous , m. & halpern , j. p. 1994
, , 90 , 1 eracleous , m. , livio , m. , halpern , j. p. & storchi - bergmann , t. 1995 , , 438 , 610 eracleous , m. , halpern , j. p. , gilbert , a. m. , newman , j. a. , & filippenko , a. v. 1997 , , 490 , 216 eracleous , m. & halpern , j. p. 2003
, , 599 , 886 eracleous , m. , lewis , k. t. & flohic , h. m. l. g. 2009 , new astronomy review , 53 , 133 flohic , h. m. l. g. , eracleous , m. 2008 , , 686 , 138 gaskell , c. m. 1996 , , 464 , l107 gezari , s. , halpern , j. p. , eracleous , m. 2007 , , 169 , 167 gilbert , a. m. , eracleous , m. , filippenko , a. v. & halpern , j. p. 1999 , in asp conf .
ser . 175 : structure and kinematics of quasar broad line regions , ed .
gaskell , c. m. , brandt , w. n. , dietrich , m. , dultzin - hacyan , d. & eracleous , m. ( san francisco : asp ) , 189 jovanovi ' c , p. , popovi ' c , l. .
2009a , arxiv:0903.0978v1 [ astro-ph.ga ] ( to appear as a chapter in the book `` black holes and galaxy formation '' by nova science publishers , inc .
jovanovi ' c , p. , popovi ' c , l. .
2009b , pasrb , 9 , 45 jovanovi ' c , p. , popovi ' c , l. .
2008 , fortschritte der physik , 56 , 456 jovanovi ' c , p. , popovi ' c , l. .
2008 , pobeo , 84 , 467 lewis , k. t. 2005 , bulletin of the american astronomical society , 37 , 1240 lewis , k. t. , & eracleous , m. 2006 , , 642 , 711 lewis , k. t. , eracleous , m. , & storchi - bergmann , t. 2010 , , 187 , 416 magorrian , j. , & tremaine , s. 1999 , , 309 , 447 maloney , p. r. , begelman , m. c. & pringle , j. e. 1996 , , 472 , 582 marziani , p. , sulentic , j. w. , calvani , m. , perez , e. , moles , m. & penston , m. v. 1993 , , 410 , 56 newman , j. a. , eracleous , m. , filippenko , a. v. & halpern , j. p. 1997
, , 485 , 570 peterson , b. m. , pogge , r. w. , wanders , i. , smith , s. m. , romanishin , w. 1995 , , 107 , 579 popovi ' c , l. . , mediavilla , e. g. , jovanovi ' c , p. , muoz , j. a. 2003 , , 398 , 975 romano , p. , marzianii , p. & dultzin - hacyan , d. 1998 , , 495 , 222 sergeev , s. g. , pronik , v. i. & sergeeva , e. a. 2000 , , 356 , 41 shapovalova , a. i. et al .
2001 , , 376 , 775 shapovalova , a. i. et al .
2009 , new astronomy review , 53 , 191 shapovalova , a. i. , et al .
2010 , arxiv:1003.2098 .
sim , s. a. , miller , l. , long , k. s. , turner , t. j. , reeves , j. n. 2010 , arxiv:1002.0544v1 [ astro-ph.he ] stalevski , m. , jovanovi ' c , p. , popovi ' c , l. .
2008 , pobeo , 84 , 491 steeghs , d. , harlaftis , e. t. & horne , k. 1997 , , 290 , l2 strateva , i. v. et al .
2003 , , 126 , 1720 strateva , i. v. brandt , w. n. , eracleous , m. , garmire , g. 2008 , , 687 , 869s storchi - bergmann , t. et al .
2003 , , 598 , 956 strubbe , l. e. , quataert , e. 2009 , , 400 , 2070 turner , t. j. , miller , l. , george , i. m. , reeves , j. n. 2006 , , 445 , 59 veilleux , s. & zheng , w. 1991 , , 377 , 89 wang , j. , & merritt , d. 2004 , , 600 , 149 wu , s .-
, wang , t .-
, & dong , x .- b .
2008 , , 389 , 213 zheng , w. , veilleux , s. & grandi , s. a. 1991 , , 381 , 41 | here we show that in the case when double peaked emission lines originate from outer parts of accretion disk , their variability could be caused by perturbations in the disk emissivity . in order to test this hypothesis
, we introduced a model of disk perturbing region in the form of a single bright spot ( or flare ) by a modification of the power law disk emissivity in appropriate way .
the disk emission was then analyzed using numerical simulations based on ray - tracing method in kerr metric and the corresponding simulated line profiles were obtained .
we applied this model to the observed h@xmath0 line profiles of 3c 390.3 ( observed in the period 1995 - 1999 ) , and estimated the parameters of both , accretion disk and perturbing region .
our results show that two large amplitude outbursts of the h@xmath0 line observed in 3c 390.3 could be explained by successive occurrences of two bright spots on approaching side of the disk .
these bright spots are either moving , originating in the inner regions of the disk and spiralling outwards by crossing small distances during the period of several years , or stationary . in both cases ,
their widths increase with time , indicating that they most likely decay . |
it is now known that our universe is hierarchical .
the largest virialized objects in the universe are clusters that are made of hundreds of galaxies .
clusters have extensively been used to test structure formation models in the universe , through their abundances and correlations@xcite .
superclusters are defined as an ensemble of clusters .
they are still assembling , and are even larger structure than clusters ; superclusters have extents of @xmath0 or even larger .
thus superclusters offer us information on late evolution of universe , and also on the transition from linear to non - linear regime .
examples of the superclusters include a901/902@xcite , ms0302 + 17@xcite , cl1604 + 43@xcite , the sculptor supercluster@xcite , and the shapley supercluster@xcite .
detailed investigations have shown that these superclusters are real physical systems where clusters are gravitationally assembling .
detailed studies of these superclusters from galaxy distributions and weak lensing also showed that the mass distribution of superclusters is in good agreement with the distribution of early - type galaxies in superclusters .
the physical state of gas in superclusters is still unknown , but it is worth mentioning that soft x - ray has been detected in the sculptor supercluster@xcite .
although most the studies have been limited to those of individual superclusters so far , the emergence of wide - field surveys / telescopes , such as the sloan digital sky survey ( sdss ) and the subaru suprime - cam , begins to make it possible to do _
statistical _ studies of superclusters .
nevertheless , there is no analytic model of superclusters that predicts , e.g. , the mass distribution . in this paper
, we construct such an analytic model of superclusters . basically , our model is a modification of the press - schechter theory@xcite , and predicts not only mass distributions of superclusters but also number distribution of clusters in each supercluster .
we also compare our model with superclusters in the sdss extracted by einasto _
et al._@xcite . throughout the paper
we assume a flat universe @xmath1 , and the hubble constant @xmath2 .
in this section , we construct the mass function of superclusters and clusters in each supercluster using the spherical model@xcite . we regard superclusters as overdense regions that are on the course of the spherical collapse .
then superclusters can be defined as the regions that have the overdensity @xmath3 , where @xmath4 is the average density within superclusters and @xmath5 is the mean density in the universe .
although superclusters are usually characterized by filamentary structure that is quite different from simple spherical overdensity , we consider this model so as to be analytically tractable .
the assumption , however , may not be so crude given the similar situation that the press - schechter theory has achieved reasonable success despite its simplicity . ) and corresponding nonlinear overdensities ( @xmath6 ) as a function of dimensionless cosmic time @xmath7 .
different lines mean different values of @xmath8 : from @xmath9 ( _ dark _ ) to @xmath10 ( _ light _ ) .
the vertical dashed line indicates the value of @xmath7 corresponding to @xmath11 . in this plot
, we adopt @xmath12 .
[ fig : sc],scaledwidth=50.0% ] consider a spherical overdensity with mass @xmath13 and radius @xmath14 .
equation of motion of the overdensity is simply given by @xmath15 by integrating this equation , we obtain @xmath16 here @xmath17 for a growing perturbation .
we define @xmath18 then equation ( [ ene ] ) reduces to @xmath19 on the other hand , the cosmological time is related with the scale factor as @xmath20 from equation ( [ col ] ) , we can derive the motion of spherical overdensity as a function of @xmath7 .
figure [ fig : sc ] shows the evolution of the size of overdense region for various values of @xmath8 .
it is shown that the corresponding overdensity @xmath6 increases monotonically as time evolves .
since the dimensionless time @xmath7 is related with the scale factor @xmath21 as equation ( [ age ] ) , at fixed time @xmath21 and for fixed @xmath6 we can determine the value of @xmath8 uniquely by solving equations ( [ col ] ) and ( [ age ] ) .
then the linear and nonlinear overdensity is easily derived : @xmath22 where @xmath23 is the linear growth rate normalized so that @xmath24 at @xmath25 . from this
, arbitrary @xmath6 is converted to the corresponding linear overdensity @xmath26 that is an essential quantity in computing the mass function . in figure
[ fig : ov ] , we plot an example of this relationship .
it shows one - to - one correspondence between @xmath6 and @xmath27 . ] ) and nonlinear ( eq . [ [ dnl ] ] ) overdensities at @xmath11 .
here we adopt @xmath12 .
[ fig : ov],scaledwidth=50.0% ] to compute the mass function of superclusters , we adopt the press - schechter formalism@xcite . from the linear overdensity @xmath26 ( eq . [ [ dlin ] ] )
, the mass function can be computed as @xmath28 where @xmath29 , and @xmath30 is the ( linear ) mass variance at @xmath31 . the conditional mass function of clusters within superclusters , that are already virialized ,
can be estimated in analog with the extended press - schechter formalism@xcite .
the number of clusters with mass @xmath13 that are in a supercluster with mass @xmath32 is given by @xmath33 , \label{cmf}\ ] ] where @xmath34 is the critical overdensity in the standard press - schechter formalism .
although the distribution of clusters in each supercluster might be affected by physical processes such as global tides and dynamical frictions@xcite , we adopt this form for simplicity .
-band luminosity density @xmath35@xcite to convert into @xmath6 .
[ fig : zdel],scaledwidth=50.0% ] in this section , we compare our analytic model with the results of einasto _ et al._@xcite who made a catalog of superclusters in the sdss early data release ( edr)@xcite .
they extracted superclusters from the density field of galaxies with a smoothing length 10@xmath36 , and also clusters from the density field with a smoothing length 0.8@xmath36 .
the superclusters and clusters are defined as the regions divided by the threshold density ( @xmath37 ) .
since the sdss edr data consists of two slices , the density filed is calculated in 2 dimensions only .
thus the total ( @xmath38-band ) luminosity of a supercluster is derived as @xmath39 where @xmath40 is the thickness of the slice , @xmath41 is the average diameter of the supercluster , and @xmath42 is the observed luminosity with the selection function of galaxies corrected . with this way 39 superclusters were identified ( see table 3 of @xcite ) . since the above definition of superclusters is slightly different from theoretical one , we need to check what the observed superclusters correspond to in our model .
we have defined superclusters as the regions that have the overdensity @xmath6 , so we compute @xmath43 for each observed supercluster .
figure [ fig : zdel ] shows @xmath6 for the sdss edr superclusters .
here we calculated @xmath6 as @xmath44 where @xmath35 is the average @xmath38-band luminosity density@xcite .
this coincides with equation ( [ dnl ] ) if the mass - to - light ratio of superclusters is the same as that of the universe .
as seen in the figure , the derived value of @xmath6 is scattered at around @xmath45 ( @xmath46 ) .
thus we adopt @xmath47 as a fiducial value , though we also examine the cases with @xmath48 to see how the uncertainty of @xmath6 would affect our results . , @xmath49 , @xmath12 , and @xmath50 .
for each panel , we changed one of these parameters while the other parameters are fixed . in all panels ,
the darker lines mean the larger values .
[ fig : lf],scaledwidth=70.0% ] first , we compare the abundance of superclusters in the sdss edr field .
to do so , the volume of the field where superclusters were identified is needed .
we simply adopt the sum of volumes of north and south slices of the edr data@xcite .
however , this may underestimate the volume because of the correction outside the slices ( eq . [ [ corr ] ] ) . to convert mass functions of superclusters to luminosity functions , the mass - to - light ratio of superclusters .
again , we assume the fiducial model @xmath49 , and estimate the sensitivity by changing @xmath51 from @xmath52 to @xmath53 .
figure [ fig : lf ] plots luminosity functions of superclusters in theory and observations .
first , the abundance of massive superclusters is sensitive to @xmath6 , thus the selection of superclusters is quite important in comparing the abundance .
next , the abundance of massive superclusters is also very sensitive to @xmath54 .
therefore , surveys of massive superclusters can be a good probe of @xmath54 .
finally , the observed number of superclusters shows reasonable agreement with the theory , particularly when we take account of several uncertainties including that of the survey volume . at smaller luminosities
the difference seems large , but this may be ascribed to the incompleteness of observations for such small objects .
einasto _ et al .
_ also estimated the number of clusters in each supercluster .
they noted that there exists a well - defined lower limit of cluster luminosities that is a function of distance to clusters ( denoted by @xmath55 ) , particularly for those at larger distances ( see figure 6 of @xcite ) .
thus we divide the supercluster sample by the distance @xmath55 . for each subsample , we assume the limiting luminosity shown in figure 6 of @xcite and calculate the number of clusters in a given supercluster by integrating equation ( [ cmf ] ) . we also convert the observed number of clusters in the edr field to the _ total _ number in the similar way as equation ( [ corr ] ) .
( @xmath56 ; _ upper left _ ) , @xmath57 ( @xmath58 ; _ upper right _ ) , @xmath59 ( @xmath60 ; _ lower left _ ) , and @xmath61 ( @xmath62 ; _ lower right _ ) . as limiting masses of clusters , we adopt `` set 1 '' in figure 6 of @xcite .
different lines mean different values of @xmath6 .
the other parameters are fixed to the fiducial values ( see the caption of figure [ fig : lf ] ) .
[ fig : cl_dis],scaledwidth=70.0% ] .
the observational data ( _ filled triangles _ ) are taken from @xcite .
for the errorbars , only statistical errors are included . for each panel , we changed one parameters while the other parameters are fixed to the fiducial values ( see the caption of figure [ fig : lf ] ) . in all panels ,
the darker lines mean the larger values .
[ fig : cl_par],scaledwidth=70.0% ] in figure [ fig : cl_dis ] , we plot the number of clusters within superclusters for different distances @xmath55 .
we find that theoretical curves and observations show good agreement with each other .
it seems that the model overpredicts the number of clusters , but this is perhaps because the assumption of the existence of lower limit of cluster luminosities becomes inaccurate for lower @xmath55 ( see figure 6 of @xcite ) .
next we examine the dependence of several parameters on the number of clusters , which is shown in figure [ fig : cl_par ] .
the figure indicates that the number of clusters in superclusters hardly depends on these parameters , though that was already clear from the expression of the conditional mass function shown in equation ( [ cmf ] ) .
thus the model predictions are quite robust , and the good agreement means that our model works well .
in this paper , we have constructed an analytic model of superclusters .
our model is based on the press - schechter theory , and defines superclusters as overdense regions that are on the course of the spherical collapse .
our model allows us to compute the mass function of superclusters , and also the number distributions of clusters in each supercluster . to test the model ,
we have compared the model with the supercluster catalog in the sdss edr data .
we have shown that the model shows in reasonable agreement with the observation .
we have also examined parameter dependences of the mass function of superclusters , and found that the mass function of massive superclusters is sensitive to @xmath54 .
thus superclusters will be a useful probe of @xmath54 if superclusters are selected more rigorously .
on the other hand , the number of clusters in superclusters shows only weak dependences on parameters
. our model would be useful not only in constraining cosmological parameters from the abundance of superclusters but also in many applications .
one of such applications is hot gas in superclusters .
recently it has been claimed that soft x - ray@xcite and sunyaev - zeldovich effects@xcite in the supercluster regions may be detected , and this might suggest the existence of relatively hot gas in superclusters . if this is true , the statistical significance of the signals that are caused by the hot gas can be computed by using our model , once the we measure the temperature and density structure of hot gas . other application is gravitational lensing ; gravitational lensing measures the total mass along the line of sight , thus a supercluster outside a cluster may affect , e.g. , the mass estimate of the cluster .
our model is useful since it allows us to estimate the distribution of the host superclusters for a given cluster . in practice
, we found that the effect is negligible when we consider superclusters with @xmath63 ; assuming a homogeneous sphere , convergence of a supercluster with @xmath64 can be estimated as @xmath65 where we assumed a source at @xmath66 and a supercluster at @xmath67 .
however , it is of interest to consider the effect more carefully , because the effect is dependent on the density profile of superclusters as well as the definition of superclusters , @xmath6 . in particular
, superclusters could have much more significant effect if superclusters have centrally concentrated density distributions .
although our model looks successful at this moment , definitely we need more tests . the more direct way to test
the model is to compare it with numerical simulations , since we know the initial condition and cosmological parameters in numerical simulations .
of course , comparison with observations is also of great importance in order to test our understanding of the universe . now
the sdss data have been becoming public@xcite , and the quite large survey region of the sdss would allow us to make three - dimensional catalog of superclusters .
the follow - up work is currently in progress , and the results will be presented elsewhere .
we thank issha kayo for useful discussions .
mo , kt , and ki are supported by jsps through jsps research fellowship for young scientists .
n. a. bahcall and r. m. soneira , astrophys . j. * 270 * , 20 ( 1983 ) | we construct an analytic model for the mass distributions of superclusters and clusters in each supercluster .
our model is a modification of the press - schechter theory , and defines superclusters as the regions that have some overdensity smaller than those for usual virialized objects .
we compare the mass functions with a catalog of superclusters in the sloan digital sky survey early data release and found that they are in reasonable agreement with each other . |
supernovae ( sne ) are associated with the deaths of stars , in particular core - collapse supernovae ( ccsne ) are associated with the deaths of massive stars , which have initial masses greater than about 8 .
sne are principally separated into two categories , those without hydrogen ( type i ) and those with ( type ii ) .
only type ia sne are thought to be thermonuclear explosions , which arise from accreting white dwarfs in binary stellar systems .
all the other sub - types are thought to be initiated by the core collapsing in massive stars .
the type of sn that occurs depends on the massive star s evolutionary stage at the time of the explosion .
the plateau subclass of type ii sne ( sne ii - p ) are thought to arise from the explosions of red supergiants ( rsgs ) , which have initial masses greater than 810 and have retained their hydrogen envelopes before core collapse @xcite . until the discovery of the red supergiant ( rsg ) that exploded as sn 2003gd @xcite , there had been no direct confirmation that sne ii - p did indeed arise from the explosions of rsgs . before this detection
there had been only two other unambiguous detections of type ii progenitors , neither of which fitted the evolutionary scenario that is commonly accepted .
these were the progenitors of the peculiar type ii - p sn 1987a , which was a blue supergiant , and the type iib sn 1993j that arose in a massive interacting binary system @xcite .
the recent discovery of sn 2005cs ( ii - p ) in m51 , a galaxy with deep multi - colour pre - explosion images from the _ hubble space telescope _
( _ hst _ ) , led to the discovery of another rsg progenitor of a ii - p @xcite .
the estimated mass of the star was @xmath8 , similar to the mass ( @xmath9 ) for the progenitor of sn 2003gd @xcite . a supergiant of mass @xmath10 was found to be coincident with the type ii - p sn 2004et by @xcite , although it is likely not to have been as cool as an m - type supergiant , and the sn itself may be peculiar .
there have been other extensive attempts to detect progenitors of nearby sne on ground- and space - based archival images e.g. @xcite , which have set upper mass limits mostly on ii - p events .
the low mass of the progenitors discovered and upper limits set has led to the suggestion that sne ii - p come only from rsgs with masses less than about 15@xcite sn 2004a is another example of a nearby sn ii - p which has _ hst _
pre - explosion images , allowing the search for a progenitor star .
sn 2004a was discovered by k. itagaki of teppo - cho , yamagata , japan on january 9.84 ut using a 0.28-m f/10 reflector .
itagaki confirmed his discovery on january 10.75 ut , with a location of r. a. @xmath11 , dec .
@xmath12 , around 22 arcsec west and 17 arcsec north of the centre of ngc 6207 .
itagaki reported that no object was visible on his observations of 2003 december 27 , which had a limiting magnitude of 18 , or any of his observations prior to this date @xcite .
itakagi s observations allow the explosion epoch to be fairly well constrained , suggesting that sn 2004a was discovered when it was quite young at less than 14d after explosion .
an optical spectrum was obtained by @xcite on january 11.8 and 11.9 ut , and showed a blue continuum with p - cygni profiles of the balmer lines , consistent with a type ii sn .
the emission features were somewhat weak suggesting that the sn was indeed young , in line with itagaki s observations .
the expansion velocity , measured from the minima of the balmer lines , was around 12000 . in _
cycle 10 , we had a snapshot programme to enhance the _
hst _ archive with 100200 wide field planetary camera 2 ( wfpc2 ) multi - colour images of galaxies within approximately 20mpc . in the future
, sne discovered in these galaxies could have pre - explosion images available to constrain the nature of the progenitor stars .
this strategy is now beginning to bear fruit , ngc 6207 was one of those targets and the pre - explosion site of sn 2004a was imaged in three filters .
in this paper we present photometric and spectroscopic data of sn 2004a in section [ sec : obs ] followed by an analysis of the photometry in section [ sec : epoch ] , where an explosion date is estimated .
we estimate the reddening towards the sn in section [ sec : red ] and obtain the expansion velocity in section [ sec : expvel ] .
the distance to ngc 6207 is not well known and only two distance estimates , which are both kinematic , exist in the literature .
we estimate the distance using two further methods and compile the distances within the literature , in an attempt to improve the situation , in section [ sec : d ] . using the distance
found we then calculate the amount of nickel synthesised in the explosion in section [ sec : ni ] .
we present the discovery of the progenitor in section [ sec : prog ] and a discussion of the implications and conclusion in sections [ sec : diss ] and [ sec : con ] , respectively . throughout this work
we have assumed the galactic reddening laws of @xcite with @xmath13 .
_ bvri _ photometry was obtained shortly after discovery from the following telescopes : the 2.0-m liverpool telescope ( lt ) , la palma ; the 4.2-m william herschel telescope ( wht ) , la palma ; and the robotic palomar 60-inch telescope ( p60derekfox / p60/ ] ; cenko et al .
2006 , in prep . ) as part of the caltech core - collapse programavishay / cccp.html ] ( cccp , * ? ? ?
* ; * ? ? ? * gal - yam et al .
2006 in prep . )
the lt observations were taken with the optical ccd camera , ratcam , using its bessel _ bv _ and sloan @xmath14 filters .
the data were reduced using the lt data reduction pipeline .
the wht observations were taken with the auxiliary port imaging camera ( aux ) , using its _ bvri _ filters , and were reduced using standard techniques within iraf .
the frames were debiased and flat - fielded using dome flats from a few nights later .
details of the p60 camera and data reduction can be found in @xcite .
a summary of these observations can be found in table [ tab : phot ] as well as the results from the sn photometry .
[ cols= " < , > , > , > , > , > , > , < " , ] [ tab : mag ]
sn 2004a appears to be a ` normal ' sn ii - p with a spectrum and light curve very like the well observed sn 1999em ( figs .
[ fig : s04a99em ] and [ fig:04a99em ] ) .
the nickel mass found in section [ sec : ni ] was also comparable to that of sn 1999em , as we would have expected .
the existence of a low luminosity , low nickel mass , sub - group of sne ii - p ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , has been suggested .
one of the proposed characteristics of this group is a rapid excess in the and colours at the end of the photospheric phase @xcite .
and @xcite noted this rapid excess in the ` faint ' sn 1997d ( see fig . [
fig : diffsnc ] ) , and it was also observed by @xcite in sn 1999eu , another ` faint ' sn , in the form of a sharp spike .
[ fig : c04a99em ] compares the colour evolutions of sn 2004a and sn 1999em .
the colour evolution of sn 2004a clearly shows such a rapid excess at the end of the plateau , which is not seen in the colour curve of sn 1999em .
[ fig : diffsnc ] shows a comparison of the colour curves of the prototypical peculiar ` faint ' sn 1997d @xcite , and the ` normal ' sne 1999em and 2003gd @xcite .
although the excess is present in the colour curve of sn 2004a , it does not show as great an excess as sn 1997d , but instead follows a similar curve to sn 2003gd with a slightly bluer tail .
there is too large a scatter in the colour curve to be of much use , but there does not appear to be any evidence of a colour excess .
the appearance of the spectrum and the light curves imply that sn 2004a is a normal sn ii - p , suggesting that this excess is not confined to the ranks of the ` faint ' sne . until we have a larger sample of sne ii - p , both ` faint ' and ` normal ' , it is difficult to confirm or rule out this as a characteristic . a comparison between the uvoir light curves of sne 2004a , 2003gd , 1999em , 1999gi and 1997d is shown in fig . [ fig : uvoir ] .
all the uvoir light curves , apart from sn 2004a , are from @xcite and a full description of how they were constructed can be found there .
the uvoir light curve of sn 2004a , which was constructed in the same way , also confirms the similarities between sn 2004a and the other normal sne ii - p .
sn 2004a differs slightly from sn 1999em in the plateau with sn 2004a being marginally fainter , whereas the tail luminosities reflect the comparable nickel mass ejected .
we compared the observed properties of sn 2004a with the progenitor mass , presented in section [ sec : prog ] , using equation ( 2 ) of @xcite .
the equations of @xcite relate the explosion energy , mass of the envelope expelled ( ) and initial radius of the star just before the explosion , to the observed quantities of sne .
@xcite used a sample of 14 supernovae to test these equations and found that they gave reasonable results .
equation ( 2 ) of @xcite , is shown here in equation ( [ equ : m ] ) , where is the ejected envelope mass , @xmath15 is the length of the plateau and @xmath16 is the velocity of the photosphere material in the middle of the plateau .
@xmath17 if we consider the uvoir light curve in fig .
[ fig : uvoir ] , the length of the plateau is @xmath1880d , although it is difficult to be confident as the sn was not observed close to explosion .
we therefore estimate the length of the plateau to be @xmath19d , where the positive error is very conservative . using the reddening estimated in section [ sec : red ] , the parameters discussed in section [ sec : scm ] for the scm at 50d post explosion and the distance estimated in section [ sec : d ]
, we found that the ejecta mass was @xmath20 .
if we assume that the compact remnant plus other mass losses from the system is @xmath21 then it suggests that @xmath22 for sn 2004a , where the upper mass limit of 23 is a hard upper limit as the error in the plateau is very conservative .
the lower limit of this is consistent with the progenitor mass , @xmath5 , found from pre - explosion images in section [ sec : prog ] .
this result may also suggest that equation ( [ equ : m ] ) overestimates the mass somewhat as the lower limit is only just consistent even though our lower limits are fairly well constrained .
the observations of progenitors do not only have important implications for stellar evolution theory , but also for the progenitor models of ` faint ' sne ii - p .
there are two very different plausible models for these , one being the low - energy explosion of massive stars ( e.g. * ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in this model
the collapsing core forms a black hole and a significant amount of fallback of material occurs .
an alternative scenario is the low - energy explosion of low - mass stars , presented by who successfully reproduced the observations of sn 1997d with an explosion of @xmath23 ergs and an ejected mass of 6 .
the findings of @xcite support the high - mass progenitor scenario . the authors find a bimodal distribution of sne in the nickel mass , plane .
a reproduction of the authors fig . 1 ( right )
is shown here in fig .
[ fig : zamp ] . we are beginning to populate this figure with direct measurements of the main - sequence mass of the progenitors , as opposed to model dependent values .
the most interesting feature of this figure , and the most controversial , is the existence of the faint branch .
the sne which populate this branch are proposed to have high initial mass progenitors .
the points with direct measurements , or robust limits , of progenitor mass from pre - explosion imaging are highlighted with the filled circles .
this figure however should be treated with caution as the theoretical estimates of from @xcite , shown with open circles in the figure , will most probably need to be revised using the value of the opacity from @xcite .
the observed nickel masses are estimated from the light curves , and are from @xcite for sn 1999br and from ( * ? ? ?
* and references therein ) for sne 2003gd and 1999em .
the @xcite nickel masses are however an input parameter of the semi - analytical light curve code and are therefore theoretical . for sn 2003gd
is from @xcite , whereas sne 1999em and 1999br are upper mass limits only and are from @xcite and @xcite , respectively .
sn 2004a comfortably sits in the ` normal ' branch of the bimodal distribution .
even though the mass limit of sn 1999br rules out a high mass progenitor for this ` faint ' sn , it does not necessarily rule out a high mass progenitor scenario for other ` faint ' sne .
it does however suggest that for at least some ` faint ' sne a low mass progenitor is likely . with more upper mass limits or direct detections of the progenitor stars of these sne
, we should be able to determine the validity of the high mass progenitor scenario .
we presented photometric and spectroscopic data of the type ii - p sn 2004a , comparing the _ bvri _ light curves with those of the well observed sn 1999em using a @xmath24-fitting algorithm .
this analysis allowed us to estimate an explosion epoch of jd @xmath25 , corresponding to a date of 2004 january 6 .
we estimated the extinction of the sn as @xmath26 , using _
acs photometry of the neighbouring stars and confirmed the reddening using the sn s colour evolution .
the expansion velocity was measured from our only spectrum , and was found to be comparable to the velocities of similar sne ii - p .
this enabled us to extrapolate the velocity evolution of sn 2004a forwards 7 days to find the velocity at a phase of 50d , to be used in the scm distance estimate . three new distances to ngc
6207 were calculated using two different methods , the standard candle method ( scm ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and the brightest supergiants method .
firstly , using the extrapolated velocity and interpolated _ vi _ magnitudes , we estimated a distance of @xmath27mpc with the scm .
we then used _ hst _ ( acs hrc ) photometry to estimate the distances of @xmath1 and @xmath28mpc using two different bsm calibrations . using these three distances and other distances within the literature we estimated an overall distance of @xmath29mpc to ngc 6207 .
this distance allowed the nickel mass synthesised in the explosion to be estimated , @xmath30 , comparable to that of sn 1999em
. the probable discovery of the progenitor of sn 2004a on pre - explosion _ hst _
wfpc2 images was presented .
the star that exploded was likely to have been a rsg with a mass of @xmath5 .
if the 4.8@xmath6 detection is not believed then the 5@xmath6 upper limit in the f814w images implies a robust upper mass limit for a rsg progenitor of 12 .
this is only the seventh progenitor star of an unambiguous core - collapse sn that has had a direct detection .
the first two progenitors discovered were those of sn 1987a , which was a bsg , and sn 1993j , which arose in a massive interacting binary system .
neither of these fitted in with the theory that rsgs explode to give type ii sne .
however , it appears that the elusive rsg progenitors are now being found , with the discovery of the progenitor of sn 2004a being the third ( after sne 2003gd and 2005cs ) .
the observations of sn 2004a were compared to those of other sne ii - p and were found to be consistent with other normal sne ii - p , although there is a small peak in the uvoir light curve at around 40d , which could just be an artifact of the photometry .
the sn observations were then compared to the mass of the progenitor using theoretical relationships which relate the mass of the envelope expelled to the observed quantities of sne , and were found also to be consistent .
we conclude that sn 2004a was a normal type ii - p sne that arose from the core - collapse induced explosion of a red supergiant of mass @xmath5 .
based on observations made with the nasa / esa hubble space telescope , obtained from the data archive at the space telescope science institute .
stsci is operated by the association of universities for research in astronomy , inc .
, under nasa contract nas 5 - 26555 .
the spectroscopic data presented were obtained at the w.m .
keck observatory , which is operated as a scientific partnership among the california institute of technology , the university of california and the national aeronautics and space administration .
the observatory was made possible by the generous financial support of the w.m .
keck foundation .
the authors wish to 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 are most fortunate to have the opportunity to conduct observations from this mountain .
some data presented were taken with the liverpool telescope at the observatorio del roque de los muchachos , la palma spain .
the authors would like to express their thanks to the research staff at caltech and palomar observatory who made the p60 automation possible .
the initial phase of the p60 automation project was funded by a grant from the caltech endowment , with additional support for this work provided by the nsf and nasa .
a. gal - yam acknowledges support by nasa through hubble fellowship grant # hst - hf-01158.01-a awarded by stsci , which is operated by aura , inc . , for nasa , under contract nas 5 - 26555 .
s. smartt acknowledges funding from pparc and the european science foundation in the forms of advanced and euryi fellowships .
m. hendry thanks pparc and queen s university , belfast for their financial support .
m. , 2004a , in measuring and modeling the universe , from the carnegie observatories centennial symposia .
carnegie observatories astrophysics series .
edited by w. l. freedman , 2004 .
pasadena : carnegie observatories , the latest version of the standardized candle method for type ii supernovae , astro - ph/0301281 m. , 2004b , in cosmic explosions . on the 10th anniversary of sn 1993j ( iau colloquium 192 ) .
springer , heidelberg , p. 535
the standard candle method for type ii supernovae and the hubble constant , astro - ph/0309122 a. , ramina m. , zampieri l. , navasardyan h. , salvo m. , fiaschi m. , 2004a , in cosmic explosions . on the 10th anniversary of sn 1993j ( iau colloquium 192 ) .
springer , heidelberg , p. 195 observational properties of type ii plateau supernovae , astro - ph/0310056 l. , ramina m. , pastorello a. , 2004 , in supernovae ( 10 years of 1993j ) , marcaide j. m. , weiler k. w. , eds , proc .
192 , springer - verlag , berlin , in press , understanding type ii supernovae , astro - ph/0310057 | we present a monitoring study of sn 2004a and probable discovery of a progenitor star in pre - explosion _ hst _ images .
the photometric and spectroscopic monitoring of sn 2004a show that it was a normal type ii - p which was discovered in ngc 6207 about two weeks after explosion .
we compare sn 2004a to the similar type ii - p sn 1999em and estimate an explosion epoch of 2004 january 6 .
we also calculate three new distances to ngc 6207 of @xmath0 , @xmath1 and @xmath2mpc .
the former was calculated using the standard candle method ( scm ) for , and the latter two from the brightest supergiants method ( bsm ) .
we combine these three distances with existing kinematic distances , to derive a mean value of . using this distance
we estimate that the ejected nickel mass in the explosion is @xmath3 .
the progenitor of sn 2004a is identified in pre - explosion wfpc2 f814w images with a magnitude of @xmath4 , but is below the detection limit of the f606w images .
we show that this was likely a red supergiant ( rsg ) with a mass of @xmath5 .
the object is detected at 4.7@xmath6 above the background noise . even if this detection is spurious
, the 5@xmath6 upper limit would give a robust upper mass limit of 12 for a rsg progenitor .
these initial masses are very similar to those of two previously identified rsg progenitors of the type ii - p sne 2004gd ( @xmath7 ) and 2005cs ( @xmath5 ) .
[ firstpage ] stars : evolution - supernovae : general - supernovae : individual : sn 2004a - galaxies : individual : ngc 6207 - galaxies : distances and redshifts |
celestial mechanics often requires approximation techniques to compute the motion of bodies over very long times .
giving reliable error estimates on these techniques is a very practical problem , of not simple solution . in this paper
we apply to a typical astronomical problem the general scheme of @xcite @xcite to estimate the error of the averaging principle , in the case of one fast angular variable @xmath3 ( in the one - dimensional torus ) and many slow variables @xmath4 ( the actions ) .
the application we consider refers to the motion of a satellite around an oblate planet : of course , oblateness gives a perturbation of the keplerian , purely radial gravitational potential .
there is a classical way to express the perturbed potential as a series of spherical harmonics , where the multipole moments of the planetary mass distribution appear as coefficients . in the case of axial symmetry , this series only involves zonal harmonics and the @xmath5-th term contains a multipole coefficient @xmath6 ( for a reminder , see subsection and references therein ) . for a slightly oblate planet , all these coefficients
are small if they are defined in a convenient , dimensionless fashion ; often , @xmath7 and one retains only the @xmath8 term of the previous expansion ; the corresponding equations of motion for the satellite are the classical `` @xmath0 problem '' . in this paper the @xmath0 problem is considered in the restricted case of polar orbits ; this problem involves a fast angular variable @xmath9 ( the angle between the polar axis and the radius vector of the satellite ) and three slow variables @xmath10 representing the parameter , the eccentricity and the argument of the pericenter for the osculating keplerian ellipse . in principle , all these variables are unknown functions of the `` physical '' time @xmath11 ; however , it is customary to consider as a `` time '' variable the angle itself or , rather , the ratio @xmath12 of the angle to @xmath13 . with this choice of the time variable , by definition
the angle evolves with the law @xmath14 $ ] ( where @xmath15 $ ] indicates equivalence mod.@xmath13 ) ; for obvious reasons , @xmath12 will be called the _ orbit counter_. the evolution of @xmath10 is described by a set of equations of the form = f^i(,[2 ] ) , ^i(0 ) = i^i_0 , [ per ] where @xmath16 is a small parameter proportional to @xmath0 ( see again subsection ) . independently of this specific application ,
a system of evolution equations like ( [ per ] ) , with any number of unknown functions @xmath17 and a small parameter @xmath18 , will be called in the sequel a _ perturbed periodic system_. to any such system one can associate an averaged system , replacing the functions @xmath19 in ( [ per ] ) with their averages over the angle @xmath20 $ ] ; the solution @xmath21 of the latter is in a function of @xmath22 .
the difference between the solutions @xmath23 of ( [ per ] ) and @xmath24 of the averaged system , on a time scale of order @xmath25 , can be estimated in a fully quantitative way with the general method of @xcite @xcite ; this is what we are going to do in the present work , for the periodic system describing the polar @xmath0 problem . as in the cited papers
, the expression `` fully quantitative '' means that our final estimate will have the form @xmath26 are computable functions , and @xmath27 is a constant , whose choice determines quantitatively the interval of observation ; we note that @xmath28 is the total number of orbits .
the availability of an algorithm to construct the estimators @xmath29 is the main difference between the approach of @xcite and the classical , qualitative result @xmath30 , see , e.g. , @xcite @xcite @xcite .
we are aware of the existence of higher order versions of the averaging method , in which the error behaves like @xmath31 for @xmath32 ( @xmath33 ; see again @xcite @xcite ) . however , the classical theory of higher order averaging gives little more than the qualitative error estimate @xmath31 ; perhaps an extension of our methods could give quantitative estimators of the form @xmath34 , but this problem is left to future work .
let us return to the functions @xmath29 in eq .
( [ formm ] ) .
our method to compute them for the @xmath0 problem is mainly analytical , but , in its final steps , it requires the numerical solution of an ode on the interval @xmath35 ; however , this numerical computation is much faster than the direct numerical solution of ( [ per ] ) ( a fact already appearing in other applications , different from the @xmath0 problem , considered in our previous papers ) . the general treatment proposed in this paper for the polar @xmath0 problem
is subsequently exemplified , choosing for @xmath16 and for the initial conditions @xmath36 the values corresponding to the earth and to the polar and cos - b satellites . in this case , the estimators @xmath29 have been computed up to @xmath37 orbits ( i.e. , one or two centuries ) , spending few seconds of cpu time on a pc . to test their reliability ,
we have computed by direct numerical integration the differences @xmath38 ; this is a more expensive operation , that we have been able to perform up to @xmath39 orbits with the same pc ; our estimators are quite satisfactory on this interval , since the functions @xmath40 are close to the envelopes of the rapidly oscillating functions @xmath41 . of course
, the treatment of real satellites of the earth should include , besides the @xmath0 gravitational term , other minor perturbations such as : gravitational forces corresponding to higher order moments of the earth , atmospheric dragging , solar wind , tidal effects due to the moon gravity .
all these effects could be treated , giving rise again to a system of the form ( [ per ] ) with more complicated perturbation components @xmath42 ( even for non polar orbits : in general , if the analysis is not restricted to the orbits in a fixed plane , the actions are not three but five ) .
presumably , the corresponding averaged system and our estimators @xmath29 for its error could be computed as well ; this is left to future work . in spite of the limitation to the polar @xmath0 problem
, we think that the results of this paper have some interest , because they show that a general method for error estimates works over very long times on a non - trivial problem .
to conclude , we describe in few words the organization of the paper . in section [ secav ] , we show how to compute the error estimators for the averaging of any periodic system ( [ per ] ) ; this illustration specializes to the periodic case the slightly more general setting of @xcite @xcite for one - frequency systems . in section [ satell ] ,
the basic facts on the motion of a satellite in a fixed plane @xmath43 are reviewed ; the outcomes are the equations of motion for @xmath44 , for any perturbation of the keplerian potential keeping the satellite on @xmath43 . in section [ polmot ] , we specialize the previous equations to the polar @xmath0 problem ; the averaged system is solved , and we construct the algorithm to compute the error estimators @xmath29 @xmath45 . in section [ seces ] , we perform the computations for the polar and cos - b satellites of the earth . some details on the constructions of sections [ satell]-[polmot ] are presented in appendices [ appkepl]-[appinv ] .
* 2a . introducing the problem . *
we consider an open set @xmath46 of ( the space of the actions ) @xmath47 and the one - dimensional torus @xmath48 : = \ { i = ( i^i)_i=1, ... ,d } ^d , : = / 2 = \ { } ; the natural projection of the real axis on the torus will be written @xmath15 : \reali \vain \bt$ ] , @xmath49 $ ] . now , let f = ( f^i)_i=1, ...
,d c^ ( , ^d ) , ( i , ) f(i , ) [ ef ] ( with @xmath50 ) ; we fix i_0 , ( 0,+ ) , [ io ] and consider the cauchy problem = f(,[2 ] ) , ( 0 ) = i_0 : [ evol ] the maximal solution ( in the future ) is a @xmath51 function @xmath52 , @xmath53 ( with @xmath54 $ ] ) . here and in the sequel ,
typeface symbols are employed for _ functions _ of @xmath12 ( or @xmath55 ) ; this allows , for example , to distinguish the above mentioned function @xmath23 from a point @xmath56 of @xmath46 .
the averaged system associated to ( [ evol ] ) is = ( ) , ( 0 ) = i_0 , [ av ] = ( ) _ i=1, ...
,d c^(,^d ) , i ( i ) : = 1 2 _ d f(i , ) ; [ sense ] the maximal solution ( in the future ) is a @xmath51 function @xmath57 . throughout the paper , we are interested in binding the difference @xmath58 .
connections with the framework of @xcite @xcite . * in these papers , we have discussed the cauchy problem \
{ d /d = f ( , ) , & ( 0 ) = i_0 + d / d = ( ) + g ( , ) , & ( 0 ) = _ 0 .
[ pert ] with @xmath59 as in ( [ ef ] ) and @xmath60 , @xmath61 , @xmath62 for all @xmath63 , @xmath64 , @xmath65 , the maximal solution being @xmath66 .
precise quantitative estimates have been derived on the difference @xmath58 , where @xmath24 is the solution of ( [ av ] ) . clearly , the perturbed periodic system ( [ evol ] ) is equivalent to a particular case of ( [ pert ] ) .
in fact , let us choose ( i ) : = 2 , g(i , ) : = 0 ; _ 0 : = 0 ; [ special ] then , the equation for @xmath9 in ( [ pert ] ) is fulfilled with ( ) : = [ 2 ] and the equation for @xmath23 in ( [ pert ] ) becomes , with this position , the cauchy problem ( [ evol ] ) .
this remark allows to apply the general results of @xcite @xcite to the problem ( [ evol ] ) ; in the sequel , we report from these papers the minimal elements enabling one to read independently the present paper , i.e. : ( i ) some basic assumptions required by our approach , and the definitions of certain auxiliary functions ; ( ii ) the final proposition from @xcite , estimating @xmath58 in terms of these auxiliary functions . *
2c . a first set of assumptions and auxiliary functions .
* we use the tensor spaces and the operations on tensors already employed in our previous works ; in particular , the spaces @xmath67 , @xmath68 and @xmath69 are made , respectively , by the the @xmath70 , @xmath71 and @xmath72 tensors over @xmath47 ; in the three cases , these are represented as families of real numbers @xmath73 , @xmath74 , @xmath75 ( @xmath76 ) . from here to the end of the section the function @xmath59 , the initial datum @xmath77 and the parameter @xmath16 of eqs .
( [ ef ] ) ( [ io ] ) are fixed .
we stipulate the following : 0.1cm(1 ) @xmath78 , @xmath79 , @xmath80 and @xmath81 are the auxiliary functions uniquely defined by the equations f = + 2 s , = 0 ; [ thef ] s = 2 , ( i,0 ) = 0 [ st ] p : = s i f ; : = vi f ; [ dq ] p = + 2 , ( i,0 ) = 0 ; [ qw ] : =
i f ; : = ^2 i^2 - ( ) ^2 .
[ em ] the barred symbols @xmath82 , @xmath83 , @xmath84 denote the averages of @xmath85 over the angle , in the sense of ( [ sense ] ) ; @xmath86 and @xmath87 indicate the jacobian and the hessian with respect to the variables @xmath4 .
explicit formulas for @xmath88 and @xmath89 are s = - , ( i , ) : = 1 2 _ 0^ d
( f(i, ) - ( i ) ) ; [ es ] v(i , ) : = 1 2 _ 0^ d
s(i, ) ; w(i , ) : = 1 2 _ 0^ d
( p(i, ) - ( i ) ) .
[ ev ] the above definitions of @xmath90 specialize the general prescriptions of @xcite @xcite to the case ( [ special ] ) , of interest in this paper .
similar comments could be added in the sequel , but they will not be repeated .
0.1cm(2 ) we introduce the open set : = \ { ( i , i ) ^d | [ i , i + i ] } , [ deg ] where @xmath91 $ ] is the closed segment in @xmath47 with the indicated extremes .
from now on , @xmath92 and @xmath93 are two functions such that , for all @xmath94 , ( i + i ) = ( i ) + ( i , i ) i , [ equel0 ] ( i + i ) = ( i ) + ( i ) i + 1 2 ( i , i ) i^2 , ^i_j k(i , i ) = ^i_k j(i , i ) .
[ equel ] @xmath95 and @xmath96 are not uniquely determined , if @xmath97 : possible choices , given by taylor s formula , are ( i , i ) : = _ 0 ^ 1 d i(i + i ) , ( i , i ) : = 2 _ 0 ^ 1 d ( 1 - ) ^2 i^2(i + i ) .
[ tayf ] ( 3 ) from now on , @xmath35 ( @xmath98 $ ] ) is a fixed interval where the solution @xmath24 of the averaged system ( [ av ] ) is assumed to exist .
we denote with @xmath99 , @xmath100 and @xmath101 , @xmath102 the solutions of the cauchy problems = ( ) , ( 0 ) = ; [ sistr ] = ( ) + ( ) , ( 0 ) = 0 ; [ sistk ] these are @xmath103 , and exist on the whole interval @xmath35 due to the linearity of the above differential equations . by standard arguments ,
one proves that @xmath104 is an invertible matrix for all @xmath105 , and derives for @xmath106 the explicit formula @xmath107 . *
. some more auxiliary functions . *
we maintain the assumptions and notations of the previous items ( 1 ) ( 2 ) ( 3 ) .
in @xcite we have introduced a second set of auxiliary functions , related to @xmath108 and to any separating system of seminorms on @xmath47 . for simplicity , here we consider the seminorms giving the absolute values of each component of vectors and tensors in @xmath47 ; what follows is an adaptation of @xcite to this choice . from now on , @xmath109 range in @xmath110 and we use einstein s summation convention on repeated indices .
0.1cm(4 ) for @xmath111 and @xmath112^d$ ] , we put b(j , ) : = \ { i ^d |
|i^i - j^i| < ^i i } .
0.1cm(5 ) @xmath113^d)$ ] is a function such that b ( ( ) , ( ) ) .
[ recall ] we put : = \ { ( , r ) [ 0,u ) [ 0,+)^d | r^i <
^i ( ) i } . [ dero ] 0.1cm(6 ) @xmath114 in @xmath115 and @xmath116 are functions such that for any @xmath105 , @xmath117 , @xmath118 , ) - ( ( ) ) ( ( ) + j , ) ) ^i | b^i ( , |j| ) , [ fb ] | ( ( ( ) + j , ) - ( ( ) ) ( + ) ( ( ) + j , ) [ fc ] @xmath119 | ^i_j((),j)| ^i_j(,|j| ) , [ fd ] | ^i_j k((),j)| e^i_j k(,|j| ) . [ fe ] in the above
, one always intends | j | : = ( | j |^i)_i=1, ...
the functions @xmath120 are assumed to be non - decreasing with respect to the variable @xmath121 , i.e. , ( , r ) , ( , r ) , r^j r^j j c^i(,r ) c^i(,r ) [ monot ] and similarly for @xmath122 , and @xmath123 .
0.1cm(7 ) given @xmath124 , we define the functions ^i c^2 ( , [ 0,+ ) ) , ^i ( , r ) : = a^i(,r ) + b^i(,r ) , [ al ] ^i c^1([0,+)^d , [ 0,+ ) ) , [ ga ] @xmath125 0.1cm(8 ) @xmath126 and @xmath127 are functions such that , for @xmath105 , | ^i_j ( ) | r^i_j ( ) , | ( ^-1)^i_j ( ) | p^i_j ( ) .
[ rp ] * 2e . the main result on general periodic systems . * in @xcite
, we have shown how to obtain estimators for @xmath58 solving a system of integral equations , or of equivalent differential equations . here
we only report the final differential reformulation , that will be used in the present paper . keeping in mind the previous items ( 1)(8 ) , we have the following statement . * proposition . * [ proprinc ] ( i ) assume there are @xmath128 , @xmath129 and @xmath130 , such that : = _ i=1, ... ,d [ ^i - ^i , ^i + ^i ] _ i=1, ... ,d ( 0,^i(0)/ ) [ setsi ] and : = _ i=1, ... ,d _ j=1^d ^i_j < 1/ , [ hj ] |^i r^j(0 , ) | ^i_j , [ hi ] | ^i(0 , ) - ^i | + ^i_j ^i < ^i [ hip ] then , there is a unique @xmath131 such that _ 0 , ( 0 , _ 0 ) = _ 0 .
[ fixedp ] ( ii ) with @xmath132 as above , let @xmath133 solve the cauchy problem = p^i_j ^j ( , , ) , _
^i(0 ) = 0 , [ tre ] = ( 1 - ( , ) ) ^-1 , i _
h ( ^h ( , ) + r^h_k p^k_j ^j ( , , ) + _ ^j ) , [ quattro ] @xmath134 for all @xmath135 , with the domain conditions 0 < ^i < ^i , ( 1 - ( , ) ) > 0 [ due ] ( in the above , @xmath136 stands for the matrix @xmath137 ( @xmath138 ) , and eq .
( [ quattro ] ) contains the matrix elements of its inverse .
we note that ( [ tre ] ) implies @xmath139 ) .
then , the solution @xmath23 of the periodic system ( [ evol ] ) exists on @xmath140 and | ^i ( ) - ^i ( ) | ^i(t ) _ _ * remarks . * [ rema2 ] for future use , it is worthy to repeat some considerations of @xcite . *
( a ) * the previous proposition mentions @xmath132 , the unique fixed point of the map @xmath141 in the set @xmath142 of eq .
( [ setsi ] ) .
the proof given in @xcite indicates that @xmath143 has lipschitz constant @xmath144 in the maximum component norm @xmath145 , with @xmath146 as in ( [ hj ] ) .
so , from the standard theory of contractions , we have the iterative construction _ 0
= _ n + l_n , l_1 , l_n : = ( 0 , l_n-1 ) .
[ iterates ] for each @xmath147 , _ 0 - l_n ( ) ^n-1 l_2 - l_1 ( 1 - ) .
one can use the above characterization of @xmath132 to compute it numerically ; in this case one finds the iterates up to a large order @xmath148 , and then approximates @xmath132 with @xmath149 . *
( b ) * in typical cases , and also in the forthcoming application to satellites , the cauchy problem ( [ tre ] ) ( [ quattro ] ) for the unknown functions @xmath150 is solved numerically ( by some standard package for odes ) . *
2f . a final comment on the functions @xmath151 .
* let us start with a remark on @xmath152 .
these functions are always multiplied by a small factor @xmath16 in the main statements on @xmath153 , i.e. , proposition [ proprinc ] ; for this reason , in applications it is generally sufficient to determine @xmath154 majorizing roughly the left - hand sides of eqs .
( [ fb])([fe ] ) .
the situation is different for the functions @xmath155 , which are not multiplied by @xmath16 ; in this case , it is important to determine them estimating accurately the left - hand side of eq .
( [ fa ] ) or , at least , the part of order zero in @xmath156 . to this purpose , we note the existence of a function @xmath157 such that , for all @xmath94 and @xmath118 , s(i + i , ) = s(i , ) + ( i , i , ) i [ equels ] ; a solution of this equation is given by taylor s formula , i.e. , ( i , i , ) : = _ 0 ^ 1 d s i(i + i , ) .
[ tays ] to go on , we write s ( ( ) + j , ) - ( ) s(i_0,_0 ) - ( ) @xmath158 which decomposes the function in the left - hand side of ( [ fa ] ) into a zero order part in @xmath156 plus a reminder controlled by @xmath159 .
now , assume there are functions @xmath160 and @xmath161 such that , for all @xmath105 , @xmath117 and @xmath118 , | ( s ( ( ) , ) - ( ) s(i_0,_0 ) - ( ) ) ^i | ( ) , [ faz ] | ^i_j((),j , ) | a^i_j ( , |j| ) .
[ faa ] then , eq . ( [ fa ] ) is fulfilled by the function a^i(,r ) : = ( ) + a^i_j(,r ) r^j ; [ defamu ] of course , this definition must be inserted in eq .
( [ al ] ) for @xmath162 . in applications
one will determine @xmath163 binding very accurately the left - hand side of eq .
( [ faz ] ) , and @xmath164 binding more roughly the left - hand side of eq .
( [ faa ] ) ; this is convenient , since in eq .
( [ defamu ] ) for @xmath165 the terms @xmath164 are multiplied by @xmath166 and this will finally take the small values @xmath167 .
the above strategy to determine @xmath168 will be employed in the application of the next sections , concerning satellite motions .
* . a general setting . * a satellite of mass @xmath169 , position @xmath170 and velocity @xmath171 is assumed to move around a planet of mass @xmath172 . as a first approximation , we assume the mass of the planet to be uniformly distributed within a sphere of center @xmath173 ; so , the gravitational potential energy ( per unit mass ) of the planet has the keplerian form v_(p ) = - g m | p - o | [ vkep ] where @xmath174 is the gravitational constant .
the corresponding gravitational force @xmath175 will be referred to as the keplerian force , and all the other forces are assumed to be small perturbations of it .
the non - keplerian force acting on the satellite ( of gravitational , or any other nature ) will be written as @xmath176 , where @xmath18 is a small dimensionless parameter , arising naturally from the analysis of this perturbation ; the parameter and the satellite mass are factored out from the perturbation , for convenience .
so , the total force acting on the satellite is ( p , p ) = m ( - v_(p ) + ( p , p ) ) .
we assume the perturbation to keep the satellite on a fixed plane @xmath43 passing through @xmath173 , where we introduce the polar coordinates @xmath177 , @xmath178 angle ( in @xmath48 ) between a fixed unit vector @xmath179 and @xmath180 .
the equations for the satellite motions in this plane can be written as a first - order system for @xmath181 and their derivatives @xmath182 , in the following way : - ^2 + g m ^2 = _ ( , , , ) , = , [ ero ] ^2 d d t + 2 = _ ( , , , ) , = , [ ete ] where @xmath183 and @xmath184 are the lagrangian components of the perturbation @xmath185 .
in particular , in the conservative case @xmath186 we have _ ( , ) = - ( , ) , _ ( , ) = - ( , ) [ eqq ] . *
. kepler elements . *
our aim is to re - express the system ( [ ero ] ) ( [ ete ] ) after a change of coordinates @xmath187 where @xmath188 are related to the unperturbed kepler problem ( as in eqs .
( [ ero ] ) ( [ ete ] ) , with @xmath189 and @xmath190 ) . for
the sake of brevity , let us introduce the abbreviations ( , , ) , ( , ) ( , , , ) .
for given @xmath191 , we define : = ; as usually , we refer to this as the osculating kepler orbit .
let @xmath192 ( circular or rectilinear trajectories being excluded ) , and define maps ( , ) p ( , ) ( 0,+ ) , e ( , ) ( 0,1 ) , y ( , ) where @xmath170 , @xmath193 , @xmath194 are , respectively , the parameter@xmath195 , the eccentricity and the argument of the pericenter for the ellipse @xmath196 ; here , r : = ( e.g. , the planetary radius ) . in the above , the standard parameter is divided by @xmath197 to obtain a dimensionless quantity ; from now on , @xmath170 itself will be called the parameter and we will refer to @xmath198 as the kepler elements of @xmath196 . as well known , @xmath199 is the set of states with @xmath200 and negative unperturbed energy : = \ { ( , ) | > 0 , 1 2 ( ^2 + ^2 ^2 ) - g m < 0 } ; [ defdd ] furthermore , p ( , ) = ^4 ^2 r g m , [ yie1 ] e ( , ) = , [ yie2 ] y ( , ) : = [ yie3 ] @xmath201 to continue , we introduce the notation i ( p , e , y ) ( i^p , i^e , i^y ) and write @xmath202 for the correspondence ( [ yie1])-([yie3 ] ) .
the mapping ( 0,+ ) ( 0,1 ) ^2 , ( , ) ( i
( , ) , ) [ maps ] is one - to - one , with inverse ( 0,+ ) ( 0,1 ) ^2 , [ inv0 ] @xmath203 ( i , ) : = e ( - y ) , [ inv2 ] ( i , ) : = ( 1 + e ( - y))^2 , [ inv3 ] ( i , ) : = r p 1 + e ( - y ) [ inv1 ] ( see appendix [ appkepl ] for more details ) . for future use , hereafter we give the apocenter @xmath204 , the pericenter @xmath205 and the orbital period for the unperturbed kepler motion on an ellipse of parameter @xmath170 and eccentricity @xmath193 ; these are _
( p , e ) = r p 1 e , t_orb(p , e ) = 2 .
[ torb ] of course , the first two equations can be inverted giving p(_+ , _ - ) = 2 _ + _ - r(_+ + _ - ) , e(_+ , _ - ) = _ + - _ - _ + + _ - . * .
the equations of motions in terms of the kepler elements .
* using the transformations ( [ yie1])([inv1 ] ) and eqs . ( [ ero ] ) ( [ ete ] ) , one easily obtains a system of differential equations for @xmath206 , namely = 2 _ ( i , ) , [ sp ] = ( - y ) _ ( i , ) [ se ] @xmath207 = - _ ( i , ) [ sy ] @xmath208 = ( 1 + e ( - y))^2 , [ ste ] _ a(i , ) : = _ a((i , ) , ) , ( a \ { , } ) .
as expected , for @xmath209 the above equations reproduce the fact that @xmath210 are constants of the motion for the unperturbed kepler problem . from now on , for technical reasons related to the application of the general averaging method we will think the `` slow '' angle @xmath194 as an element of @xmath211 rather than @xmath48 ; so , the space of kepler elements becomes : = \ { i ( p , e , y ) | p ( 0,+ ) , e ( 0,1 ) , y } .
[ delambda ] taking this viewpoint amounts to identify the vector field defined by the above equations with its lift with respect to the projection @xmath212 , \te ) \in ( 0,1 ) \times ( 0,+\infty ) \times \bt \times \bt$ ] ( recall that @xmath15 : \reali \vain \bt$ ] is the quotient map ) . * . from the physical time @xmath213 to the `` orbit counter '' @xmath214 .
* let us consider the ( maximal ) solution @xmath215 of eqs .
( [ sp])-([ste ] ) , with initial data i(0 ) = i_0 = ( p_0,e_0,y_0 ) , ( 0 ) = 0 , [ condiniz ] and denote by @xmath216 , @xmath217 the unique @xmath218 function such that ( 0 ) = 0 , ( t ) = [ 2 ( t ) ] .
[ ocounter ] then ( t ) = 2 ( t ) , and the positivity of @xmath219 ensures the function @xmath217 to be increasing .
the image of this function is an interval @xmath220 , and we denote the inverse function by t ( ) : [ 0 , _ ) [ 0,t _ ) , t ( ) . to go on ,
let us define i ( ) : = [ i(t)]_t = t ( ) for @xmath221 . then @xmath222 @xmath223 and eqs .
( [ sp])-([ste ] ) with the data ( [ condiniz ] ) yield = 4 p^2 ( 1 + e ( - y))^2 . r _ ( i , ) g m |_= [ 2 ] , p(0 ) = p_0 , [ spte ] = 2 p^2 ( - y ) ( 1 + e ( - y))^2 r^2 _ ( i , )
g m [ sete ] @xmath224 } , \qquad e(0 ) = e_0~,\ ] ] = - 2 r^2 _
( i , ) g m [ syte ] @xmath225 } , \qquad y(0 ) = y_0~;\ ] ] we note that @xmath12 , @xmath226 and @xmath227 are dimensionless quantities
( [ spte])([syte ] ) are a system of the general form ( [ evol ] ) for the function @xmath228 .
( [ ocounter ] ) indicates that @xmath229 is the number of orbits performed by the satellite from time @xmath230 to time @xmath11 ; this explains the name of _ orbit counter _ employed for the @xmath12 variable .
polar motions of a satellite around an oblate planet .
* from now on , the perturbation is due to a deviation of the planetary mass from the uniform distribution inside a sphere .
the mass distribution is assumed to be symmetric with respect to the polar axis ; @xmath173 is the midpoint between the poles , and we introduce a system of cartesian coordinates @xmath231 with origin @xmath173 , with the @xmath232 axis ( indicated by a unit vector @xmath233 ) as polar axis ( to be so small that @xmath173 can be regarded at rest . ] ) .
the total gravitational potential @xmath234 produced by the planet admits a well known expansion in zonal harmonics , namely , v(p ) = - g m |p - o| [ geopot ] where @xmath235 are the legendre polynomials , and j _ : = - 1 m r^ p_(z(q ) |q - o| ) | q - o |^ d m(q ) . in the above : @xmath172 is total mass of the planet ; @xmath197 is a typical length expressing the size of the planet , usually chosen as the equatorial radius ; @xmath235 are the legendre polynomials ; @xmath236 is the measure describing the mass distribution of the planet .
we note that @xmath237 are dimensionless due to the presence of the parameters @xmath172 and @xmath197 in their definitions ; we will refer to them as the ( dimensionless ) dipole , quadrupole , octupole , ... coefficients . for
a derivation of ( [ geopot ] ) one can refer , e.g. , to @xcite .
as for the applications of this expansion to ( polar or generic ) satellite motions , there is an enormous literature ; we only cite the classical references @xcite and @xcite . to continue , let us recall that p_1 ( ) : = , p_2 ( ) : = 1 2 ( 3 ^2 - 1 ) ; in general @xmath238 , which implies @xmath239 if @xmath5 is odd and the mass distribution is reflection - invariant with respect to the equatorial plane .
often , @xmath7 even without an exact equatorial symmetry ; in this case , the first non - negligible contribution to the expansion ( [ geopot ] ) is the @xmath8 term . from now on ,
we assume @xmath240 and neglect the terms of ( [ geopot ] ) with @xmath241 ; explicitating @xmath242 , we finally obtain v(p ) = v_(p ) + w(p ) , [ vp ] where @xmath243 is the keplerian potential and w(p ) :
= - g m r^2 | p - o |^3 ( 1 - 3 z^2(p ) |p - o|^2 ) , [ wp ] : = j_2 2 = 1 4 m r^2 ( |q - o|^2 - 3 z^2(q ) ) d m(q ) . in the sequel
we always regard eq .
( [ vp ] ) as giving the exact potential in the region where the satellite is moving , and assume @xmath18 ( and the mass distribution is uniform , i.e. , @xmath244 with @xmath245 the total volume .
then , passing to cylindrical coordinates @xmath246 we obtain @xmath247 which implies @xmath18 if @xmath248 .
as a supplementary remark , we mention that the analogous of @xmath249 employed in @xcite has a wrong sign . ] ) .
the values of @xmath16 when the planet is the earth will be reported later on . by a _ polar motion _ of the satellite , we mean one with initial position and velocity in a plane @xmath43 containing the polar axis .
such a motion stays in @xmath43 for all times , since the gravitational force at any point of @xmath43 is parallel to this plane . to analyze the motion we use on @xmath43 a system of polar coordinates @xmath250 , with @xmath3 the angle from @xmath233 . in these coordinates , w ( , )
= - g m r^2 ^3 ( 1 - 3 ^2 ) , and the motion is described by eqs .
( [ ero])([eqq ] ) . a polar motion in the domain @xmath199 of eq . ( [ defdd ] ) can be described via the kepler elements @xmath251 , and we can write down eqs .
( [ sp])([ste ] ) or ( [ spte])([syte ] ) for the present choice of the perturbation .
the explicit form of eqs .
( [ spte])([syte ] ) is = f^(i,[2 ] ) , = f^(i,[2 ] ) , = f^(i,[2 ] ) , [ eqpey ] @xmath252 depending on the function f = ( f^i)_i= , , : ^3 , ( i , ) f(i , ) , [ fp ] @xmath253 @xmath254 @xmath255 @xmath256 @xmath257 @xmath258 @xmath259 this is the system to which we will apply the general scheme of section [ secav ] .
we take as a starting point the formulation of the problem in terms of the variables @xmath260 , based on the orbit counter @xmath12 as independent variable .
so , the evolution equations have the form ( [ evol ] ) with @xmath261 as in eqs . ( [ fp ] ) . from now on , for consistency with the general notation of sections [ intro ] and [ secav ] , we write = ( , , ) for the ( maximal ) solution of these equations for given initial data
( we repeat that typeface symbols denote functions of @xmath12 or @xmath262 ; thus @xmath263 is a function , to be distinguished from a point @xmath264 of the space @xmath46 ) . throughout the section ,
indices @xmath265 range in @xmath266 . in the sequel
, we compute the averaged equations of motion and their solutions ; then we will evaluate the error of averaging with the general method of section [ secav ] . * . the averaged system . * the averages over @xmath118 of the functions ( [ fp ] ) are @xmath267 , where , for all @xmath264 , ^ ( i ) = 0 , ^ ( i ) = 0 , ^ ( i ) = - 3 p^2 .
[ mf ] so , the averaged system for @xmath268 is = 0 , d ^d = 0 , d ^d = - 3 ( ^)^2 , ( ^ , ^ , ^)(0 ) = ( p_0 , e_0 , y_0 ) , and has solution ^ ( ) = p_0 , ^()= e_0 , ^ ( ) = y_0 - 3 ^2 [ solmed ] ( which stays in @xmath46 for all @xmath269 ) . as we see , in the approximation given by averaging , the parameter and eccentricity are constant , while the argument of the pericenter varies linearly with the rescaled time @xmath270 ; this is a classical result . * . the auxiliary functions @xmath271 @xmath272 .
* these are required by our general method to evaluate the error of the averaging method ; their definitions are found in eqs .
( [ thef])([em ] ) .
we have computed all these functions by means of mathematica . in the case of @xmath273 , the components are & s^&(i , ) = - 1 p , [ sx ] + & s^&(i , ) = - 1 16 p^2 , + & s^ & ( y , ) = - 1 16 p^2 e . the expressions of @xmath274 are too long to report all of them here . as examples ,
we write down the components @xmath275 and @xmath276 , which are & v^ & ( i , ) = , + & u^ & ( i , ) = 3 128 p^5 . for future use , it is convenient to point out that ^ ( i ) & = & - 3 e^2 2 p^3 ( 2 y ) , [ compp1 ] + ^ ( i ) & = & 3 4 p^4 ( 10 e - e^3 ) ( 2 y ) , + ^ ( i ) & = & 3 16 p^4 ; ( i ) = 6 p^3 , ^i_j(i ) = 0 ; ^i_j k(i ) = 0 . to go on ,
we take @xmath159 as in ( [ tays ] ) and @xmath95 , @xmath96 as in ( [ tayf ] ) ; so , ^i(i , i , ) = _ 0 ^ 1 d
s^i i^j(i+ i , ) ; [ touses ] ^i_j(i , i ) = _ 0 ^ 1 d ^i i^j ( i + i ) ; [ touse ] ^i_j k(i , i ) = _ 0 ^ 1 d ( 1 - ) ^2 ^i i^j i^k(i + i ) , [ he1 ] the derivatives of @xmath277 , @xmath267 and @xmath278 being computed from eqs .
( [ sx ] ) ( [ mf ] ) and ( [ compp1 ] ) . in particular , we mention that ( i ) = - 18 p^4 , ( i)= 0 . * . the functions @xmath279 ; time intervals .
* these functions are the solutions of eqs .
( [ sistr ] ) ( [ sistk ] ) , and can be computed in an elementary way .
the expressions for @xmath104 and its inverse matrix are ( ) = ( 1 & 0 & 0 + 0 & 1 & 0 + + & 0 & 1 ) ; ^-1 ( ) = ( 1 & 0 & 0 + 0 & 1 & 0 + + - & 0 & 1 ) ; [ rrinv ] concerning @xmath106 , we find ^ ( ) = e^2_0 4 p_0 , [ kkx ] @xmath280~,\ ] ] @xmath281 \tau + { 20 + e^2_0 \over 16 p^2_0 } \left[\sin(2 y_0 ) - \sin(2 y_0 - { 6 \pi \over p_0 ^ 2}\ , \tau ) \right]~.\ ] ] in the above , @xmath282 ; however , from now on we put a limitation @xmath105 , for some finite @xmath27 ; consequently , the `` orbit counter '' @xmath12 will range in @xmath140 . * . the auxiliary functions @xmath283 ... , @xmath284 , @xmath285 , @xmath286 .
* these functions must be constructed so as to fulfill the inequalities in section [ secav ] .
first of all , we must find a function @xmath287 ^ 3)$ ] such that @xmath288 for all @xmath269 , with @xmath46 as in eq .
( [ delambda ] ) . the cited equation can be satisfied putting ^ ( ) : = p_0 , ^ ( ) : = ( e_0 , 1 - e_0 ) , ^ ( ) : = + for @xmath105 ( recall that @xmath289 and @xmath290 ) . from these functions ,
we define @xmath291 as in eq .
( [ dero ] ) ; this has elements @xmath292 . to go on ,
we must find a system of auxiliary functions c^([0,u ) , [ 0,+ ) ) , a^i_j , b^i , c^i , d^i_j , e^i_j k c^ ( , [ 0,+ ) ) such that eqs .
( [ faz ] ) ( [ faa ] ) and ( [ fb])([fe ] ) be satisfied . as explained in subsection 2f , each function @xmath293 should give an accurate upper bound for the left - hand side of eq .
( [ faz ] ) ; for this reason we have developed an algorithm based on numerical maximization over @xmath3 and subsequent interpolation in @xmath269 , see appendix [ appalg ] .
let us pass to the functions @xmath294 ; we refer again to subsection 2f , suggesting to find these functions by fairly rough majorizations of the left - hand sides of eqs .
( [ faa ] ) and ( [ fb])([fe ] ) .
the expressions to be bounded are very lengthy in some cases ; they were majorized with the method described in appendix [ appeae ] ( sometimes using the symbolic mode of mathematica for the necessary computations ) . here , we only report the final expressions of the majorizing functions , which are in fact @xmath269-independent ; for this reason we write @xmath295 , @xmath296 , etc . , instead of @xmath297 , @xmath298 , etc ..
it should also be noted that the dependence on @xmath121 is through the components @xmath299 , @xmath300 . in the sequel , for the sake of brevity
we put p _ :
= p_0 r^ , e _ : = e_0 r^ ; with these notations , we can take a^_(r ) & : = & 3 + 4 ^2 , a^_(r ) : = 4 p_- , a^_(r ) : = 4 p_- , [ axx ] + a^_(r ) & : = & 32 + 45 + 32 ^2 4 ^3 , a^_(r ) : = 45 + 64 8 ^2 , a^_(r ) : = 16 + 15 + 20 ^2 4 ^2 , + a^_(r ) & : = & 32 + 33 + 29 ^2 4 ^3 , a^_(r ) : = 32 + 29 ^2 8 ^2 ^2 , a^_(r ) : = 32 + 30 + 37 ^2 8 ^2 ; b^(r ) & : = & , [ bx ] + b^(r ) & : = & 6112 + 10832 + 6940 ^2 + 11372 ^3 + 1441 ^4 512 ^4 , + b^(r ) & : = & 1 256 p_0 ^ 3 ^2 ^4 ; + c^(r ) & : = & 3 , [ cx ] + c^(r ) & : = & 3 2048 ^6 ^2 , + c^(r ) & : = & 1024 p_0 ^ 3 ^3 ^6 ; d^_(r ) & : = & 9 ^2 2 ^4 , d^_(r ) : = 3 ^3 , d^_(r ) : = 3 ^2 ^3 , [ dxx ] + d^_(r ) & : = & 3 ( 10 + ^2 ) ^5 , d^_(r ) : = 3 ( 10 + 3 ^2 ) 4 ^4 , d^_(r ) : = 3 ( 10 + ^2 ) 2 ^4 , + d^_(r ) & : = & 3 ( 74 + 35 ^2 ) 4 ^5 , d^_(r ) : = 105 8 ^4 , d^_(r ) : = 15 ( 4 + ^2 ) 4 ^4 ; + e^_(r ) & : = & 18 ^4 , e^i_j k ( , r ) : = 0 . [ eyh ] from the above functions we obtain a set of @xmath301 functions @xmath302 , @xmath303 , specializing the prescriptions ( [ al ] ) ( [ ga ] ) to the present case .
explicitly , ^i ( , r ) : = a^i ( , r ) + b^i(r ) = ( ) + a^i_k(r ) r^k + b^i(r ) , [ ali ] ^i(,r , ) ^i(r , ) : = c^i(r ) + ^i_j(r ) ^+ 1 2 e^i_j k(r ) ^j ^k , [ gai ] for @xmath304 and @xmath305 . * .
the matrix @xmath306 and its inverse .
* from eq .
( [ ali ] ) , one finds that the derivatives of @xmath302 with respect to the @xmath121 variables depend on @xmath121 but not on @xmath269 . from the same equation
, we see that our matrix has components ^i_j - ( r ) = ^i_j - ( a^i_k r^j(r ) r^k + a^i_j(r ) + ( r ) ) , ( i , j= , , ) ; [ matel ] all the derivatives in the right hand side are easily obtained from eq.s ( [ axx ] ) ( [ bx ] ) .
( [ quattro ] ) , that is basic in our approach , contains the inverse of this matrix ; this could be written explicitly , but its expression is uselessly complicated .
for this reason , in the subsequent numerical computations we use an approximation giving the inverse up to a third order error @xmath307 , for @xmath308 ( ) , the matrix in which we are interested is evaluated with @xmath309 ; with this position for @xmath121 , ( [ invapp ] ) gives an inverse up to @xmath310 . ] ) . neglecting this error , we have ( 1 - ( r))^-1 = 1 + m + r^k n_(k ) + ^2 , [ invapp ] ( with a sum for @xmath311 ) , where we have introduced the matrices m : = ( & & + & & + & & ) ; [ emme ] n _ ( ) & : = & ( & & + & & + & & ) , + n _ ( ) & : = & ( & & + & & + & & ) , + n _ ( ) & : = & ( & & + & & + & & ) ; [ enne ] @xmath312 ^ _ & : = & 10208 + 27728 e_0 + 44440 e_0 ^ 2 + 46244 e_0 ^ 3 + 18369 e_0 ^ 4 , + ^ _ & : = & 14304 + 29344 e_0 + 71068 e_0 ^ 2 + 121568 e_0 ^ 3 + 65637 e_0 ^ 4 , + ^ _ & : = & 512 + 1680 e_0 + 4405 e_0 ^ 2 + 4455 e_0 ^ 3 + 3044 e_0 ^ 4 , + ^ _ & : = & 7616 + 24832 e_0 + 25096 e_0 ^ 2 + 27300 e_0 ^ 3 + 7719 e_0 ^ 4 , + ^ _ & : = & 5568 + 29376 e_0 + 33400 e_0 ^ 2 + 42444 e_0 ^ 3 + 15153 e_0 ^ 4 , + ^ _ & : = & 2048 + 2880 e_0 + 7524 e_0 ^ 2 + 5202 e_0 ^ 3 + 4385 e_0 ^ 4 .
[ qu ] for more details on the approximate inverse ( [ invapp ] ) , see appendix [ appinv ] . * .
the functions @xmath313 , @xmath314 .
* according to eq .
( [ rp ] ) , these should fulfill the inequalities | ( ) |^i_j r^i_j ( ) , | ^-1 ( ) |^i_j p^i_j ( ) . on account of eq .
( [ rrinv ] ) , we can take ( r^i_j ( ) ) : = ( p^i_j ( ) ) : = ( 1 & 0 & 0 + 0 & 1 & 0 + + & 0 & 1 ) . * .
the main result : the estimates @xmath315 from proposition [ proprinc ] .
the `` @xmath316-operation '' .
* we are finally ready to discuss the difference (t ) : = 1 [ ( t ) - ( t ) ] = 1 ( (t ) - p_0 , ( t ) - e_0 , ( t ) - ^(t ) ) [ diff ] where @xmath317 is the solution of eqs .
( [ eqpey ] ) .
we follow the scheme of proposition [ proprinc ] ; this requires to determine @xmath318 by solving a fixed point problem ^i(0 , _ 0 ) = ^i_0 . in the examples that follow @xmath132
is always found numerically , as explained in remark [ rema2](a ) .
after @xmath132 has been found , we must solve ( again numerically ) a cauchy problem for the unknown functions @xmath319 , i.e. , = p^i_j ^j ( , ) , _
^i(0 ) = 0 , [ trei ] = ( 1 - ( , ) ) ^-1 , i _ k ( ^k ( , ) + r^k_h p^h_j ^j ( , ) + _ ^j ) , [ quattroi ] @xmath134 with the domain conditions ( [ due ] ) , having in this case the form @xmath320 ( 1 - ( , ) ) > 0 .
[ duei ] if the above problem has solution on @xmath35 , our general framework grants that the solution @xmath10 of ( [ eqpey ] ) exists on @xmath140 , and gives the bounds | ^i(t ) | ^i(t ) , i.e. , a slight variation of the terminology of @xcite @xcite , we call `` @xmath321-operation '' the execution of _ all _ the numerical computations required by the present approach to obtain the estimators @xmath322 , given the initial data @xmath323 and the values of @xmath324 .
these computations include : _ _
\(a ) the numerical evaluations and subsequent interpolations , necessary to determine the functions @xmath293 @xmath45 on @xmath35 .
we have already mentioned appendix [ appalg ] , that describes everything in detail ; \(b ) the determination of the fixed point @xmath132 ; \(c ) the numerical solution of the cauchy problem ( [ trei ] ) ( [ quattroi ] ) for @xmath325 , @xmath322 . from now on
, we will indicate with @xmath326 the cpu time required to perform ( a)(b)(c ) ; this time depends on the chosen hardware and software , but is anyhow useful for comparison with other computational approaches , such as the one described hereafter . * .
the `` @xmath327-operation '' : a way to test the efficiency of the previous estimates . *
this operation was already considered in @xcite , with a role similar to the one that we ascribe to it here : checking the reliability of the @xmath321- estimates by a direct numerical treatment of the system ( [ eqpey ] ) , for @xmath328 .
the direct attack of ( [ eqpey ] ) is possible when the time scale @xmath28 is not overwhelmingly large ; for larger @xmath28 , the @xmath327-operation is too expensive , and this is just the case where the @xmath321- procedure becomes more useful . the direct attack to ( [ eqpey ] ) is performed substituting in these equations @xmath329 , @xmath330 , @xmath331 , which gives rise to the equations ( t ) = f^(p_0 + ^(t ) , e_0 + ^(t ) , ^(t ) + ^(t ) ) , [ pertel ] @xmath332 @xmath333 @xmath334 by definition , the @xmath327-operation is the numerical solution of this cauchy problem in the unknowns @xmath335 , for @xmath328 ; @xmath336 will indicate the necessary cpu time .
* introducing the examples .
* in the case of the earth , g m = 3.98600442 10 ^ 14 ^3 ^2 , r = 6.378135 10 ^ 6 , @xmath337 so , for a satellite on a polar orbit around the earth , the unperturbed apogee and perigee @xmath338 , @xmath339 and the orbital period @xmath340 are determined in this way by the initial data @xmath341 : @xmath342 in the sequel we present two examples , with initial data @xmath341 very similar to the actual data of two real satellites ( polar and cos - b ) .
the mathematica package , already mentioned in relation to the symbolic treatment of the problem , has been employed ( on a pc ) for the required numerical computations .
we repeat a comment of the introduction : the examples are not fully realistic since they do not account for perturbations different from the @xmath0 gravitational term ; nevertheless , we think they have some interest because they show the effectiveness of the method , suggesting that our approach could be applied as well including other perturbations .
each example is worked out along these lines : \(i ) firstly , we take @xmath343 .
we perform both the @xmath321- and the @xmath327-operations , using the second to test the first one .
we give figures reporting the graphs of @xmath344 ( rapidly oscillating ) and @xmath345 , for @xmath346 and @xmath328 ; the cpu times @xmath326 , @xmath336 are also indicated .
\(ii ) next , we choose @xmath347 .
the @xmath321-operation is still performed within short cpu times ; we give figures reporting @xmath326 and the graphs of the estimators @xmath348 . on the contrary , the @xmath327-operation exceeds the capabilities of the machine employed .
* example 1 .
* we take p_0 : = 3.000 , e_0 : = 0.6640 , y_0 : = 0.0000 , corresponding to _ + = 56950 , _
- = 11500 , t_orb = 17.50 .
these data are very similar to the initial conditions of the polar satellite , taking @xmath349 on april 1997 @xcite .
( [ solmed ] ) for the solution of the averaged system on any interval @xmath35 gives @xmath350 , @xmath351 , and ^ ( ) = - 1.047 .
figures 1a , 1b and 1c report the graphs of @xmath352 and @xmath353 for @xmath354 $ ] , as produced by the @xmath327- and @xmath321-operations .
we note that @xmath355 years . the cpu times ( in seconds ) for @xmath321 and @xmath327
reported in fig.1a refer to the overall calculation , also including the @xmath356 and @xmath357 components plotted in figures 1b , 1c . figures 1d , 1e and 1f report the graphs of @xmath352 and @xmath358 for @xmath359 $ ] , as produced by the @xmath321-operations .
we note that @xmath360 years . the cpu time indicated in fig.1d
also includes calculations of the @xmath356 and @xmath357 components , plotted in figures 1e , 1f . in comparison with the previous case
@xmath354 $ ] , the number of orbits is increased by a factor @xmath361 ; in spite of this , @xmath362 changes by a factor @xmath363 .
0.2 cm * example 2 .
* we take p_0 : = 1.973 , e_0 = 0.8817 , y_0 : = 0.9600 , corresponding to _
+ = 106400 , _ - = 6688 , t_orb = 37.16 .
these data are very similar to the initial conditions at launch ( august 1975 ) of the cos - b satellite @xcite .
( [ solmed ] ) for the solution of the averaged system on any interval @xmath35 gives @xmath350 , @xmath351 and ^ ( ) = 0.9600 - 2.421 .
all figures for this example give the same information as the corresponding ones of example 1 .
in particular , figures 2a , 2b and 2c refer to computations for @xmath354 $ ] , while figures 2d , 2e , 2f refer to the case @xmath364 $ ] .
we note that @xmath365 years and @xmath366 years . the cpu times ( in seconds ) for @xmath321 and @xmath327
are similar to the ones in example 1 , so the comments given therein also apply to this case . 0.4 cm 0.4 cm 0.4 cm 0.4 cm 0.4 cm 0.4 cm
we refer to the framework of subsection , having fixed @xmath367 .
the unperturbed kepler orbit @xmath196 with these initial data at time @xmath349 is an ellipse with ( dimensionless ) parameter @xmath368 , eccentricity @xmath369 and argument of the pericenter @xmath370 . in any textbook on classical mechanics ,
one finds the equations = r p 1 + e ( - y ) , [ conics0 ] = r p e ( - y ) ( 1 + e ( - y))^2 , [ conics ] r
p = ^4 ^2 g m , [ para ] e = .
[ ecc ] eq .
( [ conics0 ] ) is the representation of a conic in polar coordinates ; eq .
( [ conics ] ) follows taking the derivative of this polar representation with respect to time along the unperturbed kepler motion , i.e. , regarding @xmath188 as time independent .
( [ para ] ) ( [ ecc ] ) give the parameter and the eccentricity as functions of the angular momentum and the total energy ( per unit mass ) @xmath371 and @xmath372 .
we note that eq .
( [ conics0 ] ) and eqs .
( [ para ] ) ( [ ecc ] ) are equivalent to eq .
( [ conics0 ] ) , to the square of eq.([conics ] ) and to eq .
( [ para ] ) .
( [ conics ] ) is important because ( due to @xmath200 in @xmath199 ) it implies = ( - y ) ; [ sign]this information is essential to determine @xmath194 as a function of @xmath373 .
( [ conics0])-([sign ] ) yield the expressions for @xmath188 in eqs .
( [ yie1])-([yie3 ] ) ; from here , one also proves the bijectivity of the map @xmath374 and the expression ( [ inv0])-([inv1 ] ) for its inverse . to conclude ,
let us comment on the equations ( [ torb ] ) for the apocenter @xmath338 , the pericenter @xmath339 and the orbital period @xmath340 of the unperturbed kepler motion along an ellipse of parameter @xmath170 and eccentricity @xmath193 .
the expressions for @xmath375 are consequences of the polar equation for the ellipse ; the expression for @xmath340 follows from kepler s third law t_orb = 2 ( ) ^3/2 = 2 ( _ + 2 + _ - 2 ) ^3/2 and from the previously mentioned formulas for @xmath375 .
let us consider a general periodic system ( [ evol ] ) , and suppose we have the related functions @xmath377 , for @xmath269 within a finite interval @xmath35 .
hereafter we outline a simple , but effective scheme to construct by mixed , numerical and analytical , techniques a function @xmath378 fulfilling _ up to small errors _ the inequality ( [ faz ] ) , for any given @xmath379 ; this method is easily implemented on mathematica .
the scheme consists of four items .
\(i ) take on the torus @xmath48 a grid of equally spaced angles _ q : = , ( q = 1,2, ... ,q ) .
[ teq ] ( ii ) take in @xmath35 a grid of equally spaced instants _
n n , n = 0,1, ... ,n-1
. [ taun ] ( iii ) for each @xmath148 , compute numerically : = _ q=1, ...
,q | ( s((_n),_q ) - ( _ n ) s(i_0 , _ 0 ) - ( _ n ) ) ^i | ; [ ain ] then , for all @xmath118 , | ( s((_n ) , ) - ( ) s(i_0 , _ 0 ) - ( _ n ) ) ^i | up to an error neglected in the sequel , which is of order @xmath380 for @xmath381 .
\(iv ) interpolate the sequence @xmath382 ( @xmath383 , finding a smooth function @xmath378 such that ( _ n ) = more precisely , we define @xmath293 to be the lagrange polynomial ( restricted to @xmath35 ) such that @xmath384 for all @xmath148 ; thus ( ) = _
n=0^n-1 ( _ ^n-1 - _ m _
n - _ m ) , [ ai ] with @xmath385 and @xmath386 as in eqs .
( [ taun ] ) ( [ ain ] ) . by construction , the inequality ( [ faz ] ) @xmath387
is fulfilled , up to a small error , at all points @xmath388 ; we assume the same to happen everywhere in @xmath35 .
of course , being @xmath301 , the functions ( [ ai ] ) fulfill the regularity conditions for application of the basic proposition [ proprinc ] .
the above scheme ( i) ...
(iv ) has been employed to construct the functions @xmath293 ( @xmath389 ) in the numerical examples of section [ seces ] on the motions of satellites . in these examples , @xmath390 and @xmath391 ( with such an @xmath392 , the lagrange interpolating polynomials are computed very quickly by mathematica ) .
admittedly , the approach followed does not account for the small errors mentioned in items ( iii ) ( iv ) : throughout the paper these errors , and the ones produced by the approximate inversion of the matrix @xmath393 , are the only ones for which an analytical estimate is not provided .
as usually , the indices @xmath396 range in @xmath266 .
* finding @xmath397 . * as an example , let us illustrate the determination of @xmath398 . to this purpose , we use eq .
( [ touses ] ) with @xmath399 , @xmath400 and @xmath401 , giving ^_((),j , ) = _ 0 ^ 1 d ( p_0 + j^ , e_0 + j^ , ^()+ j^ , ) @xmath402 in the last passage , the derivative @xmath403 has been computed from eq .
( [ sx ] ) .
now we take the absolute value in the last equation , and use the elementary inequalities @xmath404 , @xmath405 , @xmath406 , @xmath407 . in this way
we get | ^_((),j , ) | 4 e_0 + |j^| p_0 - | j^| = a^_(r ) |_r^i = |j^i| , where @xmath408 ; this definition of @xmath398 agrees with eq .
( [ axx ] ) ( where we use the abbreviations @xmath409 , @xmath410 ) . * finding @xmath411 and @xmath412 . *
we illustrate the determination of @xmath413 . to this purpose
, we must first compute the @xmath414 component of the function on the left - hand side of eq .
( [ fb ] ) ( for @xmath21 and @xmath415 @xmath45 . recalling again that @xmath289 and @xmath290 for all @xmath269 , one finds ( ( ( ) + j , ) - ( ( ) ) ( ( ) + j , ) ) ^@xmath416 where the coefficients @xmath417 are polynomials in @xmath418 ; for example , p_0 1 = 256 ( e_0 + j^ ) ( -49 p_0 ^ 3 + 13 p_0 ^ 3 ( e_0 + j^)^2 - 16 ( p_0 + j^)^3 ( e_0 + j^)^2 ) .
this implies | ( ( ( ) + j , ) - ( ( ) ) ( ( ) + j , ) ) ^| @xmath419 the absolute value of each coefficient @xmath417 is easily bounded , using fairly rough estimates such as @xmath420 .
the conclusion is | ( ( ) + j , ) - ( ( ) ) ( ( ) + j , ) |^b^(r ) |_r^i = |j^i| , where @xmath413 is the function appearing in eq .
( [ bx ] ) ( which is independent of @xmath421 ) .
the determination of @xmath422 and @xmath423 is performed similarly , yielding eqs .
( [ bx ] ) ( [ cx ] ) .
* finding @xmath424 .
* as an example , let us consider the determination of @xmath425 .
we use eq .
( [ touse ] ) with @xmath426 , @xmath400 and @xmath401 , giving ^_((),j ) = _ 0 ^ 1 d ( e_0 + j^ , p_0 + j^ , ^()+ j^ ) @xmath427 in the last passage , the derivative @xmath428 has been computed from eq .
( [ compp1 ] ) .
the last equation implies | ^_((),j ) | 3 ( e_0 + @xmath425 as in eq .
( [ dxx ] ) . * finding @xmath284 . *
these functions should bind the absolute values of the components @xmath429 of @xmath96 ; these are provided by eq .
( [ he1 ] ) , with @xmath430 and @xmath401 . from the cited equation
, we see that we can take @xmath431 for @xmath432 . again from ( [ he1 ] ) , we infer | ^ _ ( ( ) , j)| = 36 | _ 0 ^ 1 d | @xmath433 with @xmath434 as in eq .
( [ eyh ] ) .
we must invert the @xmath435 matrix @xmath436 , whose elements are given by eq .
( [ matel ] ) . for @xmath437
, this equation implies 1 - ( r ) = 1 - m - r^k n_(k ) - ^2 z + o_3 ( , r ) , [ ad1 ] where we provisionally write @xmath438 and @xmath439 for the matrices of elements m^i_j : = a^i_j(0 ) , ^i_j : = a^i_k r^j(0 ) + a^i_j r^k(0 ) , z^i_j : = b^i r^j(0 ) , [ add1 ] ( @xmath440 ) . eq .
( [ ad1 ] ) , with the standard @xmath441 expansion @xmath442 , implies @xmath443 @xmath444 + @xmath445 @xmath446 . summing up , ( 1 - ( r))^-1 = 1 + m + r^k n_(k ) + ^2 + o_3 ( , r ) , : = z + m^2 .
[ ad2 ] this justifies the approximate inverse ( [ invapp ] ) if we check that the matrices @xmath438 and @xmath447 , defined here via eq.s ( [ add1 ] ) ( [ ad2 ] ) , coincide with the homonymous matrices in eq.s ( [ emme ] ) ( [ enne ] ) ( [ qu ] ) .
this is obtained substituting in ( [ add1 ] ) ( [ ad2 ] ) the explicit expressions ( [ axx ] ) ( [ bx ] ) of the functions @xmath448 and @xmath449 . *
* this work has been partially supported by the gnfm of istituto nazionale di alta matematica and by miur , research project cofin/2006 `` metodi geometrici nella teoria delle onde non lineari e applicazioni '' . 99 c. morosi , l. pizzocchero , _ on the average principle for one - frequency systems _ ,
j.phys.a:math.gen . * 39 * , 3673 - 3702 ( 2006 ) . c. morosi , l. pizzocchero , _ on the averaging principle for one - frequency systems .
seminorm estimates for the error _ , nonlinear dynam .
, doi 10.1007/s11071 - 008 - 9444 - 1 ( 2008 ) .
arnold , _ mathematical methods of classical mechanics _ , springer - verlag , new york ( 1978 ) .
p. lochak , c. meunier , _ multiphase averaging for classical systems _ , springer - verlag , new york ( 1988 ) .
sanders , f. verhulst , _ averaging methods in nonlinear dynamical systems _ , springer - verlag , new york ( 1985 ) .
wallace , _ mathematical analysis of physical problems _ , dover , new york ( 1972 ) , pages 125 - 129 .
see also the entries `` multipole expansion '' , `` gravitational multipole expansion '' and `` gravitational moment '' in eric weisstein s _ world of science _ , http://scienceworld.wolfram.com w.t .
kyner , _ a mathematical theory of the orbits about an oblate planet _ , siam j. appl
. math . * 13 * , 136 - 171 ( 1965 ) .
see the plots of orbits at the nasa site , http://pwg.gsfc.nasa.gov/polar see the orbital parameters at the esa site , http://sci.esa.int/science-e/www/object/index.cfm?fobjectid=31480 | this paper is related to our previous works @xcite @xcite on the error estimate of the averaging technique , for systems with one fast angular variable . in the cited references , a general method ( of mixed analytical and numerical type )
has been introduced to obtain precise , fully quantitative estimates on the averaging error . here
, this procedure is applied to the motion of a satellite in a polar orbit around an oblate planet , retaining only the @xmath0 term in the multipole expansion of the gravitational potential . to exemplify the method ,
the averaging errors are estimated for the data corresponding to two earth satellites ; for a very large number of orbits , computation of our estimators is much less expensive than the direct numerical solution of the equations of motion .
p h v s l o _0 _
m 1*t*^1 s 1 _ addtoresetequationsection on the averaging principle for one - frequency systems .
an application to satellite motions .
carlo morosi@xmath1 , livio pizzocchero@xmath2 + @xmath1 dipartimento di matematica , politecnico di milano , + p.za l. da vinci 32 , i-20133 milano , italy + e mail : [email protected] + @xmath2 dipartimento di matematica , universit di milano + via c. saldini 50 , i-20133 milano , italy + and istituto nazionale di fisica nucleare , sezione di milano , italy + e mail : [email protected] * keywords : * slow and fast motions , perturbations , averaging method .
* ams 2000 subject classification * : 70k65 , 70k70 , 34c29 , 70h09 , 37j40 , 70f15 . |
let @xmath0 be an oriented smooth manifold of dimension @xmath1 , @xmath2 a semisimple lie group , @xmath3 its lie algebra , @xmath4 a principal @xmath2-bundle over @xmath0 , @xmath5 a connection 1-form on @xmath4 and @xmath6 its curvature .
we assume that @xmath0 is a compact manifold without boundary .
a connection 1-form @xmath5 on @xmath4 is called _ flat _ if its curvature @xmath7 vanishes , @xmath8 locally eqs.(1.1a ) are solved trivially , and on any sufficiently small open set @xmath9 we have @xmath10 , where @xmath11 .
so , locally @xmath5 is a pure gauge .
but globally eqs.(1.1a ) are nontrivial , and finding their solutions is not an easy problem .
it is well known that in the case @xmath12 eqs.(1.1a ) are the field equations of chern - simons theories which describe flat connections on @xmath2-bundles over 3-manifolds @xmath0 .
quantum observables of these theories are topological invariants of @xmath0 @xcite . for a generalization of chern - simons theories to arbitrary dimensions , in addition to a connection @xmath5 on a @xmath2-bundle @xmath13 ,
one considers a @xmath14-form @xmath15 with values in the adjoint bundle @xmath16 . using the fields @xmath15 and @xmath7 , one introduces so - called bf theories @xcite , for which the variation of the action w.r.t . @xmath15 gives eqs.(1.1a ) , and the variation of the action w.r.t . @xmath5
gives the equations @xmath17 where @xmath18 is the covariant differential on @xmath4 .
for more details and references see @xcite .
the purpose of our paper is to describe the symmetry group of eqs.(1.1 ) arising as equations of motion in chern - simons and topological bf theories . under the symmetry group of a system of differential equations
we understand the group that maps solutions of this system into one another . only gauge ( and related to them ) symmetries of eqs.(1.1 ) forming a small subgroup in the symmetry group have been considered in the literature . in this paper
we show how _ all _ symmetries of eqs.(1.1 ) can be calculated with the help of _ deformation theory _ methods .
in fact , finding solutions and symmetries of eqs.(1.1 ) will be reduced to solving functional equations on some matrices . to illustrate the method
, we write down some nontrivial explicit solutions of these functional equations and describe symmetries corresponding to them .
in this section we recall some definitions to be used in the following and fix the notation . in sect.2.1
we introduce bundles @xmath19 of lie groups , bundles @xmath20 of lie algebras and sheaves of their sections @xcite . in considering sheaves of non - abelian groups
we follow @xcite and the papers @xcite . in sect.2.2
we recall definitions @xcite of cohomology sets of manifolds with values in the sheaves of lie groups .
let @xmath21 be a set of representatives of topological equivalence classes of @xmath2-bundles over @xmath0 .
we fix @xmath22 and choose a _ good _ covering @xmath23 of the manifold @xmath0 , which is always possible .
for such a covering @xmath24 each nonempty finite intersection @xmath25 is diffeomorphic to an open ball in @xmath26 .
let @xmath27 , @xmath28 be transition functions determining the bundle @xmath4 in a fixed trivialization .
recall that a map of a subset @xmath9 into some set is called _ locally constant _ if it is constant on each connected component of the set @xmath29 .
the fibre bundle @xmath4 is called _ locally constant _ if its transition functions @xmath30 are locally constant .
all bundles associated with such a bundle are also called locally constant . in the following
, we shall consider deformations of the locally constant bundle @xmath4 .
they are described in terms of cohomologies of the manifold @xmath0 with values in sheaves of sections of some bundles associated with @xmath4 .
below we introduce these sheaves .
let us consider the bundle of groups @xmath31 ( @xmath2 acts on itself by internal automorphisms : @xmath32 ) associated with @xmath4 . consider the sheaf @xmath33 of
_ smooth _ sections of the bundle @xmath19 and its subsheaf @xmath34 of _ locally constant _ sections .
a section @xmath35 of the sheaf @xmath34 ( @xmath33 ) over an open set @xmath9 is described by a collection @xmath36 of locally constant ( smooth ) @xmath2-valued functions @xmath37 on @xmath38 such that @xmath39 on @xmath40 .
we also consider the adjoint bundle @xmath16 of lie algebras and denote by @xmath41 the sheaf of smooth @xmath42-forms on @xmath0 with values in the bundle @xmath20 ( @xmath43 ) .
the space of sections of the sheaf @xmath41 over an open set @xmath29 is the space @xmath44 of smooth @xmath3-valued @xmath42-forms @xmath45 on @xmath38 such that @xmath46 on @xmath40 . in particular , for global sections of the sheaves @xmath41 over @xmath0 one often uses the notation @xmath47 the sheaf @xmath33 ( and @xmath34 ) acts on the sheaves @xmath48 , with the help of the adjoint representation . in particular , for any open set @xmath9 we have @xmath49 where @xmath50 , @xmath51 , @xmath52 , @xmath53 . of course ,
@xmath54 if @xmath55
. denote by @xmath56 an embedding of @xmath34 into @xmath33 .
we define a map @xmath57 given for any open set @xmath29 of the space @xmath0 by the formula @xmath58 where @xmath50 , @xmath59 .
let us also introduce an operator @xmath60 , defined for an open set @xmath9 by the formula @xmath61 where @xmath51 , @xmath62 .
in other words , the maps of sheaves @xmath63 and @xmath64 are defined by means of localization .
finally , we denote by @xmath65 the subsheaf in @xmath66 , consisting of locally defined @xmath3-valued 1-forms @xmath5 such that @xmath67 , i.e. sections @xmath68 of the sheaf @xmath69 over @xmath9 satisfy eqs.(1.1a ) .
let @xmath70 be a sheaf coinciding with either the sheaf @xmath34 or the sheaf @xmath33 introduced in sect.2.1 .
we consider a good open covering @xmath71 of @xmath0 and sections of the sheaf @xmath70 over @xmath72 . in the trivialization over @xmath73
, we can represent @xmath74 by matrices @xmath75 , and in the trivialization over @xmath76 , @xmath77 is represented by @xmath78 where @xmath79 are the transition functions of the bundle @xmath4 . if we do not write a superscript of sections of bundles , we mean that @xmath80 a _
@xmath42-cochain _ of the covering @xmath24 with values in @xmath70 is a collection @xmath81 of sections of the sheaf @xmath70 over nonempty intersections @xmath82 .
a set of @xmath42-cochains is denoted by @xmath83 ; it is a _ group _ under the pointwise multiplication . in particular , for 0-cochains @xmath84 with the coefficients in @xmath70 we have @xmath85 and for 1-cochains @xmath86 we have @xmath87 .
subsets of _ cocycles _
@xmath88 for @xmath89 are defined as follows : @xmath90 z^1({{\mathfrak u } } , s):=\{h\in c^1({{\mathfrak u } } , s ) : h_{\a\b}f_{\a\b}h_{\b\g } f_{\b\g}h_{\g\a}f_{\g\a}=1\ \mbox{on}\ u_\a\cap u_\b\cap u_\g\ne\vn\}. \tag{2.7a}\end{gathered}\ ] ] it follows from ( 2.7a ) that @xmath91 where @xmath92 ( see ( 2.5 ) ) .
we see from ( 2.6 ) that @xmath93 coincides with the group @xmath94 of global sections of the sheaf @xmath70 .
the set @xmath95 is not a group if the structure group @xmath2 of the bundle @xmath4 is a non - abelian group . for @xmath96
we define the left action @xmath97 of the group @xmath98 on the set @xmath99 by the formula @xmath100 a set of orbits of the group @xmath98 in @xmath99 is called a _ 1-cohomology set _ and denoted by @xmath101 . in other words ,
two cocycles @xmath102 are called _ equivalent _ ( cohomologous ) if @xmath103 for some @xmath104 , and the 1-cohomology set @xmath105 is a set of equivalence classes of 1-cocycles . since the chosen covering @xmath71 is acyclic for the sheaf @xmath70 , we have @xmath106 . for more details
see @xcite .
chern - simons and topological bf theories give a field - theoretic description of locally constant ( flat ) bundles which is discussed in this section . in sect.3.1
we introduce the de rham 1-cohomology set describing the moduli space of flat connections and give the de rham description of moduli of covariantly constant fields @xmath15 . using homological algebra methods
@xcite , in sect.3.2 we present the equivalent ech description of the moduli spaces of flat connections and covariantly constant fields @xmath15 .
we denote by @xmath107 the infinite - dimensional group of gauge transformations , i.e. smooth automorphisms of the bundle @xmath13 which induce the identity transformation of @xmath0 .
put another way , @xmath107 is the group of global smooth sections of the bundle @xmath19 @xcite , and therefore we have @xmath108 with the notation of sect.2 .
the space of all connections on @xmath4 is an infinite - dimensional affine space modelled on the vector space @xmath109 of global sections of the sheaf @xmath66 and therefore the space of flat connections on @xmath4 can be identified with the space @xmath110 of global sections of the sheaf @xmath69 ( see ( 2.4 ) ) .
the group ( 3.1 ) acts on the space @xmath110 on the right by the formula @xmath111 where @xmath112 , and we introduce the de rham 1-cohomology set @xmath113 as a set of orbits of the non - abelian group @xmath114 in the set @xmath110 , @xmath115 this definition is a generalization to @xmath20-valued 1-forms @xmath5 and the covariant exterior derivative @xmath18 of the standard definition of the 1st de rham cohomology group @xmath116 where @xmath117 is the exterior derivative
. * remark . * for an abelian group @xmath2 , the group ( 3.1 ) of gauge transformations and the set @xmath118 are abelian groups . in this case ,
( 3.2 ) is reduced in fact to the definition ( 3.3 ) of the standard de rham cohomology .
the definition ( 3.2 ) is nothing but the definition of the moduli space @xmath119 of flat connections on @xmath4 , @xmath120 so , the moduli space @xmath119 of flat connections on @xmath4 is the space of gauge inequivalent solutions to eqs.(1.1a ) as it should be . equations ( 1.1b ) are linear in @xmath15 . for any fixed flat connection @xmath5 , the moduli space of solutions to eqs.(1.1b )
is a vector space @xmath121 since solutions of the form @xmath122 are considered as trivial ( topological `` symmetry '' @xmath123 ) @xcite .
nontrivial solutions of eqs.(1.1b ) form the vector space @xmath124 depending on a solution @xmath5 of eqs.(1.1a ) .
the space of solutions to eqs.(1.1 ) forms a vector bundle @xmath125 , the base space of which is the space @xmath110 of solutions to eqs.(1.1a ) , and fibres of the bundle at points @xmath126 are the vector spaces @xmath124 of nontrivial solutions to eqs.(1.1b ) .
notice that the gauge group @xmath127 acts on solutions @xmath15 of eqs.(1.1b ) by the formula @xmath128 where @xmath112 .
therefore , identifying points @xmath129 and @xmath130 for any @xmath131 , we obtain the moduli space @xmath132 of solutions to eqs.(1.1 ) .
the space @xmath133 is the vector bundle over the moduli space @xmath119 of flat connections with fibres at points @xmath134\in{{\mathcal m}}_p$ ] isomorphic to the vector space @xmath124 .
we denote by @xmath134 $ ] the gauge equivalence class of a flat connection @xmath5 . by using definitions of the sheaves @xmath135 and
@xmath65 , it is not difficult to verify that the sequence of sheaves @xmath136 is exact . here
@xmath137 is a marked element of the considered sets ( the identity in the sheaf @xmath138 and zero in the sheaf @xmath65 ) . from ( 3.8 ) we obtain the exact sequence of cohomology sets @xcite @xmath139 where the map @xmath140 coincides with the canonical embedding induced by the embedding of sheaves @xmath141 . by definition the set of equivalence classes of locally constant @xmath2-bundles over @xmath0 is parametrized by the ech 1-cohomology set @xmath142 @xcite .
the kernel @xmath143 of the map @xmath140 coincides with a subset of those elements @xmath144 $ ] from @xmath145 which are mapped into the class @xmath146 of bundles @xmath147 diffeomorphic to the bundle @xmath4 defined by the 1-cocycle @xmath148 . in terms of transition functions
@xmath30 in @xmath4 and @xmath149 in @xmath147 this means that there exists a smooth 0-cochain @xmath150 such that @xmath151 i.e. cocycles @xmath148 and @xmath152 are smoothly cohomologous . by virtue of the exactness of the sequence ( 3.9 ) ,
the space @xmath153 is bijective to the quotient space ( 3.2 ) .
so we have @xmath154 i.e. there is a one - to - one correspondence between the moduli space @xmath119 of flat connections on @xmath4 and the moduli space @xmath155 of those locally constant bundles @xmath147 which are diffeomorphic to the bundle @xmath4 .
notice that to @xmath156 , representing @xmath144\in{{\mathop{\mbox{ker}}\nolimits\,}}\rho$ ] with fixed @xmath148 , there corresponds a flat connection @xmath157 representing @xmath134\in h^1_{d_a;p}(x)$ ]
. * remark . *
it is well known that the bundle @xmath13 is always diffeomorphic to the trivial bundle @xmath158 if @xmath0 is a paracompact and simply connected manifold @xcite .
it is also known that if the structure group @xmath2 is connected and simply connected , and @xmath159 , then all @xmath2-bundles over @xmath0 are topologically trivial , i.e. @xmath160 . in both the cases we have @xmath161 since now @xmath162 .
the bijections ( 3.11 ) and ( 3.12 ) are non - abelian variants of the isomorphism between ech and de rham cohomologies . the moduli space ( 3.5 ) of solutions to eqs.(1.1b ) can also be described in terms of ech cohomology .
namely , notice that for a fixed flat connection @xmath163 a general solution of eqs.(1.1b ) has the form @xmath164 where @xmath165 , and @xmath166 is an @xmath167-valued @xmath14-form satisfying the equation @xmath168 therefore the space @xmath169 of nontrivial solutions of eqs.(1.1b ) is isomorphic to the standard @xmath14th de rham cohomology group @xmath170 for forms with values in the bundle @xmath167 .
formulae ( 3.13 ) mean that the space @xmath169 is the `` dressed '' space @xmath171 , and we have @xmath172 on the other hand , it is well known that @xmath173 where @xmath174 is the @xmath14th ech cohomology group of the manifold @xmath0 with the coefficients in the sheaf @xmath175 of locally constant sections of the bundle @xmath167 .
in this section we discuss constructing solutions of eqs.(1.1 ) in some more detail .
namely , in sect.4.1 we consider a collection of group - valued smooth functions defining maps between smooth and locally constant trivializations of flat bundles .
these functions parametrize flat connections and satisfy first order differential equations .
the solution space of these equations is discussed in sect.4.1 . in the ech approach
flat bundles are described by locally constant transition functions . in sect.4.2
we discuss functional matrix equations on transition functions and the space of their solutions .
in sect.4.3 we describe the moduli space of flat connections as a double coset space and discuss the correspondence between the ech and de rham descriptions of flat connections . as before
, we fix a good covering @xmath71 of @xmath0 , trivializations of the bundle @xmath4 over @xmath176 s and locally constant transition functions @xmath30 on @xmath177 , @xmath178 , @xmath179 .
any solution of equations ( 1.1a ) on an open set @xmath176 ( a _ local _ solution ) has the form @xmath180 where @xmath181 is a smooth @xmath2-valued function on @xmath176 .
if we find solutions ( 4.1a ) for all @xmath182 , we obtain a collection @xmath165 .
if @xmath177 , then the solutions ( 4.1a ) on @xmath176 and @xmath183 are not independent and must satisfy the compatibility conditions @xmath184 on @xmath185 . from eqs.(4.1 )
the _ differential _ compatibility equations @xmath186 follow .
since we are looking for _ global _ solutions @xmath187 of eqs.(1.1a ) , eqs.(4.2 ) must be satisfied for any @xmath188 such that @xmath177 . any global solution of eqs.(1.1b ) has the form @xmath189 where @xmath190 satisfy eqs.(4.2 ) , and @xmath166 is an arbitrary ( global ) solution of eqs.(3.13b ) .
the compatibility conditions for @xmath191 and @xmath192 on @xmath177 have the form @xmath193 after substituting ( 4.3a ) into ( 4.3b ) , we obtain that @xmath194 where @xmath195 .
thus , one can easily construct solutions of eqs.(1.1b ) if one knows solutions of eqs.(1.1a ) or ( 4.2 ) .
that is why in the following we shall concentrate on describing solutions of eqs.(1.1a ) and ( 4.2 ) .
differential equations ( 4.2 ) are equations on 0-cochains @xmath196 .
we denote by @xmath197 the space of solutions @xmath190 of eqs.(4.2 ) , @xmath198 . recall that according to the definitions of sections 2 and 3 , the solution space of eqs.(1.1a ) is @xmath199 . restricting the map @xmath200 to @xmath197 , we obtain the map @xmath201 it is obvious from eqs.(4.1 ) that @xmath202 and @xmath203 with @xmath204 define the same solution @xmath205 of eqs.(1.1a ) , which leads to the bijection @xmath206 the gauge group @xmath107 acts on @xmath207 by the right multiplication : @xmath208 , @xmath209 , and the definition ( 3.4 ) of the moduli space of flat connections can be reformulated in terms of the set @xmath210 by using the bijection ( 4.5 ) , @xmath211 having a solution @xmath190 of eqs.(4.2 ) , one can introduce matrices @xmath212 defined on @xmath177 . using eqs.(4.2 )
, it is not difficult to show that @xmath213 , and therefore @xmath214 .
moreover , it is easy to see that @xmath215 s satisfy equations @xmath216 and therefore @xmath149 can be considered as transition functions of a locally constant bundle @xmath217 which is topologically equivalent to the bundle @xmath4 with the transition functions @xmath30 .
thus , we obtain a map of the data @xmath218 into the data @xmath219 , i.e. the connection @xmath5 is encoded into the transition functions @xmath152 of the locally constant bundle @xmath147 .
let us now consider `` free '' 1-cochains @xmath220 such that @xmath221 and _ functional _ equations @xmath222 on @xmath223 .
the space of solutions to eqs.(4.6a ) is isomorphic to the space @xmath224 of 1-cocycles .
denote by @xmath225 the subset of those solutions @xmath152 to eqs.(4.6a ) for which there exists a splitting @xmath226 with some @xmath165 . using equations @xmath213 ,
one can easily show that these @xmath36 will satisfy eqs.(4.2 ) , i.e. @xmath227 .
then , by introducing @xmath228 , we obtain a flat connection @xmath229 on @xmath4 .
thus , for constructing flat connections on @xmath4 , one can solve either _ differential _ equations ( 4.2 ) on @xmath230 or _ functional _ equations ( 4.6 ) on @xmath149 such that @xmath231 .
in sections 4.1 , 4.2 we showed that for the fixed transition functions @xmath27 of the locally constant bundle @xmath4 there is a correspondence @xmath233 where @xmath234 is the solution space of differential equations ( 1.1a ) , and @xmath235 is the solution space of functional equations ( 4.6 ) .
these solution spaces are not bijective .
namely , to gauge equivalent flat connections
@xmath5 and @xmath236 with @xmath237 there corresponds the same cocycle @xmath238 , since @xmath239 at the same time , the cocycles @xmath152 and @xmath240 with @xmath241 correspond to the same flat connection @xmath5 .
the correspondence ( 4.7 ) will become the bijection ( 3.11 ) if we consider the space of orbits of the group @xmath242 ( the gauge group of the model ( 1.1a ) ) in the space @xmath234 and consider the space of orbits of the group @xmath243 ( the gauge group of the model ( 4.6 ) ) in the space @xmath235 .
in other words , we should consider the pairs @xmath244)$ ] and @xmath245)$ ] , where @xmath134 $ ] is the gauge equivalence class of @xmath5 , and @xmath246 $ ] is the equivalence class of the 1-cocycle @xmath152 . in sect.3 this correspondence was described in more abstract terms .
we obtain the following correspondence between de rham and ech description of flat connections : @xmath247 \d^0\swarrow\quad\quad\searrow r \\[1ex ] z^0({{\mathfrak u}},{{\mathcal a}}_p)\quad \quad z^1_f({{\mathfrak u}},{\mathbb g}_p ) \\[1ex ] \pi\downarrow\qquad \qquad \downarrow p \\[1ex ] { { \mathcal m}}_p=\frac{z^0({{\mathfrak u}},{{\mathcal a}}_p ) } { z^0({{\mathfrak u}},{{\mathfrak s}}_p)}\longleftrightarrow \frac{z^1_f({{\mathfrak u}},{\mathbb g}_p)}{c^0({{\mathfrak u}},{\mathbb g}_p)}\qquad .
\quad \end{array } \tag{4.8}\ ] ] here a map @xmath248 is given by the formula @xmath249 , @xmath250 and @xmath251 are projections @xmath252 and a map @xmath200 is the projection ( 4.4 ) . recall that the group @xmath253 acts on the spaces @xmath254 and @xmath234 on the right , and the group @xmath243 acts on the spaces @xmath254 and @xmath235 on the left .
using a bijection of the moduli spaces of solutions to eqs.(4.2 ) , ( 1.1a ) and ( 4.6 ) , we identify the points @xmath255 and @xmath256 , where @xmath257 .
then we have @xmath258 where @xmath259 , @xmath260 , @xmath261 and @xmath262 are some ( local ) sections of the fibrations ( 4.4 ) , ( 4.9a ) , ( 4.9b ) and ( 4.9c ) , respectively .
let us emphasize that just the ambiguity of the choice of sections of the fibrations ( 4.4 ) and ( 4.9 ) leads to the ambiguity of finding @xmath152 for a given @xmath5 and @xmath5 for a given @xmath152 . as usual , one can remove this ambiguity by choosing some _ special _
sections @xmath259 , @xmath260 , @xmath261 and @xmath262 ( gauge fixing ) .
in sect.4 we have described the transformation @xmath263 .
if we now define a map @xmath264 preserving the set @xmath235 , then after using the maps @xmath265 or @xmath266 , we obtain a new flat connection @xmath267 .
below we describe this method of constructing flat connections on @xmath4 based on deformations @xmath264 of locally constant bundles .
namely , starting from the ech approach , in sect.5.1 we describe groups acting on the space of transition functions of a locally constant bundle . then , using solutions of the functional matrix equations , we proceed in sect.5.2 to the de rham description and define dressing transformations acting on the space of solutions to the field equations of chern - simons and topological bf theories .
finally , in sect.5.3 we discuss a special class of solutions of the functional matrix equations describing deformations of flat bundles . to begin with
, we shall consider groups acting on the set @xmath268 of 1-cochains .
one of such groups - the group @xmath243 - and its left action @xmath97 on @xmath268 , @xmath269 and @xmath235 have been considered in sections 2.2 and 4.3 .
the action @xmath97 of the group @xmath243 gives trivial ( gauge ) transformations of the transition functions @xmath270 , and the space of orbits of the group @xmath243 in @xmath235 is the moduli space @xmath271 .
in other words , the group @xmath243 is the stability subgroup of the point @xmath246\in{{\mathcal m}}_p$ ] . to obtain a nontrivial map of the moduli space @xmath119 onto itself , we consider the following action of the group @xmath268 on itself : @xmath272 where @xmath273 . recall that @xmath274 if @xmath275 . from the definition ( 5.1 )
it is easy to see that @xmath276 for @xmath277 .
moreover , starting from the element @xmath278 , one can obtain any other element of @xmath268 , i.e. the action @xmath279 is transitive .
we are interested in subgroups of the group @xmath268 preserving the set @xmath280 . to describe them ,
we fix @xmath281 and consider an orbit of the point @xmath282 under the action @xmath279 of the group @xmath268 .
we want to find an intersection of this orbit with the set @xmath235 , i.e. the set of all @xmath283 such that * @xmath284 , i.e. the transformation @xmath285 preserves eqs.(4.6a ) , * @xmath286 , i.e. the transformation @xmath287 preserves not only eqs.(4.6a ) , but also eqs.(4.6b )
. the condition ( i ) means that the transformation @xmath288 maps @xmath269 onto itself , which is realized , of course , not for any @xmath283 .
this condition imposes severe constraints on @xmath77 which are equivalent to the following nonlinear functional equations : @xmath289 notice that eqs.(5.2a ) are satisfied trivially if @xmath290 , since in this case @xmath291 , @xmath292 . by introducing @xmath293 and using formulae ( 2.5 )
, we can rewrite eqs.(5.2a ) in the form @xmath294 formulae ( 5.2 ) mean that @xmath295 is a 1-cocycle , @xmath296 .
the condition ( ii ) means that for the 1-cocycle @xmath297 one can find a splitting like ( 4.6b ) , i.e. there exists a 0-cochain @xmath298 such that @xmath299 this imposes additional restrictions on @xmath283 , and one of their possible resolvings is the following .
recall that for the sheaf @xmath300 of smooth sections of the bundle @xmath301 we have @xmath302 @xcite .
therefore for _ small _ deformations @xmath303 of the cocycle @xmath152 there always exists @xmath304 such that ( 5.3 ) is satisfied . as such , we shall consider elements @xmath283 that are close to the identity and satisfying eqs.(5.2 ) .
in other words , we consider a _ local group _ ( a neighbourhood of the identity in @xmath268 ) and take only those its elements @xmath77 which satisfy the conditions ( 5.2 ) .
if , in addition , we change the `` initial '' point @xmath305 for the action @xmath279 , then in the described way we shall obtain all elements of the group of bijections of the set @xmath235 . in sect.4
we have described maps @xmath306 and @xmath307 , where @xmath308 , @xmath309 , @xmath310 . in sect.5.1
we have described the map @xmath311 where @xmath312 if elements @xmath313 satisfy some conditions .
then , having found @xmath314 by formula ( 5.3 ) , we introduce @xmath315 from the equations @xmath316 and eqs.(5.3 ) it follows that @xmath317 s satisfy eqs.(4.2 ) , and @xmath318 s satisfy eqs.(4.1b ) , i.e. @xmath319 @xmath320 on @xmath177 .
thus , if we take a flat connection @xmath187 and carry out the sequence of transformations @xmath321 we obtain a new flat connection @xmath322 .
notice that by virtue of commutativity of the diagram ( 4.8 ) , one can consider the map @xmath323 instead of @xmath324 and the map @xmath265 instead of @xmath266 .
it is not difficult to verify that @xmath325 where @xmath326 formally , ( 5.7 ) looks like a gauge transformation .
but actually the transformation @xmath327 , defined by ( 5.7 ) , consists of the sequence ( 5.6 ) of transformations and is not a gauge transformation , since @xmath328 on @xmath329 .
recall that for gauge transformations @xmath330 one has @xmath331 , i.e. @xmath332 is a _
global _ section of the bundle @xmath19 , and @xmath333 is a collection of _ local _ sections @xmath334 of the bundle @xmath19 which are constructed by the algorithm described above . from formulae ( 3.13 ) , ( 4.3 ) and ( 5.7b )
it follows that the dressing transformation @xmath335 acts on any solution @xmath15 of eqs.(1.1b ) by the formula @xmath336 where @xmath337 is defined in ( 5.7b ) .
the field @xmath338 is a new solution of eqs.(1.1b ) .
as is shown above , the transformation @xmath339 is not a gauge transformation .
let us emphasize that @xmath333 depends on @xmath283 defining the transformation @xmath340 , and taking this into account we shall write @xmath341 .
the transformations ( 5.7 ) will be called the _ dressing transformations_. in this terminology we follow the papers @xcite , where the analogous transformations were used for constructing solutions of integrable equations .
it is not difficult to show that these transformations form a group .
notice that the maps ( 5.6 ) are connected with maps between the bundles @xmath342 , @xmath343 and @xmath344 .
all these bundles are diffeomorphic but not isomorphic as locally constant bundles .
a diffeomorphism of @xmath4 onto @xmath147 is defined by a 0-cochain @xmath345 , and a diffeomorphism of @xmath4 onto @xmath346 is defined by @xmath347 .
moreover , the bundles @xmath147 and @xmath346 become isomorphic as locally constant bundles after the restriction to @xmath177 : @xmath348 , but these isomorphisms are different for @xmath349 with different @xmath188 .
in other words , @xmath350 define _ local _ isomorphisms that do not extend up to the isomorphism of @xmath147 and @xmath346 as locally constant bundles over the whole @xmath0 .
matrices @xmath351 , defining the dressing transformations @xmath352 must satisfy the nonlinear functional equations ( 5.2 ) which are not so easy to solve . however , there exists an important class of solutions to these equations , which can be described explicitly .
the merit of these solutions is the fact that they do not depend on the choice of cocycles @xmath148 and @xmath152 .
notice that eqs.(5.2 ) with all indices written down have the form @xmath353 where @xmath354 means the restriction of @xmath355 defined on @xmath185 to an open set @xmath356 and analogously for all other matrices .
it is not difficult to verify that a collection @xmath357 of matrices such that @xmath358 satisfies eqs.(5.9 ) for any choice of 1-cocycles @xmath27 and @xmath220 .
the constraints ( 5.10 ) are not very severe .
they simply mean that sections @xmath355 of the sheaf @xmath34 over @xmath177 can be extended to sections of the sheaf @xmath34 over the open set @xmath359 where the summation is carried out in all @xmath188 for which @xmath177 .
in other words , from ( 5.10 ) it follows that there exists a locally constant section @xmath360 of the bundle @xmath19 such that @xmath361 @xcite . in this case
we can identify @xmath362 and @xmath363 .
such @xmath77 form a subgroup @xmath364 of the group @xmath268 .
now let us consider @xmath365 as a _
local _ group , i.e. let us take @xmath77 from the neighbourhood of the identity in @xmath365
. for such @xmath77 there always exists @xmath366 such that ( 5.3 ) is satisfied and one can introduce @xmath367 by formula ( 5.7b ) .
notice that this map @xmath368 is a homomorphism .
using @xmath367 , we can introduce a new flat connection @xmath369 and a new solution @xmath370 of eqs.(1.1b ) for any @xmath371 .
in this paper , the group - theoretic analysis of the moduli space of solutions of chern - simons and topological bf theories and nonlocal symmetries of their equations of motion has been undertaken .
we have formulated a method of constructing flat connections based on a correspondence between the non - abelian de rham and ech cohomologies .
the group of all symmetries of the field equations of chern - simons and topological bf theories and the special cohomological symmetry group have been described .
the described dressing symmetries can be lifted up to symmetries of quantum chern - simons and topological bf theories .
it would be desirable to describe and analyze representations of these symmetry groups .
oniik a.l .
, on classification of fibre bundles , _ dokl .
nauk sssr _ , 1961 ,
v.141 , 803806 ; connections without curvature and de rham theorem , _ dokl .
nauk sssr _ , 1964 ,
v.159 , 988991 ; some notions and applications of non - abelian cohomology theory , _
trudi mosk .
obsch . _ , 1967 ,
v.17 , 4588 , in russian . | we describe constructing solutions of the field equations of chern - simons and topological bf theories in terms of deformation theory of locally constant ( flat ) bundles
. maps of flat connections into one another ( dressing transformations ) are considered .
a method of calculating ( nonlocal ) dressing symmetries in chern - simons and topological bf theories is formulated .
journal of nonlinear mathematical physics 2000 , v.7 , n 4 , . [ firstpage ] |
multi - component quantum condensates with novel types of inter - component interactions are of considerable interest in contemporary physics .
for example , they are relevant to widely disparate systems , including low - dimensional spin-@xmath5 quantum antiferromagnets @xcite , bose - einstein condensates @xcite , multi - component / multi - band superconductors @xcite , and non - abelian quantum hall states and topological superconductors @xcite .
these systems have the remarkable property of possessing a mapping to a classical multi - component plasma system with highly unusual intra- and inter - component interactions .
the statistical properties of these unconventional plasmas especially their phase diagrams have important ramifications for the physics of their corresponding fractional quantum hall systems @xcite .
the statistical physics of such systems has only recently begun to be explored . in a previous paper , we investigated the metal - insulator transition in a particular version of such a plasma @xcite . in this paper , we will extend these investigations to a study of the freezing of such a plasma from a liquid to a wigner crystal . the canonical partition function of the unconventional two - component plasma that we investigate is given by @xcite @xmath6 where the potential energy @xmath7 describes two species ( components ) of particles interacting via two different types of two - dimensional ( 2d ) coulomb interactions , which are logarithmic . here
, the @xmath8 are coordinate vectors for the @xmath9 particles of component @xmath10 , which carry charge @xmath11 of the first interaction ( type 1 ) and charge @xmath12 of the second interaction ( type 2 ) .
the @xmath13 are coordinate vectors for the @xmath9 particles of component @xmath14 , which carry no charge of type 1 and charge @xmath15 of type 2 .
the term @xmath16 describes the interaction of the @xmath10-particles with a uniform density neutralizing background charge .
note that the form given in eq .
implies that the temperature @xmath17 .
this plasma is related @xcite to inner products of quantum - mechanical trial wave functions of ising - type quantum hall states , such as the moore - read pfaffian @xcite , anti - pfaffian @xcite , and bonderson - slingerland hierarchy states @xcite . for the charge values relevant to these states ,
the plasma was shown to be in its metallic liquid phase @xcite , which allows for the calculation of the braiding statistics of quasiparticle excitations of these states @xcite , confirming their conjectured non - abelian statistics .
this plasma is also related to rotating two - component bose - einstein condensates ( becs ) in two dimensions @xcite . while ref .
focused on the cases @xmath18 , which are particularly relevant for ising - type quantum hall states , here we will investigate the plasma for large values of @xmath11 . in the limit in which @xmath19 , the @xmath14-particles do not interact , and the plasma thus reduces to the standard 2d one - component coulomb plasma ( ocp ) .
it is generally believed that , at high values of @xmath11 , the ocp will be in a 2d solid state in which the charges form a triangular lattice with quasi - long - range translational and long - range orientational order @xcite , as found in the simulations in refs . .
however , some studies have claimed that there is no low - temperature ( high-@xmath11 ) crystalline state in the ocp , due to the proliferation of screened disclinations @xcite .
if we assume that the generally - held view is correct ( and we present evidence supporting this view ) , so that there is a low - temperature crystalline state , then the melting of this crystal can occur according to either of two possible scenarios .
one possibility is the kosterlitz - thouless - halperin - nelson - young theory ( kthny ) @xcite , according to which dislocation pairs unbind via a vector - defects version of the bkt transition at a berezinskii - kosterlitz - thouless ( bkt ) like transition .
the system then enters a hexatic liquid phase in which there is no translational order , but there is quasi - long - range hexatic order .
then , there is a second bkt transition at which disclination pairs unbind , hexatic order is lost , and the system enters an isotropic liquid phase .
the other possibility is a direct first - order melting transition at a lower temperature than the kthny - theory predicts @xcite .
there have been considerable efforts to investigate 2d melting , both experimentally and by numerical simulations .
some studies have found kthny transitions while others have found a weakly first - order melting transition @xcite .
it appears that the nature of 2d melting depends on details of the interatomic potential . in the case of logarithmic interactions ,
most numerical simulations find a first - order transition @xcite . before proceeding to a description of our simulations ,
we mention that , in principle , there is one other possibility : a lifshitz transition from the liquid to a striped or `` microemulsion '' phase and then later to a wigner crystal , as discussed by kivelson and spivak @xcite .
such a scenario must be considered when there is a linear coupling between the order parameter and the uniform density ( i.e. , between the order parameter at wavevector @xmath20 and the density at wavevector @xmath21 ) or , equivalently , when the first derivative of the energy with respect to the density is discontinuous at the transition . however , in our case , the order parameter is the density at non - zero wavevector , so no such linear coupling can occur .
furthermore , the order parameter vanishes on both sides of the transition since the crystalline phase is only quasi - long - range ordered , so there would be no discontinuity even if there were a linear coupling . however , even in systems to which the kivelson - spivak @xcite argument applies , there are two possible scenarios , similar to the ones that we consider : a direct first - order phase transition ( which is permitted for the case of logarithmic interactions ) and a continuous transition via one or more intermediate phases .
the system described in eqs . and is studied by means of large - scale monte carlo simulations on a sphere of radius @xmath22 . in this geometry ,
the distance between two points @xmath23 and @xmath24 is taken to be the chord length @xmath25 and the term @xmath16 is simply a uniform constant that can be disregarded .
hence , the model in eq . may be written in the form ( up to constant terms ) @xcite @xmath26 .
\label{eq : model_s_sphere}\end{aligned}\ ] ] here , @xmath27 , @xmath28 are the positions of the particles on the surface of the unit sphere .
details of the derivation , as well as on the technicalities of the monte carlo simulations , are presented in ref . .
moreover , to improve sampling at high values of @xmath11 , we used the parallel tempering algorithm @xcite , where the set of couplings was found by measuring first - passage - times as described in ref . .
in addition to the logarithmic interactions , we regularize the attractive interactions by adding a short - range hard - core repulsion such that particles are not permitted to be closer than the particle diameter @xmath29 .
hence , there is a nonzero dimensionless density @xmath30 where @xmath31 and @xmath32 is the area of the system .
first , we consider the case in which @xmath19 . this is motivated by the fact that previous studies of the ocp on the surface of a sphere are not consistent . in ref . , a freezing transition at @xmath2 was found by comparing the free energy of the solid and liquid state .
however , in ref . , the absence of a finite - temperature crystalline state was claimed and numerical evidence supporting this was provided , essentially by showing that the correlation length for crystalline order was non - divergent , @xmath33 for all @xmath34 .
we will return to this below .
the structure function is given by @xmath35 here , @xmath36 is the density - density function , where @xmath37 is the local density , and @xmath38 denotes a statistical average .
furthermore , @xmath39 denotes the positions of the point particles in the problem and @xmath40 is the fourier space vector . in this work
, we measure the azimuthal average of @xmath41 modified for a spherical geometry , given by @xcite @xmath42 \sin\theta j_0(qr\theta ) , \label{eq : structure_factor}\ ] ] where @xmath43 is the number density , @xmath22 is the radius of the sphere on which the particles live , @xmath44 is the pair distribution function with @xmath45 as the chord angle , @xmath46 is a zeroth order bessel function , and @xmath47 is the magnitude of @xmath40 . the inset of fig .
[ fig : correlation_length ] shows a plot of @xmath48 .
we assume that the correlation length @xmath49 is inversely proportional to the width of the first peak in @xmath48 , and may thus be determined by a lorentzian fit .
the procedure is identical to that used in ref . , and
the result is given in fig .
[ fig : correlation_length ] . for small values of @xmath11
our results are similar to fig . 2 in ref .
however , when @xmath50 , a value that corresponds well with the critical coupling of the freezing transition , we find a kink developing with increasing @xmath9 , that clearly violates @xmath33 . this kink is not seen in fig . 2 of ref .
however , we note that the markers of that figure exhibits large scattering .
moreover , the authors did not consider larger values of @xmath11 . as a function of @xmath11 for 8 different system sizes in the range @xmath51 .
the inset is a plot of @xmath48 for the specific case when @xmath52 and @xmath53 . ]
a hallmark of a 2d solid is that translational correlations have a power - law decay , @xmath54 , where @xmath55 is the reciprocal lattice vector , @xmath56 , @xmath57 are lattice points in the 2d solid and @xmath58 is a temperature dependent exponent @xcite .
consequently , the first - order bragg peak in @xmath59 will , as seen by eq . , scale as @xmath60 , where @xmath61 is the spatial linear extent of the system .
now , by integrating over the bragg peak of a 2d solid @xcite , the finite - size scaling of the azimuthally averaged first peak in @xmath48 , is given by @xmath62 fig . [
fig : finite_size_behavior ] shows the results for the maximum value of the first peak in @xmath48 for a wide range of system sizes and for different values of @xmath63 . as for @xmath49 , we find that the peak value also exhibits a kink at @xmath64 that should be associated with an abrupt change in the translational correlations in the plasma . indeed , when studying the finite - size behavior more closely in the lower panel of fig . [
fig : finite_size_behavior ] , the results show that when @xmath65 , @xmath66 when @xmath9 increases .
this is the behavior expected in the liquid phase , with exponentially decaying translational correlations where @xmath67 .
however , when @xmath68 , the results clearly show that there is a positive slope that develops with increasing @xmath9 , thus confirming the finite - size behavior of the 2d solid given in eq . .
when @xmath69 , it is difficult to determine whether the system is in the solid phase or not , suggesting that @xmath69 is close to the melting point of the ocp . note that in fig .
[ fig : finite_size_behavior ] , the height of the first - order peak in @xmath48 , @xmath70 when @xmath69 .
this is consistent with the 2d freezing criterion for a crystal with long - range interactions ( characterized by a divergent bulk modulus @xcite)@xcite .
, as a function of coupling @xmath63 and size @xmath9 .
the upper panel shows @xmath71 as a function of @xmath63 for eight different sizes in the range @xmath72 .
the lower panel is a log - log plot of @xmath71 as a function of @xmath9 for seven fixed values of @xmath63 .
the dashed line is a reference line that yields the expected finite - size behavior at the melting point according to kthny theory .
lines are guide to the eyes . ]
a key prediction of the kthny theory is that @xmath73 in the solid phase , where the limiting value of @xmath74 is reached at the critical point of melting from a triangular lattice to the hexatic phase @xcite . as
a result , in this scenario , @xmath71 grows more rapidly with @xmath9 than @xmath75 for all @xmath63 greater than the critical value ; @xmath71 grows as @xmath75 at the transition point ; and @xmath71 saturates in the liquid phase .
meanwhile , if the transition were first - order , the limiting value of @xmath58 would be smaller than @xmath74 , so that @xmath71 would grow more rapidly than @xmath75 at the transition point , i.e. , the slowest possible growth of @xmath71 in the crystalline phase would be faster than @xmath75 .
consequently , we expect the slope of @xmath76 vs. @xmath77 to be steeper than @xmath74 for all @xmath63 in the crystalline phase or , by the results above , for all @xmath78 . by determining the slope of @xmath76 vs. @xmath77 at @xmath79 , we could then determine if the transition is of kthny type or is first - order .
however , as may be seen in fig .
[ fig : finite_size_behavior ] , the slopes of @xmath76 vs. @xmath77 in the putative crystalline phase are _ not _ steeper than @xmath74 in our simulations . however , the slopes steepen with increasing @xmath9 , possibly converging towards the expected behavior in the thermodynamic limit .
therefore , we are unable to determine which type of transition occurs , nor whether melting proceeds via an intermediate hexatic phase ( see below ) .
it is worth emphasizing that for @xmath80 , @xmath71 appears to saturate to a finite value , as expected in a liquid , while , for @xmath78 , @xmath71 does not appear to saturate , as expected in a crystal ( although , as noted above , it does not grow as rapidly as expected ) .
therefore , the lower panel of fig .
[ fig : finite_size_behavior ] is also qualitatively consistent with a crystalline phase of the 2d ocp , which melts at @xmath79 . taken together with fig .
[ fig : correlation_length ] and the upper panel of fig .
[ fig : finite_size_behavior ] , this provides clear evidence for the existence of a low - temperature crystalline state of the ocp on a sphere , in agreement with previous studies @xcite .
our results contradict the claims made in ref . for the non - existence of a crystalline phase .
the present work considers larger values of @xmath11 and provides considerably better statistics than ref . .
the transition from the hexatic to the isotropic liquid phase is governed by fluctuations in the orientational order parameter .
these can be computed by the bond - orientational susceptibility which exhibits a peak at this transition .
when this peak is obtained at a different point than the onset of quasi - long - range translational order , it indicates the existence of an intermediate hexatic phase in between the crystalline and isotropic liquid phase @xcite . on a sphere ,
we compute the bond - orientational susceptibility @xmath81 here , we have @xmath82 furthermore , the quantity @xmath83 may be thought of as the local bond - orientational order parameter of particle @xmath84 , where all bond angles @xmath85 are measured with respect to the closest nearest neighbor in the tangential plane of particle @xmath84 .
the sum is over all @xmath86 nearest neighbors as determined by voronoi construction . in eq . , the angle @xmath87 is the bond angle of the bond to the closest nearest neighbor of particle @xmath84 , measured with respect to the @xmath88 chord in the tangential plane of particle @xmath84 .
thus , the chord line combining the two particles for every term in the sum of eq . , serves as a line of reference for bond - orientational order@xcite . in fig .
[ fig : bond - orientational_suscetibility ] , the results for @xmath89 are given .
when the system size is large ( @xmath90 ) , a peak is found .
however , the value of @xmath11 at which the peak occurs , appears to converge towards @xmath64 .
thus , with the resolution available , we can not confirm the existence of a hexatic phase since the position of the peak in @xmath89 does not appear to converge to a coupling significantly different from @xmath64 .
our findings are consistent with earlier works , which also have found no traces of an intermediate hexatic phase in these systems @xcite . as a function of @xmath11 for 7 different system sizes in the range @xmath91 .
the position of the peak appears to converge to @xmath64 . ]
we now turn our attention to the full model in eq . , i.e. , when both @xmath11 and @xmath12 are nonzero . in particular , we consider how the translational ordering of both @xmath14 and @xmath10 particles is affected as we increase the coupling constant in the second interaction - channel , @xmath12 . as for the ocp , we study translational correlations by measuring the structure factors @xmath92 , @xmath93 defined by eq . with @xmath48 , @xmath94 , @xmath95 .
in addition , we also measure the inverse dielectric constant for charges with interaction of type 2 , given by @xmath96 this quantity measures the screening properties for charges interacting with @xmath12 , and it signals a charge - unbinding transition involving @xmath10- and @xmath14-particles @xcite . in fig .
[ fig : combined_panels ] , results are given for the height of the first - order peak in the structure factor for component @xmath10 and @xmath14 , for the case when @xmath97 . apart from the fact that the height of the peak in the structure factor is much larger for the @xmath10 particles than the @xmath14-particles , the size- and @xmath63-dependence of the peaks are qualitatively very similar for the two components
. in particular , they both exhibit a kink at @xmath64 , which should be associated with melting of a 2d solid , similar to the ocp case in the upper panel of fig .
[ fig : finite_size_behavior ] . specifically , when we extract the finite - size behavior in the log - log plots in fig .
[ fig : structure_factor_both_comp ] , we find that both components exhibit @xmath66 , consistent with being in the liquid phase , when @xmath65 . when @xmath68 , the results clearly show that there is a power - law dependence on @xmath9 , consistent with the finite - size behavior of a 2d solid .
these results are consistent with the phase diagram in fig .
[ fig : plot_phase_diagram ] . and @xmath98 .
panel ( a ) shows the height of the first - order peak in the structure factor for the @xmath10 particles , @xmath99 , as a function of the coupling @xmath63 for seven different sizes in the range @xmath100 .
panel ( b ) shows the height of the first - order peak in the structure factor for the @xmath14 particles , @xmath101 , as a function of the coupling @xmath63 for seven different sizes in the range @xmath100 . in order to give an impression on how a typical structure factor looks like , the insets of panel ( a ) and ( b ) show plots of @xmath102 and @xmath103 for the specific case when @xmath104 , @xmath105 and @xmath97 . ] .
both panels show the height of the first - order peak in the structure factor as a function of size @xmath9 . in the lower panel , @xmath106 is the height of the peak of @xmath101 when we have subtracted the regular part in order to properly extract the singular finite - size behavior of @xmath92 in a log - log plot .
the solid lines are reference lines that yield the expected finite - size behavior at the melting point according to kthny theory .
lines are guide to the eyes . ] and
the dashed red line is the charge unbinding transition , at which @xmath10 and @xmath14 particles become bound together for at @xmath107 above this line .
the dotted blue line is the melting line of the wigner crystal . in phases
i and ii , the @xmath10 particles are in a liquid state ; in phases iii and iv , the @xmath10 particles form a wigner crystal . in phases
i and iii , the @xmath14 and @xmath10 particles are unbound ; in phases ii and iv , they are bound into molecules comprised of one @xmath10 and one @xmath14 particle .
see the text for details . ]
the inverse dielectric constant @xmath108 is measured to be zero to the left of the red line in fig .
[ fig : plot_phase_diagram ] .
thus , the @xmath14 and @xmath10 particles are in a metallic state regardless of the change in the structural properties when @xmath64 .
this is most salient with respect to the second type of coulomb interaction ( which has an effective strength that is determined by @xmath108 ) . in the liquid phase ,
it is clear that @xmath14 and @xmath10 particles are in a metallic state . in the crystalline phase ,
there are interstitials and vacancies in the crystal , so that a finite fraction of @xmath14 and @xmath10 particles should be considered as unbound particles that are able to screen test particles interacting with charges of type 2 , thereby leading to @xmath109 . at larger values of @xmath107 , there is a transition at which @xmath14 and @xmath10 particles are bound into molecules . for @xmath107 above this transition point , which is at @xmath110 , @xmath108 has a non - zero value , as found in ref . .
although the @xmath14 particles are able to screen the type 2 interaction when @xmath10 particles form a wigner crystal , their translational correlations exhibit signatures of a 2d solid ( as seen in fig [ fig : combined_panels ] ) , attributed to a higher probability of the @xmath14 particles to be co - centered with @xmath10 particles due to the attractive inter - component interactions of type 2 . on average , a finite fraction of the @xmath14 particles should be considered as bound to the @xmath10 particles , thus adapting to the 2d crystalline structure that is created by the strong repulsive interactions among the @xmath10 particles , when @xmath111 .
the signatures of freezing of the @xmath14-particles is thus an effect which is induced by the freezing of the @xmath10-particles .
hence , we can summarize the situation as follows , as depicted in fig .
[ fig : plot_phase_diagram ] . in phase
i , the @xmath14 and @xmath10 particles are unbound and are separately in a liquid state . in phase ii ,
the @xmath14 and @xmath10 particles are bound into molecules that are neutral with respect to the second type of coulomb interaction , and these molecules form a liquid . in phase iii ,
the @xmath14 and @xmath10 particles are unbound ; the @xmath10 particles form a wigner crystal while the @xmath14 particles form a liquid , albeit one with modulated density due to its interaction with the wigner crystal . in phase iv
, the @xmath14 and @xmath10 particles are bound into molecules forming a wigner crystal . in appendix
[ sec : transitions ] , we explain the details of how the transition lines were obtained . we consider phase iii to be a 2d counterpart of the situation that was reported for a three - dimensional system in ref . .
this work considered a two - component rotating bec with a negative dissipationless andreev - bashkin drag @xcite .
it was found that in this mixture , a situation may arise where the component with the smallest stiffness will be a _ modulated vortex liquid_. that is , the soft component breaks translational symmetry while exhibiting an unbroken symmetry in order parameter space .
the vortices of the soft component are likely to be co - centered with the vortices of the stiffest component , and will thus adapt to the spatial structure of the latter . as shown in ref .
, the mr plasma corresponds to a 2d two - component rotating bec with negative drag , where the @xmath10 component is stiffer than the @xmath14 component when @xmath112 .
in summary , we have considered the melting of an unconventional 2d two - component plasma on a sphere with particles interacting in two different channels , which may be viewed as an analogous plasma describing a non - abelian ising - type quantum hall state or a realization of a two - component two - dimensional bose - einstein condensate with inter - component non - dissipative drag . in the limiting case where there are no interactions of type 2 ( @xmath4 ) ,
the system is a standard 2d one - component plasma .
both for the one - component plasma and the unconventional two - component plasma , we find that the system freezes on a sphere for large enough inter - particle interactions .
for the two - component plasma , the @xmath14-particles do not have strong intra - component interactions , but still show signatures of forming a 2d solid .
this is attributed to the attractive inter - component interactions with the @xmath10-particles that leads to a higher probability of the @xmath14-particles to be co - centered with @xmath10-particles .
the @xmath14-particles nonetheless form a metallic state .
we have also examined the possible existence of an intermediate hexatic phase in the one - component plasma .
our results show that the value of @xmath11 where diclinations unbind and orientational order is lost , can not be distinguished from the value of @xmath11 where dislocations unbind and quasi - long - range translational order is lost .
we acknowledge useful discussions with j. s. hye , and i. b. sperstad .
e. b. , p. b. , c. n. and a. s. thank the aspen center for physics for hospitality and support under the nsf grant no .
e. v. h. thanks ntnu for financial support .
e. b. was supported by knut and alice wallenberg foundation through the royal swedish academy of sciences fellowship , swedish research council and by the national science foundation career award no .
v. g. was supported by nsf grant no . phy-0904017 . c. n. was supported in part by the darpa quest program .
l. r. acknowledges support by the nsf through dmr-1001240 and mrsec dmr-0820579 .
a. s. was supported by the norwegian research council under grants no .
205591/v30 and 216700/f20 .
the work was also supported through the norwegian consortium for high - performance computing ( notur ) .
we now discuss the determination of the phase transition lines in fig .
[ fig : plot_phase_diagram ] in more detail .
first , consider the line ( red in fig .
[ fig : plot_phase_diagram ] ) at which the @xmath10 and @xmath14 particles unbind . in this work
, we find the critical point of the bkt transition by curve - fitting the inverse dielectric constant to a logarithmic finite - size scaling relation with one free parameter ( see appendix c in ref . ) .
this means that we assume that the transition is a bkt transition as we use the bkt value of the universal jump in the finite - size scaling relation .
thus , it is a slightly less self - consistent approach than what was used in ref . , but still , one can regard this as a verification of the bkt nature , as one should not expect a good fit to the scaling relation if the transition is of a different nature .
the curve - fitting was performed according to the description in appendix c in ref . for sizes
@xmath114 , @xmath115 , @xmath116 , @xmath117 , @xmath118 , and @xmath119 , for @xmath120 , @xmath115 , and @xmath121 and for densities @xmath122 , @xmath123 , and @xmath124 . in fig . [ fig : plot_qq_sfa_eta ]
, the results for the transition point @xmath125 as a function of @xmath126 and @xmath63 are given .
we have also included the results for @xmath127 and @xmath128 from ref . .
in order to obtain a crude estimate of the transition temperature in the low density limit , we extrapolate to @xmath129 by fitting the results for finite @xmath126 to a power law @xmath130 , where @xmath125 , @xmath131 , and @xmath132 are free parameters .
the estimates we find are : @xmath133 these values are plotted in the phase diagram in fig .
[ fig : plot_phase_diagram ] , and we take the phase boundary to be the best fit straight line running through them . we now consider the wigner crystal melting transition , depicted by the blue line in fig .
[ fig : plot_phase_diagram ] .
this transition is found by measuring the value of @xmath63 at which @xmath99 attains its maximum second - derivative .
( see fig .
[ fig : finite_size_behavior ] and fig .
[ fig : combined_panels ] in the paper for example . ) in fig .
[ fig : plot_melting ] , we show the estimates of the transition point @xmath134 as a function of inverse system size @xmath135 for @xmath127 , @xmath136 , @xmath137 , and @xmath138 .
the transition points are estimated by averaging the results for @xmath90 , with errors determined by a bootstrap analysis .
the estimates we find are : @xmath139 the phase diagram in fig .
[ fig : plot_phase_diagram ] is obtained by using these values , and we take the phase boundary to be the best fit straight line running through them . | we study an unconventional two - dimensional , two - component classical plasma on a sphere , with emphasis on detecting signatures of melting transitions .
this system is relevant to ising - type quantum hall states , and is unconventional in the sense that it features particles interacting via two different two - dimensional coulomb interactions .
one species of particles in the plasma carries charge of both types @xmath0 , while the other species carries only charge of the second type @xmath1 .
we find signatures of a freezing transition at @xmath2 .
this means that the species with charge of both types will form a wigner crystal , whereas the species with charge of the second type also shows signatures of being a wigner crystal , due to the attractive inter - component interaction of the second type .
moreover , there is also a berezinskii - kosterlitz - thouless phase transition at @xmath3 , at which the two species of particles bind to form molecules that are neutral with respect to the second coulomb interaction .
these two transitions appear to be independent of each other , giving a rectangular phase diagram . as a special case , @xmath4 describes the ( conventional ) two - dimensional one - component plasma .
our study is consistent with previous studies of this plasma , and sheds new light on the freezing transition of this system . |
hi 21 cm aperture synthesis observations of nearby spiral galaxies is a mature field with over three decades of history
probably something of the order of a thousand galaxies have already been imaged .
however observers have tended to focus on bright ( @xmath3 l@xmath6 ) galaxies with hi masses @xmath7 m@xmath2 .
hi observations of faint dwarf galaxies ( m@xmath8 ) generally require comparatively long integration times , and such galaxies have hence not been studied in similar numbers .
while there have been some systematic hi surveys of dwarf galaxies ( @xcite ) , these have generally been restricted to the brighter ( m@xmath9 ) dwarfs . in hierarchical models of galaxy formation , nearby dwarf galaxies would , in some ways , be analogs of the primordial building blocks of large galaxies .
a systematic hi survey of the faintest dwarf galaxies could provide data that would be useful for a diverse range of studies , ranging from , for example , testing the predictions of cold dark matter models ( e.g @xcite ) , checking if such systems could be the host population of quasar absorption line systems ( e.g. @xcite ) etc . as the most chemically unevolved systems in the present - day galaxy population , the faintest dwarfs provide unique laboratories for understanding star formation and galaxy evolution in extreme environments , i.e. low metallicity , low dust content , low pressure , low shear , and low escape velocity ( e.g. @xcite ) . in this paper
we describe and present the first results from a giant meterwave radio telescope ( gmrt ) based hi imaging study of faint dwarf galaxies @xmath10 the faint irregular galaxies gmrt survey ( figgs ) .
the primary goal of figgs is to obtain high quality observations of the atomic ism for a large , systematically selected sample of faint , gas rich , dwarf irregular ( dirr ) galaxies .
our gmrt hi images are supplemented by single dish hi observations , hst v and i band images and ground based h@xmath11 images from the 6-m bta telescope .
additionally , the hii region abundances and h@xmath11 rotation curves are being obtained on the william herschel telescope ( wht ) , isaac newton telescope ( int ) telescopes on la palma and 6-m russian bta telescope , respectively .
this paper is organised as follows . in section [ sec : sample ]
we describe the design and the properties of the galaxy sample .
the main science drivers for figgs are described in section.[sec : science ] .
the gmrt observations are described in section [ sec : obs ] and the results of the survey are presented and discussed in section [ sec : result ] .
[ tab : figgs_sample ] [ cols="^,^,^,^,^,^,^,^,^,^,^ " , ] references : 1-@xcite 2-@xcite 3-@xcite 4-@xcite 5-@xcite 6-@xcite 7-@xcite 8-@xcite 9-@xcite 10-@xcite 11-@xcite 12-@xcite 13-@xcite 14- chengalur et al .
2008 ( in preparation ) the global hi profiles for our sample galaxies , obtained from the coarsest resolution data cubes ( see table [ tab : obs ] ) are shown in figure [ fig : spec ] .
the parameters derived from the global hi profiles for the whole figgs sample are listed in table [ tab : hiresult ] .
the columns are as follows : ( 1 ) the galaxy name , ( 2 ) the integrated hi flux along with the errorbars , ( 3 ) the central heliocentric velocity ( v@xmath12 ) , ( 4 ) the velocity width at 50% of the peak ( @xmath13 ) , ( 5 ) the hi diameter ( in arcmin ) at a column density of @xmath14 atoms @xmath15 ( d@xmath16 ) , ( 6 ) the derived hi mass ( m@xmath16 ) , ( 7 ) the hi mass - to - light ratio ( m@xmath17 ) , ( 8) the ratio of the hi diameter to the holmberg diameter . ( 9 ) the ratio of the gmrt flux to the single dish flux ( fi / fi@xmath18 ) , ( 10 ) the inclination as measured from the hi moment 0 maps ( i@xmath16 ) , and ( 11 ) the reference for the single dish fluxes . as seen in column(9 ) in table [ tab : hiresult ] , the hi flux measured from the gmrt hi profiles for most figgs galaxies , in general , agree ( within the errorbars ) with the values obtained from the single dish observations .
the average ratio of gmrt flux to single dish flux is @xmath19 .
this indicates that in general no flux was missed because of the missing short spacings in our interferometric observations .
however , for some galaxies the integrated flux derived from the gmrt observations is significantly smaller than the single dish values .
the gmrt fluxes could be lower than those obtained from single dish measurements either because of ( i ) a calibration error or ( ii ) a large fraction of hi being in an extended distribution that is resolved out , or ( iii ) the single dish flux is erroneous , possibly because of confusion with galactic emission .
however , the flux of the point sources seen in the gmrt images are in good agreement with those listed in nvss , indicating that our calibration is not at fault .
we note that in the case of kkh 12 , kkh 6 , ugc 6456 , ugca 92 and kk 41 there is a strong local hi emission at velocities very close to the systemic velocities , making it likely that the single dish integrated flux measurements for these galaxies were contaminated by blending of their hi emission with that of the galactic emission . in the case of kk 246 , its hi spectrum was near the edge of the gmrt observing band
, hence the flux could not be reliably estimated .
the gmrt integrated hi emission of the sample galaxies , obtained from the coarsest resolution data cubes ( see table [ tab : obs ] ) , overlayed on the optical digitized sky survey ( dss ) images are shown in figure [ fig : mom0 ] . the hi morphological inclinations ( i@xmath16 ) for our sample galaxies were estimated from the integrated hi maps by fitting elliptical annuli to the hi images at various resolutions . for sample galaxies which have hi disks less extended than 2 synthesised beams ( across the diameter of the galaxy ) at the lowest hi resolution ,
could in principle be derived from the higher resolution hi maps . however , for most sample galaxies ellipse fitting to the high resolution hi maps is not reliable because of clumpiness in the central high column density regions . the derived inclination ( without applying any correction for the intrinsic thickness of the hi disk )
is given in column(10 ) in table [ tab : hiresult ] .
figure [ fig : incl_figgs ] shows a comparison between the morphological inclination derived from the optical and hi isophotes of the galaxy .
no correction has been applied for the intrinsic thickness of the disk in both cases .
the solid line shows a case when both inclinations are the same .
we find that for 6 galaxies the hi inclination is significantly greater than the optical inclination ( viz .
kkh 11 , ugc 6456 , ngc 3741 , ugc 8055 , kk 230 , kk 250 ) . on the other hand , for many galaxies the optical inclination
is found to be systematically higher than the inclination derived from the hi morphology .
this result , if interpreted literally , suggest that the hi disks of these galaxies are thicker than the disks of their optical counterparts .
however , we caution that a proper analysis using deconvolved angular sizes of the the hi disks needs to be done before a firm conclusion can be drawn .
the diameter of the hi disk at a column density of @xmath20 atoms @xmath15 ( except for ugca 92 where the hi diameter is measured at @xmath21 atoms @xmath15 ) estimated from the lowest resolution integrated hi emission maps is given in column(7 ) of table [ tab : hiresult ] .
the ratio of the hi diameter to the optical ( holmberg ) diameter for the sample is also given in column(9 ) of the same table .
figure [ fig : hist_size ] shows the histogram of the derived hi extent of figgs at n@xmath22@xmath15 , normalised to the holmberg diameter of the galaxy .
the median hi extent of the figgs sample ( normalised to holmberg diameter of the galaxy ) is 2.4 . for a comparison , hunter ( 1997 ) using the data compiled from the literature for comparatively bright i
m galaxies found that the ratio of d@xmath23 is somewhat smaller , viz .
the extreme outliers in figure [ fig : hist_size ] is ngc 3741 , our figgs data show it to have an hi extent of @xmath3 8.3 times d@xmath24 ( holmberg diameter ) .
follow - up wsrt+drao+gmrt observations resulted in hi being detected out to @xmath38.8 d@xmath24 @xmath10 ngc 3741 has the most extended hi disk known .
for ngc 3741 the rotation curve could be derived out a record of @xmath344 times the disk scale length and from the last measured point of the rotation curve we estimate the dynamical mass to light ratio , m@xmath25 @xmath10 which makes it one of the `` darkest '' irregular galaxies known ( begum et al .
2005 , 2008 ) .
figure [ fig : m_hi_size ] shows a tight correlation between hi mass and the hi diameter , measured at @xmath26 @xmath15 for figgs sample .
the galaxies in figgs sample with accurate distances are shown as solid points , whereas the remaining galaxies are shown as open circle .
the best fit to the whole figgs sample shown as a solid line gives @xmath27 -0.75 cm the best fit relation was also derived using only the galaxies with trgb distances , however no significant difference was found between the best fit parameters derived in this case and that derived using the whole sample .
eqn.([eqn : mhi_dhi ] ) implies that the hi disks of the figgs sample are well described as having an constant average surface mass density @xmath28 m@xmath2 pc@xmath29 .
a tight correlation between hi mass and the size of the hi disk has been noted earlier for spiral galaxies ( e.g. @xcite ) and for brighter dwarf galaxies ( swaters 1999 ) . for these samples the hi diameter was measured at a slightly higher column density , viz . 1 m@xmath2 pc@xmath29 . for the figgs galaxies , the relationship between the hi mass and the hi diameter measured at 1 m@xmath2 pc@xmath29 is @xmath30 , for comparison , @xcite measure @xmath31 .
the fit coefficients overlap within the error bars . hence from the figgs data we find that there is at best marginal evidence for a decrease in average hi surface density with decreasing hi mass ; to a good approximation
, the disks of gas rich galaxies , ranging over 3 orders of magnitudes in hi mass , can be described as being drawn from a family with constant hi surface density .
the hi mass also correlates with the optical ( holmberg ) diameter ( shown in fig .
[ fig : m_hi_optsize ] ) , although with a larger scatter .
a linear fit with a slope and intercept of [email protected] and [email protected] , respectively is shown as a solid line . the larger scatter in the relation between m@xmath16 and the optical diameter ,
also seen in sample of brighter dwarfs ( e.g. @xcite ) , is probably indicative of a looser coupling between the gas and star formation in dwarfs , compared to that in spiral galaxies .
figure [ fig : mtol_size ] shows the hi mass to light ratio for the figgs sample plotted as a function of the hi extent , d@xmath23 . a trend of an increase in the m@xmath33 with an increase in the hi extent of the galaxies is clearly seen .
the best fit to the figgs sample shown as a solid line gives @xmath34 -0.75 cm van zee et al.(1995 ) from a hi mapping of a sample of low luminosity galaxies also found an evidence of an extended hi extent for high m@xmath33 galaxies .
figure [ fig : mtol_lb ] shows m@xmath33 for the figgs sample as a function of m@xmath35 . the same quantity for several other spiral and dwarf galaxies , spanning a range in absolute b magnitude from m@xmath36 to m@xmath37 is also plotted .
the sample from which these galaxies have been drawn are listed in the figure caption .
the galaxies in figgs sample with trgb distances are shown as solid circles , whereas the remaining figgs galaxies are shown as open circles .
the solid line shows an empirically determined upper envelope for m@xmath33 as a function of a m@xmath35 from @xcite .
this upper envelope can be interpreted as a minimum fraction of the total baryonic mass which needs to be converted into stars in order for a galaxy of a given baryonic mass to remain gravothermally stable ( @xcite ) .
it is interesting to note that except for and iv , all figgs galaxies lie much below this upper envelope .
this implies that these galaxies have converted much more baryons into stars than the minimum required for remaining stable . in this context , it is interesting to note that the average gas fraction for the figgs sample is 0.7 .
thus , for the majority of the dwarf galaxies in our sample , the baryonic mass is dominated by gas , rather than stars . in order to investigate the environmental dependence of the hi content for figgs galaxies , we plot m@xmath33 for figgs sample as a function of tidal index ( ti ) ( figure [ fig : mtol_ti ] ) . some additional gas rich galaxies with known hi extent
are also plotted in the figure .
ti is taken from @xcite and it represents the local mass density around a given galaxy , estimated using a large sample of galaxies within @xmath3 10 mpc of the milky way .
a negative value of ti for a galaxy indicates that the galaxy is isolated , whereas a positive number indicates that the galaxy is in a dense environment .
figure [ fig : mtol_ti ] .
shows that most of the figgs galaxies are in less dense environments , and that all the galaxies with high m@xmath33 ( i.e @xmath38 2.5 ) have negative tidal index i.e are isolated .
figure [ fig : size_ti ] shows the hi extent of the figgs sample , normalised to the optical ( holmberg ) radius , plotted as a function of ti . as seen in the figure , the galaxies with very extended hi disks ( @xmath39 ) are isolated . to summarize ,
we have presented the first results from the faint irregular galaxies gmrt survey ( figgs ) .
figgs is a large imaging program aimed at providing a comprehensive and statistically robust characterisation of the neutral ism properties of extremely faint , nearby , gas rich , dirr galaxies using the gmrt .
the gmrt hi data is supplemented with observations at other wavelengths .
the hi images in conjunction with the optical data will be used to investigate a variety of scientific questions including the star formation feedback on the neutral ism , threshold for star formation , baryonic tf relation and dark matter distribution in low mass galaxies .
the optical properties of the figgs sample , gmrt observations and the main science drivers for the survey are described .
the gmrt integrated hi column density maps and the hi spectra for the sample galaxies are presented .
the global hi properties of the figgs sample , derived from the gmrt observations , and their comparison with the optical properties of the sample galaxies are also presented .
a detailed comparison of the gas distribution , kinematics and star formation in the sample galaxies will be presented in the companion papers .
the observations presented in this paper were made with the giant metrewave radio telescope ( gmrt ) .
the gmrt is operated by the national center for radio astrophysics of the tata institute of fundamental research .
partial support for this work was provided by iltp grant b-3.13 .
begum , a. , chengalur , j.n . & hopp , u. , 2003 , new astronomy , 8 , 267 begum , a & chengalur , j.n . , 2003 ,
a&a , 409 , 879 begum , a & chengalur , j.n . , 2004 , a&a , 413 , 525 begum , a & chengalur , j.n . , 2004 , a&a , 424 , 509 begum , a & chengalur , j.n . , 2005 , mnras , 362 , 609 begum , a , chengalur , j.n .
& karachentsev , i. d. , 2005 , a&a , 433 , 1l begum , a , chengalur , j.n . ,
karachentsev , i. d. & sharina , m. e. , 2005 , mnras , 359 , 53l begum , a , chengalur , j.n . , karachentsev , i. d. , kaisin , s. s. & sharina , m. e. , 2006 , mnras , 365 , 1220 begum , a , chengalur , j.n . , kennicutt , r. c. , karachentsev , i. d. & janice , l. c. , 2008 , mnras , 383 , 809 bell , e. f. , mcintosh , d. h. , katz , n. & weinberg , m. d. , 2003 , apjss , 149 , 289 blanton , m. r. , geha , m. & west , a. a. 2007 ( astro - ph/0707.3813 ) bouchard , a. , jerjen , h. , da costa , g. s. & ott , j. , 2007 , aj , 133 , 261 bremnes , t. , binggeli , b. & prugniel , p. , 1999 ,
a&as , 137 , 337 bremnes , t. , binggeli , b. & prugniel , p. , 2000 , a&as , 141 , 211 broeils , a. h. & rhees , m. h. , 1997 , a&a , 324 , 877 carignan , c. , & purton , c. 1998 , apj , 506 , 125 cannon , j. m. , mcclure - griffiths , n. m. , skillman , evan d. & ct s. , 2004 , apj , 607 , 274 corbin , m. r. , kim , h. , jansen , r. a. , windhorst , r. a. & fernandes , r. c. , 2008 apj ( in press ) ( astro - ph-0710.2557v1 ) cote , s. , freeman , k. c. , carignan , c. & quinn , p. j. , 1997 , aj , 114 , 1313 ct s. , carignan , c. & freeman , k. c. , 2000 , aj , 120 , 3027 de blok , w. j. g. , bosma , a. & mcgaugh , s. , 2003 , mnras , 340 , 657 de blok , w. j. g. , 2005 , apj , 634 , 227 dekel , a. & woo , j. , 2003 , mnras , 344 , 1131 efstathiou , g. 2000 , mnras , 317 , 697 ekta , chengalur , j. n. , pustilnik , s. a. , 2006 , mnras , 372 , 853 ferguson , a. m. n. , gallagher , j. s. & wyse , r. f. g. , 2000 , aj , 120 , 821 geha , m. , blanton , m. r. , masjedi , m. & west , a. a. , 2006 , apj , 653 , 240 giovanelli , r. et al .
2005 , aj , 130 , 2613 hoffman , g. l. , salpeter , e. e. , farhat , b. , roos , t. , williams , h. & helou , g. , 1996 , apjs , 105 , 269 hopp , u. & schulte - ladbeck , r. e. , 1995 , a&as , 111 , 527 hunter , d. , 1997 , pasp , 109 , 937 hunter , d. a. & elmegreen , b. g. , 2006 , apjs , 162 , 49 huchtmeier , w. k. & richter , o. g. , 1986 , a&as , 63 , 323 huchtmeier , w. k. & skillman , e. , 1998 , a&as , 127 , 269 huchtmeier , w. k. , karachentsev , i. d. & karachentseva , v. e. , 2003 a&a , 401 , 483 kanekar , n. & chengalur , j. n. , 2005 , a&a , 429 , l51 karachentseva i. d. , prugniel , p. , vennik , j. , richter , g. m. , thuan , t. x. , & martin , j. m. , 1996 , a&as , 117 , 343 karachentsev i. d. , makarova , l. n. & andersen , m. i. , 1999 , mnras , 307 , 37l karachentsev i. d. , karachentseva v. e. , huchtmeier w. k. & makarov d. i. , 2004 , aj , 127 , 2031 karachentsev i. d. et al . , 2006 , aj , 131 , 1361 kennicutt , r. c. , 1989 , apj , 344 , 685 lo , k. y. , sargest , w. l. w. & young , k. , 1993 , aj , 106 , 507 makarova , l. , 1999 , a&as , 139 , 491 makarova , l. , karachentsev , i. d. , grebel , e. k. & barsunova , o. y. , 2002 , a&a , 384 , 71 makarova , l. , karachentsev , i. d. , grebel , e. k. , harbeck , d. , korotkova , g. g. & geisler , d. , 2005 , a&a , 433 , 751 mcgaugh , s. s. , schombert , j. m. , bothun , g. d. , de blok , w. j. g. , 2000 , apjl , 533 , 99 mcgaugh , s. s. , 2005 , apj , 632 , 859 matthews , l. d. , gallagher , j. s. , littleton , j. e. , 1995 , aj , 110 , 581 miralda - escude , j. & rees , m. j. , 1997 , apj , 478,57 navarro et al . , 2004 , mnras , 349 , 1039 parodi , b. r. , barazza , f. d. & binggeli , b. , 2002 , a&a , 388 , 29 pustilnik , s. a. & martin , j. m. , 2007 , a&a , 464 , 859 schaye , j. , 2004 , apj , 609 , 667 simon , j. d. & geha , m. , 2007 , apj , 670 , 313 simpson c. e. , hunter d. a. , nordgren t. e. , 2005 , aj , 130 , 1049 skillman , e. d. , 1987 , in star formation in galaxies , edited by c. j. lonsdale persson , nasa , 263 springob , c. m. , haynes , m. p. , giovanelli , r. & kent , b. r. , 2005 , apjs , 160 , 149 stil , j. m. & israel . f. p. , 2002 ,
a&a , 389 , 29 swarup , g. , ananthakrishnan , s. , kapahi , v.k . ,
rao , a.p .
, subrahmanya , c.r . & kulkarni , v.k .
1991 , current science , 60 , 95 swaters , r. , 1999 , phd thesis , rijksuniversiteit groningen taylor , c. l. , brinks , e. , pogge , r. w. & skillman , e. d. , 1994 aj , 107 , 971 taylor , v. a. , jansen , r. a. , windhorst , r. a. , odewahn , s. c. & hibbard , j. e. , 2005 , apj , 630 , 784 tully , b. et al . 2006 , aj , 132 , 729 van zee , l. , haynes , m. p. & giovanelli , r. , 1995 , aj , 109 , 990 van zee , l. , 2000 , apj , 119 , 2757 van den bosch , f. c. , swaters , r. a. , 2001 , mnras , 325 , 1017 verheijen , m.a.w & sancisi , r. , 2001 , a&a , 370 , 765 verheijen , m.a.w , 2001 , apj , 563 , 694 warren , b. e. , jerjen , h. & koribalski , b. s. , 2004 , aj , 128 , 1152 warren , b. e. , jerjen , h. & koribalski , b. s. , 2007 , aj , 134 , 1849 weldrake , d. t. f. , de blok , w. j. g. & walter , f. , 2003 , mnras , 340 , 12 young , l. m. , van zee , l. , lo , k. y. , dohm - palmer , r. c. , & beierle , m. e. 2003 , apj , 592 , 111 zwaan , m. a. , meyer , m. j. , steveley - smith , l. & webster , r. l. , 2005 , mnras , 359 , l30 | the faint irregular galaxies gmrt survey ( figgs ) is a giant metrewave radio telescope ( gmrt ) based hi imaging survey of a systematically selected sample of extremely faint nearby dwarf irregular galaxies .
the primary goal of figgs is to provide a comprehensive and statistically robust characterization of the neutral inter - stellar medium properties of faint , gas rich dwarf galaxies .
the figgs galaxies represent the extremely low - mass end of the dwarf irregular galaxies population , with a median m@xmath0 and median hi mass of @xmath1 m@xmath2 , extending the baseline in mass and luminosity space for a comparative study of galaxy properties .
the hi data is supplemented with observations at other wavelengths .
in addition , distances accurate to @xmath3 10% are available for most of the sample galaxies .
this paper gives an introduction to figgs , describe the gmrt observations and presents the first results from the hi observations . from the figgs data we confirm the trend of increasing hi to optical diameter ratio with decreasing optical luminosity ;
the median ratio of d@xmath4/d@xmath5 for the figgs sample is 2.4 .
further , on comparing our data with aperture synthesis surveys of bright spirals , we find at best marginal evidence for a decrease in average surface density with decreasing hi mass . to a good approximation
the disks of gas rich galaxies , ranging over 3 orders of magnitude in hi mass , can be described as being drawn from a family with constant hi surface density .
[ firstpage ] galaxies : dwarf galaxies : kinematics and dynamics radio lines : galaxies |
the discovery of thousands of extrasolar planets and planet candidates in recent years ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein and see exoplanet.org for a complete list ) , coupled with the rapidly increasing interest in the potential existence of extrasolar life , raise again in a big way the question of whether or not our solar system is special in any sense .
more specifically , we are interested in understanding whether the planetary and orbital properties in our solar system are typical or extremely unusual compared to those of extrasolar planets .
the solar system contains eight planets and two main belts ( the asteroid belt and the kuiper belt ) .
while tens of debris disks and warm dust belts ( similar perhaps to the solar system s asteroid belt ) have been observed and resolved , belts with dust masses as low as those in the solar system would currently be undetectable in extrasolar systems ( e.g. * ? ? ?
* ; * ? ? ?
consequently , we can quantitatively assess in detail how special the solar system is , only on the basis of its planetary components and properties such as its age and metallicity
. however , there are now hundreds of unresolved debris disk candidates ( e.g. * ? ? ?
* ) . of these , about two thirds of the systems
are better modelled by a two component dust disk rather than a single dust disk .
the two temperature components likely arise from two separate belts @xcite .
thus , the two belt configuration of our own solar system is plausibly fairly typical .
@xcite made an initial attempt to explore to what extent jupiter s periastron could be considered atypical compared to those of the giant planets known at the time .
their analysis , however , included only fewer than 100 exoplanets , most of which had been detected via radial velocity measurements .
consequently , selection effects dominated their conclusions a possibility fully acknowledged by the authors . in the present work we re -
examine the question of how special the solar system is .
in section [ special ] we use the much larger currently available database to consider the planetary orbital parameters .
we identify the semi major axis of the innermost planet as the most discrepant characteristic of the solar system and the low mean eccentricity as being somewhat special . in section [ prop ]
we compare the masses and densities of the planets in our solar system with those in exosolar systems . while the lack of a super earth in the solar system is somewhat unusual
, we argue that none of the characteristics identified make the solar system very special .
we discuss potential implications of our results in section 4 .
some of the apparent differences between the solar system and exoplanetary systems continue to be driven by strong selection effects that affect the sample .
we draw our conclusions in section 5 .
we first consider the statistical distribution of orbital separation and eccentricity of the observed planetary orbits . to allow for a more meaningful quantitative analysis
, we perform a @xcite transformation on the data .
this transformation makes the data closer to a normal distribution so that we can more accurately evaluate properties such as the mean and the standard deviation .
the transformation takes a skewed data set to approximate normality .
it is based on the geometric mean of the measurements and is independent of measurement units .
it is possible to do a multivariate box - cox transformation ( e.g. * ? ? ?
however , because of the selection effects associated with different parameters we choose to consider each parameter separately .
we transform the data with the function @xmath0 where @xmath1 is the parameter we are examining , such as the semi - major axis or eccentricity , and @xmath2 is a constant that depends upon the original distribution , that we discuss below .
the maximum likelihood estimator of the mean of the transformed data is @xmath3 where @xmath4 and @xmath5 is the @xmath6-th measurement of a total of @xmath7 .
similarly , the maximum likelihood estimator of the variance of the transformed data is @xmath8 we choose @xmath2 such that we maximise the log likelihood function @xmath9 this new distribution , @xmath10 , will be an exact normal distribution if @xmath11 or @xmath12 is an even integer .
we can measure how well the transformed distribution compares to a normal distribution with two parameters .
the skewness is @xmath13 where @xmath14 is the standard deviation of the distribution and thus we take @xmath15 .
the skewness measures the asymmetry of the distribution , a positive number implying the right hand tail of the distribution is longer , and a negative number that the left hand tail is longer .
furthermore , we can compare the kurtosis , @xmath16 this is a measure of the `` peakedness '' of the distribution and the heaviness of the tails .
a normal distribution has a kurtosis value of zero .
a positive value means that the distribution is tightly peaked but the tails are broad , and vice versa for a negative value . in the following subsections
we take the samples of the eccentricity and semi - major axis of the observed exoplanets and compare them to the planets in our own solar system .
there are a total of 539 exoplanets with a measured eccentricity . in the left hand panel of fig .
[ ecc ] we show the eccentricity distribution for this sample .
we find the maximum of the log likelihood function to be when @xmath17 .
the right hand panel of fig .
[ ecc ] shows the histogram of the box - cox transformed data . for our data
, we find the skewness to be @xmath18 .
for a normal distribution we expect the magnitude of the skewness to be up to @xmath19 .
thus the data are not heavily skewed .
the kurtosis is @xmath20 whereas for a normal distribution we would expect the magnitude to have values up to @xmath21 .
thus , the data have a slightly large kurtosis , or the distribution is not so tightly peaked as a normal one . for the transformed data ,
the mean is @xmath22 and the standard deviation is @xmath23 .
jupiter lies at @xmath24 from the mean .
similarly the earth lies at @xmath25 .
thus , the eccentricities of the planets in our solar system ( that range from venus at @xmath26 to mercury at @xmath27 ) are all relatively small compared to those of exoplanets , but not altogether significantly different .
recently @xcite considered how the mean eccentricity of planets in a system is correlated with the number of planets in the system .
they found a strong anti - correlation of eccentricity with multiplicity in systems observed by radial velocity .
an extrapolation of their relation up to 8 planets fits well with the mean eccentricity observed in the solar system .
furthermore , @xcite used the kepler exoplanets with asteroseismically determined stellar mean densities to derive a rather low eccentricity distribution of the multi
planet kepler systems . thus
, while the eccentricities in our solar system are low , those may be expected in a system with so many planets .
there is some bias in the exoplanet eccentricity data . for the rv planets ,
the best fit eccentricity is biased upwards from the true value leading to a reduced number of systems with a low eccentricity ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
however , the detection efficiency decreases only mildly with increasing eccentricity because despite being more difficult to detect , they have a larger rv amplitude for a fixed planet mass and semi - major axis @xcite . while planets found with the transit method require follow up observations ( for example , with rv ) to determine the eccentricity , the distribution of eccentricities is consistent with those of the rv planets @xcite . without the bias
, the eccentricities of the planets in our solar system would appear to be less special .
to date , there are a total of 1580 planets with a determined semi - major axis .
this increases up to a total of 5289 if we include planet candidates from kepler .
the false positive rate of the kepler exoplanets is low , especially for multi planet systems and the non giant planets ( e.g. * ? ? ?
* ) and thus we consider this much larger dataset also .
we find the maximum of the log likelihood function to be when @xmath28 ( @xmath29 , including planet candidates ) .
[ data ] shows the histogram of the box - cox transformed data .
the left hand panel includes only confirmed exoplanets . for this data
, we find the skewness to be @xmath30 .
this is only slightly higher than the upper value expected for a normal distribution of @xmath31 .
we find the kurtosis to be @xmath32 whereas for a normal distribution we would expect magnitudes smaller than @xmath33 .
thus , the data have a large kurtosis . for the transformed data ,
the mean is @xmath34 and the standard deviation is @xmath35 .
thus , jupiter lies at @xmath36 .
the right hand panel of fig .
[ data ] repeats this analysis but includes unconfirmed kepler planets .
the skewness for this is small at @xmath37 ( expected magnitude less than @xmath38 ) and the kurtosis is much smaller also , @xmath39 ( expected magnitude less than @xmath40 ) .
jupiter lies at @xmath41 , suggesting on the face of it that jupiter is rather special .
however , as we explain below , this is most likely a result of selection effects .
the majority of the planets in this distribution have been found by transit methods .
the planet with the largest semi - major axis found by this method is only at @xmath42 ( the planet is kic 11442793 h , * ? ? ?
kepler , is thought to be complete only for planets at least as large as the earth and for orbital periods up to a year ( e.g * ? ? ?
microlensing surveys preferentially find planets at radial distances of a few au from their host star , that is often an m dwarf ( e.g * ? ? ?
* ; * ? ? ?
this scale is dictated by the size of the einstein ring radius around the lensing star .
the radial velocity method has found planets in the range @xmath43 to @xmath44 ( e.g. * ? ? ? * ) .
direct imaging can detect planets at much larger distances ( e.g. * ? ? ?
* ; * ? ? ?
* ) , but so far only 8 planets have been detected by the method . in order to test the possibility that jupiter s outlier status is largely due to selection effects , we repeated the analysis but removed planets found by the transit method .
there remain @xmath45 planets in the sample .
we find the maximum of the log likelihood function to be when @xmath46 .
[ datano ] shows the histogram of the transformed data . for our data ,
we find the skewness to be @xmath47 .
this is less than the value expected for a normal distribution of @xmath19 . for our data
we find a high kurtosis of @xmath48 whereas for a normal distribution we would expect values with magnitude less than @xmath49 . for the transformed data , the mean is @xmath50 and the standard deviation is @xmath51 .
thus , jupiter lies at @xmath52 and continues to be somewhat of an outlier , but the trend suggests that this is most likely still due to selection effects .
the fact that direct imaging repeatedly reveals planets at separations much larger than jupiter s also may indicate that the current relative dearth of planets at large separations could be due to selection biases but more complete observations are required to test this possibility .
we should also note that @xcite used only the planet with the largest velocity semi - amplitude in each observed system in their plots .
however , they also performed the analysis with the most massive planet in each system and again with all the planets .
they reported no difference in the significance of jupiter as an outlier . in 2004
, they found that jupiter was at @xmath53 and half a sigma from its nearest neighbour .
we find that jupiter is not such an outlier as it was with the much smaller data set in 2004 , and selection effects continue to affect the distribution .
this analysis should again be repeated once we have more reliable observations around the orbital radius of jupiter .
currently , our inner solar system appears to be rather special compared to observations of exoplanet systems .
the inner edge of our solar system is at the orbit of mercury at @xmath54 , while exoplanetary systems are observed to habor planets much closer to their star .
we find that mercury lies at @xmath55 above the mean , while the earth is at @xmath56 above the mean of the distribution of confirmed exoplanet semi - major axes ( as shown in the left hand panel of fig .
[ data ] ) .
when we include all of the kepler candidates ( right hand panel of fig .
[ data ] ) , these increase to @xmath57 for mercury and @xmath58 for the earth .
thus , all of the planets in our solar system have orbital semi - major axes that are larger than the mean observed in exoplanetary systems .
however , it is possible that this could be the result of selection effects as it is easier to find planets in this region , if they are there . in terms of the radial location of observed exoplanets ,
the lack of close in planets in our solar system is the parameter that makes our solar system most special .
we should note though that if we use only the planets found by methods other than the transit , then mercury is at @xmath59 and the earth is at @xmath60 .
consequently it is difficult to say how significant this discrepancy is .
@xcite suggested that the migration of jupiter and saturn into the terrestrial planet forming region of our solar system ( down to @xmath61 ) led to the depletion of mass in @xmath62 .
the planets are then thought to migrate outwards to their current location ( see also * ? ? ?
however , whatever the formation mechanism for the giant planets in our solar system might have been , it is not thought to have been specific to our solar system , and thus neither is a depleted inner solar system .
we discuss this point further in section [ discussion ] .
finally in this section , we note that migration of planets through the protoplanetary disk or planet - planet interactions or secular interactions of a binary star could affect these distributions , especially that of the semi - major axes .
for example , it may be theoretically impossible for jupiter mass planets to form at the radial location of hot jupiters ( e.g. * ? ? ? * ) . instead
, they are supposed to form outside of the snow line and migrate inwards through the protoplanetary disk before the latter disperses ( e.g. * ? ? ?
migration could also occur by the kozai
lidov mechanism increasing the eccentricity of the planet followed by tidal circularization @xcite .
evidence obtained by the rossiter
mclaughlin effect suggests that some fraction of the hot jupiters may have been produced through dynamical interactions ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . hot
jupiter planets dominated the initial planet discoveries because they are large and close to their host star .
however , we now know that they are quite rare and orbit only about one percent of solar type stars ( e.g. * ? ? ?
opinions vary on whether in the solar system jupiter has significantly migrated . on one hand @xcite suggested that jupiter did not migrate much from its formation location , and on the other , @xcite proposed that the low mass of mars could be explained by gas - driven early migration of jupiter . @xcite
further suggested that this formation process could explain the lack of objects in our inner solar system .
@xcite assumed that planets form constantly at a radius of @xmath63 and found theoretically that about @xmath64 to @xmath65 of systems will have a jupiter mass planet that does not migrate significantly .
however , this conclusion will be affected by the presence of a dead zone ( a region of the disk with no turbulence , e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) that may slow or halt migration altogether . furthermore , the inner and outer edges of a dead zone may act as planet traps that stop migration ( e.g. * ? ? ?
* ; * ? ? ?
thus , the distribution of planet semi - major axes definitely does not represent the initial distribution at the time of planet formation .
given that jupiter mass planets are thought to form in the vicinity of jupiter s current radial location , theory suggests that the radial location of jupiter is not particularly special .
the current observational bias towards planets that are close to their host star means that it is easier to find planets that have migrated inwards , rather than those that have not , or even those that may have migrated outwards .
we expect in the future that with more complete observations of jupiter mass planets at jupiter s radial location we will be able to constrain the migration mechanisms and uncover how special jupiter really is for its small ( net ) distance of migration .
in this section we consider how the masses and densities of the planets in our solar system compare to those in exosolar systems . in this context , we discuss also the potential significance of the lack of a super - earth in our solar system .
[ mass ] shows the distribution of the approximate masses of the exoplanets that have been observed to date , where @xmath6 is the orbital inclination . for directly imaged planets ,
the mass is predicted by theoretical models of the planets evolution . for the planets that have been observed by microlensing
, it s the ratio of the planet to star mass that is measured with accuracy . ] .
the masses of the planets within our solar system are shown with arrows at the top ( but are not included in the data ) .
the masses of the gas giants fit well with those of exosolar planets , but the terrestrial planets are all on the low side .
this is most likely due ( at least partially ) to the difficulty in finding low - mass planets .
the masses of the exoplanets are strongly biased towards high mass and short period planets .
kepler has shown us that small planets are very common but the mass measurement of small mass planets is difficult and thus currently they appear to be rare .
there are two peaks in the data , the first of which is at a mass between that of the earth and that of uranus , at around @xmath66 , where @xmath67 is the mass of jupiter .
planets with a mass in the range of @xmath68 to @xmath69 ( where @xmath70 is the mass of the earth ) are known as super - earths ( e.g. * ? ? ?
our solar system does not contain any super - earths thus in that sense it is somewhat unusual .
we discuss this further in subsection [ superearth ] .
the second peak is around the mass of jupiter .
we fit the exoplanet mass data with a binormal probability density function ( pdf ) @xmath71 where @xmath72 . with a kolmogorov - smirnov ( ks )
test we find the best fitting parameters to be @xmath73 , @xmath74 , @xmath75 , @xmath76 , @xmath77 and we show the pdf as the solid line in fig . [ mass ] . with this distribution , we find that jupiter is very typical at only @xmath78 from the higher - mass peak , while saturn is at @xmath79 .
the terrestrial planets are hard to compare because there are so few data points for those small masses
. however , neptune lies at @xmath80 and uranus at @xmath58 from the lower - mass peak .
gas giants are thought to form outside of the snow line radius in the protoplanetary disk where there is more solid material available to form massive planets ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the surface density of the protoplanetary disk decreases with increasing distance from the star and the timescale to form a planet increases with radius .
thus , lower mass gas giants could preferentially form farther away from the star .
low mass and large orbital radius planets are certainly harder to detect than those with higher mass and lower orbital radius .
this can explain the lack of observed small mass planets , but also perhaps the dip in the observed distribution .
for example , if uranus and neptune were around another star , at their large orbital radii , they would also be difficult to detect .
the only method that could currently detect them is direct imaging .
however , the smallest mass planet that has been found with this method is formaulhaut b that has an approximate mass of @xmath81 @xcite .
it therefore remains a possibility that the double peaked mass distribution is solely the result of selection effects . .
the total number of exoplanets is 1516 .
, width=317 ] . ,
width=283 ] fig . [ density ] shows the approximate densities of the observed planets as a function of their mass .
the exoplanets are shown in blue and the planets in our solar system in red . while the lower mass exoplanets have a large range in their density for a given mass ( see also
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , the giant planets show a clear correlation of increasing density with mass .
thus , the super - earths may have a wide range of compositions @xcite .
despite the large spread in the data for the low mass planets , there appears to be a trend of decreasing density with increasing mass .
this could be attributed qualitatively to the peak in the theoretical radius against mass of a planet ( see e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
more recently , it has been suggested that the super - earths ( with radii in the range @xmath82 ) show two trends separated by a critical planet radius .
the smallest planets increase in density with radius while those that are larger decrease suggesting that the larger planets have a large amount of volatiles on a rocky core .
there is some uncertainty over the value of the critical radius , as estimates range from about @xmath83 to @xmath84 @xcite , if it exists at all @xcite .
the data suggest that the densities of the giant planets within our solar system are very typical of those of observed exoplanets .
the masses of the terrestrial planets in our solar system are on the edge of our current sensitivity and thus it is hard to draw any conclusions about their densities .
however , recently , @xcite found that the earth ( and venus ) can be modelled with the same ratio of iron to magnesium silicate as the low mass exoplanets observed and thus the earth may not be special in this respect .
it is interesting to examine whether our solar system s lack of a super - earth is truly unusual .
there have been several attempts to calculate an occurrence rate for super - earths taking into account the selection biases .
the results for rv observations predict an occurrence rate in the range @xmath85 in the period range @xmath86 @xcite .
the transiting planet observations imply a range of occurrence rates that is at most as high as @xmath87 in the period range @xmath88 @xcite .
more recently @xcite examined the kepler sample for planets with radii in the range @xmath89 with orbital periods in the range 50 to @xmath90 and found an occurrence rate of @xmath91 . although these results include earth - size planets and some of the periods are longer than that of mercury , the occurrence rate increases with short orbital periods , making the existence of close - in planets more likely .
the high occurrence rate of these types of planets offers perhaps the strongest argument against the solar system being very common , but even that does not necessarily make it extremely rare .
typically , systems that have an observed super - earth , have more than one and this is theoretically expected if the planets form by mergers of inwardly migrating cores ( e.g. * ? ? ?
* ; * ? ? ?
it is possible that the presence of a super - earth can affect terrestrial planet formation .
many of the super - earths observed are at small radial locations , where theoretically they could not have formed ( e.g. * ? ? ? * ; * ? ? ? * ) .
@xcite found that if a super - earth migrates sufficiently slowly through the habitable zone ( defined as the radial range of distances from the star at which a rocky planet can maintain liquid water on its surface ) then any terrestrial planet that later forms there would be volatile - rich and not very earth - like . in conclusion ,
the masses and densities of the planets of our solar system appear to be very typical of those of exoplanets .
however , the lack of a planet with a mass in the range @xmath92 , a super - earth , makes the solar system appear somewhat special .
and mercury , being the inner most planet , at @xmath93 . , width=317 ] generally , the physical properties of the planets in our solar system are quite typical when compared to those of the observed exoplanets , although the lack of a super earth is unusual
. the orbital properties , however , may be somewhat special and perhaps more conducive to life .
low eccentricity planets have a more stable temperature throughout the orbit and therefore may be more likely to host life ( e.g. * ? ? ?
* ; * ? ? ?
. furthermore , planetary systems with a low average eccentricity are more likely to have long term dynamical stability .
for example , the terrestrial planets in our solar system are expected to be dynamically stable at least until the sun becomes a red giant and engulfs the inner planets ( e.g. * ? ? ? * ) .
there are a few other factors that could , in principle at least , make our solar system special with respect to the emergence of life .
first we can consider the age .
the current age of our sun is about half the age of the disk of our galaxy , and also half of the sun s total lifetime .
thus , we expect that roughly half of the stars in our galaxy s disk are older and half are younger than our sun .
this implies that the age of our solar system is definitely not special . furthermore , @xcite considered the planet formation history of the milky way and determined that our solar system formed close to the median epoch for giant planet formation .
they also found that about @xmath94 of the currently - existing earth - like planets were already formed at the time of the earth s formation .
we also note that the fact that the solar system contains a single star does not make it particularly special , since the binary fraction in the kepler sample , for example , is about 50% @xcite .
the presence of terrestrial planets in the habitable zone around their host star appears to be quite common .
for example , @xcite examined the kepler data for m dwarfs and found that for orbital periods shorter than @xmath95 , the occurrence rate of earth - size planets in the habitable zone is around @xmath96 .
this conservative estimate could be as high as @xmath97 depending on how the habitable zone is defined .
this is consistent with radial velocity surveys that find @xmath98 potentially habitable planets per m dwarf @xcite .
thus , an earth - size planet in a habitable zone is not uncommon .
a habitable planet may require a large moon which in turn may require an asteroid collision @xcite .
thus , systems which contain an asteroid belt may be more conducive to initiating life .
however , habitability may be sensitive to the size of the asteroid belt @xcite .
as we have noted in the introduction , asteroid belts could be a common feature of planetary systems @xcite .
the metallicity of a protoplanetary disk ( and hence the host star ) determines the structure of a planetary system that forms ( e.g. * ? ? ?
* ; * ? ? ?
the higher the metallicity of a star , the more giant planets that are observed ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
however , the correlation for lower mass planets is unclear @xcite .
planets with radii less than four times that of the earth are observed around stars with a wide range of metallicities . however , the average metallicity of stars hosting small planets ( @xmath99 ) in @xcite is very close to solar .
although such planets can form at a wide range of metallicities , the fact that the average metallicity of the small planets is solar may not be a coincidence .
thus , while the metallicity of our solar system may not be especially promotive to the formation of a habitable planet , it s unclear whether the solar system is special or not in this respect . the variability of our sun has been compared to the activity of stars in the kepler sample with conflicting conclusions . @xcite and @xcite found that the sun is rather typical with only a quarter to a third of stars in the kepler sample being more active than the sun . on the other hand , @xcite found the sun to be relatively quiet with @xmath100 of stars being more active .
the difference in the conclusions stems from choices in defining the activity level of our sun and the inclusion of stars with fainter magnitudes in @xcite .
however , the studies agree that the active fraction of stars becomes larger for cooler stars .
m dwarfs have a fraction of @xmath101 that are more active than the sun .
thus , compared to other sun - like stars , our sun could be typical , but compared to cooler stars , our sun is certainly quiet . in general , there are three aspects in which the solar system differs most from other observed multi - planet systems .
first , the low mean eccentricity of the planets in the solar system maybe somewhat special , although this may be accounted for by selection effects .
secondly , there is in the total lack of planets inside mercury s orbit .
massive planets migrating through the habitable zone can change the course of planet formation in that region .
overall , however , processes that could act to clear the inner part of the solar systems ( such as giant planet migration ) , are believed to be operating within a non - negligible fraction of the exoplanet systems ( e.g. * ? ? ?
third , the lack of super - earths in our solar system is somewhat special and could have allowed the earth to become habitable .
a close - in super - earth could also affect the dynamical stability of a terrestrial planet in the habitable zone and this should be investigated in future work .
we consider the first two of these special parameters in more detail in fig .
[ average ] . for planets with measured eccentricity
, we plot the mean eccentricity of the planetary system and the semi major axis of the innermost planet observed within the system . in this parameter space
, the solar system appears to be somewhat special , but far from being rare .
although , most of the systems with three or more planets , do have a planet with an orbital semi - major axis smaller than that of mercury , this could be at least partly due to selection effects . the observed semi
major axis of the innermost planet may be close to complete but the number of planets in each system is certainly not . if the innermost planet is very close in , then it is easier to detect the planets outside of its orbit .
if on the other hand the innermost planet is farther out ( e.g. some of the small blue and red points in fig .
6 ) then additional planets will be difficult to find ( see also discussion in section [ a ] ) .
there are many factors that may be required in order to form a habitable planet .
when we multiply the probabilities for each together , we may end up with a small probability for such an event .
however , since we currently do not know which factors are truly important for life to emerge , such an exercise does not make much sense . if we consider too many details , clearly the solar system is special because all systems are different .
however , at the moment we have not identified any parameter that makes the solar system so significantly different that it would make it rare .
we find that the properties of the planets in our solar system are not so significantly special compared to those in exosolar systems to make the solar system extremely rare .
the masses and densities are typical , although the lack of a super earth sized planet appears to be somewhat unusual .
the orbital locations of our planets seem to be somewhat special but this is most likely due to selection effects and the difficulty in finding planets with a small mass or large orbital period .
the mean semi - major axis of observed exoplanets is smaller than the distance of mercury to the sun .
the relative depletion in mass of the solar system s terrestrial region may be important .
the eccentricities are relatively low compared to observed exoplanets , although the observations are biased towards finding high eccentricity planets .
the low eccentricity , however , may be expected for multi - planet systems .
thus , the two characteristics of the solar system that we find to be most special are the lack of super - earths with orbital periods of days to months and the general lack of planets inside of the orbital radius of mercury . from the perspective of habitability
the solar system does not appear to be particularly special .
if exosolar life happens to be rare it would probably not be because of simple basic physical parameters , but because of more subtle processes that are related to the emergence and evolution of life .
since at the moment we do nt know what those might be , we can allow ourselves to be optimistic about the prospects of detecting exosolar life .
we should make every possible effort to detect and characterise the atmospheres of a few dozen earth - size planets in the habitable zone , in the coming two decades .
we thank the anonymous referee for useful comments . this research has made use of the exoplanet orbit database and the exoplanet data explorer at exoplanets.org . | with the availability of considerably more data , we revisit the question of how special our solar system is , compared to observed exoplanetary systems . to this goal , we employ a mathematical transformation that allows for a meaningful , statistical comparison .
we find that the masses and densities of the giant planets in our solar system are very typical , as is the age of the solar system . while the orbital location of jupiter is somewhat of an outlier , this is most likely due to strong selection effects towards short - period planets .
the eccentricities of the planets in our solar system are relatively small compared to those in observed exosolar systems , but still consistent with the expectations for an 8-planet system ( and could , in addition , reflect a selection bias towards high - eccentricity planets ) .
the two characteristics of the solar system that we find to be most special are the lack of super - earths with orbital periods of days to months and the general lack of planets inside of the orbital radius of mercury .
overall , we conclude that in terms of its broad characteristics our solar system is not expected to be extremely rare , allowing for a level of optimism in the search for extrasolar life .
[ firstpage ] |
on the 23rd of february , 1987 , the kamiokande ii , imb , and baksan experiments observed the neutrino burst from sn1987a , which was located in the large magellanic cloud @xcite .
this was the first detection of a supernova s neutrino burst , and it introduced a new method of investigation : neutrino astronomy .
super - kamiokande ( super - k , sk ) is an imaging water cherenkov detector containing 50,000 tons of pure water ; it is the successor to the kamiokande detector .
super - k is located 1000 meters underground ( 2,700 meters of water equivalent ) in the kamioka zinc mine in the gifu prefecture of japan , at 36.4@xmath3n , 137.3@xmath3e and 25.8@xmath3n geomagnetic latitude .
the detector consists of a main inner detector and an outer veto detector .
both detectors are contained within a cylindrical stainless steel tank 39.3 m in diameter @xmath4 41.4 m in height .
the usual fiducial mass for neutrino measurements is 22.5 ktons with boundaries 2.0 m from the inner surface .
the outer detector is also a water cherenkov detector of 13,000 metric tons total mass .
it surrounds the inner detector as a 4@xmath5 solid - angle anti - detector to detect any signals coming from outside of the detector and to shield against external gamma - rays and neutrons .
the data set for the analyses was taken during two periods .
the first period started on the 1st of april , 1996 , and terminated on the 15th of july , 2001 .
a total of 11,146 photo multipliers ( pmt s ) with 20-inch diameter photocathodes provided active light collection over 40% of the entire surface of the inner detector .
this phase of the project is now referred to as super - kamiokande - i ( sk - i ) .
the second phase , super - kamiokande - ii ( sk - ii ) , started on the 10th of december , 2002 , and terminated on the 6th of october , 2005 .
a total of 5,182 20-inch pmt s , each protected by acrylic and fiber - reinforced plastic ( frp ) cases , were mounted on the inner detector , providing 19% photocathode coverage during this period . in the outer detector , a total of 1,885 8-inch pmt s were installed during both periods .
due to rapid variations in the water transparency , the data before the 31st of may , 1996 in sk - i , and the data before the 23rd of december , 2002 in sk - ii have not been used for analysis because of uncertainties in the energy calibration .
having a well - defined energy response is necessary when searching for low multiplicity event clusters , as will be described in the later sections of this paper .
theoretical calculations predict the characteristics of neutrinos expected from a supernova .
a typical core - collapse supernova explosion emits all types of neutrinos and has a total energy output of @xmath6 ergs which would generate about 10,000 sk events ( 9,000 events without neutrino oscillation ) in the case of a supernova at a distance of 10 kpc from the earth .
neutrino oscillations enhance the overall flux we observe by 10% as higher - temperature @xmath7 and @xmath8 get fully mixed into the observable @xmath9 signal .
these supernova events will be detected via the following interactions in sk , where the numbers in parentheses show the fractions of the total number of events with / without neutrino oscillations , @xmath10 and @xmath11 means the total interactions of @xmath12,@xmath13 , and their anti - neutrinos . to obtain these fractions @xcite , a supernova neutrino burst using the livermore model @xcite
is assumed . to account for neutrino oscillations in ref .
( takahashi et al .
2001),@xmath14 and @xmath15 are set to be in the favored solar neutrino lma region , and @xmath16 and @xmath17 are determined from the best oscillation fit for the atmospheric neutrinos .
for the fraction of the neutral current interactions with oxygen ( eq . [ nc ] ) , we look to ref.@xcite and obtain our number by scaling their result to the numbers in ref . @xcite . at present , based on both data and models , many of these numbers are uncertain up to a factor of two , especially the numbers of interactions with oxygen . in particular , eqs.[cc1 ] , [ cc2 ] , and [ nc ] are very dependent on the neutrino temperatures and details of mixing as discussed in the cited literature .
averaged energies of neutrinos from a standard delayed explosion supernova model are expected to be between 11 @xmath18 26 mev ( totani et al .
1998 ; thompson et al .
2003 ; sumiyoshi et al .
2005 ) for example , the calculation performed by the livermore group shows @xmath19 mev , @xmath20 mev and @xmath21 mev @xcite .
the time profile of each type of neutrino has a unique shape . during the initial @xmath1810 milliseconds , electron neutrinos from the neutronization
are released with a total energy on the order of @xmath22 ergs .
after the neutronization , all flavors of neutrinos are produced by electron - positron annihilation , and released with a total energy on the order of @xmath23 ergs with a time scale of several tens of seconds .
it still is not clear at present , however , whether or not the delayed explosion scenario is the correct picture , in which a stalled shock wave will be reheated by energy deposition from neutrinos . we expect that the observation of a galactic supernova by the sk detector with high statistics will solve this long - standing question and , furthermore , shed light on other important questions such as neutrino oscillations and neutrino mass @xcite . in this paper , we have searched for supernova neutrino bursts using the sk detector , and we present an upper limit on the supernova explosion rate within 100 kpc of the earth ( this volume contains our galaxy , the lmc , and the smc ) .
we also report the first result of search for neutronization bursts .
previously , amanda , baksan , imb , and macro have reported upper limits on the supernova explosion rate just within our galaxy @xcite , while lvd and kamiokande have reported preliminary results @xcite of the same nature .
from the 31st of may , 1996 , to the 15th of july , 2001 , and from the 24th of december , 2002 , to the 5th of october , 2005 , the livetimes of our detector for supernova searches were 1703.9 days for sk - i , and 885.3 days for sk - ii .
livetime efficiency as a function of date is shown in figure [ livtim ] over the course of the entire data - taking period the average efficiency was about 89% .
the primary cause of lost supernova livetime was scheduled calibration work .
the large losses of livetime seen in figure [ livtim ] around july 1997 , december 1998 , june 1999 , may 2001 , march 2003 , and september 2005 were due to linac energy calibration runs @xcite , though the biggest dip in 2003 was due to electronics problems which have since been corrected .
vertex and energy reconstruction techniques are the same as those used in our solar neutrino analysis @xcite .
fiducial volume for the supernova search is also 22.5 ktons , though the energy thresholds are 6.5 mev ( sk - i ) and 7.0 mev ( sk - ii ) to avoid the higher background rates associated with the lower thresholds used in the solar neutrino analysis .
data reduction steps are basically the same .
the first reduction includes removing events due to electronics noise , events with a poor reconstruction of the vertex position , and events with a vertex outside of the fiducial volume .
a spallation cut removes events which are produced by energetic muons by using a likelihood method in which both the time difference @xmath24 and distance @xmath25 between the parent muon and subsequent events are considered @xcite .
mis - reconstructed events are also removed .
after these noise reductions , the remaining event rates are 180 events / day for sk - i , and 164 events / day for sk - ii . in this section
, we introduce our general method of supernova burst search .
the procedure of data analysis is as follows : 1 .
scan event times in the data set using a sliding time - window .
we define a `` cluster '' if the number of events ( the `` multiplicity '' ) within a time - window is greater or equal to a certain threshold .
2 . check each cluster found in the first step to determine if it is a real signal from a supernova or the result of background events .
a background cluster consists of either time - correlated non - supernova events or a chance coincidence of uncorrelated low energy events .
there are two dominant background sources which produce time - correlated events .
one such source is flasher pmt s which act as sources of light , and the other is spallation product , radioactive isotopes made via interactions between energetic cosmic ray muons and oxygen nuclei . in both cases ,
the reconstructed vertices of the resulting events are spatially concentrated . on the other hand , events made by actual supernova neutrinos
should be generated uniformly in the detector volume .
therefore , to distinguish real signals from background clusters , clusters are checked by studying the correlation between the multiplicity and the events spatial distribution ( @xmath26 ) .
@xmath26 is defined by the averaged spatial distance between each event as follows : @xmath27 where @xmath28 is the multiplicity , @xmath29 is the distance between events @xmath30 and @xmath31 , and @xmath32 is the number of unique combinations . in the case of a supernova burst , @xmath26 should have a larger value than that resulting from spatially clustered events such as spallation events or flasher events .
figure [ rmean_mc ] shows @xmath26 distributions of simulated supernova events for multiplicity 2 , 3 , 4 , and 8 , which are the multiplicity thresholds of various burst searches discussed in later sections of this paper .
as shown in this figure , if the events in a cluster occur uniformly in the detector , as in supernova , then the expected value of @xmath26 tends to be large , around 1,800 cm .
the threshold of @xmath26 for a cluster with multiplicity equal to 2 is set to 750 cm , and for clusters with multiplicity greater than 2 the threshold is set to 1,000 cm .
the efficiencies for supernova events using these criteria are 94% , 96% , 99% , and 100% , respectively . in the following sections , we obtain results from three kinds of burst searches by changing the width of the time - window and the multiplicity threshold and setting additional criteria for respective searches .
the three burst searches are 1 ) distant supernova search , 2 ) supernova burst search with low energy threshold , and 3 ) neutronization burst search . in recent years , a lot of effort has been put into optical supernova searches , with the result that the number of detected supernovae has increased substantially .
the optical detection rate of supernovae suggests that theoretical calculations of supernova rates might be significantly underestimated .
more specifically , nine core - collapse supernovae have been found in nearby galaxies between 2002 and 2005 ; this is three times higher than one theoretical estimate @xcite .
this motivates us to search for neutrinos from supernovae in nearby galaxies .
since the expected total number of events at sk from a supernova in the andromeda galaxy ( @xmath33700kpc ) is around two events , for this search we set our criteria as a long time - window and low multiplicity threshold such as @xmath34 2 events / 20 seconds .
a lot of background clusters would be found in our usual solar neutrino data set @xcite with this criterion .
therefore , we reduce those backgrounds by setting a higher energy threshold because the average energy of the emitted positron from the interaction ( eq.[nebp ] ) is higher than that of most low energy background events . to find the best energy threshold ,
we calculate the detection probability of a supernova and the number of chance coincidences of the low energy background events as functions of energy threshold value .
then we take the value of energy threshold which gives us the maximum value of @xmath35 as an optimized energy threshold .
the resulting value we set is 17 mev , where the single event rate with this energy threshold is 0.762 event / day for sk - i , and 1.03 event / day for sk - ii .
figure [ ltw ] is a scatter plot of @xmath26 versus the multiplicity for each cluster which satisfies the criteria with an energy threshold of 17 mev .
three candidate clusters exist that have @xmath26 more than 1000 cm , but as shown in table [ time_table ] , the event times of those candidates actually coincide with times of mine blasting . because of the physical vibration of pmt s due to blasting
, huge electrical noise occurred , and those noises caused time - clustered a few hundreds events within a few seconds during or after blasting .
it was also confirmed that these events including the candidate cluster events have a characteristic pmt hit pattern naerly in the same area of the inner detector , and hence they should be eliminated as potential supernova events .
cccc no.1 & jul.13th 1999 19:00 & 1257.89 & 4 + no.2 & may.12th 2000 11:06 & 1928.76 & 35 + no.3 & oct.12th 2000 19:03 & 1381.76 & 2 + the background clusters which have small @xmath26 values in fig .
[ ltw ] were mainly found around times of calibration work , and more importantly the positions of these clusters were consistent with the positions of the calibration sources themselves . that these clusters made it into our supposedly all non - calibration data sample
is therefore judged to be the result of occasional operator error in assigning the proper run type during transitions between normal and calibration data taking .
( note that such mis - labeled runs are excluded in our usual solar neutrino data analysis , but in order to maximize supernova livetime a wider variety of runs were included in this supernova analysis . ) therefore , no real supernova signal was observed during the data - taking periods in both sk - i and sk - ii . in this section , we set various time - windows of 0.5 , 2 , and 10 seconds with a lower energy threshold of 6.5 mev for sk - i and 7.0 mev for sk - ii to search for signals from a supernova in this lower energy region .
in fact , the data of sn1987a imply @xmath36 mev at kamiokande and 11.1 mev at imb @xcite while @xmath20 mev to take the livermore group model as an example @xcite . since a lot of background events due to spallation or flasher pmt s still remain in the lower energy regions , we set criteria of higher multiplicity for each time - window as @xmath34 3 events / 0.5 seconds , @xmath34 4 events / 2.0 seconds , and @xmath34 8 events/ 10 seconds . if a cluster satisfies the requirement @xmath26 , we closely check data around the cluster for example all events within @xmath3720 seconds around the cluster so that we get information from as many supernova neutrinos as possible .
figure [ ml_rm_data ] shows the correlation between @xmath26 and multiplicity for obtained clusters from sk - i and sk - ii .
there were only three clusters which had large @xmath26 values in sk - i .
as mentioned in the previous section , however , all three candidate clusters consist of the same events as in the previous section which were found during periods of mine blasting .
the same criteria are applied to sk - ii data , and the correlation of the multiplicity and @xmath26 of the candidate clusters was studied .
there was no cluster with @xmath38 1000 cm , but two candidate clusters had @xmath26 just below 1000 cm . to make sure that those candidates were not real signals , cosmic ray muon events around the cluster
were studied , and event displays of candidate events were checked . as a result ,
one of the candidates was found to be an accidental coincidence of a low energy event with spallation events .
although most spallation events have already been removed by the spallation cut , a small fraction of spallation products are still present at lower multiplicities .
the other candidate cluster was comprised of false events caused by a flasher tube . as in the previous search
, it was confirmed that the background clusters with small @xmath26 were mostly due to mislabeled calibration work . in conclusion
, the remaining candidate clusters were all caused by mine blasting , and so there is no clear evidence for any supernova neutrino burst in the data obtained by sk - i as well as sk - ii .
therefore , we will present an upper limit at the 90% confidence level for the rate of supernova explosions in section 3 .
we conducted another burst search with shorter time - windows to investigate the short - lived neutronization burst of @xmath39 events from a supernova . prior to the core explosion
, many @xmath39 s are emitted via the reaction @xmath40 as the shock wave propagates into the exploding star s outer core .
the shock wave dissociates nuclei into free nucleons on which the cross section of electron capture is larger than that on nuclei ; the resulting burst of @xmath39 s thus forms the so - called neutronization burst .
the duration of the neutronization burst is on the time scale of the shock wave propagation , which is less than 10 milliseconds .
hence , if such a short time window is set , the @xmath39 s can be dominant , and the expected number of neutronization burst neutrinos which will be observed at sk is between one and six ( depending on neutrino oscillation models ) in the case of a supernova at a distance of 10 kpc from the earth @xcite .
it should be mentioned that even if no supernova explosion is observed in the galaxy during that period , it still might be possible to observe only the neutronization burst .
for example , if a black hole forms shortly after the neutronization stage , then the main burst of supernova neutrinos might not be able to escape from the black hole @xcite . based on the above theoretical expectations , we set time - windows of 1 , 10 , and 100 milliseconds .
the multiplicity threshold is two events for each time window ; @xmath34 2 events / 1 msec , @xmath34 2 events / 10 msec , or @xmath34 2 events / 100 msec . because recoil electrons will have lower energies from neutrino
electron scattering ( eq.[nee ] ) which is the dominant interaction in this case , we use the same sample as the sk - i solar neutrino analysis @xcite , whose energy threshold is 5 mev and livetime is 1496 days , and we use the same sample as in previous sections for sk - ii . since the threshold of multiplicity for a candidate cluster is two events , a lot of backgrounds exist even after the @xmath26 cut .
therefore , we need another variable to reduce background . in this case , the directional information of observed events is a strong tool because the recoil electrons have almost the same direction as the incident neutrinos , which means that events from a real neutronization burst should have roughly the same reconstructed direction . to check the isotropy of events in a candidate cluster ,
a new variable is defined as follows : @xmath41 where @xmath42 is a reconstructed direction vector of events in a cluster , and @xmath28 is the multiplicity of the cluster . by this definition , @xmath43 will be close to 1 in the case of a real supernova cluster .
figure [ sumdir ] shows the @xmath43 distributions of supernova monte carlo and background monte carlo .
we set the threshold of @xmath43 for a supernova candidate cluster to 0.75 as this is the point at which the signal - to - noise ratio exceeds unity ; the efficiency for real supernova events is estimated to be 84% .
table [ tab : shtw ] shows the number of observed candidates during the periods of sk - i and sk - ii after @xmath26 and @xmath43 cuts .
there is good agreement between the number of observed clusters and the number of expected backgrounds which can be estimated by @xmath44^{j-1}}{(j-1)!}\ ] ] where @xmath45 is threshold of multiplicity , @xmath46 and @xmath47 are single event rate and live time of each data taking period ( maximum 24 hours ) , and @xmath24 is the time - window for each case .
we can use a poisson - based estimate for the background because , after the @xmath26 cut , most of the background clusters are due to chance coincidences of low energy events such as solar neutrino events , flasher events , and spallation events in the data samples .
since there were no candidates with stricter criteria : @xmath34 3 events / 1 msec , @xmath34 3 events / 10 msec , or @xmath34 3 events / 100 msec , also in agreement with the expected background as shown in table [ tab : shtw ] , we conclude that no signal from a real neutronization burst was observed during this period .
we have searched the sk - i and sk - ii data for neutrino burst signals from supernova explosions .
we conclude that no real signals of supernova bursts occurred during the data taking periods between late may 1996 and early october 2005 , which corresponds to a total livetime of 2589.2 days .
super - k also performed an all sky search for transient astrophysical neutrinos in the gev - tev energy region and did not find anything @xcite , so we can rule out the detection of high energy neutrinos from any supernovae as well .
we can evaluate the performance of super - k as a supernova watcher based on these results .
we simulate neutrino events in the tank to estimate detection probability of a supernova as a function of distance from the earth .
the incident neutrinos are assumed to be emitted by a supernova of the model used by the livermore group @xcite .
the detection probability of a supernova at a certain distance is determined as a probability in which one simulated neutrino burst satisfies each criterion given in previous sections after basic data reduction . as shown in figure [ dprb ] , full detection probability is maintained out to around 100 kpc .
therefore , the upper limit at 90% c.l .
for the supernova explosion rate out to 100 kpc within which our galaxy , the lmc , and the smc may be found is determined to be 0.32 per year by combining the results from sk - i and sk - ii . while the probability for the burst search with lower energy thresholds goes down rapidly to almost 0.0 at 700 kpc
the distance to the andromeda galaxy the probability of the distant supernova search is still 0.075 at this distance , which demonstrates the benefit of conducting a long time - window search in addition to the usual burst search .
we have also performed , for the first time , a systematic search for neutrinos from neutronization bursts .
these could occur in isolation in the case of early black hole formation following a core collapse .
however there was no such signal observed in the data set with a total livetime of 2,381.3 days .
the authors gratefully acknowledge the cooperation of the kamioka mining and smelting company .
super - k has been built and operated from funds provided by the japanese ministry of education , culture , sports , science and technology , the u.s .
department of energy , and the u.s . national science foundation .
this work was partially supported by the korean research foundation ( bk21 ) , the korean ministry of science and technology , and the national science foundation of china .
99 abe , k. _ et al , , _ 2006 , astrophys .
j. * 652 * , 198 .
ahrens , j. _ et al , , _ 2002 , astropart .
* 16 * , 345 .
alekseev , e. n. _ et al . , _ 1987 , pisma zh . eksp .
fiz * 45 * , 461 .
[ jetp lett .
* 45 * , 589 ( 1987 ) ] alekseev , e. n. _ et al . , _ 2002 , zh .
fiz * 95 * , 10 .
[ 2002 , j. exp .
phys * 95 * , 10 ] ambrosio , m. _ et al . , _ 2004 ,
j. c * 37 * , 265 .
ando , s. _ et al . , _
2005 , phys .
lett . * 95 * , 171101 .
bionta , r. m. _ et al . , _ imb collaboration , 1987 , phys .
lett . * 58 * , 1494 .
dadykin , v. l. _ et al . , _ 1987 , pisma zh
fiz * 45 * , 464 .
[ 1987 , jetp lett . *
45 * , 593 ] dye , s. t. _ et al . , _ 1989 ,
* 62 * , 2069 .
hirata , k. s. _ et al . , _
1987 , phys .
lett . * 58 * , 1490 .
hosaka , j. _ et al .
, _ 2006 , phys .
d * 73 * , 112001 jegerlehner , b. _ et al . , _ 1996 ,
d * 54 * , 1194 .
langanke , k. _ et al . , _ 1996 ,
lett . * 76 * , 2629 lvd coll , 2003 , proc .
28th icrc h.e 2.3.9 , tokyo .
nakahata , m. _ et al . , _ 1999 ,
methods phys .
sect . a * 421 * , 113 .
raffelt , g. g. nucl .
suppl ) 2002 , * 110 * , 254 .
sumiyoshi , k. _ et al . , _
2005 , astrophys .
j. * 629 * , 922 .
sumiyoshi , k. _ et al .
, _ 2006 , phys . rev .
* 97 * , 91101 .
suzuki , y. 1993 , _ in proc.of the international symposium on neutrino astrophysics : frontiers of neutrino astrophysics _ , edited by y. suzuki and k. nakamura , ( universal academy press inc . ,
tokyo ) , number 5 in frontier science series , p.61 .
thompson , t. a. _ et al . , _ 2003 ,
j. * 592 * , 434 .
totani , t. _ et al .
_ , 1998 , phys .
* 80 * , 2039 .
takahashi , k. _ et al . , _
2001 , phys .
d * 64 * , 93004 .
ccccc + @xmath34 2events/1msec & 1 & 2.10 & 0 & 0.125 + @xmath34 2events/10msec & 19 & 19.1 & 0 & 1.25 + @xmath34 2events/100msec & 194 & 191 & 10 & 12.5 + + @xmath34 3events/1msec & 0 & 9.90@xmath48 & 0 & 1.65@xmath49 + @xmath34 3events/10msec & 0 & 9.78@xmath50 & 0 & 1.65@xmath51 + @xmath34 3events/100msec & 0 & 9.78@xmath52 & 0 & 1.65@xmath53 | the result of a search for neutrino bursts from supernova explosions using the super - kamiokande detector is reported .
super - kamiokande is sensitive to core - collapse supernova explosions via observation of their neutrino emissions .
the expected number of events comprising such a burst is @xmath0 and the average energy of the neutrinos is in few tens of mev range in the case of a core - collapse supernova explosion at the typical distance in our galaxy ( 10 kiloparsecs ) ; this large signal means that the detection efficiency anywhere within our galaxy and well past the magellanic clouds is 100% .
we examined a data set which was taken from may , 1996 to july , 2001 and from december , 2002 to october , 2005 corresponding to 2589.2 live days .
however , there is no evidence of such a supernova explosion during the data - taking period . the 90% c.l .
upper limit on the rate of core - collapse supernova explosions out to distances of 100 kiloparsecs is found to be 0.32 sn @xmath1 year@xmath2 . |
at present information on the characteristics of hadronic interactions in fragmentation region is still scarce or missing and experiments with ` roman pots ' are anticipated to improve the situation .
some of this information , in principle , could be obtained with the use of the data on cosmic ray ( cr ) muon and hadron spectra , provided primary spectra are known with high precision , but that is not the case .
the obvious obstacle here is that at high energies primary cosmic ray ( pcr ) fluxes , measured in direct experiments , themselves are functionals of various interaction parameters plus their accuracy is appreciably affected by additional systematic effects @xcite . in the series of our papers
@xcite we underlined , that these effects can lead to underestimation of light nuclei fluxes and thus can explain discrepancy between measured and calculated muon fluxes for @xmath0 gev .
preliminary data of atic-2 @xcite , covering the gap between magnetic spectrometer and emulsion chamber experiments , seems to be in concordance with our conclusions , but situation is more complicated in fact , as further consideration will show .
atic-2 slope of proton spectrum @xmath1 @xcite for primary energies below 10 tev is in remarkable contradiction with previously measured values @xmath2 and @xmath3 by runjob @xcite and jacee @xcite experiments correspondingly , but this discrepancy is removed by steepening in atic-2 proton spectrum at energies above 10 tev .
these new data were already exploited in extensive calculations of muon and hadron fluxes with number of interaction models at different atmospheric depths and zenith angles in @xcite where it was shown , that their use allows to get reasonable agreement with the most of the data under appropriate choice of hadronic interaction parameters .
in fact , it is not possible to obtain concordant conclusions on primary spectra or hadronic interactions parameters coming even from much smaller subset of the experimental data . in this paper
we demonstrate this on the basis of analysis of the data on muon flux at vertical direction and only one set of the data on hadron flux of eas - top @xcite .
first , we use the data of sea - level and underground experiments to obtain conclusions on behavior of muon spectrum at sea - level in the energy range 40 gev10 tev . further we analyze influence of uncertainties in the muon data and interaction parameters on properties of reconstructed primary proton fluxes . and
finally we show , that such different in approach and characteristics interaction models as sibyll 2.1 @xcite and nexus 3.97 @xcite can bring to hardly distinguishable predictions on muon and hadron fluxes .
depth - intensity relation , needed for reconstruction of sea - level muon spectrum , may be obtained via numerical solution of one - dimensional transport equation . in adjoint approach
this equation has the following form @xcite @xmath4 here @xmath5 is survival probability of muon with energy @xmath6 , being born at the distance @xmath7 from detector , @xmath8 total interaction cross - section , @xmath9 differential cross - sections for processes of ionization , bremsstrahlung , pair production and photonuclear interaction correspondingly , @xmath10 detector sensitivity function .
the numerical method , applied for solution of this equation @xcite , allows to avoid any approximations ( such as continuous losses one ) and to obtain muon intensities at large depths of matter with account of fluctuations in all muon interaction processes .
accuracy of our calculations of muon survival probabilities and intensities was thoroughly examined in @xcite and comparison with the results of monte - carlo codes mum @xcite and music @xcite is presented in fig .
[ fig : vsmum ] . anticipating further discussion of sea - level muon spectrum behavior it is necessary to note , that our calculations give upper estimate of muon flux at large depths in comparison with mum , because of use of @xmath111% lower muon energy losses @xcite . to describe the data on muon intensity underground and at sea - level , and to estimate influence of uncertainty in muon flux data on reconstruction of primary proton flux , we used two parameterizations in the simple form , proposed in work of bogdanova et al .
original fit for the vertical from @xcite @xmath12 provides good agreement with the data at sea - level ( fig .
[ fig : param ] ) , but leads to underestimation of the muon flux for the depths below 6 km w.e .
[ fig : strock ] ) . to match better underground data for the depths 26 km w.e .
we shall also apply modified fit with slightly @xmath13% increased intensity in multi - tev region @xmath14 as it is seen from fig .
[ fig : strock ] , for depths from 4 km w.e . up to 8 km w.e . , corresponding to @xmath15 tev median muon energies at sea level , use of this spectrum provides good agreement with the data of lvd @xcite , bno @xcite and frejus @xcite collaborations and leads to underestimation of data of macro @xcite and soudan @xcite experiments .
our consideration will touch muon energies only above 40 gev , to exclude different effects , not related to high - energy hadronic interactions features , such as geomagnetic effect , influence of uncertainties in low - energy interaction models , or even absorption of low energy muons in ground as in case of l3+c detector .
% variation of total muon energy losses .
standard rock ( muon spectrum at sea level from @xcite ) : dashed line present work , open circles music @xcite . ] ) , dashed line for muon spectrum ( [ eq : mbogdanova ] ) .
neutrino induced muon contribution is taken from @xcite . ]
average numbers of hadrons @xmath16 and muons @xmath17 with energies above @xmath18 in eas from primary nucleon of energy @xmath19 were obtained with the help of one - dimensional hybrid code conex @xcite in regime of cascade equations solution for interaction models qgsjet 01 @xcite , sibyll 2.1 @xcite , nexus 3.97 @xcite and qgsjet - ii-03 @xcite .
to get differential spectrum of hadrons ( or in the same way of muons ) for some energy @xmath20 the following simple formula had been used : @xmath21/2\delta e.\ ] ] here @xmath22 is the integral spectrum of hadrons @xmath23 for primary nucleon spectrum @xmath24 .
the interval width @xmath25 must be chosen to provide the difference between integral intensities in ( [ eq : diff ] ) to be much larger then calculations error .
test computations for @xmath26 and observation levels 820 and 1030 g/@xmath27 for energy @xmath28 tev brought to differential flux values lying within @xmath29 from each other .
change of upper integration limit @xmath30 in formula for @xmath22 from @xmath31 to @xmath32 gives less then 1% increase of differential flux . and ,
the last , increase of number of primary energy bins in @xmath33 from 10 to 20 per order introduces @xmath34 variation to differential flux value . the listed error
sources partially compensate each other and the total error of our calculations does not exceed 3% .
the calculation were performed for the set of energies , coinciding with the set from eas - top experiment paper @xcite .
as a basic model of pcr nuclei spectra the parameterizations from @xcite were chosen .
nuclei with @xmath35 were treated in the framework of the superposition model , high accuracy of this approach is well known and was checked by our calculations with the use of conex both for muons and hadrons once again .
comparison of the calculated muon fluxes with the experimental data , presented in fig .
[ fig : param ] , reveal familiar picture of high energy muon deficit .
the reasons of its appearance were considered in our previous papers @xcite and they still hold true regardless of the fact , that three more interaction models were included in our analysis .
all interaction codes , except qgsjet 01 , satisfactory describe data on muon flux only up to @xmath36100 gev and then one by one fail to do it . accounting that such muon energies correspond to primary energies above 1 tev , studied with balloon(satellite)-borne emulsion chambers , we related muon deficit to underestimation of primary light nuclei fluxes , taking place in these experiments @xcite .
unfortunately , disagreement between the models in the muon fluxes also appears at energies around 100 gev , thus making impossible precise reconstruction of primary nucleon spectrum for @xmath37 tev .
in fact , in such conditions there are no reasons to rule out any of the models , except qgsjet 01 , which leads to remarkable disagreement with the experiment even in the range of reliable magnetic spectrometers data on pcr and muon spectra . to find out , why the models differ in the predicted muon fluxes let us consider quite characteristic energy 1.29 tev , where discrepancies between the models reach appreciable values and the data on muons from underground installations are yet quite reliable .
contributions of primary protons to the differential flux of muons of the given energy , presented in fig .
[ fig : inclusmu ] show , that spread in muon fluxes between the interaction models is entirely due to uncertainties in the description of @xmath38-spectra in fragmentation region @xmath39 . since inclusive muon flux is sensitive nearly only to the characteristics of the very first primary particle interaction , hence , the harder these spectra are in the particular model , the larger muon intensity its use leads to .
for the lower values of @xmath40 , i.e. for @xmath41 tev , all the models give practically the same muon yields .
as noted above , in view of uncertain situation with primary spectra for @xmath37 tev , one can not give preference to any of the models in comparison with the others .
if to demand the minimal disagreement with the direct measurements data on pcr spectra , then obviously sibyll 2.1 satisfies this requirement the best , or , on the other hand , one could say that it provides the most acceptable description of @xmath38 production spectra in @xmath42-air collisions in fragmentation region .
we shall discuss this affirmation in detail below .
from the previous consideration it is clear , that reconstructed fluxes of protons shall be higher than measured in emulsion chamber experiments , but can be comparable with recently obtained data of atic-2 group @xcite .
we performed the reconstruction simply by picking up of appropriate primary proton flux parameters in order to minimize deviation of the obtained muon fluxes from the parameterizations ( [ eq : bogdanova ] ) and ( [ eq : mbogdanova ] ) .
, to muon flux parameterizations ( [ eq : bogdanova ] ) and ( [ eq : mbogdanova]).,title="fig:",scaledwidth=48.0% ] , to muon flux parameterizations ( [ eq : bogdanova ] ) and ( [ eq : mbogdanova]).,title="fig:",scaledwidth=48.0% ] first , let us consider an attempt to minimize maximal deviation from spectra ( [ eq : bogdanova]),([eq : mbogdanova ] ) for @xmath43 gev10 tev in assumption , that primary proton spectrum can be described by single power law function @xmath44 in the entire energy range 100 gev500 tev . in fig . [
fig : primp ] and in table [ tab : pp ] one can find the results of this reconstruction , upper and lower lines for each muon spectrum parametrizations in the fig . [
fig : primp ] correspond to low and high helium flux fits @xcite .
as expected , the obtained spectra are flatter , than measured by runjob and jacee groups , but , except for qgsjet 01 model , agree well with atic-2 results @xcite .
figure [ fig : fitratio ] shows , that in this case it is possible to achieve agreement with the fitted muon spectrum within 10% , but only sibyll 2.1 reproduces its shape correctly , with other models it is not possible to get right muon spectrum slope variation .
we should also note , that this behavior turned out to be insensitive to the choice of helium flux parametrization . as a result
, there is a dip in the ratio of the muon flux , obtained from fitting of proton spectrum , to the parameterizations ( [ eq : bogdanova]),([eq : mbogdanova ] ) for energies around 1 tev , where the underground data were already underestimated ( overestimation of muon flux at higher energies does not compensate this effect completely ) , and growth in small energies range , which brings to contradiction with low energy @xmath45 gev data .
parametrization ( [ eq : mbogdanova ] ) was also projected to make muon spectrum slope variation less sharp , but this lead to problems with its fitting in @xmath46 gev energy range , as it is seen from fig .
[ fig : fitratio ] ( right panel ) for sibyll 2.1 and nexus 3.97 models .
the problem with muon spectrum shape matching is better illustrated by upper left panel of fig .
[ fig : primpbr ] , where it is shown , that correct reproduction of muon spectrum for energies below 1 tev with single power law proton spectrum leads to appreciable overestimation of muon flux at higher energies .
there are three possible explanations or solutions of this problem .
first , the discrepancy can be completely removed by choice of appropriate interaction parameters , e.g. similar to those in sibyll 2.1 .
another argument , which can be given is that the data on muon flux for energies above 1 tev are not so definite to claim their inconsistency with the calculations , but it does not look well supported by underground data ( see fig . [
fig : strock ] and calculations @xcite ) . and
the last possibility is to assume , that primary proton spectrum is not monotonous and either has sharp break or slowly changing exponent @xmath47 .
let us consider the latter assumption , which finds experimental @xcite and theoretical @xcite justifications , in more detail .
the results for the simple case with break ( fig .
[ fig : primpbr ] ) , which allows to achieve correct description of muon spectrum shape with right asymptotic and deviation in flux value @xmath48% , show , that small difference between spectra ( [ eq : bogdanova ] ) and ( [ eq : mbogdanova ] ) results not only in different proton intensities , but also in break positions .
the latter lies for parametrization ( [ eq : mbogdanova ] ) in the primary energy range 1015 tev , the change in power index reaches appreciable values up to @xmath49 for qgsjet 01 and qgsjet ii models ( table [ tab : pp ] ) .
proton spectra , obtained from muon flux ( [ eq : mbogdanova ] ) , with qgsjet ii and nexus 3.97 models are in the best agreement with atic-2 data , while sibyll 2.1 provides intermediate between atic-2 and emulsion chambers experiments slope value .
spectra , reconstructed from parametrization ( [ eq : bogdanova ] ) , have breaks at 36 tev and in case of qgsjet ii proton flux poorly agrees with experiments at primary energies around 100 gev . evidently , the latter problems are explained by too low , in comparison with underground data , muon flux and this parametrization is considered here mostly for estimation of sensitivity of primary spectrum features to the choice of reference muon flux .
it is necessary to note , that due to low sensitivity of differential muon flux to helium and heavier groups of primary nuclei it is impossible to derive any conclusions on presence of the break in these pcr components .
for illustration let us consider example of calculations for qgsjet 01 and high helium flux , where the break in proton spectrum is positioned at @xmath50 tev and change of power index is equal to 0.14 ( see table [ tab : pp ] ) .
introduction of rigidity - dependent break in he spectrum at @xmath51 per nucleus of the same value @xmath52 gives remarkable discrepancy between calculated muon spectrum and parameterization ( [ eq : mbogdanova ] ) only for energies above 7 tev , which reaches 10% at 20 tev . to correct this asymptotic behavior it suffices to reduce @xmath53 to 0.11 simultaneously for protons and helium without change of the break position , and
thus we get proton spectrum lying well within corridor between parameterizations for high and low helium fits , shown in fig .
[ fig : primpbr ] .
hence , this corridor covers all possible cases of he flux behavior ( with or without break ) , provided the helium flux stays within limits , given in @xcite . summarizing we can say , that primary proton spectrum shape turns out to be sensitive to the choice of interaction model and allows presence of break at 1015 tev with @xmath54 up to 0.15 , which can be slightly softened though , if to allow presence of the same break in other pcr components spectra .
lccccccc + & & + model & & & break & + & @xmath55 & @xmath56 & @xmath55 & @xmath56 & ( tev ) & @xmath55 & @xmath56 + qgsjet 01 & 1.42 & 2.675 & 0.93 & 2.620 & 15 & 3.25 & 2.750 + qgsjet ii & 1.04 & 2.650 & 0.69 & 2.595 & 10 & 2.39 & 2.730 + nexus 3.97 & 1.22 & 2.680 & 1.01 & 2.655 & 13 & 1.78 & 2.715 + sibyll 2.1 & 1.29 & 2.695 & & & & & + + & & + model & & & break & + &
@xmath55 & @xmath56 & @xmath55 & @xmath56 & ( tev ) & @xmath55 & @xmath56 + qgsjet 01 & 1.45 & 2.690 & 0.95 & 2.635 & 15 & 3.65 & 2.775 + qgsjet ii & 1.08 & 2.670 & 0.69 & 2.610 & 14 & 3.18 & 2.770 + nexus 3.97 & 1.25 & 2.700 & 1.01 & 2.670 & 12 & 2.14 & 2.750 + sibyll 2.1 & 1.37 & 2.720 & & & & & +
comparison of our calculations of hadron spectrum for the primary spectra from @xcite ( high helium flux ) with the recent measurements , performed by eas - top collaboration @xcite , is presented in fig .
[ fig : eastop ] ( left panel ) . first , let us note the following facts .
below 100 gev all calculated spectra have breaks , caused by non - perfect matching of low - energy interaction model gheisha to the high - energy models .
shape of the measured hadron spectra also breaks at energies above 4 tev and the data become less definite , thus in the forthcoming analysis we are going to consider the data for energies from 129 gev to 4 tev . for these energies
qgsjet 01 , qgsjet ii and sibyll 2.1 quite reasonably reproduce the shape of the measured hadron spectrum , nexus 3.97 leads to spectrum with almost constant power index .
one can see , that the most consistent description of the data for specified energies provide qgsjet 01 and sibyll 2.1 .
in contrast with the muons there are no energy range , where the models agree on the hadron fluxes and the reasons of this disagreement are not as simply to point out as in the case with muons .
the most important characteristics in this analysis are total inelastic cross section , determining chances of primary particle to survive , shapes of inclusive spectra @xmath57 , @xmath58 , @xmath59 in the very forward region , responsible for substantial process of leading particles production ( see fig . [
fig : hinclus ] for the listed spectra ) .
let us briefly outline the major conclusions , which one may come to in the given situation .
nexus 3.97 gives the lowest fluxes as of hadrons in total , so of nucleons and mesons ( fig .
[ fig : mesnuc_sp ] ) , and this happens in spite of the lowest inelastic cross - section values .
inclusive spectrum @xmath60 immediately helps to figure out , that incident protons in nexus 3.97 have comparably low chances to save most of their energy in collision and this leads to such low nucleon flux , the same can be said about meson flux and production of pions by pions .
similarly , from comparison of the inclusive spectra , it can be easily understood , why qgsjet ii gives the highest hadron flux .
note , that sibyll 2.1 concedes to qgsjet ii in hadron intensity mostly because of less effective production of leading neutrons in @xmath42-air collisions and due to the larger total interaction cross - section .
thus , from analysis of the data on hadron flux it is difficult to imply any strict constraints on inclusive spectra shapes , since mechanism of hadron spectrum formation is more sophisticated than that in the case of muon spectrum .
sibyll 2.1 and qgsjet 01 display quite different behaviors of the relevant inclusive spectra and total interaction cross - sections , but both models almost equally succeed in description of the eas - top data ( i.e. produce close hadron fluxes ) .
the given standard approach to analysis of situation , in fact , is of little sense , since it is based on assumption about validity of primary spectra in form of fits from @xcite , which was called into question in our previous discussion . in this case
it is logical to analyze interaction models self - consistency , i.e. their ability to give correct estimates of several cr components at once .
provided we know behavior of primary proton spectra for every model , required to match the data on muon flux , we may check how these proton spectra agree with the data on hadrons . in fig .
[ fig : eastop ] ( right panel ) we give hadron intensities , calculated for primary proton spectra with breaks from fig .
[ fig : primpbr ] ( for sibyll 2.1 see fig .
[ fig : primp ] ) , corresponding to muon spectrum parametrization ( [ eq : mbogdanova ] ) . after increase of primary nucleon flux , dictated by the data on muons , one can see ,
that the best agreement with eas - top measurement provide nexus 3.97 and sibyll 2.1
. it would be interesting to note that two models with different philosophies and inclusive spectra give the most self - consistent results on muons and hadrons , but , of course , this conclusion must be taken with much care , since it is based on the single set of data and we have only indirect indications on the accuracy of this set , e.g. such as agreement of primary proton fluxes , obtained by eas - top and kascade teams ( the latter is derived from flux of unaccompanied hadrons @xcite ) .
if to try to perform the same analysis with the variety of the data , obtained at different atmospheric altitudes and zenith angles , no consistent notions of such kind will be obtained as can be easily seen from calculations , presented in @xcite .
the progress in cr and high - energy physics , achieved during last 1015 years allowed to turn from statements about satisfactory ( qualitative ) concordance between different kinds of data to investigation of more fine effects . as an example
, we managed to show that reconstructed from the data on vertical muon flux primary proton spectra have not only expected interaction model dependent intensities , but also model - dependent shapes .
it is demonstrated , that application of qgsjet 01 , qgsjet ii and nexus 3.97 models brings to proton spectrum with break at 1015 tev and power index @xmath47 before break close to that , measured in atic-2 experiment .
nevertheless one can see , that absolute proton flux for qgsjet 01 is hardly compatible with any data of direct experiments , and the break for all these three models is more moderate , compared to what can be inferred from atic-2 data , which though become less definite right in the break region . on the other hand sibyll 2.1 allows to reproduce shape of the muon spectrum with single power law proton spectrum , which is in reasonable agreement with both emulsion chamber and atic-2 data within experimental errors .
further improvement of our understanding of the situation , which is of primary astrophysical interest , can be achieved via experimental study as of muon cr component characteristics with water and ice neutrino telescopes and so of inclusive spectra @xmath61 in fragmentation region .
reduction of uncertainties in the latter component with the use of the data on primary spectra , hadron and muon components , does not look possible , because of 1 ) poor correlation between muon and hadron production mechanisms , 2 ) ambiguity of existing cr experimental data and 3 ) possibility to realize self - consistent description of the data on muons and hadrons with the models , having remarkably differing inclusive spectra and underlying philosophies .
authors greatly acknowledge conex team and personally tanguy pierog for their kind permission to use conex cross - section tables and for technical support .
we are grateful to anonymous referee for constructive remarks , which helped us to improve the manuscript .
+ this work was supported in part by rfbr grant 07 - 02 - 01154-a . | it is shown , that primary proton spectrum , reconstructed from sea - level and underground data on muon spectrum with the use of qgsjet 01 , qgsjet ii , nexus 3.97 and sibyll 2.1 interaction models , demonstrates not only model - dependent intensity , but also model - dependent form . for correct reproduction of muon spectrum shape primary proton flux should have non - constant power index for all
considered models , except sibyll 2.1 , with break at energies around 1015 tev and value of exponent before break close to that obtained in atic-2 experiment . to validate presence of this break understanding of inclusive spectra behavior in fragmentation region in p - air collisions should be improved , but we show , that it is impossible to do on the basis of the existing experimental data on primary nuclei , atmospheric muon and hadron fluxes . |
recent observation of resonances near the open charm threshold which decay into a specific charmonium state , @xmath1 or @xmath2 , and two or even one pion with no other observed conspicuous decay channels@xcite suggests@xcite that these resonances in fact contain the specific charmonium state ` coated ' by an excited light - hadron matter .
it has been further argued@xcite that such hadro - charmonium systems are likely to arise due to binding of charmonium inside sufficiently excited light resonances through a qcd analog of the van der waals interaction . in this paper
we investigate in some detail the interaction of heavy quarkonium with excited light mesons within the simplest soft wall model of holographic qcd@xcite which provides linear behavior of regge trajectories@xcite and a reasonable behavior of a heavy - quark potential@xcite . in this approach
the mesons can be considered as modes of the flavor gauge field in 5d theory ( vector mesons ) or a wilson line of this gauge field along the fifth coordinate ( @xmath3-meson ) .
large - spin mesons can be considered within this approach as well .
we also have to introduce heavy quarkonium in this framework .
it is clear that the corresponding heavy degrees of freedom are localized near the boundary at @xmath4 at scales of order @xmath5 , where @xmath6 is the mass of the heavy quark . in the leading approximation we shall substitute the heavy meson by the point - like source .
the essential local operator on the boundary can be deduced from the physical arguments and is approximated by the trace of the stress tensor . since this operator interacts with the dilaton we have a point - like dilaton source in the 5d theory . on the other hand the light exited meson with large spin
is mainly localized on the ir wall scale@xcite .
although it is better described by a long open string , this approximation is also reliable .
thus the appropriate configuration consists of a local source at the boundary @xmath4 and extended meson at the @xmath7 interacting via dilaton exchange .
it is shown that the interaction is strong enough to yield the bound state at a sufficiently large spin @xmath8 of the light meson .
another feature which has to be explained is the phenomenologically apparent suppression of the decays of hadro - quarkonium into states with pairs of heavy mesons with open flavor . in the discussed picture
such decays would correspond to dissociation of heavy quarkonium due to forces from light degrees of freedom .
we argue that in the limit of large heavy quark mass @xmath6 such dissociation can be described by semiclassical tunneling and is suppressed by the factor @xmath9 . within the holographic approach
this process is a generalization of the picture discussed in @xcite .
clearly , the suppression of the dissociative decays is in qualitative agreement with the available observations .
the subsequent material in the paper is organized as follows . in sec.2
a description of the interaction of heavy quarkonium with soft gluonic field is recapitulated and formulated in terms of the trace of the stress tensor in low - energy qcd . in sec.3
the interaction of the heavy quarkonium with light meson is described in terms of the soft wall holographic model and in sec.4 it is shown that this interaction necessarily results in existence of bound states for sufficiently high orbital excitation of the light meson . in sec.5
we consider the dissociation process for decays of the bound states into heavy meson pairs with open flavor and argue that such process generally carries a semiclassical suppression at large mass of the heavy quark .
finally , in sec.6 we summarize our conclusions .
due to the binding between heavy quark and antiquark , the states of heavy quarkonium are relatively compact in the standard hadronic scale set by @xmath10 . for this reason
the interaction of quarkonium with gluonic fields inside a light meson can be considered in terms of multipole expansion@xcite .
the leading nonrelativistic ( in the heavy quarks ) contribution to the van der waals type interaction then arises in the second order in the chromoelectric dipole term , and can be described by the effective hamiltonian@xcite h_eff= -1 2 ^(q |q ) e_i^a e_i^a , [ hdiag ] where @xmath11 is the chromoelectric field and @xmath12 is the chromo - polarizability , depending on particular state of the @xmath13 system , in complete analogy with the description in terms of polarizability of the interaction of atoms with long - wave electric field .
the values of the chromo - polarizability for charmonium and bottomonium levels below the open flavor threshold are all real and positive@xcite .
the numerical values are so far known only for the off - diagonal polarizability , describing the observed @xmath14 transitions in charmonium and bottomonium .
the ` reference ' values are@xcite @xmath15gev@xmath16 in charmonium and @xmath17gev@xmath16 in bottomonium .
it is fully expected@xcite that proper diagonal values of the chromo - polarizability in both systems are larger than these ` reference ' values , especially for excited states .
the quoted values correspond to the normalization convention for the gluonic field strength with the qcd coupling @xmath18 included in the field strength @xmath19 . in what follows
we consider the interaction ( [ hdiag ] ) with the operator @xmath20 replaced by @xmath21 , with @xmath22 being the chromomagnetic field . in doing
so we consider attraction of quarkonium into gluonic medium which is weaker by the contribution of the manifestly sign - definite term @xmath23 than the actual interaction described by the hamiltonian ( [ hdiag ] ) .
this clearly leads to a conservative treatment of the problem of existence of bound states@xcite .
furthermore , the operator @xmath24 is related to the trace of the stress tensor @xmath25 in qcd with ( three ) light quarks in the chiral limit by the well known anomaly relation : _
^=(g^2 ) 4 g^4 ( f_^a)^2 - 9 32 ^2 ( f_^a)^2 , [ anom ] where we replace the qcd beta function by its first one - loop term for definiteness of subsequent numerical estimates , while our general conclusion about existence of bound states does not depend on this replacement
. therefore the form of the van der waals interaction of heavy quarkonium with light hadronic matter that we consider here is h_w =- c _ ^ [ hw ] with @xmath26 $ ] .
we consider the limit of large @xmath6 so that the kinetic energy of the motion of heavy quarkonium is totally neglected , and we choose the position of quarkonium ( in the three dimensional space ) at the origin @xmath27 . in a holographic model heavy quarkonium
is localized at @xmath4 , so that the interaction ( [ hw ] ) corresponds to a localized at @xmath28 source of dilaton .
in a holographic soft wall model@xcite an orbitally excited light meson with spin @xmath8 is described by normalizable solutions @xmath29 of the five - dimensional equation : ( _ ^- ^2 z^2 + z^2 + 2 s -2 + s^2 - 1/4 z^2 ) ( z , x ) = 0 .
[ 5deq ] upon substitution of a four - dimensional plane wave @xmath30 this results in the eigenvalue problem for the masses @xmath31 of the mesons : ( - d^2 z^2 + z^2 + 2 s -2 + s^2 - 1/4 z^2 ) _
n(z ) = m_n^2 _ n(z ) , [ zeig ] where @xmath32 is the excitation number at a given @xmath8 .
the spectrum of @xmath33 is then directly found from the obvious relation of this problem to a two - dimensional harmonic oscillator , and is given by@xcite m_n^2 = 4(n+s ) , [ rspec ] thus reproducing linear regge trajectories .
it can be mentioned that here dimensionless units are used , corresponding to the coefficient of the soft wall " term @xmath34 equal to one . in normal units the energy scale is set by a coefficient @xmath35 proportional to the string tension . in the presence of a static source of the dilaton field located at @xmath36 the equation ( [ 5deq ] )
is generally no longer translationally invariant in @xmath37 , and the problem becomes a ( 3 + 1 ) dimensional eigenvalue problem for the energy @xmath38 : _
n(z , x ) = _ n^2 _ n(z , x ) , [ 5deqm ] where @xmath39 is the three - dimensional laplacian and @xmath40 is the potential resulting from the dilaton propagation from the source and interaction with the light - hadron string : v(z , x ) = g(z ) d(z , x ) , [ potgen ] with @xmath41 being the dilaton propgator from the boundary ( @xmath4 , @xmath42 ) to the bulk ( @xmath43 ) , integrated over time ( static source ) , @xmath44 is the strength of the source determined by eq.([hw ] ) , and @xmath45 is the vertex for the dilaton - string interaction .
the @xmath46 dependence of the latter vertex arises due to the assumed in eq.([5deq ] ) @xmath46 dependent conformal symmetry breaking term @xmath34 .
we will show that for the purpose of present discussion there is in fact no need to determine each of the factors in eq.([potgen ] ) separately , and that the potential @xmath40 can be found using the normalization condition for the stress tensor and the fact that the propagation in the bulk of the ( conformal dimension @xmath47 ) dilaton field from a static source is described by the operator -_x - ^2 z^2 + z^2 + 2 + 15 4 z^2 .
[ dilop ] the @xmath46 part of this operator corresponds to the motion of two - dimensional harmonic oscillator with angular momentum @xmath48 , and the spatial ( @xmath49 ) part corresponds to a free motion .
thus one can readily find the propagator as an integral over the proper time @xmath50 of the corresponding evolution kernel@xcite and write v(z , x ) = f(z ) _ 0^ d ( 1 2 ) ^3/2 ( - x^2 2 ) e^- ^3 ( - z^2 2 ) .
[ pot1 ] the function @xmath51 here includes the vertex factor @xmath45 and the source strength @xmath44 from eq.([potgen ] ) as well as normalization and @xmath46-dependent metric factors in the propagator @xmath52 .
the function @xmath51 can be found by comparing the matrix elements of the hamiltonian ( [ hw ] ) and the potential ( [ pot1 ] ) over the plane - wave in @xmath53 solutions of the equation ( [ 5deq ] ) .
indeed for _ any _ eigenstate in eq.([zeig ] ) the normalization of the trace of the stress tensor requires _
n(z , x ) | _ ^ | _ n(z , x ) = 2 m_n^2 _ n(z , x ) | _ n(z , x ) , [ norm ] which determines the average of the interaction ( [ hw ] ) over those states . on the other hand ,
the average of the potential over the same states is given by its integral over @xmath37 : _
n(z , x ) | v(z , x ) | _ n(z , x ) = _ n(z ) | v(z , x ) d^3 x | _ n(z ) = _ n(z ) | f(z ) 4 e^- z^2 / 2 z^4 | _
n(z ) , [ avpot ] where in the latter expression the result of explicit calculation of the integral over @xmath37 of the expression from eq.([pot1 ] ) is used .
the latter average corresponds to the condition ( [ norm ] ) for _ any _ eigenstate equation ( [ zeig ] ) this is the virial theorem for harmonic oscillator : the average potential energy is one half of the total . ] if the averaged in the last expression in eq.([avpot ] ) function of @xmath46 is @xmath54 .
taking into account also the factor @xmath55 in the hamiltonian ( [ hw ] ) one can finally write v(z , x ) = -c ( z^2+s-1 ) z^4 _ 0^ d ( 1 2 ) ^3/2 ( - x^2 2 ) e^- ^3 ( - z^2 2 e^- ) [ potf ] with @xmath56 being the dimensionless value of @xmath57 .
the problem of existence of bound states between the heavy quarkonium and light mesons is thus reduced to the problem of existence of localized in @xmath37 solutions of the eigenvalue equation ( [ 5deqm ] ) .
general physical arguments suggest@xcite that such solutions exist at a given @xmath58 and sufficiently large excitation in @xmath32 or / and @xmath8 . in practice a full analysis in @xmath46 and @xmath53 of states excited in @xmath32
is significantly complicated by the presence of multitude of lower energy states with lower @xmath32 and nonzero spatial momentum @xmath59 .
for this reason we analyze here in some detail the excitation of the light meson in the spin @xmath8 while keeping @xmath60 .
the treatment of such excitations is greatly simplified due to the fact that bound state in question is the ground state in the ` radial ' equation ( [ 5deqm ] ) at a given @xmath8 , so that the corresponding eigenfunction @xmath61 has no zeros .
we thus use a straightforward variational procedure to establish existence of bound states at any nonzero @xmath58 , provided that @xmath8 is sufficiently large , and then find numerically the onset of the binding as a function of the coefficient @xmath58 at few low values of @xmath8 . for a variational treatment of the binding problem
we choose the _ ansatz _ for the eigenfunction in a factorized form _ 0(z , x)= z^s+1 2 e^- z^2/2 ( x ) , [ ansatz ] where the @xmath46 dependence is that of the ground - state wave function of the @xmath46 equation ( [ zeig ] ) at a given @xmath8 . substituting this variational function in eq.([5deqm ] )
one finds that the upper bound for the energy of the ground state of the system is then given by the eigenvalue @xmath38 found from the @xmath53 equation ( x ) = ( ^2 - 4s ) ( x ) , [ varxi ] with the effective potential @xmath62 considering large @xmath8 one can readily see that the factor @xmath63 effectively cuts off the contribution of values of @xmath50 above @xmath64 , so that the range of the potential behaves as @xmath65 . on the other hand
the integral over @xmath37 of the potential is proportional to @xmath66 , which is weaker than the overal factor @xmath67 in front of and in the integral in eq.([uofx ] ) .
therefore a bound state exists at sufficiently large @xmath8 for any nonzero value of the coupling strength @xmath58 , which is necessarily positive , due to the positivity of the chromo - polarizability .
the results of such variational treatment of the problem of binding are confirmed by an explicit numerical solution of the two dimensional ( in @xmath46 and @xmath68 ) schrdinger equation with the potential ( [ potf ] ) . for a fixed @xmath8
we find that a localized in @xmath53 ground state solution appears starting from a critical ( @xmath8 dependent ) value of @xmath58 .
the behavior of the ` binding energy ' @xmath69 as a function of @xmath58 is shown in fig.1 for few values of @xmath8 .
one can readily see that the critical value of @xmath58 decreases with @xmath8 .
phenomenological applicability of our conclusions to actual charmonium or bottomonium critically depends on the values of the chromo - polarizability for specific quarkonium resonances . using the estimate for the scale factor @xmath35 as@xcite @xmath70gev@xmath71 we find for the ` reference ' value of @xmath72 : @xmath73 . according to the data presented in fig.1 such value of @xmath58
would correspond to the appearance of binding at @xmath74 .
however , as already mentioned , the diagonal chromo - polarizability of the @xmath2 resonance is likely to be somewhat larger , so that bound hadro - charmonium states may exist already at @xmath75 .
the notion of hadro - quarkonium necessarily assumes that the heavy quark and antiquark stay bound together as quarkonium inside the host light - matter resonance . in other words the forces from the light quarks
do not split the heavy quarkonium , which would result in the decay of the whole system into final states with pairs of heavy flavored mesons , which decays are conspicuously suppressed@xcite for the discussed @xmath76 and @xmath77 resonances .
the problem of dissociation of the heavy quarkonium can in fact be treated , to an extent , in the limit of large mass @xmath6 , in a semiclassical approximation .
generally the problem can be formulated as that of a reconnection of strings in terms of a holographic correspondence , using , instead of the braneless soft wall model , the underlying scheme with the flavor branes localized in the radial coordinate ( for a recent review see e.g. ref.@xcite ) .
the position of the flavor brane in the radial direction is fixed by the quark mass of the corresponding flavor .
the mesons in this approach correspond to the open strings with ends on the flavor branes .
the ends can be on the same flavor brane , in which case the meson involves quarks of one flavor , or on the different flavor branes which corresponds to a meson built from different types of quarks .
any such model involves a horizon in the radial coordinate which reflects the emergent qcd scale in the holographic picture .
the flavor brane corresponding to the light meson is close to the horizon while the brane corresponding to the heavy flavor is far from it .
a bound state of hadro - quarkonium thus involves two strings .
one string with the ends on the light - flavor brane(s ) and the other on the heavy - flavor one .
the decay into states with open heavy flavor mesons corresponds to the reconnection of the strings into the configuration with two open strings both connecting the heavy - flavor brane with the light - flavor one(s ) .
this process is somewhat similar to the previously considered decay of a single excited meson@xcite via a reconnection of open string ends , which has been interpreted as a schwinger type pair production and treated semiclassically .
the discussed here reconnection of the strings between light and heavy flavor branes can similarly be traced to an _ induced _ production of heavy quark pair near the threshold , which approach will be presented elsewhere , and which can be useful in a more elaborate than here treatment . the leading at large @xmath6 behavior of the probability of heavy quarkonium dissociation can be evaluated by a simple nonrelativistic quantum - mechanical consideration , using a potential model for quarkonium combined with a stringy picture at larger separation between quarks .
indeed a potential @xmath78 between heavy quark and antiquark in quarkonium generically is assumed to be of the form@xcite :
v_q(r ) = v_c(r ) + r , [ potmod ] where the linear term is associated with the stringy behavior at long distances , while the term @xmath79 is the short - distance ` perturbative ' part , e.g. in the one - gluon exchange picture this is a coulomb - like potential @xmath80 . in the hadro - quarkonium configuration
the total energy of the string is @xmath81 , where @xmath82 is the distance between the heavy quark and the antiquark , and @xmath83 is the length of the light - meson string . when the string reconnects between heavy and light quarks , there is no string at all between the heavy quark and antiquark , so that the minimal energy of the reconnected configuration is @xmath84 .
therefore the dependence of the minimal ( over the relative orientations ) energy on @xmath82 becomes in fact given by v_hq(r ) = v_c(r ) -
r , [ vhq ] and this effective potential presents a barrier , tunneling through which results in dissociation of the heavy quarkonium .
the tunneling rate then can be estimated as ( -2 critically depends on the presently unknown position of the energy of the quarkonium below the top of the barrier . in a generic case of this energy gap
being of order @xmath10 this coefficient is parametrically of order one , and the rate is strongly suppressed for heavy quarks .
to summarize . we have considered the van der waals type interaction between a heavy quarkonium and an excited light meson within a soft wall holographic model of qcd . in this model the heavy quarkonium acts as a source of the dilaton field localized at the boundary .
the interaction of the light matter with the dilaton field is then described by a 5d equation with the potential derived from the dilaton propagation from the boundary to the bulk .
we have shown by a variational estimate that such interaction necessarily results in bound hadro - quarkonium states localized near the source at sufficiently high spin excitation of the light meson .
we have also found by a numerical calculation that bound states arise at low values of the spin already for moderate values of the strength of the source , which values are close to those expected for charmonium phenomenology .
furthermore , it is argued that the hadro - quarkonium resonances are metastable with respect to dissociation into final states with open - flavor heavy mesons in the limit of large heavy quark mass .
the behavior of the hadro - quarkonium states argued in the present paper agrees well with the observed properties of some of the y and z resonances near the open charm threshold .
the work of s.d . and m.b.v .
is supported in part by the doe grant de - fg02 - 94er40823 .
the work of a.g . is supported in part by the grants rfbr-06 - 02 - 17382 , intas-1000008 - 7865 , pics-07 - 0292165 .
a.g . acknowledges the warm hospitality at the william i. fine theoretical physics institute where this work was done . c. z. yuan _ et al . _
[ belle collaboration ] , phys .
lett . * 99 * , 182004 ( 2007 ) ; g. pakhlova _ et al . _ [ belle collaboration ] , phys .
d * 77 * , 011103 ( 2008 ) ; b. aubert _ et al .
_ [ babar collaboration ] , arxiv:0710.1371 [ hep - ex ] .
j. erlich , e. katz , d. t. son and m. a. stephanov , `` qcd and a holographic model of hadrons , '' phys .
lett . * 95 * , 261602 ( 2005 ) [ arxiv : hep - ph/0501128 ] .
a. karch , e. katz , d. t. son and m. a. stephanov , `` linear confinement and ads / qcd , '' phys . rev .
d * 74 * , 015005 ( 2006 ) [ arxiv : hep - ph/0602229 ] .
r. r. metsaev , arxiv : hep - th/0002008 .
o. andreev and v. i. zakharov , `` heavy - quark potentials and ads / qcd , '' phys .
d * 74 * , 025023 ( 2006 ) [ arxiv : hep - ph/0604204 ] .
m. kruczenski , l. a. p. zayas , j. sonnenschein and d. vaman , `` regge trajectories for mesons in the holographic dual of large - n(c ) qcd , '' jhep * 0506 * , 046 ( 2005 ) [ arxiv : hep - th/0410035 ] .
k. peeters , j. sonnenschein and m. zamaklar , `` holographic decays of large - spin mesons , '' jhep * 0602 * , 009 ( 2006 ) [ arxiv : hep - th/0511044 ] . k. gottfried , phys . rev
. lett . * 40 * , 598 ( 1978 ) . | we consider the recently suggested model for some resonances near the open charm threshold as bound states of charmonium inside excited light mesons . it is argued in the soft - wall holographic model of qcd that such states of heavy quarkonium necessarily exist at sufficiently large spin of the light meson .
the bound state is provided by the dilaton exchange through the 5d bulk .
we also argue that the decay of such bound systems into mesons with open heavy flavors due to splitting of the heavy quarkonium can be treated as semiclassical tunneling and is suppressed .
this behavior is in agreement with the known relative suppression of the decay of the discussed charmonium - like resonances into channels with @xmath0 mesons .
* william i. fine theoretical physics institute + university of minnesota + * itep - th-18/08 + ftpi - minn-08/12 + umn - th-2643/08 + april 2008 + * holographic hadro - quarkonium + * school of physics and astronomy , university of minnesota , + minneapolis , mn 55455 , usa , + * a. gorsky + * institute of theoretical and experimental physics , moscow , 117218 , russia + and + * m.b .
voloshin + * william i. fine theoretical physics institute , university of minnesota , + minneapolis , mn 55455 , usa + and + institute of theoretical and experimental physics , moscow , 117218 , russia + |
while the first states containing the @xmath1 quark , the @xmath2 resonances , were discovered in hadronic interactions at fermilab @xcite , the study of @xmath0 mesons for many years has been largely the province of @xmath3 colliders . with the reconstruction of large numbers of exclusive @xmath0 decays by the cdf collaboration @xcite , that situation now has the potential to change .
the present talk outlines some of the ways in which hadron colliders can exploit their innate advantage of large @xmath0 production rates to provide information complementary and in many cases superior to that available in electron - positron collisions .
we begin in sec .
ii with a discussion of the cabibbo - kobayashi - maskawa ( ckm ) matrix @xcite and its role in cp violation as currently observed in the neutral kaon system .
the role of @xmath0 mesons in checking this picture is outlined in sec .
we comment specifically on @xmath4 mixing in sec .
iv and on systematics of lifetime differences in sec . v. the identification of the flavor of a neutral @xmath0 meson by means of correlations with charged pions produced nearby in phase space , via fragmentation and/or resonances ,
is reviewed in sec .
decays of @xmath0 mesons to pairs of hadrons containing light ( @xmath5 ) quarks can provide information on phases of ckm elements , as noted in sec .
we mention some specific issues associated with studies in hadron colliders in sec .
viii , and conclude in sec .
an analogue of the @xmath13 system in the absence of cp ( or time - reversal ! ) violation is provided by two resonant circuits with equal natural frequencies and damping terms , coupled through a resistor , so that the long - lived oscillations will be those in which the two systems are in phase with one another and no energy is dissipated in the resistor
. emulation of cp violation is harder .
one may do so using asymmetric exchange of radiated energy between obliquely polarized antennas through a region in which faraday rotation is occurring .
thus , for example , vlf ( very - low - frequency ) radio waves are subject to different attenuation when propagating in the ionosphere from east to west and from west to east @xcite . a more `` table - top '' version of such a phenomenon could utilize the hall effect . in a parametrization @xcite in which the rows of the ckm matrix are labelled by @xmath14 and the columns by @xmath15
, we have @xmath16~~.\ ] ] the third quark family ensures that there will be a non - trivial phase , taken here to be that of @xmath17 .
the elements on and below the diagonal are determined by unitarity .
the phase of @xmath18 is appreciable , while those of @xmath19 and @xmath20 are negligible for present purposes .
these phases allow the standard @xmath21 interaction to generate cp violation as a higher - order weak effect .
the four parameters are measured in the following manner @xcite : \(1 ) the parameter @xmath22 is measured by a comparison of strange particle decays with muon decay and nuclear beta decay , leading to @xmath23 , where @xmath24 is the cabibbo angle .
\(2 ) the dominant decays of @xmath1-flavored hadrons occur via the element @xmath25 .
the lifetimes of these hadrons and their semileptonic branching ratios then lead to an estimate @xmath26 , or @xmath27 .
\(3 ) the decays of @xmath1-flavored hadrons to charmless final states allow one to measure the magnitude of the element @xmath17 and thus to conclude that @xmath28 .
\(4 ) the phase of @xmath17 , arg @xmath29 , must be determined with the help of indirect information associated with contributions of higher - order diagrams involving the top quark .
the unitarity of v and the fact that @xmath30 and @xmath31 are very close to 1 allow us to write @xmath32 , or , dividing by a common factor of @xmath33 , @xmath34 .
the point @xmath35 thus describes in the complex plane one vertex of a triangle [ with angle @xmath36 in a conventional notation @xcite ] whose other two vertices are @xmath37 ( with angle @xmath38 ) and @xmath39 ( with angle @xmath40 ) .
we obtain a set of constraints on the parameters @xmath41 and @xmath42 in the ckm matrix in the following manner .
more details may be found in ref .
@xcite ; we have updated some values .
\(1 ) charmless @xmath0 decays lead to the estimate @xmath43 and hence to the constraint @xmath28 mentioned above .
the limits correspond to the dotted semicircles with center ( 0,0 ) in fig . 1 .
\(2 ) recent averages @xcite for the @xmath44 mixing parameter @xmath45 ps@xmath46 and the @xmath47 lifetime @xmath48 ps can be combined to yield @xmath49 . if interpreted in terms of a box diagram for @xmath50 ( dominated by the top quark ) , this value leads to an estimate for @xmath51 reducing to @xmath52 .
the corresponding limits are shown as the dashed lines in fig . 1
. \(3 ) the parameter @xmath53 can be reproduced with a cp - violating @xmath11 mixing term due to the box diagram for @xmath12
. the contribution of top quarks alone would be proportional to i m @xmath54 ; with a small additional charm contribution , one finds @xmath55 , whose limits are denoted by the solid hyperbolae in fig . 1 .
here we have also used the recent average of cdf @xcite and d0 @xcite top quark masses , @xmath56 gev/@xmath57 , vacuum saturation factors @xmath58 and @xmath59 , a @xmath0 meson decay constant @xmath60 mev , and a qcd correction for the box diagram @xmath61 .
a large region centered about @xmath62 , @xmath63 is permitted .
nonetheless , it could be that the cp violation seen in kaons is due to an entirely different source , perhaps a superweak mixing of @xmath64 and @xmath65 @xcite . in that case
one could probably still accommodate @xmath66 ( see fig . 1 ) , and hence a real ckm matrix .
in order to confirm the predicted nonzero value of @xmath42 , we turn to @xmath0 mesons for help .
the error in the @xmath67 ckm element is dominated by theoretical uncertainty in extracting information from the measured @xmath0 lifetime and semileptonic branching ratio .
even more theoretical uncertainty is associated in extraction of @xmath68 from a small portion of the semileptonic decay spectrum involving leptons beyond the endpoint for @xmath69 . in the long run a measurement of the branching ratio for @xmath70 or @xmath71 , though challenging , will be helpful in extracting the product @xmath72 .
the expected branching ratios are about @xmath73 for @xmath74 and @xmath75 for @xmath76 .
one should not forget the possibility of eventually getting information on @xmath77 through the decay @xmath78 , which is expected to have a partial width of @xmath79 gev for @xmath56 gev/@xmath57 and @xmath80 .
the error on @xmath51 extracted from @xmath81 for nonstrange neutral @xmath0 mesons is dominated by uncertainty in @xmath82 . in sec .
iv we shall mention some ways in which this uncertainty may be reduced , both by direct calculation and by comparison of mixing in nonstrange and strange @xmath0 mesons .
two main avenues for detecting cp violating in systems involving @xmath1 quarks involve 1 ) decays to cp eigenstates @xcite , and 2 ) decays to cp non - eigenstates . in both cases , partial rates for particle and antiparticle decays
are compared , but experimental aspects and interpretations differ . _
1 . in decays
to cp eigenstates _
, one compares the partial rate for a decay of an initial @xmath47 with that for an initial @xmath83 .
the interference of direct decays ( such as @xmath84 ) and those involving mixing ( such as @xmath85 ) gives rise to rate asymmetries which can be easily interpreted in terms of the angles @xmath86 .
thus , if we define @xmath87 we have , in the limit of a single direct contribution to decay amplitudes , @xmath88 where @xmath89 .
this limit is expected to be very good for @xmath90 , but some correction for penguin contributions is probably needed for @xmath91 .
the value @xmath92 is nearly optimum to maximize the coefficient of @xmath93 . for the range of parameters noted in fig . 1
, we expect @xmath94 , i.e. , an asymmetry of a definite sign , and @xmath95 , i.e. , nearly any asymmetry within the possible limits imposed by the factor @xmath96 . thus , the @xmath90 asymmetry is likely to provide a consistency check , while the @xmath91 asymmetry should be more useful in specifying the parameters @xmath97 ( unless it lies outside the expected limits ) . in order to employ this method
it is necessary to know the flavor of the produced @xmath0 meson .
we shall remark on one possible `` tagging '' method in sec .
_ 2 . in decays
to cp non - eigenstates _ , the asymmetry is proportional to @xmath98 , where @xmath99 is a difference between weak phases of two interfering amplitudes , while @xmath100 is the corresponding strong phase shift difference in the two channels .
such channels might , for instance , be the @xmath101 and @xmath102 amplitudes for @xmath103 .
here one is able to compare decays of charged @xmath0 mesons with those of their antiparticles , so the identification of the flavor of the decaying meson does not pose a problem . on the other hand
, @xmath100 is generally expected to be small and quite uncertain for the energies characteristic of @xmath0 decays .
we shall outline in sec .
vii some recent progress in using decays of charged @xmath0 mesons to provide information on ckm phases _ without _ necessarily having to observe a cp - violating decay rate asymmetry .
as mentioned , the dominant error in extracting @xmath51 from @xmath44 mixing is associated with uncertainty in @xmath82 .
lattice gauge theory calculations @xcite indicate @xmath104 mev . a quark model calculation @xcite uses isospin splittings in @xmath105 and @xmath0 mesons to estimate the heavy quark light antiquark wave function at zero separation , obtaining @xmath106 mev with a negligible experimental error but a systematic error which is hard to estimate .
although no direct information on @xmath82 is available , we can check some of the predictions of the above models for @xmath105 mesons .
several experimental determinations @xcite of the @xmath107 decay constant have appeared in the last two years , leading to @xmath108 mev .
the estimate of ref .
@xcite finds @xmath109 or @xmath110 mev , not far below the present upper limit of 290 mev @xcite .
lattice estimates for the ratio in eq .
( [ eqn : su ] ) range between 0.8 and 0.9 .
the mixing of neutral strange @xmath0 mesons is expected to involve a ratio @xmath111 .
observation of this quantity poses a challenge but would provide very useful information on @xmath112 , since the corresponding ckm elements are assumed well known from unitarity : @xmath80 , @xmath113 . in turn
, one would estimate @xmath82 from @xmath112 using eq .
( [ eqn : su ] ) .
we show in table 1 the dependence of @xmath114 on @xmath112 and @xmath115 . we have updated a similar table in ref .
@xcite on the basis of more recent information on the @xmath116 lifetime and the top quark mass .
the dominant source of error on @xmath114 is primarily @xmath112 . our value of @xmath82 and eq .
( [ eqn : su ] ) imply @xmath117 mev .
.dependence of mixing parameter @xmath114 on top quark mass and @xmath116 decay constant . [ cols="^,^,^,^",options="header " , ] the opal @xcite , delphi @xcite , and aleph @xcite collaborations have now observed @xmath118 correlations which can be interpreted in terms of the predicted @xmath119 states . the opal data are shown in fig .
similar plots have been presented by delphi and aleph .
opal also sees a @xmath120 correlation . in all experiments
one is able to measure only the effective mass of a @xmath118 system . if a @xmath121 decays to @xmath122 , the photon in the @xmath123 decay
is lost , leading to an underestimate of the @xmath121 mass by about 46 mev but negligible energy smearing @xcite .
thus , the contributions to the @xmath118 mass distribution of @xmath124 and @xmath125 resonances with spacing @xmath126 can appear as three peaks , one at @xmath127 due to @xmath128 , one at @xmath129 mev due to @xmath130 , and one at @xmath131 mev due to @xmath132 .
the opal collaboration fits their @xmath118 mass distribution either with one peak with @xmath133 mev/@xmath57 and width @xmath134 mev , or two resonances , a @xmath124 candidiate at 5725 mev/@xmath57 with width @xmath135 mev and a @xmath125 candidiate at 5737 mev/@xmath57 with width @xmath136 mev .
the widths , mass splittings , and branching ratios to @xmath118 and @xmath122 in this last fit are taken from ref .
@xcite , and only the overall mass and production cross sections are left as free parameters . the opal @xmath120 mass distribution is fit either with a single resonance at @xmath137 mev/@xmath57 with width @xmath138 mev , or two narrow resonances , a @xmath124 candidate at 5874 mev/@xmath57 and a @xmath125 candidate at 5886 mev/@xmath57 .
the fitted masses of the nonstrange and strange resonances are respectively 30 mev lower and 40 mev higher than the predictions of ref .
the difference could well be due to additional contributions to the nonstrange channel from the lower - lying @xmath139 and @xmath140 states or from nonresonant fragmentation .
the corresponding strange states might lie below @xmath141 threshold .
the opal results imply that @xmath142 of the observed @xmath143 mesons are accompanied by a `` tagging @xmath144 '' arising from @xmath145 decay . by isospin reflection
, one should then expect @xmath142 of @xmath47 to be accompanied by a `` tagging @xmath146 '' arising from @xmath147 decay .
this is good news for the possibility of `` same - side tagging '' of neutral @xmath0 mesons .
[ another @xmath148 of the observed @xmath143 mesons are accompanied by a `` tagging @xmath149 . ''
the isospin - reflected kaon is neutral , and unsuitable for tagging .
] the delphi data can be fit with a single peak having a mass of @xmath150 mev and width @xmath151 mev .
the number of @xmath152 per @xmath1 jet is quoted as @xmath153 .
the aleph peak occurs at @xmath154 mev/@xmath57 , with gaussian width @xmath155 mev/@xmath57 .
the aleph @xmath121 signal is characterized by a production rate @xmath156 where the first error is statistical , the second is systematic , and the third is associated with uncertainty in ascribing the peak to the contribution of various resonances . multiplying the delphi and aleph @xmath121 rates by the isospin factor of 2/3 to compare with the opal result , we find complete agreement among the three . in principle
it should be possible to calibrate the correlations between charged pions and neutral @xmath0 s by comparing them with the isospin - reflected correlations betwen charged pions and _ charged _ @xmath0 s ( whose flavor may be easier to identify ) .
thus , the enhancement of the non - exotic @xmath157 channel with respect to the exotic @xmath158 channel should be the same as that of the non - exotic @xmath159 channel with respect to the exotic @xmath160 channel .
what can spoil this relation ?
i. dunietz and i @xcite have explored several instances in which care is warranted in making this comparison .
some of the differences could be real , but there are many sources of potential instrumental error against which one has concrete remedies .
interaction with the producing system _ can lead to final states which need not be invariant under isospin reflection .
for example , although a pair of gluons would produce a @xmath161 pair with isospin @xmath162 , the fragmentation process could involve picking up quarks from the producing system ( e.g. , proton or antiproton fragments ) in a manner not invariant with respect to @xmath163 substitution .
similarly , the production of a @xmath0 meson through diffractive dissociation of a proton ( which has more valence @xmath164 quarks than @xmath165 quarks ) need not be invariant under isospin reflection .
misidentification of associated charged kaons as pions _ can lead one to overestimate the charged @xmath0 charged pion correlations .
one expects @xmath166 correlations , as seen by opal , but not @xmath167 correlations .
as mentioned , the isospin reflection of a charged kaon is neutral , and would not contribute to a correlation between charged particles and neutral @xmath0 s .
pions in the decay of the associated @xmath0 _ will not be produced in an isospin - reflection - symmetric manner .
one must be careful not to confuse them with primary pions .
different time - dependent selection criteria _ for charged and neutral @xmath0 s can lead one to mis - estimate the mixing of neutral @xmath0 s with their antiparticles .
( i thank p. derwent for pointing this out . )
it is possible to make an unfortunate cut on @xmath47 lifetime which enhances the mixing considerably with respect to the value obtained by integrating over all times .
overestimates of particle identification efficiencies _ can lead to confusion in identification of the flavor of a neutral @xmath0 through the decay @xmath168 .
it is possible , especially for @xmath169 and @xmath144 with equal laboratory momenta , to confuse them with @xmath146 and @xmath149 , while still keeping them in a @xmath170 peak .
the cdf collaboration at fermilab has been studying charged pion - @xmath0 correlations ever since the reports of ref .
@xcite appeared , but no public announcement of these results has yet appeared .
the intent of the present subsection is _ not _ to provide excuses for this failure to report results , but rather to provide ideas for ways to view the data in which such correlations are likely to be robust
. they should certainly be present in hadron collider data .
in this section we turn to decays of @xmath0 mesons to cp non - eigenstates .
we have suggested @xcite that relations between decays of charged @xmath0 s to pairs of light pseudoscalar mesons based on flavor su(3 ) @xcite could provide information on weak phases by means of _ rate measurements alone_. the latest chapter in this story has been written since the workshop .
two years ago the cleo collaboration @xcite presented evidence for a combination of @xmath172 and @xmath91 decays , generically known as @xmath173 . on the basis of 2.4 fb@xmath46 of data ,
the most recent result @xcite is @xmath174 .
although one still can not conclude that either decay mode is nonzero at the @xmath175 level , the most likely solution is roughly equal branching ratios ( i.e. , about @xmath176 ) for each mode .
other results @xcite of the cleo collaboration on related modes include the upper bounds @xmath177 , @xmath178 , and @xmath179 .
interesting levels for the last two modes @xcite are probably around @xmath180 , and probably @xmath181 or less for @xmath182 . with good particle identification and a factor of several times more data
, it appears that cleo will be able to make a systematic study of decay modes with two light pseudoscalars .
what can it teach us ?
we mentioned earlier that rate asymmetries in the decays @xmath183 and @xmath184 could provide information on the weak angle @xmath36 , as long as a single quark subprocess dominated the decay .
additional contributions from penguin diagrams @xcite can be taken into account by means of an isospin triangle construction @xcite involving the relation @xmath185 , and the corresponding relation for the charge - conjugate processes . here
we define amplitudes such that a partial width is always proportional to the square of an amplitude with the same coefficient .
a similar amplitude quadrangle applies to the decays @xmath103 @xcite : @xmath186 .
when combined with the corresponding relation for @xmath83 and @xmath187 decays , and used in conjunction with the time - dependence of the decays @xmath188 , these quadrangles are useful in extracting the weak phase @xmath36 . in examining su(3 )
relations among @xmath189 amplitudes , where @xmath190 is a pseudoscalar meson , we found that one of the diagonals of the amplitude quadrangle for @xmath103 ( corresponding to an amplitude with isospin @xmath191 ) could be related to the purely @xmath192 amplitude for @xmath193 .
we obtained the relation @xmath194 where @xmath195 .
the @xmath196 amplitude is expected to be dominated by a ( gluonic ) penguin contribution , involving mainly the combination of ckm elements @xmath197 , whose electroweak phase is @xmath198 .
the electroweak phase of the @xmath199 amplitude is just @xmath200 .
thus , in the absence of strong - interaction phase shift differences , the shape of the amplitude triangle would give @xmath38 .
one could account for strong - interaction phases by comparing the amplitude triangle for @xmath143 decays with that for @xmath187 decays @xcite .
we estimated @xcite that in order to measure @xmath38 to @xmath201 one would need a sample including about 100 events in the channels @xmath202 .
the analyses of ref .
@xcite assumed that the only penguin contributions to @xmath0 decays were gluonic in nature .
consequently , one could treat the flavor - dependence in terms of an effective @xmath203 or @xmath204 transition since the gluon couples to light quarks in a flavor - symmetric manner .
thus , the @xmath102 amplitude in @xmath103 ( the diagonal of the @xmath171 amplitude quadrangle mentioned above ) was due entirely to the cabibbo - suppressed `` tree - diagram '' process @xmath205 , whose weak phase was well - specified .
it was pointed out @xcite that in certain penguin - dominated @xmath0 decays such as @xmath206 , electroweak penguin amplitudes were large enough to compete favorably with the tree amplitude in the @xmath102 channel .
in contrast to gluonic penguins , the virtual photon or @xmath207 emitted in an electroweak penguin diagram does not couple to light quarks in a flavor - symmetric manner , and possesses an @xmath208 component .
specifically , if one decomposes amplitudes into isospin channels , @xmath209 @xmath210 deshpande and he @xcite find , in a specific calculation , that @xmath211 @xmath212 where the first term in each equation is the `` tree '' contribution ( of lowest order in electroweak interactions ) , while the second term is the penguin contribution .
only the electroweak penguin contributes to the @xmath102 amplitude , but with magnitude comparable to the tree contribution .
the electroweak penguin spoils the relation of eq .
( [ eqn : tri ] ) .
recently several of us re - examined the effects of su(3 ) breaking @xcite and electroweak penguins @xcite , to see if one could extract electroweak penguin effects _ directly from the data_. since our previous su(3 ) decomposition gave a complete set of reduced amplitudes , electroweak penguins only changed the interpretation of these amplitudes , so that a separation of electroweak penguin effects was not possible merely on the basis of su(3 ) .
as pointed out by deshpande and he @xcite , certain amplitudes ( notably those for @xmath213 are expected to be dominated by electroweak penguins .
we noted that the @xmath171 amplitude quadrangle could be written in such a manner that one if its diagonals was equal to @xmath214 , where @xmath215 denotes an unmixed octet member .
the shape of the quadrangle , shown in fig . 3 ,
is uniquely determined , up to possible discrete ambiguities .
the case of octet - singlet mixtures in the @xmath42 requires us to replace the @xmath216 by the appropriate coefficient ; one can show that the su(3 ) singlet contribution of the @xmath42 is unimportant in this case .
the quadrangle has been written in such a way as to illustrate the fact @xcite that the @xmath217 amplitude receives only penguin contributions in the absence of @xmath218 corrections .
the weak phases of @xmath219 penguins , which are dominated by a top quark in the loop , are expected to be @xmath198 .
we have oriented the quadrangle to subtract out the corresponding strong phase , and define corresponding strong phase shift differences @xmath220 with respect to the strong phase of the @xmath221 amplitude .
the @xmath222 amplitude is composed of two parts , as noted above .
we can rewrite it slightly as @xmath223 the corresponding charge - conjugate quadrangle has one diagonal equal to @xmath224 so that one can take the difference to eliminate the electroweak penguin contribution : @xmath225 the quantity @xmath226 can be related to the @xmath192 @xmath8 amplitude in order to obtain @xmath227 .
specifically , if we neglect electroweak penguin effects in @xmath199 ( a good approximation ) , we find that latexmath:[\[\label{eqn : pipi } thus , we can extract not only @xmath227 , but also a strong phase shift difference @xmath229 , by comparing eqs .
( [ eqn : diff ] ) and ( [ eqn : pipi ] ) .
if such a strong phase shift difference exists , the @xmath0 and @xmath230 quadrangles will have different shapes , and cp violation in the @xmath0 system will already have been demonstrated .
the challenge in utilizing the amplitude quadrangle in fig .
3 is to measure @xmath231 , which has been estimated @xcite to be only @xmath232 ! very recently ( since the workshop ) deshpande and he @xcite have pointed out that the amplitude triangle @xmath233 implied by the su(3 ) relations of refs .
@xcite , where @xmath215 is the octet component of the @xmath42 , permits one to specify the quantity @xmath234 in eq .
( [ eqn : diff ] ) and extract @xmath38 as above .
here there is some delicacy associated with the su(3 ) singlet component of the physical @xmath42 .
@xmath235 final states _ ( @xmath236 ) are characterized by more graphs ( and hence more reduced su(3 ) amplitudes ) , since one no longer has the benefit of bose statistics as in @xmath237 decays .
there still exist quadrangle relations in @xmath238 and @xmath239 decays , however .
remarkably , if @xmath240 gluonic penguin diagrams ( small in any case ) are approximately equal for the cases in which the spectator quark ends up in a vector meson and in a pseudoscalar [ see ref .
@xcite for details ] , the previous quadrangle construction still holds if we replace @xmath241 with @xmath242 , @xmath243 with @xmath244 , and @xmath245 with @xmath246 .
deshpande and he @xcite predict @xmath247 , which probably means that the quadrangle reduces to two nearly overlapping triangles ( whose shapes will consequently be difficult to specify ) . on the other hand , in @xmath235
decays , the effects of electroweak penguins then may not be so important if the dominant processes are characterized by branching ratios of order @xmath176 as in @xmath189 decays .
hints of some signals have been seen in some @xmath235 channels in the latest cleo data @xcite , but only upper limits are being quoted .
these are fairly close to theoretical expectations in the case of some @xmath248 channels .
@xmath249 final states _
satisfy bose statistics .
since the total angular momentum of the decaying particle is zero , the ( space ) @xmath250 ( spin ) part of the @xmath249 wave function will be symmetric , as in @xmath251 final states @xcite .
thus , there should exist amplitude relations for each relative orbital angular angular momentum @xmath252 .
if one @xmath252 value dominates the decays , such relations might be tested using triangles constructed of square roots of decay rates , as in the @xmath251 case .
in this section we comment on miscellaneous issues which deserve more attention if hadron collider experiments are to make a dent in the problem of cp violation in the @xmath0 meson system .
as we mentioned earlier , if different minimum decay lengths are adopted for selection of different @xmath47 decay modes , the data sample will contain different admixtures of @xmath47 and @xmath83 in different decay modes .
apparent inconsistencies can show up in charged pion
neutral @xmath0 correlations .
a trigger based on the presence of secondary vertices may be the only way to pick out elusive decays such as @xmath253 in sufficent numbers for cp studies . in the comparison of @xmath253 and @xmath103 decay modes
, we have seen that good discrimination between pions and kaons is essential .
some distinctive processes of @xmath0 production and decay involve large number of kaons .
consider , for example the production of a @xmath254 with a kaon `` tag . ''
if the subsequent decay is @xmath255 , a total of four kaons in the final state ( counting the `` tagging '' kaon ) must be identified . in the production of a @xmath161 pair in high - energy @xmath3 collisions , as at lep ,
the @xmath1 and @xmath256 are separated from one another by many units of rapidity , with many hadrons filling the gap in between . in hadronic @xmath257 production
, there may be a continuum of rapidity separations between @xmath1 and @xmath256 , ranging from the very small ( as in fragmentation of a hard gluon to @xmath161 ) to the very large ( as in back - to - back production of @xmath1 and @xmath256 jets ) .
the `` same - side '' tagging method is likely to be more relevant for widely separated @xmath1 and @xmath256 .
one must learn more about the hadrons which fill the gap between a @xmath1 and a @xmath256 in hadronic collisions , and tailor one s tagging methods accordingly .
the most promising advantage for the study of @xmath0 physics at the tevatron ( and ultimately , at the lhc ) consists of the high rate of @xmath1 production .
the relatively high @xmath161 effective mass expected in such processes implies that `` same - side '' tagging of neutral @xmath0 mesons in order to identify cp - violating rate asymmetries deserves serious study .
a study of lifetime differences of hadrons containing @xmath1 quarks at an accuracy of several percent will permit the checking of detailed predictions of these differences .
the innate advantage ( based on rate ) of the fermilab detectors for identification of tagged neutral @xmath0 mesons became clear several years ago when the first reconstructed @xmath0 mesons were presented by the cdf collaboration .
however , `` same - side '' tagging results from cdf have been nearly two years in gestation
. meanwhile , the recent identification of `` @xmath121 '' mesons at lep has brought to mind the last of satchel paige s six rules for staying young @xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ do nt look back .
something may be gaining on you .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
i wish to thank m. gronau , o. f. hernndez , and d. london for enjoyable collaborations on some of the topics menioned here ; c. quigg and m. shochet for constructive comments ; f. bedeschi , d. cassel , m. feindt , r. fleischer , r. forty , l. k. gibbons , r. kowalewski , h. lipkin , s. schael , p. sphicas , and f. wrthwein for helpful correspondence ; p. derwent and d. gerdes for help with the last reference ; and m. worah for comments on the manuscript .
this work was supported by the united states department of energy under grant no .
de fg02 90er40560 .
# 1#2#3am . j. phys .
* # 1 * , # 2 ( # 3 ) # 1#2#3ann .
( n.y . ) * # 1 * , # 2 ( # 3 ) # 1#2#3acta phys .
polonica * # 1 * , # 2 ( # 3 ) # 1#2#3ann .
nucl . part .
sci . * # 1 * , # 2 ( # 3 ) # 1#2#3bull .
am . phys .
soc . * # 1 * , # 2 ( # 3 ) # 1#2#3comments on nucl .
phys . * # 1 * , # 2 ( # 3 ) 89_cp violation , _ edited by c. jarlskog ( world scientific , singapore , 1989 ) 79_proceedings of the 1979 international symposium on lepton and photon interactions at high energies , _ fermilab , august 23 - 29 , 1979 , ed . by t.
b. w. kirk and h. d. i. abarbanel ( fermi national accelerator laboratory , batavia , il , 1979 87_proceeding of the 1987 international symposium on lepton and photon interactions at high energies , _ hamburg , 1987 , ed . by w. bartel and r. rckl ( nucl .
b , proc . suppl .
3 ) ( north - holland , amsterdam , 1988 ) # 1#2#3 * # 1 * , # 2 ( # 3 ) 72_proceedings of the xvi international conference on high energy physics _ , chicago and batavia , illinois , sept . 6 13 , 1972 , edited by j. d. jackson , a. roberts , and r. donaldson ( fermilab , batavia , il , 1972 ) # 1#2#3int
phys . a * # 1 * , # 2 ( # 3 ) 87_selected topics in electroweak interactions _ ( proceedings of the second lake louise institute on new frontiers in particle physics , 15 21 february , 1987 ) , edited by j. m. cameron ( world scientific , singapore , 1987 ) 85_proceedings of the international symposium on lepton and photon interactions at high energy , _ kyoto , aug .
19 - 24 , 1985 , edited by m. konuma and k. takahashi ( kyoto univ . ,
kyoto , 1985 ) # 1#2#3mod .
lett . a * # 1 * , # 2 ( # 3 ) # 1#2#3nuovo cim . * # 1 * , # 2 ( # 3 ) # 1#2#3nucl
. phys . * # 1 * , # 2 ( # 3 ) # 1#2#3#4pisma zh .
* # 1 * , # 2 ( # 3 ) [ jetp lett . *
# 1 * , # 4 ( # 3 ) ] # 1#2#3phys . lett . * # 1 * , # 2 ( # 3 ) # 1#2#3phys .
b * # 1 * , # 2 ( # 3 ) # 1#2#3phys .
# 1 * , # 2 ( # 3 ) # 1#2#3phys .
d * # 1 * , # 2 ( # 3 ) # 1#2#3phys .
# 1 * , # 2 ( # 3 ) # 1#2#3phys . rep . * # 1 * , # 2 ( # 3 ) # 1#2#3prog .
# 1 * , # 2 ( # 3 ) # 1#2#3rev .
phys . * # 1 * , # 2 ( # 3 ) # 1 rp # 1 9025th international conference on high energy physics , singapore , aug . 2 - 8 , 1990 87_proceedings of the salt lake city meeting _ ( division of particles and fields , american physical society , salt lake city , utah , 1987 ) , ed .
by c. detar and j. s. ball ( world scientific , singapore , 1987 ) 89_proceedings of the xivth international symposium on lepton and photon interactions , _ stanford , california , 1989 , edited by m. riordan ( world scientific , singapore , 1990 ) 82_proceedings of the 1982 dpf summer study on elementary particle physics and future facilities _ , snowmass , colorado , edited by r. donaldson , r. gustafson , and f. paige ( world scientific , singapore , 1982 ) 90_research directions for the decade _
( proceedings of the 1990 summer study on high energy physics , june 25july 13 , snowmass , colorado ) , edited by e. l. berger ( world scientific , singapore , 1992 ) 90_testing the standard model _ ( proceedings of the 1990 theoretical advanced study institute in elementary particle physics , boulder , colorado , 327 june , 1990 ) , edited by m. cveti and p. langacker ( world scientific , singapore , 1991 ) # 1#2#3#4yad . fiz . *
# 1 * , # 2 ( # 3 ) [ sov .
# 1 * , # 4 ( # 3 ) ] # 1#2#3#4#5#6zh .
teor . fiz . *
# 1 * , # 2 ( # 3 ) [ sov .
- jetp * # 4 * , # 5 ( # 6 ) ] # 1#2#3zeit .
c * # 1 * , # 2 ( # 3 ) j. rosner , in _ the future of high sensitivity charm experiments ( proceedings of the charm2000 workshop ) _
, fermilab , 79 june 1994 , edited by d. m. kaplan and s. kwan ( fermilab - conf-94/190 , fermilab , batavia , il , 1994 ) , p. 297 . c. t. hill , in _ proceedings of the workshop on @xmath0 physics at hadron accelerators ( op .
, p. 127 ; c. quigg , _ ibid .
_ , p. 443 ; e.
eichten , c. t. hill , and c. quigg , , and in proceedings of the the charm2000 workshop ( see ref .
32 ) , pp . 345 , 355 . satchel paige , as quoted by richard donovan , collier s , 1953 .
paige was the first black pitcher in the major leagues , pitching for the cleveland indians from 1948 to 1951 and for the st .
louis browns from 1951 until his retirement at the age of nearly 50 in 1953 . | some aspects of @xmath0 physics relevant for experiments in hadron colliders are discussed .
these include the determination of parameters of the ckm matrix and confirmation of its role in cp violation , studies of mixing of nonstrange and strange @xmath0 mesons , lifetimes of hadrons containing @xmath1 quarks , the use of `` same - side '' tagging of neutral @xmath0 mesons via correlations with charged pions through fragmentation or resonances , and the determination of ckm phases through the study of decays of @xmath0 mesons to pairs of light hadrons . |
extradimensions are the backbone of present theoretical physics . during the last years
, there has been a considerable interest in kaluza - klein like scenarios suggesting that our usual spacetime could be just a slice of a larger dimensional manifold .
recent advances in string theory have thus postulated the existence of `` braneworlds '' ( @xmath1-hyperdimensional surfaces embedded in a @xmath2 dimensional manifold ) in which we are living on [ 1,2 ] .
several issues like the hierarchy between the electroweak and the planck scales [ 3,4 ] as well as the idea that the post - inflationary epoch of our universe was preceded by the collision of d3-branes [ 5 ] , for instance , have been successfully addressed in such multidimensional approaches .
+ in braneworld models , it is generally assumed that the standard model particles can not freely propagate into unseen dimensions and must be constrained to live on a @xmath3 submanifold [ 1,2 ] .
such conservative point of view adopts the idea that no interaction ( except gravitation ) can propagate into the bulk or between adjacent branes [ 6,7,8 ] .
the problem is then to find a physically reasonable mechanism that could effectively constrain the particles to stay confined in a lower dimensional spacetime sheet [ 1,2,9 - 20 ] . in some string theory inspired models ,
the confinement arises as a natural consequence of the fact that particles are open strings whose endpoints are attached on dp - branes [ 1,2 ] .
other approaches postulate the existence of domain walls in which standard model particles could be entrapped [ 9 - 20 ] . among earlier works [ 9,10,11 ] ,
rubakov and shaposhnikov [ 10 ] suggested long time ago that the wave function of fermion zero modes could concentrate near domain walls , generating 4d massless chiral fermions attached to them .
this idea has attracted much interest over the last years as it can solve in a very elegant way the hierarchy problem [ 3,4 ] and the proton decay problem : the extradimensional separation between chiral fermions generates exponentially suppressed couplings between them [ 21 - 23 ] .
nevertheless , most of these scenarios suffer from a significant drawback as they require the existence of external non - gravitational forces like scalar fields .
since the existence of such fields is still the subject of debate [ 24 ] , the credibility of these scenarios can be reasonably questioned .
+ in parallel , some attempts have tried to extend the hypothetical graviton capability of moving through the bulk [ 1,6,7,8 ] to the case of massive particles as well [ 12,15,16 ] . as a result , there could be a possibility for highly massive or energetic particles to acquire a non zero extradimensional momentum and escape from the branes to propagate into the bulk .
+ considering these two opposite approaches , it is clear that the question of whether standard model particles are totally or only partially confined on 4d hypersurfaces is still an open issue .
+ furthermore , all those approaches perpetuate the tradition inherited from relativity by assuming that the whole universe behaves like a smooth continuum .
however , there have been some recent attempts to develop models where the continuous extra dimensions are substituted by discrete dimensions [ 25 - 29 ] . in those approaches ,
the extra - dimensions are replaced by a finite number of points and the whole universe can be seen as made up of a collection of 4d sheets . besides keeping the physical richness of multidimensional spacetimes , such multi - sheeted approaches provide also a nice framework where the standard model may arise from pure geometry [ 25 ] . in recent works , at the crossroad between brane models and non commutative two - sheeted spacetimes , present authors have studied the quantum dynamics of fermions in a @xmath4 universe [ 30 - 32 ] . such a model can be seen as a formal extension of the kaluza - klein spacetime with a fifth extradimension restricted to only two points .
it was emphasised that despite the discrete bulk structure , both spacetime sheets are still connected at the quantum level .
this connection was shown to be related to the specific geometrical structure of the bulk with a coupling strength partly dictated by the magnitude of the electromagnetic gauge fields of both sheets .
+ in this paper , we propose to discuss the predictions of our model for the issue concerning the fermion localization on the 4d sub - manifolds . in section 2
we briefly summarize the mathematics underlying our approach and the main results of previous works are reported . in section 3 , we detail the mechanisms by which fermions are effectively confined on the 4d sheets .
it is demonstrated that the confining effect arises from the combination of the discrete bulk structure and the environmental interactions . to the authors knowledge ,
the mechanism described here is completely new and presents the advantage of not relying neither on any exotic trapping mechanism involving negative bulk energy nor on any scalar fields .
+ besides , in section 4 , we show that particles can have access granted to the other spacetime sheet in some specific circumstances .
this situation which is analogous to a motion through the bulk in braneworld models is illustrated by considering a resonant mechanism forcing the particle to oscillate between the two sheets .
this mechanism which might be experimentally investigated at present times and energy scales is worth being studied as it could become a practical tool for the experimental search of extradimensions .
in refs . [ 30 - 32 ] , a model aiming at describing the quantum dynamics of fermions in a two - sheeted spacetime was described .
it corresponds formally to the product of a four continuous manifold times a discrete two points space , i.e. @xmath5 .
+ such a universe can be pictured as a five dimensional universe where the fifth dimension is replaced by a discrete two points space with coordinates @xmath6 .
each point is then endowed with its own four - dimensional sheet and both sheets are separated by a distance @xmath7 . at that point , it is important to stress that this distance is just a phenomenological one and does not necessarily match the concept of distance we are familiar with .
the interaction between both sheets is reflected by the existence of a coupling strength @xmath8 ( see figure 1 ) given by the inverse distance between the sheets . in the abovementioned references , it was demonstrated that the model could be built by considering two different approaches which are now briefly described .
+ between the sheets is related to the distance which separates them in the extra - space.,title="fig:",width=7,height=4 ] the first approach was mainly based on the work of connes [ 25 ] , viet and wali [ 26,27 ] and relies on a non - commutative definition of the exterior derivative acting on the product manifold . due to the specific geometrical structure of the bulk , this operator is given by @xmath9 where the term @xmath10 has the dimension of mass ( in @xmath11 units ) and acts as a finite difference operator along the discrete dimension .
in the model discussed in ref . [ 30 ] and contrary to previous works , @xmath10 was considered as a constant geometrical field ( and not the higgs field [ 25 ] ) . as a consequence
a mass term was introduced and a two - sheeted dirac equation was derived @xmath12 it can be noticed that by virtue of the two - sheeted structure of spacetime , the wave function @xmath13 of the fermion is split into two components , each component living on a distinct spacetime sheet . obviously , the equation ( 2 ) can also be derived from the @xmath14 lagrangian _
l _ defined as @xmath15 where @xmath16 is the two - sheeted adjoint spinor of @xmath17 and with @xmath18 and @xmath19 the adjoint spinors respectively of @xmath20 and @xmath21 .
such a lagrangian can be also written using the following expended form @xmath22 at first sight , the doubling of the wave function can be seen as a reminiscence of the hidden - sector concept . while it is true that hidden sector models and present approach share several common points , it is equally true that they differ by the number of spacetime sheets they consider .
for instance , the so - called mirror matter approach , considers only one 4d manifold and justifies for the left / right parity by introducing new internal degrees of freedom to particles ( see for instance ref . [ 33 ] ) . in the present work
, it can be noticed that the number of particle families remains unchanged but the particles , thanks to their five dimensional nature , have access granted to the two 4d sheets .
so , instead of considering new internal degrees of freedom , we simply allow the particles to move freely in an extended spacetime .
+ the second approach developed in ref . [ 31 ] started from the usual covariant dirac equation in 5d . by assuming a discrete structure of the bulk , with two four - dimensional submanifolds , the extradimensional derivative has to be replaced by a finite difference counterpart @xmath23 then , as in the non - commutative approach , the dirac equation breaks down into a set of two coupled differential equations similar to eq .
( 2 ) [ 31 ] .
+ the incorporation of the electromagnetic field in the model can be done quite easily [ 30 - 32 ] .
the usual u(1 ) gauge field must be substituted by an extended u(1)@xmath24u(1 ) gauge relevant for the discrete @xmath25 structure of the universe .
in addition , each sheet possesses its own current and charge density distribution as source of the electromagnetic field .
the most general form for the new gauge ( to be incorporated within the two - sheeted dirac equation such that @xmath26 ) is then defined by ( see also refs . [
28 ] and [ 29 ] where such a gauge was also considered ) @xmath27 on the two sheets live then the distinct @xmath28 and @xmath29 fields . in the present model , the component @xmath30
can not be associated with the usual higgs field encountered in gsw model .
@xmath30 is the coupling of the photons fields of the two sheets .
since this term introduces obvious complications unless to be weak enough compared to @xmath31 ( most notably it leads to unusual transformations laws of the electromagnetic field which are difficult to reconcile with observations ) it is preferable to set @xmath32 .
the electromagnetic fields of both sheets are then completely decoupled and each sheet is endowed with its own electromagnetic structure .
note that the consequence of having @xmath33 would have been to couple each charged particle with the electromagnetic field of both sheets , irrespective of the localization of this particle in the bulk .
for instance a particle of charge @xmath34 localized in the sheet @xmath35 would have been sensitive to the electromagnetic field of the sheet @xmath36 with an effective charge @xmath37 ( @xmath38 ) .
this kind of exotic interactions has been considered previously in literature within the framework of the mirror matter paradigm and is not covered by the present paper [ 34,35 ] .
another relevant consequence of considering @xmath32 is that the photons fields are now totally trapped in their original sheets : photons belonging to a given sheet are not able to go into the other sheet ( as a noticeable consequence , the structures belonging to a given sheet are invisible from the perspective of an observer located in the other sheet ) .
the classical gauge field transformation which reads @xmath39 can be easily extended to fulfil the two - sheeted requirements . a solution consistent with
the above hypothesis @xmath40 shows that the gauge transformation is degenerated and reduced to a single @xmath41 which must be applied to both sheets simultaneously [ 30,31 ] . by setting @xmath40 and by considering the same gauge transformation in the two sheets we can get photons fields @xmath42 which behave independently from each other and in accordance with observations [ 30 - 32 ] . after introducing the gauge field into the @xmath25 dirac equation and taking the non relativistic limit ( following the standard procedure ) ,
a two - sheeted pauli like equation can be derived [ 30 - 32 ] ( using now natural units ) @xmath43 where @xmath44 and @xmath45 correspond to the wave functions in the @xmath35 and @xmath36 sheets respectively .
the hamiltonian @xmath46 is a block - diagonal matrix where each block is simply the classical pauli hamiltonian expressed in both sheets @xmath47 and such that @xmath28 and @xmath48 denote the magnetic vector potentials in the sheet @xmath35 and @xmath36 respectively .
the same convention is applied for the magnetic fields @xmath49 .
@xmath50 is the magnetic moment of the particle with @xmath51 the gyromagnetic factor and @xmath52 the magneton .
in addition to these `` classical '' terms , the two - sheeted hamiltonian comprises also new terms involving the discrete structure of the bulk .
@xmath53 is a constant geometrical term proportional to the square of the coupling constant @xmath54 and @xmath55 is another geometrical coupling involving the gauge fields of the two sheets .
this last term is proportional to @xmath10 .
since @xmath54 must be tiny in order for the model to be consistent with the experimental observations and measurements , the term @xmath53 can be neglected in a first approximation [ 30 - 32 ] and the non - commutative [ 30 ] approaches of the two - sheeted spacetime predict almost the same physics .
the only difference arises from a difference about the expression of the @xmath56 term . in the finite difference approach
an off diagonal part proportional to @xmath57 appears into @xmath53 .
this term was demonstrated to be responsible of particle oscillations between the two spacetime sheets , even in the case of a free particle [ 31 ] .
nevertheless , as explained in [ 31 ] , @xmath53 can be neglected since it is proportional to @xmath54 , an obvious tiny value in order for the model to be consistent with known experimental results . by contrast , in the non - commutative case , the @xmath53 term is free of off diagonal terms , and thus it can be simply eliminated through a rescaling of the energy scale . ] . as a consequence , @xmath55 imparts most of the new physics of the model .
this contribution takes the form [ 30 - 32 ] @xmath58 \label{10}\ ] ] as a consequence , the remaining coupling term @xmath55 arises through the magnetic vector potentials @xmath28 and @xmath48 and the magnetic moment @xmath50 . in eq .
( 10 ) , @xmath59 is a constant and @xmath60 stands for the isogyromagnetic factor [ 31 ] . from a theoretical point of view
, the neglect of the qed corrections leads to @xmath61 for an electron [ 31 ] . as a consequence @xmath59 must be equal to @xmath62 .
although it was suggested in ref .
[ 31 ] that @xmath63 could differ slightly from @xmath51 when taking into account qed , these corrections will not be considered in the present paper . as a consequence , in the proton or neutron cases , we consider also @xmath64 and use the standard experimental values of @xmath51 ( i.e. 5.58 and -3.82 respectively ) [ 31 ] .
it is worth being noticed that @xmath65 can be seen as an extradimensional magnetic field .
the similarities between @xmath55 and the classical term @xmath66 become more obvious when considering the presence of the magnetic moment in eq .
the coupling term @xmath55 arises from the geometrical proximity between the two sheets . as a consequence ,
even if these sheets do not share any 4d link , the particles can possibly move from one sheet to the other one .
such a delocalization between the two sheets can be seen as a major drawback of the model but we are now going to show that environmental interactions constrain the capability of moving between the sheets .
an illustrative example can be given .
let us consider a fermion moving in a constant magnetic vector potential @xmath67 created in the first sheet .
assuming that @xmath46 can be expressed in its diagonal form , the particle dynamics reduces to @xmath68 \mathbf{+}i\hbar \omega \left [ \begin{array}{cc } 0 & -\sigma _ z \\ \sigma _ z & 0 \end{array } \right ] \right\ } \left| \psi ( t)\right\rangle \label{11}\ ] ] where @xmath69 .
provided that the particle is originally located in the sheet @xmath35 and unpolarized , it is straightforward to show that the probability to find the particle in the @xmath36 sheet is @xmath70 with @xmath71 it is worth being noticed that the difference @xmath72 between the eigenvalues @xmath73 and @xmath74 of the hamiltonian @xmath46 plays a fundamental role as it governs the particle ability of reaching the second sheet ( i.e. of moving in the discrete space ) .
for that reason , it is convenient to recast it in the following form @xmath75 with @xmath76 in eq .
( 13 ) , the gravitational and electrostatic contributions have been included in the potential in order to give @xmath77 the most general form .
+ in the following , we restrict ourselves to the case of an irrotational magnetic vector potential such that @xmath78 . from eq .
( 12 ) , it is then clear that if @xmath79 , i.e. @xmath80 , the particle oscillates freely between the two sheets with a time periodicity @xmath81 .
note that during the oscillations , the particle which is a @xmath82 entity , behaves sometimes as a four dimensional @xmath83 one . indeed , when the maximum of the probability is reached ( i.e. 1 ) , the particle becomes totally localized in a four dimensional submanifold . from this point of view , it is clear that the particle is not spontaneaously constrained to stay on a specific sheet .
such ability to move between the sheets , can be seen as a discrete counterpart of motions in the bulk sometimes discussed in brane - world models . +
although the particles are able to oscillate between the sheets , four dimensional localization can still occur in this model .
we start by considering @xmath84 i.e. there is now an effective potential @xmath77 applied onto the particle .
then , it is trivial to see that the amplitude and the period of oscillations change .
the higher the potential is , the weaker the oscillation amplitude is .
therefore , in presence of a strong enough potential , the particle is now confined on its sheet and can not reach the other one . since , the potential is in fact the sum of all external influences acting on the particle , it becomes obvious that the confinement arises mainly from environmental interactions .
such an explanation of the matter confinement in a multidimensional spacetime is elegant in several aspects : first , it does not require any ad - hoc bulk field nor any extradimensional gravitational trapping mechanism [ 9 - 23 ] .
secondly , it is quite remarkable that the model suggests that the confinement of fermions may arise from usual 4d interactions ( particle scattering , em interactions , gravitational fields ... ) .
+ note that the matter confinement could perhaps be insured even for large value of @xmath10 provided that the intensity of the environmental interactions are significant .
+ let us consider more closely the electrostatic and gravitational contributions in @xmath72 .
with regards to the electrostatic potentials , the neutrality of matter implies that they can be neglected at our scale or any larger scale . at smaller scale
however , as inside an atom for instance , the typical potential energies undergone by an electron are about @xmath85 ev , which is a quite small value . the same can not be said for the gravitational potentials .
we consider the typical example of a single neutron interacting with a larger object like the earth , the sun or the milky way core ( which is responsible for the sun attraction around the galactic center ) .
if the acceleration induced by the earth on a neutron is @xmath86 , then it is @xmath87 and @xmath88 for the sun and the milky way core respectively .
although the gravitational forces exerted by a very distant object can be neglected , the gravitational potentials can become huge as we can convince oneself easily . for a neutron in the lab reference frame ,
the potential energy arising from the earth / particle interaction is @xmath89 ev ( absolute value ) , this potential reaches @xmath90 ev when considering the sun influence and it rises up to @xmath91 ev for the milky way core .
although the induced acceleration by such a massive object is negligible in an earth lab , its effect on the potential energy of particles is considerable .
it is of course related to the fact that the gravitational potential varies like @xmath92 and the force like @xmath93 with @xmath94 the distance between the particle and the influencing massive object . in such circumstances , it appears that the main contribution to the confinement is the gravitational potential exerted by massive objects or clusters of matter .
until now we have only discussed the gravitational contribution of the mass located in our sheet but from eq .
( 13 ) , it is clear that the masses located in the neighboring sheet also constrain the particle oscillations .
unfortunately , to assess the gravitational influence of the masses located in the other sheet , we need some kind of gravitational telescope which does not exist at present time . + let us now calculate the effective probability for a single particle , to oscillate between the two sheets .
we are going to assume a coupling constant of the order @xmath95 m@xmath96 ( which is a quite huge value ) .
typically , the magnetic potential can be related to a current intensity @xmath97 through @xmath98 ( @xmath99 is the vacuum permeability ) .
therefore , intensity can be considered as a good indicator of the reachable magnetic potentials .
we thus consider an intensity of @xmath100 a and a global confining potential of @xmath91 ev . for a neutron , the eq .
( 12 ) indicates that the maximum amplitude of the oscillations of probability is @xmath101 .
this value is so tiny that it is hard to believe that any particle oscillations might be observed in the physical world .
hence , in all practical cases , the confining effect is so strong that the particles are constrained to move on their sheet only .
such a result is interesting because it shows that the model does not violate any current observations .
however , it is unsatisfactory because it seems to prohibit any attempt to confirm the existence of a coupling between the neighboring sheets as predicted by this model
. a closer investigation of @xmath55 suggests however a way to enhance the probability amplitude .
obviously , a first way to enhance the probability of oscillations is to increase the intensity of the vector potential and/or to decrease the gravitational trapping potential .
unfortunately , since gravity can not be shielded and since magnetic potential is limited by the reachable current intensity , these solutions remain poorly effective . previously , the similarity between @xmath55 and the term @xmath102 was noted . since this last term
is traditionally involved in magnetic resonance , it is legitimate to search for a two - sheeted counterpart of this mechanism through which particle oscillations might occur .
let us again consider the two - sheeted pauli equation ( eq .
by analogy with magnetic resonance , we are now going to consider a particle initially located in the first sheet and subjected to a rotating curl - free magnetic vector potential .
the irrotational character of the magnetic potential precludes the existence of a related magnetic field and thus prevents the existence of a hamiltonian contribution of the form @xmath103 .
let us choose @xmath104 and @xmath105 with the magnetic vector potential rotating in the @xmath106 plane , i.e. @xmath107 \label{15}\ ] ] for simplicity reasons , it is assumed that @xmath46 can be expressed in its diagonal form .
the eigenstates of the hamiltonian @xmath46 are defined as @xmath108 \text { , } \left| \psi _ { -1/2}^{+}\right\rangle = \left [ \begin{array}{c } 0 \\ 1 \\ 0 \\ 0 \end{array } \right ] \label{16}\ ] ] for particles located in the first sheet with an energy equals to @xmath73 ( assuming both spin states @xmath109 ) , and @xmath110 \text { , } \left| \psi _ { -1/2}^{-}\right\rangle = \left [ \begin{array}{c } 0 \\ 0 \\ 0 \\ 1 \end{array } \right ] \label{17}\ ] ] for particles located in the second sheet with an energy equals to @xmath74 ( assuming both spin states @xmath109 ) .
taking into account these assumptions , eq .
( 8) becomes @xmath111 + i\omega _
p\hbar \left [ \begin{array}{cc } 0 & -\mathbf{\sigma \cdot e } \\
\mathbf{\sigma \cdot e } & 0 \end{array } \right ] \right\ } \left| \psi ( t)\right\rangle \label{18}\end{aligned}\ ] ] with @xmath112 .
it is then possible to develop the general solution as @xmath113 this solution can be inserted into eq . ( 18 ) to give the following system @xmath114 where @xmath115 .
this system can be trivially solved by analytical means . using the initial conditions
@xmath116 , @xmath117 and @xmath118 with @xmath119 ( i.e. the particle is initially localized in the @xmath35 sheet ) , one obtains @xmath120 if the spin was in the @xmath121 state at @xmath122 ( @xmath123 and @xmath124 ) , it can then be shown using eq .
( 24 ) that the probability @xmath125 to find the particle in the second sheet is given by @xmath126 and in addition , the particle is then in the down spin state .
by contrast , if the spin was in the @xmath127 state at @xmath122 we get for the probability @xmath125 @xmath128 and now the particle is in the spin up state in the second sheet .
( 25 ) corresponds to a resonant process : as the potential vector rotates with the angular frequency @xmath129 the probability @xmath125 presents a maximal amplitude equals to one , independently of the coupling constant and of the magnetic vector potential amplitude .
in addition , the probability oscillates with an angular frequency given by @xmath130 .
the resonance width at the half - height is @xmath131 .
the weaker the coupling constant @xmath10 is , the narrower the resonance is .
by contrast , the greater the magnetic vector potential @xmath132 is , the broader the resonance is .
note that eq .
( 26 ) is similar to eq .
( 25 ) except that it describes an anti - resonance associated with a counter - rotating vector potential .
( 25 ) and ( 26 ) remind those found in magnetic resonance ( mr ) where the spin orientation is influenced by the combination of a static and an oscillating magnetic field . in the present case ,
the matter exchange between the sheets arises from the resonant interaction between the rotating curl - free magnetic vector potential and the spin ( via the magnetic moment ) .
note that the diagonal contribution of the hamiltonian in eq .
( 18 ) plays a role similar to the spin / static magnetic field interaction found in mr .
but if @xmath133 stands for the difference of magnetic energy between two spin states in mr , the difference @xmath72 in our model arises from a completely different origin . as a consequence ,
when the angular frequency @xmath134 of the magnetic vector potential matches the confinement potential @xmath135 ( see previous section ) the particle is resonantly transferred from one sheet to the other one . in practice ,
any rotating vector potential induces electric field and possibly magnetic field that can skew dramatically any experimental study by increasing the contribution of @xmath72 .
even if we work with a neutral particle ( a neutron for instance ) , the use of an irrotational magnetic vector potential is necessary to avoid any undesirable magnetic field .
this requirement may be difficult to fulfill and an electromagnetic wave is not necessarily relevant for this purpose .
therefore , obtaining the required conditions for a resonant transfer remains a major challenge although it could perhaps be accessible with our current technology .
in this paper , the quantum behavior of massive particles has been studied within the framework of the two - sheeted spacetime model introduced in earlier works .
the model predicts that fermions oscillate between the two spacetime sheets in presence of irrotational vector potentials .
it is shown however that environmental interactions constrain very tightly these oscillations and lead to a four - dimensional localization of the fermion species . yet , we predict that the environmental confining effect can be overcome through a resonant mechanism which might be investigated at a laboratory scale .
the study of such a mechanism could be relevant for demonstrating the existence of another spacetime sheet and confirms that our spacetime is just a sheet embedded in a more complex manifold .
m. sarrazin , f. petit , `` quantum dynamics of particles in a discrete two - branes world model : can matter particles exchange occur between branes ? '' , acta phys .
b36 ( 2005 ) 1933 - 1950 ( hep - th/0409083 ) m. sarrazin , f. petit , `` artificially induced positronium oscillations in a two - sheeted spacetime : consequences on the observed decay processes '' , int
a21 ( 2006 ) 6303 - 6314 ( hep - th/0505014 ) | in recent papers , a model of a two - sheeted spacetime @xmath0 was introduced and the quantum dynamics of massive fermions was studied in this framework . in the present study , we show that the physical predictions of the model are perfectly consistent with observations and most important , it can solve the puzzling problem of the four - dimensional localization of the fermion species in multidimensional spacetimes .
it is demonstrated that fermion localization on the sheets arises from the combination of the discrete bulk structure and environmental interactions .
the mechanism described in this paper can be seen as an alternative to the domain wall localization arising in continuous five dimensional spacetimes .
although tightly constrained , motions between the sheets are , however , not completely prohibited . as an illustration , a resonant mechanism through which fermion oscillations between the sheets might occur is described . |
the architecture of concurrent software usually exhibits some forms of modularity . for example , concurrent algorithms are encapsulated in libraries and complex algorithms are often constructed using libraries of simpler ones .
this lets developers benefit from ready - made libraries of concurrency patterns and high - performance concurrent data structures , such as @xmath0 for java and threading building blocks for c++ . to simplify reasoning about concurrent software
, we need to exploit the available modularity .
in particular , in reasoning about a client of a concurrent library , we would like to abstract from the details of a particular library implementation .
this requires an appropriate notion of library correctness .
correctness of concurrent libraries is commonly formalised by _ linearizability _
@xcite , which fixes a certain correspondence between the library and its specification .
the latter is usually just another library , but implemented atomically using an abstract data type ; the two libraries are called _ concrete _ and _ abstract _ , respectively . a good notion of linearizability
should validate an _ abstraction theorem _
@xcite : the behaviours of any client using the concrete library are contained in the behaviours of the client using the abstract one .
this makes it sound to replace a library by its specification in reasoning about its client .
classical linearizability assumes a complete isolation between a library and its client , with interactions limited to passing values of a given data type as parameters or return values of library methods .
this notion is not appropriate for low - level heap - manipulating languages , such as c / c++ .
there the library and the client run in a shared address space ; thus , to prove the whole program correct , we need to verify that one of them does not corrupt the data structures used by the other .
type systems @xcite and program logics @xcite usually establish this using the concept of _ ownership _ of data structures by a program component : the right to access a data structure is given only to a particular component or a set of them .
when verifying realistic programs , this ownership of data structures can not be assigned statically ; rather , it should be _ transferred _ between the client and the library at calls to and returns from the latter .
the times when ownership is transferred are not determined operationally , but set by the proof method : as ohearn famously put it , `` ownership is in the eye of the asserter '' @xcite . however , ownership transfer reflects actual interactions between program components via the heap , e.g. , alternating accesses to a shared area of memory .
such interactions also exist in high - level languages providing basic memory protection , such as java . in this case
, we need to ensure that a client does not subvert a library by accessing a memory object after its ownership was transferred to the latter . for an example of ownership transfer between concurrent libraries and their clients consider a memory allocator accessible concurrently to multiple threads .
we can think of the allocator as owning the blocks of memory on its free - list ; in particular , it can store free - list pointers in them .
having allocated a block , a thread gets its exclusive ownership , which allows accessing it without interference from the other threads .
when the thread frees the block , its ownership is returned to the allocator .
trying to write to a memory cell after it was freed has dire consequences . as another example , consider any container with concurrent access , such as a concurrent set from @xmath0 or threading building blocks .
a typical use of such a container is to store pointers to a certain type of data structures .
however , when verifying a client of the container , we usually think of the latter as holding the ownership of the data structures whose addresses it stores @xcite .
thus , when a thread inserts a pointer to a data structure into a container , its ownership is transferred from the thread to the container .
when another thread removes a pointer from the container , it acquires the ownership of the data structure the pointer identifies .
if the first thread tries to access a data structure after a pointer to it has been inserted into the container , this may result in a race condition .
unlike a memory allocator , the container code usually does not access the contents of the data structures its elements identify , but merely ferries their ownership between different threads .
for this reason , correctness proofs for such containers @xcite have so far established their classical linearizability , without taking ownership transfer into account .
we would like to use the notion of linearizability and , in particular , an abstraction theorem to reason about the above libraries and their clients in isolation , taking into account only the memory that they own . to this end , we would like the correctness of a library to constrain not only pointers that are passed between it and the client , but also the contents of the data structures whose ownership is transferred .
so far , there has been no notion of linearizability that would allow this .
in the case of concurrent containers , we have no way of using classical linearizability established for them to validate an abstraction theorem that would be applicable to clients performing ownership transfer .
this paper fills in these gaps .
* contributions . * in this paper , we generalise linearizability to a setting where a library and its client execute in a shared address space , and boundaries between their data structures can change via ownership transfers .
linearizability is usually defined in terms of _ histories _ , which are sequences of calls to and returns from a library in a given program execution , recording parameters and return values passed . to handle ownership transfer
, histories also have to include descriptions of memory areas transferred .
however , in this case , some histories can not be generated by any pair of a client and a library : while generating histories of a library we should only consider its executions in an environment that respects ownership .
for example , a client that transfers an area of memory upon a call to a library not communicating with anyone else can not then transfer the same area again before getting it back from the library upon a method return .
we propose a notion of _ balancedness _ that characterises those histories that treat ownership transfer correctly .
we then define a _
linearizability relation _ between balanced histories , matching histories of a concrete and an abstract library ( section [ sec : observ ] ) .
this definition does not rely on a particular model of program states for describing memory areas transferred in histories , but assumes an arbitrary model defined by a _ separation algebra _ @xcite . by picking a model with so - called _ permissions _
@xcite , we can allow clients and libraries to transfer non - exclusive rights to access certain memory areas in particular ways , instead of transferring their full ownership .
this makes our results applicable even when libraries and their clients share access to some areas of memory .
our definition of balancedness for arbitrary separation algebras relies on a new formalisation of the notion of a _ footprint _ of a state , describing the amount of permissions the state includes ( section [ sec : prelim ] ) .
the rest of our technical development relies on a novel compositional semantics for a language with libraries that defines the denotation of a library or a client considered separately in an environment that communicates with the component correctly via ownership transfers ( sections [ sec : prog ] and [ sec : semantics ] ) .
in particular , the semantics allows us to generate the set of all histories that can be produced by a library solely from its code , without considering all its possible clients .
this , in its turn , allows us to lift the linearizability on histories to libraries and establish the abstraction theorem ( section [ sec : refinement ] ) . on the way
, we also obtain an insight into the original definition of linearizability without ownership transfer , showing a surprising relationship between one of the ways of its formulation and the plain subset inclusion on the sets of histories produced by concrete and abstract libraries .
we note that the need to consider ownership transfer makes the proof of the abstraction theorem highly non - trivial .
this is because proving the theorem requires us to convert a computation with a history produced by the concrete library into a computation with a history produced by the abstract one , which requires moving calls and returns to different points in the computation . in the setting without ownership transfer ,
these actions are thread - local and can be moved easily ; however , once they involve ownership transfer , they become global and the justification of their moves becomes subtle , in particular , relying on the fact that the histories involved are balanced ( section [ sec : rearr - proof ] ) . to avoid having to prove the new notion of linearizability from scratch for libraries that do not access some of the data structures transferred to them , such as concurrent containers
, we propose a _ frame rule for linearizability _
( section [ sec : frame ] ) .
it ensures the linearizability of such libraries with respect to a specification with ownership transfer given their linearizability with respect to a specification without one .
we provide a glossary of notation at the end of the paper .
our results hold for a class of models of program states called _ separation algebras _ @xcite , which allow expressing the dynamic memory partitioning between libraries and clients .
a * _ separation algebra _ * is a set @xmath1 , together with a partial commutative , associative and cancellative operation @xmath2 on @xmath1 and a unit element @xmath3 . here
commutativity and associativity hold for the equality that means both sides are defined and equal , or both are undefined .
the property of cancellativity says that for each @xmath4 , the function @xmath5 is injective .
we think of elements of a separation algebra @xmath1 as _ portions _ of program states and the @xmath2 operation as combining such portions .
the partial states allow us to describe parts of the program state belonging to a library or the client .
when the @xmath2-combination of two states is defined , we call them * _ compatible_*. incompatible states usually make contradictory claims about the ownership of memory .
we sometimes use a pointwise lifting @xmath6 of @xmath2 to sets of states : for @xmath7 we let @xmath8 .
elements of separation algebras are often defined using partial functions .
we use the following notation : @xmath9 means that the function @xmath10 is defined on @xmath11 , @xmath12 means that it is undefined on @xmath11 , @xmath13 denotes the set of arguments on which @xmath10 is defined , @xmath14 $ ] denotes a nowhere - defined function , and @xmath15 $ ] denotes the function that has the same value as @xmath10 everywhere , except for @xmath11 , where it has the value @xmath16 .
we also write @xmath17 for an expression whose value is irrelevant and implicitly existentially quantified .
below is an example separation algebra @xmath18 : @xmath19 a ( partial ) state in this model consists of a finite partial function from allocated memory locations to the values they store .
the @xmath2 operation on @xmath18 is defined as the disjoint function union @xmath20 , with the nowhere - defined function @xmath14 $ ] as its unit .
thus , @xmath2 combines disjoint pieces of memory .
more complicated separation algebras do not split memory completely , instead allowing heap parts combined by @xmath2 to overlap .
this is done by associating so - called _ permissions _
@xcite with memory cells in the model , which do not give their exclusive ownership , but allow accessing them in a certain way .
types of permissions range from read sharing @xcite to accessing memory in an arbitrary way consistent with a given specification @xcite .
we now give an example of a separation algebra with permissions of the former kind .
we define the algebra @xmath21 as follows : @xmath22 ; \qquad { { \sf ram}}_\pi = { { \sf loc}}\rightharpoonup_{\it fin } ( { { \sf val}}\times { { \sf perm}}).\ ] ] a state in this model consists of a finite partial function from allocated memory locations to values they store and so - called _
permissions_numbers from @xmath23 $ ] that show `` how much '' of the memory cell belongs to the partial state @xcite .
the latter allow a library and its client to share access to some of memory cells .
permissions in @xmath21 allow only read sharing : when defining the semantics of commands over states in @xmath21 , the permissions strictly less than @xmath24 are interpreted as permissions to read ; the full permission @xmath24 additionally allows writing .
the @xmath2 operation on @xmath21 adds up permissions for memory cells .
formally , for @xmath25 , we write @xmath26 if : @xmath27 if @xmath26 , then we define @xmath28 otherwise , @xmath29 is undefined . the unit for @xmath2 is the empty heap @xmath14 $ ] .
this definition of @xmath2 allows us , e.g. , to split a memory area into two disjoint parts .
it also allows splitting a cell with a full permission @xmath24 into two parts , carrying read - only permissions @xmath30 and agreeing on the value stored in the cell .
these permissions can later be recombined to obtain the full permission , which allows both reading from and writing to the cell .
since we develop all our results for an arbitrary separation algebra , by instantiating it with algebras similar to @xmath21 , we can handle cases when a library and its client share access to some memory areas .
consider an arbitrary separation algebra @xmath1 with an operation @xmath2 .
we define a partial operation @xmath31 , called * _ state subtraction _ * , as follows : @xmath32 is a state in @xmath1 such that @xmath33 ; if such a state does not exist , @xmath32 is undefined .
the cancellativity of @xmath2 implies that @xmath32 is determined uniquely , and hence , the @xmath34 operation is well - defined .
when reasoning about ownership transfer between a library and a client , we use the @xmath2 operation to express a state change for the component that is receiving the ownership of memory , and the @xmath34 operation for the one that is giving it up .
[ prop - diff ] for all @xmath35 , if @xmath29 and @xmath36 are defined , then @xmath37 our definition of linearizability uses a novel formalisation of a _ footprint _ of a state , which , informally , describes the amount of memory or permissions the state includes .
[ defn - foot ] a * _ footprint _ * of a state @xmath38 in a separation algebra @xmath1 is the set of states @xmath39 in the following , @xmath40 ranges over footprints .
the function @xmath41 computes the equivalence class of states with the same footprint as @xmath38 . in the case of @xmath18
, we have @xmath42 for every @xmath43 .
thus , states with the same footprint contain the same memory cells .
definitions of @xmath41 for separation algebras with permissions are more complicated , taking into account not only memory cells present in the state , but also permissions for them . in the case of the algebra @xmath21 , for @xmath44 we have @xmath45 in other words , states with the same footprint contain the same memory cells with the identical permissions ; in the case of memory cells on read permissions , the states also have to agree on their values .
let @xmath46 be the set of footprints in a separation algebra @xmath1 .
we now lift the @xmath2 and @xmath34 operations on @xmath1 to @xmath47 .
first , we define the operation @xmath48 for adding footprints .
consider @xmath49 and @xmath50 such that @xmath51 and @xmath52 .
if @xmath53 is defined , we let @xmath54 ; otherwise @xmath55 is undefined .
the @xmath56 operation is well - defined .
consider @xmath49 and @xmath57 such that @xmath51 and @xmath52 .
take another pair of states @xmath58 such that @xmath59 and @xmath60 .
thus , @xmath61 and @xmath62 , which implies : @xmath63 furthermore , if @xmath64 and @xmath53 are defined , then for all @xmath65 , @xmath66 hence , @xmath67 , so that @xmath56 is well - defined . for @xmath18
, @xmath56 is just the pointwise lifting of the disjoint function union @xmath68 . to define a subtraction operation on footprints , we use the following condition . [ foot - canc ]
the @xmath2 operation of a separation algebra @xmath1 is * _ cancellative on footprints _ * when for all @xmath69 , if @xmath53 and @xmath64 are defined , then @xmath70 for example , the @xmath2 operations on @xmath18 and @xmath21 satisfy this condition .
when the @xmath2 operation of an algebra @xmath1 is cancellative on footprints , we can define an operation @xmath71 of * _ footprint subtraction _ * as follows .
consider @xmath72 .
if for some @xmath73 , we have @xmath51 , @xmath52 and @xmath74 , then we let @xmath75 .
when such @xmath76 do not exist , @xmath77 is undefined .
the @xmath78 operation is well - defined .
consider @xmath79 and @xmath80 such that @xmath81 , @xmath82 , @xmath83 and @xmath84 .
we have : @xmath85 since @xmath86 , by definition [ foot - canc ] this implies @xmath87 , so that @xmath78 is well - defined . for @xmath18 ,
the @xmath78 operation is defined as follows . for @xmath88 , take any @xmath89 such that @xmath90 and @xmath91 .
then @xmath92 , if @xmath93 ; otherwise , @xmath94 is undefined .
we say that a footprint @xmath95 is * _ smaller _ * than @xmath96 , written @xmath97 , when @xmath77 is defined .
the @xmath56 and @xmath78 operations on footprints satisfy an analogue of proposition [ prop - diff ] .
[ prop - delta ] for all @xmath98 , if @xmath99 and @xmath100 are defined , then @xmath101 in the rest of the paper , we fix a separation algebra @xmath1 with the @xmath2 operation cancellative on footprints .
in the following , we consider descriptions of computations of a library providing several methods to a multithreaded client .
we fix a set @xmath102 of thread identifiers and a set @xmath103 of method names . as we explained in section
[ sec : intro ] , a good definition of linearizability has to allow replacing a concrete library implementation with its abstract version while keeping client behaviours reproducible . for this
, it should require that the two libraries have similar client - observable behaviours .
such behaviours are recorded using _ histories _ , which we now define in our setting .
[ def : intact ] an * _ interface action _ * @xmath104 is an expression of the form @xmath105 or @xmath106 , where @xmath107 , @xmath108 and @xmath109 .
we denote the sets of all call and return actions by @xmath110 and @xmath111 , and the set of all interface actions by @xmath112 .
an interface action records a call to or a return from a library method @xmath113 by thread @xmath114 .
the component @xmath38 in @xmath105 specifies the part of the state transferred upon the call from the client to the library ; @xmath38 in @xmath115 is transferred in the other direction .
for example , in the algebra @xmath18 , the annotation @xmath116 $ ] implies the transfer of the cell at the address @xmath117 storing @xmath118 . in the algebra @xmath21 ,
@xmath119 $ ] implies the transfer of a read permission for this cell . a * _ history _ * @xmath120 is a finite sequence of interface actions such that for every thread @xmath114 , its projection @xmath121 to actions by @xmath114 is a sequence of alternating call and return actions over matching methods that starts from a call action . in the following ,
we use the standard notation for sequences : @xmath122 is the empty sequence , @xmath123 is the @xmath124-th element of a sequence @xmath125 , @xmath126 is the prefix of @xmath125 of length @xmath127 , and @xmath128 is the length of @xmath125 .
not all histories make intuitive sense with respect to the ownership transfer reading of interface actions . for example , let @xmath129 and consider the history in figure [ fig : hist](a ) .
the history is meant to describe _ all _ the interactions between the library and the client . according to the history
, the cell at the address @xmath130 was first owned by the client , and then transferred to the library by thread @xmath24 . however , before this state was transferred back to the client , it was again transferred from the client to the library , this time by thread @xmath131 .
this is not consistent with the intuition of ownership transfer , as executing the second action requires the cell to be owned both by the library and by the client , which is impossible in @xmath18 . as we show in this paper ,
histories that do not respect the notion of ownership , such as the one above , can not be generated by any program , and should not be taken into account when defining linearizability .
we now use the notion of footprints of states from section [ sec : prelim ] to characterise formally the set of histories that respect ownership .
a finite history @xmath120 induces a partial function @xmath132 , which tracks how a computation with the history @xmath120 changes the footprint of the library state : @xmath133{@{}r@{\ } c@{\ } l@{\ \ } l@ { } } { \llbracket \varepsilon \rrbracket^\sharp}l & = & l ; \\[2pt ] { \llbracket h \psi \rrbracket^\sharp}l & = & { \llbracket h \rrbracket^\sharp}l \circ \delta(\sigma ) , & \mbox{if } \psi = ( \ _ , { { \sf call}}\ \_(\sigma ) ) \wedge { ( { \llbracket h \rrbracket^\sharp}l \circ \delta(\sigma)){\mathpunct{\downarrow } } } ; \\[2pt ] { \llbracket h \psi \rrbracket^\sharp}l & = & { \llbracket h \rrbracket^\sharp}l { \mathop{{\backslash}\!\!{\backslash}}}\delta(\sigma ) , & \mbox{if } \psi = ( \ _ , { { \sf ret}}\ \_(\sigma ) ) \wedge { ( { \llbracket h \rrbracket^\sharp}l{\mathop{{\backslash}\!\!{\backslash}}}\delta(\sigma)){\mathpunct{\downarrow } } } ; \\[2pt ] { \llbracket h \psi \rrbracket^\sharp}l & = & \mbox{undefined } , & \mbox{otherwise}. \end{array}\ ] ] using this function , we characterise histories respecting the notion of ownership as follows . [
defn : balance ] a history @xmath120 is * _ balanced _ * from @xmath134 if @xmath135 is defined .
we call subsets of @xmath136 * _ interface sets_*. an interface set can be used to describe all the behaviours of a library relevant to its clients . in the following ,
@xmath137 ranges over interface sets . to keep client behaviours reproducible when replacing a concrete library by an abstract one
, we do not need to require the latter to reproduce the histories of the former exactly : the histories generated by the two libraries can be different in ways that are irrelevant for their clients .
we now introduce a _ linearizability relation _ that matches a history of a concrete library with that of the abstract one that yields the same client - observable behaviour .
[ lin ] the * _ linearization relation _ * @xmath138 on histories is defined as follows : @xmath139 holds if there exists a bijection @xmath140 such that @xmath141 & ( h(i ) = ( \ _ , { { \sf ret}}\ \ _ ) \wedge h(j ) = ( \_,{{\sf call}}\ \ _ ) ) ) ) { \implies } \rho(i ) < \rho(j ) ) .
\end{array}\ ] ] we lift @xmath138 to @xmath142 as follows : @xmath143 holds if @xmath144 and @xmath139 . finally , we lift @xmath138 to interface sets as follows : @xmath145 holds if @xmath146 thus , a history @xmath120 is linearized by a history @xmath147 when the latter is a permutation of the former preserving the order of actions within threads and non - overlapping method invocations .
the duration of a method invocation is defined by the interval from the method call action to the corresponding return action ( or to the end of the history if there is none ) .
an interface set @xmath148 is linearized by an interface set @xmath149 , if every history in @xmath148 may be reproduced in a linearized form by @xmath149 without requiring more memory .
we now discuss the definition in more detail .
definition [ lin ] treats parts of memory whose ownership is passed between the library and the client in the same way as parameters and return values in the classical definition @xcite : they are required to be the same in the two histories .
in fact , the setting of the classical definition can be modelled in ours if we pass parameters and return values via the heap .
let @xmath129 and let us fix distinct locations @xmath150 for @xmath151 meant for the transfer of parameters and return values
. then histories of the classical definition are represented in our setting by histories where all actions are of the form @xmath152 ) ) \mbox { or } ( t,{{\sf ret}}\ m([{{\sf arg}}_t : { \it retval } ] ) ) , \mbox { where } { \it param } , { \it retval } \in { { \sf val}}.\ ] ] the novelty of our definition lies in restricting the histories considered to balanced ones , which are the only ones that can be produced by programs ( we formalise this fact in section [ sec : semantics ] ) .
the notion of balancedness also plays a key role in proving the abstraction theorem in the presence of ownership transfer ( section [ sec : rearr ] ) .
the fact that the linearizability relation allows us to permute actions by different threads lets us arrange method invocations into a linear sequence .
for example the history in figure [ fig : hist](b ) is linearized by that in figure [ fig : hist](c ) .
the former might correspond to a concurrent stack implementation , where threads @xmath24 and @xmath131 pass parameters and return values via locations @xmath24 and @xmath131 , respectively .
histories such as the one in figure [ fig : hist](c ) , where a call to every method is immediately followed by the corresponding return , are called * _ sequential_*. sequential histories correspond to abstract libraries with every method implemented atomically .
when the histories in figures [ fig : hist](b ) and [ fig : hist](c ) are members of interface sets defining the behaviour of concurrent and atomic stack implementations , the linearizability relationship between them allows us to justify that the call to pop by thread @xmath24 in figure [ fig : hist](b ) can return @xmath153 , since this behaviour can be witnessed by the valid history of the atomic implementation in figure [ fig : hist](c ) .
( in fact , the pop would also be allowed to return @xmath154 , since the resulting history would be linearized by a sequential history with the push of @xmath154 preceding the pop . ) the requirement that the order of non - overlapping method invocations be preserved is inherited from the classical notion of linearizability @xcite . as shown by filipovi et al .
@xcite , this requirement is essential to validate an abstraction theorem for clients that can communicate via shared client - side variables .
for example , since in figure [ fig : hist](b ) the push of @xmath155 returns before the push of @xmath154 is called , the order between these method invocations has to stay the same in any linearizing history , such as the one in figure [ fig : hist](c ) . following filipovi et al .
@xcite , we do not require the abstract history @xmath147 to be sequential , like in the classical definition of linearizablity .
this allows our definition to compare behaviours of two concurrent library implementations .
we also allow a concrete history to contain calls without matching returns , arising , e.g. , because the corresponding method invocation did not terminate . in this case
, we require the same behaviour to be reproduced in the abstract history @xcite , which is possible because the latter does not have to be sequential .
for example , the history in figure [ fig : hist](d ) is linearized by that in figure [ fig : hist](e ) .
this yields a simpler treatment of non - terminating calls than the use of completions in the classical definition of linearizability @xcite .
definition [ lin ] requires that the initial footprint of an abstract history @xmath147 be smaller than that of the concrete history @xmath120 .
this requirement is standard in data refinement @xcite : it ensures that , when we replace a concrete library by an abstract one in a program , the library - owned memory stays disjoint from the client - owned one .
it does not pose problems in practice , as the abstract library generating @xmath147 usually represents some of the data structures of the concrete library abstractly and , hence , more concisely .
so far we have defined the notion of linearizability on interface sets without taking into account library implementations that generate them . in the rest of the paper , we develop this notion for libraries written in a particular programming language and prove an abstraction theorem , which guarantees that a library can be replaced by another library linearizing it when we reason about its client program .
we consider a simple concurrent programming language : @xmath156 { \mathcal{l } } & : : = & \{m = c;\ \ldots;\ m = c \ } \\[2pt ] { \mathcal{s } } & : : = & { \sf let}\ { \mathcal{l}}\ { \sf in}\ c \parallel \ldots \parallel c \end{array}\ ] ] a program consists of a single * _ library _ * @xmath157 implementing methods @xmath108 and its * _ client _ * @xmath158 , given by a parallel composition of threads . the language is parameterised by a set of * _ primitive commands _ * @xmath159 , meant to execute atomically . commands also include method calls @xmath108 , sequential composition @xmath160 , nondeterministic choice @xmath161 and finite iteration @xmath162 .
we use @xmath163 and @xmath164 instead of conditionals and while loops for theoretical simplicity : as we show below , the latter can be defined in the language as syntactic sugar .
methods do not take arguments and do not return values , as these can be passed via special locations on the heap associated with the identifier of the thread calling the method ( section [ sec : observ ] ) .
we disallow nested method calls .
we also assume that every method called in the program is defined by the library , and thus call @xmath165 a * _ complete program_*. an * _ open program _ * is a library @xmath157 without a client , or a client @xmath166 without a library implementation : @xmath167\ { \sf in}\ c\parallel\ldots\parallel c \\[2pt ] { \mathcal{p}}&::= & { \mathcal{s}}\mid { \mathcal{c}}\mid { \mathcal{l}}\end{array}\ ] ] in @xmath166 , we allow the client to call methods that are not defined in the program ( but belong to the missing library ) .
an open program represents a library or a client considered in isolation .
the novelty of the kind of open programs we consider here is that we allow them to communicate with their environment via ownership transfers .
we now define a way to specify a contract this communication follows .
a * _ predicate _ * is a set of states from @xmath1 , and a * _ parameterised predicate _ * is a mapping from thread identifiers to predicates .
we use the same symbols @xmath168 for ordinary and parameterised predicates ; it should always be clear from the context which one we mean .
when @xmath169 is a parameterised predicate , we write @xmath170 for the predicate obtained by applying @xmath169 to a thread @xmath114 . both kinds of predicates can be described syntactically , e.g. , using separation logic assertions @xcite .
we define possible ownership transfers between components with the aid of * _ method specifications _ * @xmath171 , which are sets of hoare triples @xmath172 , at most one for each method . here
@xmath169 and @xmath173 are parameterised predicates such that @xmath170 describes pieces of state transferred when thread @xmath114 calls the method @xmath113 , and @xmath174 , those transferred at its return .
note that the intention of the pre- and postconditions in method specifications is only to identify the areas of memory transferred ; in other words , they describe the `` type '' of the returned data structure , but not its `` value '' . as usual for concurrent algorithms , a complete specification of a library is given by its abstract implementation ( section [ sec : refinement ] ) . for example , as we discussed in section [ sec : intro ] , when programmers store pointers in a concurrent container , they often intend to transfer the ownership of the data structures these pointers identify at calls to and returns from the container s methods . in figure [ fig : impl](a ) we give an example of such a container a bounded stack represented by an array . for readability , we write examples in c instead of the minimalistic language introduced above .
the library protects the array by a lock ; more complicated algorithms allow a higher degree of concurrency @xcite .
take @xmath129 and for @xmath175 let @xmath176 denote the set of states representing all well - formed data structures of a certain type allocated at the address @xmath11 .
for example , for objects with a single integer field we have @xmath177 \mid y \in { { \sf val}}\}$ ] .
then the specification of the stack when it stores pointers to such data structures can be given as follows : [ container ] l \{x .
arg_t x * obj(x ) } push \{arg_t ( arg_t * obj(x ) ) } ; + \{arg_t _ } pop \{x .
arg_t x * ( ( x = empty emp ) ( x = empty ( x)))}. here we use the separation logic syntax to describe predicates parameterised by the thread identifier @xmath114 .
thus , @xmath178 denotes the empty heap @xmath14 $ ] , @xmath179 the heap @xmath180 $ ] , and @xmath2 the combination of heaps with disjoint domains . in the following we also use the assertion @xmath181 for @xmath182 , denoting all the heaps with the domain @xmath183 .
we use distinguished locations @xmath184 , @xmath185 to pass parameters and return values . according to the specification
, the stack gets the ownership of an object when a pointer to it is pushed , and gives it up when the pointer is popped .
@l@|@l@ void * stack[size ] ; int count = 0 ; // count of // elements stored lock array_lock ; // protects // the array and the count int push(void * arg ) lock(array_lock ) ; if ( count = = size ) unlock(array_lock ) ; return full ; stack[count++ ] = arg ; unlock(array_lock ) ; return ok ; void * pop ( ) lock(array_lock ) ; if ( count = = 0 ) unlock(array_lock ) ; return empty ; void * obj = stack[count ] ; unlock(array_lock ) ; return obj ; & struct node node * prev , * next ; ; node * free_list ; // a cyclic doubly- // linked list with a sentinel node lock * list_lock ; // protects the list void free(void * arg ) node * block = ( node*)arg ; lock(list_lock ) ; block->prev = free_list ; block->next = free_list->next ; free_list->next->prev = block ; free_list->next = block ; unlock(list_lock ) ; void * alloc ( ) lock(list_lock ) ; if ( free_list->next = = free_list ) unlock(list_lock ) ; return null ; node * block = free_list->next ; free_list->next = block->next ; block->next->prev = free_list ; unlock(list_lock ) ; return block ; + + ( a ) & ( b ) now take @xmath186 and let @xmath187 we specify an allocator managing blocks of _ size _ memory cells as follows : [ alloc ] l \{x .
arg_t x * ( x .. (x+_size_-1 ) _ ) } free \{arg_t _ } ; + \{arg_t _ } alloc \{x .
arg_t x * ( ( x = 0 emp ) ( x = 0 ( x .. (x+_size_-1 ) _ ) ) ) } .
the specification corresponds to the ownership transfer reading of allocator calls explained in section [ sec : intro ] . in figure
[ fig : impl](b ) we give an example allocator corresponding to the specification ( we have omitted initialisation code from the figure ) .
note that , unlike the stack , the allocator does access the blocks of memory transferred to it by the client , since it stores free - list pointers inside them . to define the semantics of ownership transfers unambiguously ( section [ sec : semantics ] ) , we require pre- and postconditions in method specifications to be _ precise _ @xcite . a predicate @xmath188 is * _ precise _ * if for every state @xmath38 there exists at most one substate @xmath189 satisfying @xmath190 , i.e. , such that @xmath191 and @xmath192 for some @xmath193 . since the @xmath2 operation is cancellative , when such a substate @xmath189 exists , the corresponding substate @xmath193 is unique and is denoted by @xmath194 .
a parameterised predicate @xmath190 is precise if so is @xmath195 for every @xmath114 . informally , a precise predicate carves out a unique piece of the heap .
for example , assuming the algebra @xmath18 , the predicate @xmath196\}$ ] and those used in the allocator specification are precise .
however , the predicate @xmath197 , [ \,]\}$ ] is not : when @xmath116 $ ] we can take either @xmath198 $ ] and @xmath199 $ ] , or @xmath200 $ ] and @xmath201 $ ] .
a * _ specified open program _ * is of the form @xmath202 or @xmath203 . in the former
, the specification @xmath171 describes all the methods that @xmath166 may call . in the latter
, @xmath171 provides specifications for the methods in the open program that can be called by its external environment . in both cases
, @xmath171 specifies the type of another open program that can fill in the hole in @xmath166 or @xmath157 .
when we are not sure which form a program has , we write @xmath204 . in this case , if @xmath205 does not have a client , then @xmath171 is empty ; if @xmath205 does not have a library , then @xmath206 is empty ; and if @xmath205 is complete , then both @xmath171 and @xmath206 are empty . for specified open programs @xmath207\ { \sf in}\ c_1\parallel\ldots\parallel c_n\ ] ] and @xmath208 agreeing on the specification @xmath171 of library methods , we denote by @xmath209 the complete program @xmath210 we now discuss primitive commands in more detail .
consider the set @xmath211 of subsets of @xmath1 with a special element @xmath212 used to denote an error state , resulting , e.g. , from dereferencing an invalid pointer .
we extend the @xmath2 operation on @xmath213 to @xmath214 by letting @xmath215 for all @xmath216 .
we assume an interpretation of every primitive command @xmath217 as a transformer @xmath218 , which maps pre - states to states obtained when thread @xmath219 executes @xmath220 from a pre - state .
the fact that our transformers are parameterised by @xmath114 allows atomic accesses to areas of memory indexed by thread identifiers .
this idealisation simplifies the setting in that it lets us do without special thread - local or method - local storage for passing method parameters and return values .
some typical primitive commands are : @xmath221 = e ' , \qquad { \sf assume}(e),\ ] ] where expressions @xmath222 are defined as follows : @xmath223 \mid e+e \mid { -}e \mid { ! e } \mid \ldots\ ] ] here @xmath224 refers to the identifier of the thread executing the command , @xmath225 $ ] returns the contents of the address @xmath222 in memory , and @xmath226 is the c - style negation of an expression @xmath222it returns @xmath24 when @xmath222 evaluates to @xmath118 , and @xmath118 otherwise .
the @xmath227 command filters out all the input states where @xmath222 evaluates to @xmath118 .
hence , after @xmath227 is executed , @xmath222 always has a non - zero value . using it ,
the standard commands for conditionals and loops can be defined in our language as follows : @xmath228 for the above commands @xmath220 and @xmath107 , we define @xmath229 using the transition relation @xmath230 in figure [ fig : transfer ] : @xmath231 in the figure , @xmath232 denotes the result of evaluating the expression @xmath222 in the state @xmath38 with the current thread identifier @xmath114 .
when this evaluation dereferences illegal memory addresses , it results in the error value @xmath212 .
we define @xmath233 similarly , but using the transition relation @xmath234 in figure [ fig : transfer - permission1 ] .
the transformers formalise the semantics of permissions explained in section [ sec : prelim ] : permissions less than @xmath24 allow reading , and the full permission @xmath24 additionally allows writing . note that @xmath235 yields an empty set of post - states when its condition evaluates to zero , leading to the program getting stuck . thus , even though , when executing the if statement ( [ if - assm ] ) , both branches of the non - deterministic choice will be explored , only the branch where the assume condition evaluates to true will proceed further .
@xmath236 \sigma , [ e ] = e ' & \leadsto_t & \sigma{[{\llbracket e \rrbracket}_{\sigma , t } : { \llbracket e ' \rrbracket}_{\sigma , t } ] } , & \mbox{if}\ { \llbracket e \rrbracket}_{\sigma , t } \in { { \sf dom}}(\sigma ) , { \llbracket e ' \rrbracket}_{\sigma , t}\in { { \sf
val}}\\[2pt ] \sigma , { [ e ] } = e ' & \leadsto_t & \top , & \mbox{if}\ { \llbracket e \rrbracket}_{\sigma , t } \not\in { { \sf dom}}(\sigma ) \mbox { or } { \llbracket e ' \rrbracket}_{\sigma , t } = \top \\[2pt ] \sigma , { { \sf assume}}(e ) & \leadsto_t & \sigma , & \mbox{if}\ { \llbracket e \rrbracket}_{\sigma , t } \in { { \sf val}}- \{0\ } \\[2pt ] \sigma , { { \sf assume}}(e ) & \leadsto_t & \top , & \mbox{if}\ { \llbracket e \rrbracket}_{\sigma , t } = \top \end{array}\ ] ] @xmath237 \sigma , { [ e ] } = e ' & \leadsto_t & \sigma{[{\llbracket e \rrbracket}_{\sigma , t } : ( { \llbracket e ' \rrbracket}_{\sigma , t},1 ) ] } , & \mbox{if}\ \sigma({\llbracket e \rrbracket}_{\sigma , t } ) = ( \ _ , 1 ) , { \llbracket e ' \rrbracket}_{\sigma , t } \in { { \sf val}}\\[2pt ] \sigma , { [ e ] } = e ' & \leadsto_t & \top , & \mbox{if the above condition does not hold } \\[2pt ] \sigma , { { \sf assume}}(e ) & \leadsto_t & \sigma , & \mbox{if}\ { \llbracket e \rrbracket}_{\sigma , t } \in { { \sf val}}- \{0\ } \\[2pt ] \sigma , { { \sf assume}}(e ) & \leadsto_t & \top , & \mbox{if}\ { \llbracket e \rrbracket}_{\sigma , t } = \top \end{array}\ ] ] for our results to hold , we need to place some restrictions on the transformers @xmath238 for every primitive command @xmath217 and thread @xmath185 : footprint preservation : : @xmath239 .
strong locality : : @xmath240 .
footprint preservation prohibits primitive commands from allocating or deallocating memory .
this does not pose a problem , since in the context of linearizability , an allocator is just another library and should be treated as such .
the strong locality of @xmath238 says that , if a command @xmath220 can be safely executed from a state @xmath189 , then when executed from a bigger state @xmath53 , it does not change the additional state @xmath193 and its effect depends only on the state @xmath189 and not on the additional state @xmath193 .
the strong locality is a strengthening of the locality property in separation logic @xcite : @xmath241 locality rules out commands that can check if a cell is allocated in the heap other than by trying to access it and faulting if it is not allocated . for example , let @xmath242 and consider the following transformer @xmath243 : @xmath244\ } \mbox { else } \{\sigma\}.\ ] ] the transformer @xmath245 defines the denotation of a ` command ' that writes @xmath118 to the cell at the address @xmath24 if it is allocated and acts as a no - op if it is not .
this violates locality .
indeed , take @xmath246 $ ] and @xmath247 $ ] . then @xmath248 )
= \{[1 : 0]\}\ ] ] and @xmath249 ) * \{[1:1]\}= \{[\,]\ } * \{[1:1]\}=\{[1:1]\}.\ ] ] hence , @xmath250 does not hold . while locality prohibits the command from changing the additional state , it permits the effect of the command to depend on this state @xcite .
the strong locality forbids such dependencies . to see this , consider another ` command ' defined by the following transformer @xmath243 : @xmath251 @l@ l@ f^t ( ) = & + & + & .
@xmath251 the command does not access the cell at the address @xmath131 , since it does not fault if the cell is not allocated .
however , when the cell is allocated , the effect of the command depends on its value .
it is easy to check that @xmath245 is local .
however , it is not strongly local , since for @xmath252 $ ] and @xmath253 $ ] , we have @xmath254 ) = \{[1 : 0 , 2:0]\}\ ] ] and @xmath255 ) * \{[2:0]\ } = \{[1:0 ] , [ 1:1]\ } * \{[2:0]\ } = \{[1:0 , 2:0 ] , [ 1:1 , 2:0]\},\ ] ] so that @xmath256 does not hold .
the property of strong locality subsumes the one of contents independence used in situations similar to ours in previous work on data refinement in a sequential setting @xcite . the transformers for standard commands , except memory ( de)allocation ,
satisfy the conditions of footprint preservation and strong locality .
we now give the semantics to complete and open programs . in the latter case
, we define component - local semantics that include all behaviours of an open program under any environment satisfying the specification associated with it . in section [ sec : refinement ] , we use these to lift linearizability to libraries and formulate the abstraction theorem .
programs in our semantics denote sets of _ traces _ , recording every step in a computation .
these include both internal actions by program components and calls and returns .
we define program semantics in two stages .
first , given a program , we generate the set of the possible execution traces of the program .
this is done solely based on the structure of its statements , without taking into account restrictions arising from the semantics of primitive commands or ownership transfers .
the next step filters out traces that are not consistent with the above restrictions using a trace evaluation process and , for open programs , annotates calls and returns appropriately .
traces consist of _ actions _ , which include primitive commands performed internally by a component and calls or returns , possibly annotated with states .
thus actions , include all interface actions @xmath104 from definition [ def : intact ] .
the set of * _ actions _ * is defined as follows : @xmath257 where @xmath107 , @xmath108 and @xmath217 . a * _ trace _ * @xmath125 is a finite sequence of actions such that for every thread @xmath114 , the projection of @xmath125 to @xmath114 s call and return actions is a sequence of alternating call and return actions over matching methods that starts from a call action .
we classify actions in a trace as those performed by the client and the library based on whether they happen inside a method . for a trace @xmath125 and an index @xmath258 ,
an action @xmath123 is a * _ client action _ * if @xmath259 for some thread @xmath114 and a primitive command @xmath220 and @xmath260 an action @xmath123 is a * _ library action _
* if @xmath259 for some @xmath114 and @xmath220 but @xmath123 is not a client action , that is , @xmath261 a trace is a * _ client trace _ * , if all of its actions of the form @xmath262 are client actions ; it is a * _ library trace _ * , if they all of them are library actions . in the following
, @xmath263 denotes client traces , @xmath264 library traces , and @xmath125 arbitrary ones .
we write @xmath265 for the projection of @xmath125 to client , call and return actions , @xmath266 for that to library , call and return actions , and @xmath267 for that to call and return actions .
consider a program @xmath204 and let @xmath268 be the set of methods implemented by its library or called by its client .
we define the trace set @xmath269 of @xmath205 in figure [ fig : trace ] .
we first define the trace set @xmath270 of a command @xmath271 , parameterised by the identifier @xmath114 of the thread executing it and a mapping @xmath272 giving the trace set of the body of every method that @xmath271 can call when executed by a given thread .
the trace set of a client @xmath273 is obtained by interleaving traces of its threads .
@xmath274 { \llparenthesis\hspace{1pt } c_1 + c_2 \hspace{1pt}\rrparenthesis}_t \eta & = & { { \llparenthesis\hspace{1pt } c_1 \hspace{1pt}\rrparenthesis}_t \eta } \cup { { \llparenthesis\hspace{1pt } c_2 \hspace{1pt}\rrparenthesis}_t \eta } \\[2pt ] { \llparenthesis\hspace{1pt } c^ * \hspace{1pt}\rrparenthesis}_{t}\eta & = & \big({\llparenthesis\hspace{1pt } c \hspace{1pt}\rrparenthesis}_t \eta\big)^ * \\[2pt ] { \llparenthesis\hspace{1pt } m \hspace{1pt}\rrparenthesis}_t \eta & = & \{(t , { { \sf call}}\ m)\ , \tau\ , ( t , { { \sf ret}}\ m ) \mid \tau \in \eta(m , t)\ } \\[2pt ] { \llparenthesis\hspace{1pt } c_1 ; c_2 \hspace{1pt}\rrparenthesis}_t \eta & = & \{\tau_1 \tau_2 \mid \tau_1 \in { \llparenthesis\hspace{1pt } c_1 \hspace{1pt}\rrparenthesis}_t \eta \wedge \tau_2 \in { \llparenthesis\hspace{1pt } c_2 \hspace{1pt}\rrparenthesis}_t \eta\ } \\[2pt ] { \llparenthesis\hspace{1pt } c_1 \parallel \ldots \parallel c_n \hspace{1pt}\rrparenthesis } \eta & = & \bigcup\{\tau_1 \parallel \ldots \parallel \tau_n \mid \forall t \in \ { 1,\ldots , n\}.\ , \tau_t \in { \llparenthesis\hspace{1pt } c_t \hspace{1pt}\rrparenthesis}_t \eta\ } \end{array } \\ \\ \begin{array}{@{}r@{}c@{}l@ { } } { \llparenthesis\hspace{1pt } { \sf let}\ \{m = c_m \mid m \in m\}\ { \sf in}\ c_1 \parallel\ldots \parallel c_n \hspace{1pt}\rrparenthesis } & = & { { \sf prefix}}({\llparenthesis\hspace{1pt } c_1 \parallel\ldots
\parallel c_n \hspace{1pt}\rrparenthesis } ( { \boldsymbol{\lambda}}(m , t).\ , { \llparenthesis\hspace{1pt } c_m \hspace{1pt}\rrparenthesis}_t ( \ _ ) ) ) \\[2pt ] { \llparenthesis\hspace{1pt } \gamma : { \sf let}\ [ -]\ { \sf in}\ c_1\parallel\ldots\parallel c_n \hspace{1pt}\rrparenthesis } & { } = { } & { { \sf prefix}}({\llparenthesis\hspace{1pt } c_1\parallel\ldots\parallel c_n \hspace{1pt}\rrparenthesis } ( { \boldsymbol{\lambda}}(m , t).\ , \{\varepsilon\ } ) ) \\[2pt ] { \llparenthesis\hspace{1pt } \{m = c_m \mid m \in \{m_1,\ldots , m_j\}\ } : \gamma \hspace{1pt}\rrparenthesis } & = & { } \\[2pt ] \multicolumn{3}{@{}l@ { } } { \hfill \begin{array}{@{}r@ { } } { { \sf prefix}}(\bigcup\nolimits_{k\ge 1 } { \llparenthesis\hspace{1pt } c_{{\sf mgc}}\parallel \ldots ( \mbox{$k$ times } ) \ldots \parallel c_{{\sf mgc}}\hspace{1pt}\rrparenthesis}({\boldsymbol{\lambda}}(m , t).\ , { \llparenthesis\hspace{1pt } c_m \hspace{1pt}\rrparenthesis}_t ( \ _ ) ) ) \\[2pt ] ( \mbox{where } c_{{\sf mgc}}= ( m_1 + \ldots + m_j)^ * ) \end{array } } \end{array } \end{array}\ ] ] the
trace set @xmath275 of a complete program is that of its client computed with respect to a mapping @xmath276 associating every method @xmath113 with the trace set of its body @xmath277 .
since we prohibit nested method calls , @xmath278 does not depend on @xmath279 .
we prefix - close the resulting trace set to take into account incomplete executions . in particular
, this allows the thread scheduler to be unfair : a thread can be preempted and never scheduled again . a program @xmath202 generates client traces @xmath280 , which do not include internal library actions .
this is achieved by associating an empty trace with every library method .
finally , a program @xmath281 generates all possible library traces @xmath282 .
this is achieved by running the library under its _ most general client _ , where every thread executes an infinite loop , repeatedly invoking arbitrary library methods .
@xmath283 & & \mbox{else } \{\tau ' \mid \exists \tau \in { \llparenthesis\hspace{1pt } \gamma \vdash { \mathcal{p } } : \gamma ' \hspace{1pt}\rrparenthesis}.\ , ( \ _ , \tau ' ) \in { \llbracket \gamma \vdash \tau : \gamma ' \rrbracket}\sigma\ } \end{array } $ ] @xmath284 { \llbracket \gamma \vdash \tau\varphi : \gamma ' \rrbracket}\sigma & { } = { } & \begin{array}[t]{@{}l@ { } } \mbox{if $ ( { \llbracket \gamma \vdash \tau : \gamma ' \rrbracket}\sigma = \top)$ then $ \top$ } \\[2pt ] \mbox{else if $ ( \exists ( \sigma',\ _ ) \in { \llbracket \gamma \vdash \tau : \gamma ' \rrbracket}\sigma.\ , { \llbracket \gamma \vdash \varphi : \gamma ' \rrbracket}\sigma ' = \top)$ then $ \top$ } \\[2pt ] \mbox{else $ \{(\sigma'',\tau'\varphi ' ) \mid\exists \sigma'.\,(\sigma',\tau ' ) \in { \llbracket \gamma \vdash \tau : \gamma ' \rrbracket}\sigma \wedge ( \sigma'',\varphi ' ) \in { \llbracket \gamma \vdash \varphi : \gamma ' \rrbracket}\sigma'\}$ } \end{array } \end{array } $ ] @xmath285 { \llbracket ( t,{{\sf call}}\ m ) \rrbracket}\sigma & { } = { } & \{(\sigma , ( t,{{\sf call}}\ m))\ } \\[2pt ] { \llbracket ( t,{{\sf ret}}\ m ) \rrbracket}\sigma & { } = { } & \{(\sigma , ( t,{{\sf ret}}\ m))\ } \\[2pt ] { \llbracket ( t,{{\sf call}}\ m ) : ( \{p\}\ { m}\ \{q\ } ) , \gamma ' \rrbracket}\sigma & { } = { } & \{(\sigma * \sigma_p , ( t,{{\sf call}}\ m(\sigma_p ) ) ) \mid \sigma_p \in p_t \wedge ( \sigma * \sigma_p){\mathpunct{\downarrow}}\ } \\[2pt ] { \llbracket ( t,{{\sf ret}}\ m ) : ( \{p\}\ { m}\ \{q\ } ) , \gamma ' \rrbracket}\sigma & { } = { } & \mbox{if } ( \sigma { \mathop{\backslash}}q_t){\mathpunct{\uparrow}}\mbox { then } \top \mbox
{ else } \{(\sigma { \mathop{\backslash}}q_t , ( t,{{\sf ret}}\ m(\sigma { \mathop{\backslash}}{(\sigma { \mathop{\backslash}}q_t)})))\ } \\[2pt ] { \llbracket ( \{p\}\ { m}\ \{q\ } ) , \gamma \vdash ( t,{{\sf call}}\ m ) \rrbracket}\sigma & { } = { } & \mbox{if } ( \sigma { \mathop{\backslash}}p_t){\mathpunct{\uparrow}}\mbox { then } \top \mbox { else } \{(\sigma { \mathop{\backslash}}p_t , ( t,{{\sf call}}\ m(\sigma { \mathop{\backslash}}{(\sigma { \mathop{\backslash}}p_t)})))\ } \\[2pt ] { \llbracket ( \{p\}\ { m}\ \{q\ } ) , \gamma \vdash ( t,{{\sf ret}}\ m ) \rrbracket}\sigma & { } = { } & \{(\sigma * \sigma_q , ( t,{{\sf ret}}\ m(\sigma_q ) ) ) \mid \sigma_q \in q_t \wedge ( \sigma * \sigma_q){\mathpunct{\downarrow}}\ } \end{array } $ ] the set of traces generated using @xmath286 may include those not consistent with the semantics of primitive commands or expected ownership transfers .
we therefore define the meaning of a program @xmath287 by evaluating every trace in @xmath288 from a given initial state to determine whether it is feasible . for open programs
, this process also annotates calls and returns in a trace with states transferred .
the formal definition of @xmath289 is given in figure [ fig : eval ] with the aid of a trace evaluation function @xmath290 .
given an initial state , this either yields multiple final states and annotated traces , or fails and produces @xmath212 . if the resulting set of state - trace pairs is empty , then the trace is infeasible and is discarded .
if the evaluation produces @xmath212 on any trace from @xmath291 , then the program has no semantics for the given initial state and its denotation is defined to be @xmath212 .
the evaluation of @xmath125 is defined inductively on its length using a function @xmath292 that evaluates a single action @xmath293 .
we explain this function by considering separately the cases of a complete program , open program with a library and open program with a client .
the evaluation @xmath294 for an action in a complete program has the standard semantics , with the effects of primitive commands computed using their transformers from section [ sec : prog ] . in this case ,
calls and returns are left unannotated , since no ownership transfers to or from the external environment are performed .
the function @xmath295 gives a * _ library - local _ * semantics to the program @xmath296 , in the sense that it generates library traces under any client respecting @xmath206 .
when a method @xmath113 from @xmath206 is called by thread @xmath114 , the library receives the ownership of any state consistent with the method precondition @xmath170 .
this state has to be compatible with that of the library .
after the method returns , the library has to give up the piece of state satisfying its postcondition .
since @xmath174 is precise , this piece of state is determined uniquely . the evaluation faults if the state to be transferred is not available ; thus , a library has no semantics if it violates the contract with its client given by @xmath206 .
this also ensures that the histories produced by a library are balanced .
[ safe2wb ] if @xmath297 , then @xmath298 is balanced from @xmath299 .
the function @xmath300 gives a * _ client - local _ * semantics to @xmath202 , in the sense that it generates traces of this client assuming any behaviour of the library consistent with @xmath171 .
when a thread @xmath114 calls a method @xmath113 in @xmath171 , it transfers the ownership of a piece of state satisfying the method precondition @xmath170 to the library being called . as before
, this piece is defined uniquely , because preconditions are precise . when such a piece of state is not available , the evaluation faults .
this ensures that client respects the method specifications of the libraries it uses . when the method returns , the client receives the ownership of an arbitrary piece of state satisfying its postcondition @xmath174 , compatible with the current state of the client .
we now formulate a lemma , used in the proof of the abstraction theorem ( section [ sec : refinement ] ) , that states the connection between the library - local and client - local semantics on one side and the semantics of complete programs on the other .
we start by introducing some auxiliary definitions .
a program @xmath204 is * _ safe _ * at @xmath38 , if @xmath301 ; @xmath205 is safe for @xmath302 , if it is safe at @xmath38 for all @xmath303 . for a set of initial states
@xmath304 , let @xmath305 we define an operator @xmath306 combining the resulting sets @xmath307 and @xmath308 of state - trace pairs produced by the client - local and library - local semantics into a set corresponding to the complete program : @xmath309 where @xmath310 and @xmath311 is a function on traces that erases the state annotations from their interface actions .
[ prop : sem : local - global ] assume @xmath202 and @xmath312 safe for @xmath313 and @xmath314 , respectively .
then @xmath209 is safe for @xmath315 and @xmath316 the lemma shows that the set of traces produced by @xmath209 can be obtained by combining pairs of traces with the same history produced by @xmath166 and @xmath157 .
note that , since the semantics of @xmath209 does not annotate calls and returns with the states transferred , in @xmath317 we have to erase these annotations from the local traces @xmath263 or @xmath318 before comparing the traces with @xmath125 .
unpacking the definition of @xmath319 and using the fact that @xmath320 from lemma [ prop : sem : local - global ] we get the following two corollaries . [ lemma - clientlocal ] assume @xmath202 and @xmath312 safe for @xmath313 and @xmath314 , respectively .
then @xmath209 is safe for @xmath315 and @xmath321 [ lemma - clientlocal2 ] if @xmath202 and @xmath312 are safe for @xmath313 and @xmath314 , respectively , then @xmath322 corollary [ lemma - clientlocal ] can be viewed as carrying over properties of the local semantics , such as safety , to the global one , and in this sense is the statement of the soundness of the former with respect to the latter .
the corollary also confirms that the client defined by @xmath282 and @xmath323 is indeed most general , as it reproduces library behaviours under any possible clients .
corollary [ lemma - clientlocal2 ] carries over properties of the global semantics to the local ones , stating the adequacy of the latter .
lemma [ prop : sem : local - global ] is proved in appendix [ sec : proofs ] .
most of the proof deals with maintaining a splitting of the state of @xmath324 into the parts owned by @xmath157 and @xmath166 , which changes during ownership transfers .
the proof relies crucially on the safety of the client and the libraries and the strong locality property of primitive commands . in more detail
, safety is defined by considering executions of a component in the library - local or the client - local semantics .
these execute the component code only on the memory it owns , whose amount only changes with ownership transfers to and from its environment according to method specifications . because of the strong locality property , commands fault when accessing memory cells that are not present in the state they are run from , and their execution does not depend on any additional memory that might be present in the state .
hence , when we use a component inside a complete program , its safety guarantees that the component code does not touch the part of the heap belonging to other components in the program , and its execution is not affected by the state of such components .
this guarantees that the behaviour a component produces as part of the complete program can be reproduced when we execute it in isolation and vice versa , allowing us to establish lemma [ prop : sem : local - global ] . in practice , the safety of a program can be established using existing program logics , such as separation logic @xcite .
we are now in a position to define the notion of linearizability on libraries and prove the central technical result of this paper the abstraction theorem .
we define linearizability between specified libraries @xmath312 , together with their sets of initial states @xmath325 .
first , using the library - local semantics of section [ sec : semantics ] , we define the interface set describing all the behaviours of a library @xmath157 when run from initial states in @xmath325 : @xmath326 [ lin2 ] consider @xmath327 and @xmath328 safe for @xmath314 and @xmath329 , respectively .
we say that @xmath330 * _ is linearized by _
* @xmath331 , written @xmath332 , if , according to definition [ lin ] , @xmath333 for an interface set @xmath149 we say that @xmath330 * _ is linearized by _
* @xmath149 , written @xmath334 , if @xmath335 thus , @xmath330 is linearized by @xmath331 if every history generated by the library - local semantics of the former may be reproduced in a linearized form by the library - local semantics of the latter without requiring more memory .
the relation @xmath336 allows us to specify a library by another piece of code , but possibly simpler than the original one .
for example , the stack and the allocator from figure [ fig : impl ] with method specifications ( [ container ] ) and ( [ alloc ] ) can be specified by the libraries in figure [ fig : spec ] .
the libraries replace the array and the linked list in the implementations by the abstract data types of a sequence and a set ( we assume a trivial extension of the @xmath18 algebra from section [ sec : prelim ] to allow memory cells to store values of such types ) .
thus , the abstract libraries use less memory than the concrete ones . instead of using locking , all operations on the abstract data types
are done atomically ; formally , we assume primitive commands corresponding to the code in the atomic blocks .
@l@|@l@ sequence < void * > stack ; int push(void * arg ) atomic if ( nondet ( ) ) return full ; else add_to_head(stack , arg ) ; return ok ; void * pop ( ) atomic if ( !
isempty(stack ) ) void * obj = head(stack ) ; stack = tail(stack ) ; return obj ; else return empty ; & set < void * > free_list ; void free(void * arg ) atomic add(free_list , arg ) ; void * alloc ( ) atomic if ( ! isempty(free_list ) ) node * block = ( node*)take(free_list ) ; block->next = nondet ( ) ; block->prev = nondet ( ) ; return block ; else return 0 ; + + ( a ) & ( b ) the other relation @xmath334 introduced in definition [ lin2 ] allows us to specify a library directly by an interface set , without fixing a piece of code generating it .
the interface set @xmath149 can still be simpler than that of @xmath330 , e.g. , containing only sequential histories .
even though the two forms of defining linearizability may seem very similar , as we show in section [ sec : rearr ] , their mathematical properties are fundamentally different .
we now formulate two variants of the abstraction theorem , corresponding to the two ways of specifying libraries ( we prove them in section [ sec : rearr ] ) .
[ thm2 ] if @xmath337 @xmath327 , @xmath328 , @xmath202 are safe for @xmath314 , @xmath329 , @xmath325 , respectively , and @xmath338 , then @xmath337 @xmath339 and @xmath340 are safe for @xmath341 and @xmath342 , respectively , and @xmath343 thus , when reasoning about a client @xmath344 of a library @xmath345 , we can soundly replace @xmath345 by a library @xmath346 linearizing it : if a safety property over client traces holds of @xmath347 , it will also hold of @xmath344 . in practice , we are usually interested in * _ atomicity abstraction _ * , a special case of this transformation when methods in @xmath346 are atomic .
an instance is replacing one of the libraries from figure [ fig : impl ] by its specification from figure [ fig : spec ] .
the requirement that @xmath166 be safe in the theorem restricts its applicability to well - behaved clients that do not access memory owned by the library : you can not replace a library by another one if the client can access its internal data structures and thereby `` look inside the box '' .
similarly , the safety of the libraries ensures that they can not corrupt the data structures owned by the client .
the other version of the abstraction theorem , allowing library specification by an interface set , guarantees that replacing a library by its specification leaves all the original client behaviours reproducible modulo the following notion of trace equivalence .
client traces @xmath263 and @xmath348 are * _ equivalent _ * , written @xmath349 , if @xmath350 for all @xmath185 and the projections of @xmath263 and @xmath348 to non - interface actions are identical .
[ thm - spec ] if @xmath337 @xmath327 and @xmath202 are safe for @xmath314 and @xmath325 , respectively , and @xmath351 , then @xmath337 @xmath339 is safe for @xmath341 and @xmath352 the theorem shows that client behaviours of @xmath339 can be reproduced by the client - local semantics of @xmath166 projected to histories in @xmath149 with initial footprints compatible with initial client states .
note that @xmath353 in theorem [ thm2 ] implies that @xmath354 , i.e. , @xmath340 can reproduce the history of @xmath339 exactly .
in contrast , theorem [ thm - spec ] does not guarantee this , since @xmath355 does not imply @xmath356 ; we only know that the projection to non - interface actions is reproduced .
we discuss the reason for this discrepancy below .
the key component used for establishing theorem [ thm2 ] is the rearrangement lemma : if @xmath139 , then every execution trace of a library producing @xmath147 can be transformed into another trace of the same library that differs from the original one only in the order of interface actions and produces @xmath120 , instead of @xmath147 .
hence , the library specification can simulate any behaviour of its implementation the client can expect .
[ cor - rearr ] if @xmath357 and @xmath312 is safe at @xmath358 , then @xmath359 the proof of the lemma is highly non - trivial and is a subject of section [ sec : rearr - proof ] .
we point out lemma [ cor - rearr ] would not hold had we included unbalanced histories in our definition of linearizability . to show this ,
take @xmath129 and consider the histories in figure [ fig : cex ] . in figure
[ fig : cex](b ) the library receives the cell @xmath130 from the client , then returns it and then receives it again .
even though the history in figure [ fig : cex](a ) is linearized by that in figure [ fig : cex](b ) , the former is not balanced , and by proposition [ safe2wb ] , can not be produced by @xmath157 .
this shows that lemma [ cor - rearr ] does not hold for unbalanced @xmath120 .
note that we have @xmath360 for any history @xmath120 . as a consequence , from lemma [ cor - rearr ]
we obtain the following surprising result , stating that linearizability between libraries is equivalent to inclusion between the sets of histories they produce .
[ cor : subset ] if @xmath327 and @xmath361 are safe for @xmath362 and @xmath363 , respectively , then @xmath364 this fact is not a consequence of ownership transfer and also holds for the classical notion of linearizability .
intuitively , lemma [ cor - rearr ] , and hence , corollary [ cor : subset ] hold due to a closure property of the semantics of the language from section [ sec : prog ] .
namely , in this and other programming languages , there may always be a delay between the point when a library method is called and when it starts executing and , conversely , when it ends executing and when the control returns to the client .
for example , when executing the code in figure [ fig : spec](a ) , there may be delays between a call to push , the execution of the atomic block and the return from push .
hence , this library can produce both of the histories in figures [ fig : hist](b ) and [ fig : hist](c ) .
due to this property of the program semantics , a trace from @xmath365 , e.g. , one producing the history in figure [ fig : hist](c ) , will stay valid if we execute some of the calls in it earlier and returns later , like in figure [ fig : hist](b ) .
this is also ( usually ) safe given the ownership transfer reading of calls and returns in the library - local semantics defined by @xmath366 : it just means that the library receives state from the client earlier and gives it up later .
the proof of lemma [ cor - rearr ] uses such transformations on a history to `` de - linearize '' it , e.g. , transforming the history in figure [ fig : hist](c ) into that in figure [ fig : hist](b ) .
the above closure property is also the reason for theorem [ thm2 ] guaranteeing that @xmath340 can reproduce the history of @xmath339 exactly . given that replacing a library @xmath345 by its linearization @xmath346 does not simplify its interface set , can theorem [ thm2 ] really simplify reasoning about a complete program @xmath339 ?
fortunately , the answer is yes , since the point of the theorem is to simplify _ the code _ of this program .
for example , replacing the library in figure [ fig : impl](a ) by the one in figure [ fig : spec](a ) allows us to pretend in reasoning about a complete program that changes to the library state , shared between different threads , are atomic , and thus consider fewer possible thread interleavings .
calls and returns in such a complete program are merely thread - local operations that do not complicate reasoning . in section
[ sec : example ] , we discuss an example of using theorem [ thm2 ] to simplify proofs of complicated algorithms . specifying a library by an interface set @xmath149 instead of code , as in theorem [ thm - spec ] , does not allow us to get results such as lemma [ cor - rearr ] and corollary [ cor : subset ] , since the set @xmath149 is not guaranteed to satisfy any closure properties .
for example , it might contain only sequential histories , where every call is immediately followed by the corresponding return without a delay . in fact , @xmath149 has to be simpler than the interface set of @xmath345 for theorem [ thm - spec ] to be useful , since this is what the theorem replaces @xmath345 by . fortunately , to prove theorem [ thm - spec ] we can exploit a closure property of the client - local semantics , formalised by the following variant of the rearrangement lemma : any client trace can be transformed into an equivalent one with a given history linearizing the history of the original one
. [ thm ] if @xmath143 and @xmath202 is safe at @xmath38 , then @xmath367 intuitively , the lemma holds because , in the client - local semantics , it is safe to execute calls later and returns earlier . like lemma [ cor - rearr ] , this lemma would not hold if we allowed @xmath147 to be unbalanced . in summary ,
when a library is specified by the code of its abstract implementation , the ability to linearize a concrete history while looking for a matching abstract one allowed by definition [ lin ] is not strictly needed .
however , it is indispensable when the library is specified directly by a set of histories .
we were able to obtain this insight into the original definition of linearizability by formalising the guarantees the linearizability of a library provides to its clients as abstraction theorems .
using lemmas [ cor - rearr ] and [ thm ] , we now prove the two versions of the abstraction theorem .
[ [ proof - of - theoremthm2 . ] ] _ proof of theorem [ thm2 ] .
_ + + + + + + + + + + + + + + + + + + + + + + + + + + the safety of @xmath339 and @xmath340 follows from corollary [ lemma - clientlocal ] .
take @xmath368 .
we transform the trace @xmath369 of @xmath339 into a trace @xmath370 of @xmath340 with the same client projection using the local semantics of @xmath345 , @xmath346 and @xmath166 .
namely , we first apply corollary [ lemma - clientlocal ] to generate a pair @xmath371 of a library - local initial state and a trace and a client - local pair @xmath372 , such that [ eq - decomp ] = _ * c**_*l*^1 client(_1)=ground ( ) history ( ) = history(_1 ) . since @xmath373 , for some @xmath374
, we have @xmath375 which implies @xmath376 . by lemma [ cor - rearr ]
, @xmath377 can be transformed into a trace @xmath378 such that @xmath379 since @xmath376 and @xmath380 , we have @xmath381 .
we then use corollary [ lemma - clientlocal2 ] to compose the library - local trace @xmath378 with the client - local one @xmath263 into a trace @xmath370 such that @xmath382 the above proof scheme can be described mnemonically as ` decompose , rearrange , compose ' .
we reuse its first two steps to prove theorem [ thm - spec ] .
[ [ proof - of - theoremthm - spec . ] ] _ proof of theorem [ thm - spec ] . _ +
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + take @xmath383 .
like in the proof of theorem [ thm2 ] , we apply corollary [ lemma - clientlocal ] to generate @xmath384 and @xmath385 such that ( [ eq - decomp ] ) holds . since @xmath386 ,
for some @xmath387 , we have @xmath388 then by lemma [ thm ] , @xmath263 can be transformed into a trace @xmath348 , such that @xmath389 since @xmath390 , we thus have @xmath391 . furthermore , since @xmath392 and @xmath393 , we have @xmath394 . hence , @xmath348 and @xmath96 are the required trace and footprint . our preliminary investigations show that linearizability with ownership transfer can be established by generalising existing proof systems for proving classical linearizability based on separation logic @xcite .
the details of such a generalisation are out of the scope of this paper ; we plan to report on it in the future .
the abstraction theorem is not just a theoretical result : it enables compositional reasoning about complex concurrent algorithms that are challenging for existing verification methods .
for example , the theorem can be used to justify vafeiadis s compositional proof ( * ? ? ? * section 5.3 ) of the multiple - word compare - and - swap ( mcas ) algorithm implemented using an auxiliary operation called rdcss @xcite ( the proof used an abstraction of the kind enabled by theorem [ thm2 ] without justifying its correctness ) .
if the mcas algorithm were verified together with rdcss , its proof would be extremely complicated .
fortunately , we can consider mcas as a client of rdcss , with the two components performing ownership transfers between them .
the abstraction theorem then makes the proof tractable by allowing us to verify the linearizability of mcas assuming an atomic specification of the inner rdcss algorithm .
we only give the proof of lemma [ cor - rearr ] , as that of lemma [ thm ] is completely symmetric .
the proof transforms @xmath395 into @xmath318 by repeatedly swapping adjacent actions in it according to a certain strategy to make the history of the trace equal to @xmath120 .
the most subtle place in the proof is swapping @xmath396 to yield @xmath397 where @xmath398 .
this case is subtle for the following reason .
let the state of the library @xmath157 before the return action be @xmath399 ; then @xmath400 and the state of the library after executing the return and the call is @xmath401 . for the swapping to be possible , we need @xmath402 ; then by proposition [ prop - diff ] [ theta - swap ] ( _ 1 ) * _ 2 = ( _ 2)_1 , which can be used to establish that the resulting trace is still produced by @xmath157 . however , @xmath402 is not guaranteed if the history @xmath120 is arbitrary .
for example , take @xmath403 and let @xmath120 and @xmath147 be defined by figures [ fig : cex](a ) and [ fig : cex](b ) .
since @xmath120 is unbalanced , it can not be produced by any library , and hence , we can not swap @xmath404))\,(2 , { { \sf call}}\ m_2([10:0]))\ ] ] in @xmath147 . in our proof
we use the fact that the history @xmath120 is balanced to show that a situation in which we can not swap a return followed by a call while transforming @xmath395 into @xmath318 can not happen .
this is non - trivial , as the problematic situation can potentially happen midway through the transformation .
we only know that the target history @xmath120 of @xmath318 is balanced , but this does not straightforwardly imply that the histories of the intermediate traces obtained while transforming @xmath395 into @xmath318 are , since these histories might be quite different from @xmath120 . inferring their balancedness from that of @xmath120 represents the most challenging part of the proof .
we therefore first do the proof under an assumption that allows swapping a return followed by a call easily and consider the general case later .
this lets us illustrate the overall idea of the proof , which is then reused in the additional part of the proof dealing with the challenge presented by the general case .
namely , we make the following assumption : [ assm ] @xmath129 and for any @xmath405 and interface actions @xmath406 and @xmath407 in @xmath408 , if @xmath398 , then @xmath409 . for example , this holds when states transferred between the client and the library are always thread - local .
it is easy to check that in @xmath18 , if @xmath400 , @xmath410 and @xmath411 , then @xmath402 and thus ( [ theta - swap ] ) holds .
hence , ( [ assm ] ) allows us to justify swapping a return followed by a call in a trace easily .
we now proceed to prove lemma [ cor - rearr ] under this assumption . in our proof
, we use the assumption in a single place , which we note explicitly ; the rest of the proof is independent from it .
below we sometimes write @xmath412 instead of @xmath138 to make the bijection @xmath413 used to establish the relation between histories in definition [ lin ] explicit . for a bijection @xmath413 between histories
@xmath120 and @xmath147 , we write @xmath414 if @xmath413 is an identity on the first @xmath127 actions in @xmath120 .
take @xmath415 and consider a trace @xmath416 .
assume histories @xmath417 such that @xmath418 and @xmath419 , so that @xmath120 is balanced from @xmath299 and @xmath147 from @xmath420 .
we prove that there exists a trace @xmath421 such that @xmath422 . to this end
, we define a finite sequence of steps that transforms @xmath395 into a such a trace @xmath318 .
the main idea of the transformation is to make progressively longer prefixes of the trace have histories coinciding with prefixes of @xmath120 .
namely , the transformation is done in stages , and on stage @xmath423 we obtain a trace @xmath424 , where @xmath425 .
every one of these traces is such that for some prefix @xmath426 of @xmath427 we have : @xmath428 we let @xmath429 , so that the above conditions are initially satisfied . thus , during the transformation , progressively longer prefixes @xmath426 of @xmath427 have histories coinciding with prefixes of @xmath120 , while the linearizability relation between the history @xmath120 and that of @xmath427 is preserved .
we then take @xmath430 as the desired trace @xmath318 .
the trace @xmath431 is constructed from the trace @xmath427 by applying the following lemma for @xmath432 , @xmath433 , @xmath434 , @xmath435 , @xmath436 and @xmath437 .
[ lemma ] assume ( [ assm ] ) holds .
consider a history @xmath438 and a trace @xmath439 such that @xmath440 then there exist traces @xmath378 and @xmath441 such that @xmath442 and @xmath443 to prove lemma [ lemma ] , we convert @xmath444 into @xmath445 by swapping adjacent actions in the trace a finite number of times while preserving its properties of interest .
these transformations are described by the following proposition , which formalises the closure properties of the library - local semantics we alluded to in section [ sec : rearr ] .
[ prop ] let @xmath312 be safe at @xmath446 and consider @xmath447 and a history @xmath448 such that @xmath449 . then swapping any two adjacent actions @xmath450 in @xmath408 executed by different threads such that 1 .
@xmath451 , @xmath452 ; or 2 .
@xmath453 , @xmath452 , @xmath454 precedes @xmath455 in @xmath448 , and ( [ assm ] ) holds ; or 3 .
@xmath456 , @xmath457 yields a trace @xmath458 such that @xmath459 for the bijection @xmath460 defined as follows . if @xmath461 or @xmath462 , then @xmath463
otherwise , let @xmath124 be the index of @xmath455 in @xmath448 .
then @xmath464 , @xmath465 and @xmath466 for @xmath467 .
since , in the library - local semantics , the library gains state at a call and gives it up at a return , intuitively , the transformation in the proposition allows the library to gain state earlier ( i , ii ) and give it up later ( iii ) .
the assumption that @xmath454 precede @xmath455 in case ( ii ) is needed to ensure that the transformation does not violate the linearizability relation .
the proof of case ( ii ) is the only place where the assumption ( [ assm ] ) is used .
[ [ proof - sketch - for - propositionprop . ] ] _ proof sketch for proposition [ prop ] . _ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + consider @xmath468 and let @xmath469 .
the proof of the required linearizability relationship is trivial .
it therefore remains to show that @xmath470 .
we know that for some @xmath471 we have @xmath472 and @xmath473 .
let @xmath474 and @xmath475 , where @xmath476 and @xmath477 correspond to @xmath455 and @xmath454 .
it is easy to see that @xmath478 .
it therefore remains to show that @xmath479 .
the proof proceeds by case analysis on the kind of actions @xmath455 and @xmath454 .
the justification of the case when @xmath455 is a return and @xmath454 is a call follows from ( [ assm ] ) by the argument given earlier . out of the remaining cases ,
we only consider a single illustrative one : @xmath480 and @xmath481 for @xmath398 .
assume @xmath482 .
then for some @xmath399 we have @xmath483 , @xmath484 and @xmath485 . by the footprint preservation property , we get @xmath486 . then by the strong locality property , @xmath487 hence , @xmath488 .
thus , footprint preservation and strong locality guarantee that a call can be safely executed earlier than a primitive command .
[ [ proof - of - lemmalemma . ] ] _ proof of lemma [ lemma ] . _ + + + + + + + + + + + + + + + + + + + + + + + + + from ( [ 1 ] ) and ( [ 2 ] ) it follows that @xmath489 for some traces @xmath490 and @xmath491 , where @xmath413 maps @xmath104 in @xmath438 to the @xmath104 action shown in @xmath377 ; see figure [ fig : diag ] .
we consider two cases .
@xmath492 . let @xmath114 be the thread executing @xmath104 .
by ( [ 2 ] ) we have @xmath493 .
then , since @xmath444 is a library trace , ( [ 1 ] ) implies that there are no actions by thread @xmath114 in @xmath490 .
furthermore , for any return action @xmath293 in @xmath490 , the action in @xmath438 corresponding to it according to @xmath413 is in @xmath494 .
thus , we can move the action @xmath104 to the position between @xmath495 and @xmath490 by swapping it with adjacent actions a finite number of times as described in proposition [ prop](i , ii ) . as a result
, we obtain the trace @xmath496 .
conditions ( [ 4])([5 ] ) then follow from proposition [ prop](i , ii ) for @xmath497 and @xmath498 .
assume that @xmath490 contains a call action @xmath293 , so that it precedes the return action @xmath104 in @xmath500 . then by ( [ 1 ] ) and ( [ 2 ] ) the action in @xmath501 corresponding to @xmath293 according to @xmath413 is in @xmath494 and thus follows @xmath104 in @xmath438 .
this violates the preservation of the order of non - overlapping method invocations required by ( [ 2 ] ) .
hence , there are no call actions in @xmath490 .
since @xmath444 is a library trace , this implies that for any action @xmath502 in @xmath490 there are no actions by the thread @xmath114 in @xmath490 following @xmath293 .
thus , we can move all return actions in the subtrace @xmath490 of @xmath444 to the position between @xmath104 and @xmath491 by swapping them with adjacent actions a finite number of times as described in proposition [ prop](iii ) .
we thus obtain the trace @xmath503 , where @xmath504 consists of all return actions in @xmath490 , and @xmath505 of the rest of actions in the subtrace ; in particular , @xmath505 does not contain any interface actions .
conditions ( [ 4])([5 ] ) then follow from proposition [ prop](iii ) for @xmath506 and @xmath507 .
this completes the proof of lemma [ cor - rearr ] under the assumption ( [ assm ] ) .
let us now lift this assumption and consider the only place in the proof of the lemma that relies on it that when we swap a return followed by a call using proposition [ prop](ii ) in case 1 in the proof of lemma [ lemma ] .
let us now identify precise conditions under which this situation happens .
let @xmath508 and let the adjacent return action with which we are trying to swap it be @xmath509 .
let @xmath510 be the target history @xmath438 from lemma [ lemma ] and @xmath511 be the history of the trace in which we are trying to swap the return and the call .
then the two histories are of the form [ 2hist ] r@ c@ l s^1 & = & s ( t_1,call m_1(_1 ) ) s_1 ( t_2,ret m_2(_2 ) ) s_2 ; + s^2 & = & s s_1 ( t_2,ret m_2(_2 ) ) ( t_1,call m_1(_1 ) ) s_2 for some @xmath512 . furthermore , from the conditions of lemma [ lemma ]
, we have : a. [ cond0 ] @xmath513 ; b. @xmath510 and @xmath511 are balanced from some @xmath95 and @xmath96 , respectively , such that @xmath514 ; and c. @xmath515 , where @xmath516 and @xmath413 maps @xmath517 and @xmath509 in @xmath510 to the corresponding actions shown in @xmath511 .
as the following proposition shows , we can always do the desired transformation if the history [ hist1 ] s s_1 ( t_1,call m_1(_1 ) ) ( t_2,ret m_2(_2 ) ) s_2 resulting from swapping the return and the call in @xmath511 is balanced from @xmath96 .
[ cor - call - ret ] let @xmath312 be safe at @xmath446 .
consider traces @xmath518 and @xmath519 if @xmath520 is balanced from @xmath521 , then @xmath522 .
thus , the only problematic case we have is when the history ( [ hist1 ] ) is not balanced from @xmath96 .
we summarise all the conditions under which such case can happen in the following definition .
[ def - confl ] histories @xmath510 and @xmath511 of the form ( [ 2hist ] ) are * _ conflicting _ * if the conditions ( i)(iii ) above are satisfied and the history ( [ hist1 ] ) is not balanced from @xmath96 . given proposition [ cor - call - ret ] , the only case when the transformation in the proof of lemma [ lemma ] can fail to convert the trace is when @xmath501 and the history currently being transformed are conflicting .
thus , with the assumption ( [ assm ] ) lifted , lemma [ lemma ] turns into [ lemma2 ] consider a history @xmath438 , and a trace @xmath523 such that ( [ 1 ] ) and ( [ 2 ] ) hold .
then either @xmath438 and another history composed of actions from @xmath500 are conflicting , or there exist traces @xmath378 and @xmath441 such that @xmath442 and ( [ 4 ] ) and ( [ 5 ] ) hold . we now show that no conflicting pairs of histories exist , hence guaranteeing that lemma [ lemma2 ] can always be used to construct @xmath431 from @xmath427 in transforming @xmath395 into @xmath318 .
this completes the proof of lemma [ cor - rearr ] in the general case .
we first discuss the main idea of the proof .
the history @xmath510 in ( [ 2hist ] ) is similar to ( [ hist1 ] ) in that the call precedes the return .
we would like to use the fact that @xmath510 is balanced to prove that so is ( [ hist1 ] ) , thereby yielding a contradiction . as we noted at the beginning of this section , this is not straightforward due to the differences in the form of the histories @xmath510 and @xmath511 other than the precedence of the two call and return actions .
we resolve this problem by adjusting the strategy we used above to transform @xmath395 into @xmath318 under the assumption ( [ assm ] ) to iron out the differences between @xmath510 and @xmath511 .
in particular , we use a variant of the transformation that , when the process of moving the call action @xmath104 to the left in @xmath444 gets stuck ( figure [ fig : diag ] ) , leaves the corresponding action in @xmath438 unmatched and continues bringing the rest of the trace @xmath444 in sync with the target history .
[ call - ret ] there are no conflicting pairs of histories .
consider histories @xmath510 and @xmath511 satisfying the conditions in definition [ def - confl ] . since @xmath511 is balanced from @xmath96 , @xmath524 is defined .
assume @xmath525 is defined .
since @xmath526 is defined , by proposition [ prop - delta ] , we have : @xmath527 then ( [ hist1 ] ) is balanced from @xmath96 , contradicting our assumptions . hence , @xmath528 .
a call action in @xmath529 can not be in @xmath530 : in this case it would follow @xmath509 in @xmath510 , but precede it in @xmath511 , contradicting @xmath531 . hence , all call actions in @xmath529 are in @xmath532 .
let @xmath533 , where @xmath534 is the minimal prefix of @xmath532 containing all call actions from @xmath529 . then @xmath535 s^2 & = & s s'_1\ , ( t_2,{{\sf ret}}\ m_2(\sigma_2))\ , ( t_1,{{\sf call}}\ m_1(\sigma_1 ) ) \,s'_2 . \end{array}\ ] ] if @xmath534 is non - empty , any return action in it precedes its last call action , which is also in @xmath529 . since @xmath531
, such a return action also has to be in @xmath529 .
thus , all return actions in @xmath534 are in @xmath529 .
the traces @xmath510 and @xmath511 are of the following more general form , obtained by letting @xmath536 and @xmath537 : @xmath538 s^2 & = & s'_0 s'_1\ , ( t_2,{{\sf ret}}\ m_2(\sigma_2))\ , ( t_1,{{\sf call}}\ m_1(\sigma_1 ) ) \,s'_2 , \end{array}\ ] ] where @xmath337 @xmath510 and @xmath511 are balanced from some @xmath95 and @xmath96 , respectively , such that @xmath514 ; @xmath539 and @xmath540 are identical , except @xmath539 may have some extra call actions ; @xmath515 ; @xmath541 maps all call actions in @xmath529 to actions in @xmath534 ; @xmath413 maps all return actions in @xmath534 to actions in @xmath529 ; @xmath541 maps actions in @xmath540 to those in @xmath539 , in particular @xmath542 to an action in @xmath539 , and @xmath509 to the same action shown in @xmath510 ; and @xmath543 .
we denote this form by ( * f * ) .
the additional call actions in @xmath539 are the ones for which the transformation in lemma [ lemma ] failed .
the conditions relating @xmath534 and @xmath529 imply that @xmath534 may have more calls than @xmath529 , and @xmath529 more returns than @xmath534 .
thus , intuitively , @xmath544 gains more state than @xmath545 , including that transferred by @xmath517 , and @xmath545 gives up more than @xmath544 . in the following , we use this and the fact that @xmath510 is balanced from a bigger footprint than @xmath511 to show that @xmath546 , thereby yielding a contradiction . to this end
, we describe a process that transforms the histories @xmath510 and @xmath511 into another pair of histories satisfying the conditions above , but such that @xmath534 is strictly smaller . repeatedly applying this process
, we can make @xmath534 empty , obtaining histories satisfying ( * f * ) : [ hist - temp ] l s_0 s_4 ( t_2,ret m_2(_2 ) ) s_2 ; + s_0 s_1 ( t_2,ret m_2(_2 ) ) ( t_1,call m_1(_1 ) ) s_2 .
in particular , @xmath543 . before describing the transformation process ,
we show that , given the above pair of histories , we can obtain a contradiction .
we use the following simple proposition , proved in appendix [ sec : proof - foot ] .
[ prop - foot ] assume @xmath448 is identical to @xmath547 , except it may have extra calls , and @xmath448 and @xmath547 are balanced from @xmath95 and @xmath96 , respectively , such that @xmath548 .
then the @xmath56-combination @xmath549 of footprints of states transferred at the extra call actions in @xmath448 is defined , @xmath547 is balanced from @xmath95 and @xmath550 consider the histories in ( [ hist - temp ] ) . since all calls from @xmath529 are in @xmath551 , @xmath529 contains only returns .
since the histories are balanced from @xmath95 and @xmath96 , respectively , @xmath552 and @xmath553 are defined .
the history @xmath539 is identical to @xmath540 , except it may have extra calls . by proposition [ prop - foot ] , the @xmath56-combination of footprints of states transferred at the extra call actions in @xmath539 is defined . since an action @xmath517 is in @xmath539 , but not in @xmath540 , this combination is of the form @xmath554 for some @xmath549 ; hence , @xmath555 therefore , @xmath556 is defined . since @xmath529 contains only return actions , @xmath557 where @xmath558 is the @xmath559-combination of the footprints of states transferred at these actions .
this implies @xmath560 where both expressions are defined .
but then so is @xmath561 hence , @xmath562 is defined , contradicting the opposite fact established above .
this contradiction implies that a conflicting pair of histories does not exist
. now assume arbitrary histories @xmath510 and
@xmath511 satisfying ( * f * ) : @xmath538 s^2 & = & s'_0 s'_1\ , ( t_2,{{\sf ret}}\ m_2(\sigma_2))\ , ( t_2,{{\sf call}}\ m_1(\sigma_1 ) ) \,s'_2 , \end{array}\ ] ] we show that from these we can construct another pair of histories satisfying ( * f * ) , but with @xmath534 strictly smaller .
we use the same transformation as in the proof of lemma [ lemma ] .
when this transformation gets stuck , we obtain another pair of histories of the form ( * f * ) , but again with a smaller @xmath534 .
let us make a case split on the next action in @xmath534 .
@xmath337 @xmath563 , such that the action corresponding to @xmath105 according to @xmath413 is not in @xmath529 . in this case
we let @xmath564 and @xmath565 .
thus , the call action @xmath105 unmatched in @xmath529 becomes part of @xmath539 .
@xmath563 , such that the action corresponding to @xmath105 according to @xmath413 is in @xmath529 .
let @xmath566 , so that @xmath567 s^2 & = & s'_0 s'_3\ , ( t , { { \sf call}}\ m(\sigma))\ , s'_4\ , ( t_2,{{\sf ret}}\ m_2(\sigma_2))\ , ( t_2,{{\sf call}}\ m_1(\sigma_1 ) ) \,s'_2 . \end{array}\ ] ] using the transformations from case 1 in the proof of lemma [ lemma ] , we can try to move the action @xmath105 to the position between @xmath540 and @xmath568 while preserving the balancedness of the history .
if this succeeds , we construct a new pair of histories of the form ( * f * ) by letting @xmath564 , @xmath569 , @xmath570 and @xmath571 .
otherwise , we get a pair of conflicting histories , which are of the form ( * f * ) but with a smaller @xmath534 . again , in this case the unmatched call action @xmath105 becomes part of @xmath539 .
then the action corresponding to @xmath573 according to @xmath413 is also in @xmath529 , so that @xmath574 : @xmath575 s^2 & = & s'_0 s'_3\ , ( t , { { \sf ret}}\ m(\sigma))\ , s'_4\ , ( t_2,{{\sf ret}}\ m_2(\sigma_2))\,(t_2,{{\sf call}}\ m_1(\sigma_1))\ , s'_2 . \end{array}\ ] ] using the transformations from case 2 in the proof of lemma [ lemma ] , we can move the return action to the position between @xmath540 and @xmath568 while preserving the balancedness of the history .
we thus obtain a pair of histories : @xmath576 s^2 & = & s'_0\ , ( t , { { \sf ret}}\ m(\sigma))\ , s'_3 s'_4\ , ( t_2,{{\sf ret}}\ m_2(\sigma_2))\,(t_2,{{\sf call}}\ m_1(\sigma_1 ) ) \,s'_2 . \end{array}\ ] ] then we can let @xmath577 , @xmath569 , @xmath578 and @xmath571 .
libraries such as concurrent containers are used by clients to transfer the ownership of data structures , but do not actually access their contents .
we show that for such libraries , the classical linearizability implies linearizability with ownership transfer .
[ def : gamma2 ] a method specification @xmath579 * _ extends _ * a specification @xmath580 , if @xmath581 .
for example , the method specification ( [ container ] ) of the stack in figure [ fig : impl](a ) extends the following specification : [ container - val ] l \{x .
arg_t x } push \{arg_t arg_t } ; + \{y .
arg_t y } pop \{x .
arg_t x}. according to this specification ,
push just receives an arbitrary pointer @xmath11 as a parameter ; in contrast , the specification ( [ container ] ) additionally mandates that the object the pointer identifies be transferred to the library .
we now identify conditions under which the linearizability between a pair of libraries satisfying @xmath171 entails that of the same libraries satisfying an extended method specification @xmath206 .
this yields a result somewhat analogous to the frame rule of separation logic @xcite .
we start by introducing some auxiliary definitions .
we first define operations for mapping between histories corresponding to extended and non - extended method specifications . for the method specification @xmath171 from definition [ def : gamma2 ] , we define operations @xmath582 and @xmath583 on interface actions as follows : @xmath584 \llfloor(t , { { \sf ret}}\ m(\sigma))\rrfloor_{\gamma } & = & ( t , { { \sf ret}}\ m(\sigma { \mathop{\backslash}}{(\sigma{\mathop{\backslash}}q^m_t ) } ) ) ; \\[2pt ] \llceil ( t , { { \sf call}}\ m(\sigma ) ) \rrceil_{\gamma } & = & ( t , { { \sf call}}\ m(\sigma{\mathop{\backslash}}p^m_t ) ) ; \\[2pt ] \llceil ( t , { { \sf ret}}\ m(\sigma ) ) \rrceil_{\gamma } & = & ( t , { { \sf ret}}\ m(\sigma{\mathop{\backslash}}q^m_t ) ) ; \end{array}\ ] ] otherwise , the result is undefined .
thus , @xmath585 selects the part of the state in @xmath104 that is required by @xmath171 and @xmath586 the extra piece of state not required by it .
we then lift @xmath587 and @xmath588 to traces by applying them to every interface action . given a history @xmath589 produced by a library @xmath281 , we need to be able to check that the library does not modify the extra pieces of state not required by the original method specification @xmath171 , which are given by @xmath590 . to this end
, we define an evaluation function similar to @xmath591 from section [ sec : observ ] , which is meant to be applied to @xmath592 . for an interface action
@xmath104 we define @xmath593 as follows : @xmath594 { \langle ( t,{{\sf ret}}\;m(\sigma_0 ) ) \rangle}\sigma & = & \mbox{if } { ( \sigma { \mathop{\backslash}}\sigma_0){\mathpunct{\downarrow } } } \mbox { then } \sigma { \mathop{\backslash}}\sigma_0 \mbox { else } \top .
\end{array}\ ] ] we then define the evaluation @xmath595 of a history @xmath120 as follows : @xmath596 thus , if the evaluation does not fail , then the history respects the notion of ownership and the pieces of state transferred to the library are returned to the client unmodified .
[ thm : frame ] assume 1 .
@xmath206 extends @xmath171 ; 2 . for all @xmath597 , @xmath598 and @xmath599
are safe for @xmath600 and @xmath601 , respectively ; 3 . @xmath602 ; and 4 . for every @xmath603 , @xmath604 and @xmath605
, we have @xmath606 . then @xmath607 .
the theorem allows us to establish the linearizability relation with respect to the extended specification @xmath206 given the relation with respect to @xmath171 .
this enables the use of the abstraction theorem for clients performing ownership transfer .
however , the theorem does not guarantee the safety of the libraries with respect to @xmath206 for free , because it is not implied by their safety with respect to @xmath171 .
intuitively , this is because @xmath206 extends both preconditions and postconditions in @xmath171 ; hence , not only does it guarantee to the library that the client will provide extra pieces of state at calls , but it also requires the library to provide ( possibly different ) extra pieces of state at returns . for example , @xmath171 might assign the specification @xmath608 to every method @xmath113 , and @xmath206 , the specification @xmath609 . unless a library already has all the memory required by the postconditions in @xmath206 in its initial state
, it has no way of satisfying @xmath206 .
this situation is in contrast to the frame rule of separation logic @xcite , which guarantees the safety of a piece of code with respect to an extended specification .
however , the frame rule requires the latter specification to extend both pre- and postconditions with the same piece of state , so that the code returns it immediately after termination . in our setting , a library can return the extra state to its client after a different method invocation and , possibly , in a different thread . finally ,
condition ( 4 ) in theorem [ thm : frame ] ensures that the extra memory required by postconditions in @xmath206 comes from the extra memory provided in its preconditions and the extension of the initial state , not from the memory transferred according to @xmath171 .
it can be shown that for the library @xmath345 in figure [ fig : impl](a ) , the library @xmath346 in figure [ fig : spec](a ) and method specification @xmath171 defined by ( [ container - val ] ) , we have @xmath610 .
it is also not difficult to prove ( e.g. , using separation logic ) that @xmath611 and @xmath612 are safe .
condition ( 4 ) in theorem [ thm : frame ] is satisfied , since the proof of safety of @xmath611 would use only the extra state provided in the preconditions of @xmath206 to provide the extra state required by its postconditions .
hence , by theorem [ thm : frame ] we have @xmath613 .
however , theorem [ thm : frame ] is not applicable to the memory allocators in figures [ fig : impl](b ) and [ fig : spec](b ) : since the allocator implementation in figure [ fig : impl](b ) stores free - list pointers inside the memory blocks , it is unsafe with respect to the variant of the method specification ( [ alloc ] ) that does not transfer their ownership .
the proof of theorem [ thm : frame ] relies on the following two lemmas , proved in appendices [ proof : frame - out ] and [ proof : frame - in ] , that convert between library traces corresponding to extended and original method specifications .
the first lemma shows that for a trace @xmath318 produced by @xmath614 , the trace @xmath615 can be produced by @xmath203 .
the safety of the library with respect to @xmath171 and condition ( 4 ) from theorem [ thm : frame ] guarantee that the smaller preconditions specified by @xmath171 are enough for the library to execute safely and that the extra pieces of state in @xmath206 do not influence its execution .
[ frame - out ] if @xmath616 , @xmath617 is safe at @xmath446 , and @xmath618 , then @xmath619 .
the other lemma gives conditions under which we can conclude that a trace @xmath318 is produced by @xmath281 given that @xmath620 is produced by @xmath312 .
[ frame - in ] assume @xmath621 , @xmath622 , @xmath298 is balanced from @xmath623 for @xmath624 , and @xmath625 . then @xmath626 .
[ [ proof - of - theoremthmframe . ] ] _ proof of theorem [ thm : frame ] . _ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + consider a trace @xmath627 , where @xmath628 and @xmath629 . then by ( 4 ) we have @xmath630 , and hence by lemma [ frame - out ] we have @xmath631 . since @xmath602 , for some @xmath632 and @xmath633 we have @xmath634 by lemma [ cor - rearr ] , there exists @xmath635 such that @xmath636 .
let @xmath441 be the trace @xmath378 with its interface actions replaced so that they form the history @xmath637 . then @xmath638 and @xmath639 .
since @xmath640 , we have @xmath641 . hence , by lemma [ frame - in ] , @xmath642 , from which the required follows .
the original definition of linearizability @xcite was proposed in an abstract setting that did not consider a particular programming language and implicitly assumed a complete isolation between the states of the client and the library .
furthermore , at the time it was not clear what the linearizability of a library entails for its clients .
filipovi et al .
@xcite were the first to observe that linearizability implies a form of contextual refinement ; technically , their result is similar to our lemma [ thm ] , but formulated over a highly idealistic semantics . in a previous work @xcite , we generalised their result to a compositional proof method , formalised by an abstraction theorem , that allows one to replace a concrete library by an abstract one in reasoning about a complete program .
this paper is part of our recent push to propose notions of concurrent library correctness for realistic programming languages .
so far we have developed such notions together with the corresponding abstraction theorems for supporting reasoning about liveness properties @xcite and weak memory models @xcite .
all these results assumed that the library and its client operate in disjoint address spaces and , hence , are guaranteed not to interfere with each other and can not communicate via the heap .
lifting this restriction is the goal of the present paper .
although the basic proof structure of theorems [ thm2 ] and [ thm - spec ] is the same as in @xcite , the formulations and proofs of the abstraction theorem and the required lemmas here have to deal with technical challenges posed by ownership transfer that did not arise in previous work .
first , their formulations rely on the novel forms of client - local and library - local semantics ( section [ sec : semantics ] ) that allow a component to communicate with its environment via ownership transfers . proving lemma [ prop : sem : local - global ]
then involves a delicate tracking of a splitting between the parts of the state owned by the library and the client , and how ownership transfers affect it .
second , the key result needed to establish the abstraction theorem is the rearrangement lemma ( lemmas [ cor - rearr ] and [ thm ] ) .
what makes the proof of this lemma difficult in our case is the need to deal with subtle interactions between concurrency and ownership transfer that have not been considered in previous work .
namely , changing the history in the lemma requires commuting ownership transfer actions ; justifying the correctness of these transformations is non - trivial and relies on the notion of history balancedness that we propose .
these differences notwithstanding , we hope that techniques for handling ownership transfer proposed in this paper can be combined with the ones for handling other types of client - library interactions considered so far @xcite .
recently , there has been a lot of work on verifying linearizability of common algorithms ; representative papers include @xcite . all of them proved classical linearizability , where libraries and their clients exchange values of a given data type and do not perform ownership transfers .
this includes even libraries , such as concurrent containers , that are actually used by client threads to transfer the ownership of data structures .
the frame rule for linearizability we propose ( theorem [ thm : frame ] ) justifies that classical linearizability established for concurrent containers entails linearizability with ownership transfer .
this makes our abstraction theorem applicable , enabling compositional reasoning about their clients .
turon and wand @xcite have proposed a logic for establishing refinements between concurrent modules , likely equivalent to linearizability @xcite .
their logic considers libraries and clients residing in a shared address space , but not ownership transfer . as a result
, they do not support separate reasoning about a library and its client in realistic situations of the kind we consider .
elmas et al .
@xcite have developed a system for verifying concurrent programs based on repeated applications of atomicity abstraction .
they do not use linearizability to perform the abstraction .
instead , they check the commutativity of an action to be incorporated into an atomic block with _ all _ actions of other threads . in particular , to abstract a library implementation in a program by its atomic specification , their method would have to check the commutativity of every internal action of the library with all actions executed by the client code of other threads .
thus , the method of elmas et al .
does not allow decomposing the verification of a program into verifying libraries and their clients separately .
in contrast , our abstraction theorem ensures the atomicity of a library under _ any _ safe client .
the most common approach of decomposing the verification of concurrent programs is using _ thread - modular _ reasoning methods , which consider every thread in the program in isolation under some assumptions on its environment @xcite .
however , a single thread would usually make use of multiple program components .
this work goes further by allowing a finer - grain _ intrathread - modular _ reasoning : separating the verification of a library and its client , the code from both of which may be executed by a single thread .
note that this approach is complementary to thread - modular reasoning , which can still be used to carry out the verification subtasks , such as establishing the linearizability of libraries and proving the safety of clients .
thread - modular techniques do enable a restricted form of intrathread - modular reasoning , since they allow reasoning about the control of a thread in a program while ignoring the possibility of its interruption by the other threads .
hence , they allow considering a library method called by the thread in isolation , e.g. , by using the standard proof rules for procedures .
however , such a decomposition is done under fixed assumptions on the environment of the thread and thus does not allow , e.g. , increasing the atomicity of the environment s actions . as the example of mcas shows ( section [ sec : example ] )
, this is necessary to deal with complex algorithms .
ways of establishing relationships between different sequential implementations of the same library have been studied in _ data refinement _
@xcite , including cases of interactions via ownership transfer @xcite .
our results can be viewed as generalising data refinement to the concurrent setting . moreover ,
when specialised to the sequential case , they provide a more flexible method of performing it in the presence of the heap and ownership transfer than previously proposed ones . in more detail , the way we define client safety ( section [ sec : semantics ] ) is more general some of the ways used in data refinement @xcite .
there , it is typical to fix a ( precise ) invariant of a library and check that the client does not access the area of memory fenced off by the invariant .
here we do not require an explicit library invariant , using the client - local semantics instead : since primitive commands fault when accessing non - existent memory cells , the safety of the client in this semantics ensures that it does not access the internals of the library .
we note that the approach requiring an invariant for library - local data structures does not generalise to the concurrent setting : while a precise invariant for the data structures _ shared _ among threads executing library code is not usually difficult to find , the state of data structures _
local _ to the threads depends on their program counters .
thus , an invariant insensitive to program positions inside the library code often does not exist .
such difficulties are one of reasons for using client- and library - local semantics in this paper .
finally , we note that the applicability of our results is not limited to proving existing programs correct : they can also be used in the context of formal program development . in this case , instead of _ abstracting _ an existing library to an atomic specification while proving a complete program , the abstraction theorem allows _ refining _ an atomic library specification to a concrete concurrent implementation while developing a program top - down @xcite .
our work thus advances the method of atomicity refinement to a setting with concurrent components sharing an address space and communicating via ownership transfers .
we would like to thank anindya banerjee , josh berdine , xinyu feng , hongjin liang , victor luchangco , david naumann , peter ohearn , matthew parkinson , noam rinetzky and julles villard for helpful comments .
gotsman was supported by the eu fet project advent .
yang was supported by epsrc . 10 d. amit , n. rinetzky , t. w. reps , m. sagiv , and e. yahav .
comparison under abstraction for verifying linearizability . in _ cav07 : conference on computer aided verification _ , volume 4590 of _ lncs _ , pages 477490 .
springer , 2007 .
r .- j . back . on correct refinement of programs .
, 23(1):4968 , 1981 . a. banerjee and d. a. naumann .
ownership confinement ensures representation independence in object - oriented programs .
, 52(6):894960 , 2005 .
mark batty , mike dodds , and alexey gotsman .
library abstraction for c / c++ concurrency . in _
popl13 : symposium on principles of programming languages _ , pages 235248 .
acm press , 2013 .
r. bornat , c. calcagno , p.
ohearn , and m. parkinson .
permission accounting in separation logic . in _
popl05 : symposium on principles of programming languages _ , pages 259270 .
acm press , 2005 .
s. burckhardt , a. gotsman , m. musuvathi , and h. yang .
concurrent library correctness on the tso memory model . in _
esop12 : european symposium on programming _ , volume 7211 of _ lncs _ , pages 87107 .
springer , 2012 . c. calcagno , p.
ohearn , and h. yang . local action and abstract separation logic . in _
lics07 : symposium on logic in computer science _ , pages 366378 .
ieee , 2007 .
d. g. clarke , j. noble , and j. m. potter .
simple ownership types for object containment . in _
ecoop01 : european conference on object - oriented programming _ , volume 2072 of _ lncs _ , pages 5376 .
springer , 2001 .
m. dodds , x. feng , m. parkinson , and v. vafeiadis .
deny - guarantee reasoning . in _
esop09 : european symposium on programming _ , volume 5502 of _ lncs _ , pages 363377 .
springer , 2009 .
t. elmas , s. qadeer , a. sezgin , o. subasi , and s. tasiran .
simplifying linearizability proofs with reduction and abstraction . in _
tacas10 : conference on tools and algorithms for the construction and analysis of systems _ , volume 6015 of _ lncs _ , pages 296311 .
springer , 2010 .
t. elmas , s. qadeer , and s. tasiran . a calculus of atomic actions . in _
popl09 : symposium on principles of programming languages _ , pages 215 .
acm press , 2009 .
i. filipovi , p. ohearn , n. rinetzky , and h. yang .
abstraction for concurrent objects .
, 411(51 - 52):43794398 , 2010 .
i. filipovi , p. ohearn , n. torp - smith , and h. yang .
blaiming the client : on data refinement in the presence of pointers .
, 22(5):547583 , 2010 .
a. gotsman and h. yang .
liveness - preserving atomicity abstraction . in _
icalp11 : international colloquium on automata , languages and programming _ , volume 6756 of _ lncs _ , pages 453465 .
springer , 2011 .
alexey gotsman , madanlal musuvathi , and hongseok yang .
show no weakness : sequentially consistent specifications of tso libraries . in _
disc12 : symposium on distributed computing _ ,
volume 7611 of _ lncs _ , pages 3145 .
springer , 2012 .
t. harris , k. fraser , and i. pratt . a practical multi - word compare - and - swap operation . in _
disc02 : symposium on distributed computing _ ,
volume 2508 of _ lncs _ , pages 265279 .
springer , 2002 .
m. herlihy and n. shavit . .
morgan kaufmann , 2008 .
m. herlihy and j. m. wing .
linearizability : a correctness condition for concurrent objects .
, 12(3):463492 , 1990 . c. a. r. hoare .
proof of correctness of data representations .
, 1:271281 , 1972 .
c. b. jones . specification and design of ( parallel )
programs . in _ ifip congress _ , pages 321332 , 1983 . c. b. jones . splitting atoms safely . , 375(1 - 3):109119 , 2007 .
i. mijajlovic and h. yang .
data refinement with low - level pointer operations . in _
aplas05 : asian symposium on programming languages and systems _ ,
volume 3780 of _ lncs _ , pages 1936 .
springer , 2005 .
p. ohearn .
resources , concurrency and local reasoning . , 375(1 - 3):271307 , 2007 .
a. pnueli . in transition from global to modular temporal reasoning about programs . in _ logics and models of concurrent systems _ ,
pages 123144 .
springer , 1985 .
j. c. reynolds .
types , abstraction and parametric polymorphism . in _ ifip congress _ , pages 513523 , 1983 . j. c. reynolds .
logic : a logic for shared mutable data structures . in _
lics02 : symposium on logic in computer science _ , pages 5574 .
ieee , 2002 .
a. turon and m. wand . a separation logic for refining concurrent objects . in _
popl11 : symposium on principles of programming languages _
, pages 247258 .
acm press , 2011 .
v. vafeiadis .
modular fine - grained concurrency verification .
phd thesis .
university of cambridge , 2008 .
v. vafeiadis .
automatically proving linearizability . in _
cav10 : conference on computer aided verification _
, volume 6174 of _ lncs _ , pages 450464 .
springer , 2010 .
we prove the required by induction on the length of @xmath448 . if @xmath448 is empty , then so is @xmath547 and @xmath643 . assume the statement of the proposition is valid for all histories @xmath448 of length less than @xmath644 .
consider a history @xmath645 of length @xmath646 and a corresponding history @xmath547 satisfying the conditions of the proposition .
we now make a case split on the type of the action @xmath104 .
@xmath337 @xmath104 is a call transferring @xmath446 that is not in @xmath547 .
then @xmath539 and @xmath547 are identical except @xmath539 may have extra calls .
hence , by the induction hypothesis for @xmath539 and @xmath547 , @xmath547 is balanced from @xmath95 , @xmath647 and @xmath648 @xmath104 is a call transferring @xmath446 also present in @xmath547 .
then @xmath649 , where @xmath540 and @xmath539 are identical except @xmath539 may have extra calls .
hence , by the induction hypothesis for @xmath539 and @xmath540 , we have @xmath650 in particular , @xmath547 is balanced from @xmath95 . by the induction hypothesis for @xmath539 and @xmath540
, we also have @xmath651 . from this
we get @xmath652 @xmath104 is a return transferring @xmath446 .
then it is also present in @xmath547 , so that @xmath653 , where @xmath539 and @xmath540 are identical except @xmath539 may have extra calls .
then by the induction hypothesis for @xmath539 and @xmath540 , we have : @xmath654 since @xmath655 is balanced from @xmath96 , @xmath656 is defined .
furthermore , by the induction hypothesis for @xmath539 and @xmath540 , we also have @xmath651 .
hence , @xmath657 is defined as well . by proposition
[ prop - delta ] , we then have : @xmath658 in particular , @xmath547 is balanced from @xmath95 . from @xmath651 , it also follows that @xmath659 before delving into the proof of lemma [ prop : sem : local - global ] , we prove three important lemmas about our semantics that justify its key steps .
the first concerns the evaluation of a call or a return action : intuitively , it says that the evaluation of such an action by the client matches that by the library .
[ lem : preservation ] let @xmath171 be a method specification , @xmath660 states , and @xmath293 an action describing a call to or return from a method specified in @xmath171 such that @xmath661 then for all @xmath662 , @xmath663 if furthermore @xmath664 is defined , then we have @xmath665 consider @xmath666 satisfying the conditions in the lemma .
we show the lemma only for the case when @xmath293 is a call action : for some @xmath667 , we have @xmath668 and @xmath669 .
the proof for the other case is symmetric . to show the first claim of the lemma , consider @xmath662 such that @xmath670 by the definition of the action evaluation
, there exist @xmath671 such that @xmath672 hence , @xmath673 let us move on to the second claim of the lemma . since @xmath674 , @xmath170 is precise and the @xmath2 operator is cancellative , there exists a unique splitting @xmath675 of @xmath446 such that @xmath676
. let @xmath677 .
then @xmath678 hence , @xmath679 the second lemma describes the decomposition and composition properties of trace evaluation .
[ lem : comp : eval ] consider traces @xmath680 without interface actions such that @xmath681 .
for all states @xmath660 , if @xmath682 then @xmath683 and @xmath684 consider @xmath685 satisfying the assumptions .
we prove the lemma by induction on the length of @xmath125 .
the base case of @xmath125 being the empty sequence is trivial .
now suppose that @xmath686 for some @xmath687 .
then there exist @xmath348 and @xmath395 such that @xmath688 by the assumption of the lemma , we have that @xmath689 hence , by the induction hypothesis , we have that @xmath690 and @xmath691 from it follows that : @xmath692 next , we prove that @xmath693 for the sake of contradiction , suppose this disequality does not hold . because of , there exists @xmath694 such that @xmath695 by , this implies the existence of @xmath696 such that @xmath697 we split cases based on the relationships among @xmath263 , @xmath348 , @xmath318 and @xmath395 . 1 . if @xmath698 and @xmath699 , then @xmath700 for some @xmath701
. by , @xmath702 ,
so that @xmath703 . hence , by the strong locality of @xmath704 , @xmath705 , so that @xmath706 . but this contradicts .
2 . if @xmath707 and @xmath708 , then @xmath700 for some @xmath701 .
this case is symmetric to the previous one .
3 . if @xmath698 and @xmath708 , then @xmath293 is a call or a return action . then , by the definition of evaluation , @xmath709 . this gives the desired contradiction .
the remainder of the proof is again done by a case analysis on the relationships among @xmath263 , @xmath348 , @xmath318 and @xmath395 .
we consider three cases .
\1 . @xmath698 and @xmath699 . in this case , @xmath710 for some @xmath701 . as shown in @xmath711 , @xmath712 .
pick @xmath694 such that [ alexey1 ] ( , ) (_0 * _ 1 ) . by the definition of the trace evaluation and the induction hypothesis in ,
there exist @xmath713 , @xmath714 , @xmath715 and @xmath716 such that [ alexey2 ] ( , ) ( _ 0 * _ 1 ) ( _ 0 , ) _ 0 ( _ 1 , ) : _ 1 cover( , , ) .
then @xmath717 we have @xmath718 , because @xmath719 and @xmath720 .
hence @xmath721 and , furthermore , @xmath722 .
hence , by the strong locality of @xmath704 , there exists @xmath723 such that @xmath724 this implies @xmath725 from what we have shown so far , it follows that @xmath726 thus , [ alexey3 ] ( _ 0 * _ 1 ) \ { ( , ) _ 0 : _
1}. to show the other inclusion and the part of the lemma , consider @xmath727 such that @xmath728 then for some @xmath729 we have @xmath730 by the definition of the evaluation function , there exists @xmath723 such that @xmath731 by the induction hypothesis in and , @xmath732 now by the footprint preservation property of @xmath704 , the first conjunct above implies that @xmath733 , which proves @xmath734 . by the strong locality of @xmath704 , @xmath735 from what we have shown above it follows that @xmath736 hence , [ alexey4 ] ( _ 0 * _ 1 ) \ { ( , ) _ 0 : _ 1}. \2 . @xmath707 and @xmath708 .
this case is symmetric to the previous one .
\3 . @xmath737 and @xmath708 . in this case
@xmath293 is a call to or a return from a method in @xmath171 . as shown in ,
pick @xmath694 such that ( [ alexey1 ] ) holds . by the definition of evaluation and the induction hypothesis in ,
there exist @xmath713 , @xmath714 , @xmath715 and @xmath716 such that ( [ alexey2 ] ) holds .
but @xmath738 furthermore , @xmath739 .
hence , by lemma [ lem : preservation ] and the definition of the evaluation , there exist @xmath740 such that @xmath741 this in turn implies that @xmath742 from what we have shown so far , it follows that @xmath743 thus , ( [ alexey3 ] ) holds . to show the other inclusion and the part of the lemma , consider @xmath744 such that @xmath745 we need to show that @xmath746 in particular , this establishes ( [ alexey4 ] ) .
since @xmath747 and @xmath748 is a call to or a return from a method in @xmath171 , @xmath749 we now use the induction hypothesis in and and derive that @xmath750 but @xmath751 and @xmath752 .
hence , by lemma [ lem : preservation ] , @xmath753 by the definition of evaluation , @xmath754 from what we have shown , it follows that @xmath755 as required .
the following lemma shows that the trace - set generation of our semantics also satisfies the decomposition and composition properties .
[ lem : comp : tracesemantics ] @xmath756 let @xmath757\ { \sf in}\ c_1\parallel{\ldots}\parallel c_n ; \ \ , { \mathcal{l}}= \{m = c_m \mid m \in \{m_1,{\ldots},m_j\}\ } ; \ \ , c_{{\sf mgc}}= ( m_1 + { \ldots } + m_j)^*.\ ] ] first , consider @xmath758 . by the definition of the semantics , for some trace @xmath759 , @xmath125 is a prefix of @xmath759 , @xmath760 and all actions in @xmath759 are done by some thread @xmath761 .
then @xmath762 since all actions in @xmath759 are done by some thread @xmath761 , we have @xmath763 hence , @xmath764 since @xmath265 is a prefix of @xmath765 and @xmath266 is a prefix of @xmath766 , this implies @xmath767 furthermore , @xmath768 , as desired .
assume now that @xmath769 then for some traces @xmath348 and @xmath395 , @xmath263 is a prefix of @xmath348 , @xmath318 is a prefix of @xmath395 , and @xmath770 the definition of our semantics in figure [ fig : trace ] allows us to choose @xmath395 in such a way that for some trace @xmath759 , @xmath125 is a prefix of @xmath759 and @xmath771
. then @xmath772 and @xmath773 .
thus , @xmath774 , which implies @xmath758 , as desired .
[ [ proof - of - lemmapropsemlocal - global . ] ] _ proof of lemma [ prop : sem : local - global ] . _ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + we first show that @xmath209 is safe for @xmath775 .
pick states @xmath776 such that @xmath777 by lemma [ lem : comp : tracesemantics ] , there exist traces @xmath778 such that [ alexey10 ] : cover ( , , ) . by our assumptions ,
@xmath202 and @xmath312 are safe for @xmath779 and @xmath362 , respectively .
hence , @xmath720 and @xmath780 . by lemma [ lem :
comp : eval ] , these disequalities imply that @xmath781 .
we have just shown the safety of @xmath209 for @xmath782 .
next , we show that @xmath783 pick @xmath784
. then for some @xmath660 we have @xmath785 by lemma [ lem : comp : tracesemantics ] , there are @xmath778 such that ( [ alexey10 ] ) holds .
we use lemma [ lem : comp : eval ] and deduce that for some @xmath786 we have @xmath787 furthermore , @xmath788 , @xmath789 and @xmath790 .
hence , @xmath791 as desired . finally , we prove that @xmath792 pick @xmath793 . by the definition of the @xmath319 operator and our semantics , there exist @xmath794 such that @xmath795 by the definition of our semantics , @xmath796 and @xmath797 . because of this and @xmath798 we have @xmath681 . by lemma [ lem :
comp : tracesemantics ] , this implies @xmath799 .
also , by lemma [ lem : comp : eval ] , we have that @xmath800 hence , @xmath801 as desired .
consider @xmath616 .
then there exist @xmath189 and @xmath802 such that @xmath803 .
we show that for some @xmath193 we have @xmath804 we proceed by induction on the length of @xmath408 .
the base case of @xmath805 is trivial .
assume that the above holds for some @xmath806 and consider @xmath807 such that @xmath808 we show that for some @xmath809 we have @xmath810 we consider three cases , depending on the type of the actions @xmath293 and @xmath748 .
@xmath337 @xmath811 .
then @xmath812 . since @xmath617 is safe at @xmath446 , @xmath813
hence , by the strong locality property , we have @xmath814 then @xmath815 for some @xmath816 .
@xmath817 and @xmath818 , where @xmath819 and @xmath820 .
then @xmath821 hence , the required holds for @xmath822 .
@xmath823 and @xmath824 , where @xmath825 and @xmath826 .
then @xmath827 since @xmath617 is safe at @xmath446 , @xmath828 for some @xmath809 , so that @xmath829 by the cancellativity of @xmath2 , this entails @xmath830 we also know that @xmath831 , so that @xmath832 is defined .
hence , @xmath833 . in the following , we extend the @xmath587 operation to non - interface actions by assuming that it does not change them .
consider @xmath834 satisfying the conditions of the lemma . then @xmath298 is balanced from @xmath623 for @xmath835 , and there exist @xmath38 and @xmath802 such that @xmath836 .
we show that @xmath837 by induction on the length of @xmath408 .
the base case of @xmath838 is trivial .
assume that the above holds for some @xmath839 and consider @xmath840 such that @xmath841 we show that @xmath842 we consider three cases , depending on the type of the actions @xmath293 and @xmath748 .
@xmath337 @xmath811 .
then @xmath843 and @xmath844 .
since @xmath845 by the footprint preservation property , @xmath846 . then by the strong locality property , @xmath847 @xmath817 and @xmath848 , where @xmath819 and @xmath849 .
in this case we have @xmath850 .
since @xmath851 is balanced from @xmath623 , we have @xmath852
. then @xmath853 @xmath823 and @xmath854 , where @xmath825 and @xmath855 . in this case
we have @xmath856 .
we know @xmath857 .
thus , @xmath858 .
since @xmath859 is defined , so is @xmath860
[ cols= " < , < " , ] | linearizability is a commonly accepted notion of correctness for libraries of concurrent algorithms .
unfortunately , it assumes a complete isolation between a library and its client , with interactions limited to passing values of a given data type .
this is inappropriate for common programming languages , where libraries and their clients can communicate via the heap , transferring the ownership of data structures , and can even run in a shared address space without any memory protection . in this paper
, we present the first definition of linearizability that lifts this limitation and establish an abstraction theorem : while proving a property of a client of a concurrent library , we can soundly replace the library by its abstract implementation related to the original one by our generalisation of linearizability .
this allows abstracting from the details of the library implementation while reasoning about the client .
we also prove that linearizability with ownership transfer can be derived from the classical one if the library does not access some of data structures transferred to it by the client . |
a _ maximal _ surface in a @xmath4dimensional lorentzian manifold is a spacelike surface with zero mean curvature . here by _
spacelike _ we mean that the induced metric from the ambient lorentzian metric is a riemannian metric on the surface .
the mathematical interest of maximal surfaces resides in the fact that they locally maximize area among all nearby surfaces having the same boundary @xcite .
furthermore , maximal surfaces and , more generally spacelike surfaces with constant mean curvature , have also a great importance in general relativity @xcite .
parabolicity is a concept that lies in the borderline between several branches of mathematics , such as riemannian geometry , stochastic analysis , partial differential equations and potential theory .
let us recall that a riemannian surface @xmath5 with non - empty boundary , @xmath6 , is said to be _ parabolic _ if every bounded harmonic function on @xmath7 is determined by its boundary values .
it is interesting to observe that the parabolicity of @xmath7 is equivalent to the existence of a proper nonnegative superharmonic function on @xmath7 ( see beginning of section [ s3 ] ; for details see @xcite and @xcite ) .
classically , a riemannian surface without boundary is called parabolic if it does not admit a nonconstant negative subharmonic function .
in fact , in @xcite we considered that definition of parabolicity .
however , along this work we will reserve the term _ parabolicity _ for riemannian surfaces with non - empty boundary , and we will use the term _ recurrence _ for riemannian surfaces without boundary . a riemannian surface @xmath5 without boundary is called _ recurrent _ if for every nonempty open set @xmath8 with smooth boundary , @xmath9 is parabolic ( as a surface with boundary ) .
it is worth pointing out that the concept of recurrence is equivalent to the classical concept of parabolicity ( for surfaces without boundary ) .
this follows from the observation that a riemannian manifold without boundary is recurrent precisely when almost all brownian paths are dense in the manifold ( for the details , see @xcite and ( * ? ? ?
* theorem 5.1 ) ) . in @xcite fernndez and lpez
have recently proved that properly immersed maximal surfaces with non - empty boundary in the lorentz - minkowski spacetime @xmath10 are parabolic if the lorentzian norm on the maximal surface in @xmath10 is eventually positive and proper ( see also section 4 in @xcite ) .
motivated by that work , in this paper we study some parabolicity criteria for maximal surfaces immersed into a lorentzian product space of the form @xmath11 , where @xmath1 is a connected riemannian surface and @xmath11 is endowed with the lorentzian metric @xmath12 here @xmath13 and @xmath14 denote the projections from @xmath15 onto each factor , and @xmath16 is the riemannian metric on . for simplicity ,
we will simply write @xmath17 and we will denote by @xmath18 the 3-dimensional product manifold @xmath11 endowed with that lorentzian metric . observe that in particular , when @xmath19 is the flat euclidean plane what we obtain is the well known @xmath20 lorentz - minkowski 3-dimensional spacetime . a natural generalization of the lorentzian norm on a surface in @xmath10 to the lorentzian product consists in considering the function @xmath21 where the function @xmath22 measures the distance on the factor @xmath23 to a fixed point @xmath24 and @xmath25 is the height function of the surface @xmath26 ( for the details , see section [ s3 ] ) . in this context ,
our main result is theorem [ teorlap ] which asserts that given a complete riemannian surface @xmath27 with non - negative gaussian curvature , then every maximal surface in with non - empty boundary and such that the function @xmath28 is eventually positive and proper is parabolic . in particular , and as an application of this result , we deduce that every maximal graph over a starlike domain @xmath3 is parabolic .
this allows us to give an alternative proof of the non - parametric version of our calabi - bernstein result ( * ? ? ?
* theorem 4.3 and corollary 4.4 ) for maximal entire graphs in ( see corollary [ cb ] ) .
the authors would like to heartily thank v. palmer , j. prez and s. verpoort for several explanations and useful comments during the preparation of this paper .
a smooth immersion @xmath29 of a connected surface @xmath30 is said to be a spacelike surface if @xmath31 induces a riemannian metric on , which as usual is also denoted by @xmath32 . in that case , since @xmath33 is a unitary timelike vector field globally defined on the ambient spacetime @xmath34 , then there exists a unique unitary timelike normal field @xmath35 globally defined on which is in the same time - orientation as @xmath36 , so that @xmath37 we will refer to @xmath35 as the future - pointing gauss map of , and we will denote by @xmath38 $ ] the smooth function on given by @xmath39 . observe that the function @xmath40 measures the hyperbolic angle @xmath41 between the future - pointing vector fields @xmath35 and @xmath36 along . indeed , they are related by @xmath42 .
let @xmath43 and @xmath44 denote the levi - civita connections in and , respectively .
then the gauss and weingarten formulae for the spacelike surface are given by [ eqgauss ] _
xy=_xy - n and [ eqwein ] ax=-_xn , for every tangent vector fields @xmath45 .
here @xmath46 stands for the shape operator ( or second fundamental form ) of with respect to its future - pointing gauss map @xmath35 . the height function @xmath25 of a spacelike surface
@xmath47 is the projection of onto @xmath48 , that is , @xmath49 .
observe that the gradient of @xmath14 on is @xmath50 so that the gradient of @xmath51 on @xmath7 is @xmath52 throughout this paper , for a given vector field @xmath53 along the immersion , we will denote by @xmath54 its tangential component ; that is , @xmath55 in particular , @xmath56 and we easily get [ normgradh ] h^2=^2 - 1 , where @xmath57 denotes the norm of a vector field on . since @xmath36 is parallel on we have that [ eqaux ] _ x _
t=0 for every tangent vector field @xmath58 .
writing @xmath59 along the surface and using gauss and weingarten formulae , we easily get from that [ eqaux2 ] _ x h = ax for every @xmath58 .
therefore the laplacian on of the height function is given by @xmath60 where @xmath61 is the mean curvature of relative to @xmath35 .
in particular , [ laph ] h=0 for every maximal surface in . given any function @xmath62 , we can consider its lifting @xmath63 defined by @xmath64 in turn , we associate to @xmath65 the function @xmath66 given by @xmath67 . in this context , the laplacian on of @xmath68 can be expressed in terms of the laplacian @xmath69 of @xmath70 and the differential operators of @xmath71 as follows .
[ lemmaaux ] along a spacelike surface we have that [ lemma ] =| |+2h_m+^2 ( n^,n^ ) where @xmath72 , and @xmath73 and @xmath74 denote the gradient and the hessian operators on @xmath75 , respectively .
since @xmath76 , we get from and that the hessian operators of @xmath70 and @xmath68 satisfy @xmath77 for every @xmath58 .
therefore , it can be easily seen that [ rellap ] ||=-2h-|^2|(n , n ) .
observe now that , as the function @xmath70 does not depend on @xmath78 , then @xmath79 .
thus , @xmath80 and @xmath81 so that the lemma follows directly from .
our main result in this section generalizes ( * ? ? ? * theorem 3.1 ) ( see also ( * ? ? ? * theorem 4.2 ) ) to the case of maximal spacelike surfaces in @xmath82 , when is a complete riemannian surface with non - negative gaussian curvature . in that case , consider the function @xmath83 defined by @xmath84 where @xmath85 is a fixed point .
observe that @xmath86 almost everywhere .
actually , @xmath87 is smooth on @xmath88 , where @xmath89 stands for the cut locus of @xmath90 .
as is well - known , @xmath91 and @xmath89 is a null set .
following our notation above , let @xmath92 denote the lifting of @xmath87 to , and for a given spacelike surface , let @xmath22 stands for the restriction of @xmath93 to @xmath7 , @xmath94 .
consider @xmath95 the projection of @xmath7 on @xmath75 .
it is not difficult to see that @xmath96 , where @xmath32 stands for the riemannian metric on induced from the lorentzian ambient space .
that means that @xmath97 is a local diffeomorphism which increases the distance between the riemannian surfaces and , and by ( * ? ? ?
* chapter viii , lemma 8.1 ) @xmath97 is a covering map(for the details , see lemma 3.1 in @xcite ) .
therefore , @xmath98 and the function @xmath22 is smooth almost everywhere in .
the proof of our main result is based on the following technical lemma .
[ lemadebil ] let @xmath99 be a riemannian surface with non - empty boundary , @xmath100 .
if there exists a proper continuous function @xmath101 which is eventually positive and superharmonic , then @xmath7 is parabolic . as is usual , by eventually we mean here a property that is satisfied outside a compact set .
the proof follows the ideas of the proof of an analogous criterium for proper smooth functions given by meeks and prez @xcite . for the sake of completeness
, we sketch it here . it suffices to see that if @xmath102 is a bounded harmonic function on which vanishes on the boundary , @xmath103 , then @xmath102 is constant zero .
let @xmath104 be a compact subset such that @xmath68 is positive and superharmonic on @xmath105 .
it suffices to show that @xmath106 .
let us assume that there exists a point @xmath107 such that @xmath108 .
since @xmath109 , there is a constant @xmath110 with @xmath111 .
since @xmath102 is harmonic , the function @xmath112 is superharmonic on @xmath105 . besides , since @xmath68 is proper and positive on @xmath105 , then @xmath113 is positive outside the compact subset @xmath114 where @xmath115 . finally , since @xmath116
, @xmath113 must reach its minimum on @xmath105 at an interior point , which is a contradiction by the minimum principle for superharmonic functions .
now we are ready to prove our main result .
[ teorlap ] let @xmath27 be a complete riemannian surface with non - negative gaussian curvature .
consider @xmath7 a maximal surface in with non - empty boundary , @xmath100 , and assume that the function @xmath117 defined by @xmath118 is eventually positive and proper .
then @xmath7 is parabolic .
it is worth pointing out that the assumption on the non negativity of the gaussian curvature of @xmath75 is necessary .
actually , let @xmath119 and consider @xmath120 a connected domain with smooth boundary .
then , for a fixed @xmath121 , @xmath122 is trivially a non - parabolic maximal surface in @xmath123 on which @xmath124 is eventually positive and proper .
let @xmath125 , and consider @xmath126 .
@xmath127 is compact because @xmath124 is eventually positive and proper .
as is well known , parabolicity is not affected by adding or removing compact subsets , so that is parabolic if and only if @xmath128 is parabolic .
the function @xmath129 is a proper positive function on @xmath128 .
therefore , in order to prove that @xmath128 is parabolic it suffices to see that @xmath130 is superharmonic on @xmath128 .
equivalently , it suffices to see that @xmath130 is superharmonic on the dense subset @xmath131 where it is smooth . in what follows
, we will work on that subset @xmath132 . from
and we get [ laph2 ] h^2=2hh+2h^2=2(^2 - 1 ) . on the other hand , as the function @xmath93 does not depend on @xmath78 then @xmath133 and @xmath134 . therefore ,
[ lapr21 ] ||r^2(x , t)=2 |r(x , t)||r(x , t)+2||r(x , t)^2= 2((x)(x)+1 ) , since , as is well - known , @xmath135 .
applying now lemma [ lemmaaux ] to @xmath136 we get [ lapr22 ] r^2=||r^2+^2 ^2(n^,n^)= 2(r+1)+^2 ^2(n^,n^ ) . to study the last term of we
will compute first the hessian of @xmath137 on @xmath75 at a point @xmath138 .
for @xmath139 we have @xmath140 in particular , for @xmath141 of unit length @xmath142 we get @xmath143 as any @xmath144 can be decomposed as @xmath145 we finally obtain @xmath146 therefore , along the surface @xmath132 we have that @xmath147 and becomes [ lapr2 ] r^2=r(1+_m^2)+1+_m^2 .
now , from and we get that @xmath148 as @xmath27 is complete and has non - negative gaussian curvature , by the laplacian comparison theorem we have that @xmath149 , so that @xmath150 on @xmath132 . using this in , we obtain that @xmath151 since @xmath152 .
on the other hand , @xmath153 , and so @xmath154 observe that @xmath155 and @xmath59 .
taking into account that @xmath156 and @xmath157 , it follows from here that @xmath158 which jointly with implies that @xmath159 therefore , from and we finally get [ lap ] = ( -^2)0 , which means that @xmath160 is a proper positive superharmonic function on @xmath132 . then , @xmath132 is parabolic , and @xmath26 is also parabolic .
it is interesting to look for some natural conditions under which the assumptions of theorem [ teorlap ] are satisfied . in this context
, we have the following result . [ propwa ]
let @xmath27 be a complete riemannian surface and let be a proper spacelike immersion which eventually lies in @xmath161 for some @xmath162 .
then the function @xmath21 is eventually positive and proper on .
let @xmath27 be a complete riemannian surface with non - negative gaussian curvature .
then every proper maximal immersion with non - empty boundary which eventually lies in @xmath163 for some @xmath162 , is parabolic . since @xmath164 eventually lies in @xmath163
, then there exists a compact set @xmath165 such that @xmath166 and @xmath167 on @xmath168 . in order to see that @xmath124 is proper
, it suffices to prove that @xmath169)$ ] is compact for every @xmath170 .
let @xmath171 defined by @xmath172 , so that @xmath173
. then @xmath169)=f^{-1}(\bar{\phi}^{-1}([0,b]))$ ] .
since @xmath31 is proper , then it suffices to prove that @xmath174)\subset\mathcal{w}_a$ ] is compact .
observe that for every @xmath175 one has @xmath176 therefore , for every @xmath177)\subset\mathcal{w}_a$ ] we have that @xmath178 this implies that @xmath174)\subset \bar{b}(x_0,c)\times [ -ac , ac]$ ] , where @xmath179 denotes the geodesic disc on @xmath75 of radius @xmath180 centered at @xmath90 ( see figure 1 ) . since @xmath181 $ ] is compact , the result follows . )
$ ] is compact under the assumptions of proposition [ propwa].,width=264 ] on the other hand , recall that a riemannian surface @xmath5 _ without _ boundary is called recurrent if for every nonempty open set @xmath8 with smooth boundary , @xmath9 is parabolic ( as a surface with boundary ) . therefore , as another consequence of our theorem [ teorlap ] we can state the following . [ coro5 ]
let @xmath27 be a complete riemannian surface with non - negative gaussian curvature , and let @xmath7 be a maximal surface in without boundary , @xmath182 .
if the function @xmath21 is eventually positive and proper on @xmath7 , then @xmath7 is recurrent . for a proof
, simply observe that if @xmath8 is a nonempty open set with smooth boundary , then the function @xmath124 restricted to @xmath9 is also eventually positive and proper on @xmath9 , and therefore the maximal surface with boundary @xmath9 is parabolic by our main result .
let @xmath183 be a connected domain . every smooth function @xmath184 determines a graph over @xmath185 given by @xmath186 .
the metric induced on @xmath185 from the lorentzian metric on the ambient space via @xmath187 is given by [ gu ] = _ m - du^2 .
therefore , @xmath187 is a spacelike surface in if and only if @xmath188 everywhere on @xmath185 .
a graph is said to be entire if @xmath189 .
consider @xmath190 a connected domain and let @xmath191 .
we will say that @xmath185 is starlike with respect to @xmath90 if for every @xmath192 there exists a ( non - necessarily unique ) minimizing geodesic segment from @xmath90 to @xmath138 which is contained in @xmath185 . obviously , if is a complete riemannian surface , then itself is starlike with respect to any of its points .
[ lemageom ] let @xmath27 be a complete riemannian surface and let @xmath193 be a spacelike graph over a domain @xmath185 which is starlike with respect to some point @xmath191 .
then the function @xmath21 is eventually positive and proper on @xmath193 .
we may assume without loss of generality that @xmath194 . since @xmath193 is homeomorphic to @xmath185 ( via the standard embedding @xmath195 ) , and the thesis of our result is topological , it is equivalent to prove that the function @xmath196 is eventually positive and proper on @xmath185 .
consider @xmath197 firstly , we will prove that @xmath102 is positive for every @xmath198 .
that is , we are going to see that @xmath199 for a given @xmath200 , consider @xmath201\rightarrow\omega$ ] a minimizing geodesic segment such that @xmath202 , @xmath203 and @xmath204 .
let @xmath205 , where @xmath206 .
@xmath193 is a spacelike surface , so @xmath207 is a non - vanishing spacelike vector , that is , @xmath208 therefore , @xmath209 for every @xmath210 , and integrating we obtain [ u ] -(x)<u(x)<(x ) .
consequently , @xmath211 and @xmath212 for every @xmath192 , @xmath200 .
it remains to prove that @xmath102 is proper .
let us consider on @xmath213 the standard riemannian metric , @xmath214 , and let us denote by @xmath215 the distance related to this riemannian metric .
let us see now that [ dist1 ] _
+ ( ( x , t),)= \{(x)-t,(x)+t}=((x)-|t| ) for every @xmath216 .
observe that @xmath217 can be decomposed into @xmath218 where @xmath219 therefore , @xmath220 expression is clear for @xmath221 ( and necessarily @xmath222 ) . for a given @xmath200 , let @xmath223\rightarrow\omega$ ]
be a minimizing geodesic segment such that @xmath202 , @xmath224 .
we will compute first @xmath225 . since @xmath226 is minimizing , for every @xmath227
$ ] we have @xmath228 , so that @xmath229 and @xmath230 observe that this expression attains its minimum at @xmath231 , and [ dist3 ] _ + ( ( x , t),((s_0),s_0))=|(x)-t|= ( ( x)-t ) .
we claim that @xmath225 is given by .
in fact , for every @xmath232 we have that @xmath233 and so @xmath234 therefore , [ dist4 ] _ + ( ( x , t),)=((x)-t ) .
let @xmath192 , @xmath200 , and let @xmath223\rightarrow\omega$ ] be a minimizing geodesic segment such that @xmath202 , @xmath224 .
. then @xmath237 and by we have that [ dist2 ] _ + ( ( ( s),u(s)),)= ( s-|u(s)| ) .
this implies that @xmath238 is a positive increasing function for @xmath239 .
therefore , if we choose @xmath240 such that the geodesic disc @xmath241 , then it follows that [ dist6 ] _
+ ( ( x , u(x)),)>0 xb _ , where @xmath242 ( see figure 2 ) .
now we are ready to prove that @xmath102 is proper on @xmath185 .
since @xmath243 with @xmath244 compact , it suffices to prove that @xmath245 is proper on @xmath246 .
let @xmath247 be the standard embedding , @xmath248 , and let @xmath249 defined by @xmath172 .
observe that @xmath31 is trivially proper for if @xmath250 is compact , then @xmath251 is compact , where @xmath252 and @xmath253 are the projections .
by we have that @xmath254 therefore , @xmath255 .
note that the map @xmath256 is proper .
in fact , for every compact set @xmath257 we see that @xmath258 is compact .
therefore , it suffices to show that @xmath259 is proper , or , equivalently , that for every @xmath170 , @xmath260)=\mathcal{u}\cap\bar{\phi}^{-1}([0,b])$ ] is compact .
let @xmath261)$ ] .
since @xmath262 , by we obtain that @xmath263 therefore , since @xmath264 , this gives @xmath265 that is @xmath266 this implies that @xmath260)\subset\overline{b}(x_0,c)\times[\sqrt{2}\varepsilon - c , c-\sqrt{2}\varepsilon]$ ] is compact .
this finishes the proof of proposition [ lemageom ] . as another consequence of our main result ( theorem [ teorlap ] )
, we can give a new proof of the following calabi - bernstein theorem , first established in ( * ? ? ?
* theorem 4.3 ) ( see also @xcite for another approach to the parametric version of that result , first established in ( * ? ? ? * theorem 3.3 ) ) . [ cb ]
let @xmath1 be a complete riemannian surface with non - negative gaussian curvature .
then any entire maximal graph @xmath187 in is totally geodesic .
in addition , if @xmath267 at some point on , then @xmath268 is constant . by applying proposition [ lemageom ] with @xmath189
we know that the function @xmath21 is eventually positive and proper on @xmath187 .
therefore , by corollary [ coro5 ] we have that @xmath187 is recurrent .
equivalently , any negative subharmonic function on the surface @xmath187 must be constant .
the proof then follows as in the proof of ( * ? ? ?
* theorem 3.3 ) . for the sake of completeness ,
we briefly sketch it here .
recall that @xmath269 .
a careful computation using codazzi equation and the maximality of @xmath187 gives that @xmath270 and @xmath271 where @xmath272 and @xmath273 stands for the gaussian curvature of along @xmath187 .
this implies that @xmath274 that is , @xmath275 is a negative subharmonic function on the recurrent surface @xmath187 , and hence it must be constant .
therefore , @xmath276 and @xmath277 on @xmath187 , where @xmath278 .
thus , @xmath187 is totally geodesic and , if @xmath279 at some point on @xmath187 ( equivalently , @xmath267 at some point on ) , then it must be @xmath280 , which means that @xmath51 is constant and @xmath187 is a slice .
i. fernndez and f.j .
lpez , _ on the uniqueness of the helicoid and enneper s surface in the lorentz - minkowski space @xmath282 _ , to appear in trans .
available at http://arxiv.org/abs/0707.1946 | in this paper we establish some parabolicity criteria for maximal surfaces immersed into a lorentzian product space of the form @xmath0 , where @xmath1 is a connected riemannian surface with non - negative gaussian curvature and @xmath0 is endowed with the lorentzian product metric @xmath2 . in particular , and as an application of our main result , we deduce that every maximal graph over a starlike domain @xmath3 is parabolic .
this allows us to give an alternative proof of the non - parametric version of the calabi - bernstein result for entire maximal graphs in . |
numerically computed magnetohydrodynamic ( mhd ) equilibria are the starting point of a wide class of numerical solvers that are used to study mhd stability , transport , and heating and current drive in magnetic fusion devices @xcite .
static mhd equilibria of toroidally axisymmetric configurations are described by the grad - shafranov ( g - s ) equation @xcite , a nonlinear , second - order elliptic partial differential equation .
stationary equilibria with purely toroidal flows are determined by solving a close variant of the g - s equation @xcite , the only difference being that for the latter the pressure term does not only depend on the poloidal flux function @xmath0 , but also on the radial variable @xmath1 .
numerical codes to solve the g - s equation have been developed since the early days of the magnetic fusion program @xcite . nevertheless , the development of optimized g - s codes remains a topic of active research , for three main reasons .
first , g - s solvers must be able to properly resolve complex two - dimensional geometries @xcite , with boundaries that may have a corner , corresponding to a magnetic field x - point @xcite .
second , g - s solvers must be fast .
this criterion is particularly relevant in the context of multiphysics integrated simulations @xcite .
several of these multiscale , multiphysics solvers already include , or will eventually include , in their iterative procedure a step in which the equilibrium configuration is self - consistently updated .
a reasonable requirement is that the calculation of the updated equilibrium takes a negligible amount of time and computing resources as compared to the computationally intensive transport , mhd stability and plasma heating solvers .
third , g - s solvers must be accurate .
the solution of the g - s equation is the poloidal flux @xmath0 , but the physical quantities of interest , such as the magnetic field , the safety factor , the magnetic shear , the magnetic curvature , and the current density are all functions of partial derivatives of @xmath0 .
since there always is some loss of accuracy when computing derivatives , a high level of accuracy for @xmath0 is desired . in this article , we present the new grad - shafranov code ecom ( equilibrium solver via conformal mapping ) .
ecom is a fixed boundary , direct solver written in fortran 77/90 that is based on three key elements : 1 ) the formulation of the g - s equation as a nonlinear poisson problem ; 2 ) a spectrally accurate numerical method to compute the conformal map from the smooth plasma cross section of interest to the unit disk ; 3 ) a fast , high order poisson solver on the unit disk @xcite .
its main novelty lies in the last two aspects discussed in the paragraph above , namely accuracy and speed .
regarding the first point , we demonstrate in this article that ecom has better convergence properties than popular g - s solvers based on finite elements @xcite . in the finite element approach the numerical error of
the solution decays as a power of grid size , i.e. @xmath2 where @xmath3 is an integer and @xmath4 is the number of grid points in one direction . often , for magnetic fusion applications @xmath5 @xcite , although @xmath6 was recently demonstrated @xcite .
in contrast , convergence in ecom is exponential : the error decays as @xmath7 for some real number @xmath8 .
just as importantly , in ecom the rate of convergence for the derivatives of @xmath0 is the same as that of @xmath0 , whereas in the finite element approach the derivatives of @xmath0 converge more slowly than @xmath0 @xcite .
remarkably , in ecom numerical accuracy is not obtained at the expense of computational complexity and speed . for a given grid size ,
our solver is faster than finite element solvers and less demanding in terms of memory .
we will show that a drawback of relying on conformal mapping is that ecom often requires a denser grid than fem based solvers to achieve a desired accuracy . yet despite this , we find that for tokamak geometries and medium to high number of grid points , ecom is more accurate than fem based equilibrium codes at equal run time .
this article follows an earlier article @xcite , in which we gave a detailed description of our new numerical algorithm for solving the g - s equation .
the focus here is different .
one of the main motivations is to present extensions recently added to our g - s solver that make it a practical tool readily usable in fusion applications .
the new capabilities of our equilibrium solver include the possibility of computing equilibria with arbitrary toroidal flow profiles , the possibility of specifying current and pressure profiles in various ways , as well as the evaluation of the key physical quantities that are required as inputs in stability , transport and heating codes .
since ecom is a direct solver that calculates @xmath0 on a prescribed grid for the poloidal cross section @xcite , we put a particular emphasis on the accurate computation of the contours of constant flux , and of flux surface quantities such as the safety factor and the magnetic shear .
a second motivation for this article is to perform detailed comparisons between the popular g - s code chease @xcite and ecom , and to assess the merits of each solver .
the structure of the article is as follows . in section [ sec : gs ]
we briefly review the numerical algorithm we use to solve the g - s equation @xcite . in section [ sec : post ] , we give a detailed presentation of the equilibrium quantities ecom computes during the postprocessing phase , and of the numerical methods we implemented to calculate these quantitites with high accuracy . in section [ sec : accuracy ] we evaluate the speed , accuracy , and convergence properties of our solver , and compare them to those of chease @xcite . in section [ sec : torflow ] we explain how ecom computes stationary equilibria with toroidal flows , and in section [ sec : dis ] we summarize our main findings , discuss the current limitations of ecom and future plans . appendix [ sec : miller ] presents our method to calculate the miller parametrization @xcite of a numerically computed flux contour , and appendix [ sec : namelist ] contains a table with all the important variables in ecom , along with a short description for each of them .
in this section , we briefly review the numerical algorithm used in ecom to solve the g - s equation .
a more detailed presentation of each of the steps described below can be found in @xcite .
the grad - shafranov equation is given by @xmath9 where @xmath10 is the usual cylindrical coordinate system associated with the toroidal geometry , @xmath11 is the poloidal magnetic flux , @xmath12 is the permeability of free space , @xmath13 is the plasma pressure , and @xmath14 , with @xmath15 the toroidal component of the magnetic field . once the free functions @xmath13 and @xmath16 are given and eq .
( [ gs1 ] ) is solved with appropriate boundary conditions , the magnetic field @xmath17 and the current density @xmath18 can be computed according to the following formulae @xmath19 eq .
( [ gs1 ] ) is a second - order elliptic nonlinear partial differential equation for @xmath0 .
ecom solves the fixed boundary problem associated with this equation .
specifically , the boundary curve @xmath20 enclosing the plasma domain @xmath21 of interest is an input to the solver , and ecom solves eq .
( [ gs1 ] ) with the dirichlet data @xmath22 on @xmath20 , where @xmath23 is a constant .
this formulation is particularly convenient for multiphysics theoretical studies of the influence of shaping on plasma performance @xcite .
two types of inputs can be used in ecom to determine the geometry of @xmath20 .
one option is to give an exact representation of the plasma boundary , for example in the form of parametric equations @xcite .
when such a representation is not available , one can also give the coordinates @xmath24 of discrete points on the boundary . at the moment
, ecom can only treat smooth plasma boundaries , and can therefore not compute equilibria whose plasma boundary has a separatrix .
the functional dependence on @xmath0 of the pressure and toroidal magnetic field profiles is either prescribed or determined from transport equations . in both cases , it is an input to ecom . in general , these profiles are such that eq .
( [ gs1 ] ) is nonlinear , and for a wide class of profiles eq .
( [ gs1 ] ) has to be solved as an eigenvalue problem @xcite .
this means that eq .
( [ gs1 ] ) has to be solved by iterating on @xmath0 @xcite . in ecom
, this is done as follows . a normalized flux @xmath25 is defined by @xmath26 , where @xmath27 and @xmath28 are the poloidal flux at the magnetic axis and the last closed flux surface respectively , so that @xmath29 at the magnetic axis and @xmath30 at the last closed flux surface .
the pressure and toroidal magnetic field profiles are also normalized and expressed in terms of @xmath25 according to : @xmath31 defining @xmath32 , eq .
( [ gs1 ] ) then becomes @xmath33 where @xmath34 plays the role of an eigenvalue .
in ecom , there are several options to specify the profiles @xmath35 and @xmath36 , with corresponding namelist parameter iptype for the pressure and iftype for the poloidal current . if iptype=1 or iftype=1 , the profiles are given by an explicit formula in terms of @xmath25 . in ecom , we often use @xmath37 , as is also done in chease @xcite , where the constants @xmath38 , @xmath39 , and @xmath40 are specified in the namelist . likewise
, we often use @xmath41 .
different expressions can be easily implemented , such as formulae describing a steep pressure pedestal @xcite . if iptype=2 or iftype=2 , the profiles are given by a set of data points and the corresponding values of @xmath25 or of the minor radius .
the value of the namelist variable iptable determines whether the tabulated values of the profiles are in terms of @xmath25 or of the minor radius .
if iptable=0 , the numerical tables of @xmath35 and @xmath42 are specified in terms of discrete values of @xmath25 .
if iptable=1 , @xmath43 and @xmath44 instead of their derivatives are specified by tables in terms of @xmath25 . if iptable=2 , the numerical tables of @xmath45 and @xmath46 are specified in terms of the normalized minor radius @xmath47 . in section
[ sec : post ] , we explain in detail how ecom accurately computes @xmath35 and @xmath42 starting from such tables .
furthermore , ecom offers the possibility to choose between three different definitions for the minor radius @xmath47 , corresponding to three different values of the namelist variable irho . if irho=0 , the minor radius is defined by @xmath48 , where @xmath49 is the outer location of the flux surface @xmath25 at @xmath50 and @xmath51 is the location of the magnetic axis .
if irho=1 , @xmath47 is defined by @xmath52 where @xmath53 is the inner location of the flux surface @xmath25 at @xmath50 .
if irho=2 , @xmath47 is defined by @xmath54 .
as already mentioned , eq .
( [ eigen ] ) must be solved iteratively .
g - s solvers usually iterate on the operator @xmath55 @xcite .
one of the key ideas in ecom is to iterate on the laplacian operator @xmath56 instead .
this can be done without any loss of generality , and does not incur any additional computational cost .
the advantage is that at a given iteration , one now has to solve poisson s equation , and one can therefore rely on the larger body of numerical methods developed for fast high order poisson solvers in two dimensions . to solve for @xmath25 and the smallest eigenvalue @xmath34 in eq.([eigen ] ) , ecom uses a modified version of the inverse iteration method @xcite .
specifically , if @xmath57 and @xmath58 are known at the iteration step @xmath59 , then @xmath60 and @xmath61 are computed according to @xmath62\label{poisson1}\\ & \psi^{(i+1)}=\frac{\tilde{\psi}^{(i+1)}}{||\tilde{\psi}^{(i+1)}||_{\infty}}\label{psiup}\\ & \lambda^{(i+1)}=\frac{\lambda_{i}}{||\tilde{\psi}^{(i+1)}||_{\infty}}.\label{lambdaup}\end{aligned}\ ] ] where @xmath63 is the extremum of @xmath64 in the domain @xmath21 . in ecom ,
the iterative process terminates when @xmath65 for some predetermined small @xmath66 . in the numerical results presented in this article
, we typically had @xmath67 . in order to keep consistency across different grid sizes ,
@xmath63 is not only calculated for all values of @xmath64 on the mesh , but instead over the entire domain @xmath21 . in ecom
, this is done by finding the location where @xmath68 with the newton - raphson method .
it requires knowledge of the hessian matrix , whose values away from grid points are evaluated by interpolation , based on the high order fourier and chebyshev representations ecom uses for the poisson solver on the unit disk .
the extremum of @xmath25 on the grid is used to provide a very good initial guess , so that in practice very few newton steps are subsequently required to find @xmath69 in @xmath21 .
it is known empirically that iterative schemes such as the one above converge faster when the right - hand side of the partial differential equation is slowly varying . for better convergence
we thus scale the unknown function @xmath25 as @xmath70 . replacing @xmath25 with @xmath71 , eq.([poisson1 ] )
becomes @xmath72\label{poisson2}\ ] ] unlike eq .
( [ poisson1 ] ) , the right - hand side of eq .
( [ poisson2 ] ) does not have derivatives of @xmath25 , and it is therefore smoother . note that while one solves for @xmath71 in the poisson step , the normalization steps ( [ psiup ] ) and ( [ lambdaup ] ) are still computed in terms of @xmath25 .
thus far , we have explained how ecom treats the grad - shafranov equation as a nonlinear poisson problem .
we now briefly describe how ecom computes the solution of the poisson equation ( [ poisson2 ] ) at each iteration , i.e. with fixed right - hand side , on a domain @xmath21 of fusion interest and with the dirichlet boundary condition @xmath73 on @xmath20 .
the poisson solver in ecom is based on two elements : 1 ) a spectrally accurate numerical method to compute the conformal map from the plasma domain @xmath21 to the unit disk ; 2 ) a fast , high order poisson solver on the unit disk .
conformal mapping is an effective method for solving poisson s equation because a conformal map transforms a laplacian operator into another laplacian operator , with a scale factor @xcite .
consider the generic poisson equation @xmath74 the conformal map @xmath75 from @xmath21 to the unit disc @xmath76 and its inverse map @xmath77 .
solving eq .
( [ poisson_generic ] ) is equivalent to solving the following poisson problem in @xmath76 : @xmath78 where @xmath79 , and the functions @xmath80 and @xmath81 should be seen as the real and imaginary parts of the inverse map @xmath82 . clearly , solving eq .
( [ poisson_mapped ] ) is easier than solving eq .
( [ poisson_generic ] ) , provided one has a way to calculate the inverse map @xmath82 at both boundary and interior points of @xmath76 . in ecom , this is done as follows .
ecom first computes the forward map @xmath83 for points on @xmath20 that are equispaced in arc length through the kerzman - stein integral equation based on the szeg kernel @xcite .
using oversampling and interpolation , ecom then uses the boundary values of the forward map to calculate the inverse map @xmath80 and @xmath81 for points on the boundary of @xmath76 that are equispaced in the polar angle @xmath84 .
finally , ecom computes @xmath82 for points in the interior of @xmath76 using the cauchy integral formula and the fast fourier transform @xcite .
ecom relies on a somewhat naive implementation of the kerzmann - stein integral equation for the computation of @xmath83 on @xmath20 , that requires @xmath85 work , where @xmath86 is the number of discretization points on @xmath20 .
there exist methods resulting in an asymptotic @xmath87 run time @xcite , but they are not currently implemented in ecom because the computation only needs to be done once , and because other steps in ecom are more expensive .
the computation of the inverse map at interior points is based on the cauchy integral formula and the fast fourier transform , and results in a run time complexity of @xmath88 , where @xmath89 is the number of radial grid points and @xmath90 the number of angular grid points for the mesh that ecom uses to solve poisson s equation on @xmath76 the boundary @xmath20 can be defined in several ways , that are specified by the namelist variable ibtype in ecom . if ibtype=0 , the plasma boundary corresponds to the contour @xmath91 of a solovev equilibrium we discuss in section [ sec : solovev ] , parametrized by eqs .
( [ solob1])-([solob2 ] ) . if ibtype=1 , the boundary is specified by the miller parametrization @xcite given by eqs .
( [ millerb1])-([millerb2 ] ) of section [ sec : exnl ] , and the elongation @xmath92 and triangularity @xmath93 must then be specified .
if ibtype=2 , the boundary is specified by a set of discrete points @xmath94 .
this allows the computation of equilibria specified by experimental data and is also the method of choice to compute up - down asymmetric equilibria .
the conformal mapping routine requires that the points on @xmath20 be equispaced in arc length .
when ibtype=0 or ibtype=1 , these points are easily calculated from the parametric equations for the boundary .
when ibtype=2 , ecom uses lagrange interpolation to compute these points . in principle , the conformal map has to be computed only once , at the beginning of the iterative procedure corresponding to eqs .
( [ psiup])-([poisson2 ] ) .
however , as one might physically expect , computing flux surface quantities is much more convenient if the point that is mapped to the center of @xmath76 coincides with the magnetic axis .
to facilitate the calculation of these quantities , which takes place after the g - s equation is solved , ecom recomputes the conformal map several times within the poisson iterations to adjust the center of @xmath76 to the magnetic axis .
we have empirically observed that the conformal map only needs to be recomputed a few times .
the riemann mapping theorem guarantees the existence and uniqueness of an analytic map between any simply connected plasma cross section @xmath21 and the unit disk .
however , this does not mean that conformal mapping is a practical numerical method for any arbitrary plasma shape .
the issue is that points on the boundary of @xmath76 that are equispaced in the angle @xmath84 are not necessarily mapped , under the inverse map , to points that resolve the boundary of @xmath21 in the desired fashion .
it is well known , for example , that if @xmath21 is an elongated ellipse , uniformly spaced points in @xmath84 on the boundary of @xmath76 correspond to a distribution of points on the boundary of @xmath21 which is sparse on the curved parts and crowded on the flat parts of the ellipse @xcite . in the remainder of this article
, we will call this phenomenon the crowding effect " .
it has two direct implications for ecom .
first , ecom can only treat in a robust manner domains that have a smooth boundary , and can therefore not be used to compute equilibria with a magnetic x - point on the plasma boundary .
second , ecom is particularly efficient for plasma shapes that are not too elongated , as is the case for conventional tokamaks . as elongation is increased , the high order convergence properties are maintained for @xmath0 and its derivatives , but a higher number of grid points is required to reach a certain level of accuracy @xcite .
we will go back to this point in section [ sec : dis ] .
we finish this section by describing how ecom solves poisson s equation on the unit disk , @xmath95 as required in eq .
( [ poisson_mapped ] ) .
the solver uses separation of variables in the usual polar coordinates @xmath96 and expands @xmath97 and @xmath98 as fourier series @xmath99 substituting these expressions into poisson s equation , we get the following ordinary differential equation for each @xmath100 : @xmath101 where the boundary condition at @xmath102 is obtained by requiring the regularity of the solution at this point . for each @xmath100 , a solution of eq .
( [ ode ] ) that does not satisfy the boundary condition at @xmath103 can be written in terms of convolutions with the green s function associated with eq .
( [ ode ] ) that has the proper behavior at @xmath102 and @xmath104 : @xmath105 for each @xmath100 , the general solution to the homogeneous equation @xmath106 satisfying the regularity condition at @xmath102 can also be written explicitly : @xmath107 where @xmath108 is a constant to be determined from the boundary condition at @xmath103 .
setting @xmath109 , we can then write the solution of eq .
( [ ode ] ) satisfying the proper boundary conditions as follows : @xmath110 one of the major advantages of the green s function formulation and the formula ( [ radial_sol ] ) used in ecom is that partial derivatives of @xmath97 can be calculated explicitly from the formulae in eqs .
( [ particular ] ) and ( [ homosol ] ) @xcite .
numerical differentiation is never required , which is one of the main reasons why our numerical method leads to partial derivatives of @xmath25 that have the same order of convergence as @xmath25 . in ecom ,
the computation and sum of the fourier series are done with the fast fourier transform . the angular grid on the unit disk is uniformly spaced in the polar angle @xmath84 , to guarantee the spectral accuracy of the representation for smooth data .
the number of grid points in the @xmath84 direction is @xmath90 . in the radial direction
, ecom uses a piecewise chebyshev grid .
specifically , the interval @xmath111 $ ] is divided into @xmath112 subintervals , and on each of these subintervals a chebyshev grid of order @xmath113 is constructed .
the number of grid points in the radial direction is then @xmath114 .
@xmath115 is the default setting in ecom .
the convolutions with the green s function in eq .
( [ particular ] ) are computed with a 16th order gaussian quadrature rule .
reference @xcite describes ways to avoid the computational issues associated with the rapid growth and decay of the monomials @xmath116 and @xmath117 for large @xmath100 , as well as the recursive algorithm used to compute these integrals in @xmath118 work .
the run time complexity of the poisson solver on the disk is @xmath88 : ecom computes @xmath118 ffts of size @xmath90 at a cost of @xmath88 and solves @xmath90 radial ordinary differential equations at a cost of @xmath119 .
the flow diagram in figure [ ecomdiagram ] presents a condensed view of the iterative scheme used in ecom , as a summary of section [ sec : gs ] .
the initialization step corresponds to the specification of the grid resolution and of the values of the namelist variables presented in this article .
the parameters @xmath120 , @xmath121 and @xmath122 in figure [ ecomdiagram ] refer to discretizations used during postprocessing , after the g - s equation is solved , and are defined in section [ sec : post ] .
the main purpose of postprocessing is to compute equilibrium quantities that play a key role in heating and current drive , stability and transport calculations , and to scale the normalized solution @xmath25 to the physical flux @xmath0 .
in addition , ecom can compute the parallel current density using reduced models for the ohmic current and the bootstrap current , check the mercier criterion for each flux surface @xcite and the global troyon limit @xcite , and calculate the miller parametrization that best fits a flux surface chosen by the user .
the evaluation of flux quantities requires integration along contours of constant poloidal magnetic flux @xmath25 .
a flux coordinate system @xmath123 is more convenient for such integrals than the @xmath124 coordinate system used by ecom to solve the g - s equation .
for postprocessing , ecom thus constructs the following flux coordinate system .
@xmath125 is chosen to be the usual toroidal angle @xmath125 , with @xmath126 , while the poloidal angle @xmath127 is defined by @xmath128 if @xmath129 , @xmath130 if @xmath131 , where @xmath132 is the four - quadrant inverse tangent and @xmath133 is the position of the magnetic axis . in terms of the @xmath123
coordinates , the flux surface average of a function @xmath134 @xcite , written @xmath135 , is defined by @xmath136 where @xmath137 is the jacobian of the transformation between cartesian coordinates and the @xmath123 coordinate system : @xmath138}\left[(r - r_0)\frac{\partial \psi}{\partial r}+(z - z_0)\frac{\partial \psi}{\partial z}\right]\end{aligned}\ ] ] as will soon be apparent , several physical quantities are best expressed in terms of the three functions @xmath139 , @xmath140 , and @xmath141 defined by @xmath142 rewriting eq .
( [ eigen ] ) using flux coordinates , a simple relationship can be obtained that links @xmath143 , @xmath144 , and @xmath145 .
indeed , eq . ( [ eigen ] ) takes the following form in flux coordinates @xcite : @xmath146 taking the flux surface average of @xmath147 eq .
( [ eq : gs_flux ] ) , we obtain the desired relationship between @xmath143 , @xmath144 , and @xmath145 : @xmath148 the plasma volume inside the flux surface @xmath25 is @xmath149 and the total plasma volume is @xmath150 .
the toroidal current within the flux surface @xmath25 is @xmath151 where we have used eq .
( [ gsrsec3 ] ) and @xmath152 .
the total toroidal current in @xmath21 is @xmath153 .
the volume averaged pressure @xmath154 is given by @xmath155 where we have used integration by part , combined with @xmath156 , @xmath157 , and @xmath158 .
the total beta is @xmath159 where @xmath160 is the vacuum field at the point @xmath161 , with @xmath162 .
our definitions of the poloidal beta and of the internal inductance are the same as jardin s @xcite : @xmath163 the total poloidal magnetic field energy is @xmath164 finally , for each flux surface , the safety factor is defined as @xmath165}{2\pi}\int_0^{2\pi } d\theta \frac{j}{r^2}=\lambda\frac{f[\psi(\psi)]}{2\pi } i_c\label{q1}.\end{aligned}\ ] ] the normalized radius can be used to find the differential volume @xmath166 and the differential flux @xmath167 , which are often used in transport or mhd analysis .
also , the magnetic shear can be defined in terms of @xmath47 by @xmath168 to evaluate the integrals in eq .
( [ int123 ] ) numerically , one needs to find the location of the desired flux contours , which in general do not coincide with the @xmath94 grid of the poisson solver , and then integrate the integrands along these contours . in ecom
, this is done with the following three steps : ( i ) ecom first determines the radial location of @xmath120 flux surfaces for each angle @xmath84 in the unit disk @xmath169 ; ( ii ) ecom then evaluates the integrands at the corresponding @xmath94 points in @xmath21 , and interpolates the integrands to a grid that is equispaced in the angle @xmath127 ; ( iii ) ecom finally computes the integrals along contours of constant @xmath25 in the domain @xmath21 using the trapezoidal rule . since the numerical methods that are used for interpolation and integration in steps ( i)-(iii ) are spectrally accurate , ecom computes the location of the flux contours and the integrals in eq .
( [ int123 ] ) without significant loss of accuracy , as we will demonstrate in section [ sec : accuracy ] .
we now describe steps ( i ) , ( ii ) , and ( iii ) in more detail . for a description of step ( i ) , we put ourselves in the situation in which the g - s solver has computed the values @xmath170 of the flux @xmath25 on the grid of the unit disk @xmath76 , and we imagine that we want to determine the location of the contour @xmath171 on @xmath76 .
ecom does this as follows . for each angle @xmath172
, ecom first finds the chebyshev subinterval of the radial grid for which @xmath173 , where @xmath174 is the index of the first chebyshev point in that radial subinterval .
once the subinterval is found , ecom constructs a local continuous approximation @xmath175 of @xmath25 in the radial direction from the known values @xmath176 on the subinterval of interest and the chebyshev grid for that subinterval .
specifically , @xmath177 is written as the following sum : @xmath178 where the functions @xmath179 are the chebyshev polymials associated with the chebyshev grid of the subinterval , and where the coefficients @xmath180 are given by the expression @xmath181 where @xmath182 is the kronecker delta .
ecom then uses the expansion in eq .
( [ eq : cheby_rep ] ) to find the radial position satisfying @xmath183 with a newton - raphson iterative method : @xmath184 this root finding process usually converges in a few iterations , typically @xmath185 , and the total cost to find the location of @xmath186 contours at @xmath90 angles is @xmath187 . finally ,
after the radial position of a given contour is found , ecom uses @xmath188 and @xmath189 for @xmath190 and chebyshev representations analogous to eq .
( [ eq : cheby_rep ] ) to accurately evaluate @xmath191 and @xmath192 at the location of the flux contours . at the end of step ( i )
the radial position of the specified flux contours are known for each angle @xmath172 in @xmath76 , and so are the values of the integrands in eq .
( [ int123 ] ) at these points . through the backward map ,
all these quantities are also known in @xmath21 . in eq .
( [ int123 ] ) , one integrates quantities that are @xmath193-periodic in @xmath127 over the period @xmath194 $ ] .
numerically , this can be done very accurately with a trapezoidal - rule quadrature , provided that the integrands are known on an equispaced @xmath127 grid . since the equispaced @xmath84 grid in @xmath76
is not mapped to an equispaced @xmath127 grid , the goal of step ( ii ) is to interpolate the quantities computed in step ( i ) on a grid in @xmath21 that is equispaced in @xmath127 . because of the crowding effect that is inherent to the conformal mapping technique , the angular grid resulting from the inverse map underresolves certain regions of @xmath21 . for accurate
interpolation , it is thus desirable to first oversample the integrands to be interpolated .
the oversampling is done by refining the equispaced @xmath84 grid of @xmath76 using the fft , assuming band - limited integrands . as a result of this ,
the integrands are known at @xmath195 angular grid points , where @xmath196 is the oversampling factor . in principle , these values could then be mapped back to @xmath21 , and interpolated . in practice
, however , the backward mapping of the @xmath195 grid points for @xmath120 contours from @xmath169 to @xmath21 requires @xmath197 operations , and results in significant computational time when @xmath196 is as large as desired for accurate interpolation .
note that the run - time complexity is not of the form @xmath198 as it was for the backward mapping of the @xmath96 grid points in @xmath76 because the calculation of the backward map for the flux contours can not be trivially accelerated by the fft .
the reason for this is that the radial location of the flux contours varies as a function of @xmath84 . in order to reduce the computational cost , we only compute the backward map for @xmath121 points among the @xmath195 oversampled points , with @xmath199 , and chosen such that their mapped positions are the closest to the target grid of @xmath122 points equispaced in @xmath127 .
these @xmath121 points are found by computing the backward map of a small number of points lying on the contour of interest in @xmath76 with increasing angle @xmath84 across the interval @xmath194 $ ] , and using @xmath200-th order lagrange interpolation to construct an approximation of the function @xmath201 on that contour .
we usually take @xmath202 and @xmath203 .
once the @xmath121 non - equispaced points are found , we use trigonometric interpolation for a periodic function @xcite to interpolate the integrands at the @xmath121 points to the @xmath122 equispaced points .
this requires @xmath204 work to find the barycentric factors , and @xmath205 work to interpolate at @xmath122 points . once step ( ii )
is completed , step ( iii ) is straightforward .
the contour integrals in eq .
( [ int123 ] ) are computed from the integrands on the equispaced @xmath127 grid using the trapezoidal rule . since the integrands are smooth and periodic in @xmath127 , and since the @xmath127 grid is uniform , the trapezoidal rule is spectrally accurate @xcite .
the required work for the trapezoidal - rule quadrature is very small , @xmath206 . a single solution of the normalized form of the g - s equation as given in eq .
( [ eigen ] ) can describe an infinite sequence of axisymmetric equilibria that have a different total toroidal current @xmath207 and a different safety factor @xmath208 at the magnetic axis . to understand these degrees of freedom , consider that ecom has just computed the eigenvector - eigenvalue solution @xmath209 of eq .
( [ eigen ] ) .
the normalized total toroidal current @xmath210 can be calculated from this solution according to eq .
( [ torcur1 ] ) .
all is then needed to obtain an equilibrium with the desired total toroidal current @xmath211 is the simple rescaling @xmath212 , which is equivalent to rescaling @xmath0 . from the definitions in eq .
( [ normal_prof ] ) , it is clear that @xmath213 and @xmath36 are also scaled by @xmath34 , and must be rescaled as well : @xmath214 and @xmath215 .
once @xmath34 is fixed , there still is a degree of freedom for the determination of @xmath216 , because @xmath216 depends on @xmath217 instead of @xmath218 , as can be seen in eq .
( [ q1 ] ) . in ecom
, this degree of freedom can be removed by specifying the value of either the poloidal current or the safety factor at a certain radial location .
there are several options for the scaling of the normalized equilibrium in ecom , with corresponding namelist variable iscale .
they are summarized in table [ tab : iscale ] . if iscale=0 or iscale=1 , @xmath34 is not rescaled , so that the total toroidal current is the normalized total toroidal current @xmath210 . when iscale=0 , the degree of freedom associated with @xmath216 is removed by specifying either @xmath219 or @xmath220 .
for @xmath219 , the additional namelist variable ifpol needs to be set to 0 , while for @xmath220 ifpol needs to be set to 1 .
when iscale=1 , @xmath221 is specified .
if iscale=2 or iscale=3 , @xmath34 is rescaled so that the total toroidal current is adjusted to the desired toroidal current @xmath211 .
the choices for constraining the @xmath216 profile are the same as before : when iscale=2 , either @xmath219 or @xmath220 is given , depending on the value of ifpol , and when iscale=3 , @xmath221 is given . finally , a last option to fix @xmath34 in ecom is to specify @xmath221 and either @xmath219 or @xmath220 .
this option corresponds to iscale=4 . [ cols="^,^,^",options="header " , ]
13 natexlab#1#1bibnamefont # 1#1bibfnamefont # 1#1citenamefont # 1#1url # 1`#1`urlprefix[2]#2 [ 2][]#2 c.r .
sovinec , a.h .
glasser , t.a .
gianakon , d.c .
barnes , r.a .
nebel , s.e .
kruger , d.d .
schnack , s.j .
plimpton , a. tarditi , m.s .
chu , and the nimrod team , _ journal of computational physics _ * 195 * , 355 ( 2004 ) j.r .
cary , j. candy , j. cobb , r.h .
cohen , t. epperly , d.j .
estep , s. krasheninnikov , a. d. malony , d.c .
mccune , l. mcinnes , a. pankin , s. balay , j.a .
carlsson , m.r .
fahey , r.j .
groebner , a.h .
hakim , s.e .
kruger , m. miah , a. pletzer , s. shasharina , s. vadlamani , d. wade - stein , t.d .
rognlien , a. morris , s. shende , g.w .
hammett , k. indireshkumar , a. yu . pigarov and h. zhang , _ journal of physics : conference series _ * 180 * , 012056 ( 2009 ) c.s .
chang , s. ku , p. diamond , m. adams , r. barreto , y. chen , j. cummings , e. dazevedo , g. dif - pradalier , s. ethier , l. greengard , t.s .
hahm , f. hinton , d. keyes , s. klasky , z. lin , j. lofstead , g. park , s. parker , n. podhorszki , k. schwan , a. shoshani , d. silver , m. wolf , p. worley , h. weitzner , e. yoon and d. zorin , _ journal of physics : conference series _ * 180 * , 012057 ( 2009 ) i. voitsekhovitch , b. alper , m. brix , r.v .
budny , p. buratti , c.d .
challis , j. ferron , c. giroud , e. joffrin , l. laborde , t.c .
luce , d. mccune , j. menard , m. murakami , j.m .
park , and jet - efda contributors , _ nuclear fusion _ * 49 * , 055026 ( 2009 ) s.d .
scott , p.h .
diamond , r.j .
fonck , r.j .
goldston , r.b .
howell , k.p .
jaehnig , g. schilling , e.j .
synakowski , m.c .
zarnstorff , c.e .
bush , e. fredrickson , k.w .
hill , a.c .
janos , d.k .
mansfield , d.k .
owens , h. park , g. pautasso , a.t .
ramsey , j. schivell , g.d .
tait , w.m .
tang , and g. taylor , _ physical review letters _ * 64 * , 531 ( 1990 ) l.l .
lao , k.h .
burrell , t.s .
casper , v.s .
chan , m.s .
chu , j.c .
deboo , e.j .
doyle , r.d .
durst , c.b .
forest , c.m .
greenfield , r.j .
groebner , f.l .
hinton , y. kawano , e.a .
lazarus , y.r .
linliu , m.e .
mauel , w.h .
meyer , r.l .
miller , g.a .
navratil , t.h .
osborne , q. peng , c.l .
rettig , g. rewoldt , t.l .
rhodes , b.w .
rice , d.p .
schissel , b.w .
stallard , e.j .
strait , w.m .
tang , t.s .
taylor , a.d .
turnbull , r.e .
waltz , and the diiid team , _ physics of plasmas _ * 3 * 1951 ( 1996 ) j.e .
menard , r.e .
bell , e.d .
fredrickson , d.a .
gates , s.m .
kaye , b.p .
leblanc , r. maingi , s.s .
medley , w.park , s.a .
sabbagh , a. sontag , d. stutman , k. tritz , w. zhu , and the nstx research team , _ nuclear fusion _ * 45 * 539 ( 2005 ) | we present ecom ( equilibrium solver via conformal mapping ) , a fast and accurate fixed boundary solver for toroidally axisymmetric magnetohydrodynamic equilibria with or without a toroidal flow .
ecom combines conformal mapping and fourier and integral equation methods on the unit disk to achieve exponential convergence for the poloidal flux function as well as its first and second partial derivatives . as a consequence of its high order accuracy , for dense grids and tokamak - like
elongations ecom computes key quantities such as the safety factor and the magnetic shear with higher accuracy than the finite element based code chease [ h. ltjens _ et al .
_ , computer physics communications 97 , 219 ( 1996 ) ] at equal run time .
ecom has been developed to provide equilibrium quantities and details of the flux contour geometry as inputs to stability , wave propagation and transport codes . |
saturable absorption ( sa ) is an extreme nonlinear phenomenon that consists in the quenching of optical absorption under high - intensity illumination .
this effect , which is an inherent property of photonic materials , constitutes a key element for passive mode - locking ( pml ) in laser cavities @xcite , where continuous waves are broken into a train of ultrashort optical pulses .
most materials undergo saturable absorption at very high optical intensities , in close proximity to their optical damage threshold .
currently , state - of - the - art semiconductor - based sa mirrors are routinely employed for pml lasers @xcite .
however , these mirrors operate in a narrow spectral range , are poorly tunable , and require advanced fabrication techniques . recently , carbon nanomaterials have emerged as an attractive , viable , and cost - effective alternative for the development of next - generation pml lasers .
for example , carbon nanotubes ( cnts ) undergo sa at rather modest light intensities , while their operation wavelength ( determined by the energy band gap ) can be manipulated by tuning their diameter @xcite .
broadband operation has been demonstrated by using an ensemble of cnts with a wide distribution in diameter , at the expense of higher linear loss from off - resonance tubes @xcite .
graphene overcomes this limitation thanks to its peculiar conical band structure , which gives rise to broadband resonant sa at remarkably low light intensity @xcite that can further be tuned by means of an externally applied gate voltage @xcite .
graphene - based sa components have been used to achieve pml ultrafast laser operation @xcite , broadband tunability @xcite , and quality - factor switching @xcite .
graphene multilayers have also been employed to generate large energy pulses @xcite and to achieve pml in fiber lasers with normal dispersion @xcite .
in addition , recent theoretical investigations predict single - mode operation of random lasers by embedding graphene flakes in a gain medium @xcite . despite the rising interest in sa of graphene for developing the above mentioned applications , a transparent and accurate theory of graphene sa is still missing .
this is most likely due to the fact that sa is inherently a non - perturbative phenomenon , which can not be accurately described by standard perturbative approaches used to calculate nonlinear graphene conductivities @xcite .
furthermore , theoretical approaches relying on the boltzmann transport equation ( accounting only for intraband electron dynamics ) @xcite are inadequate to simulate sa of graphene , which arises due to the interplay of intraband and interband dynamics . in principle
, sa can be modeled through time - domain integration of the single - particle density matrix equations in graphene nanoislands @xcite .
however , the computational demand of this approach becomes unaffordable when modeling nanostructures with dimensions @xmath0 nm .
then , in this regime , the single - particle massless dirac fermion ( mdf ) picture @xcite appears to be an appropriate theoretical environment for describing graphene sa . in this article , we calculate intraband and interband contributions to sa of extended graphene by non - perturbatively solving the single - particle dirac equation for mdfs in the presence of an external electromagnetic field .
we further investigate the dependence of the intensity - saturated graphene conductivity on doping , temperature , and optical frequency .
interestingly , we find a remarkably low intensity threshold for sa , which is consistent with available experimental reports @xcite .
our calculations indicate a strong quenching of absorption depth produced by electrical doping ( which can be controlled through gating ) , as well as a weak dependence on electron temperature .
the sa behavior predicted in the mdf picture for extended graphene is found to be in excellent agreement with atomistic time - domain simulations of graphene nanoribbons based on the tight - binding / single - particle density matrix formalism @xcite .
we understand that the present findings are highly relevant for the future development of graphene - based pml fibre lasers and single - mode random lasers .
we consider a monochromatic optical field @xmath1 of angular frequency @xmath2 and complex amplitude @xmath3 impinging at normal incidence on a self - standing extended graphene sheet [ see fig .
[ fig1](a ) ] . at visible and lower frequencies , electrons in this material
behave as mdfs , with their temporal evolution governed by the single - particle dirac equation @xcite @xmath4 where @xmath5 is the electron momentum , @xmath6 is the time derivative , @xmath7 is the fermi velocity , @xmath8 is the speed of light in vacuum , @xmath9 is the reduced planck constant , @xmath10 is the two - dimensional ( 2d ) pauli - matrix vector , and @xmath11 is the @xmath5- and time - dependent two - component spinor accounting for the quantum states in the upper and lower dirac - cone bands . here , we introduce an electron quasi - momentum @xmath12 that coincides with the unperturbed momentum ( i.e. , @xmath13 ) in the absence of external illumination . in this case , eq .
( [ diraceq ] ) admits spinor eigenvectors @xmath14 where @xmath15 is the unperturbed conical dispersion of upper ( @xmath16 ) and lower ( @xmath17 ) energy bands [ see fig .
[ fig1](b ) ] , while @xmath18 identifies the momentum direction , such that @xmath19 , and a spatial dependence @xmath20 ( normalized to the sheet area @xmath21 ) is understood in the spinor . ,
amplitude @xmath3 , and intensity @xmath22 ) normally impinging on a self - standing graphene sheet . *
( b ) * illustration of the temporal displacement of the dirac cone along the @xmath23 direction due to the oscillatory optical field .
the central cone represents the unperturbed conical dispersion of graphene ( @xmath24 ) , while left and right cones indicate the maximum achievable displacement.,scaledwidth=50.0% ] in the presence of the impinging optical field , we use the customary minimal electron - light coupling prescription to write the electron quasi - momentum as @xmath25 , where @xmath26 is the electron charge and @xmath27 is the potential vector in the coulomb gauge ( @xmath28 ) . without loss of generality , we assume the external light to be linearly polarized along the in - plane @xmath29 direction , so that @xmath30 and @xmath31
. therefore , the incident field induces an oscillatory shift of the unperturbed bands @xmath24 along @xmath32 around @xmath33 [ see fig .
[ fig1](b ) ] . following the non - perturbative approach developed by ishikawa @xcite , which we review in this section for the sake of completeness , we write the time - dependent spinor as a linear combination of the instantaneous upper- and lower - cone states @xmath34 where @xmath35 @xmath36 is a global dynamical phase , and @xmath37\}$ ] is the time - dependent direction angle of the electron quasi - momentum @xmath12 .
we now insert the ansatz provided by eq .
( [ ansatzeq ] ) into eq .
( [ diraceq ] ) and define the interband coherence @xmath38 and the population difference @xmath39 . without making any approximations , we can then rewrite the two - dimensional dirac equation for mdfs in the form of generalized bloch equations ( gbes ) @xcite , @xmath40 [ gbeqs ] the carrier current is thus obtained by subtracting the valence band current ( assuming the valence band is fully populated ) from the total electron current , since it depends only on the carrier occupation , resulting in @xcite @xmath41 { \bf \hat{x } } + \nonumber \\ & + \left [ ( n_{\bf p}+1)\sin \theta_{\bf p } + 2 \cos \theta_{\bf p } { \rm i m } \left\ { \rho_{\bf p } { \rm e}^{-2i\omega_{\bf p } } \right\ } \right ] { \bf \hat{y}}. \nonumber\end{aligned}\ ] ] we emphasize that this expression is obtained non - perturbatively , without any approximation beyond the mdf picture .
the @xmath42 terms account for the intraband current , while the remaining terms , which depend on the coherence @xmath43 , arise from interband dynamics @xcite .
the macroscopic induced surface current @xmath44 is finally obtained by integrating over all electron momenta , @xmath45 where @xmath46 account for spin and valley degeneracies , and the integral extends over the entire 2d @xmath5 plane .
we remark that , although the carrier current @xmath47 has both @xmath48 and @xmath49 components , the integrated macroscopic current remains polarized along @xmath48 and thus parallel to the external field .
this theoretical framework describes the optical response of graphene in a single - electron picture , but it still neglects inelastic electron transitions that are produced for example by impurity scattering and coupling to phonons . in the following sections ,
these interactions are introduced phenomenologically through an effective electron lifetime .
additionally , we argue that the response of graphene , quantified as a @xmath50 absorption at low intensities , is sufficiently weak as to ignore coulomb self - interaction among induced charges @xcite .
we focus first on the intraband contribution to saturable absorption , i.e. , neglecting interband dynamics .
this contribution is dominant in doped graphene for photon energies smaller than twice the fermi level ( i.e. , @xmath51 ) , for which interband transitions are prohibited by the pauli exclusion principle .
consequently , in this regime , one need not solve explicitly the gbes , as it is sufficient to set @xmath52 and @xmath53 , where @xmath54\}$ ] is the fermi - dirac occupation number , @xmath55 is the chemical potential , @xmath56 is the boltzmann constant , and @xmath57 is the electron temperature .
the resulting macroscopic current then reduces to @xmath58 \cos \theta_{\bf p}(t ) d^2{\bf p } \label{intraintegraleq}.\ ] ] now , in order to account for inelastic electron collisions , we modify heuristically the definition of the direction angle @xmath59 , assuming that the intraband electron quasi - momentum satisfies the differential equation @xmath60 where @xmath61 is the characteristic inelastic collision time .
we assume a value @xmath62fs throughout this work , which is consistent with recent experiments @xcite .
we find eq .
( [ pitauintra ] ) to admit the straightforward analytical solution @xmath63 , where @xmath64 ( lower panel ) induced in extended graphene by a harmonic external electric field ( upper panel ) with maximum amplitude @xmath65 and optical frequency @xmath66 , where @xmath67 nm is the optical excitation wavelength . in the lower panel
we provide results for incident intensities @xmath68 , 100 , and 1000gw@xmath69@xmath70 , compared with the maximum achievable current density @xmath71 in the @xmath72 limit ( dashed curve ) .
we assume a fermi energy @xmath73ev and zero temperature.,scaledwidth=50.0% ] this modified expression effectively accounts for intraband electron collisions , and in particular , the direction angle of the electron momentum now becomes @xmath74\}$ ] .
incidentally , the @xmath75 limit of @xmath76 readily reduces to the expression @xmath77 that was used in sec .
[ theoreticalmodel ] . in the limit of vanishing temperature ( @xmath78 )
the chemical potential coincides with the fermi energy @xmath79 , the upper band occupation distribution becomes @xmath80 , where @xmath81 is the heaviside step function , and the integral in eq .
( [ intraintegraleq ] ) can be solved analytically [ see appendix [ a1 ] ] : @xmath82 ^ 2 } \label{intracurrentt0 } \\ & \times\int_0^{\pi/2 } \cos\phi\left[\sqrt{1+f(t)\cos\phi}-\sqrt{1-f(t)\cos\phi}\right]d\phi , \nonumber\end{aligned}\ ] ] where @xmath83 is the fermi momentum , @xmath84 ^ 2\}$ ] , and the integral over @xmath18 can be expressed in terms of generalized jacobi elliptic functions @xcite ( see appendix [ a1 ] ) .
we emphasize that the intraband surface current density in eq .
( [ intracurrentt0 ] ) is highly nonlinear when @xmath85 , where @xmath86 is the intraband saturation field .
additionally , the series expansion of eq .
( [ intracurrentt0 ] ) in powers of the electric field amplitude @xmath87 , @xmath88 fully reproduces the result obtained by means of the boltzmann transport equation @xcite . in order to illustrate intraband saturable absorption of extended graphene , we plot in fig .
[ fig2 ] the intraband current density calculated from eq .
( [ intracurrentt0 ] ) for several values of the incident light intensity @xmath89 , where @xmath90 is the vacuum permittivity .
notice that , when @xmath91 , where @xmath92 is the saturation intensity ( e.g. , @xmath93gw@xmath69@xmath70 at @xmath67 nm ) , the current density also saturates , acquiring a square - like - wave temporal profile .
this can be verified analytically upon examination of the @xmath94 limit of eq .
( [ intracurrentt0 ] ) , which yields the maximum achievable surface current density in doped graphene , @xmath95{\bf \hat{x}},\ ] ] where @xmath96 is the density of doping electrons .
then , as a consequence of current saturation , intraband absorption also saturates .
we describe this effect quantitatively by defining an intraband absorption coefficient as the ratio of the time average of the absorbed power , which is simply evaluated over a single optical cycle , to the incident - light intensity , @xmath97 this quantity reaches its maximum value @xmath98 $ ] at low intensities and zero temperature .
in contrast , it vanishes as @xmath99 in the limit of large incident intensity @xmath22 .
we further examine the @xmath22 dependence of @xmath100 for several fermi energies [ fig .
[ fig3](a ) ] and electron temperatures [ fig . [ fig3](b ) ] by numerically solving eqs .
( [ intraintegraleq ] ) and ( [ intraabscoeffeq ] ) .
a strong dependence on fermi energy is observed , as well as large thermal modulation at high electron temperatures exceeding @xmath101k .
we find that the depth of thermal modulation in the intraband absorption is roughly proportional to the fermi energy and vanishes in undoped graphene .
typical intensities at which the intraband absorption saturates are of the order @xmath102gw@xmath69@xmath70 , depending on the doping level . note that the intraband saturation intensity @xmath103 scales as the inverse square of the optical wavelength ( @xmath104 ) , thus changing by few orders of magnitude within the optical and near - infrared spectrum .
we now turn our attention to the effect of interband transitions on saturable absorption , neglecting intraband dynamics .
this contribution becomes dominant for photon energies @xmath105 ( see below ) .
interestingly , while the optical momentum @xmath106 was found in sec .
[ intrabandsat ] to produce sizable corrections to the intraband dynamics , it does not affect interband dynamics significantly near its resonant contribution at @xmath107 , where the optical momentum is negligible [ i.e. , @xmath108 . we consequently neglect @xmath109 , so that eqs .
( [ gbeqs ] ) reduce to @xmath110 + \frac{4e}{p } { \rm re } \left\ { e_0 { \rm e}^{-i\omega t } \right\ } \sin \phi \;{\rm im}\{\gamma_{\bf p}\ } , \label{beq2}\end{aligned}\ ] ] [ beqs ] where @xmath111 , @xmath112 , and we have introduced a phenomenological relaxation time @xmath61 that encompasses the effect of numerous ultrafast decay channels for out - of - equilibrium electrons into hot carriers and phonons @xcite .
we assume for simplicity a single effective relaxation time @xmath61 , even though polarization dephasing and electron - hole recombination can take place over different time scales and their actual temporal dependences remain uncertain .
additionally , we use the same symbol @xmath61 for this quantity as in the intraband contribution ( sec .
[ intrabandsat ] ) , although we remark that the relative importance of different channels can differ substantially in both cases .
interestingly , eqs .
( [ beqs ] ) coincide with the traditional bloch equations for graphene and other two - band systems @xcite , which are routinely used to describe saturable absorption and other two - band effects , such as for example self - induced transparency @xcite . to describe interband saturable absorption ,
we adopt the steady - state ansatz @xmath113 using these expressions and neglecting third - harmonic terms , eqs .
( [ beqs ] ) lead to @xmath114 [ gmms ] where @xmath115 and @xmath116 . while an analytical solution to eqs .
( [ gmms ] ) is readily obtained , we omit the resulting expressions , which are rather involved and do not provide much physical insight . the interband current density @xmath117 is then obtained from this solution and eq .
( [ curdenseq ] ) through the expression @xmath118 d^2{\bf p } \right\ } { \bf \hat{x}}. \nonumber \end{aligned}\ ] ] in 2d @xmath5 space at @xmath119k for a chemical potential @xmath120ev .
the electron momentum components @xmath32 and @xmath121 are normalized to @xmath122 . *
( b ) * cut along the @xmath121 axis of ( a ) for several chemical potentials @xmath123ev , with @xmath121 scaled to @xmath124 .
the light intensity and wavelength are @xmath125gw@xmath69@xmath70 and @xmath126 nm in both plots .
we assume a phenomenological relaxation time @xmath62fs.,scaledwidth=45.0% ] we finally solve the 2d integral in @xmath5 space numerically . because we neglect third - harmonic terms
, @xmath127 oscillates in time around zero with the same angular frequency @xmath2 as the external field , while @xmath128 oscillates with angular frequency @xmath129 around the steady - state out - of - equilibrium population difference @xmath130 .
consequently , our approach enables the explicit calculation of the out - of - equilibrium occupation distribution of optically induced free - carriers @xmath131/2 $ ] [ see figs .
[ fig4](a , b ) , where we plot @xmath132 for several chemical potentials @xmath55 at fixed incident light intensity @xmath133 gw@xmath69@xmath70 and wavelength @xmath67 nm ] . as a function of incident light intensity @xmath22 for a wavelength of @xmath134 nm for * ( a ) * several fermi energies @xmath135 at zero temperature and * ( b ) * several electron temperatures @xmath57 and fixed chemical potential @xmath136 ev . the inelastic relaxation time is set to @xmath62fs in all plots.,scaledwidth=45.0% ] the @xmath5-space steady - state carrier occupation @xmath132
is characterized by two lobes along the @xmath49 axis , peaked at @xmath137 and surrounding the thermal fermi - dirac distribution of electrically doped electrons [ see fig .
[ fig4](a ) ] . when @xmath138 , the distribution @xmath132 stays thermalized , as interband absorption is inhibited by the pauli exclusion principle . in particular , after irradiation by an ultrashort optical pulse , @xmath132 quickly evolves within a time of @xmath139fs to a different thermal distribution corresponding to an increased electron temperature @xcite .
eventually , it relaxes to the original unperturbed distribution @xmath140 within a time scale @xmath141ps due to electron - phonon scattering @xcite . of undoped graphene on incident light intensity @xmath22 and inelastic relaxation time @xmath61 at zero temperature .
the white dashed curve represents the saturation intensity @xmath142 for which @xmath143.,scaledwidth=47.0% ] as a consequence of the optically induced out - of - equilibrium occupation @xmath132 in the upper band , interband absorption also saturates . in simple terms , at high intensities the electron - hole recombinations produced by inelastic collisions balance the light - driven interband transitions , thus leading to absorption quenching .
similar to sec .
[ intrabandsat ] , we attempt to quantitatively describe this effect by defining an interband absorption coefficient as the ratio of the time - averaged absorbed power over an optical cycle to the impinging intensity , @xmath144 we calculate the intensity - dependent interband absorption coefficient @xmath145 for several fermi energies [ fig .
[ fig5](a ) ] and electron temperatures [ fig .
[ fig5](b ) ] by numerically solving eqs . ( [ interbandcurrenteq ] ) and ( [ interabscoeffeq ] ) . for low intensities and zero temperature , we reproduce the universal absorption law of undoped graphene , @xmath146 @xcite , where @xmath147 is the fine - structure constant .
this result reflects a dispersionless linear conductivity @xmath148 .
in contrast , the interband absorption vanishes as @xmath149 in the limit of large incident intensities @xmath22 .
this high - intensity behavior is similar to intraband absorption ( see sec .
[ intrabandsat ] ) .
we also observe a strong dependence of @xmath145 on the fermi energy , as well as a large thermal modulation at high electron temperatures @xmath150k .
it should be noted that the modulation depth is much higher for interband than for intraband absorption , and additionally , it can be efficiently controlled by changing the fermi energy .
( left scale ) as a function of optical wavelength @xmath151 ( horizontal axis ) and relaxation time @xmath61 ( color scale ) at zero temperature .
available experimental results for the saturation intensity are indicated by symbols @xcite ( see legend).,scaledwidth=47.0% ] the light intensity required to achieve saturated absorption can be affected by the actual value of the inelastic collision time . for large @xmath61 ,
optically pumped electrons are expected to produce pauli blocking during a longer time , therefore reducing the interband saturation intensity @xmath152 . a plot of @xmath153 as a function of @xmath22 and @xmath61 [ fig .
[ fig6](a ) , calculated for @xmath67 nm ] confirms this intuition and further reveals that @xmath152 varies over a wide range ( @xmath154mw@xmath69@xmath70 ) when @xmath61 evolves within a range compatible with reported measurements of the electron inelastic lifetime in graphene samples of different qualities . but more importantly , the actual values of @xmath152 are much smaller than the characteristic intraband saturation intensities @xmath103 derived in sec .
[ intrabandsat ] .
the interband saturation intensity @xmath152 heavily depends on the optical wavelength @xmath151 [ see fig .
[ fig7 ] , where we plot @xmath155 for several relaxation times @xmath61 ] , varying by several orders of magnitude within the optical and near - infrared spectrum . , solid curves and symbols ) and the analytical mdf model for the interband absorption coefficient ( @xmath153 , dashed curves ) as a function of incident light intensity @xmath22 for a wavelength of @xmath134 nm and several fermi energies @xmath135 at zero temperature.,scaledwidth=45.0% ]
although we have provided a comprehensive theory of saturable absorption in extended graphene , practical devices operating on this principle must be finite in size .
this modifies the electronic structure of the material and thus its optical response .
while this effect is only expected to play a significant role in structures with dimensions less than a few tens of nanometers @xcite , we contrast next the above theoretical description for extended graphene with results obtained from an atomistic approach for finite structures .
specifically , following the methods of refs.@xcite , we simulate the intensity - dependent optical response of one - dimensional graphene nanoribbons by numerically solving the single - particle density matrix equation of motion in the time domain , using a tight - binding hamiltonian for the @xmath157-band electronic structure along with a self - consistent electron - electron ( hartree ) interaction potential ( see appendix [ a2 ] for further details ) .
remarkably , for a nanoribbon of only @xmath158 nm width , we find the atomistic simulations of the intensity - dependent absorbed power under cw illumination to be in excellent quantitative agreement with that for extended graphene based on the mdf picture ( see fig .
[ fig8 ] , where we compare results from the two methods for several fermi energies at an excitation wavelength of @xmath134 nm ) . despite this agreement on the saturation intensity ,
the unsaturated absorption coefficients in the weak intensity limit differ slightly , presumably due to finite - size effects that are captured in the atomistic approach but inherently absent in the mdf picture .
interestingly , for a fermi energy of @xmath159ev , the atomistic approach predicts an intensity - dependent increase in absorption rather than saturation .
this discrepancy could originate in two - photon absorption processes : although one - photon absorption ( for which the analytical derivation of the previous section is valid ) is quenched at @xmath160 m and @xmath161ev , two - photon absorption leads to an increase in absorption at very high intensities .
a detailed description of two - photon absorption goes however beyond the scope of the present discussion .
figure [ fig9 ] depicts the results of atomistic simulations for the dependence of the absorption coefficient @xmath162 on the ribbon width @xmath163 and the impinging light intensity @xmath22 , which are shown to converge for widths of a few tens of nanometers .
this corroborates that finite - size effects can be indeed neglected for larger graphene structures ( i.e. , flakes of 10 s nm in size ) . at two different optical wavelengths : @xmath164 nm ( red curves and symbols ) and @xmath126 nm ( blue curves and symbols ) .
experimental results are taken from refs .
@xcite ( rombic markers ) and @xcite ( circle markers).,scaledwidth=45.0% ]
we are now ready to compare our theoretical results with experimental data available in the literature .
most experimental studies exploit graphene as a saturable absorber for pml applications , and thus focus on undoped or poorly doped samples , for which only interband absorption becomes relevant .
surprisingly , we find enormous variations in reported measurements , even for experiments conducted with the same light wavelength , so we attribute this dispersion in observed results to the different qualities of the graphene samples and experimental conditions .
this intuition is supported by the calculations presented in fig .
[ fig7 ] ( see above ) , in which the saturation intensity for sa is shown to be highly dependent on the intrinsic relaxation time @xmath61 , which is directly affected by the graphene sample quality .
in contrast , the saturation intensity is less sensitive to doping ( for @xmath165 ) , although these parameters have a considerable effect on the modulation depth . in fig .
[ fig7 ] we compare the calculated saturation intensity ( color plot ) to experimental measurements ( symbols ) for several optical wavelengths and relaxation times , while in fig .
[ fig10 ] we contrast the predicted intensity - dependent absorption coefficient ( curves ) with experimental data ( symbols ) .
we find good agreement with several experimental results @xcite by assuming reasonable values for the relaxation times , ranging from @xmath166fs to @xmath167fs , while other experiments @xcite can only be reproduced either with extremely low relaxation times @xmath168fs , presumably corresponding to low - quality samples , or with extremely high relaxation times @xmath169ps .
we also note that a non - vanishing background absorption , which does not saturate within the considered range of illumination intensities , is usually present in the aforementioned experimental results [ see fig .
[ fig10 ] ] .
this effect appears to be dependent on the number of graphene layers @xcite and is found to vanish for pristine single - layer extended graphene @xcite .
in summary , we have developed a non - perturbative description of both intraband and interband contributions to saturable absorption in extended graphene by modeling conduction electrons as 2d massless dirac fermions coupled to an external optical field .
the dirac equation describing the time evolution of these electrons is recast in the form of generalized bloch equations , which we solve analytically .
remarkably , the interband contribution presents saturation at unusually low light intensities when compared with other materials , or with the intraband saturation .
additionally , we find a significant dependence of these effects on fermi energy , providing an active mechanism of control over saturable absorption in the carbon layer via electrical doping modulation .
in contrast , the electron temperature @xmath57 does not play a relevant role , unless @xmath170 exceeds the fermi energy or the incident photon energy .
nonetheless , @xmath57 can reach that regime under intense optical pumping , and therefore , this is an effect that must be considered when illuminating with long , intense light pulses .
we attribute the strong saturable absorption in graphene to its peculiar electronic band structure , and in particular , to the combination of its linear dispersion relation and the vanishing of the density of states at the dirac point . consequently
, both doping and optical transitions in the presence of strong optical fields produce large modifications in the populations of electronic states , thus resulting in substantial variations in momentum and producing radical changes in the conductivity .
we therefore expect a similarly low threshold for saturable absorption in other nanoscale materials that present dirac cones or other exotic electronic structures characterized by a low density of electronic states at the fermi level .
band - structure engineering is then a direction to explore in nanographenes , fullerenes , and carbon nanotubes , as well as in other van der waals atomic - layer materials and its derivatives .
interband absorption dominates under the conditions explored in currently available experiments , which are in good agreement with our analytical theory .
in contrast , intraband saturation absorption is not widely studied in the literature , and we predict it to emerge only at very high light intensities @xmath171gw@xmath69@xmath70 , depending on the doping level . however
, the intraband saturation intensity scales as the inverse square of the optical wavelength , and thus should become important at terahertz frequencies , particularly if it is enhanced by localized graphene plasmons in nano - islands / ribbons @xcite . indeed ,
while we find good agreement with an atomistic description of saturable absorption , including finite - size effects under nonresonant conditions , we anticipate that in general the enhanced near - fields associated with localized plasmons in graphene nanostructures should contribute significantly to reduce the saturation intensity and enlarge the modulation depth of sa .
a detailed study of plasmon - enhanced sa is still needed to explore how far down the saturation intensity can be pushed .
overall , the extraordinarily low intensity threshold for saturable absorption in graphene , combined with its electrical tunability , offers great potential for photonic applications such as mode - locking in graphene - clad fibre lasers and graphene - based random lasers .
this work has been partially supported by the spanish mineco ( mat2014 - 59096-p and sev2015 - 0522 ) and the european commission ( graphene flagship cnect - ict-604391 and fp7-ict-2013 - 613024-grasp ) .
here , we provide further details on the derivation of the intraband current [ eq . ( [ intracurrentt0 ] ) ] . at zero temperature the chemical potential coincides with the fermi energy ( @xmath79 ) and @xmath172 ( we assume doping with holes ) . inserting this and the expression given in the main text for @xmath173 into eq .
( [ intraintegraleq ] ) , we find @xmath174 ^ 2+p_y^2}}. \nonumber\end{aligned}\ ] ] now , we perform the @xmath32 integral analytically and define @xmath84 ^ 2\}$ ] to obtain eq .
( [ intracurrentt0 ] ) . finally , the @xmath18 integral in eq .
( [ intracurrentt0 ] ) gives rise to the analytical expression @xmath175 ^ 2 } \;{\cal i}[f(t ) ] , \end{aligned}\ ] ] where @xmath176 \right .
\nonumber \\ & \left .
+ \frac{f^2 - 1}{\sqrt{1-f } } \left [ { \cal f } \left ( \frac{\pi}{4}\left|\frac{2f}{f-1}\right . \right ) - { \cal f }
\left ( 0\left|\frac{2f}{f-1}\right . \right ) \right ] \right .
\nonumber \\ & \left .
+ \sqrt{1+f } \left [ { \cal e } \left ( \frac{\pi}{4}\left|\frac{2f}{f+1}\right .
\right ) - { \cal e } \left ( 0\left|\frac{2f}{f+1}\right .
\right ) \right ] \right .
\nonumber \\ & \left .
+ \sqrt{1-f } \left [ { \cal e } \left ( \frac{\pi}{4}\left|\frac{2f}{f-1}\right .
\right ) - { \cal e } \left ( 0\left|\frac{2f}{f-1}\right .
\right ) \right]\right\ } \end{aligned}\ ] ] and @xmath177 are the generalized jacobi elliptic functions of first and second kind @xcite .
we consider a graphene ribbon containing @xmath178 unit cells with period @xmath179 along its direction of translational symmetry . following a previously reported procedure @xcite
, we construct the corresponding one - electron wave functions from a tight - binding hamiltonian @xmath180 ( assuming a nearest - neighbor hopping energy of 2.8ev ) as @xmath181 , where @xmath182 denotes the band index , @xmath183 is the in - plane bloch wave vector along the ribbon , @xmath184 is the 2p carbon orbital at site @xmath185 in unit cell @xmath186 , and @xmath187 are complex expansion coefficients .
the optical response is simulated via direct numerical integration of the single - electron density matrix equation of motion , @xmath188-\frac{1}{2\tau}\left(\rho-\rho^0\right),\ ] ] where @xmath189 is the self - consistent electric potential , with @xmath190 denoting the coulomb interaction between atom @xmath191 in unit cell @xmath186 and atom @xmath192 in unit cell @xmath193 .
relaxation in eq .
( [ rhoeom ] ) brings us back to the equilibrium density matrix @xmath194 at a rate @xmath195 . in the state representation ,
the relaxed matrix elements @xmath196 are constructed by populating electron states according to the fermi - dirac distribution ( see occupation numbers @xmath197 ) .
here we consider light impinging along the graphene plane normal , with polarization @xmath198 directed across the ribbon [ i.e. , the incident light field is @xmath199 . upon integration of eq .
( [ rhoeom ] ) , the time - dependent elements @xmath200 yield the induced dipole moment per unit length of the nanoribbon @xmath201 the absorption coefficient is then calculated from the optical cross - section normalized to the ribbon area , @xmath202 , where @xmath163 is the nanoribbon width and @xmath203 is the fourier transform of @xmath204 over a single optical cycle of the incident cw field . | saturable absorption is a non - perturbative nonlinear optical phenomenon that plays a pivotal role in the generation of ultrafast light pulses .
here we show that this effect emerges in graphene at unprecedentedly low light intensities , thus opening avenues to new nonlinear physics and applications in optical technology .
specifically , we theoretically investigate saturable absorption in extended graphene by developing a non - perturbative single - particle approach , describing conduction - electron dynamics in the atomically - thin material using the two - dimensional dirac equation for massless dirac fermions , which is recast in the form of generalized bloch equations . by solving the electron dynamics non - perturbatively
, we account for both interband and intraband contributions to the intensity - dependent saturated conductivity and conclude that the former dominates regardless of the intrinsic doping state of the material .
the results are in excellent agreement with atomistic quantum - mechanical simulations including higher - band effects .
additionally , we find that the modulation depth of saturable absorption in graphene can be electrically manipulated through an externally applied gate voltage .
our results are relevant for the development of graphene - based optoelectronic devices , as well as for applications in mode - locking and random lasers . |
the interaction of ultraintense lasers with jets of molecular clusters or nanodroplets ( with typical size in the range @xmath0 nm and containing @xmath1 particles ) is a central research topic @xcite , with important applications , such as tabletop nuclear fusion for compact neutron sources @xcite , or the laboratory investigations of nucleosynthesis reactions , relevant to astrophysical scenarios @xcite .
clustered media can be regarded as sparse distributions of tiny solid targets , a peculiar configuration that allows for both a deep penetration of the laser radiation and a strong laser - matter coupling with many individual , overdense targets , thus providing extremely efficient energy absorption @xcite .
when hit by an ultraintense laser beam , the neutral atoms in a cluster are promptly ionized ( cf . ref .
@xcite for a detailed analysis of the concurring ionization mechanisms in different laser / cluster configurations ) and a dense `` nanoplasma '' @xcite is formed .
the free electrons then absorb energy from the laser pulse @xcite and start expanding , causing the formation of strong electric fields , which lead to efficient ion acceleration , as first predicted by dawson @xcite . when the energy transferred to the electrons is much smaller than the electrostatic energy stored in the ion core , charge separation is localized to regions much smaller than the cluster @xcite , which then remains quasi - neutral and undergoes a hydrodynamic - like expansion @xcite ; in opposite conditions ( e.g. with small deuterium clusters exposed to extremely intense laser radiation ) the electrons suddenly escape from the cluster and the remaining bare - ion distribution undergoes a pure coulomb explosion ( ce ) @xcite . in intermediate situations ,
the expansion dynamics is a mixture of the phenomenology of the two limits , with the expansion process being strongly dependent on the self - consistent dynamics of ions and trapped electrons @xcite . when increasing the laser energy , or when lowering the cluster size and density , the expansion conditions vary smoothly from quasi - neutral , hydrodynamic - like regimes to pure ce regimes , as confirmed by particle - in - cell ( pic ) simulations @xcite of the self - consistent laser - cluster interaction and by kinetic or fluid modeling of the expansion of finite - size , non - quasi - neutral plasma expansions @xcite .
therefore , the expansion regime can be controlled by regulating the amount of energy transferred to the electrons @xcite , which can be obtained with appropriately shaped laser beams .
an important example is the irradiation of homonuclear deuterium clusters with two sequential laser pulses having different intensities @xcite . in this way , one can taylor the expansion dynamics so as to induce overrunning between ions and the consequent formation of expanding shells containing multiple ion flows . within such structures (
here denoted as `` shock shells '' , following the terminology in @xcite ) , the relative velocities between deuterium ions from different flows can be sufficiently high for energetic collisions and _ dd _ fusion reactions to occurr @xcite .
since these intracluster reactions occur early in the expansion , and before different exploding clusters overlap , they are expected to produce a time resolved burst of fusion neutrons before the bulk neutron signal due to intercluster reactions @xcite . in the case of heteronuclear clusters , e.g. deuterium - tritium clusters ,
ion species having different charge - to - mass ratios expand with different velocities making intracluster , interspecies reactions possible also with standard single - pulse irradiation @xcite .
this paper provides an organic review of the work on controlled expansions of clusters and nanoplasmas published in @xcite , complemented with novel simulation results obtained with a recently developed molecular - dynamics technique @xcite . in the following
, the transition to the ce regime is analyzed with a self - consistent kinetic model ( section [ sec : expansion ] ) , the concept of shock shell is briefly reviewed ( section [ sec : shocks ] ) , and the possibility of achieving intracluster nuclear reactions in homonuclear clusters @xcite by driving a slow expansion followed by a sudden ce @xcite is explored ( section [ sec : reactions ] ) .
in order to analyze the influence of the electrons on the expansion dynamics , a kinetic model for the collisionless expansion of a spherically symmetric nanoplasma has been developed @xcite , based on the assumptions that the electrons are nonrelativistic and resorting to the large mass disparity between electrons and ions .
the expansion process is divided in two stages : an initial charging transient with frozen ions , and the long term expansion of both ions and electrons . in the second stage ,
the expansion dynamics is described self - consistently by following the radial motion of the cold ions , while accounting for the three - dimensional dynamics of the hot electrons using a sequence of `` ergodic '' equilibrium configurations @xcite , represented by stationary solutions of the vlasov equation that depend on the total energy @xmath2 only .
the electron density instantaneously in equilibrium with the electrostatic potential @xmath3 can be written as @xmath4 where @xmath5 indicates functional dependence on @xmath3 , @xmath6 is the energy distribution of the electrons , and @xmath7^{\frac{1}{2}}{\mathrm{d}r}^{\prime}$ ] is the probability density of finding an electron having total energy @xmath8 at the radial position @xmath9 .
the self - consistent potential @xmath3 satisfies the nonlinear poisson equation @xmath10 $ ] . by resorting to the theory of adiabatic invariants for time varying hamiltonians @xcite , a closed set of equations
is then obtained in the form @xmath11 where @xmath12 and @xmath13 are the ion mass and charge , respectively .
equations and , determining the ion trajectory @xmath14 ( @xmath15 : initial radius ) and the evolution of the electron energy @xmath16 ( @xmath17 : initial energy ) , are coupled with the nonlinear poisson equation . in eq . , the relation @xmath18 has been used , where @xmath19 is the energy distribution of the electrons resulting from the initial charging transient , a fast process that can not be described by equations
the self - consistent shape of @xmath19 is determined by a virtual charging transient , in which the ions stay immobile , while an external potential barrier , initially confining the electrons , is gradually moved from @xmath20 to infinity , with a series of small radial displacements ( the validity of this procedure has been confirmed by particle - in - cell simulations @xcite and ad - hoc solutions of the vp model @xcite ) .
the initial , nonequilibrium electron distribution is assumed to be a maxwellian , with temperature @xmath21 , so that the whole expansion dynamics is fully determined by the single dimensionless parameter @xmath22 ( @xmath23 : maximum ion energy attainable from the ce of a uniform ion sphere with radius @xmath20 and total charge @xmath24 ; @xmath25 : initial electron debye length ) , which accounts for both the initial electron temperature and the cluster size and density . the positive charge buildup at the ion front , @xmath26 , and the mean kinetic energy of the trapped electrons , @xmath27 , as functions of @xmath28 ,
are displayed in fig .
[ fig : qeq_teq ] .
simple fits for @xmath26 and @xmath27 are found as @xmath29 and @xmath30 $ ] , where @xmath31 . for @xmath32 ( i.e. @xmath33 ) ,
the fits reduce to @xmath34 , thus recovering the theoretical results for planar expansions @xcite , and to @xmath35 . when the initial equilibrium is reached and the ions are allowed to expand , gaining kinetic energy , the electrons cool down and the charge buildup decreases ( until a ballistic regime is reached for both species @xcite ) . depending on the value of @xmath28 ,
such behavior strongly affects the ion dynamics and the resulting distribution of ion energy @xcite : for @xmath36 , the spectrum is monotonic as in a ce ; for @xmath37 , it exhibits a local maximum far from the cutoff energy , thus being qualitatively different from the ce case . in this respect ,
the transition value @xmath38 can be considered as the lower bound for the validity of the ce approximation .
the cutoff energy is fit by @xmath39 , which , for @xmath40 , reduces to @xmath41 ; for @xmath37 , the location of the maximum is fit by the power - law @xmath42 .
these scaling laws , valid for any combination of @xmath20 , @xmath43 , and @xmath21 , can be useful to interpret experimental data @xcite .
in the absence of electrons , the explosion dynamics of a spherical distribution of cold ions is described by equation , which reduces to @xmath44 , where @xmath45 is the number of ions enclosed by a sphere of radius @xmath9 at time @xmath46 .
if the initial profile of ion density is uniform and step - like ( i.e. @xmath47 for @xmath48 ) , the repulsive electric field grows linearly for @xmath49 , reaching its maximum at the outer boundary , which causes the outer ions to be always faster than the inner ones .
thus , ions never overtake each other , @xmath45 is conserved along the ion trajectories ( so is the total energy of each ion ) , and the equation of motion can be integrated analytically @xcite , yielding , for ions initially at rest , the expansion velocity @xmath50^{1/2}r_0\omega{_{\mathrm{pi}}}$ ] , where @xmath51 is the initial ion plasma frequency . the radial trajectory is then given , in implicit form , by @xmath52^{1/2 } + \log[\xi{_{\mathrm{i}}}^{1/2}+(\xi{_{\mathrm{i}}}-1)^{1/2 } ] = ( 2/3)^{1/2}\omega{_{\mathrm{pi}}}t$ ] , where @xmath53 is the expansion factor . in this solution
, @xmath54 is independent of @xmath15 , the @xmath55 phase - space profile is always a straight line with equation @xmath56^{1/2}\omega{_{\mathrm{pi}}}r$ ] , and the ion density maintains its step - like form , decreasing in time as @xmath57 . the asymptotic ion energy distribution is @xmath58 ( cf .
[ fig : ionspectrum_t0 ] ) . if , more realistically , nonuniform density profiles are considered , the expansion features change qualitatively @xcite : when the initial density is a decreasing function of @xmath9 , the repulsive coulomb field reaches its maximum within the ion sphere @xcite , leading to ion overtaking and to the formation of multiple - flow regions ( shock shells ) with characteristic multi - branched phase - space profile as in fig .
[ fig : dp ] ( since the ion trajectories are no longer independent of one another , the above analytical solution no longer holds @xcite ) .
the physical interest of these shock shells resides in the appearance of multiple flows with large relative velocities within a single exploding cluster , which can lead to energetic ion - ion collisions and to intracluster reactions in homonuclear clusters @xcite .
an effective strategy to produce large - scale shock shells in a controlled fashion is combining different expansion regimes so that a slow hydrodynamic - like expansion , providing a smoothly decreasing ion - density profile , is followed by an abrupt ce .
this is achievable with a double - pump irradiation scheme as in fig .
[ fig : dp ] , wherein a weak pulse ( intensity : @xmath59 ; duration : @xmath60 ) is followed by a strong pulse ( intensity : @xmath61 ; duration : @xmath62 ) , with suitable time delay @xmath63 : @xmath59 must be sufficiently high to ionize the atoms , creating a nanoplasma , but not so high as to expel a significant fraction of the electrons from the cluster , whereas @xmath64 must be high enough to drive a sudden ce . in these conditions ,
a pronounced shock shell is formed , whose features are determined by the key double - pump parameters , namely @xmath59 and @xmath63 , which strongly affects the density profile of the cluster when being hit by the second laser . the effectiveness of the technique has been demonstrated @xcite by resorting to two- and three - dimensional pic simulations performed using the osiris 2.0 framework @xcite , closely matching realistic physical scenarios .
the ability of producing multiple flows with high relative velocities within single exploding clusters makes the phenomenon attractive as a possible way to induce intracluster fusion reactions in homonuclear deuterium clusters . as pointed out in @xcite
, the rates for intracluster reactions can be significantly higher than those for intercluster reactions , because the typical densities within a shock shell ( @xmath65 @xmath66 ) are higher than those within the hot plasma filament resulting from the exploded clusters ( @xmath67 @xmath66 ) .
however , it is crucial to consider also that the expanding shock - shell stays appreciably dense only for a very brief time ( @xmath68 fs ) , much shorter than the typical disassembly time of the plasma filament ( @xmath69 ps ) : as shown in @xcite , the rates of intracluster reactions exhibit a sharp , time - resolved peak right after the shock - shell formation . once the expansion dynamics is known , the number of intracluster fusion reactions is obtained by summing over all possible contributions from collisions between ions belonging to different velocity branches in a shock shell ( cf . fig .
[ fig : dp ] ) .
the number of reactions per unit time and volume , @xmath70 , is given by @xmath71 where @xmath72 is the cross section for _ dd _ fusion , while @xmath73 and @xmath74 indicate , respectively , the ion density and velocity on the @xmath75th branch .
the number of intracluster fusion reactions , @xmath76 , is obtained by integrating @xmath70 over time and space , as @xmath77 where @xmath78 , @xmath79 , and @xmath80 are the shock - shell boundaries and formation time , respectively . in @xcite ,
the influence of the double - pump parameters on @xmath76 was analyzed using a simple 1d model wherein the laser field of the first pulse gradually strips the initially neutral cluster of a part of its electrons according to a cluster barrier suppression ionization ( cbsi ) model ( cf . ref .
this provided preliminary estimates for the optimal combinations of delay and intensities that maximize @xmath76 , and suggested that , for very large clusters ( @xmath81 nm ) , the intracluster reaction yield can become comparable with the intercluster neutron yield , with @xmath82 of the fusion reactions arising from intracluster collisions @xcite . accurate calculations of the intracluster reaction yields achievable with the double - pump irradiation of homonuclear deuterium clusters are currently being performed resorting to three - dimensional molecular - dynamics simulations based on the recently developed scaled electron and ion dynamics ( seid ) technique @xcite , for different values of laser intensities , durations , and delays .
preliminary results indicate that , with @xmath59 in the range @xmath83 w@xmath84 , @xmath60 in the range @xmath85 fs , @xmath86 w@xmath84 , @xmath87 fs , and @xmath63 in the range @xmath88 fs , the number of intracluster reaction per cluster can be as high as @xmath89 with a cluster size @xmath90 nm . these values are comparable with the estimate in @xcite ; furthermore , as predicted in @xcite , it is found that the optimal intensities of the first pulse are significantly lower , typically by an order of magnitude , than those estimated with the cbsi model .
calculations based on the seid technique are being performed also in the case of heteronuclear deuterium - tritium clusters and nanodroplets irradiated by a single laser pulse , for either homogenous mixtures or layered targets composed by a deuterium core surrounded by a tritium shell .
owing to the higher value of the deuterium - tritium fusion cross section for the energy range considered , higher number of intracluster reactions , on the order of @xmath91 , are obtained with respect to the pure - deuterium case .
detailed results of these numerical simulations will be presented in a future publication @xcite .
the ergodic model has been used to simulate the collisionless expansion of spherical nanoplasmas driven by energetic electrons over a wide range of values of plasma size , plasma density , and initial electron thermal energy , combined in the single dimensionless parameter @xmath28 : a qualitative change in the shape of the asymptotic energy spectrum of the ions when approaching the coulomb explosion regime has been identified and accurate fit laws for the relevant expansion features have been provided .
the douple - pump irradiation technique to induce the formation of multiple ion flows during the expansion of large homonuclear deuterium clusters or nanodroplets has been described , and its applicability to produce intracluster _
dt _ fusion reactions has been proved resorting to three - dimensional numerical simulations performed with a recently developed scaled molecular dynamics technique .
work partially supported by fct ( portugal ) through grants pdct / poci/66823/2006 , sfrh / bd/39523/2007 , and sfrh / bpd/34887/2007 , and by the european community - new and emerging science and technology activity under the fp6 `` structuring the european research area '' programme ( project euroleap , contract number 028514 ) . part of the simulations discussed here were performed using the ist cluster ( ist / lisbon ) .
99 ditmire t , tisch jwg , springate e , mason mb , hay n , smith ra , marangos j and hutchinson mhr 1997 _ nature _ * 386 * 54 ditmire t , zweiback jwg , yanovsky vp , cowan te , hays g and wharton kb 1999 _ nature _ * 398 * 489 .
zweiback j , smith ra , cowan te , hays g , wharton kb , yanovsky vp and ditmire t 2000 _ phys .
lett . _ * 84 * 2634 .
zweiback j , cowan te , smith ra , hartley jh , howell r , steinke ca , hays g , wharton kb , crane jk and ditmire t 2000 _ phys .
lett . _ * 85 * 3640 .
madison kw , patel pk , price d , edens a , allen m , cowan te , zweiback j and ditmire t 2004 _ phys .
plasmas _ * 11 * 270 .
grillon g , balcou p , chambaret jp _
_ 2002 _ phys .
lett . _ * 89 * 065005 .
madison kw , patel pk , allen m , price d , fitzpatrick r and ditmire t 2004 _ phys . rev .
a _ * 70 * 053201 .
last i and jortner j 2006 _ phys .
lett . _ * 97 * 173401 .
heidenreich a , jortner j and last i 2006 _ proc .
_ * 103 * 10589 .
last i and jortner j 2008 _ phys .
a _ * 77 * 033201 .
ditmire t , smith ra , tisch jwg and hutchinson mhr 1997 _ phys .
lett . _ * 78 * 3121 .
last i and jortner j 2004 _ j. chem .
phys . _ * 120 * 1336 .
ditmire t , donnelly t , rubenchik am , falcone rw , and perry md 1996 _ phys .
a _ * 53 * 3379 .
mulser p , kanapathipillai m , and hoffmann dhh 2005 _ phys .
lett . _ * 95 * 103401 .
dawson jm 1964 _ phys .
fluids _ * 7 * 981 .
fukuda y , kishimoto y , masaki t and yamakawa k 2006 _ phys .
a _ * 73 * 031201(r ) .
mora p 2003 _ phys .
lett . _ * 90 * 185002 .
mora p 2005 _ phys
. plasmas _ * 12 * 112102 .
mora p 2005 _ phys .
e _ * 72 * 056401 .
murakami m , basko mm 2006 _ phys . plasmas _ * 13 * 012105 .
ditmire t , springate e , tisch jwg , shao yl , mason mb , hay n , marangos jp and hutchinson mhr 1998 _ phys .
a _ * 57 * 369 .
milchberg hm , mcnaught sj and parra e 2001 _ phys . rev .
e _ * 64 * 056402 .
_ 1998 _ phys .
lett . _ * 80 * 261 .
zweiback j , cowan te , hartley jh , howell r , wharton kb , crane jk , yanovsky vp , hays g , smith ra , and ditmire t 2002 _ phys . plasmas _ * 9 * 3108 . liu cs and tripathi vk 2003 _ phys . plasmas _ * 10 * 4085 .
kaplan ae , dubetsky by and shkolnikov pl 2003 _ phys .
* 91 * 143401 .
peano f , peinetti f , mulas r , coppa g and silva lo 2006 _ phys .
_ * 96 * 1750021 .
kishimoto y , masaki t and tajima t 2002 _ phys .
plasmas _ * 9 * 589 .
peano f 2005 _ laser - induced coulomb explosion of large deuterium clusters _ ( turin : politecnico di torino ) .
peano f , coppa g , peinetti f , mulas r and silva lo 2007 _ phys .
e _ * 75 * 066403 .
peano f , martins jl , fonseca ra , silva lo , coppa g , peinetti f and mulas 2007 _ phys .
plasmas _ * 14 * 056704 .
peano f , fonseca ra , and silva lo 2005 _ phys . rev
_ * 94 * 033401 .
peano f , fonseca ra , martins jl and silva lo 2006 _ phys . rev .
* 73 * 053202 .
li h , liu j , wang ch , ni g , kim chj , li r and xu zh 2007 _ j. phys .
b _ * 40 * 3941 .
last i and jortner j 2007 _ phys .
a _ * 75 * 042507 .
ott e 1979 _ phys .
lett . _ * 42 * 1628 .
t. grismayer , p. mora , j. c. adam , and a. hron 2008 _ phys .
e _ * 77 * 066407 .
je , auer pl and allen je 1975 _ j. plasma phys . _ * 14 * 89 .
manfredi g , mola s and feix mr 1993 _ phys . fluids _
b * 90 * 388 .
sakabe s , shimizu s , hashida m _
_ 2004 _ phys .
a _ * 69 * 023203 .
hirokane m , shimizu s , hashida m , okada s , okihara s , sato f , iida t and sakabe s 2004 _ phys .
a _ , * 69 * 063201 .
sakabe s , shirai k , shimizu s , hashida m and masuno s 2006 _ phys .
a _ * 74 * 043205 .
kovalev vf , popov ki , bychenkov vy and rozmus w 2007 _ phys
. plasmas _ * 14 * 053103 .
parks pb , cowan te , stephens rb and campbell em 2001 _ phys .
a _ * 63 * 063203 .
li h , liu j , wang c , ni g , li r and xu z 2006 _ phys .
a _ * 74 * 023201 .
fonseca ra _
notes comp .
sci . _ * 2331 * ( heilderberg : springer - verlag , 2002 ) 342 .
last i , jortner j , peano f , silva lo _ et al_. , _ in preparation_. | the expansion of laser - irradiated clusters or nanodroplets depends strongly on the amount of energy delivered to the electrons and can be controlled by using appropriately shaped laser pulses . in this paper ,
a self - consistent kinetic model is used to analyze the transition from quasineutral , hydrodinamic - like expansion regimes to the coulomb explosion ( ce ) regime when increasing the ratio between the thermal energy of the electrons and the electrostatic energy stored in the cluster .
it is shown that a suitable double - pump irradiation scheme can produce hybrid expansion regimes , wherein a slow hydrodynamic expansion is followed by a fast ce , leading to ion overtaking and producing multiple ion flows expanding with different velocities .
this can be exploited to obtain intracluster fusion reactions in both homonuclear deuterium clusters and heteronuclear deuterium - tritium clusters , as also proved by three - dimensional molecular - dynamics simulations . |
clusters of galaxies , the largest virialized systems known , provide one of the most powerful tools in studying the structure and evolution of the universe .
clusters highlight the large scale structure of the universe @xcite ; they trace the evolution of structure with time @xcite ; they constrain the amount and distribution of dark and baryonic matter @xcite ; they reveal important clues about the formation and evolution of galaxies @xcite ; and they place critical constraints on cosmology @xcite . in fact , clusters of galaxies place some of the most powerful constraints on cosmological parameters such as the mass density of the universe and the amplitude of mass fluctuations . in spite of their great value and their tremendous impact on understanding the universe , systematic studies of clusters of galaxies are currently limited by the lack of large area , accurate , complete , and objectively selected catalogs of optical clusters , and by the limited photometric and redshift information for those that do exist . the first comprehensive catalog of clusters of galaxies ever produced , the abell catalog of rich clusters @xcite , was a pioneering project that provided a seminal contribution to the study of extragalactic astronomy and to the field of clusters of galaxies . while galaxy clustering had been recognized before abell , the data were fragmentary and not well understood .
both abell s catalog , as well as zwicky s @xcite independent catalog , were obtained by visual inspection of the palomar observatory sky survey plates .
these catalogs have served the astronomical community for nearly half a century and were the basis for many of the important advances in cluster science ( see references above ; also abell s centennial paper , @xcite ) . at the beginning of the new century , the need for a new comprehensive catalog of optical clusters one that is automated , precise , and objectively selected , with redshifts that extend beyond the [email protected] limit of the abell catalog has become apparent .
there have been recent advances in this direction , including large area catalogs selected by objective algorithms from digitized photographic plates ( @xcite for the lick catalog ; @xcite for the edcc catalog ; @xcite and @xcite for the apm catalog ) , as well as small area , deep digital surveys of distant clusters ( e.g. , the 5 deg@xmath2 palomar distant cluster survey , @xcite ; 100 deg@xmath2 red - sequence cluster survey , @xcite ; and 16 deg@xmath2 kpno deeprange survey , @xcite ) .
a particularly important advance for optical surveys has been the inclusion of accurate ccd - based color information for galaxy selection .
the inclusion of color in cluster selection greatly reduces the problems of density projection which have long plagued optical selection of clusters .
good examples of color - based optical selection include the 100 deg@xmath2 red - sequence cluster survey @xcite and the sdss selection described in this work .
surveys of x - ray clusters and observations of the sunyaev - zeldovich effect in clusters have and will continue to provide important data that is complementary to the optical observations of clusters of galaxies .
these methods identify rich systems that have developed an extensive hot intracluster medium . while excellent for selection of massive , well developed clusters , these methods have thresholds which are sensitive to the evolution of the hot intracluster medium , both with cosmic time and with the richness of the objects . in this sense
, optical selection has the important complementary advantage of being able to identify galaxy clustering across a wide range of system richness and time evolution .
the sloan digital sky survey ( sdss ; @xcite ) will provide a comprehensive digital imaging survey of 10@xmath15 deg@xmath2 of the north galactic cap ( and a smaller , deeper area in the south ) in five bands ( @xmath16 , @xmath17 , @xmath18 , @xmath19 , @xmath20 ) , followed by a spectroscopic multi - fiber survey of the brightest one million galaxies ( [ clusterselection ] ) . with high photometric precession in 5 colors and a large area coverage ( comparable to the abell catalog ) , the sdss survey will enable state - of - the - art cluster selection using automated cluster selection methods .
nearby clusters ( to z @xmath14 0.05 - 0.1 ) can be selected directly in 3-dimensions using redshifts from the spectroscopic survey .
the imaging survey will enable cluster selection to [email protected] and beyond using the 5 color bands of the survey .
in the range [email protected] - 0.3 , the 2d cluster selection algorithms work well , with only small effects due to selection function ( for the richest clusters ) . in the nearest part of this range , [email protected] - 0.15 , the sdss spectroscopic data can also be useful for cluster confirmation and for redshift determination .
even poor clusters can be detected with high efficiency in this redshift range . for [email protected] - 0.5 ,
2d selection works well , but selection function effects become important , especially for poorer clusters .
several cluster selection algorithms have recently been applied to @xmath1400 deg@xmath2 of early sdss imaging commissioning data in a test of various 2d cluster selection techniques .
these methods , outlined in [ clusterselection ] , include the matched - filter method @xcite , and the red - sequence color - magnitude method , maxbcg @xcite , as well as a cut and enhance method @xcite and a multicolor technique ( c4 ; @xcite ) .
each method can identify clusters of galaxies in sdss data to [email protected] , with richness thresholds that range from small groups to rich clusters , and with different selection functions .
since each algorithm uses different selection criteria that emphasize different aspects of clusters , the lists of clusters found by different techniques will not be identical . in this paper we present a catalog of 799 clusters of galaxies in the redshift range z = 0.05 - 0.3 from 379 deg@xmath2 of sdss imaging data .
the catalog was constructed by merging lists of clusters found by two independent 2d cluster selection methods : hybrid matched filter and maxbcg . we compare the results from the two techniques and investigate the nature of clusters they select .
we derive scaling relations between cluster richness and observed cluster luminosity and cluster velocity dispersion .
we use the scaling relations to combine appropriate subsamples of these lists to produce a conservative merged catalog ; the catalog is limited to a richness threshold specified in [ catalog ] ; the threshold corresponds to clusters with a typical velocity dispersion of @xmath21 400 km s@xmath8 .
the average space density of the clusters is @xmath22 mpc@xmath13 .
a flat lcdm cosmology with @xmath23= 0.3 and a hubble constant of h@xmath24 = 100 @xmath25 km s@xmath8 mpc@xmath8 with @xmath25 = 1 is used throughout .
the current work represents preliminary tests of selection algorithms on early sdss commissioning data .
the results will improve as more extensive sdss data become available .
the sdss imaging survey is carried out in drift - scan mode in five filters , @xmath16 , @xmath17 , @xmath18 , @xmath19 , @xmath20 , to a limiting magnitude of @xmath2623 @xcite .
the spectroscopic survey will target nearly one million galaxies to approximately @xmath2617.7 , with a median redshift of [email protected] @xcite , and a small , deeper sample of @xmath110@xmath27 luminous red galaxies to @xmath2819 and [email protected] @xcite . for more details of the sdss survey see @xcite , @xcite , @xcite , @xcite and @xcite .
cluster selection was performed on 379 deg@xmath2 of sdss commissioning data , covering the area @xmath29(2000 ) = 355 to 56 , @xmath30(2000 ) = -1.25 to 1.25 ; and @xmath29(2000 ) = 145.3 to 236.0 , @xmath30(2000)= -1.25 to 1.25 ( runs 94/125 and 752/756 ) .
the limiting magnitude of galaxies used in the cluster selection algorithms was conservatively selected to be @xmath2621 ( where @xmath18 is the sdss petrosian magnitude ) . at this magnitude limit , star - galaxy separation is excellent @xcite .
the clusters of galaxies studied in this paper were selected from these imaging data using a matched filter method @xcite and an independent color - magnitude maximum - likelihood brightest cluster galaxy method ( maxbcg ; @xcite ) .
these methods are briefly described below .
the matched filter method hmf ( hybrid matched filter ; @xcite ) is a hybrid of the matched filter @xcite and the adaptive matched filter techniques @xcite .
this method identifies clusters in imaging data by finding peaks in a cluster likelihood map generated by convolving the galaxy survey with a filter based on a model of the cluster and field galaxy distributions .
the cluster filter is composed of a projected density profile model for the galaxy distribution ( a plummer law profile is used here ) , and a luminosity function filter ( schechter function ) .
the filters use the typical parameters observed for galaxy clusters ( e.g. , core radius r@xmath31 = 0.1 @xmath9 mpc , cutoff radius r@xmath32=1 @xmath33mpc , and luminosity function parameters m@xmath34 and @xmath35 for @xmath25 = 1 ) .
the hmf method identifies the highest likelihood clusters in the imaging data ( @xmath18-band ) and determines their estimated redshift ( z@xmath0 ) and richness ( @xmath3 ) ; the richness @xmath3 is derived from the best - fit cluster model that satisfies l@xmath36(@xmath37 mpc ) = @xmath3l@xmath38 , where l@xmath36 is the total cluster luminosity within 1 @xmath9 mpc radius ( at z@xmath0 ) , and l@xmath39 .
a relatively high threshold has been applied to the cluster selection ( @xmath405.2 , @xcite ) ; the selected clusters have richnesses @xmath3@xmath7 20 - 30 ( i.e. , l@xmath36(@xmath41 mpc ) @xmath42 ) .
this threshold is below the typical abell richness class 0 .
the maxbcg method ( @xcite ) is based on the fact that the brightest cluster galaxy ( bcg ) generally lies in a narrowly defined space in luminosity and color ( see , e.g , @xcite ) . for each sdss galaxy ,
a `` bcg likelihood '' is calculated based on the galaxy color ( @xmath43 and @xmath44 ) and magnitude ( m@xmath45 , in @xmath19-band ) .
the bcg likelihood is then weighted by the number of nearby galaxies located within the color - magnitude region of the appropriate e / s0 ridgeline ; this count includes all galaxies within 1 @xmath9 mpc projected separation that are fainter than m@xmath45 and brighter than the magnitude limit m@xmath45(lim ) = -20.25 , and are located within 2-@xmath46 of the mean observed color scatter around the e / s0 ridgeline ( i.e. , @xmath47 ) .
the combined likelihood is used for cluster identification .
the likelihood is calculated for every redshift from z = 0 to 0.5 , at 0.01 intervals ; the redshift that maximizes the cluster likelihood is adopted as the cluster redshift .
since bcg and elliptical galaxies in the red ridgeline possess very specific colors and luminosities , their observed magnitude and colors provide excellent photometric redshift estimates for the parent clusters .
the richness estimator , @xmath5 , is defined as the number of red e / s0 ridgeline galaxies ( within the 2-@xmath46 color scatter as discussed above ) that are brighter than m@xmath45(lim ) = -20.25 ( i.e. , 1 mag fainter than l@xmath48 in the @xmath19-band ; @xmath25 = 1 ) , and are located within a 1 @xmath9 mpc projected radius of the bcg . the hmf and maxbcg methods focus on different properties of galaxy clusters : hmf finds clusters with approximately plummer density profiles and a schechter luminosity function , while maxbcg selects groups and clusters dominated by red @xmath49 galaxies .
we compare the results of these cluster selection algorithms in the following sections and merge the clusters into a single complementary self - consistent catalog .
when comparing different catalogs , uncertainties in cluster estimated redshift , position , richness , and selection function , in addition to the different nature of each cluster selection algorithm , render the comparisons difficult . even selecting the richest clusters from each catalog will not provide a perfect match , mostly due to the noisy estimate of richness and its sharp threshold . in this section
we briefly summarize the main comparisons of the cluster redshift , position , and richness estimators for the hmf and maxbcg methods .
the accuracy of cluster redshift estimates for each method was determined using comparisons with measured redshifts from the sdss spectroscopic data .
a comparison of the estimated and spectroscopic redshifts for hmf and maxbcg clusters with z@xmath0 = 0.05 - 0.3 and richnesses @xmath3@xmath4 40 ( hmf ) and @xmath5@xmath4 13 ( maxbcg ) is shown in figures 1 and 2 .
a spectroscopic match is considered if the spectroscopic galaxy is located at the position of the bcg . for these relatively high richness clusters
we find a redshift uncertainty of @xmath50 = 0.014 for maxbcg ( from 382 cluster matches ) and @xmath50 = 0.033 for hmf ( from 237 cluster matches ; there are fewer hmf matches since a spectroscopic match is defined at the bcg position so as to minimize noise ) . a direct comparison between the hmf and maxbcg estimated cluster redshifts , using a positional matching criterion defined below , is shown in figure 3 . the positional accuracy of cluster centers is determined by comparing hmf - maxbcg cluster pairs ( in the above z = 0.05 - 0.3 sample ) with pairs in random catalogs .
the comparison shows significant excess of cluster matches over random for projected cluster separations of @xmath14 0.5 @xmath9 mpc , with a tail to @xmath1 1 @xmath9 mpc ( figure 4 ) .
these excess pairs represent real cluster matches ; their distribution provides a measure of the typical offset between the cluster centers determined in the two methods .
the offsets follow a gaussian distribution with a dispersion of 0.175 @xmath9 mpc ( figure 4 ) .
comparison of clusters identified by different selection methods depends not only on the positional and redshift uncertainties discussed above , and on the different selection function inherent to each catalog , but also on the uncertainties in the richness estimates .
the difference in selection functions and the uncertainties in richness estimates are the main cause of the relatively low matching rates among different samples ( see [ catalog ] ) .
the richness scatter is important because each cluster sample is cut at a specific richness threshold ; since the observed richness function is steep and the richness scatter is significant , a richness threshold causes many clusters to scatter across the threshold .
this scatter has a strong effect on cluster sample comparisons .
we illustrate the effect by monte carlo simulations of two identical cluster samples with different noisy richness estimators ( figure 5 ) . placing a richness threshold on each sample
, we obtain richness limited subsamples . for an intrinsic richness function of @xmath51(richness)@xmath52 ( see [ abundance ] ) , and a 30 @xmath53 scatter in richness , the overlap of the two samples is only 54 @xmath53 .
any difference in selection functions , which can be nearly a factor of @xmath12 in the two methods used here , will further reduce the apparent overlap .
this simple model provides an estimate for how large we might expect the overlap between two otherwise identical cluster samples to be .
it is important to bear this in mind as we make direct comparisons of cluster catalogs in subsequent sections . how do the hmf and maxbcg cluster richness estimates compare with each other ?
cluster richness estimates describe , in one form or another , how populated or luminous a cluster is : either by counting galaxy members within a given radius and luminosity range , or by estimating total cluster luminosity . in general , this measure also reflects the mass of the cluster , its velocity dispersion , and temperature .
while richness correlates well on average with other parameters ( e.g. , rich clusters are more luminous and more massive than poor clusters ) , individual cluster richness estimates exhibit large scatter .
this scatter is due to the sharp luminosity threshold in the richness galaxy count , uncertainties in the background corrections , uncertainties in the estimated redshift and center of the cluster , sub - structure in clusters , and other effects .
still , optical richness estimators provide a basic measure of a cluster population ; richnesses have been determined for all clusters in the above catalogs .
the two richness estimates obtained by the cluster selection algorithms described above are @xmath5 for maxbcg and @xmath3 for hmf ( [ clusterselection ] ) .
@xmath5 is the number of red ( e / s0 ) ridgeline galaxies located within 1 @xmath33mpc of the bcg galaxy and are brighter then m@xmath45(lim ) = -20.25 .
the richness @xmath3 is determined by the hmf fine likelihood for each cluster and reflects the best - fit cluster model luminosity within 1 @xmath33mpc radius , l@xmath36(@xmath41 mpc)= @xmath3 l@xmath48 ( [ clusterselection ] ; see @xcite and @xcite ) . in comparing these richnesses , differences in
the estimated redshifts and cluster centers introduce additional scatter on top of any intrinsic variations .
figure 6 presents the observed relation between @xmath3 and @xmath5 for clusters with z@xmath0 = 0.05 - 0.3 and @xmath5@xmath4 13 . while the scatter is large , as expected from the monte carlo simulations ( figure 5 ) , a clear correlation between the mean richnesses is observed . the best - fit relation between @xmath5 ( as determined for the maxbcg clusters with @xmath5@xmath413 ) and
the mean @xmath3 ( for the matching hmf clusters ) is : @xmath54 the error - bars reflect uncertainties on the mean best - fit .
( this relation differs somewhat if both richnesses are determined at the maxbcg - selected cluster positions and redshifts or at the hmf - selected clusters ; see , e.g. , @xcite ) .
the ratio @xmath3/@xmath5 decreases somewhat with @xmath5 ; we find @xmath3/@xmath5 @xmath55 2 for @xmath5@xmath720 , increasing to @xmath3/@xmath5 @xmath1 3 for lower richnesses .
a comparison of the richness estimates @xmath3 and @xmath5 with directly observed cluster luminosities and velocity dispersions is discussed in the following section .
we derive preliminary scaling relations between cluster richness estimates and directly observed mean cluster luminosity and cluster velocity dispersion .
this enables a direct physical comparison between the independent catalogs and allows proper merging of the two samples .
it also provides a physical calibration of the cluster richness estimates in terms of their mean luminosity , velocity dispersion , and hence mass .
the observed cluster luminosities can be directly obtained from the sdss imaging data using population subtraction . by comparing the galaxy population in regions around cluster centers to that in random locations we can determine the properties of galaxies in and around the clusters as well as the cluster luminosities .
since the redshifts of the sdss clusters are relatively accurate , we can determine cluster luminosities in physical units i.e. , in solar luminosities within a metric aperture .
the multi - color sdss data also allow us to apply accurate k - corrections to cluster galaxy magnitudes .
we determine the luminosity of a cluster by measuring the total luminosity of all galaxies within 0.6 @xmath33mpc of the cluster center .
we use all hmf and maxbcg clusters in the redshift range 0.05 @xmath56 z @xmath56 0.3 with richness @xmath3@xmath4 30 ( hmf ) and @xmath5@xmath4 10 ( maxbcg ) . for each cluster , we extract all galaxies within a projected radius of 0.6 @xmath33mpc of the cluster center , and compute a k - corrected absolute magnitude for each galaxy according to its type ( following @xcite ) .
we then sum the total luminosity ( @xmath18-band ) within the absolute magnitude range of -23.0@xmath57 - 19.8 .
we determine the background contribution to this total luminosity by selecting five random locations away from the cluster area ( within the same sdss stripe ) , each with the same angular extent ; we extract galaxies within these regions , k - correct them as if they were at the cluster redshift , and subtract the resulting mean luminosity ( within the same magnitude range ) from that of the cluster .
this process allows a determination of the variance in the background correction and yields an estimate of cluster luminosity within a radius of 0.6 @xmath33mpc and within the luminosity range -23.0@xmath57 - 19.8 ( corresponding to approximately 1.3 mag below hmf s @xmath58 ) .
we denote this luminosity @xmath59 .
a hubble constant of @xmath25 = 1 and a flat lcdm cosmology with @xmath23= 0.3 are used to determine cluster distances and luminosities .
details of this analysis , along with tests and a variety of related population subtraction results , will be presented in a forthcoming paper @xcite . for greater accuracy , and to minimize the spread due to redshift uncertainty , all clusters with a given richness are stacked and their mean luminosity @xmath59 determined .
these stacked luminosities are presented as a function of cluster richness in figures 7 and 8 for the hmf and maxbcg clusters . a strong correlation between richness and mean
luminosity is observed ; this is of course expected , since both @xmath5 and @xmath3 represent cluster richnesses which broadly relate to luminosity ( [ clustercomparison ] ) .
the best - fit power - law relations to the binned mean luminosities are : @xmath60 the few highest richness points ( @xmath3@xmath61 80 , @xmath5@xmath61 33 ) exhibit large scatter due to their small numbers .
inclusion of these points does not change the fits ; we find @xmath59 = 1.6 @xmath5 ( for maxbcg , @xmath5@xmath410 ) and @xmath59 = 0.015 @xmath3@xmath62 ( for hmf , @xmath3@xmath4 30 ) .
the non - linearity observed in the @xmath63-@xmath3 relation at high @xmath3 reflects the fact that the measured cluster luminosity @xmath63 corrects for an underestimate in @xmath3 at high richness seen in simulations @xcite ; the luminosity @xmath63 measures the true cluster luminosity , independent of any uncertainty in cluster richness estimates .
the luminosity @xmath59 is the cluster luminosity down to a magnitude of -19.8 . to convert this luminosity to a total cluster luminosity
, we integrate the cluster luminosity function from -19.8@xmath64 down to the faintest luminosities .
the luminosity function of hmf clusters ( within r = 0.6 @xmath9 mpc ) is observed to have schechter function parameters of @xmath65 and m@xmath66 , and maxbcg has @xmath67 and m@xmath68 ( @xmath25 = 1 ; @xcite ) . integrating these luminosity functions from -19.8 down to zero luminosity yields correction factors of 1.42 ( for hmf ) and 1.34 ( for maxbcg ) for the added contribution of faint galaxies to the total cluster luminosity .
the total mean cluster luminosities are therefore given by equations 2 and 3 multiplied by these correction factors , yielding @xmath69 the sdss spectroscopic survey includes spectra of galaxies brighter than @xmath18 = 17.7 @xcite , with a median redshift of z = 0.1 , as well as spectra of the ` luminous red galaxy ' ( lrg ) sample that reaches to r @xmath55 19 and z @xmath1 0.5 @xcite .
for some rich clusters at low redshift , it is possible within the sdss spectroscopic data to directly measure the cluster velocity dispersion .
here we compare these velocity dispersions , together with velocity dispersions available from the literature ( for some of the abell clusters within the current sample ; [ abell ] ) , to cluster richnesses ; this provides an independent physical calibration of richness .
the correlation between the observed cluster velocity dispersion and cluster richness is presented in figure 9 .
we use cluster velocity dispersions of 20 clusters determined from the sdss spectroscopic survey ( for clusters with @xmath130 to 160 redshifts ) using a gaussian fit method , as well as from several abell clusters available in the literature ( abell 168 , 295 , 957 , 1238 , 1367 , 2644 ; @xcite ) . even though the number of clusters with measured velocity dispersion is not large and the scatter is considerable , a clear correlation between median velocity dispersion and richness
is observed , as expected ( figure 9 ) .
the best - fit relations are : @xmath70 also shown in figure 9 , for comparison , are all stacked sdss spectroscopic data for the galaxy velocity differences in the clusters ( relative to the bcg velocity ) , subtracted for the mean observed background , as a function of richness .
these are obtained using the best gaussian fit to the stacked velocity data , after background subtraction .
the results are consistent with the directly observed @xmath46-@xmath3 and @xmath46-@xmath5 relations discussed above .
the velocity scaling relations ( equations 6 and 7 ) provide an important calibration of cluster richness versus mean cluster velocity dispersion ( and thus mass ) . also shown in the figures , for comparison ,
are the @xmath6-richness relations derived from the observed mean @xmath71-@xmath3 and @xmath71-@xmath5 correlations ( section [ lumscale ] , figures 7 - 8) . here
the luminosity @xmath72 is converted to mass , m@xmath73 , using the typical observed m / l ratio relevant for these clusters and the observed relation between @xmath74 and @xmath6 based on calibration using gravitational lensing observations ( see @xcite for details ) .
good agreement exists between these independent scaling relations .
larger samples , when available , will further improve this important calibration .
the independent scaling relations discussed above are consistent with each other .
the directly observed mean @xmath3-@xmath5 relation ( equation 1 ) is in agreement with the observed luminosity - richness relations , @xmath75-@xmath3 and @xmath75-@xmath5 ( equations 4 and 5 ) .
both relations the luminosity - richness relations and the @xmath3-@xmath5 relation yield , independently , @xmath3@xmath76 , and reproduce the observed total luminosity relations discussed above .
this consistency is illustrated by the solid and dashed lines in figure 6 which represent , respectively , the observed mean @xmath3-@xmath5 relation and the one obtained from the mean luminosty - richness relations ( @xmath77-@xmath3 and @xmath77-@xmath5 ) .
the third independent relation , velocity dispersion versus richness ( equations 6 and 7 ) , is also consistent with the above results ; this is illustrated by the dotted curve in figure 6 .
the non - linearity observed in the @xmath63(@xmath3 ) @xmath1 @xmath3@xmath2 relation ( equation 3 and discussion below it ; figure 7 ) , and the similar non - linearity observed in the @xmath3@xmath1 @xmath5@xmath78 relation ( i.e. , @xmath5@xmath1 @xmath3@xmath2 ; equation 1 ; figure 6 ) , are consistent with the velocity scaling relation , @xmath46 = 10.2 @xmath3 ( equation 6 ) , since the latter implies that cluster mass ( within a fixed radius ) is m @xmath79 ; this is consistent with the observed @xmath80 .
the maxbcg relations are also self - consistent , with a linear @xmath81 @xmath5 , @xmath82 @xmath5@xmath83 , and hence m @xmath84 . in both cases , @xmath85 is nearly constant in fact , slightly increasing with @xmath63 as expected ( e.g. , @xcite ) .
the consistency of the scaling relations is illustrated in figure 6 .
a summary of the mean quantitative scaling relations betweem @xmath3 , @xmath5 , velocity dispersion , luminosity , and mass ( within 0.6 @xmath9 mpc ) is presented in table 1 .
we use the scaling relations derived above ( [ scaling ] ) to define a conservative merged catalog of clusters of galaxies from the early sdss commissioning data based on the maxbcg and the hybrid matched - filter samples .
the merged bh catalog is limited to clusters within the redshift range z@xmath0 = 0.05 - 0.3 and richness above the threshold listed below , over the 379 deg@xmath86 area ( [ clusterselection ] ) .
a total of 799 clusters are listed in the catalog .
the clusters are selected using the following criteria : 1 .
z@xmath0 = 0.05 - 0.3 2 .
richness threshold of @xmath3@xmath440 ( for hmf clusters ) and @xmath5@xmath413 ( for maxbcg clusters ) .
these thresholds are comparable to each other and correspond to a mean cluster velocity dispersion of @xmath87 400 km s@xmath8 and luminosity l@xmath88 l@xmath89 ; the related mass is approximately m@xmath90 m@xmath11(see table 1 ) .
clusters that overlap between the two methods are considered as single clusters if they are separated by @xmath561 @xmath33mpc ( projected ) and @xmath560.08 in estimated redshift ( 2.5-@xmath91 ) .
overlap clusters are listed as a single cluster , on a single line , but include the relevant parameters from both the hmf and maxbcg selection ( position , redshift , richness ) .
this is done in order to provide complete information about the clusters and allow their proper use with the independent hmf and maxbcg selection functions .
for each cataloged cluster ( hmf with @xmath3@xmath4 40 or maxbcg with @xmath5@xmath4 13 ) we include cluster matches ( i.e. , overlaps with separations as defined above ) that reach beyond the richness or redshift thresholds of the catalog .
for example , an hmf cluster with @xmath3@xmath4 40 and z = 0.30 may list as a match a maxbcg cluster with @xmath5@xmath92 13 and/or z = 0.22 to 0.38 ( i.e. , @xmath93z @xmath56 0.08 ) . a lower limit of @xmath5@xmath46 is set for all matches . while not part of the @xmath3@xmath440 , @xmath5@xmath413 catalog , such matches with @xmath5@xmath9213 and @xmath3@xmath9240 clusters are listed in order to provide full information of possible matches , considering the large uncertainty in the richness parameter .
( if there is more than one match per cluster , we select the one with the closest separation ) . some of the matches , especially at low richness ( @xmath5@xmath1410 ) and large separation ( @xmath1 1 @xmath9 mpc or @xmath93z @xmath1 0.08 ) , may be coincidental .
clusters that do not overlap are listed as separate clusters and are so noted . the catalog is presented in table 2 .
listed in the catalog , in order of increasing right - ascension , are the following : sdss cluster number ( column 1 ) , method of detection ( h for hmf , b for maxbcg ; lower case ( h , b ) represents cluster matches that are outside the catalog richness or redshift thresholds , i.e. , @xmath3@xmath9240 , @xmath5@xmath9213 , [email protected] ; column 2 ) , hmf @xmath29 and @xmath30 ( in degrees 2000 ; column 3 - 4 ) , hmf estimated redshift ( column 5 ) , hmf cluster richness @xmath3 ( column 6 ) .
columns 7 - 10 provide similar information for the maxbcg detection , if the cluster so detected : @xmath29 and @xmath30 ( 2000 ; column 7 - 8) , maxbcg redshift estimate ( column 9 ) , and richness estimate @xmath5 ( column 10 ) .
an sdss spectroscopic redshift that matches the cluster , if available , is listed in column 11 ( mainly for the bcg galaxy ) .
column 12 lists matches with abell and x - ray clusters .
all the noras x - ray clusters and 53 of the 58 abell clusters in this area are identified in the catalog ; the additional five abell clusters are identified by the combined hmf and maxbcg techniques but are below the catalog richness threshold ( see [ abell ] ) .
the catalog contains 436 hmf clusters ( @xmath3@xmath4 40 ) , 524 maxbcg clusters ( @xmath5@xmath4 13 ) , and a total merged catalog ( as defined above ) of 799 clusters ( at @xmath94 = 0.05 - 0.3 ) .
some clusters are false - positive detections ( i.e. , not real clusters ) ; the false - positive rate is discussed below .
the overlap between the independent hmf and maxbcg clusters within the above redshift range is 81@xmath53 ( of the hmf clusters , accounting for all matches to @xmath5@xmath46 ) .
this overlap rate is consistent with expectations based on the selection functions and false - positive rates for the hmf and maxbcg clusters ( see below ) and the effects of redshift and positional uncertainties .
the overlap rate increases to @xmath7 90@xmath53 with more liberal matching criteria ( e.g. , separation larger than 1 @xmath9 mpc and/or larger than 0.08 in redshift ) . the overlap rate drops , as expected , when the richness restriction of the matching sample is tightened ( e.g. , the matching rate is 37@xmath53 if only @xmath5@xmath413 matches are considered for @xmath3@xmath440 hmf clusters ; this is consistent with expectations based on monte carlo richness simulations , [ clustercomparison ] ) .
the richest clusters , hmf with @xmath3@xmath452 , are matched at a higher rate , as expected : 90@xmath53 match with @xmath5@xmath46 maxbcg clusters and 61@xmath53 match with @xmath5@xmath413 clusters .
a summary of the catalog cluster distribution by redshift and richness is presented in table 3 .
selection functions for the independent hmf and maxbcg clusters have been determined from simulations and are presented as a function of redshift and richness in figure 10 ( for hmf ; @xcite ) and figure 11 ( for maxbcg ; @xcite ) ( see above refereneces for more details ) .
the richest clusters are nearly complete and volume limited to z @xmath14 0.3 , while the @xmath3@xmath1 40 hmf clusters are only @xmath140@xmath53 complete at z @xmath1 0.3 .
the selection functions need to be properly accounted for in any statistical analysis of the current samples .
some systems are false - positive detections ( i.e. , non - real clusters ) .
the false - positive detection rates for the clusters have been estimated from simulations @xcite as well as from visual inspection .
the false - positive rate is found to be small ( @xmath1410@xmath53 ) for the @xmath5@xmath413 maxbcg and @xmath3@xmath440 hmf clusters ( z = 0.05 - 0.3 ) .
all detections are included in the catalog , including false - positive detections , in order to avoid unquantitative visual selection .
some maxbcg systems are found to be small clumps of red galaxies in the outskirts of richer hmf clusters .
some un - matched hmf and maxbcg systems are in fact parts of the same larger cluster split into separate listings because of the @xmath93z @xmath560.08 and the 1 @xmath9 mpc separation cutoff .
this can result from uncertainties in z@xmath0 and from the different definitions of cluster center ( i.e. , hmf clusters typically center on a mean high density region , while maxbcg clusters center on a likely bcg galaxy ) .
the splittings may also represent sub - structure in clusters .
occasionally , a single hmf or maxbcg cluster may be split by the selection algorithm into two separate systems , which may represent sub - clustering .
some systems may be part of an extended galaxy overdensity region rather than true condensed virialized clusters ; this is less likely for the richer systems . the scaling relations between richness , luminosity and velocity dispersion ( [ scaling ] )
suggest that @xmath3@xmath7 40 and @xmath5@xmath7 13 clusters correspond to approximately @xmath6@xmath7 400 km s@xmath8 , and @xmath3@xmath7 60 and @xmath5@xmath7 30 clusters correspond to @xmath6@xmath7 600 km s@xmath8 , i.e. , rich clusters .
the mean calibrations are summarized in table 1 .
the distribution of clusters on the sky is mapped for the catalog clusters in figure 12 .
all clusters with 0.05 @xmath56 z @xmath56 0.3 , richness @xmath3@xmath4 40 ( for hmf ) and @xmath5@xmath4 13 ( for maxbcg ) , and their matching clusters are shown .
the abell clusters located in the survey area are also shown ( see [ abell ] ) .
a 1 @xmath33mpc radius circle is presented around the center of each cluster ; this helps visualize possible matches that may be offset in their center position due to uncertainties in cluster centers and the different definition of `` center '' ( [ clustercomparison ] ) , or may represent sub - structure within more extended regions .
images of a sample of cataloged clusters representing a wide range of redshift ( [email protected] - 0.3 ) and richness ( @xmath3@xmath740 , @xmath5@xmath713 ) are presented as examples in figure 13 .
a total of 58 abell clusters @xcite are located in the current survey region .
the sdss bh catalog includes 53 ( 91@xmath53 ) of these clusters ( listed in the last column of table 2 ) , using the matching requirement of a projected separation of less than 1 @xmath9 mpc .
( since many of abell clusters have no measured redshifts , no redshift information is used . )
most matches are at separations typically @xmath14 0.2 @xmath9 .
the five additional abell clusters not listed in the catalog are all detected by the combined hmf and maxbcg methods , but are below the catalog threshold ; these are a116 ( @xmath3= 29 , @xmath5= 9 ) , a237 ( @xmath3= 35 , @xmath5= 7 ) , a295 ( @xmath5= 11 ) , a2051 ( @xmath5= 11 ) , a2696 ( @xmath5= 11 ) .
this matching rate is consistent with the expected selection function of the hmf and maxbcg methods .
eight clusters from the noras x - ray cluster catalog @xcite lie in the sdss bh area and redshift region .
all eight x - ray clusters are detected and included in our catalog ; maxbcg detects all eight clusters ( with 2 below the threshold of @xmath5= 13 ) , and hmf detects seven of the clusters ( all within the catalog threshold of @xmath3@xmath4 40 ) .
details of the comparison are given in table 4 .
the observed distribution of cluster abundance as a function of richness the cluster richness function is presented in figure 14 .
the observed cluster counts are corrected for the relevant hmf and maxbcg selection functions .
here each cluster is corrected by the selection function appropriate for its richness and redshift ( for each method ; see figures 10 , 11 ) and by the false - positive expectation rate ( [ catalog ] ) .
the corrected count is divided by the sample volume to produce a volume - limited cluster abundance as a function of richness .
smaller corrections for richness and redshift uncertainties are not included ; these will reduce the cluster abundances by @xmath95 to @xmath96 for @xmath3@xmath140 to @xmath160 ( see @xcite ) .
the results show a steeply declining richness function with increasing richness , as expected . the richness function of the hmf - selected and maxbcg - selected clusters are consistent with each other when properly corrected for the different selection functions and scaled by the richness scaling relation .
the richness function indicates a cluster abundance of @xmath12 mpc@xmath13 for @xmath3@xmath7 40 and @xmath5@xmath7 13 clusters ( @xmath97 400 @xmath98 ) .
these abundances are in general good agreement with abell clusters and with other richness or temperature function observations when properly scaled by the relevant richness scaling relations ( e.g. , @xcite ) .
the mass function of sdss clusters was recently determined by @xcite ( for z = 0.1 - 0.2 , using an extension of the current catalog to slightly lower richnesses ) , yielding consistent results for the hmf and maxbcg subsamples .
the mass function was used by @xcite to place strong cosmological constraints on the mass density parameter of the universe , @xmath23 , and the amplitude of mass fluctuations , @xmath99 : @xmath23= 0.19 @xmath100 and @xmath99 = 0.9 @xmath101 .
we compare two independent cluster selection methods used on 379 deg@xmath2 of early sdss commissioning data : matched - filter ( hmf ) and the color - magnitude maxbcg .
we clarify the relation between the methods and the nature of clusters they select .
hmf selects clusters that follow a typical density profile and luminosity function , while maxbcg selects clusters dominated by bright red galaxies quite different selection criteria .
we determine scaling relations between the observed cluster richness , luminosity , and velocity dispersion .
we use the above scaling relations to combine appropriate subsamples of the hmf and maxbcg clusters and produce a conservative merged catalog of 799 clusters of galaxies at z@xmath0 = 0.05 - 0.3 above richness threshold of @xmath3@xmath4 40 ( hmf ) and @xmath5@xmath4 13 ( maxbcg ) ( [ catalog ] ) .
this threshold corresponds to clusters with a typical mean velocity dispersion of @xmath6@xmath7 400 km s@xmath8 , total @xmath18-band luminosity l@xmath102 l@xmath11 and mass m@xmath103 m@xmath11 ( within a radius of 0.6 @xmath9 mpc ) .
this threshold reflects clusters that are poorer than abell richness class 0 .
the average space density of the clusters is @xmath12 clusters / mpc@xmath104 . using the relevant selection functions
, we determine the cluster richness function ; we find it to be a steeply declining function of cluster abundance with increasing richness .
we compare the cataloged clusters with the abell and x - ray clusters located in the survey region ; they are all detected ( with 5 of the 58 abell clusters below the above merged richness cuts ) .
the relevant selection functions for the catalog clusters are provided .
the catalog can be used for studies of individual clusters , for comparisons with other objects ( e.g. , x - ray clusters , sz clusters , agns ) , and in statistical analyses ( when properly corrected for the relevant selection functions ) . as an example , we determined the mass function of clusters ( see @xcite ) and used it to place powerful constarints on the mass - density parameter of the universe and the amplitude of mass fluctuations ; we find @xmath23= 0.19 @xmath100 and @xmath99 = 0.9 @xmath101 .
the current work represents preliminary results from early sdss commissioning data ( 4@xmath53 of the ultimate sdss survey ) .
the results will greatly improve as more extensive sdss data become available .
the sdss is a joint project of the university of chicago , fermilab , the institute for advanced study , the japan participation group , the johns hopkins university , los alamos national laboratory , the max - planck - institute for astronomy ( mpia ) , the max - planck - institute for astrophysics ( mpa ) , new mexico state university , university of pittsburgh , princeton university , the united states naval observatory , and the university of washington .
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/. tim mckay acknowledges support from nsf pecase grant ast 9708232 .
abell , g. o. 1958 , , 3 , 211 abell , g. o. , corwin , h. g. , & olowin , r. p. 1989
, , 70 , 1 annis , j. , et al . 2003a , in preparation .
annis , j. , et al . 2003b , in preparation .
bahcall , n. a. 1977 , , 15 , 505 bahcall , n. a. & soneira , r. m. 1983 , , 270 , 20 bahcall , n. a. & soneira , r. m. 1984 , , 277 , 27 bahcall , n. a. 1988 , , 26 , 631 bahcall , n. a. & cen , r. 1992 , , 398 , l81 bahcall , n. a. , lubin , l. m. , & dorman , v. 1995 , , 447 , l81 bahcall , n. a. , fan , x. , & cen , r. 1997 , , 485 , l53 bahcall , n. a. & fan , x. 1998 , , 504 , 1 bahcall , n. a. , ostriker , j. p. , perlmutter , s. , & steinhardt , p. j. 1999 , science , 284 , 1481 bahcall , n. a. 1999 , , 525 , c873 bahcall , n. a. , cen , r. , dav ' e , r. , ostriker , j. p. , & yu , q. 2000 , , 541 , 1 bahcall , n. a. et .
2003 , , vol.585 , in press blanton , m. r. , lupton , r .h . , maley , f .m .
, young , n. , zehavi , i. , & loveday , j. 2002 , aj , in press , astro - ph/0105535 b " ohringer , h. et al . 2000 , , 129 , 435 carlberg , r. g. et al .
1997 , , 485 , l13 carlstrom , j. et al .
2001 , iap conference 2000 , constructing the universe with clusters of galaxies , edt .
f. durret and d. gerbal ( astro - ph/0103480 ) connolly , a. j. , csabai , i. , szalay , a. s. , koo , d. c. , kron , r. g. , & munn , j. a. 1995 , , 110 , 2655 croft , r. a. c. , dalton , g. b. , efstathiou , g. , sutherland , w. j. , & maddox , s. j. 1997 , , 291 , 305 csabai , i. , connolly , a. j. , szalay , a. s. , & budav ' ari , t. 2000 , , 119 , 69 dalton , g. b. , efstathiou , g. , maddox , s. j. , & sutherland , w. j. 1994 , , 269 , 151 dressler , a. 1984 , , 22 , 185 donahue , m. & voit , g. m. 1999 , , 523 , l137 ebeling , h. , voges , w. , bohringer , h. , edge , a. c. , huchra , j. p. , & briel , u. g. 1996 , , 281 , 799 eisenstein , d. , et al .
2001 , , 122 , 2267 eke , v. r. , cole , s. , & frenk , c. s. 1996 , , 282 , 263 fischer , p. & tyson , j. a. 1997 , , 114 , 14 fukugita , m. , ichikawa , t. , gunn , j. e. , doi , m. , shimasaku , k. , & schneider , d. p. 1996
, , 111 , 1748 gladders , m. d. & yee , h. k. c. 2000 , , 120 , 2148 goto , t. , et al .
2002 , aj , 123 , 1807 gunn , j. e. & dressler , a. 1988 , assl vol . 141 : towards understanding galaxies at large redshift , 227 gunn , j. e. et al .
1998 , , 116 , 3040 hansen , s. , et al .
2003 , in preparation ( senior thesis , univ . of michigan , 2003 ) henry , j. p. , gioia , i. m. , maccacaro , t. , morris , s. l. , stocke , j. t. , & wolter , a. 1992 , , 386 , 408 henry , j. p. 2000 , , 534 , 565 hoessel , j. g. , & schneider , d. p. 1985 , aj , 90 , 1648 hogg , d. w. et al .
1998 , , 115 , 1418 hogg , d. w. et al . 2001 , , 122 , 2129 huchra , j. p. , geller , m. j. , henry , j. p. , & postman , m. 1990 , , 365 , 66 ikebe , y. , reiprich , t. h. , b " ohringer , h. , tanaka , y. & kitayama , t. 2002 , a&a , 383 , 773 joffre , m. et al . 2000 , , 534 , l131 kepner , j. , fan , x. , bahcall , n. , gunn , j. , lupton , r. , & xu , g. 1999 , , 517 , 78 kim , r. , et al .
2002 , 123 , 20 .
kim , r. , et al .
2003 , in preparation klypin , a. a. & kopylov , a. i. 1983 , soviet astronomy letters , 9 , 41 lumsden , s. l. , nichol , r. c. , collins , c. a. , & guzzo , l. 1992 , , 258 , 1 lupton , r. , et al .
2001 , in preparation mazure , a. , et al .
1996 , a&a , 310 , 31 miller , c. , et al .
2003 , in preparation pier , j. r. et al .
2002 , aj , in press , astro - ph/0211375 postman , m. , huchra , j. p. , & geller , m. j. 1992 , , 384 , 404 postman , m. , lubin , l. m. , gunn , j. e. , oke , j. b. , hoessel , j. g. , schneider , d. p. , & christensen , j. a. 1996 , , 111 , 615 postman , m. , lauer , t. r. , oegerle , w. , donahue , m. 2002 , submitted to rosati , p. , stanford , s. a. , eisenhardt , p. r. , elston , r. , spinrad , h. , stern , d. , & dey , a. 1999 , , 118 , 76 rosati , p. , borgani , s. , & norman , c. 2002 , ara&a , 40 , 539 scranton , r. , et al .
2002 , , 579 , 48 shectman , s. a. 1985 , , 57 , 77 sheldon , e. s. et al . 2001 , , 554 , 881 slinglend , k. , batuski , d. , miller , c. , haase , s. , michaud , k. , hill , j. m. , 1998 , apjs , 115 , 1s smail , i. , ellis , r. s. , dressler , a. , couch , w. j. , oemler , a. j. , sharples , r. m. , & butcher , h. 1997 , , 479 , 70 smith , j. a. et al . 2002 ,
aj , 123 , 2121 squires , g. & kaiser , n. 1996 , , 473 , 65 stoughton , c. , et al .
2002 , , 123 , 485 strauss , m. , et al .
2002 , , 124 , 1810 trumper , j. 1983 , mitteilungen der astronomischen gesellschaft hamburg , 60 , 255 voges , w. et al . 1999 , , 349 , 389 voges , w. et al .
2000 , , 7432 , 1 white , s. d. m. , navarro , j. f. , evrard , a. e. , & frenk , c. s. 1993 , , 366 , 429 white , s. d. m. , efstathiou , g. , & frenk , c. s. 1993 , , 262 , 1023 york , d. g. et al .
2000 , , 120 , 1579 zwicky , f. 1957 , berlin : springer , 1957 zwicky , f. , herzog , e. , & wild , p. 1968
, pasadena : california institute of technology ( cit ) , 1961 - 1968 = 0.05 - 0.3 , @xmath3@xmath440 ( hmf ) , @xmath5@xmath413 ( maxbcg ) , and their matches ] .
dotted circles are maxbcg clusters , solid circles are hmf clusters , and bold solid circles are abell clusters in the survey area .
all circles have a radius of 1 @xmath9 mpc at the cluster redshift .
( a redshift of 0.1 is assumed for the abell clusters , many of which have no measured redshift . ) [ f12 ] ] + bh148 ( z=0.045 ; h , b ; @xmath3=55.1 ; @xmath5=12 ; a168 ; rxc0114.9 ) 0.31 in bh563 ( z=0.082 ; h , b ; @xmath3=59.3 ; @xmath5=27 ) 0.2 in + bh1 ( z=0.090 ; h ; @xmath3=45.1 ; a2644 ; [ rxc2341.1 ] ) 0.92 in bh541 ( z=0.090 ; b ; @xmath5=13 ) 0.2 in + bh100 ( z=0.117 ; h , b ; @xmath3=43.0 ; @xmath5=17 ; a101 ) 0.94 in bh269 ( z=0.121 ; h , b ; @xmath3=55.2 ; @xmath5=19 ; a867 ) 0.2 in + bh645 ( z=0.127 ; h , b ; @xmath3=52.5 ; @xmath5=15 ) 1.28 in bh650 ( z=0.139 ; h , b ; @xmath3=87.2 ; @xmath5=66 ; a1882 ) 0.2 in + bh479 ( z=0.157 ; h , b ; @xmath3=67.6 ; @xmath5=25 ) 1.28 in bh187 ( z=0.174 ; h , b ; @xmath3=67.2 ; @xmath5=24 ; a315 ) 0.2 in + bh152 ( z=0.175 ; h , b ; @xmath3=46.7 ; @xmath5=32 ; a181 ; rxc0121.9 ) 0.29 in bh379 ( z=0.185 ; h , b ; @xmath3=76.0 ; @xmath5=27 ; a1191 ) 0.2 in + bh35 ( z=0.210 ; h , b ; @xmath3=62.2 ; @xmath5=26 ) 1.34 in bh40 ( z=0.210 ; b ; @xmath5=33 ; rxc0020.1 ) 0.2 in + bh695 ( z=0.219 ; h , b ; @xmath3=66.2 ; @xmath5=16 ) 1.28 in bh481 ( z=0.225 ; h , b ; @xmath3=79.0 ; @xmath5=22 ; a1525 ) 0.2 in + bh126 ( z=0.264 ; h , b ; @xmath3=61.8 ; @xmath5=22 ; a142 ) 0.94 in bh94 ( z=0.268 ; h , b ; @xmath3=69.0 ; @xmath5=16 ) 0.2 in + bh119 ( z=0.276 ; h , b ; @xmath3=77.4 ; @xmath5=35 ) 1.30 in bh131 ( z=0.286 ; h , b ; @xmath3=50.9 ; @xmath5=25 ; rxc0106.8 ) 0.2 in rrrrr 30 & 8 & 300 & 1.6 & 0.3 + 40 & 13 & 400 & 2.8 & 0.5 + 50 & 20 & 500 & 4.4 & 0.9 + 60 & 29 & 610 & 6.3 & 1.4 + 70 & 40 & 720 & 8.5 & 2.0 + @xmath5 & @xmath3 & @xmath105 & @xmath72 & @xmath74 + & & ( @xmath106 ) & ( @xmath107 ) & ( @xmath108 ) + + 10 & 35 & 350 & 2.1 & 0.35 + 15 & 43 & 430 & 3.2 & 0.6 + 20 & 50 & 500 & 4.4 & 0.9 + 25 & 56 & 560 & 5.5 & 1.2 + 30 & 61 & 620 & 6.6 & 1.5 + 40 & 70 & 730 & 8.7 & 2.2 + lcrrrrrrrrrrr 1 & h & 355.271 & 0.088 & 0.09 & 45.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & a2644 , ( rxc 2341.1 ) + 2 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 355.277 & -0.002 & 0.205 & 16 & 0.190 & + 3 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 355.342 & 0.333 & 0.244 & 16 & 0.278 & + 4 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 355.916 & 0.424 & 0.190 & 13 & 0.186 & + 5 & h , b & 355.945 & 0.263 & 0.24 & 64.4 & & 355.899 & 0.331 & 0.268 & 38 & 0.000 & + 6 & h , b & 356.518 & -0.185 & 0.25 & 46.0 & & 356.520 & -0.186 & 0.257 & 17 & 0.266 & + 7 & h , b & 356.604 & 0.732 & 0.21 & 49.1 & & 356.607 & 0.728 & 0.218 & 20 & 0.000 & + 8 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 356.847 & -0.164 & 0.227 & 19 & 0.000 & + 9 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 356.944 & -0.106 & 0.238 & 20 & 0.263 & + 10 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 357.021 & 0.540 & 0.281 & 14 & 0.000 & + + 11 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 357.574 & 0.188 & 0.242 & 13 & 0.000 & + 12 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 357.709 & 0.573 & 0.225 & 13 & 0.000 & + 13 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 357.977 & 0.553 & 0.295 & 13 & 0.271 & + 14 & h , b & 358.195 & 1.187 & 0.24 & 42.6 & & 358.194 & 1.193 & 0.238 & 11 & 0.000 & + 15 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 358.272 & 0.761 & 0.284 & 13 & 0.000 & + 16 & h , b & 358.381 & 0.673 & 0.22 & 40.2 & & 358.391 & 0.651 & 0.199 & 18 & 0.228 & + 17 & h , b & 358.535 & 0.993 & 0.20 & 34.4 & & 358.531 & 0.992 & 0.227 & 13 & 0.241 & + 18 & h , b & 358.784 & 0.690 & 0.27 & 42.8 & & 358.849 & 0.727 & 0.286 & 14 & 0.272 & + 19 & h , b & 359.419 & 0.184 & 0.26 & 45.8 & & 359.417 & 0.184 & 0.325 & 6 & 0.000 & + 20 & h , b & 359.988 & 0.696 & 0.08 & 41.0 & & 359.995 & 0.702 & 0.077 & 10 & 0.084 & + + 21 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 0.211 & 0.785 & 0.229 & 13 & 0.000 & + 22 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 0.219 & 0.804 & 0.286 & 17 & 0.000 & + 23 & h , b & 0.354 & -0.036 & 0.21 & 44.0 & & 0.360 & -0.029 & 0.247 & 24 & 0.248 & + 24 & h , b & 0.516 & 0.359 & 0.22 & 36.5 & & 0.508 & 0.347 & 0.284 & 13 & 0.301 & + 25 & h , b & 0.671 & -0.521 & 0.22 & 40.4 & & 0.686 & -0.510 & 0.203 & 15 & 0.195 & + 26 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 0.844 & 1.058 & 0.081 & 13 & 0.079 & + 27 & h , b & 1.239 & 0.963 & 0.09 & 32.6 & & 1.204 & 0.951 & 0.094 & 17 & 0.100 & + 28 & h , b & 1.655 & -0.320 & 0.28 & 40.3 & & 1.614 & -0.277 & 0.329 & 7 & 0.000 & + 29 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 2.030 & 1.094 & 0.242 & 14 & 0.000 & + 30 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 2.102 & 0.261 & 0.299 & 13 & 0.000 & + + 31 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 3.715 & -0.264 & 0.153 & 13 & 0.155 & + 32 & h , b & 4.076 & 0.852 & 0.28 & 43.8 & & 4.067 & 0.871 & 0.275 & 11 & 0.000 & + 33 & h , b & 4.172 & -0.451 & 0.07 & 37.8 & & 4.177 & -0.445 & 0.070 & 13 & 0.063 & + 34 & h , b & 4.191 & -1.142 & 0.16 & 43.7 & & 4.228 & -1.133 & 0.151 & 19 & 0.153 & + 35 & h , b & 4.412 & -0.873 & 0.23 & 62.2 & & 4.404 & -0.854 & 0.207 & 26 & 0.210 & + 36 & h , b & 4.627 & -0.764 & 0.14 & 36.7 & & 4.622 & -0.778 & 0.190 & 18 & 0.191 & + 37 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 4.792 & -0.856 & 0.183 & 15 & 0.191 & + 38 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 5.019 & 0.200 & 0.194 & 20 & 0.207 & + 39 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 5.041 & 0.941 & 0.220 & 15 & 0.213 & + 40 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 5.067 & 0.079 & 0.220 & 33 & 0.210 & rxc 0020.1 + + 41 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 5.089 & -0.253 & 0.212 & 17 & 0.210 & + 42 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 5.241 & 0.155 & 0.212 & 29 & 0.216 & + 43 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 5.348 & -0.826 & 0.168 & 22 & 0.168 & + 44 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 5.427 & -0.876 & 0.107 & 27 & 0.105 & a23 + 45 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 5.471 & -0.670 & 0.299 & 13 & 0.239 & + 46 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 5.622 & -0.719 & 0.162 & 13 & 0.162 & + 47 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 5.723 & 0.601 & 0.262 & 13 & 0.000 & + 48 & h , b & 5.767 & -0.123 & 0.16 & 68.0 & & 5.752 & -0.155 & 0.157 & 29 & 0.158 & a25 + 49 & h , b & 5.784 & 0.975 & 0.26 & 42.2 & & 5.794 & 0.976 & 0.216 & 9 & 0.227 & + 50 & h , b & 5.803 & 0.540 & 0.30 & 49.6 & & 5.799 & 0.589 & 0.368 & 7 & 0.000 & + + 51 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 5.872 & 0.199 & 0.268 & 20 & 0.000 & + 52 & h , b & 5.908 & 0.049 & 0.29 & 47.7 & & 5.902 & 0.048 & 0.249 & 8 & 0.000 & + 53 & h , b & 5.915 & -0.795 & 0.08 & 31.4 & & 5.976 & -0.725 & 0.072 & 14 & 0.064 & + 54 & h , b & 6.038 & -0.466 & 0.26 & 45.6 & & 5.996 & -0.511 & 0.271 & 17 & 0.000 & + 55 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 6.114 & 0.414 & 0.260 & 15 & 0.260 & + 56 & h , b & 6.071 & 0.141 & 0.32 & 53.6 & & 6.124 & 0.121 & 0.295 & 15 & 0.000 & + 57 & h , b & 6.144 & -0.226 & 0.25 & 47.0 & & 6.076 & -0.223 & 0.275 & 12 & 0.231 & + 58 & h & 6.492 & 0.122 & 0.25 & 53.2 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 59 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 6.511 & -0.271 & 0.275 & 14 & 0.231 & + 60 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 6.565 & 1.221 & 0.290 & 17 & 0.000 & + + 61 & h , b & 6.707 & -0.266 & 0.23 & 38.9 & & 6.703 & -0.269 & 0.247 & 23 & 0.231 & + 62 & h , b & 6.774 & -0.113 & 0.26 & 42.4 & & 6.778 & -0.115 & 0.201 & 17 & 0.000 & + 63 & h , b & 6.951 & 0.162 & 0.28 & 52.1 & & 6.871 & 0.175 & 0.207 & 17 & 0.000 & + 64 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 7.088 & -0.017 & 0.209 & 24 & 0.216 & + 65 & h , b & 7.092 & 0.616 & 0.29 & 44.5 & & 7.091 & 0.589 & 0.336 & 13 & 0.000 & + 66 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 7.168 & -0.483 & 0.240 & 13 & 0.230 & + 67 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 7.179 & 0.269 & 0.223 & 15 & 0.223 & + 68 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 7.185 & -0.096 & 0.216 & 24 & 0.217 & + 69 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 7.254 & 0.882 & 0.229 & 18 & 0.000 & + 70 & h , b & 7.301 & 0.961 & 0.32 & 64.2 & & 7.303 & 0.962 & 0.281 & 18 & 0.000 & + + 71 & h , b & 7.309 & -0.181 & 0.08 & 56.2 & & 7.368 & -0.213 & 0.079 & 17 & 0.060 & + 72 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 7.328 & -0.182 & 0.281 & 13 & 0.000 & + 73 & h , b & 7.622 & -0.940 & 0.23 & 44.0 & & 7.631 & -0.931 & 0.199 & 15 & 0.000 & + 74 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 7.737 & -0.169 & 0.209 & 16 & 0.225 & + 75 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 7.887 & -0.008 & 0.199 & 15 & 0.219 & + 76 & h & 8.048 & -0.673 & 0.15 & 40.6 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 77 & h , b & 8.182 & -0.897 & 0.33 & 45.9 & & 8.220 & -0.945 & 0.286 & 13 & 0.000 & + 78 & h , b & 8.320 & -0.631 & 0.29 & 50.0 & & 8.321 & -0.633 & 0.275 & 10 & 0.261 & + 79 & h , b & 8.488 & -0.747 & 0.26 & 50.8 & & 8.488 & -0.744 & 0.329 & 10 & 0.000 & + 80 & h , b & 8.593 & 0.863 & 0.17 & 46.1 & & 8.604 & 0.810 & 0.166 & 23 & 0.191 & + + 81 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 8.625 & 0.994 & 0.242 & 13 & 0.000 & + 82 & h , b & 8.744 & -1.199 & 0.23 & 44.3 & & 8.757 & -1.205 & 0.199 & 7 & 0.212 & + 83 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 8.980 & 0.356 & 0.253 & 20 & 0.259 & + 84 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 9.022 & 0.158 & 0.253 & 23 & 0.000 & + 85 & h , b & 9.068 & 0.445 & 0.20 & 39.1 & & 9.083 & 0.405 & 0.231 & 20 & 0.258 & + 86 & h , b & 9.258 & -0.334 & 0.23 & 41.9 & & 9.231 & -0.342 & 0.264 & 14 & 0.265 & + 87 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 9.287 & 0.174 & 0.188 & 13 & 0.194 & + 88 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 9.299 & 0.096 & 0.264 & 14 & 0.259 & + 89 & h , b & 9.457 & 0.069 & 0.22 & 52.3 & & 9.418 & 0.028 & 0.242 & 20 & 0.256 & + 90 & h , b & 9.800 & -0.322 & 0.20 & 42.4 & & 9.811 & -0.324 & 0.231 & 11 & 0.216 & + + 91 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 9.852 & -0.214 & 0.209 & 19 & 0.219 & + 92 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 10.061 & -0.123 & 0.177 & 14 & 0.216 & + 93 & h , b & 10.137 & -0.227 & 0.26 & 53.6 & & 10.138 & -0.233 & 0.209 & 16 & 0.238 & + 94 & h , b & 10.710 & 0.218 & 0.27 & 69.0 & & 10.700 & 0.231 & 0.286 & 16 & 0.268 & + 95 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 10.814 & 0.229 & 0.231 & 14 & 0.000 & + 96 & h , b & 11.001 & 0.112 & 0.19 & 52.6 & & 11.005 & 0.113 & 0.264 & 13 & 0.217 & + 97 & h , b & 11.444 & -0.881 & 0.13 & 33.5 & & 11.553 & -1.020 & 0.090 & 17 & 0.111 & a95 + 98 & h , b & 11.567 & 0.010 & 0.12 & 52.2 & & 11.601 & 0.002 & 0.090 & 19 & 0.116 & + 99 & h , b & 11.594 & -0.154 & 0.18 & 46.7 & & 11.593 & -0.155 & 0.199 & 16 & 0.218 & + 100 & h , b & 11.882 & -0.875 & 0.10 & 43.0 & & 11.881 & -0.882 & 0.111 & 17 & 0.117 & a101 + + 101 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 12.469 & 1.077 & 0.297 & 22 & 0.302 & + 102 & h & 12.986 & -1.146 & 0.16 & 46.3 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 103 & h , b & 13.436 & -0.781 & 0.09 & 36.2 & & 13.443 & -0.780 & 0.133 & 24 & 0.138 & a112 + 104 & h , b & 14.086 & -1.178 & 0.05 & 40.3 & & 14.071 & -0.953 & 0.122 & 10 & 0.140 & a119 + 105 & h , b & 14.048 & -1.023 & 0.05 & 36.7 & & 14.259 & -0.875 & 0.068 & 28 & 0.044 & a119 + 106 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 15.303 & -0.578 & 0.188 & 13 & 0.192 & + 107 & h , b & 15.387 & 0.542 & 0.13 & 33.2 & & 15.308 & 0.574 & 0.199 & 14 & 0.199 & + 108 & h , b & 15.356 & -0.068 & 0.09 & 39.1 & & 15.359 & -0.078 & 0.090 & 13 & 0.112 & a130 + 109 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 15.371 & -0.262 & 0.209 & 13 & 0.193 & + 110 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 15.372 & 0.760 & 0.199 & 14 & 0.200 & + + 111 & h , b & 15.569 & -1.152 & 0.26 & 40.2 & & 15.546 & -1.165 & 0.286 & 6 & 0.000 & + 112 & h , b & 15.672 & 1.145 & 0.20 & 57.7 & & 15.679 & 1.136 & 0.155 & 16 & 0.144 & + 113 & h , b & 15.694 & 0.250 & 0.28 & 53.8 & & 15.689 & 0.183 & 0.286 & 9 & 0.000 & + 114 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 15.718 & -0.900 & 0.286 & 18 & 0.000 & + 115 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 15.827 & -0.897 & 0.253 & 14 & 0.000 & + 116 & h , b & 15.831 & 0.312 & 0.23 & 45.6 & & 15.847 & 0.319 & 0.188 & 16 & 0.238 & + 117 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 15.943 & -0.860 & 0.242 & 14 & 0.000 & + 118 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 16.018 & -0.435 & 0.275 & 30 & 0.279 & + 119 & h , b & 16.227 & 0.059 & 0.23 & 77.4 & & 16.231 & 0.060 & 0.308 & 35 & 0.276 & + 120 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 16.293 & -0.160 & 0.231 & 15 & 0.262 & + + 121 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 16.430 & -0.777 & 0.209 & 13 & 0.000 & + 122 & h , b & 16.447 & -0.340 & 0.27 & 60.8 & & 16.396 & -0.331 & 0.340 & 16 & 0.326 & + 123 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 16.479 & 0.093 & 0.242 & 22 & 0.248 & + 124 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 16.484 & 0.715 & 0.264 & 16 & 0.251 & + 125 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 16.506 & 0.361 & 0.242 & 13 & 0.252 & + 126 & h , b & 16.544 & 0.863 & 0.25 & 61.8 & & 16.549 & 0.914 & 0.264 & 22 & 0.000 & a142 + 127 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 16.552 & 1.139 & 0.220 & 14 & 0.189 & + 128 & h , b & 16.671 & 0.290 & 0.23 & 48.0 & & 16.748 & 0.255 & 0.275 & 15 & 0.270 & + 129 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 16.678 & -0.324 & 0.286 & 15 & 0.278 & + 130 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 16.703 & 0.550 & 0.253 & 13 & 0.000 & + + 131 & h , b & 16.712 & 1.058 & 0.22 & 50.9 & & 16.711 & 1.070 & 0.286 & 25 & 0.000 & rxc 0106.8 + 132 & h , b & 16.772 & -0.304 & 0.33 & 49.0 & & 16.758 & -0.382 & 0.253 & 25 & 0.000 & + 133 & h , b & 16.871 & 0.142 & 0.25 & 42.8 & & 16.878 & 0.156 & 0.297 & 15 & 0.296 & + 134 & h , b & 16.897 & 0.517 & 0.26 & 42.4 & & 16.976 & 0.515 & 0.264 & 15 & 0.000 & + 135 & h , b & 16.972 & 0.351 & 0.29 & 50.0 & & 16.968 & 0.357 & 0.329 & 17 & 0.313 & + 136 & h , b & 17.129 & -0.153 & 0.26 & 41.0 & & 17.087 & -0.155 & 0.264 & 15 & 0.291 & + 137 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 17.346 & 0.908 & 0.253 & 15 & 0.264 & + 138 & h , b & 17.367 & 1.029 & 0.25 & 41.6 & & 17.299 & 1.013 & 0.242 & 13 & 0.000 & + 139 & h , b & 17.384 & -0.893 & 0.21 & 49.1 & & 17.407 & -0.922 & 0.144 & 8 & 0.109 & + 140 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 17.545 & 1.105 & 0.188 & 14 & 0.000 & + + 141 & h , b & 17.664 & 1.056 & 0.17 & 34.6 & & 17.666 & 1.067 & 0.155 & 19 & 0.178 & + 142 & h & 17.871 & 0.553 & 0.29 & 52.7 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 143 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 17.885 & -0.027 & 0.286 & 18 & 0.000 & + 144 & h , b & 17.948 & -0.012 & 0.18 & 44.4 & & 17.953 & -0.018 & 0.209 & 10 & 0.178 & + 145 & h , b & 18.019 & -0.730 & 0.14 & 36.5 & & 18.112 & -0.777 & 0.166 & 17 & 0.179 & + 146 & h , b & 18.230 & 0.988 & 0.13 & 42.0 & & 18.306 & 1.109 & 0.111 & 9 & 0.133 & ( large spiral ) + 147 & h , b & 18.613 & -0.851 & 0.13 & 37.2 & & 18.562 & -0.913 & 0.155 & 16 & 0.000 & + 148 & h , b & 18.742 & 0.287 & 0.05 & 55.1 & & 18.816 & 0.213 & 0.068 & 12 & 0.045 & a168 , rxc 0114.9 + 149 & h , b & 18.787 & 0.404 & 0.05 & 42.6 & & 18.816 & 0.213 & 0.068 & 12 & 0.045 & a168 , rxc 0114.9 + 150 & h , b & 19.098 & -0.134 & 0.15 & 36.0 & & 19.102 & -0.133 & 0.155 & 16 & 0.177 & + + 151 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 19.885 & -1.185 & 0.177 & 15 & 0.186 & + 152 & h , b & 20.506 & 0.333 & 0.16 & 46.7 & & 20.511 & 0.335 & 0.166 & 32 & 0.175 & a181 , rxc 0121.9 + 153 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 20.652 & -0.814 & 0.177 & 22 & 0.173 & + 154 & h , b & 21.629 & 0.571 & 0.34 & 66.8 & & 21.624 & 0.573 & 0.286 & 16 & 0.000 & + 155 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 21.962 & 0.453 & 0.297 & 14 & 0.287 & + 156 & h , b & 22.049 & -0.671 & 0.26 & 42.0 & & 22.020 & -0.665 & 0.231 & 10 & 0.256 & + 157 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 22.371 & 0.154 & 0.286 & 16 & 0.286 & + 158 & h , b & 22.886 & 0.561 & 0.07 & 36.6 & & 22.823 & 0.505 & 0.101 & 15 & 0.081 & a208 + 159 & h , b & 22.920 & 1.007 & 0.32 & 43.6 & & 22.899 & 0.936 & 0.297 & 15 & 0.293 & + 160 & h , b & 23.251 & 0.938 & 0.11 & 35.3 & & 23.228 & 0.976 & 0.122 & 15 & 0.133 & + + 161 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 23.379 & 0.830 & 0.297 & 18 & 0.000 & + 162 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 23.724 & -0.665 & 0.090 & 14 & 0.086 & + 163 & h , b & 23.833 & 0.343 & 0.19 & 40.0 & & 23.731 & 0.388 & 0.144 & 13 & 0.153 & + 164 & h & 25.749 & 0.906 & 0.26 & 41.0 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 165 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 25.819 & 0.973 & 0.166 & 14 & 0.167 & + 166 & h , b & 26.233 & 0.643 & 0.16 & 33.0 & & 26.237 & 0.642 & 0.188 & 14 & 0.211 & + 167 & h & 26.423 & 0.189 & 0.20 & 43.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 168 & h , b & 26.944 & 0.186 & 0.23 & 41.4 & & 26.867 & 0.187 & 0.177 & 13 & 0.185 & + 169 & h & 27.035 & 0.009 & 0.30 & 41.3 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 170 & h , b & 27.051 & 0.333 & 0.14 & 27.6 & & 27.069 & 0.358 & 0.166 & 16 & 0.206 & + + 171 & h , b & 27.073 & 0.196 & 0.14 & 24.1 & & 27.078 & 0.166 & 0.199 & 16 & 0.204 & + 172 & h , b & 27.895 & -1.119 & 0.20 & 42.7 & & 27.869 & -1.018 & 0.231 & 6 & 0.000 & + 173 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 28.075 & 1.083 & 0.209 & 19 & 0.226 & + 174 & h , b & 28.195 & 0.997 & 0.23 & 73.4 & & 28.175 & 1.007 & 0.264 & 29 & 0.230 & a267 , rxc 0152.7 + 175 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 28.210 & 1.212 & 0.199 & 13 & 0.227 & + 176 & h , b & 28.299 & 1.035 & 0.08 & 47.6 & & 28.197 & 1.105 & 0.068 & 16 & 0.059 & + 177 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 28.357 & -1.160 & 0.231 & 34 & 0.242 & a268 + 178 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 29.071 & 1.051 & 0.068 & 19 & 0.079 & a279 + 179 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 29.129 & 0.803 & 0.220 & 27 & 0.220 & + 180 & h & 29.324 & -0.122 & 0.28 & 50.0 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + + 181 & h , b & 29.488 & -0.614 & 0.16 & 50.3 & & 29.497 & -0.724 & 0.177 & 29 & 0.187 & a284 + 182 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 29.695 & -0.695 & 0.188 & 18 & 0.191 & + 183 & h , b & 30.129 & 0.581 & 0.18 & 56.9 & & 30.126 & 0.582 & 0.144 & 12 & 0.164 & + 184 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 31.126 & 0.305 & 0.177 & 15 & 0.173 & a299 + 185 & h , b & 31.484 & 0.025 & 0.12 & 27.5 & & 31.486 & 0.019 & 0.166 & 15 & 0.173 & + 186 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 32.040 & -1.059 & 0.264 & 13 & 0.000 & + 187 & h , b & 32.531 & -1.035 & 0.18 & 67.2 & & 32.556 & -1.090 & 0.177 & 24 & 0.174 & a315 + 188 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 32.646 & -0.609 & 0.286 & 23 & 0.284 & + 189 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 32.727 & -1.157 & 0.166 & 24 & 0.173 & + 190 & h , b & 32.735 & 1.178 & 0.26 & 49.7 & & 32.736 & 1.168 & 0.308 & 15 & 0.000 & + + 191 & h , b & 32.799 & 0.180 & 0.23 & 37.6 & & 32.794 & 0.192 & 0.188 & 13 & 0.112 & + 192 & h , b & 32.908 & 0.386 & 0.26 & 46.3 & & 32.906 & 0.387 & 0.297 & 14 & 0.000 & a321 + 193 & h , b & 32.931 & 0.691 & 0.30 & 48.4 & & 32.951 & 0.722 & 0.253 & 9 & 0.000 & + 194 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 32.986 & 0.426 & 0.220 & 14 & 0.000 & + 195 & h , b & 33.174 & 0.480 & 0.14 & 42.6 & & 33.175 & 0.475 & 0.177 & 15 & 0.149 & + 196 & h , b & 33.468 & 0.475 & 0.18 & 41.4 & & 33.464 & 0.467 & 0.177 & 19 & 0.182 & + 197 & h , b & 33.549 & -0.193 & 0.17 & 48.7 & & 33.553 & -0.189 & 0.144 & 15 & 0.141 & + 198 & h , b & 34.002 & 0.867 & 0.25 & 40.2 & & 33.964 & 0.870 & 0.209 & 22 & 0.212 & + 199 & h , b & 34.090 & 0.267 & 0.28 & 44.6 & & 34.085 & 0.266 & 0.209 & 12 & 0.243 & + 200 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 34.666 & -0.209 & 0.275 & 13 & 0.000 & + + 201 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 34.675 & 0.114 & 0.275 & 20 & 0.272 & + 202 & h , b & 34.751 & -0.586 & 0.26 & 30.0 & & 34.723 & -0.529 & 0.286 & 13 & 0.273 & + 203 & h , b & 35.139 & 0.468 & 0.31 & 48.8 & & 35.137 & 0.468 & 0.286 & 23 & 0.272 & + 204 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 35.877 & 1.134 & 0.231 & 14 & 0.000 & + 205 & h , b & 35.957 & 1.039 & 0.18 & 36.0 & & 35.980 & 0.978 & 0.220 & 13 & 0.000 & + 206 & h , b & 36.304 & -1.064 & 0.20 & 45.5 & & 36.309 & -1.086 & 0.166 & 8 & 0.169 & + 207 & h , b & 36.687 & -1.123 & 0.34 & 52.1 & & 36.685 & -1.132 & 0.286 & 19 & 0.000 & + 208 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 36.892 & -0.922 & 0.286 & 13 & 0.000 & + 209 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 37.118 & -0.913 & 0.253 & 14 & 0.300 & + 210 & h , b & 37.293 & -1.124 & 0.25 & 37.1 & & 37.284 & -1.045 & 0.286 & 14 & 0.000 & + + 211 & h , b & 38.623 & -0.271 & 0.23 & 37.9 & & 38.661 & -0.296 & 0.231 & 16 & 0.245 & + 212 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 38.936 & -0.947 & 0.220 & 15 & 0.000 & + 213 & h , b & 39.032 & -1.002 & 0.27 & 47.3 & & 39.036 & -1.007 & 0.231 & 18 & 0.251 & + 214 & h , b & 39.084 & -0.608 & 0.25 & 41.6 & & 39.084 & -0.591 & 0.253 & 7 & 0.244 & + 215 & h , b & 39.116 & 0.229 & 0.21 & 31.3 & & 39.144 & 0.242 & 0.264 & 17 & 0.270 & + 216 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 39.588 & 0.282 & 0.199 & 13 & 0.000 & + 217 & h , b & 40.213 & -0.930 & 0.18 & 35.2 & & 40.213 & -0.932 & 0.231 & 19 & 0.240 & + 218 & h , b & 40.604 & -0.855 & 0.24 & 37.0 & & 40.606 & -0.865 & 0.220 & 13 & 0.000 & + 219 & h , b & 40.806 & -1.028 & 0.20 & 45.4 & & 40.801 & -1.020 & 0.220 & 22 & 0.239 & + 220 & h , b & 40.935 & -0.240 & 0.29 & 42.8 & & 40.934 & -0.236 & 0.341 & 19 & 0.377 & + + 221 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 41.030 & -0.883 & 0.231 & 14 & 0.000 & + 222 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 41.418 & -0.723 & 0.177 & 25 & 0.182 & a381 + 223 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 42.296 & -0.057 & 0.177 & 19 & 0.175 & + 224 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 42.301 & -0.835 & 0.264 & 13 & 0.272 & + 225 & h , b & 42.392 & 0.051 & 0.22 & 49.2 & & 42.438 & 0.102 & 0.177 & 13 & 0.185 & + 226 & h , b & 43.194 & 1.081 & 0.14 & 40.6 & & 43.192 & 1.085 & 0.133 & 10 & 0.137 & + 227 & h , b & 43.554 & 1.229 & 0.29 & 50.8 & & 43.528 & 1.243 & 0.275 & 11 & 0.000 & + 228 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 43.599 & 0.248 & 0.297 & 17 & 0.361 & + 229 & h , b & 43.759 & 1.203 & 0.20 & 46.0 & & 43.713 & 1.204 & 0.199 & 13 & 0.000 & + 230 & h , b & 44.644 & -0.082 & 0.29 & 48.7 & & 44.669 & -0.088 & 0.308 & 6 & 0.000 & + + 231 & h , b & 44.883 & 0.239 & 0.13 & 37.9 & & 44.886 & 0.232 & 0.188 & 29 & 0.193 & + 232 & h , b & 45.750 & -0.199 & 0.18 & 44.5 & & 45.748 & -0.193 & 0.166 & 9 & 0.158 & + 233 & h , b & 46.065 & 1.003 & 0.17 & 44.4 & & 46.109 & 1.046 & 0.144 & 17 & 0.153 & a411 + 234 & h , b & 46.386 & -1.074 & 0.27 & 57.2 & & 46.369 & -1.039 & 0.340 & 20 & 0.000 & + 235 & h , b & 46.571 & -0.141 & 0.11 & 42.7 & & 46.572 & -0.140 & 0.101 & 21 & 0.110 & a412 + 236 & h , b & 47.046 & 1.029 & 0.29 & 38.6 & & 47.074 & 0.963 & 0.253 & 13 & 0.000 & + 237 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 47.113 & -0.353 & 0.275 & 13 & 0.000 & + 238 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 47.272 & 0.038 & 0.286 & 20 & 0.000 & + 239 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 48.011 & -0.678 & 0.231 & 13 & 0.000 & + 240 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 48.020 & -0.689 & 0.286 & 17 & 0.000 & + + 241 & h , b & 48.578 & -0.613 & 0.09 & 25.7 & & 48.573 & -0.610 & 0.101 & 13 & 0.115 & + 242 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 48.663 & -0.879 & 0.242 & 19 & 0.000 & + 243 & h & 50.193 & 0.178 & 0.30 & 49.4 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 244 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 50.553 & 0.323 & 0.286 & 13 & 0.305 & + 245 & h , b & 50.585 & 0.432 & 0.14 & 39.5 & & 50.594 & 0.437 & 0.122 & 16 & 0.131 & + 246 & h & 50.777 & 0.882 & 0.28 & 40.7 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 247 & h , b & 51.194 & 1.022 & 0.23 & 36.7 & & 51.194 & 1.017 & 0.297 & 20 & 0.000 & + 248 & h , b & 51.277 & 0.080 & 0.26 & 36.6 & & 51.234 & 0.068 & 0.220 & 16 & 0.213 & + 249 & h , b & 52.072 & 0.733 & 0.30 & 58.6 & & 52.062 & 0.751 & 0.308 & 25 & 0.315 & + 250 & h , b & 52.429 & 0.236 & 0.28 & 52.5 & & 52.446 & 0.265 & 0.297 & 21 & 0.320 & + + 251 & h & 52.983 & 0.469 & 0.29 & 40.9 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 252 & h , b & 53.006 & -0.794 & 0.16 & 46.4 & & 53.017 & -0.799 & 0.144 & 14 & 0.160 & + 253 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 53.262 & 0.386 & 0.286 & 19 & 0.332 & + 254 & h , b & 53.685 & -0.733 & 0.32 & 78.1 & & 53.602 & -0.753 & 0.253 & 21 & 0.267 & + 255 & h , b & 53.605 & -1.166 & 0.25 & 61.7 & & 53.571 & -1.157 & 0.329 & 7 & 0.000 & + 256 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 53.918 & -0.539 & 0.264 & 13 & 0.270 & + 257 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 53.993 & -0.620 & 0.253 & 17 & 0.279 & + 258 & h & 54.295 & -0.881 & 0.28 & 40.6 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 259 & h , b & 54.507 & 0.486 & 0.28 & 46.6 & & 54.493 & 0.483 & 0.329 & 18 & 0.323 & + 260 & h , b & 55.125 & -0.100 & 0.25 & 35.9 & & 55.114 & -0.102 & 0.231 & 15 & 0.000 & + + 261 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 55.124 & 0.063 & 0.242 & 17 & 0.000 & + 262 & h & 145.785 & -0.212 & 0.29 & 55.6 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 263 & h & 145.854 & 0.050 & 0.13 & 54.9 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & a861 + 264 & h , b & 146.087 & 0.284 & 0.12 & 28.7 & & 146.006 & 0.146 & 0.148 & 20 & 0.127 & ( a861 ) + 265 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 146.022 & -0.900 & 0.277 & 13 & 0.000 & + 266 & h , b & 146.114 & 0.477 & 0.22 & 43.4 & & 146.106 & 0.471 & 0.209 & 6 & 0.000 & + 267 & h , b & 146.147 & -0.732 & 0.30 & 67.9 & & 146.191 & -0.737 & 0.305 & 6 & 0.000 & + 268 & h & 146.178 & 0.957 & 0.28 & 43.9 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 269 & h , b & 146.316 & 0.583 & 0.14 & 55.2 & & 146.275 & 0.581 & 0.138 & 19 & 0.121 & a867 + 270 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 146.425 & 0.567 & 0.295 & 13 & 0.263 & + + 271 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 146.781 & 0.738 & 0.284 & 19 & 0.262 & + 272 & h & 146.782 & 0.713 & 0.13 & 45.2 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 273 & h , b & 147.314 & 0.896 & 0.20 & 42.4 & & 147.318 & 0.822 & 0.255 & 12 & 0.225 & + 274 & h , b & 147.316 & 0.340 & 0.23 & 56.2 & & 147.330 & 0.253 & 0.286 & 11 & 0.000 & + 275 & h , b & 147.710 & -0.881 & 0.31 & 57.2 & & 147.709 & -0.888 & 0.290 & 21 & 0.271 & + 276 & h & 147.901 & 0.567 & 0.26 & 40.7 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 277 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 148.179 & 0.757 & 0.273 & 13 & 0.000 & + 278 & h & 148.212 & -0.728 & 0.24 & 43.2 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 279 & h , b & 148.225 & 0.351 & 0.22 & 44.9 & & 148.233 & 0.358 & 0.247 & 11 & 0.255 & + 280 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 148.293 & 0.204 & 0.153 & 13 & 0.163 & + + 281 & h , b & 148.399 & 0.596 & 0.10 & 50.2 & & 148.377 & 0.449 & 0.083 & 15 & 0.080 & a892 + 282 & h , b & 148.622 & -1.111 & 0.19 & 41.8 & & 148.542 & -1.195 & 0.146 & 15 & 0.138 & + 283 & h & 148.727 & 0.309 & 0.22 & 44.6 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 284 & h , b & 149.139 & -0.354 & 0.11 & 41.3 & & 149.161 & -0.358 & 0.077 & 20 & 0.088 & + 285 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 149.704 & 1.081 & 0.081 & 14 & 0.079 & + 286 & h , b & 149.747 & -0.193 & 0.14 & 42.8 & & 149.746 & -0.206 & 0.175 & 25 & 0.170 & + 287 & h , b & 149.750 & -0.976 & 0.24 & 40.5 & & 149.768 & -1.026 & 0.308 & 8 & 0.000 & + 288 & h & 149.849 & 1.240 & 0.21 & 45.2 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 289 & h & 149.850 & 1.030 & 0.19 & 49.9 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 290 & h , b & 149.875 & 0.325 & 0.14 & 35.5 & & 149.867 & 0.287 & 0.188 & 16 & 0.169 & + + 291 & h , b & 149.957 & 0.815 & 0.20 & 51.4 & & 149.949 & 0.819 & 0.233 & 17 & 0.219 & + 292 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 150.108 & 0.234 & 0.216 & 20 & 0.221 & + 293 & h , b & 150.282 & -0.087 & 0.06 & 55.9 & & 150.297 & -0.133 & 0.094 & 15 & 0.090 & a912 + 294 & h , b & 150.593 & -0.775 & 0.12 & 29.8 & & 150.470 & -0.878 & 0.124 & 14 & 0.136 & + 295 & h , b & 150.552 & -1.222 & 0.29 & 43.7 & & 150.557 & -1.210 & 0.334 & 7 & 0.000 & + 296 & h , b & 151.261 & -0.706 & 0.12 & 42.7 & & 151.262 & -0.709 & 0.192 & 28 & 0.196 & a919 + 297 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 151.540 & -0.740 & 0.166 & 13 & 0.188 & + 298 & h , b & 151.862 & -1.010 & 0.26 & 43.2 & & 151.904 & -1.027 & 0.338 & 11 & 0.377 & + 299 & h , b & 151.927 & -0.548 & 0.18 & 40.3 & & 151.926 & -0.552 & 0.216 & 13 & 0.221 & + 300 & h , b & 151.962 & 0.534 & 0.08 & 56.7 & & 151.886 & 0.594 & 0.096 & 33 & 0.000 & a933 + + 301 & h , b & 152.284 & -0.257 & 0.06 & 39.9 & & 152.266 & -0.280 & 0.066 & 13 & 0.068 & + 302 & h , b & 152.438 & 0.402 & 0.13 & 37.4 & & 152.451 & 0.484 & 0.183 & 13 & 0.186 & + 303 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 152.463 & -0.312 & 0.190 & 15 & 0.102 & + 304 & h , b & 152.804 & -0.066 & 0.06 & 45.6 & & 152.833 & -0.023 & 0.061 & 8 & 0.000 & + 305 & h , b & 152.868 & -0.693 & 0.21 & 37.9 & & 152.888 & -0.717 & 0.229 & 14 & 0.216 & + 306 & h , b & 152.973 & -0.472 & 0.19 & 54.3 & & 152.987 & -0.474 & 0.188 & 25 & 0.000 & + 307 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 152.980 & -0.403 & 0.260 & 25 & 0.000 & + 308 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 153.087 & -0.588 & 0.186 & 19 & 0.000 & + 309 & h , b & 153.379 & 0.755 & 0.30 & 47.9 & & 153.352 & 0.772 & 0.292 & 10 & 0.000 & + 310 & h , b & 153.426 & -0.129 & 0.14 & 80.6 & & 153.437 & -0.120 & 0.094 & 21 & 0.101 & a954 + + 311 & h & 153.430 & -0.907 & 0.07 & 71.6 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 312 & h , b & 153.530 & -0.895 & 0.06 & 58.6 & & 153.607 & -0.872 & 0.057 & 35 & 0.048 & a957 + 313 & h , b & 153.638 & 0.317 & 0.08 & 43.1 & & 153.655 & 0.342 & 0.079 & 11 & 0.071 & + 314 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 153.711 & 0.286 & 0.233 & 13 & 0.271 & + 315 & h , b & 153.889 & -0.866 & 0.17 & 46.8 & & 153.895 & -0.879 & 0.181 & 16 & 0.180 & + 316 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 153.902 & 1.150 & 0.297 & 14 & 0.000 & + 317 & h , b & 153.976 & 0.011 & 0.09 & 41.9 & & 154.015 & 0.047 & 0.109 & 10 & 0.094 & + 318 & h , b & 154.451 & -0.036 & 0.07 & 45.4 & & 154.508 & 0.034 & 0.085 & 17 & 0.071 & + 319 & h , b & 154.705 & 0.385 & 0.06 & 38.8 & & 154.908 & 0.332 & 0.090 & 14 & 0.096 & + 320 & h , b & 154.964 & -0.638 & 0.07 & 32.9 & & 154.934 & -0.638 & 0.087 & 15 & 0.094 & + + 321 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 155.016 & -0.793 & 0.120 & 13 & 0.127 & + 322 & h , b & 155.276 & 0.415 & 0.27 & 48.2 & & 155.277 & 0.478 & 0.319 & 14 & 0.308 & + 323 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 155.707 & -0.054 & 0.286 & 14 & 0.311 & + 324 & h & 156.003 & -0.236 & 0.24 & 48.6 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 325 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 156.288 & -0.565 & 0.157 & 15 & 0.158 & + 326 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 156.464 & -0.588 & 0.146 & 22 & 0.159 & + 327 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 156.481 & -0.443 & 0.179 & 18 & 0.170 & + 328 & h , b & 156.577 & 0.706 & 0.05 & 29.9 & & 156.589 & 0.488 & 0.101 & 13 & 0.098 & + 329 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 157.009 & -0.485 & 0.199 & 21 & 0.000 & + 330 & h , b & 157.131 & -0.731 & 0.18 & 51.6 & & 157.147 & -0.752 & 0.192 & 23 & 0.222 & + + 331 & h , b & 157.383 & -0.472 & 0.19 & 47.3 & & 157.311 & -0.486 & 0.199 & 8 & 0.000 & + 332 & h , b & 157.602 & -1.136 & 0.23 & 43.3 & & 157.675 & -1.170 & 0.231 & 8 & 0.000 & + 333 & h , b & 157.692 & -0.790 & 0.29 & 46.6 & & 157.686 & -0.792 & 0.273 & 8 & 0.000 & + 334 & h & 158.736 & 0.794 & 0.26 & 40.0 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 335 & h , b & 159.355 & -0.685 & 0.11 & 46.3 & & 159.374 & -0.678 & 0.098 & 12 & 0.096 & + 336 & h , b & 159.841 & -1.097 & 0.27 & 42.5 & & 159.837 & -1.104 & 0.334 & 10 & 0.385 & + 337 & h , b & 159.973 & 0.958 & 0.07 & 40.7 & & 159.848 & 1.023 & 0.105 & 12 & 0.111 & + 338 & h , b & 160.264 & -0.729 & 0.10 & 44.0 & & 160.257 & -0.746 & 0.142 & 17 & 0.138 & + 339 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 160.371 & -0.633 & 0.142 & 14 & 0.136 & + 340 & h , b & 160.551 & -0.126 & 0.14 & 43.7 & & 160.601 & -0.138 & 0.129 & 16 & 0.138 & + + 341 & h & 160.607 & -0.955 & 0.28 & 45.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 342 & h , b & 160.814 & 0.846 & 0.14 & 57.4 & & 160.761 & 0.905 & 0.124 & 21 & 0.126 & a1078 + 343 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 160.819 & 0.579 & 0.124 & 21 & 0.100 & + 344 & h , b & 160.892 & -0.126 & 0.29 & 49.7 & & 160.883 & -0.125 & 0.332 & 8 & 0.383 & + 345 & h & 160.953 & -0.934 & 0.18 & 42.4 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 346 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 160.966 & 1.062 & 0.114 & 28 & 0.116 & a1080 + 347 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 161.200 & -0.136 & 0.290 & 15 & 0.000 & + 348 & h , b & 161.514 & -0.158 & 0.27 & 45.8 & & 161.515 & -0.161 & 0.233 & 6 & 0.000 & + 349 & h , b & 161.607 & -1.045 & 0.22 & 40.6 & & 161.612 & -1.041 & 0.229 & 6 & 0.000 & + 350 & h , b & 162.136 & 1.139 & 0.21 & 50.6 & & 162.134 & 1.140 & 0.216 & 6 & 0.247 & + + 351 & h , b & 162.092 & 0.844 & 0.12 & 30.8 & & 162.219 & 0.853 & 0.070 & 14 & 0.038 & + 352 & h , b & 162.471 & 1.171 & 0.05 & 29.8 & & 162.440 & 1.130 & 0.111 & 15 & 0.098 & + 353 & h & 164.514 & 0.403 & 0.09 & 40.3 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 354 & h , b & 164.515 & 0.665 & 0.29 & 46.5 & & 164.472 & 0.663 & 0.338 & 6 & 0.000 & + 355 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 164.593 & 0.338 & 0.273 & 15 & 0.243 & + 356 & h , b & 165.060 & -0.446 & 0.21 & 44.4 & & 165.064 & -0.444 & 0.240 & 9 & 0.247 & + 357 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 165.065 & -0.647 & 0.253 & 15 & 0.253 & + 358 & h , b & 165.240 & -0.774 & 0.29 & 51.4 & & 165.250 & -0.768 & 0.288 & 8 & 0.000 & + 359 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 165.597 & -0.184 & 0.275 & 18 & 0.254 & + 360 & h , b & 165.667 & -1.128 & 0.14 & 42.9 & & 165.656 & -1.112 & 0.157 & 9 & 0.155 & a1148 + + 361 & h , b & 165.722 & -0.042 & 0.28 & 52.7 & & 165.724 & -0.097 & 0.323 & 7 & 0.000 & + 362 & h , b & 165.901 & -0.137 & 0.27 & 40.5 & & 165.947 & -0.193 & 0.342 & 6 & 0.385 & + 363 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 165.930 & 0.484 & 0.268 & 15 & 0.273 & + 364 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 165.964 & -0.350 & 0.266 & 14 & 0.000 & + 365 & h , b & 166.020 & 1.074 & 0.16 & 42.1 & & 165.991 & 1.042 & 0.142 & 6 & 0.153 & + 366 & h , b & 166.065 & -0.237 & 0.30 & 57.8 & & 166.081 & -0.214 & 0.275 & 13 & 0.000 & + 367 & h , b & 166.377 & 0.877 & 0.06 & 38.5 & & 166.355 & 0.938 & 0.114 & 13 & 0.123 & + 368 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 166.611 & -0.592 & 0.290 & 14 & 0.279 & + 369 & h , b & 166.865 & -0.466 & 0.30 & 50.8 & & 166.876 & -0.463 & 0.288 & 12 & 0.270 & + 370 & h , b & 166.952 & 0.772 & 0.20 & 46.2 & & 166.948 & 0.777 & 0.231 & 18 & 0.000 & + + 371 & h & 166.964 & 1.027 & 0.23 & 49.4 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 372 & h , b & 167.019 & 0.255 & 0.18 & 59.8 & & 167.026 & 0.282 & 0.194 & 21 & 0.203 & + 373 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 167.048 & 0.823 & 0.194 & 14 & 0.200 & + 374 & h , b & 167.252 & 0.743 & 0.26 & 56.6 & & 167.268 & 0.759 & 0.314 & 14 & 0.000 & + 375 & h & 167.253 & -0.122 & 0.11 & 41.7 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 376 & h & 167.419 & 0.760 & 0.22 & 42.8 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 377 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 167.629 & -0.479 & 0.231 & 16 & 0.241 & + 378 & h , b & 167.781 & 1.107 & 0.09 & 63.4 & & 167.750 & 1.127 & 0.090 & 17 & 0.097 & a1189 + 379 & h , b & 167.794 & 0.761 & 0.17 & 76.0 & & 167.797 & 0.752 & 0.177 & 27 & 0.185 & a1191 + 380 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 167.882 & 0.891 & 0.177 & 19 & 0.185 & + + 381 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 167.895 & -0.672 & 0.275 & 16 & 0.000 & + 382 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 167.975 & -0.583 & 0.253 & 17 & 0.000 & + 383 & h , b & 168.115 & 0.640 & 0.14 & 40.7 & & 168.125 & 0.641 & 0.190 & 10 & 0.194 & + 384 & h , b & 168.164 & -0.941 & 0.29 & 54.2 & & 168.147 & -0.940 & 0.321 & 14 & 0.000 & + 385 & h , b & 168.400 & -0.339 & 0.05 & 25.9 & & 168.503 & -0.336 & 0.092 & 23 & 0.103 & + 386 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 168.599 & -0.550 & 0.251 & 15 & 0.280 & + 387 & h & 168.807 & 0.408 & 0.26 & 43.7 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 388 & h , b & 169.201 & -0.612 & 0.22 & 52.5 & & 169.177 & -0.574 & 0.188 & 8 & 0.186 & + 389 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 169.376 & -0.596 & 0.290 & 19 & 0.276 & + 390 & h , b & 169.714 & -0.797 & 0.29 & 51.8 & & 169.718 & -0.792 & 0.249 & 16 & 0.000 & + + 391 & h , b & 170.137 & -0.171 & 0.07 & 23.1 & & 170.331 & -0.221 & 0.087 & 18 & 0.101 & + 392 & h , b & 170.239 & -0.374 & 0.11 & 37.2 & & 170.363 & -0.414 & 0.111 & 16 & 0.068 & + 393 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 170.429 & 0.992 & 0.094 & 15 & 0.102 & + 394 & h , b & 170.662 & 0.463 & 0.10 & 58.0 & & 170.673 & 0.652 & 0.092 & 31 & 0.071 & a1236 + 395 & h , b & 170.722 & 1.114 & 0.08 & 58.8 & & 170.716 & 1.113 & 0.081 & 20 & 0.073 & a1238 + 396 & h , b & 170.731 & 0.035 & 0.20 & 45.3 & & 170.717 & 0.033 & 0.216 & 6 & 0.209 & + 397 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 170.826 & -0.232 & 0.260 & 13 & 0.000 & + 398 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 171.019 & 0.917 & 0.096 & 14 & 0.100 & + 399 & h & 171.518 & -0.473 & 0.25 & 42.6 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 400 & h , b & 171.642 & -0.337 & 0.23 & 56.3 & & 171.655 & -0.431 & 0.159 & 7 & 0.145 & + + 401 & h , b & 171.816 & -0.403 & 0.27 & 48.1 & & 171.820 & -0.404 & 0.310 & 17 & 0.333 & + 402 & h & 171.838 & -0.857 & 0.29 & 42.5 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 403 & h , b & 171.874 & 0.147 & 0.11 & 46.5 & & 171.844 & 0.136 & 0.120 & 23 & 0.134 & + 404 & h , b & 172.001 & -0.610 & 0.29 & 43.2 & & 171.988 & -0.615 & 0.336 & 8 & 0.000 & + 405 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 172.086 & -0.540 & 0.295 & 13 & 0.000 & + 406 & h , b & 172.318 & 0.638 & 0.14 & 47.8 & & 172.322 & 0.650 & 0.118 & 9 & 0.126 & + 407 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 172.320 & -0.712 & 0.299 & 19 & 0.282 & + 408 & h , b & 172.560 & 0.429 & 0.24 & 40.3 & & 172.520 & 0.416 & 0.277 & 8 & 0.000 & + 409 & h , b & 172.910 & 0.221 & 0.13 & 42.0 & & 172.996 & 0.195 & 0.131 & 8 & 0.129 & + 410 & h , b & 173.637 & -0.301 & 0.29 & 52.8 & & 173.630 & -0.311 & 0.341 & 18 & 0.341 & + + 411 & h , b & 174.926 & 0.766 & 0.29 & 47.6 & & 174.925 & 0.762 & 0.340 & 21 & 0.357 & + 412 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 175.554 & 0.633 & 0.290 & 16 & 0.306 & + 413 & h , b & 175.674 & 0.482 & 0.27 & 56.5 & & 175.684 & 0.468 & 0.292 & 19 & 0.000 & + 414 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 175.755 & -0.787 & 0.286 & 14 & 0.000 & + 415 & h , b & 175.946 & 0.353 & 0.29 & 64.2 & & 175.947 & 0.337 & 0.238 & 10 & 0.000 & + 416 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 176.238 & 0.061 & 0.122 & 13 & 0.093 & + 417 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 176.252 & -0.566 & 0.281 & 14 & 0.259 & + 418 & h & 176.273 & -0.561 & 0.18 & 42.5 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 419 & h , b & 176.316 & -0.834 & 0.20 & 43.7 & & 176.244 & -0.823 & 0.135 & 7 & 0.156 & + 420 & h , b & 176.419 & 0.397 & 0.26 & 47.1 & & 176.425 & 0.393 & 0.275 & 14 & 0.261 & + + 421 & h , b & 176.545 & -1.096 & 0.12 & 55.1 & & 176.534 & -1.121 & 0.146 & 12 & 0.119 & a1376 + 422 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 176.996 & 1.077 & 0.247 & 15 & 0.233 & + 423 & h , b & 177.585 & -0.594 & 0.11 & 55.0 & & 177.578 & -0.609 & 0.135 & 28 & 0.138 & a1392 + 424 & h , b & 177.995 & -0.409 & 0.19 & 45.1 & & 177.988 & -0.391 & 0.244 & 6 & 0.257 & + 425 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 178.084 & -0.508 & 0.236 & 13 & 0.244 & + 426 & h , b & 178.570 & -0.125 & 0.23 & 50.5 & & 178.566 & -0.127 & 0.271 & 19 & 0.000 & + 427 & h , b & 178.774 & -0.580 & 0.14 & 50.9 & & 178.807 & -0.526 & 0.140 & 17 & 0.132 & a1411 + 428 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 178.824 & -0.332 & 0.216 & 14 & 0.244 & + 429 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 179.003 & 0.004 & 0.098 & 15 & 0.106 & + 430 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 179.012 & -0.495 & 0.264 & 13 & 0.000 & + + 431 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 179.041 & -0.326 & 0.253 & 28 & 0.260 & + 432 & h , b & 179.065 & -0.206 & 0.09 & 49.4 & & 179.097 & -0.364 & 0.109 & 16 & 0.106 & a1419 + 433 & h , b & 179.306 & -0.486 & 0.22 & 42.2 & & 179.276 & -0.463 & 0.236 & 12 & 0.000 & + 434 & h , b & 179.551 & -1.047 & 0.11 & 47.5 & & 179.671 & -1.067 & 0.111 & 17 & 0.131 & + 435 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 179.603 & -0.067 & 0.251 & 14 & 0.000 & + 436 & h , b & 179.432 & 0.021 & 0.11 & 33.8 & & 179.603 & -0.074 & 0.109 & 14 & 0.107 & + 437 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 179.804 & 0.516 & 0.194 & 16 & 0.174 & + 438 & h , b & 179.809 & -0.053 & 0.22 & 41.3 & & 179.781 & -0.023 & 0.225 & 8 & 0.000 & + 439 & h , b & 179.971 & 1.114 & 0.17 & 42.3 & & 179.957 & 1.079 & 0.192 & 12 & 0.200 & + 440 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 180.041 & -0.597 & 0.166 & 16 & 0.169 & + + 441 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 180.186 & -0.156 & 0.168 & 27 & 0.166 & + 442 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 180.194 & -0.021 & 0.186 & 14 & 0.165 & + 443 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 180.363 & -0.447 & 0.148 & 18 & 0.169 & + 444 & h & 180.411 & -0.489 & 0.23 & 70.0 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 445 & h , b & 180.434 & -0.186 & 0.21 & 71.4 & & 180.420 & -0.202 & 0.162 & 29 & 0.173 & a1445 + 446 & h , b & 180.453 & 0.635 & 0.30 & 49.9 & & 180.417 & 0.701 & 0.260 & 9 & 0.000 & + 447 & h , b & 180.596 & 0.772 & 0.23 & 47.9 & & 180.627 & 0.713 & 0.229 & 10 & 0.000 & + 448 & h & 180.924 & 1.037 & 0.17 & 44.6 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 449 & h , b & 180.984 & 0.484 & 0.20 & 42.7 & & 180.986 & 0.481 & 0.207 & 10 & 0.210 & + 450 & h & 181.654 & 0.300 & 0.27 & 45.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + + 451 & h , b & 181.664 & 1.201 & 0.27 & 43.9 & & 181.664 & 1.215 & 0.268 & 10 & 0.253 & + 452 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 181.686 & 0.529 & 0.262 & 15 & 0.280 & + 453 & h & 181.848 & 1.141 & 0.14 & 41.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 454 & h , b & 181.872 & 0.238 & 0.23 & 43.7 & & 181.877 & 0.223 & 0.297 & 8 & 0.000 & + 455 & h , b & 182.468 & -0.564 & 0.14 & 55.2 & & 182.478 & -0.558 & 0.181 & 24 & 0.181 & + 456 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 182.642 & -0.412 & 0.186 & 16 & 0.181 & + 457 & h , b & 182.652 & 1.076 & 0.30 & 57.3 & & 182.615 & 1.064 & 0.284 & 10 & 0.294 & + 458 & h & 182.695 & 1.205 & 0.29 & 46.0 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 459 & h , b & 182.833 & 1.181 & 0.25 & 44.4 & & 182.831 & 1.086 & 0.188 & 7 & 0.000 & + 460 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 183.371 & -0.113 & 0.295 & 15 & 0.000 & + + 461 & h , b & 183.430 & -0.576 & 0.26 & 54.2 & & 183.428 & -0.634 & 0.260 & 16 & 0.000 & + 462 & h , b & 183.568 & -0.442 & 0.21 & 49.8 & & 183.564 & -0.449 & 0.260 & 19 & 0.245 & + 463 & h , b & 183.520 & -0.895 & 0.08 & 30.5 & & 183.607 & -0.948 & 0.085 & 14 & 0.082 & + 464 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 183.643 & 0.791 & 0.223 & 17 & 0.251 & + 465 & h , b & 183.928 & 1.167 & 0.29 & 51.0 & & 183.922 & 1.127 & 0.212 & 7 & 0.000 & + 466 & h , b & 183.990 & 0.718 & 0.23 & 40.6 & & 183.959 & 0.724 & 0.236 & 16 & 0.000 & + 467 & h , b & 184.107 & -0.069 & 0.22 & 48.2 & & 184.103 & -0.080 & 0.299 & 24 & 0.276 & + 468 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 184.125 & 0.692 & 0.299 & 20 & 0.306 & + 469 & h , b & 184.138 & -0.776 & 0.15 & 41.9 & & 184.205 & -0.742 & 0.081 & 12 & 0.071 & + 470 & h , b & 184.387 & 0.396 & 0.26 & 53.6 & & 184.428 & 0.339 & 0.290 & 17 & 0.000 & + + 471 & h & 184.415 & -0.525 & 0.23 & 46.4 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 472 & h , b & 184.486 & -0.848 & 0.09 & 43.1 & & 184.466 & -0.905 & 0.138 & 19 & 0.000 & + 473 & h , b & 184.685 & -1.037 & 0.14 & 53.3 & & 184.707 & -1.047 & 0.116 & 13 & 0.116 & + 474 & h & 184.755 & -0.735 & 0.21 & 51.5 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 475 & h & 184.924 & 1.045 & 0.28 & 43.9 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 476 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 184.937 & 0.216 & 0.260 & 15 & 0.000 & + 477 & h , b & 185.081 & -0.043 & 0.29 & 47.5 & & 185.117 & -0.059 & 0.312 & 8 & 0.000 & + 478 & h , b & 185.253 & 1.143 & 0.21 & 50.1 & & 185.235 & 1.036 & 0.177 & 7 & 0.000 & + 479 & h , b & 185.387 & 0.222 & 0.16 & 67.6 & & 185.430 & 0.329 & 0.157 & 25 & 0.159 & + 480 & h , b & 185.442 & -0.409 & 0.25 & 51.2 & & 185.443 & -0.412 & 0.268 & 19 & 0.000 & + + 481 & h , b & 185.497 & -1.137 & 0.23 & 79.0 & & 185.542 & -1.196 & 0.225 & 22 & 0.259 & a1525 + 482 & h & 185.668 & -0.941 & 0.18 & 43.7 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 483 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 185.797 & 0.515 & 0.297 & 13 & 0.000 & + 484 & h , b & 186.073 & -0.509 & 0.22 & 46.6 & & 186.072 & -0.519 & 0.251 & 11 & 0.000 & + 485 & h , b & 186.100 & -0.398 & 0.18 & 40.8 & & 186.183 & -0.346 & 0.159 & 19 & 0.158 & + 486 & h , b & 186.141 & 0.936 & 0.21 & 62.4 & & 186.119 & 0.927 & 0.238 & 22 & 0.000 & a1533 + 487 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 186.155 & 0.220 & 0.275 & 13 & 0.263 & + 488 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 186.191 & -0.431 & 0.273 & 15 & 0.257 & + 489 & h , b & 186.299 & -0.442 & 0.22 & 63.1 & & 186.354 & -0.554 & 0.155 & 8 & 0.155 & + 490 & h , b & 186.302 & -0.013 & 0.23 & 40.2 & & 186.248 & -0.070 & 0.275 & 13 & 0.000 & + + 491 & h , b & 186.359 & 0.704 & 0.22 & 60.3 & & 186.364 & 0.710 & 0.240 & 18 & 0.236 & + 492 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 186.533 & -0.604 & 0.168 & 17 & 0.157 & + 493 & h , b & 186.706 & -0.645 & 0.17 & 42.0 & & 186.687 & -0.622 & 0.159 & 25 & 0.159 & + 494 & h & 186.830 & 1.135 & 0.21 & 41.2 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 495 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 187.345 & -0.012 & 0.295 & 13 & 0.303 & + 496 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 187.742 & -1.198 & 0.292 & 13 & 0.000 & + 497 & h , b & 187.854 & 0.143 & 0.20 & 40.2 & & 187.860 & 0.142 & 0.138 & 10 & 0.135 & + 498 & h , b & 188.078 & -0.707 & 0.27 & 68.1 & & 188.082 & -0.710 & 0.196 & 17 & 0.192 & + 499 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 188.415 & 0.326 & 0.181 & 13 & 0.000 & + 500 & h , b & 188.571 & 0.660 & 0.23 & 50.6 & & 188.598 & 0.573 & 0.199 & 8 & 0.205 & + + 501 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 188.738 & -0.794 & 0.255 & 14 & 0.248 & + 502 & h & 188.746 & -0.815 & 0.17 & 53.8 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 503 & h & 188.977 & 1.200 & 0.28 & 45.6 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 504 & h , b & 189.259 & -0.810 & 0.18 & 43.6 & & 189.289 & -0.814 & 0.216 & 12 & 0.234 & + 505 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 189.325 & 1.096 & 0.255 & 14 & 0.242 & + 506 & h , b & 189.432 & -0.674 & 0.16 & 44.4 & & 189.414 & -0.623 & 0.140 & 9 & 0.138 & + 507 & h , b & 189.479 & -0.280 & 0.14 & 64.2 & & 189.481 & -0.275 & 0.142 & 32 & 0.139 & a1577 + 508 & h , b & 189.655 & -0.860 & 0.17 & 42.9 & & 189.653 & -0.858 & 0.233 & 13 & 0.231 & + 509 & h , b & 189.773 & 0.597 & 0.26 & 40.4 & & 189.708 & 0.619 & 0.323 & 10 & 0.344 & + 510 & h , b & 189.842 & 0.171 & 0.22 & 42.1 & & 189.847 & 0.169 & 0.247 & 15 & 0.249 & + + 511 & h , b & 189.943 & -1.164 & 0.24 & 42.0 & & 189.921 & -1.226 & 0.170 & 6 & 0.188 & + 512 & h , b & 190.040 & 0.555 & 0.25 & 42.3 & & 190.104 & 0.583 & 0.203 & 7 & 0.216 & + 513 & h , b & 190.185 & 1.186 & 0.25 & 41.3 & & 190.243 & 1.117 & 0.207 & 7 & 0.000 & + 514 & h , b & 190.660 & 0.119 & 0.27 & 83.8 & & 190.627 & 0.177 & 0.329 & 16 & 0.000 & + 515 & h , b & 190.682 & -1.224 & 0.12 & 44.1 & & 190.684 & -1.232 & 0.153 & 23 & 0.172 & + 516 & h , b & 190.781 & -0.757 & 0.13 & 44.6 & & 190.790 & -0.784 & 0.131 & 11 & 0.143 & + 517 & h , b & 190.896 & -1.230 & 0.14 & 36.8 & & 190.858 & -1.177 & 0.157 & 31 & 0.168 & + 518 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 190.911 & 0.547 & 0.074 & 18 & 0.064 & + 519 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 191.082 & 0.747 & 0.081 & 20 & 0.082 & + 520 & h , b & 191.195 & -1.016 & 0.11 & 39.2 & & 191.192 & -1.020 & 0.157 & 19 & 0.147 & + + 521 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 191.573 & -0.283 & 0.133 & 14 & 0.127 & + 522 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 191.643 & -0.228 & 0.249 & 14 & 0.242 & + 523 & h , b & 191.720 & 0.306 & 0.05 & 43.1 & & 191.713 & 0.297 & 0.096 & 32 & 0.089 & + 524 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 191.927 & -0.137 & 0.111 & 24 & 0.091 & + 525 & h , b & 191.996 & -0.349 & 0.06 & 24.2 & & 192.071 & -0.261 & 0.087 & 19 & 0.089 & + 526 & h , b & 192.391 & -0.763 & 0.17 & 41.0 & & 192.390 & -0.765 & 0.201 & 14 & 0.194 & + 527 & h , b & 193.413 & 0.067 & 0.29 & 46.7 & & 193.450 & 0.101 & 0.358 & 9 & 0.000 & + 528 & h & 194.003 & 0.466 & 0.27 & 44.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 529 & h & 194.113 & -0.209 & 0.29 & 44.2 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 530 & h , b & 195.172 & 1.052 & 0.09 & 46.3 & & 195.169 & 1.069 & 0.090 & 19 & 0.082 & + + 531 & h & 195.582 & 0.233 & 0.20 & 43.9 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 532 & h , b & 195.818 & -0.757 & 0.23 & 55.1 & & 195.900 & -0.821 & 0.186 & 10 & 0.208 & + 533 & h , b & 195.974 & 0.074 & 0.25 & 45.3 & & 195.956 & 0.138 & 0.292 & 8 & 0.000 & + 534 & h & 196.455 & -0.589 & 0.29 & 49.0 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 535 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 196.591 & -0.828 & 0.186 & 13 & 0.190 & + 536 & h & 196.780 & 1.204 & 0.25 & 50.4 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 537 & h , b & 196.963 & 1.185 & 0.29 & 47.7 & & 196.977 & 1.155 & 0.251 & 6 & 0.000 & + 538 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 197.659 & -0.657 & 0.087 & 14 & 0.086 & + 539 & h & 197.826 & -1.234 & 0.20 & 49.3 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 540 & h , b & 197.830 & 0.070 & 0.16 & 42.1 & & 197.770 & -0.046 & 0.094 & 8 & 0.096 & + + 541 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 197.903 & -0.483 & 0.096 & 13 & 0.090 & + 542 & h , b & 198.049 & -0.988 & 0.09 & 54.0 & & 198.057 & -0.975 & 0.090 & 30 & 0.083 & a1692 + 543 & h , b & 198.223 & 1.016 & 0.07 & 46.1 & & 198.228 & 1.036 & 0.063 & 8 & 0.072 & + 544 & h , b & 198.403 & 0.563 & 0.31 & 60.3 & & 198.395 & 0.567 & 0.251 & 14 & 0.000 & + 545 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 198.576 & 0.363 & 0.251 & 13 & 0.264 & + 546 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 198.638 & -0.466 & 0.271 & 14 & 0.273 & + 547 & h , b & 198.676 & 0.051 & 0.20 & 41.8 & & 198.658 & 0.156 & 0.268 & 7 & 0.000 & + 548 & h , b & 198.706 & 1.046 & 0.29 & 44.0 & & 198.698 & 1.034 & 0.321 & 8 & 0.000 & + 549 & h & 198.709 & 1.177 & 0.26 & 40.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 550 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 198.777 & -0.677 & 0.279 & 13 & 0.292 & + + 551 & h , b & 198.897 & 1.100 & 0.11 & 29.3 & & 198.848 & 0.959 & 0.148 & 14 & 0.142 & + 552 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 198.887 & -0.645 & 0.277 & 16 & 0.000 & + 553 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 198.916 & -1.075 & 0.275 & 14 & 0.000 & + 554 & h , b & 199.049 & -0.930 & 0.10 & 48.5 & & 199.057 & -0.913 & 0.090 & 7 & 0.111 & + 555 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 199.078 & -0.948 & 0.249 & 19 & 0.000 & + 556 & h , b & 199.106 & 0.863 & 0.08 & 41.0 & & 199.136 & 0.870 & 0.090 & 11 & 0.080 & + 557 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 199.130 & -1.087 & 0.275 & 19 & 0.274 & + 558 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 199.298 & -0.923 & 0.279 & 14 & 0.290 & + 559 & h , b & 199.475 & 1.085 & 0.14 & 41.3 & & 199.510 & 1.114 & 0.077 & 9 & 0.110 & + 560 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 199.528 & -1.179 & 0.203 & 16 & 0.000 & + + 561 & h , b & 199.661 & -0.810 & 0.07 & 38.6 & & 199.557 & -0.626 & 0.094 & 18 & 0.110 & + 562 & h , b & 199.702 & 0.692 & 0.16 & 46.2 & & 199.697 & 0.680 & 0.209 & 11 & 0.220 & + 563 & h , b & 199.813 & -0.932 & 0.06 & 59.3 & & 199.820 & -0.995 & 0.068 & 27 & 0.082 & + 564 & h , b & 200.031 & -0.264 & 0.23 & 42.8 & & 200.046 & -0.269 & 0.303 & 14 & 0.284 & + 565 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 200.094 & -0.402 & 0.220 & 14 & 0.000 & + 566 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 200.218 & -0.516 & 0.209 & 14 & 0.000 & + 567 & h , b & 200.152 & 1.237 & 0.09 & 30.8 & & 200.224 & 1.220 & 0.142 & 13 & 0.148 & + 568 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 200.329 & -1.150 & 0.236 & 15 & 0.224 & + 569 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 200.356 & -0.693 & 0.109 & 17 & 0.108 & + 570 & h , b & 200.374 & -1.031 & 0.13 & 40.2 & & 200.228 & -0.941 & 0.063 & 14 & 0.048 & + + 571 & h , b & 200.936 & 1.052 & 0.09 & 45.5 & & 200.952 & 1.113 & 0.109 & 23 & 0.105 & + 572 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 201.068 & 0.808 & 0.092 & 13 & 0.108 & + 573 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 201.128 & 1.018 & 0.286 & 13 & 0.000 & + 574 & h & 201.131 & -1.195 & 0.29 & 45.5 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 575 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 201.478 & -0.306 & 0.181 & 13 & 0.189 & + 576 & h , b & 201.515 & -0.427 & 0.18 & 53.1 & & 201.517 & -0.442 & 0.183 & 21 & 0.181 & + 577 & h & 201.569 & 0.229 & 0.13 & 44.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & rxc 1326.2 + 578 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 201.602 & 0.036 & 0.098 & 13 & 0.083 & ( rxc 1326.2 ) + 579 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 201.899 & -0.945 & 0.183 & 23 & 0.184 & + 580 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 202.046 & -1.005 & 0.214 & 15 & 0.200 & + + 581 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 202.084 & 0.438 & 0.244 & 15 & 0.260 & + 582 & h , b & 202.086 & 0.436 & 0.13 & 43.0 & & 202.093 & 0.472 & 0.148 & 6 & 0.000 & + 583 & h , b & 202.605 & 0.735 & 0.21 & 51.6 & & 202.613 & 0.748 & 0.192 & 10 & 0.217 & + 584 & h & 202.691 & 1.197 & 0.16 & 48.3 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 585 & h , b & 202.801 & 1.005 & 0.32 & 79.7 & & 202.796 & 1.083 & 0.299 & 15 & 0.000 & + 586 & h & 202.833 & -1.168 & 0.29 & 45.2 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 587 & h & 203.115 & 1.153 & 0.09 & 40.9 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 588 & h , b & 203.505 & 1.196 & 0.26 & 59.8 & & 203.511 & 1.216 & 0.233 & 15 & 0.248 & + 589 & h , b & 203.536 & 0.798 & 0.26 & 47.5 & & 203.483 & 0.728 & 0.253 & 9 & 0.000 & + 590 & h , b & 203.573 & -0.951 & 0.15 & 45.7 & & 203.682 & -0.898 & 0.083 & 10 & 0.000 & + + 591 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 203.577 & -0.957 & 0.262 & 17 & 0.270 & + 592 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 203.596 & -0.216 & 0.271 & 21 & 0.267 & + 593 & h , b & 203.601 & -0.299 & 0.34 & 59.8 & & 203.596 & -0.310 & 0.292 & 20 & 0.245 & + 594 & h & 203.605 & 0.994 & 0.29 & 62.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 595 & h , b & 203.640 & 0.352 & 0.27 & 46.5 & & 203.659 & 0.303 & 0.332 & 10 & 0.000 & + 596 & h , b & 204.201 & -0.215 & 0.25 & 41.6 & & 204.217 & -0.144 & 0.251 & 6 & 0.000 & + 597 & h , b & 204.390 & 0.796 & 0.29 & 43.0 & & 204.393 & 0.769 & 0.255 & 9 & 0.000 & + 598 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 204.515 & -0.317 & 0.236 & 14 & 0.233 & + 599 & h , b & 204.591 & 0.035 & 0.26 & 47.1 & & 204.630 & -0.042 & 0.251 & 12 & 0.249 & + 600 & h , b & 204.602 & 0.252 & 0.28 & 47.6 & & 204.578 & 0.301 & 0.227 & 13 & 0.000 & + + 601 & h , b & 204.887 & -0.283 & 0.14 & 56.8 & & 204.859 & -0.267 & 0.140 & 17 & 0.146 & + 602 & h , b & 205.064 & 0.032 & 0.14 & 46.9 & & 205.074 & 0.032 & 0.144 & 12 & 0.145 & + 603 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 205.122 & -0.212 & 0.268 & 15 & 0.000 & + 604 & h & 205.137 & -0.616 & 0.20 & 55.5 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 605 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 205.155 & -0.633 & 0.297 & 19 & 0.285 & + 606 & h , b & 205.350 & -1.013 & 0.23 & 42.5 & & 205.356 & -1.042 & 0.286 & 14 & 0.287 & + 607 & h & 205.853 & -0.190 & 0.20 & 40.2 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 608 & h , b & 205.949 & 0.973 & 0.17 & 52.4 & & 205.943 & 0.979 & 0.233 & 13 & 0.229 & + 609 & h , b & 205.958 & 0.757 & 0.16 & 41.7 & & 205.984 & 0.722 & 0.231 & 6 & 0.000 & + 610 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 206.154 & -0.120 & 0.290 & 14 & 0.000 & + + 611 & h , b & 206.291 & 0.123 & 0.05 & 40.7 & & 206.377 & 0.343 & 0.096 & 13 & 0.089 & + 612 & h , b & 206.410 & -0.160 & 0.08 & 30.9 & & 206.315 & -0.142 & 0.083 & 15 & 0.087 & + 613 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 206.345 & 0.206 & 0.271 & 20 & 0.256 & + 614 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 206.451 & 0.226 & 0.271 & 22 & 0.276 & + 615 & h , b & 206.633 & 0.129 & 0.27 & 55.5 & & 206.639 & 0.117 & 0.249 & 7 & 0.000 & + 616 & h & 207.483 & 0.625 & 0.18 & 45.6 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 617 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 207.640 & -0.412 & 0.299 & 14 & 0.360 & + 618 & h & 207.780 & -0.515 & 0.28 & 47.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 619 & h & 207.901 & 0.110 & 0.30 & 44.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 620 & h & 208.005 & 0.435 & 0.26 & 48.5 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + + 621 & h , b & 208.169 & -1.050 & 0.12 & 35.9 & & 208.193 & -1.027 & 0.129 & 13 & 0.151 & + 622 & h , b & 208.178 & 0.134 & 0.11 & 30.7 & & 208.359 & 0.047 & 0.107 & 15 & 0.116 & + 623 & h , b & 208.414 & 0.965 & 0.24 & 41.8 & & 208.415 & 0.969 & 0.303 & 9 & 0.000 & + 624 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 208.508 & 0.242 & 0.229 & 14 & 0.251 & + 625 & h , b & 208.591 & -1.025 & 0.14 & 58.8 & & 208.600 & -1.044 & 0.146 & 21 & 0.146 & + 626 & h , b & 209.250 & -0.904 & 0.21 & 42.6 & & 209.245 & -0.894 & 0.172 & 11 & 0.196 & + 627 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 209.426 & 1.003 & 0.162 & 13 & 0.168 & + 628 & h , b & 209.975 & -1.127 & 0.24 & 42.0 & & 209.971 & -1.192 & 0.297 & 6 & 0.000 & + 629 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 210.024 & 0.361 & 0.164 & 14 & 0.167 & + 630 & h , b & 210.058 & 0.245 & 0.15 & 42.5 & & 210.064 & 0.230 & 0.181 & 14 & 0.188 & + + 631 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 210.065 & 0.280 & 0.244 & 13 & 0.253 & + 632 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 210.091 & -0.149 & 0.170 & 13 & 0.190 & + 633 & h & 210.168 & -1.023 & 0.28 & 45.0 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 634 & h , b & 210.184 & -0.515 & 0.18 & 47.1 & & 210.128 & -0.464 & 0.177 & 12 & 0.132 & + 635 & h , b & 210.382 & 0.056 & 0.19 & 49.3 & & 210.380 & 0.058 & 0.168 & 12 & 0.188 & + 636 & h , b & 210.719 & 1.231 & 0.23 & 47.0 & & 210.775 & 1.216 & 0.253 & 16 & 0.255 & + 637 & h , b & 210.769 & -0.237 & 0.29 & 60.0 & & 210.772 & -0.192 & 0.297 & 6 & 0.000 & + 638 & h , b & 210.789 & 0.415 & 0.20 & 43.0 & & 210.787 & 0.413 & 0.181 & 16 & 0.182 & + 639 & h , b & 210.829 & 0.096 & 0.15 & 51.0 & & 210.873 & 0.112 & 0.162 & 19 & 0.189 & + 640 & h , b & 210.964 & -0.519 & 0.19 & 42.6 & & 210.915 & -0.446 & 0.177 & 13 & 0.000 & + + 641 & h , b & 211.284 & -0.149 & 0.25 & 53.6 & & 211.310 & -0.129 & 0.247 & 12 & 0.000 & + 642 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 211.774 & 0.258 & 0.277 & 13 & 0.258 & + 643 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 212.733 & 0.613 & 0.155 & 23 & 0.178 & + 644 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 212.946 & 0.406 & 0.251 & 13 & 0.262 & + 645 & h , b & 213.086 & 0.060 & 0.12 & 52.5 & & 213.091 & 0.077 & 0.105 & 15 & 0.127 & + 646 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 213.099 & 0.640 & 0.205 & 19 & 0.262 & + 647 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 213.191 & 0.783 & 0.227 & 15 & 0.000 & + 648 & h & 213.265 & 0.356 & 0.30 & 44.8 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 649 & h , b & 213.460 & 0.855 & 0.29 & 45.1 & & 213.452 & 0.805 & 0.273 & 9 & 0.000 & + 650 & h , b & 213.616 & -0.376 & 0.10 & 87.2 & & 213.740 & -0.350 & 0.144 & 66 & 0.139 & a1882 + + 651 & h , b & 213.815 & 0.184 & 0.07 & 26.9 & & 213.736 & 0.206 & 0.129 & 18 & 0.127 & + 652 & h , b & 213.778 & 1.015 & 0.27 & 45.2 & & 213.837 & 0.994 & 0.260 & 7 & 0.000 & + 653 & h , b & 213.906 & 1.022 & 0.26 & 42.0 & & 213.837 & 0.994 & 0.260 & 7 & 0.000 & + 654 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 213.944 & -0.992 & 0.122 & 14 & 0.149 & + 655 & h , b & 213.963 & 0.256 & 0.13 & 49.4 & & 213.954 & 0.260 & 0.129 & 15 & 0.126 & rxc 1415.8 + 656 & h , b & 214.025 & -0.723 & 0.05 & 20.3 & & 214.022 & -0.586 & 0.114 & 20 & 0.000 & + 657 & h , b & 214.096 & 0.765 & 0.23 & 41.7 & & 214.082 & 0.812 & 0.301 & 7 & 0.000 & + 658 & h , b & 214.197 & 0.602 & 0.26 & 42.1 & & 214.196 & 0.607 & 0.292 & 11 & 0.000 & + 659 & h , b & 214.391 & -0.495 & 0.11 & 40.7 & & 214.477 & -0.521 & 0.116 & 10 & 0.148 & + 660 & h , b & 214.620 & 0.791 & 0.28 & 47.1 & & 214.618 & 0.785 & 0.341 & 27 & 0.343 & + + 661 & h , b & 214.914 & -0.450 & 0.10 & 41.8 & & 214.796 & -0.469 & 0.120 & 11 & 0.134 & + 662 & h , b & 215.017 & 0.997 & 0.16 & 51.7 & & 215.036 & 0.990 & 0.157 & 9 & 0.171 & + 663 & h , b & 215.111 & -0.103 & 0.20 & 50.3 & & 215.117 & -0.109 & 0.262 & 13 & 0.000 & + 664 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 215.257 & 0.196 & 0.275 & 17 & 0.278 & + 665 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 215.310 & -0.035 & 0.238 & 14 & 0.000 & + 666 & h , b & 215.403 & -0.346 & 0.06 & 44.6 & & 215.423 & -0.338 & 0.068 & 7 & 0.052 & + 667 & h , b & 215.410 & 0.848 & 0.14 & 52.2 & & 215.417 & 0.780 & 0.142 & 11 & 0.146 & + 668 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 215.831 & -0.539 & 0.299 & 20 & 0.278 & + 669 & h , b & 216.038 & 0.517 & 0.19 & 49.7 & & 216.014 & 0.499 & 0.133 & 6 & 0.125 & + 670 & h & 216.078 & -0.211 & 0.15 & 45.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + + 671 & h , b & 216.103 & -0.682 & 0.16 & 41.0 & & 216.111 & -0.685 & 0.201 & 6 & 0.176 & + 672 & h , b & 216.241 & 1.155 & 0.23 & 57.3 & & 216.243 & 1.154 & 0.238 & 6 & 0.304 & + 673 & h , b & 216.318 & -1.050 & 0.09 & 41.0 & & 216.281 & -1.128 & 0.127 & 6 & 0.134 & + 674 & h , b & 216.376 & 0.340 & 0.11 & 45.8 & & 216.365 & 0.338 & 0.140 & 15 & 0.134 & + 675 & h , b & 216.624 & 0.839 & 0.11 & 51.6 & & 216.568 & 0.838 & 0.138 & 12 & 0.125 & + 676 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 216.639 & -0.109 & 0.271 & 15 & 0.000 & + 677 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 216.643 & 1.103 & 0.288 & 15 & 0.281 & + 678 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 216.841 & 0.946 & 0.299 & 20 & 0.291 & + 679 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 216.843 & -0.202 & 0.284 & 15 & 0.000 & + 680 & h , b & 216.849 & -0.200 & 0.19 & 60.6 & & 216.754 & -0.175 & 0.242 & 13 & 0.000 & + + 681 & h , b & 216.976 & 0.479 & 0.23 & 44.3 & & 216.979 & 0.453 & 0.203 & 6 & 0.000 & + 682 & h & 216.980 & -1.234 & 0.17 & 42.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 683 & h , b & 217.073 & 0.730 & 0.23 & 44.7 & & 217.190 & 0.708 & 0.159 & 8 & 0.103 & + 684 & h , b & 217.210 & -0.357 & 0.26 & 51.7 & & 217.203 & -0.406 & 0.312 & 15 & 0.279 & + 685 & h , b & 217.669 & 0.245 & 0.05 & 36.2 & & 217.419 & 0.366 & 0.061 & 18 & 0.056 & + 686 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 217.644 & 1.069 & 0.131 & 14 & 0.132 & + 687 & h , b & 217.688 & 0.802 & 0.18 & 63.9 & & 217.673 & 0.825 & 0.188 & 7 & 0.205 & + 688 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 217.745 & 0.405 & 0.133 & 15 & 0.131 & + 689 & h & 217.956 & -0.289 & 0.27 & 48.0 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 690 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 218.347 & -0.794 & 0.214 & 14 & 0.229 & + + 691 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 218.439 & -0.632 & 0.227 & 17 & 0.000 & + 692 & h , b & 218.597 & -0.615 & 0.20 & 54.2 & & 218.629 & -0.572 & 0.205 & 10 & 0.221 & + 693 & h , b & 218.789 & -0.293 & 0.20 & 42.5 & & 218.797 & -0.293 & 0.216 & 14 & 0.221 & + 694 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 218.895 & -0.709 & 0.251 & 14 & 0.257 & + 695 & h , b & 219.055 & -0.251 & 0.18 & 66.2 & & 219.057 & -0.269 & 0.207 & 16 & 0.219 & + 696 & h , b & 219.168 & 1.164 & 0.30 & 46.2 & & 219.134 & 1.125 & 0.329 & 7 & 0.000 & + 697 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 219.212 & -0.700 & 0.218 & 16 & 0.215 & + 698 & h , b & 219.241 & -1.072 & 0.10 & 43.3 & & 219.084 & -1.130 & 0.090 & 8 & 0.106 & + 699 & h , b & 219.396 & -0.333 & 0.13 & 58.9 & & 219.371 & -0.267 & 0.144 & 26 & 0.135 & a1938 + 700 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 219.489 & -0.708 & 0.271 & 13 & 0.000 & + + 701 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 219.566 & 1.013 & 0.227 & 13 & 0.260 & + 702 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 219.662 & -0.595 & 0.266 & 13 & 0.000 & + 703 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 219.762 & -0.188 & 0.177 & 13 & 0.181 & + 704 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 219.775 & -0.581 & 0.281 & 17 & 0.298 & + 705 & h , b & 219.888 & 0.539 & 0.12 & 46.3 & & 219.917 & 0.498 & 0.146 & 8 & 0.138 & + 706 & h , b & 219.947 & -0.130 & 0.22 & 42.9 & & 220.030 & -0.104 & 0.225 & 7 & 0.000 & + 707 & h , b & 220.040 & 0.717 & 0.21 & 42.3 & & 220.046 & 0.724 & 0.249 & 7 & 0.259 & + 708 & h , b & 220.324 & -0.984 & 0.17 & 40.5 & & 220.324 & -0.988 & 0.148 & 10 & 0.184 & + 709 & h , b & 220.542 & -0.748 & 0.11 & 37.0 & & 220.436 & -0.827 & 0.148 & 20 & 0.143 & + 710 & h , b & 220.440 & 0.559 & 0.27 & 48.1 & & 220.476 & 0.600 & 0.340 & 13 & 0.321 & + + 711 & h , b & 220.826 & -0.667 & 0.14 & 39.6 & & 220.862 & -0.721 & 0.144 & 14 & 0.150 & + 712 & h & 220.866 & -0.330 & 0.19 & 45.9 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 713 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 220.897 & -0.352 & 0.275 & 16 & 0.291 & + 714 & h & 220.989 & -1.141 & 0.22 & 40.6 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 715 & h , b & 220.962 & -0.954 & 0.19 & 37.3 & & 221.009 & -0.947 & 0.133 & 15 & 0.149 & + 716 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 221.038 & 0.178 & 0.299 & 33 & 0.296 & + 717 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 221.227 & 0.142 & 0.277 & 25 & 0.000 & + 718 & h & 221.236 & -1.173 & 0.23 & 46.8 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 719 & h , b & 221.328 & 0.115 & 0.29 & 70.3 & & 221.334 & 0.115 & 0.297 & 22 & 0.294 & + 720 & h , b & 221.632 & 0.739 & 0.20 & 40.2 & & 221.750 & 0.752 & 0.162 & 15 & 0.171 & + + 721 & h , b & 221.649 & -1.023 & 0.26 & 52.2 & & 221.651 & -1.015 & 0.183 & 10 & 0.000 & + 722 & h , b & 221.856 & -1.210 & 0.19 & 47.4 & & 221.848 & -1.196 & 0.138 & 13 & 0.145 & + 723 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 222.918 & -0.779 & 0.297 & 14 & 0.000 & + 724 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 223.554 & -0.237 & 0.238 & 16 & 0.000 & + 725 & h , b & 224.130 & 0.424 & 0.29 & 48.3 & & 224.130 & 0.431 & 0.295 & 7 & 0.000 & + 726 & h , b & 224.282 & 0.119 & 0.21 & 48.0 & & 224.283 & 0.121 & 0.251 & 18 & 0.000 & + 727 & h , b & 224.889 & -0.019 & 0.29 & 52.1 & & 224.896 & -0.006 & 0.255 & 13 & 0.000 & + 728 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 225.001 & -0.025 & 0.244 & 13 & 0.000 & + 729 & h & 225.070 & -1.225 & 0.26 & 41.9 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 730 & h , b & 225.983 & -0.129 & 0.22 & 43.1 & & 225.984 & -0.128 & 0.231 & 20 & 0.000 & + + 731 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 226.090 & -0.090 & 0.231 & 18 & 0.000 & + 732 & h , b & 226.419 & 0.657 & 0.19 & 41.8 & & 226.361 & 0.617 & 0.229 & 19 & 0.225 & + 733 & h , b & 226.594 & 0.464 & 0.22 & 40.1 & & 226.580 & 0.468 & 0.218 & 9 & 0.218 & + 734 & h , b & 226.726 & 0.589 & 0.27 & 43.0 & & 226.724 & 0.583 & 0.279 & 8 & 0.000 & + 735 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 226.729 & -0.136 & 0.207 & 16 & 0.234 & + 736 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 226.740 & 0.114 & 0.220 & 15 & 0.000 & + 737 & h , b & 226.765 & 1.120 & 0.16 & 36.8 & & 226.761 & 1.114 & 0.203 & 15 & 0.185 & + 738 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 226.783 & -0.280 & 0.223 & 16 & 0.237 & + 739 & h , b & 226.901 & 0.035 & 0.19 & 74.0 & & 226.906 & 0.040 & 0.249 & 34 & 0.232 & + 740 & h , b & 227.114 & -0.266 & 0.08 & 54.0 & & 227.107 & -0.266 & 0.077 & 16 & 0.090 & a2026 + + 741 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.216 & 0.686 & 0.236 & 16 & 0.233 & + 742 & h , b & 227.217 & 0.021 & 0.22 & 58.4 & & 227.207 & 0.037 & 0.244 & 21 & 0.000 & + 743 & h , b & 227.259 & 0.385 & 0.22 & 71.0 & & 227.250 & 0.391 & 0.238 & 28 & 0.247 & + 744 & h , b & 227.279 & -1.202 & 0.28 & 67.4 & & 227.275 & -1.207 & 0.348 & 19 & 0.342 & + 745 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.290 & 1.141 & 0.247 & 21 & 0.000 & + 746 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.330 & -0.239 & 0.275 & 20 & 0.000 & + 747 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.367 & 0.899 & 0.268 & 17 & 0.000 & + 748 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.388 & -0.333 & 0.212 & 16 & 0.000 & + 749 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.421 & 1.154 & 0.273 & 15 & 0.000 & + 750 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.596 & -0.760 & 0.284 & 13 & 0.000 & + + 751 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.613 & 0.021 & 0.236 & 13 & 0.000 & + 752 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.619 & 1.029 & 0.292 & 17 & 0.000 & + 753 & h , b & 227.822 & -0.111 & 0.08 & 65.5 & & 227.821 & -0.130 & 0.101 & 42 & 0.092 & a2030 + 754 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.846 & 0.983 & 0.286 & 14 & 0.287 & + 755 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.870 & 0.472 & 0.288 & 13 & 0.302 & + 756 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 227.898 & 0.098 & 0.218 & 19 & 0.217 & + 757 & h , b & 227.977 & 0.621 & 0.16 & 47.0 & & 227.971 & 0.620 & 0.120 & 12 & 0.123 & + 758 & h & 227.980 & -1.247 & 0.25 & 58.5 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 759 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 228.073 & 0.723 & 0.216 & 14 & 0.219 & + 760 & h , b & 228.102 & -0.953 & 0.22 & 59.4 & & 228.003 & -0.948 & 0.207 & 21 & 0.212 & + + 761 & h , b & 228.109 & 0.159 & 0.23 & 45.8 & & 228.132 & 0.122 & 0.214 & 7 & 0.215 & + 762 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 228.118 & -0.073 & 0.194 & 14 & 0.215 & + 763 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 228.126 & -0.251 & 0.203 & 13 & 0.215 & + 764 & h , b & 228.196 & 0.936 & 0.25 & 64.0 & & 228.204 & 0.938 & 0.223 & 19 & 0.218 & + 765 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 228.206 & 0.814 & 0.129 & 14 & 0.000 & + 766 & h , b & 228.232 & -0.673 & 0.27 & 46.3 & & 228.281 & -0.674 & 0.314 & 8 & 0.000 & + 767 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 228.306 & 1.156 & 0.238 & 13 & 0.000 & + 768 & h , b & 228.324 & 0.273 & 0.20 & 50.1 & & 228.247 & 0.228 & 0.218 & 17 & 0.223 & + 769 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 228.381 & 1.073 & 0.251 & 13 & 0.218 & + 770 & h , b & 228.397 & -0.442 & 0.25 & 49.9 & & 228.398 & -0.458 & 0.277 & 11 & 0.251 & + + 771 & h , b & 228.421 & -0.201 & 0.29 & 42.8 & & 228.391 & -0.226 & 0.312 & 9 & 0.000 & + 772 & h , b & 228.195 & 0.381 & 0.05 & 23.0 & & 228.457 & 0.359 & 0.090 & 14 & 0.092 & + 773 & h , b & 228.588 & 1.179 & 0.11 & 36.0 & & 228.597 & 1.192 & 0.148 & 15 & 0.125 & + 774 & h , b & 228.677 & 0.271 & 0.11 & 40.8 & & 228.669 & 0.271 & 0.122 & 12 & 0.138 & + 775 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 229.053 & -0.801 & 0.122 & 28 & 0.118 & + 776 & h , b & 229.082 & 0.111 & 0.13 & 79.5 & & 229.083 & 0.060 & 0.120 & 39 & 0.117 & a2050 + 777 & h & 229.137 & -0.381 & 0.14 & 41.8 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 778 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 229.242 & -1.111 & 0.116 & 22 & 0.117 & + 779 & h , b & 229.497 & 0.227 & 0.05 & 19.2 & & 229.290 & 0.129 & 0.114 & 20 & 0.119 & + 780 & h , b & 229.326 & -0.708 & 0.11 & 73.2 & & 229.351 & -0.738 & 0.120 & 44 & 0.116 & a2053 + + 781 & h , b & 229.537 & 1.099 & 0.25 & 41.5 & & 229.589 & 1.065 & 0.247 & 6 & 0.000 & + 782 & h & 229.736 & -0.331 & 0.27 & 50.1 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 783 & h , b & 229.811 & -0.086 & 0.24 & 49.7 & & 229.816 & -0.066 & 0.212 & 6 & 0.205 & + 784 & h , b & 230.476 & -0.396 & 0.26 & 46.3 & & 230.466 & -0.389 & 0.240 & 10 & 0.000 & + 785 & h & 230.556 & 1.024 & 0.21 & 43.7 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 786 & h , b & 230.611 & -0.733 & 0.30 & 47.7 & & 230.589 & -0.661 & 0.288 & 9 & 0.000 & + 787 & h , b & 230.991 & 1.035 & 0.14 & 77.3 & & 230.979 & 1.064 & 0.138 & 16 & 0.077 & a2066 + 788 & h , b & 231.280 & 1.168 & 0.26 & 59.3 & & 231.282 & 1.160 & 0.277 & 13 & 0.279 & + 789 & h , b & 231.694 & 0.814 & 0.12 & 44.3 & & 231.675 & 0.891 & 0.135 & 10 & 0.116 & + 790 & h , b & 231.885 & -0.856 & 0.13 & 41.1 & & 231.853 & -0.704 & 0.129 & 10 & 0.000 & + + 791 & h , b & 232.308 & -0.250 & 0.06 & 48.9 & & 232.303 & -0.252 & 0.087 & 16 & 0.089 & + 792 & h & 232.329 & 0.927 & 0.27 & 52.0 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 793 & h , b & 232.641 & -0.803 & 0.06 & 41.7 & & 232.639 & -0.809 & 0.070 & 10 & 0.077 & + 794 & b & 0.000 & 0.000 & 0.00 & 0.0 & & 232.764 & -0.201 & 0.288 & 15 & 0.000 & + 795 & h , b & 233.274 & -0.766 & 0.13 & 82.6 & & 233.265 & -0.771 & 0.148 & 53 & 0.149 & + 796 & h , b & 233.367 & -0.378 & 0.20 & 49.2 & & 233.359 & -0.386 & 0.257 & 16 & 0.261 & + 797 & h , b & 234.891 & 0.590 & 0.05 & 34.1 & & 234.877 & 0.853 & 0.109 & 14 & 0.085 & + 798 & h & 235.196 & 0.924 & 0.09 & 40.2 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + 799 & h & 235.544 & 1.180 & 0.26 & 47.0 & & 0.000 & 0.000 & 0.000 & 0 & 0.000 & + rrrrr 40 - 50 & 297 & @xmath109 & 13 - 20 & 402 + 50 - 60 & 97 & @xmath109 & 20 - 30 & 103 + 60 - 70 & 25 & @xmath109 & 30 - 40 & 15 + @xmath470 & 17 & @xmath109 & @xmath440 & 4 + total(@xmath3@xmath440 ) & 436 & @xmath109 & total(@xmath5@xmath413 ) & 524 + z@xmath110 & n@xmath36(hmf ) & @xmath109 & z@xmath111 & n@xmath36(maxbcg ) + & ( @xmath3@xmath440 ) & @xmath109 & & ( @xmath5@xmath413 ) + + 0.05 - 0.1 & 45 & @xmath109 & 0.05 - 0.1 & 52 + 0.1 - 0.2 & 150 & @xmath109 & 0.1 - 0.2 & 173 + 0.2 - 0.3 & 241 & @xmath109 & 0.2 - 0.3 & 299 + total(z=0.05 - 0.3 ) & 436 & @xmath109 & total(z=0.05 - 0.3 ) & 524 + rrrrrrrrr rxc j0020.1 + 0005 & 5.044 & 0.092 & 0.212 & 1.08 & 0.220 & 33 & & + rxc j0106.8 + 0103 & 16.709 & 1.055 & 0.253 & 8.65 & 0.286 & 25 & 0.22 & 50.9 + rxc j0114.9 + 0024 & 18.740 & 0.406 & 0.044 & 0.92 & 0.068 & 12 & 0.05 & 42.6 + rxc j0121.9 + 0021 & 20.493 & 0.358 & 0.175 & 1.85 & 0.166 & 32 & 0.16 & 46.7
+ rxc j0152.7 + 0100 & 28.182 & 1.016 & 0.227 & 9.37 & 0.264 & 29 & 0.23 & 73.4 + rxc j1326.2 + 0013 & 201.573 & 0.225 & 0.082 & 1.65 & 0.098 & 13 & 0.13 & 44.1 + rxc j1415.8 + 0015 & 213.965 & 0.259 & 0.125 & 0.83 & 0.129 & 15 & 0.13 & 49.4 + rxc j2341.1 + 0018 & 355.276 & 0.315 & 0.110 & 1.03 & 0.140 & 8 & 0.09 & 45.1 + | we present a catalog of 799 clusters of galaxies in the redshift range z@xmath0 = 0.05 - 0.3 selected from @xmath1400 deg@xmath2 of early sdss commissioning data along the celestial equator .
the catalog is based on merging two independent selection methods a color - magnitude red - sequence maxbcg technique ( b ) , and a hybrid matched - filter method ( h ) .
the bh catalog includes clusters with richness @xmath3@xmath4 40 ( matched - filter ) and @xmath5@xmath4 13 ( maxbcg ) , corresponding to typical velocity dispersion of @xmath6@xmath7 400 km s@xmath8 and mass ( within 0.6 @xmath9 mpc radius ) @xmath10 m@xmath11 .
this threshold is below abell richness class 0 clusters .
the average space density of these clusters is @xmath12 mpc@xmath13 .
all noras x - ray clusters and 53 of the 58 abell clusters in the survey region are detected in the catalog ; the 5 additional abell clusters are detected below the bh catalog cuts . the cluster richness function is determined and found to exhibit a steeply decreasing cluster abundance with increasing richness .
we derive observational scaling relations between cluster richness and observed cluster luminosity and cluster velocity dispersion ; these scaling relations provide important physical calibrations for the clusters
. the catalog can be used for studies of individual clusters , for comparisons with other sources such as x - ray clusters and agns , and , with proper correction for the relevant selection functions , also for statistical analyses of clusters . |
in situations of practical interest , physical systems evolve in time under the action of well studied operators , one common example is provided by the diffusion processes .
sampling of such an evolving system can be done by sensors or measurement devices that are placed at various locations and can be activated at different times . in practice , increasing the spatial sampling density is usually much more expensive than increasing the temporal sampling rate ( @xcite ) .
given the different costs associated with spatial and temporal sampling , we aim to reconstruct any states in the evolutionary process using as few sensors as possible , but allow one to take samples at different time levels .
besides , in some cases , obtaining samples at a sufficient rate at any single time level may not even be possible , spatiotemporal sampling may resolve this issue by oversampling in time .
a natural question is whether one can compensate for insufficient spatial sampling densities by oversampling in time .
this setting has departed from the classical sampling theory , where the samples are taken simultaneously at only one time level , see @xcite .
dynamical sampling is a newly proposed mathematical sampling framework .
it involves studying the time - space patterns formed by the locations of the measurement devices and the times of their activation . mathematically speaking ,
suppose @xmath3 is an initial distribution that is evolving in time satisfying the evolution rule : @xmath9 where @xmath10 is a family of evolution operators satisfying the condition @xmath11 .
dynamical sampling asks the question : when do coarse samplings taken at varying times @xmath12 contain the same information as a finer sampling taken at the earliest time ?
one goal of dynamical sampling is to find all spatiotemporal sampling sets @xmath13 such that the certain classes of signals @xmath3 can be recovered from the spatiotemporal samples @xmath14 . in the above cases ,
the evolution operators are assumed to be known .
it has been well - studied in the context of various evolutionary systems in a very general setting , see @xcite .
another important problem arises when the evolution operators are themselves unknown or partially known . in this case , we are interested in finding all spatiotemporal sampling sets and certain classes of evolution operators so that the family @xmath10 or their spectrum and even the unknown states can be identified .
we call such a problem the unsupervised system identification problem in dynamical sampling
. applications to wireless sensor networks ( wsn ) is a natural setting for dynamical sampling .
wsn are widely used in many industries , including the health , military , and environmental industries , see @xcite for numerious examples . in wsn ,
a huge amount of sensors are distributed to monitor a physical field such as the pollution , temperature or pressure .
the goal is to exploit the evolutionary structure and the placement of sensors to reconstruct an unknown field .
however , the current approaches and algorithms do not make use of the evolutionary structure of the field , see @xcite .
it is not always possible to place sampling devices at the desired locations .
then we may be able to recover the desired information by placing the sensors elsewhere and use the evolution process to recover the signals at the relevant locations .
in addition , dynamical sampling will make the reconstruction cheaper since we use a reduced numer of sensors . in this subsection , we state an instance of the unsupervised system identification problem of dynamical sampling in an infinite dimensional setting .
let @xmath15 be an unknown initial spatial signal and the evolution operator @xmath1 be given by an unknown convolution filter @xmath2 such that @xmath16 . at time @xmath17
, the signal @xmath3 evolves to be @xmath18 , where @xmath19 .
we call this evolutionary system spatially invariant . given the spatiotemporal samples with both @xmath3 and @xmath1 unknown , we would like to recover as much information about them as we can under the given various priors . here
we first study the case of uniform subsampling . without loss of generality , we assume a positive odd integer @xmath20 ( @xmath21 ) to be the uniform subsampling factor . at time level @xmath22 , we uniformly undersample the evolving state @xmath23 and get the spatiotemporal data @xmath24 which is a sequence in @xmath4 .
it is obvious that at any single time level @xmath22 , we can not determine the state @xmath23 from the measurement @xmath25 the problem we are going to consider can be summarized as follows : _ under what conditions on @xmath26 and @xmath3 , can @xmath8 and @xmath3 be recovered from the spatiotemporal samples @xmath27 , or equivalently , from the set of measurement sequences @xmath28 _ ? in @xcite ,
aldroubi and krishtal consider the recovery of an unknown @xmath29 matrix @xmath30 and an unknown initial state @xmath31 from coarse spatial samples of its successive states @xmath32 .
given an initial sampling set @xmath33 , they employ techniques related to krylov subspace methods to show how large @xmath34 should be to recover all the eigenvalues of @xmath30 that can possibly be recovered from spatiotemporal samples @xmath35 .
our setup is very similar to the special case of regular invariant dynamical sampling problem in @xcite . in this special case
, they employ a generalization of the well known prony method that uses these regular undersampled spatiotemporal data first for the recovery of the filter @xmath8 . then by using techniques developed in @xcite ,
they show how to recover the initial state from these spatiotemporal samples . in this paper
, we will address the infinite dimensional analog of this special case and provide more algorithms . in @xcite ,
peter and plonka use a generalized prony method to reconstruct the sparse sums of the eigenfunctions of some known linear operators .
our generalization of prony method shares some similar spirits with it , but deals with a fundamentally different problem . in sparse fourier transformation , see @xcite , the idea is to uniformly undersample the fixed signal with different factors so that one can group subsets of fourier space together into a small number of bins to isolate frequencies , then take an aliasing - based search by chinese remainder theorem so that one can recover the coefficients and the frequencies . in our case , intuitively , one can think of recovering of the shape of an evolving wave by observing the amplitude of its aliasing version at fixed coarse locations over a long period of time as opposed to acquiring all of the amplitudes at once , then by the given priors , one can achieve the perfect reconstructions . other similar work
include the the slepian - wolf distributed source coding problem @xcite and the distributed sampling problem in @xcite .
our problem , however , is very different from the above in the nature of the processes we study .
distributed sampling problem typically deals with two signals correlated by a transmission channel .
we , on the other hand , can observe an evolution process at several instances and over longer periods of time .
in the following , we use standard notations . by @xmath36 , we denote the set of all positive integers . the linear space of all column vectors with @xmath37 complex components is denoted by @xmath38 .
the linear space of all complex @xmath39 matrices is denoted by @xmath40 for a matrix @xmath41 , its transpose is denoted by @xmath42 , its conjugate - transpose by @xmath43 and its moore - penrose pseudoinverse by @xmath44 a square matrix @xmath45 is abbreviated to @xmath46 .
its infinity norm is defined by @xmath47 for a vector @xmath48 , the @xmath49 diagonal matrix built from @xmath50 is denoted by @xmath51 .
we define the infinity norm @xmath52 .
it is easy to see that @xmath53 further we use the known submatrix notation coincides with matlab . for example , @xmath54 is the submatrix of @xmath55 obtained by extracting rows 1 through @xmath37 and columns 2 through @xmath56 , and @xmath57 means the last column vector of @xmath55 .
the minimal annihilating polynomial of a square matrix @xmath58 is @xmath59 $ ] , if it is the monic polynomial of smallest degree among all the monic polynomials @xmath60 such that @xmath61 .
we will denote the degree of @xmath59 $ ] by @xmath62 .
let the monic polynomial @xmath63=\sum\limits_{k=0}^{m-1}p_kz^k+z^m$ ] , the companion matrix of @xmath63 $ ] is defined by @xmath64}=\left ( \begin{array}{ccccc } 0&0 & \cdots&0&-p_0\\ 1 & 0 & \cdots&0&-p_1 \\ 0 & 1 & \cdots&0 & -p_2\\ \vdots&\vdots&\ddots&\vdots&\vdots\\ 0 & 0 & \cdots&1 & -p_{m-1 } \end{array } \right).\ ] ]
let @xmath65 be @xmath66 distinct complex numbers , denote @xmath67^t$ ] , the @xmath68 vandermonde matrix generated by @xmath69 is defined by @xmath70 for a sequence @xmath71 , we define its fourier transformation to be the function on the torus @xmath72 @xmath73 the remainder of the paper is organized as follows : in section [ noisefree ] , we discuss the noise free case . from a theoretical aspect ,
we show that we can reconstruct a `` typical low pass filter '' @xmath8 and the initial signal @xmath3 from the dynamical spatiotemporal samples @xmath74 almost surely , provided @xmath75 . for the case
when both @xmath8 and @xmath3 are of finite impulse response and an upper bound of their support is known , we propose a generalized prony method algorithm to recover the fourier spectrum of @xmath8 .
in section [ stabilityanalysis ] , we provide a perturbation analysis of this algorithm .
the estimation results are formulated in the rigid @xmath76 norm and give us an idea of how the performance depends on the system parameters @xmath77 and @xmath20 . in section
[ ne ] , we do several numerical experiments to verify some estimation results . in section [ othernumericalmethods ] , we propose several other algorithms such as generalized matrix pencil method , generalized esprit method and cadzow denoising methods to improve the effectiveness and robustness of recovery .
the comparison between algorithms is illustrated by a numerical example in section [ numericalexamples ] .
finally , we summarize the work in section [ conclusion ] .
we consider the recovery of a frequently encountered case in applications when the filter @xmath2 is a typical low pass filter " so that @xmath78 is real , symmetric and strictly decreasing on @xmath79 $ ] .
an example of such a typical low pass filter is shown in figure 1 .
the symmetry reflects the fact that there is often no preferential direction for physical kernels and monotonicity is a reflection of energy dissipation .
[ lp ] without loss of generality , we also assume @xmath8 is a normalized filter , i.e. , @xmath80 . in this section , we assume the spatiotemporal data @xmath81 is exact . define the downsampling operator @xmath82 by @xmath83 then @xmath84 due to the poisson summation formula and the convolution theorem , we have the lemma below for the downsampling operator .
[ psf]the fourier transform of each measurement sequence @xmath85 at @xmath86 is @xmath87[timeseries ] let @xmath88 be an integer satisfying @xmath75 , we define the @xmath89 column vector @xmath90^{t},\ ] ] and build the hankel matrices @xmath91,\ ] ] @xmath92.\ ] ] for @xmath93 , we introduce the notations @xmath94^t$ ] and @xmath95^t$ ] .
[ hankelrecovery ] let @xmath88 be an integer satisfying @xmath75 . 1 .
then the rectangular hankel matrices can be factorized in the following form : @xmath96where @xmath97 .
the vandermonde matrix @xmath98 and @xmath99 are given in the way as indicated in definition [ v ] .
2 . assume the entries of @xmath100 are all nonzero .
the rank of the hankel matrix @xmath101 can be summarized as follows : @xmath102 3 . assume the entries of @xmath100 are all nonzero .
for @xmath103 , the vector @xmath104^t$ ] is the unique solution of the linear system @xmath105if and only if the polynomial @xmath106=\sum\limits_{k=0}^{m-1}q_k(\xi)z^k+z^m\]]with coefficients given by @xmath107 is the minimal annihilating polynomial of the diagonal matrix @xmath108 . in other words ,
the polynomial @xmath109 $ ] has all @xmath110 as roots .
moreover , if @xmath63 $ ] is a monic polynomial of degree @xmath20 , then @xmath111}=\mathbf{h}_{n - m , m}^\xi(1)\ ] ] if and only if @xmath63 $ ] is the minimal annihilating polynomial of @xmath108 .
\(1 ) by lemma [ psf ] , for @xmath112 , we have the identity : @xmath113 hence the first identity follows by the definition of @xmath101 .
notice that for @xmath114 , @xmath115 the second identity follows similarly .
+ ( 2 ) by the symmetric and monotonicity condition of @xmath116 on @xmath117 , we have @xmath118 since @xmath75 , @xmath119 . by our assumptions
, @xmath120 is invertible .
the rank of hankel matrix @xmath101 can be computed by its factorization results in ( 1 ) .
+ ( 3 ) if @xmath121 , then the diagonal matrix @xmath108 has @xmath20 distinct eigenvalues consist of @xmath122 the minimal annihilating polynomial of @xmath108 is of degree @xmath20 .
suppose @xmath109=\sum\limits_{k=0}^{m-1}q_k(\xi)z^k+z^m$ ] is the minimal annihilating polynomial of @xmath108 , @xmath123=0.$ ] in other words , @xmath124 then @xmath125 conversely , if @xmath107 is the solution of linear system , let the monic polynomial given by @xmath126 be @xmath109 $ ] , then by the computation process of , we have @xmath127 \mathbf{x}(\xi)=0.\ ] ] since @xmath128 is full column rank , @xmath123 \mathbf{x}(\xi)=0.$ ] by the fact that @xmath123 $ ] is diagonal and @xmath100 has no zero entries , we know @xmath109 $ ] is a monic annihilating polynomial of @xmath108 .
the minimality is followed by counting its degree .
if @xmath63 $ ] is a monic annihilating polynomial of @xmath108 , by computations , it is easy to show the identity is an equivalent formulation with the identity . in the case of @xmath129 or @xmath130 , if @xmath120 is invertible , then the coefficient vector of the minimal annihilating polynomial of @xmath108 @xmath131 is the unique solution of the following linear system : @xmath132 where @xmath133.$ ] let @xmath134 denote the lebesgue measure on @xmath117 , and @xmath135 be a subclass of @xmath4 defined by @xmath136 clearly , @xmath135 is a dense class of @xmath4 under the norm topology . in noise free scenario ,
we show that we can recover @xmath8 and @xmath3 provided that our initial state @xmath137 .
[ thm8.3]let @xmath137 be the initial state and the evolution operator @xmath1 be a convolution operator given by @xmath138 so that @xmath78 is real , symmetric , and strictly decreasing on @xmath139 $ ] . then @xmath8 and @xmath3 can be recovered from the set of measurement sequences @xmath140 defined in @xmath141 when @xmath75 .
since fourier transformation is an isometric isomorphism from @xmath4 to @xmath142 , we can look at this recovery problem on the fourier domain equivalently .
we are going to show that the regular subsampled data @xmath74 contains enough information to recover the fourier spectrum of @xmath8 on @xmath117 up to a measure zero set . by our assumptions of @xmath3
, there exists a measurable subset @xmath143 of @xmath117 with @xmath144 , so that @xmath120 is an invertible matrix for @xmath145 . let @xmath146 , if @xmath147 , by ( 3 ) of proposition [ hankelrecovery ] , we can recover the minimal annihilating polynomial of @xmath108 . now to recover the diagonal entries of @xmath108 , it amounts to finding the roots of this minimal annihilating polynomial and ordering them according to the monotonicity and symmetric condition on @xmath116 . in summary , for each @xmath147 , we can uniquely determine @xmath148 .
note @xmath149 , and hence we can recover the fourier spectrum of @xmath8 up to a measure zero set .
the conclusion is followed by applying the inverse fourier transformation on @xmath78 .
once @xmath8 is recovered , we can recover @xmath3 from the spatiotemporal samples @xmath150 using techniques developed in @xcite .
theorem [ thm8.3 ] addresses the infinite dimensional analog of theorem 4.1 in @xcite .
if we do nt know anything about @xmath8 in advance , with minor modifications of the above proof , one can show the recovery of the range of @xmath116 on a measurable subset of @xmath117 , where the measure of this subset is 1 .
let @xmath151 , the support set of @xmath8 is defined by @xmath152 .
if @xmath153 is a finite set , @xmath8 is said to be of finite impulse response .
in particular , if @xmath3 is of finite impulse response , then @xmath137 .
now if both @xmath3 and @xmath8 are of finite impulse response , and we know an upper bound @xmath154 such that @xmath153 and @xmath155 are contained in @xmath156 , then we can compute the value of the fourier transformation of @xmath74 at any @xmath157 from the proof of theorem [ thm8.3 ] , we can give an algorithm similar to the classical prony method to recover @xmath148 almost surely , given @xmath158 chosen uniformly from @xmath117 .
it is summarized in algorithm @xmath159
@xmath160 , @xmath154 , @xmath161 , @xmath162 compute the fourier transformation of the measurement sequences @xmath74 and build the hankel matrix @xmath163 and the vector @xmath164 .
compute the solution of the overdetermined linear system : @xmath165 form the polynomial @xmath109=\sum\limits_{k=0}^{m-1}q_k(\xi)z^k+z^m$ ] and find its roots , this can be done by solving the standard eigenvalue problem of its companion matrix .
order the roots by the monotonicity and symmetric condition of @xmath116 to get @xmath166 .
+ @xmath167 [ cor1 ] in addition to the assumptions of theorem [ thm8.3 ] ,
if both @xmath8 and @xmath3 are of finite impulse response with support contained in @xmath168 for some @xmath169 , then it is enough to determine @xmath8 and @xmath3 after we recover @xmath170 at @xmath171 distinct locations by algorithm [ algo : prony ] . under these assumptions , we know
@xmath172 suppose @xmath173 are recovered , we set up the following linear equation @xmath174 note that @xmath175 is a chebyshev system on @xmath176$](see @xcite ) , and hence has a unique solution
. then we can recover @xmath3 by solving the linear system @xmath177 for finitely many @xmath158s , which finishes the proof . for @xmath178
, @xmath179 is called the @xmath180 approximation of @xmath2 if @xmath181 if @xmath182 , @xmath183 otherwise .
the @xmath180 approximation error , denote by @xmath184 , is defined by @xmath185 . by corollary
@xmath186 , if @xmath187 for some @xmath188 , reconstructing @xmath8 amounts to finding its finitely many fourier spectrums .
it is natural to consider an approximation scheme for those filters with @xmath184 decaying rapidly to 0 as @xmath171 increase .
fix @xmath178 , let @xmath189 be the solved fourier spectrum set and denote @xmath190 .
let @xmath191 be the interpolation matrix defined by @xmath192 and the interpolation vector be defined by @xmath193 .
denote @xmath194 , we approximate @xmath195 by @xmath196 , which is the solution of @xmath197 and define the approximation error by @xmath198 .
let @xmath199 be the vector defined by @xmath196 , it follows that @xmath200 .
consequently , it is important to understand the dependence of @xmath201 on the choice of @xmath202 and the value of @xmath184 .
[ vandermonde]for arbitrary @xmath203 with @xmath204 if @xmath205 , let @xmath206 be the @xmath207 vandermonde matrix with nodes @xmath208 defined as @xmath209 there holds @xmath210 [ app1]let @xmath195 , @xmath211 and @xmath196 be defined as above . there holds @xmath212 where @xmath213 moreover , @xmath214 let @xmath215 , so we have the equations @xmath216 by our assumptions , @xmath217 hence @xmath218 now it suffice to study the relation between @xmath219 and parameters @xmath190 .
observe @xmath220 therefore @xmath221 now use the norm estimation results of inverse vandermonde matrix in @xcite , the conclusion follows .
proposition @xmath222 sheds some light on how to choose nodes and how small @xmath184 should be to make @xmath201 small .
if the we choose @xmath223 such that @xmath224 has very small minimum separation among its nodes , @xmath225 may grow exponentially when @xmath171 increase . on the other hand , if we choose @xmath226 , then @xmath227 .
in fact , since our nodes of @xmath228 areon the unit circle , the upper bound on estimation might be very rough and @xmath228 could be well conditioned . for condition number of vandermonde matrix on the unit circle
, we refer to @xcite for details .
in previous sections , we have shown that if we are able to compute the spectral data @xmath229 at @xmath158 , then we can recover the fourier spectrum @xmath148 by algorithm [ algo : prony ] .
however , we assume the spectral data are noise free . a critical issue still remains .
we need to analyze the accuracy of the solution achieved by algorithm [ algo : prony ] in the presence of noise .
mathematically speaking , assume the measurements are given by @xmath230 compared to @xmath141 so that @xmath231 for all @xmath93 .
given an estimation for @xmath232 , how large can the error be in the worst case for the output parameters of algorithm [ algo : prony ] in terms of @xmath233 , and the system parameters @xmath234 and @xmath20 .
most importantly , we need to understand analytically what kind of effects that the subsampling factor @xmath20 will impose on the performance of the algorithm [ algo : prony ] . in this section , for simplicity , we choose @xmath235 to meet the minimal requirement . in this case , the hankel matrix @xmath163 is a square matrix and the vectors @xmath236 are of length @xmath20 . we denote them by two new notations : @xmath237 and @xmath238 .
our perturbation analysis will consist of two steps .
suppose our measurements are perturbed from @xmath239 to @xmath240 .
for any @xmath158 , we firstly measure the perturbation of @xmath107 in terms of @xmath76 norm .
secondly we measure the perturbation of the roots .
it is well known that the roots of a polynomial are continuously dependent on the small change of its coefficients , see proposition [ polyp ] .
hence , for a small perturbation , although the roots of the perturbed polynomial @xmath241 $ ] may not be real , we can order them according to their modulus and have a one to one correspondence with the roots of @xmath242 $ ] . before presenting our main results in this section ,
let us introduce some useful notations and terminologies .
[ def6]let @xmath243 , consider the set @xmath244 that consists of m distinct nodes . 1 . for @xmath245 ,
the separation between @xmath246 with other @xmath247 nodes is measured by @xmath248 2 . for @xmath249 , the @xmath250-th elementary symmetric function generated by the @xmath20 nodes is denoted by @xmath251 for @xmath252 , the @xmath250-th elementary symmetric function generated by @xmath247 nodes with @xmath253 missing is denoted by @xmath254 .
the following proposition measures the perturbation of the polynomial roots in terms of the perturbation of its coefficients and is the key to our perturbation analysis .
[ polyp ] let @xmath255 be a root of multiplicity @xmath256 of the @xmath171-th order polynomial @xmath63.$ ] for all @xmath257 , let @xmath258=p[z]+\epsilon \delta p[z],$ ] where @xmath259 $ ] is a polynomial of order lower than @xmath171 .
suppose that @xmath260 \neq 0.$ ] then there exists a positive @xmath261 such that for all @xmath262 there are exactly @xmath263 roots of @xmath258 $ ] , denoted @xmath264 , which admit the first - order fractional expansion @xmath265 where @xmath266 is an arbitrary @xmath263-th root of the complex number @xmath267}{\frac{1}{m_k ! } p^{(m_k)}[z_k]}.\ ] ] [ prop3]let the perturbed measurements @xmath240 be given with an error satisfying @xmath268 .
let @xmath269 and @xmath270 be given by @xmath271 in the same way as in and .
assume @xmath237 is invertible and @xmath233 is sufficient small so that @xmath269 is also invertible .
denote by @xmath272 the solution of the linear system @xmath273 .
let @xmath241 $ ] be the prony polynomial formed by @xmath274 and @xmath275 be its roots , then we have the following estimates as @xmath276 , @xmath277 where @xmath278 . as a result
, we achieve the following first order estimation @xmath279 where @xmath280 . note that linear system
@xmath281 is perturbed to be @xmath282 by our assumptions , we have @xmath283 @xmath284 define @xmath285 , by simple computation , @xmath286 hence if @xmath276 , we obtain @xmath287 now we can easily get an estimation of @xmath76 norm of @xmath288 @xmath289 since @xmath290 are the roots of @xmath109 $ ] , using vieta s formulas(see @xcite ) , we know @xmath291 let @xmath292 $ ] be the polynomial of degree less than or equal to @xmath247 defined by the vector @xmath293 . using proposition [ polyp ] , and denote by @xmath294 $ ] the derivative function of @xmath109 $ ] , for @xmath295 , we conclude @xmath296 } { ( q^{\xi})^{'}[\hat a(\frac{\xi+i}{m})]}+o(\epsilon^2 ) \lvert \\ & \leq \frac{\lvert\lvert \delta \mathbf{q}(\xi ) \rvert\rvert_{\infty } ( \sum\limits_{k=0}^{m-1}\lvert \hat a ^{k}(\frac{\xi+i}{m})\rvert)}{\prod\limits_{j \neq i \atop 0\leq j\leq m-1 }
\lvert \hat a(\frac{\xi+j}{m})-\hat a(\frac{\xi+i}{m } ) \rvert}+o(\epsilon^2)\\ & \leq c_i(\xi)\lvert \lvert \mathbf{h}_m^{-1}(\xi ) \rvert\rvert_{\infty}(1+m\mathop{max}_{1 \leq k\leq m}\lvert \sigma_{k}(\xi)\rvert)\epsilon+o(\epsilon^2 ) , \end{split}\ ] ] where @xmath297 .
therefore it is important to understand the relation between the behavior of @xmath298 and our system parameters , i.e , @xmath8 , @xmath20 and @xmath3 .
next , we are going to estimate @xmath298 and reveal their connection with the spectral properties of @xmath8,@xmath3 and the subsampling factor @xmath20 .
[ theorem3 ] assume @xmath237 is invertible , we have the lower bound estimation @xmath299 where @xmath300 , and the upper bound estimation @xmath301 firstly , we prove the lower bound for @xmath298 .
denote the vandermonde matrix @xmath302 by the abbreviated @xmath303 .
suppose @xmath304 is the inverse of @xmath303 , by the inverse formula for a standard vandermonde matrix , @xmath305 let @xmath306 be the standard basis for @xmath307 and @xmath308 for @xmath309 .
since @xmath310 , we conclude that @xmath311 .
@xmath312 on the other hand , using the factorization and the upper bound norm estimation for the inverse of a vandermonde matrix in @xcite , we show that @xmath313 as an application of proposition @xmath314 , the following theorem sheds some light on the dependence of @xmath315 on m. [ cor2 ] if @xmath316 for every @xmath93 , then @xmath317 .
therefore , @xmath318 as @xmath319 .
we show this by proving @xmath320 .
note @xmath321 . by , @xmath322 the conclusion follows .
let @xmath323 .
note that @xmath324 since every entry of @xmath325 is contained in @xmath326 $ ] , the chebyshev points on @xmath326 $ ] maximize the determinant of vandermonde matrix , see @xcite .
therefore , by the formula for the determinant of a vandermonde matrix on the chebyshev points in @xcite , we get @xmath327 by @xmath328 , @xmath329 which implies that @xmath330 hence by @xmath331 @xmath332 by our proof , we also see that @xmath315 grows at least geometrically when m increases . summarizing
, our results in this section suggest that 1 . for @xmath333 , the accuracy of recovering the node @xmath246 not only depends on its separation with other nodes @xmath334(see definition [ def6 ] ) , but also depends on the global minimal separation @xmath335 among the nodes .
fix @xmath336 , our estimations and suggest that error @xmath337 in the worst possible case could be proportional to @xmath338 .
our numerical experiment suggests this is sharp , see figure 2 ( c ) and ( d ) .
the accuracy of recovering all nodes is inversely proportional to the lowest magnitude of @xmath339 .
3 . increasing @xmath20 may result in amplifying the error caused by the noise significantly . since by the proof of theorem @xmath340
, @xmath341 grows at least geometrically when @xmath20 increases .
thus , when m increases , the infinity norm of @xmath342 gets bigger and our solutions become more likely less robust to noise , see figure 2 ( a ) and ( b ) .
in this section , we provide some simple numerical simulations to verify some theoretical accuracy estimations in section [ stabilityanalysis ] .
suppose our filter @xmath8 is around the center of radius 3 .
for example , let @xmath343 such that @xmath344 , @xmath345 such that @xmath346 we choose @xmath347 . 1 . in this experiment , we choose 9 points @xmath348=0.49:0.001:0.498 $ ] and calculate @xmath349 and the perturbed @xmath350 for @xmath351 , where @xmath352 is defined as in @xmath141 and @xmath353 .
2 . use algorithm [ algo : prony ] to calculate the roots of @xmath109 $ ] and the perturbed roots of @xmath241 $ ] respectively , then compute @xmath354 for @xmath355 3 . choose @xmath356 and @xmath357 ,
we compute @xmath358 for different m. in this subsection , we plot several figures to reflect the experiment results .
the @xmath3-axis of the figure 2 ( a)@xmath359(e ) are set to be 1:9 , which represent @xmath360 1 .
* the dependence of @xmath361 on the infinity norm of @xmath362*. since the points @xmath363 are more and more closer to @xmath130 , we expect the infinity norm of @xmath364 to get sufficiently larger and larger
. note that @xmath20 and @xmath3 are fixed , the quantity @xmath364 is the only significantly large item in the error estimations .
we plot the value of @xmath365 and @xmath366 for @xmath367 in figure 2 ( a ) and ( b ) .
they exhibit almost the same behaviour and grows proportionally .
this indicates that the bigger @xmath365 is , the bigger @xmath366 is . 2 .
* sharpness of estimation and * .
our estimation and suggest that error @xmath368 in the worst possible case could be proportional to @xmath338 .
we plot the value of @xmath369 and @xmath370 for @xmath367 in figure 2 ( c ) and ( d ) .
it is indicated that @xmath371 grows approximately proportionally to the growth of @xmath372 , which suggests the sharpness of estimation and .
it is worthy to mention that the curve of @xmath366 coincides with the curve of @xmath369 , and the curve of @xmath373 coincides with the curve of @xmath372 . since in this experiment , @xmath20 and @xmath3
are fixed , this also suggests that the quantity @xmath374 essentially decides the accuracy .
the bigger the quantity is , the less accuracy the algorithm is .
the infinity norm of @xmath375*. recall in this experiment , we choose @xmath376 and @xmath356 .
we plot the value of @xmath377 for different @xmath20 .
the results are presented in figure 2 ( f ) .
the @xmath378axis is set to be logarithmic .
it is shown that @xmath298 grows geometrically .
in the following subsections , we will investigate the data structure of the hankel matrix built from the spatiotemporal samples and present two algorithms based on the classical matrix pencil method and esprit estimation method to our case .
these two classical methods are well known for their better numerical stability than the original prony method .
let @xmath379 and @xmath88 be two integers satisfying @xmath380 and @xmath381 .
similarly , we define the @xmath382 column vector @xmath383^{t } , \ ] ] and form the rectangular hankel matrices @xmath384,\ ] ] @xmath385 similar to the case @xmath386 , for @xmath387 @xmath388 recall that the superscripts `` @xmath389 '' and @xmath390 " will denote the conjugate transpose and the pseudoinverse . the following lemma provides a foundation with the generalized matrix pencil method .
let @xmath391 be two postive integers s.t .
assume @xmath103 and @xmath120 is invertible .
the solutions to the generalized singular eigenvalue problem : @xmath393 subject to @xmath394 , which denotes the column space of @xmath395 are @xmath396 @xmath397 for @xmath398 the proof can be done by the factorization @xmath399 and a similar manner with the proof of theorem 2 in @xcite .
[ mtx ] let @xmath391 be two postive integers s.t .
assume @xmath103 and @xmath120 is invertible .
the @xmath400 matrix @xmath401 has @xmath402 and @xmath403 zeros as eigenvalues .
left multiplying by @xmath404 , we have @xmath405 by property of pseudoinverse , @xmath406 is the orthogonal projection onto the @xmath407 . since @xmath408 , it is easy to see that the set @xmath166 are @xmath20 eigenvalues of @xmath409 since the rank of @xmath401 is @xmath410 , @xmath401 has @xmath403 zero eigenvalues .
it is immediate to see that one advantage of the matrix pencil method is the fact that there is no need to compute the coefficients of the minimal annihilating polynomial of @xmath108 . in this way
, we just need to solve a standard eigenvalue problem of a square matrix @xmath401 . in order to compute @xmath401 , inspired by idea of algorithm 5 for svd based matrix pencil method in @xcite
, we can employ the singular value decomposition(svd ) of the hankel matrices .
in addition to the conditions of proposition [ mtx ] , given the svd of the hankel matrix , @xmath411 then @xmath412 this can be shown by direct computations and noticing that @xmath413 has only @xmath20 nonzero singular values .
we summarize the generalized matrix pencil method in algorithm [ algo : matrixpencil ] . note that the amount of computation required by algorithm [ algo : matrixpencil ] depends on the free parameter @xmath379 .
numerical experiments show that the choice of @xmath379 greatly affects the noise sensitivity of the eigenvalues . in terms of the noise sensitivity and computation cost
, the good choice for @xmath379 is between one third of @xmath88 and two thirds of @xmath88 @xcite . in our numerical example , we choose @xmath379 to be around one third of @xmath88 .
@xmath392 , @xmath154 , @xmath161 , @xmath162 compute the fourier transformation of the measurement sequences @xmath74 and build the hankel matrix @xmath413 and compute its svd @xmath414 compute the eigenvalues of @xmath415 .
delete @xmath403 smallest values in modulus ( zeros in noise free case ) from the eigenvalues .
order the rest eigenvalues by the monotonicity and symmetric condition of @xmath116 to get @xmath166 .
+ @xmath167 original esprit method relies on a particular property of vandermonde matrices known as the rotational invariance @xcite . by the factorization results , we have seen that the hankel data matrix @xmath413 containing successive spatiotemporal data of the evolving states is rank deficient and that its range space , known as the signal subspace , is spanned by vandermonde matrix generated by @xmath402 .
hence we can generalize the idea and present the generalized esprit algorithm based on svd for estimating the @xmath416 in our case .
we summarize it in algorithm [ algo : esprit ] @xmath417 , @xmath154 , @xmath74 , @xmath162 compute the fourier transformation of the measurement sequences @xmath74 and form the hankel matrix @xmath418 compute the svd of @xmath419 compute the @xmath420 spectral matrix @xmath421 by solving the linear system @xmath422 and estimate the eigenvalues of @xmath421 .
order the eigenvalues by the monotonicity and symmetric condition of @xmath116 to get @xmath166 .
+ @xmath167 it has been shown in the previous sections the hankel matrix @xmath423 has two key properties in the noise free case under appropriate hypothesis : 1 .
it has rank @xmath20 .
2 . it is toeplitz . in the noisy case ,
these two properties are not initially satisfied simultaneously .
@xmath413 is very sensitive to noise , numerical experiments show that even very small noise ( @xmath424 ) will change its rank dramatically . to further improve robustness
, we use an iterative method devised by cadzow @xcite to preprocess the noisy data and guarantee to build a hankel matrix with above two key properties . in our context
, it can be summarized in algorithm [ algo : cadzow ] .
@xmath417 , @xmath425 , @xmath162 build the hankel matrix @xmath426 from @xmath425 and preform the svd .
let @xmath427 be its singular values , @xmath428 set @xmath233 to be a small positive number .
enforce the rank @xmath20 of @xmath426 by setting the @xmath429 smallest singular values to zero .
enforce the toeplitz structure on @xmath426 by averaging the entries along the diagonals .
extract the denoised fourier data @xmath425 from the first column and the last row of @xmath426 denoised fourier data @xmath425 and hankel matrix @xmath418 the procedure of algorithm [ algo : cadzow ] is guaranteed to converge to a matrix which exhibits the desired two key properties @xcite .
the iterations stop whenever the ratio of the @xmath430-th singular value to the @xmath20-th one , falls below a predetermined threshold .
since algorithm [ algo : prony ] does not perform well when noise is big , we can combine the algorithm [ algo : prony ] and algorithm [ algo : cadzow ] to recover the fourier spectrum of @xmath8 and improve the performance . in our numerical example
, we choose @xmath386 .
in this section , we present a numerical example to illustrate the effectiveness and robustness of the proposed algorithms . let the filter @xmath431 so that @xmath432 .
@xmath116 is approximately gaussian on @xmath433.$ ] let the initial signal @xmath3 be a conjugate symmetric vector given by @xmath434 and @xmath435 so that @xmath436 .
the subsampling factor @xmath20 is set to be 5 .
given the fourier data of the spatiotemporal samples @xmath437 , we add independent uniform distributed noise @xmath438 to the fourier data @xmath439 for @xmath440 . recall that @xmath441 , we define the relative error @xmath442 for @xmath443 . the best case error is set to be @xmath444 and the worst case error is set to be @xmath445 besides , we define the mean square error @xmath446 then we apply our algorithm [ algo : prony ] , algorithms [ algo : prony]+algorithm [ algo : cadzow ] , algorithm [ algo : matrixpencil ] and algorithm [ algo : esprit ] to the case when @xmath447 . for several parameters @xmath88 and @xmath379 , the resulting errors ( average over 100 experiments ) are presented in table [ aggiungi ] . as the bound @xmath233 in the algorithms we use @xmath448 .
it is shown in the table that increasing the temporal samples , i.e. @xmath88 , will help reduce the error .
the new proposed algorithms have better performance than algorithm [ algo : prony ] , if given more spatiotemporal data .
.numerical results [ cols="^,^,^,^,^,^,^",options="header " , ]
in this paper , we investigate the conditions under which we can recover a typical low pass convolution filter @xmath2 and a vector @xmath15 from the combined regular subsampled version of the vector @xmath449 defined in @xmath141 , where @xmath16 .
we show that if one doubles the amount of temporal samples needed in @xcite to recover the signal propagated by a known filter , one can almost surely solve the problem even if the filter is unknown .
we firstly propose an algorithm based on the classical prony method to recover the finite impulse response filter and signal , if an upper bound for their support is known .
in particular , we have done a first order perturbation analysis and the estimates are formulated in very simple geometric terms involving fourier spectral function of @xmath450 and @xmath20 , shedding some light on the structure of the problem .
we get a lower bound estimation for infinity norm of @xmath364 in terms of @xmath451 then we propose several other algorithms , which can make use of more temporal samples and increase the robustness to noise .
the potential applications includes the one - chip sensing : sensors inside chips for accurate measurements of voltages , currents , and temperature ( e.g. , avoid overheating any area of the chip ) , sources localization of an evolving state and time - space trade off(e.g .
, sound field acquisition using microphones)etc .
the author would like to thank akram aldroubi for his endless encouragements and very insightful comments in the process of creating this work .
the author is also indebted to ilya kristhal and yang wang for their helpful discussions related to this work .
a. aldroubi , j. davis and i. krishtal , _ exact reconstruction of signals in evolutionary systems via spatiotemporal trade - off_. journal of fourier analysis and applications , volume 21 , issue 1 , february 2015 , pp .
1131 .
a. aldroubi , i. krishtal and e. weber , _ finite dimensional dynamical sampling : an overview_. excursions in harmonic analysis .
vol 3 , appl .
anal . , birkhuser / springer , new york , 2015(r .
balan , m. begue , j. benedetto , w. czaja and k. okodujou , editors ) . to appear . j. christensen and g. lafsson , _ sampling in spaces of bandlimited functions on commutative spaces_. in applied and numerical harmonic analysis book excursions in harmonic analysis , volume 1 , 2013 , pp .
3569 .
r. roy , a. paulraj , and t. kailath , _ esprit - a subspace rotation ap- proach to estimation of parameters of cisoids in noise _ , ieee trans .
speech , signal process .
34 * , no .
1986 , pp . 13401342 .
r. aceska and s. tang , _ dynamical sampling in hybrid shift invariant spaces_. in operator methods in wavelets , tilings , and frame(v .
furst , k. kornelsen , e. weber , editors ) , contemporary mathematics , american mathematics society , providence , vol 626(2014 ) . to appear .
r. aceska , a. aldroubi , j. davis and a. petrosyan , _ dynamical sampling in shift invariant spaces_. in commutative and noncommutative harmonic analysis and applications ( a. mayeli , a. iosevich , p. jorgensen , g. lafsson , editors ) .
contemporary mathematics , american mathematical society , providence , vol 603(2013 ) , pp . 139148 .
g. reise , g. matz , and k. grchenig , _ distributed field reconstruction in wireless sensor networks based on hybrid shift - invariant spaces _ , signal processing , ieee transactions on * 60 * ( 2012 ) , no .
10 , pp .
54265439 .
yingbo hua and t. k. sarar , _ matrix pencil method for estimating parameters of exponentially damped/ undamped sinusoids in noise _ , ieee trans .
speech , signal process .
, vol . * 38 * , may .
1990 , pp . 814824 .
g. reise and g. matz , _ reconstruction of time - varying fields in wireless sensor networks using shift - invariant spaces : interative algorithms and impact of sensor localization errors _ , proc .
spawc , marrakech , morocco , ( 2010 ) , pp . 1 - 5
. y. lu and m. vetterli , _ spatial super - resolution of a diffusion field by temporal oversampling in sensor networks_. in acoustics , speech and signal processing , 2009 .
icassp 2009 .
ieee international conference on , april 2009 , pp .
2249 2252 .
a. hormati , o. roy , y. lu and m. vetterli , _ distributed sampling of signals linked by sparse filtering : theory and applications_. ieee transactions on signal processing , vol .
3 , march 2010,pp .
10951109 .
r. badeau , b. david and g. richard , _ high - resolution spectral analysis of mixtures of complex exponentials modulated by polynomials_. ieee transactions on signal processing , vol54 , no .
4 , april 2006 , pp 13411350 .
w. gautschi , _ how ( un)stable are vandermonde systems_. asymptotic and computational analysis ( winnipeg , mb , 1989 ) lecture notes in pure and appl . math .
124 , dekker , new york , 1990 , pp . | we consider the problem of spatiotemporal sampling in a discrete infinite dimensional spatially invariant evolutionary process @xmath0 to recover an unknown convolution operator @xmath1 given by a filter @xmath2 and an unknown initial state @xmath3 modeled as avector in @xmath4 .
traditionally , under appropriate hypotheses , any @xmath3 can be recovered from its samples on @xmath5 and @xmath1 can be recovered by the classical techniques of deconvolution . in this paper
, we will exploit the spatiotemporal correlation and propose a new spatiotemporal sampling scheme to recover @xmath1 and @xmath3 that allows to sample the evolving states @xmath6 on a sub - lattice of @xmath5 , and thus achieve the spatiotemporal trade off .
the spatiotemporal trade off is motivated by several industrial applications @xcite .
specifically , we show that @xmath7 contains enough information to recover a typical low pass filter " @xmath8 and @xmath3 almost surely , in which we generalize the idea of the finite dimensional case in @xcite .
in particular , we provide an algorithm based on a generalized prony method for the case when both @xmath8 and @xmath3 are of finite impulse response and an upper bound of their support is known .
we also perform the perturbation analysis based on the spectral properties of the operator @xmath1 and initial state @xmath3 , and verify them by several numerical experiments .
finally , we provide several other numerical methods to stabilize the method and numerical example shows the improvement .
+ + |
according to standard cosmology , neutrinos are probably one of the most abundant particles of the universe .
the universe is filled with a sea of relic neutrinos that decoupled from the rest of the matter within the first few seconds after the big bang .
it is excessively difficult to measure relic neutrinos since the interactions of their cross sections with matter are tremendously suppressed .
besides , it is crucial to detect relic neutrinos in order to test the neutrino aspects of the big bang model of cosmology but it would seem impossible with present methods . however , some indirect evidences of the relic sea may be observed . for example , weiler @xcite have shown that the uhe cosmic neutrinos may interact with relic neutrinos via the following reactions occurring on the z resonance : @xmath1 in such an event , an uhe cosmic neutrino has energy @xmath2 ev .
therefore , the interaction of relic neutrinos and uhe cosmic neutrinos would have significant cross section .
the high energy photon - neutrino interactions are very important in astrophysics , high energy cosmic ray physics and cosmology . from yang s theorem @xcite
the leading term of the cross section for the @xmath0 process is very small due to the vector - axial vector nature of the weak coupling when the neutrinos are massless .
it is shown that @xmath3 , where @xmath4 is the electron mass and @xmath5 is the photon energy in the center of the mass frame where the cross section for the @xmath0 process is in the order of @xmath6 and @xmath7 is the w boson mass @xcite .
the dimension-8 effective lagrangian for the photon - neutrino interaction in standard model ( sm ) is as follows @xcite @xmath8f_{\mu\lambda}f^{\nu\lambda } \label{l1}\end{aligned}\ ] ] where @xmath9 is the neutrino field , @xmath10 is the electroweak gauge coupling , @xmath11 is the photon field tensor , @xmath12 is the fine structure constant and @xmath13 is the following @xmath14.\end{aligned}\ ] ] the equation ( [ l1 ] ) can be rewritten as following format @xcite , @xmath15 here @xmath16 and @xmath17 are the stress - energy tensor of the neutrinos and photons which are given below , @xmath18,\end{aligned}\ ] ] @xmath19 for the sm , the photons and neutrinos decouple , i.e. , @xmath0 process at a temperature @xmath20 gev within one micro second after the big bang @xcite .
when decoupling temperature is reduced to the qcd phase transition ( @xmath21 mev ) , some remnants of the photons circular polarization can possibly be retained in the cosmic microwave background @xcite which can be considered as an evidence for the relic neutrino background . for reducing the decoupling temperature , the cross section for the @xmath22 process should be increased .
this can be done via the models which are beyond the sm .
for instance , contribution of large extra dimensions to these process have been calculated in ref .
they have shown that the inclusion of the extra dimension effects did not provide large enough high energy neutrinos to scatter from relic neutrinos in this process , but concluded that the photon decoupling temperature can be significantly reduced . also , in ref .
@xcite , it has been remarked that unparticle physics can lower decouple temperature below the @xmath23 .
the sm has been successful in describing the physics of the electroweak interactions and , it is consistent with experiments .
however , some questions are still left unanswered , such as , the number of fermion generation and fermion mass spectrum have not been exhibited by the sm .
attractive explanations are provided by models assuming composite quarks and leptons .
the existence of excited states of the leptons and quarks is a natural consequence of these models and their discovery would provide convincing evidence of a new scale of matter . in this model , charged and neutral leptons can be considered as a heavy lepton sharing leptonic quantum number with the corresponding sm lepton .
they should be regarded as the composite structures which are made up of more fundamental constituents .
therefore , excited neutrinos can be considered to spin-@xmath24 bound states or , including three spin-@xmath24 or spin-@xmath24 and spin @xmath25 substructure .
all composite models have an underlying substructure which is characterized by a scale @xmath26 .
the interaction between spin @xmath24 excited fermions , gauge bosons and the sm fermions can be described by the @xmath27 invariant effective lagrangian as follows @xcite , @xmath28 in these expressions , @xmath29 with @xmath30 being the dirac matrices , @xmath31 and @xmath32 are the field strength tensors of the @xmath33 and @xmath34 , @xmath35 and @xmath36 are the generators of the corresponding gauge group , @xmath10 and @xmath37 are standard electroweak and strong gauge couplings .
@xmath26 is the scale of the new physics responsible for the existence of excited neutrinos and @xmath38 , @xmath39 scale the @xmath33 and @xmath34 couplings , respectively .
the effective lagrangian can be rewritten in the physical basis , @xmath40 first term in the above equation is a purely diagonal term with @xmath41 and second term is a non - abelien part , which involves triple as well as quartic vertices with @xmath42 @xmath43 @xmath44 @xmath45 the chiral @xmath46 ( @xmath47 ) interaction term can be found as follows @xmath48 where @xmath49 is the momentum of the gauge boson , @xmath50 is the electroweak coupling parameter , @xmath51 is defined for photon by @xmath52 where we have assumed that @xmath53 . up to now , searches have not found any signal for excited neutrinos at the colliders .
the current mass limits on excited neutrinos are @xmath54 gev at the lep @xcite and @xmath55 gev assuming @xmath56 at the hera @xcite .
excited neutrinos have been also studied for hadron colliders @xcite and next linear colliders @xcite . in these studies
, it has been obtained that excited neutrinos masses up to @xmath57 tev can be detected at the lhc . in this paper
, we examine the effect of the excited neutrinos on the interaction of the uhe cosmic and relic neutrinos for the @xmath0 process .
the sm contribution to @xmath0 process have been calculated in refs.@xcite with using equation ( [ l1 ] ) . from this effective lagrangian
, the squared amplitude for the sm can be obtained in terms of mandelstam invariants u and t as follows , @xmath58 . the new physics ( np ) contribution comes from t and u channels excited neutrino exchange . the analytical expressions for the polarization summed amplitudes square for np , sm and np interference terms are given below , @xmath59 @xmath60 where @xmath61 .
therefore , the whole squared amplitude can be calculated as follows , @xmath62 . in the limit @xmath63
, the amplitude turns into following formation , @xmath64 we have calculated the cross section with / without approximation cross section .
we have seen that results are closely for different @xmath65 and @xmath66 .
therefore , the differential cross section for @xmath0 process can be obtained by using @xmath67 then , the total cross section can be found from the equation ( [ dcs ] ) as follows , @xmath68 in fig .
( [ fig1 ] ) , we have plotted the total cross sections as a function of the center of mass energy @xmath69 for both the @xmath70 and total cross sections when @xmath71 gev .
these cross sections are obtained for the two excited neutrinos , with masses of @xmath72 gev , @xmath73 gev . also , in fig.([fig2 ] ) we have showed the cross sections for @xmath72 gev and three scale of the new physics : @xmath71 gev , @xmath74 gev and @xmath75 gev .
this figure shows similar behavior with fig .
( [ fig1 ] ) . extra contribution to the @xmath76 cross section from excited neutrino exchange influences the decoupling temperature .
the temperature at which the this process ceases to take place can be found from the the reaction rate per unit volume , @xmath77 the terms in equation ( [ roo ] ) are as following : @xmath78 and @xmath79 are the momentum of the neutrino and antineutrino ; @xmath80 and @xmath81 their energies ; t is the temperature ; @xmath82 is the flux .
the @xmath83 can be given in terms of @xmath84 in the center of mass frame by using of invariance of @xmath85 @xmath86 @xmath87 where @xmath88 and @xmath89 is the angle between @xmath78 and @xmath79 .
equation ( [ roo ] ) can be found @xmath90 where @xmath91 and @xmath92
. then the reaction rate per unit volume has been obtained , @xmath93^{2}\end{aligned}\ ] ] where @xmath94 is the riemann zeta function .
the interaction rate @xmath95 is obtained by dividing @xmath96 by the neutrino density @xmath97 at temperature @xmath98 .
thus we have found , @xmath99 multiplying equation ( [ rr ] ) by the age of the universe , @xmath100 at least one interaction to occur is @xmath101 .
the solution of the following equation gives the decoupling temperature , @xmath102 if @xmath103 and @xmath104 are replaced in the above equation , then following equation can be found , @xmath105 fig .
( [ fig3 ] ) shows solution of the this equation .
we have analyzed the contribution of excited neutrinos on the interaction of relic neutrinos with uhe cosmic neutrinos via the @xmath76 process .
it is shown that excited neutrino contribution to total cross section of the @xmath76 process is significant depending on the @xmath106 , @xmath107 .
we have seen that for the appropriate values of these parameters , the sm and total cross sections can be distinguished from each other in the specific center of mass energy regions . for decreasing decoupling temperature , the total cross section of the @xmath76 should be increased . if the @xmath65 or new physics parameter @xmath108 decrease then the total cross section increases .
therefore , @xmath109 can be decreased significantly . for different values of @xmath65 ,
@xmath109 have been shown in fig.([fig3 ] ) as a function of the new physics parameter @xmath108 . as seen from this figure
, our obtained values of the decoupling temperature can decrease the value of the sm decoupling temperature ( @xmath110 gev ) . | we examine the effect of excited neutrinos on the annihilation of relic neutrinos with ultra high energy cosmic neutrinos for the @xmath0 process .
the contribution of the excited neutrinos to the neutrino - photon decoupling temperature are calculated .
we see that photon - neutrino decoupling temperature can be significantly reduced below the obtained value of the standard model with the impact of excited neutrinos . |
imagine a sequence of worlds in which the light quark masses are continuously varied away from their physical values .
it has been argued that heavy nuclei will disassociate for a @xmath5 increase in the sum @xmath6
one can anticipate that before this happens , the deuteron will become unbound .
this is because the deuteron binding energy is less than the average binding energy per nucleon . here , we study this problem quantitatively and provide details about how the deuteron would respond to variations in its potential induced by changes in @xmath6 .
the deuteron is an example of a hadronic molecule .
such a two - hadron composite , bound by the strong interactions and whose mass lies slightly below the associated two - hadron threshold , should be describable with a nonrelativistic schrdinger equation .
we can picture the two hadrons as interacting via a meson exchange potential @xcite .
hadronic , as opposed to quark , degrees of freedom should be appropriate for these situations . in this paper
, we consider the @xmath7-wave radial schrdinger equation , [ schr ] u(r ) = 0 , where @xmath8 is the reduced mass of the two hadrons and @xmath9 . in section ii
, our focus will be on the form of the potential energy function @xmath10 , especially as regards the contributions from multi - pion exchanges . in section iii ,
we carry out a numerical study of how varying the potential energy affects deuteron binding .
our conclusions and plans for future study are presented in section iv .
although the deuteron state , and more generally the two - nucleon potential , are long - studied areas of research , they continue to attract theoretical attention .
there has been much recent activity about how the nuclear potential behaves in the chiral limit of zero light quark mass .
this work has motivated our interest in deuteron binding . by design
our description is simple , _
e.g. _ we omit tensor interactions and thus our ` deuteron ' has no d - wave component @xcite .
we comment further on this aspect of our work in the conclusion .
spatial potential energy functions for nonrelativistic calculations are obtained from particle exchange diagrams in quantum field theory .
a nonrelativistic reduction of the exchange amplitude is followed by fourier transformation from momentum to position space ( see eqs . ( [ vq])-([vr ] ) below ) . for example , photon exchange between an electron and a proton leads to a number of separate effects ( collectively the _ breit - fermi _ interaction @xcite ) starting with the long range coulomb potential .
in addition to generating long range potentials , such exchange amplitudes also generally give rise to short range or even local contributions . in this paper , we restrict our attention to only the long range part . for convenience , we shall cast the potential energy in a dispersive form @xcite .
the strongest components of the two - nucleon potential arise from the isoscalar ( @xmath11 ) two - pion and three - pion exchange channels .
a smaller isovector two - pion component must also be included . for isoscalar ( i=0 ) potentials ,
the two - pion ( scalar ) exchange gives rise to an attractive interaction while the three - pion ( vector ) interaction is repulsive .
a momentum space version is given in terms of a corresponding spectral function @xmath12 , v_i^(i=0 ) ( q^2 ) = 1 _ _
i^2^ d ^2 _ i^(i=0 ) ( ) ^2 + q^2 , [ vq ] where @xmath13 and @xmath14 .
the spatial potential energie @xmath15 corresponding to eq .
( [ vq ] ) is v_i^(i=0 ) ( r ) = 1 4 ^2 r _ _ i^2^ d ^2 e^-r _ i^(i=0 ) ( ) .
[ vr ] the above unsubtracted dispersive forms are taken from refs . , where behavior at relatively low energies ( @xmath16 gev ) is studied .
the issue of high energy behavior and subtractions is considered in ref .
however , the choice of a subtraction scale at sufficiently high energy is not expected to alter the basic conclusions of refs . , which are anchored by the constraints of chiral perturbation theory @xcite .
for the isoscalar vector channel , it is a good approximation to take @xmath17 , where @xmath18 mev .
this is also appropriate for the isovector two - pion exchange channel , where now @xmath19 , with @xmath20 mev . in these two cases ,
the association between an exchanged vector meson of mass @xmath21 and a spatial yukawa potential @xmath22 is natural and obvious .
the isoscalar two - pion exchange channel is more subtle .
the nucleon - nucleon potential is known to have a strong attractive contribution of intermediate range .
this arises from an exchange potential in the scalar , isoscalar channel , due to the exchange of a pair of pions in the @xmath24 channel .
there is now a description of this which , although not explicitly given in terms of quark interactions , does nonetheless respect the underlying qcd dynamics .
the approach is ( i ) to use chiral perturbation theory to describe the low energy part of the momentum - space potential , and ( ii ) to invoke two - pion rescattering via the omnes function at higher energies in order to include the constraint of unitarity .
the chiral component of @xmath25 turns out , as is generally the case in chiral perturbation theory , to depend on a small number of low energy constants whose numerical values are obtained phenomenologically . to the extent that these constants depend on any meson , it is actually @xmath26 that contributes
the chiral form of @xmath25 grows with energy and at some point requires modification in order to respect unitarity .
such a modification can be accomplished @xcite by introducing the omnes function , & & ^(i=0 ) ( ) = , [ omnes ] where @xmath27 is the s - wave two - pion phase shift in the @xmath11 channel . in this manner ,
the two ingredients of chiral perturbation theory and two - pion rescattering provide a sound theoretical basis for understanding the two - pion @xmath28 exchange potential . in the intermediate energy range @xmath29 mev , the @xmath28 two - pion phase shift does not pass through @xmath30 and thus ( unlike @xmath2 and @xmath26 ) has no obvious association with a resonant state . the dispersive spatial potential of eq .
( [ vr ] ) would seem to require a broad spectral function in this case .
however , it can be shown ( see especially fig .
3(b ) of ref .
@xcite ) that _ as a numerical recipe _
the effect of nonresonant @xmath28 two - pion scattering is reproduced by using @xmath31 with @xmath32 mev . in our work ,
we adopt this approach and refer to it hereafter as the ` @xmath1 exchange potential ' .
we stress that our use of ` @xmath1 ' is simply a convenient shorthand and not meant to commit us to any specific model of this state @xcite .
let us now consider the two - nucleon interaction in terms of four exchange potentials , [ deutpotl ] v^(nn ) = v_^(nn ) + v_^(nn ) + v_^(nn ) + v_^(nn ) .
each of these is taken to have the yukawa form , [ yuk ] v^(nn)(r ) = _ i _ i g_i^24 e^- m_i r r , where @xmath33 for repulsion and @xmath34 for attraction .
all that remains is to fix the strength of each potential . for detailed discussions of potential energy fits to nuclear binding , see refs .
. we will assume the following values for the @xmath35 and @xmath36 couplings @xcite , [ sig - omeg ] g_^2 4 16.6 10.3 12.9 .
our strategy will be to fit @xmath37 to the experimental deuteron binding energy @xcite [ bed ] be_deut = 2.22457 mev and see how it compares to the range in eq .
( [ sig - omeg ] ) .
we use this approach because a precise fit will allow us to study the effect on @xmath38 of modifying the potential energy function .
the pion potential energy is obtained by starting from the field theory interaction with pseudoscalar coupling [ ps ] l _ = i f _ |n _ 5 n and reducing the pion exchange graph to its nonrelativistic limit . upon keeping only the yukawa dependence of eq .
( [ yuk ] ) , one obtains the isospin - dependent , spin - dependent potential energy [ relate ] v_^(nn ) ( r ) = f^2 _ 4 ( m_2 m_n ) ^2 ( 2 * i*^2 - 3 ) e^- m_r r . for the deuteron channel ( @xmath39 , @xmath40 ) this reduces to [ gpi ] v_(r)^(nn ) = - g_^24 e^- m_r r , with [ pioncoup ] g_^2 f^2 _
( m_2 m_n ) ^2 0.073 .
observe that @xmath41 is tiny relative to @xmath42 and @xmath43 . for the @xmath26 interaction
the lagrangian is [ vnn ] l_nn = g _ |n _ ^n . where we take @xmath44 @xcite .
the @xmath3 component of the potential energy is isospin dependent , [ vrho ] v_^(nn ) ( r ) = g^2 _ 4 ( 2 * i*^2 - 3 ) e^- m_r r . of the four component potentials , only the @xmath2 piece is repulsive in the @xmath11 channel .
it is mainly the interplay of the large @xmath2 repulsion and @xmath1 attraction that results in the deuteron binding .
although the coefficient of the @xmath2 potential is larger than that of the attractive @xmath1 , the lighter mass of the @xmath1 allows it to dominate over @xmath2 at intermediate distance scales .
the @xmath2 provides the repulsive core . the potential energy which includes all four components is displayed in fig .
[ fig : deut - b ] .
when we use as input potential energies @xmath45 , @xmath46 , @xmath47 and @xmath48 to fit the value of @xmath49 to the deuteron binding energy , the value obtained for @xmath42 is [ sigma ] g_^2 4 10.83 .
this is in accord with the range given above in eq .
( [ sig - omeg ] ) .
the two - nucleon system has isospins @xmath50 and spins @xmath51 , and if the two nucleons are in an s - wave , the @xmath24 and @xmath52 configurations are ruled out by fermi - dirac statistics .
the deuteron has @xmath11 , @xmath53 , which leaves finally the combination @xmath54 .
the @xmath54 channel has a virtual bound state , as evidenced by the large scattering length @xmath55 fm .
since @xmath1 and @xmath2 are isoscalars , a potential energy with only these exchanges would imply equal binding in both the @xmath56 and @xmath54 channels .
however , the @xmath3 contribution is attractive in the deuteron channel but repulsive ( at one third the strength ) in the @xmath54 channel .
our numerical study obtains the desired result of no binding for the @xmath54 channel upon using the @xmath1 coupling strength of eq .
( [ sigma ] ) .
anthropics can be viewed as relating to the class of theories ( including possibly string theory and chaotic inflation ) in which spacetime has a domain structure , each domain with its own set of fundamental parameters ( ckm mixing angles , lepton and quark masses , _ etc _ ) @xcite .
such a description is often referred to as the _ multiverse _ @xcite .
this theoretical possibility provides an interesting framework for viewing deep issues of physics , such as the cosmological constant @xcite , the higgs vacuum expectation value @xcite and the light quark masses @xcite .
previous studies have revealed that even a modest variation of fundamental parameters can induce qualitative changes in physical systems @xcite .
works like these can help determine the ` physical ' windows in parameter space which enable universes like our own to exist .
this is in contrast with ` theoretical ' parameter windows which can be associated with a given underlying theory , _
e.g. _ string dynamics . if the physical window is small compared to the theoretical window , then it is not unreasonable to assign a flat probability distribution across the physical window and thus to study the relative liklihood of that our universe should exist within a given theoretical framework @xcite .
.[tab : rstab ] dependence of deuteron binding energy on the @xmath57 coupling as expressed directly ( first two columns ) or in terms of the ratios @xmath58 , @xmath59 defined in eq .
( [ ratio ] ) . [ cols="^,^,^,^",options="header " , ] in this paper , we are concerned with the light quark masses
@xcite show that increasing @xmath6 decreases the @xmath57 coupling whereas decreasing @xmath6 has the opposite effect .
for example , it is estimated that taking @xmath60 produces a @xmath61 increase in @xmath49 while increasing @xmath62 to twice its physical value decreases @xmath49 by @xmath63 .
the source of this sensitivity to @xmath64 ( and ultimately to @xmath6 ) lies mainly in the threshold contribution in the dispersive integrals of eqs .
( [ vq]),([vr ] ) .
by contrast , the @xmath65nn and @xmath66nn couplings are relatively unaffected by such changes in the light quark masses because their spectral functions are large only well above threshold . 1
. _ unbinding of the deuteron via weakening of @xmath45 _ : in our description , the two - nucleon potential is mainly the competition between the attractive @xmath67 and the repulsive @xmath68 .
it is not hard to upset this balance and thus unbind the deuteron by varying some combination of the potential energies , _
e.g. _ by reducing @xmath45 ( keeping all other interactions fixed ) , increasing @xmath69 ( again keeping all other interactions fixed ) , _
etc_. however , the former is the only variation associated with the mass values of the light quarks . to help keep track of how the deuteron binding energy responds to the variation in @xmath70 , we find it useful to define the following ratios , [ ratio ] r^(coup)[g_nn ] r^(be)[be_deut ] . from the numerical values in table [ tab : rstab ] , we see that the deuteron becomes unbound if @xmath45 is reduced in strength by about @xmath4 .
this is to be compared with the @xmath71 required to disassociate heavy nuclei @xcite .
2 . _ alternative variations _ : there can , in principle , be other parameter variations . for the sake of completeness
, we briefly consider a few of these .
it is obvious that increasing @xmath45 will increase the deuteron binding energy .
however , there is another consequence , which turns out to somewhat alter the physical world .
we find that an increase of about @xmath72 in @xmath45 , with other potentials held fixed , suffices to first produce binding the @xmath54 nn channel .
varying the @xmath66 and @xmath65 potentials would likewise have physical consequences , _
e.g. _ with @xmath45 fixed , an increase in strength of @xmath69 by about @xmath73 would result in unbinding the deuteron , this kind of variation is not associated in any obvious way with the fundamental parameters of the standard model and thus has no connections with our anthropic analysis .
our aim in this paper has been to explore the dependence of deuteron binding on variations in @xmath6 in terms of a simple model involving just central potentials .
the central potentials were constructed with care , by exploiting up - to - date results from studies of nuclear binding .
this description is admittedly in contrast with the traditional approach which uses the one - pion - exchange ( ope ) tensor potential @xcite .
we feel that these two procedures are not incompatible .
consider a description containing the huge @xmath74 and @xmath66 potentials along with the tensor ope .
if the @xmath74 and @xmath66 effects cancel even more than in our calculation ( which would involve a slight change in @xmath67 ) , the deuteron binding would be given largely in terms of the tensor ope .
even so , deuteron binding would still feel the changes induced by @xmath6 largely via the effect on @xmath67 ( as described in this paper ) .
we leave for future study a more ambitious description which includes all these elements in a schrdinger equation context , plus insights from chiral perturbation theory .
thus , given the deuteron potential energies of eq .
( [ deutpotl ] ) , we have shown ( _ cf _ table [ tab : rstab ] ) that a slight reduction in @xmath67 results in unbinding the deuteron .
even a slight weakening of the attractive component results in a large effect . as a result
, the deuteron is a sensitive system for studying how physics would respond to changes in @xmath6 .
the crucial links in the chain of logic which connects these are ( i ) the chiral relation , @xmath75 , which relates the pion mass to the light quark masses and ( ii ) the dispersive formula eq .
( [ vq ] ) which relates @xmath67 to the pion mass @xcite .
the @xmath4 decrease in @xmath67 needed to unbind deuterons is less than the estimated @xmath71 needed to unbind heavy nuclei @xcite .
it is abundantly clear that if conditions forbade atoms or nuclei from existing , then there would be no universe as we know it .
the absence of the deuteron , although itself just a solitary quantum particle , would likewise be dramatic .
the deuteron first appears in cosmology as part of big bang nucleosynthesis ( bbn ) in the early universe , [ bbn0 ] p + n d + . before deuteron formation ,
free neutrons are copious ( about one neutron per five protons ) , but after formation free neutrons are absent ( the number of free neutrons equals the number of deuterons at about 200 seconds after the big bang ) .
once deuterons are formed , they take part in a number of additional bbn reactions , such as fusion with protons and neutrons , [ bbn1 ] d + p ^3he + d + n ^3h + , or with each other , [ bbn2 ] d + d ^3h + p d + d ^3he + n .
these lead to the production of @xmath76 , [ bbn3 ] although trace amounts of heavier nuclei are formed , the above relations display the basic yield of bbn once the deuteron is formed .
it is clear that , although just a solitary particle , the deuteron occupies a central role in bbn .
there are some additional issues , not yet mentioned , which we leave for further study . as regards bbn , it would be interesting to study more carefully what happens under conditions where the deuteron becomes slightly unbound . in a different direction , by using a modification of the meson exchange approach described in this paper , we can consider the possibility of molecular structures in the meson sector
. of particular interest are the states @xmath77 and @xmath78 , whose nearness to the @xmath79 threshold has suggested a molecular interpretation @xcite .
we will report on this study elsewhere @xcite .
this work was supported in part by the u.s .
national science foundation under grant phy0555304 .
we thank g. colon and e. thompson for help with the figure .
discussions with john donoghue were useful , informative , and above all , interesting .
the author thanks daniel phillips for his critical review of the manuscript and various insightful remarks .
for other works in this area , see also s. r. beane and m. j. savage , nucl .
a * 713 * , 148 ( 2003 ) [ arxiv : hep - ph/0206113 ] ; v. v. flambaum and r. b. wiringa , phys .
c * 76 * , 054002 ( 2007 ) [ arxiv:0709.0077 [ nucl - th ] ] .
two classics which describe dynamics based on meson exchange potentials are j. j. sakurai , annals phys . * 11 * , 1 ( 1960 ) for particle physics and r. v. reid , annals phys . * 50 * , 411 ( 1968 ) for nuclear physics .
see e. epelbaum , u. g. meissner and w. gloeckle , nucl .
a * 714 * , 535 ( 2003 ) [ arxiv : nucl - th/0207089 ] and references cited therein .
s. r. beane and m. j. savage , nucl .
a * 717 * , 91 ( 2003 ) [ arxiv : nucl - th/0208021 ] .
j. f. donoghue , phys .
b * 643 * , 165 ( 2006 ) [ arxiv : nucl - th/0602074 ] ; for an instructive reference , see t. e. o. ericson and m. rosa - clot , ann .
nucl . part .
sci . * 35 * , 271 ( 1985 ) .
w. n. cottingham , r. vinh mau , phys .
* 130 * , 735 ( 1963 ) ; w. n. cottingham , m. lacombe , b. loiseau , j. m. richard and r. vinh mau , phys .
d * 8 * , 800 ( 1973 ) ; m. chemtob , j. w. durso and d. o. riska , nucl .
b * 38 * , 141 ( 1972 ) .
j. f. donoghue , phys .
c * 74 * , 024002 ( 2006 ) [ arxiv : nucl - th/0603016 ] .
a. adams , a. jenkins and d. oconnell , arxiv:0802.4081 [ hep - ph ] .
j. f. donoghue , to appear .
g. colangelo , j. gasser and h. leutwyler , nucl .
b * 603 * , 125 ( 2001 ) [ arxiv : hep - ph/0103088 ] .
for an example where the @xmath74 is treated as a physical meson , see y. b. ding _ et al . _ , j. phys .
g * 30 * ( 2004 ) 841 [ arxiv : hep - ph/0402109 ] .
for example , see d. jido , e. oset and j. e. palomar , nucl .
a * 709 * , 345 ( 2002 ) [ arxiv : nucl - th/0202070 ] . for a book - length treatment , see l. susskind , _ the cosmic landspace : string theory and the illusion of intelligent design _ , ( little , brown and company , new york ny 2005 ) .
j. donoghue , _ the fine - tuning problems of particle physics and anthropic mechanisms _ , pp . 231 - 246 , contribution to _ universe or multiverse _ , b. j. carr ( ed . ) , ( cambridge university press , cambridge , england 2007 ) .
s. weinberg , phys .
lett . * 59 * , 2607 ( 1987 ) .
v. agrawal , s. m. barr , j. f. donoghue and d. seckel , phys .
lett . * 80 * , 1822 ( 1998 ) [ arxiv : hep - ph/9801253 ] ; phys .
d * 57 * , 5480 ( 1998 ) [ arxiv : hep - ph/9707380 ] . | we consider a description of the deuteron based on meson exchange potentials .
a key feature is the inclusion of the @xmath0 two - pion intermediate state ( ` @xmath1 ' ) as a significant component of the inter - nucleon potential energy . in this approach ,
deuteron binding is seen to be predominantly a consequence of @xmath1 and @xmath2 exchange , with a secondary role played by @xmath3 .
we explore sensitivity of two - nucleon binding to changes in the potential and thereby obtain an anthropic constraint that the deuteron unbinds for a modest decrease ( about @xmath4 ) in the attractive @xmath1 potential . |
the discovery of quark - hadron duality in electron - nucleon interactions @xcite rises a natural question if the same phenomenon can be seen also in neutrino - nucleon scattering .
unfortunately the available data is not yet precise enough to discuss the problem on the experimental level . at present
the only possibility is to analyze the existing theoretical models of resonance production .
this approach was adopted by sato and lee @xcite who discovered that for their model of @xmath4 production the quark - hadron duality is seen : the resonance peaks of structure functions calculated at @xmath5 gev@xmath3 slide along the dis structure functions ( with cteq6 pdfs ) calculated at @xmath6 gev@xmath3 , both as functions of the nachtmann variable . in this contribution
we wish to report an investigation of the rein - sehgal ( rs ) model @xcite of resonance production .
the rs model is used by almost all monte carlo generators of neutrino interactions and understanding its properties is of practical value .
the model describes resonance production in the region of hadronic invariant mass @xmath7 below @xmath8 gev by summing contributions from 18 resonances .
it contains also a non - resonant background fine tuned in order to get a good agreement with the existing single pion production ( spp ) data .
we calculate the structure functions @xmath9 as they are defined by the rs model @xcite .
they turn out to be linear combinations of cross sections for three polarization states of the intermediate @xmath7 boson .
the aim of the rs model is to describe only spp channels and the elasticity coefficients are used in order to select exclusive spp channels from the overall resonance production cross section .
the non - resonant background is adjusted only for spp channels .
the analysis of duality requires the knowledge of the inclusive cross section in the resonance region . in the spirit of the rs model
one should add contributions from inelastic channels taking care of complicated interference patterns . in order to make the analysis simpler we introduce 1@xmath10 functions which are defined as probabilities that in the given region of the kinematically allowed space the final state is that of spp .
we calculate them numerically for each spp channel separately using the lund fragmentation and hadronization routines .
all the details of our approach can be found in @xcite .
functions defined in eq . [ 1pion_def ] with resonance elasticity factors taken from pdg @xcite . in the case of @xmath11-neutron scattering
we show the sum of two functions corresponding to @xmath12 and @xmath13 reactions [ elasticity],width=302 ] it turns out that @xmath15 do not depend on @xmath16 .
their plots for cc @xmath17 channels are presented in fig .
[ elasticity ] . in the case of neutron
we show the sum of 1@xmath10 functions corresponding to two exclusive spp channels . on the same plot
we show also the available data on elasticity of @xmath18 and @xmath4 resonances .
we see that 1@xmath10 functions provide satisfactory average description .
we select the physically relevant region in the @xmath16 variable as @xmath28 gev@xmath3 by looking at double differential cross section @xmath29 for neutrino energy in a few gev region , see fig . 2 @xcite . having in mind mc generators of events
the quark - hadron duality should ensure the smooth passage of the cross section for res and dis contributions in the relevant kinematical transition region .
structure function for cc neutrino reactions on proton , neutron and isoscalar target .
@xmath1 for rs model is calculated at @xmath30 gev@xmath3 and for dis at @xmath6 gev@xmath3 .
[ slizganie],width=302 ] in fig .
[ slizganie ] we show @xmath1 structure functions for cc reactions defined by the rs model for three values of @xmath16 : @xmath31 gev@xmath3 and also dis @xmath1 structure functions calculated at @xmath32 gev@xmath3 .
the targets are from the top to the bottom : proton , neutron and their average isoscalar one .
the plots for the dis part are done with grv94 pdfs @xcite .
rs structure functions are yet not re - scaled by 1@xmath10 functions .
similar plots for @xmath1 at @xmath33 gev@xmath3 were shown before in @xcite .
we see that in the case of proton and isoscalar target typical manifestation of local quark - hadron duality : the @xmath4 peak slides along the dis curve .
the duality does not seem to apply neither to other resonances nor to neutron target even in the case of the prominent @xmath4 resonance . in fig .
[ 1pion_in_action ] we show the role played by 1@xmath10 functions . on the same plot
both rein - sehgal and 1@xmath10 function re - scaled res @xmath1 structure functions for cc reaction on neutron are presented .
we see that the modifications apply mostly to the region of @xmath7 close to 2 gev .
[ ratios ] presents the functions defined in eq .
[ ratio_q2=const ] for three different integration regions i.e. for three definitions of the resonance region : @xmath34 where @xmath35 gev .
we see that the choice @xmath36 gev makes the functions slowly varying in the wide region of @xmath16 .
the behavior for small values of @xmath37 gev@xmath3 is very different in agreement with the predictions made in @xcite .
for the proton target the best choice is @xmath36 gev while for other two targets it is preferable to choice the resonance region as more confined at the price of significant variations with @xmath16 .
the choice @xmath36 gev is a natural one for the rs model .
but it is suggested in @xcite that the rs model underestimates the cross section for @xmath38 gev .
we conclude that with the choice @xmath36 gev the duality holds very well for the proton and badly for other targets .
it seems to be difficult to have duality for all the targets simultaneously .
p. lipari , _ calculation of the neutrino cross section .
open problems , lines of research _ a talk given at _ third international workshop on neutrino - nucleus interactions in the few - gev region _
, assergi , march 17 - 21 , 2004 .
a. bodek , _ nuint02 conference summary : modelling quasi - elastic , resonance and inelastic neutrino and electron scattering on nucleons and nuclei _ , a talk at _ second international workshop on neutrino - nucleus interactions in the few - gev region _ , irvine , dec 12 - 15 , 2002 . | quark - hadron duality in neutrino - nucleon reactions is investigated under the assumption that cross sections in the resonance region are given by the rein - sehgal model .
the quantitative analysis of the duality is done by means of appropriate integrals of the structure functions in the nachtmann variable . we conclude that with the definition of the resonance region @xmath0 gev ) the duality holds for neutrino - proton reaction @xmath1 structure function for @xmath2 gev@xmath3 and it is absent for neutrino - neutron reaction . |
metric theories of gravity offer the singular beauty of endowing spacetime with a symmetric , second - rank tensor field @xmath1 that couples universally to all non - gravitational fields .
this unique operational geometry is dependent upon the validity of the einstein equivalence principle ( eep ) , which states that the outcomes of nongravitational test experiments performed within a local , freely falling frame are independent of the frame s location ( local position invariance , lpi ) and velocity ( local lorentz invariance , lli ) in a background gravitational field .
non metric theories break this universality by coupling additional gravitational fields to matter , and so violate either lpi , lli or both .
limits on lpi or lli are imposed by gravitational redshift and atomic physics experiments respectively , by comparing atomic energy transitions that are sensitive to these symmetries .
laser experiments have set stringent limits on violations of lli ( to a precision of @xmath2 ) @xcite , while the next generation of gravitational redshift experiments could reach a precision up to @xmath3 @xcite .
these experiments probe transitions that are predominantly sensitive to nuclear electrostatic energy , although violations of wep / eep due to other forms of energy ( virtually all of which are associated with baryonic matter ) have also been estimated @xcite . however there
exist many other physical systems , dominated by primarily non - baryonic energies , for which the validity of the eep is comparatively less well understood @xcite . included in the list of such systems
are photons of differing polarization @xcite , antimatter systems @xcite , neutrinos @xcite , mesons @xcite , massive leptons , hypothesized dark matter , second and third generation matter , and quantum vacuum energies . in order to establish the universal behavior of gravity , it is important to empirically confront the eep with as diverse a range of non - gravitational interactions as is possible . amongst the most intriguing of the list of physical systems noted above
are those for which potential violations of the eep can arise in quantities dependent upon vacuum energy shifts , which are peculiarly quantum - mechanical in origin ( _ i.e. _ do not have a classical or semi - classical description ) .
the empirical validity of the eep in physical systems where radiative corrections are non - negligible remains an open question .
such systems provide an interesting empirical regime for a confrontation between gravitation and quantum mechanics .
perhaps the most useful framework for further such testing of the eep is provided by quantum electrodynamics ( qed ) .
we considered this approach in a previous paper @xcite by analyzing the behavior of lamb shift transition energies within the context of a wide class of nonmetric theories of gravity .
the lamb shift , along with the anomalous magnetic moments ( @xmath0 factors ) of elementary particles , presents the most compelling evidence in support of qed .
it is therefore natural to extend the nonmetric analysis to the @xmath0 anomaly .
we consider in this paper the possibility of using measurements of anomalous magnetic moments of elementary particles as a possible test of the eep .
the high precision attained in @xmath0 experiments motivated the early work of newman _
@xcite to use such experiments to set new bounds on the validity of special relativity .
similarly , we expect ( and shall subsequently demonstrate ) that such experiments could impose stringent and qualitatively new limits on the parameter space of nonmetric theories .
our interpretation of those experiments is substantially different from theirs as we explicitly include violations of the eep in the computation of @xmath0 .
these effects were not considered in ref .
@xcite , which assumed that violations of special relativity arose only from the dynamical equation governing the motion of the fermion .
we consider the class of non - metric theories described by the formalism @xcite , following the approach given in ref .
@xcite in developing gravitationally modified ( gm ) qed .
this formalism encompasses a wide class of nonmetric theories of gravity , and deals with the dynamics of charged particles and electromagnetic fields in a static , spherically symmetric gravitational field .
it assumes that the ( classical ) non - gravitational laws of physics can be derived from an action : @xmath4 where @xmath5 , @xmath6 , and @xmath7 are the rest mass , charge , and world line of particle @xmath8 , @xmath9 , @xmath10 , @xmath11 , @xmath12 .
the metric is assumed to be [ tmumet ] ds^2 = t(r)dt^2 - h(r)(dr^2+r^2d^2 ) where @xmath13 , @xmath14 , @xmath15 , and @xmath16 are arbitrary functions of the ( background ) newtonian gravitational potential @xmath17 , which approaches unity as @xmath18 .
for an arbitrary non - metric theory , these functions will depend upon the type of matter , _
i.e. _ the species of particle or field coupling to gravity . the functions @xmath15 and @xmath16 parameterize the ` photon metric ' , whereas @xmath13 and @xmath14 parameterize the ` particle metric ' in the static , spherically symmetric case .
although we shall generically employ the notation @xmath13 and @xmath14 throughout this paper , it should be kept in mind that these functions shall in general have one set of values for electrons , another set for muons , another for protons , _
etc._. universality of gravitational coupling in the particle sector implies that the @xmath13 and @xmath14 functions are species independent .
it is an empirical question as to whether or not such universality holds for all particle species .
the stringent limits on universality violation set by previous experiments @xcite have only been with regards to the relative gravitational couplings in the baryon / photon sector of the standard model . for the leptonic sector relevant to our considerations , relatively little is known @xcite . in order to study atomic systems
, this classical action can be generalized for quantum mechanical systems @xcite , [ q1 ] s = d^4x ( i+e a - m ) + d^4x(e^2-c^2b^2 ) , where local natural units are used , @xmath19 , and @xmath20 with the subscript `` 0 '' denoting the functions evaluated at @xmath21 , the origin of the local frame of reference .
this action emerges upon replacing the point - particle part of the action in ( [ 1 ] ) with the dirac lagrangian , expanding the parameters about the origin , neglecting their spatial variation over atomic distance scales , and rescaling coordinates and fields .
these operations have the effect of treating spatial and temporal derivatives and gauge couplings in the fermionic sector on the same footing , and so we need not add additional terms of the form @xmath22 times an arbitrary constant , as such terms can be re - absorbed into a definition of the parameters which appear in ( [ q1 ] ) above , as noted below .
the action ( [ q1 ] ) refers to the preferred frame , as defined by the rest frame of the external gravitational field @xmath23 . in order to analyze effects in systems moving with velocity @xmath24 with respect to that frame
we need to transform the fields and coordinates in ( [ q1 ] ) via the corresponding lorentz transformations .
this gives @xmath25\nonumber.\end{aligned}\ ] ] where @xmath26 is the electromagnetic 4-current associated with some external source , @xmath27 , and @xmath28 is a dimensionless ( and species - dependent ) parameter that measures the degree to which lpi / lli is broken .
this parameter scales with the magnitude of the dimensionless newtonian potential , which is expected to be much smaller than unity for actual experiments .
we are therefore able to compute effects of the terms in eq .
( [ q2 ] ) that break local lorentz invariance via a perturbative analysis about the familiar and well - behaved @xmath29 or @xmath30 limit .
we therefore consider gravitationally modified quantum electrodynamics ( gmqed ) based on the action ( [ q2 ] ) . the fermion sector and
the interaction term do not change with respect to the metric case , and so neither do the fermion propagator and the vertex rule .
all the non - metric effects are accounted for in the pure electromagnetic sector of the action , which in turns modifies the photon propagator . after a proper choice of the gauge fixing term involved in the quantization of the photon field ,
it can be shown @xcite that the photon propagator takes the form up to first order in @xmath31 ( in momentum space ) : @xmath32 , \ ] ] where @xmath33 is the minkowski tensor with a signature ( + - - - ) ; @xmath34 and @xmath35 ; henceforth @xmath36
. therefore eq .
( [ 11 ] ) , along with the unmodified fermion propagator @xmath37 and vertex rule , may be used as the basis of the feynman rules of gmqed .
the radiative corrections affecting these quantities are defined in terms of the photon self energy @xmath38 , fermion self energy @xmath39 , and vertex function @xmath40 respectively .
these insertions involve the calculation of loop integrals as given by the feynman rules up to a given order .
the addition of more parameters to the theory also entails new renormalizations beyond those of the wavefunctions , charge and mass of the fermion .
the parameters appear as functions of @xmath41 and @xmath42 , and must then be correspondly redefined . in units
where @xmath43 ( @xmath44 ) , eep - violating corrections only appear in the electromagnetic sector of the action ( as terms proportional to @xmath31 ) .
however a more general choice is @xmath45 , for which the particle sector of the lagrangian density is of the form _ d=(p - v - m)+_0(p_0-v_0)^0with @xmath46 . in the moving frame
this is ( up to a constant ) [ apbd ] l_d&=&(p - v - m ) + & + & _ 0 ^ 2(p - v ) . from ( [ apbd ] )
we see that quantum corrections of the form _
d=(_0^(1)p-_0^(2 ) v)can still be expected .
it is straightforward to show that gauge invariance will guarantee @xmath47 .
hence , in order to renormalize the mass and the parameters , we have to include counterterms of the form [ coun1 ] m+_0(p - v ) , which consequently participate in the redefinition of the fermion self energy and vertex function .
additionally , the electromagnetic sector will induce quantum fluctuations of the form : [ coun2 ] _ em = a^\{(k^2-^2(k)^2 ) _ -^2__k^2}a^corresponding to the renormalization of the parameters , or equivalently @xmath48 , entailing a renormalization of the photon self energy .
this summarizes the procedure for performing perturbative calculations in gmqed as employed previously in ref .
@xcite in computing the lamb shift .
this calculation was complicated by the boundedness of the electron to the hydrogenic atom under consideration . here , in the @xmath0 case , we have to deal only with free leptonic states . in sec .
ii we evaluate the ( one loop ) radiative corrections to the elastic scattering of a free electron by an external electromagnetic field . sec .
iii relates the scattering amplitude with experimental observables describable in terms of the g-2 anomaly .
this requires a derivation of the relativistic equation of motion for the electron spin in the presence of a magnetic field .
this follows from a classical treatment for the electron and magnetic field .
quantum effects are introduced in the modified hamiltonian only ( as are non metric effects ) .
the connection with possible lli / lpi violating experiments is presented at the end of this section .
discussion of the results and general comments are given in sec .
more details about the loop calculation are shown in the appendix .
we shall consider the lowest order radiative correction to the elastic scattering of electrons by a static external field @xmath49 .
these one loop contributions can be summarized in terms of the feynman diagrams illustrated in fig . 1 .
the feynman amplitudes for the diagrams follow from the feynman rules giving the result : [ ampli ] ^(p,p)=u(p)\{^+p^+l^}u(p ) where [ vertex ] ^(p,p)&=&d^4k^ is_f(p-k)^is_f(p - k ) ^i g_(k ) + [ pola ] p^(p,p)&=&^i g_(q)i
^(q ) + [ leg ] l^(p,p)&=&i(p)is_f(p)^+^is_f(p)i(p ) with [ self ] i(p)&=&d^4k
i g_(k)^is_f(p - k)^ + [ pola1 ] i^(q)&=&(-)d^4k^is_f(k+q ) ^is_f(k ) and @xmath50 .
we refer to eqs .
( [ vertex ] ) , ( [ pola ] ) , and ( [ leg ] ) as the vertex , polarization , and leg contributions , which respectively correspond to diagrams ( a ) , ( b ) and ( c ) plus ( d ) .
we also note that expressions ( [ vertex ] ) , ( [ self ] ) , ( [ pola1 ] ) represent the one loop corrections to the vertex , fermion and photon self energy parts respectively . given the form of the photon propagator it is convenient to introduce : ^=(1+)^_0+^2^_where the subscript 0 " denotes the ( known ) result coming from the standard part of the photon propagator , and @xmath31 " for the part proportional to @xmath51 in ( [ 11 ] ) [ fp ] g_^= + _ in the remainder of this section we consider this part of the propagator only , omitting the @xmath31 " label in the corresponding expressions . the procedure for evaluating the loop integrals is equivalent to that of standard ( or metric ) qed . we need to regularize them first and then renormalize the parameters , which include the parameter along with the fermion charge and mass .
the regularization of the photon propagator is carried out using [ regu ] -_^2^^2,-2_^2^^2 , with the assumed limits @xmath52 and @xmath53 , and the parameter renormalization by the inclusion of the corresponding counterterms to each loop integral .
details about this procedure and the corresponding calculations are given in the appendix .
we quote the final result for the loop integrals : [ self - f ] ( p)&=&(p - m)\ { -^2 } + & -&\{(p - m)+(p - m ) } + o((p - m)^2 ) + [ vertex - f ] ^(p,p)&=&\{^ [ -^2(()^2 + - ( ) ^2 ) .
+ & + & + & -&^ ( + ) + ( -)^ + & -&^-^+^+ ( + + ) ^ + & + & ( - ) -(+)}+o(q^3 ) + [ pola - f ] ^(q)&=&-(q^2 ^ -q^q^)+o(q^6 ) where we have implicitly assumed that ( [ vertex - f ] ) is acting on free spinors . the ward identity [ wardi ] = ^(p , p )
. is a consequence of gauge invariance , and therefore it holds even in the absence of lorentz invariance .
it is straightforward to check that ( [ self - f ] ) and ( [ vertex - f ] ) satisfy ( [ wardi ] ) .
the evaluation of ( [ ampli ] ) is also straightforward once the loop integrals have been calculated .
we just comment on the computation of the leg correction , which is ambiguous since it contains terms like @xmath54 , which are indeterminate . to obtain an unambiguous result , we must explicitly introduce a damping factor , which is necessary for the correct definition of the initial and final `` bare '' states .
details of this adiabatic approach are presented in appendix b. the final result for the leg correction is l^&=&\{^ + & + & ^-^+ ^ } note that this part gives a contribution to the total amplitude that _ can not _
be removed after renormalization .
furthermore , the gauge invariance of the feynman amplitude which is manifest as q=0 requires the presence of such terms , a condition that is not satisfied by the vertex contribution only . the final result for the scattering amplitude is [ final ] ^=f^+g^+i^with [ ff ] f^&=&\{^-^ + & + & ^+^ } + [ gg ] g^&=&\{-^ -^-^+(+)^ } + [ ii ] i^&=&\{^+(- ) -(+ ) } the various terms in ( [ final ] ) distinguish the different contributions to the scattering amplitude . in ( [ ff ] ) we group terms of order @xmath55 or higher .
@xmath56 accounts for terms of order @xmath57 at least , and @xmath58 for the gauge terms or those who give no contribution to the amplitude .
note that the remaining infrared divergence in @xmath59 can be understood in terms of soft photon radiation , analogous to the metric case . in the next section
we will use the above results to compute the @xmath0 anomaly .
to lowest order the feynman amplitude associated with the elastic scattering of an electron by a static external field is [ zero ] ieu(p)a(q)u(p ) .
the radiative correction of order @xmath60 to this process is given by [ uno ] ieu(p)\{(1+)_0a+^2_a}u(p ) where @xmath61 represents the ( known ) metric result and @xmath62 represents the contribution from ( [ final ] ) . in the nonrelativistic limit of slowly moving particles ( @xmath63 ) and a static magnetic field , it is straightforward to show that ea(q ) & & -b + e_0a&&-()b + e_a&&ega [ xigg ] with @xmath56 given by ( [ gg ] ) , which is the dominant term as @xmath64 . in order to simplify this contribution , we consider a constant magnetic field @xmath65 , that is @xmath66 , in which case qa-(buu -bu^2 ) where we have neglected the terms that mix the large and small spinor components .
similarly , we can show qa-(buu -bu^2 ) and in the non relativistic limit qaqa - band qa - b . if we put everything together in ( [ gg ] ) : ega-\{b(+u^2 ) -buu } as a cross - check on the above result , we take the limit @xmath67 obtaining @xmath68 , which is the required limit consistent with the structure of eq .
( [ fp ] ) in that case .
the previous result is the contribution of ( [ xigg ] ) to ( [ uno ] ) , which added to ( [ zero ] ) , give us the relevant part of the hamiltonian as [ inter ] h_=-\{sb+_*subu } + o(^2)o(^2 ) -^ijs_ib_j with [ gs ] g&&\{2 + } + [ gu ] _ * g_*&&-^2 where we have identified @xmath69 , and @xmath70 .
the @xmath71 parameters account for the coupling strength between the magnetic field and spin .
we see that @xmath72 generalizes the gyromagnetic ratio of a fermion analogous to the manner in which the anomalous mass tensor generalizes the mass of a particle @xcite .
we therefore identify the parameters @xmath73 with the components of the _ anomalous gyromagnetic ratio tensor _ of the fermion in the class of theories .
note that the presence of preferred frame effects induces a qualitatively new form of interaction between the spin and magnetic field which is quantified by @xmath74 . here , instead of coupling with each other , they both couple independently to the fermion velocity relative to the preferred frame .
this interaction stems purely from radiative corrections , and would be absent in any tree - level analysis of gmqed .
hence , eq . ( [ inter ] ) describes the interaction ( as seen from the particle rest frame ) between the particle spin and an external homogeneous magnetic field . from this
we can extract the energy difference between electrons with opposite spin projection in the direction of the magnetic field as : [ eshift ] e_=- where @xmath75 is the angle between the magnetic field and the preferred frame velocity .
the influence of the radiative corrections ( coming from @xmath0 and @xmath76 ) in ( [ eshift ] ) is negligible in comparison to the dominant factor of 2 in @xmath77 .
since we want to single out the effects of the non - metric corrections , it is more interesting to study the precession of the spin or , more specifically , the oscillation of the longitudinal spin polarization . in the metric case , this frequency is proportional to the factor @xmath0 , and so it is a distinctive signature of radiative corrections .
the observable quantity in the @xmath0 experiments is actually the electron polarization , which is proportional to the quantum mechanical expectation value of @xmath78 , that is , @xmath79 . using ehrenfest s theorem , a quantum mechanical solution for the motion of @xmath79
is obtained from the equation [ spin1 ] |_r.f .=- i[s , h_]=s where the primed variables are referred explicitly to the particle rest frame ( @xmath80 ) .
note that the preferred frame effect will show distinctly as a temporal variation of the spin component parallel to the magnetic field .
in general we want to know the spin precession relative to some specific laboratory system , with respect to which the particle is moving with some velocity @xmath81 .
this frame need not _ a - priori _ be the previously defined preferred frame , and so @xmath82 .
since the formalism does not change ( locally ) the fermion electromagnetic field interaction , we expect that a charged particle in the presence of an homogeneous magnetic field will satisfy the equation [ spin2 ] = _ c with the cyclotron frequency @xmath83 and @xmath84 . relating ( [ spin1 ] ) to the laboratory system yields [ spin3 ] |_lab=|_r.f .
+ _ ts due to thomas precession , with @xmath85 .
this frequency is kinematic in origin and it is a consequence of the non - commutativity of the lorentz transformations . relating the primed variables in ( [ spin1 ] ) to the laboratory ones by a lorentz transformation gives [ spin5 ] |_lab = s_s with _ s = b+(1-)_c+ _ * ( bu)u where we have set @xmath86 and considered ( for simplicity ) the case of orbital motion perpendicular to the magnetic field ( @xmath87 ) in the above .
note that the spin precession about @xmath88 is no longer parallel to the magnetic field ( axial direction ) , but has a component parallel to @xmath24 that comes from radiative and nonmetric effects . at this point
it becomes necessary to define the preferred coordinate system .
there are several candidates ( such as the rest frame of the cosmic microwave background ) for this frame @xcite .
to study this issue it is sufficient to assume that the laboratory system ( earth ) moves with a non - relativistic velocity ( @xmath89 ) with respect to the preferred frame , and so we can identify @xmath90 in order to single out the effects of radiative corrections , we study the spin precession relative to the rotational motion of the electron , that is : [ set1 ] |_rot = s_d with @xmath91 and @xmath92 , where the first two components are perpendicular to @xmath65 ( lower index ) but parallel and perpendicular to @xmath81 ( upper index ) , and the last one parallel to @xmath65 . in the following we refer to the difference frequency ( @xmath93 ) as the anomalous frequency ( given its connection with the anomalous magnetic moment in the metric case ) .
it is convenient to rewrite : _
d=_a+_a^*(v_+ ) with [ anomal ] _
a=(g+g_*v^2 ^ 2 - 2)b and @xmath94 ; where @xmath75 represents the angle between @xmath95 and the magnetic field , and @xmath96 the component of the velocity perpendicular to @xmath97 .
in @xmath98 we group all the terms parallel to the magnetic field that contribute to the anomalous frequency ( including nonmetric effects ) .
the remaining terms perpendicular to @xmath97 arise from nonmetric effects only , and produce a temporal variation of the spin component parallel to the magnetic field .
this effect is absent in the metric case , and so represents a qualitatively new manifestion of possible eep violation . in general
we are interested in solving ( [ set1 ] ) for the cases @xmath99 or @xmath100 so that @xmath101 or @xmath102 , but is otherwise constant . since @xmath103 is proportional to @xmath31 , we can perturbatively solve for each component in ( [ set1 ] ) .
taking , for example , the initial condition @xmath104 we find [ set4 ] s_^&=&s_at s_^=s_at + s_&=&s(1-_at)+s vwhere we have chosen a coordinate system where @xmath105 so that v = b+ x , = y _ ct - x _ ct and assumed that any rotation related to @xmath75 is negligible in comparison to other frequencies involved in the problem ( @xmath98 or @xmath106 ) .
the fact that @xmath98 was ( in the metric case ) proportional to @xmath0 , motived the very precise @xmath0 experiments which were designed to specifically measure that anomalous frequency .
we see that this frequency is modified by from its metric value by the additional terms present in ( [ anomal ] ) .
if we assume that the eep - violating contributions to @xmath98 are bounded by the current level of precision for anomalous magnetic moments @xcite , then the discrepancy between the best empirical and theoretical values for the electron yields the bounds [ belectron ] |_e^-| < 3.510 ^ -8 |_e^- - _ e^+|
< 10 ^ -9 the latter following from a comparison of positron and electron magnetic moments . for muons , a similar analysis yields [ bmuon ]
anomaly is lower than the electron one , the slightly stronger bound in ( [ bmuon ] ) arises because the experiments are carried out for high - velocity muons @xcite .
to our knowledge these bounds on violation of gravitational universality are the most stringent yet noted for leptonic matter .
newman _ et .
_ analyzed the @xmath0 experiments @xcite in order to find new bounds for the validity of special relativity .
they assumed that the parameter @xmath107 involved in the electron motion had a different value ( @xmath108 ) from that which arises kinematically ( in thomas precession and lorentz transformations ) .
the equivalent equation for ( [ anomal ] ) is in that case _
a^nfrs=(- ) and by comparing with two electron @xmath0 experiments , one at electron relativistic energy ( @xmath109 ) and the other nearly at rest ( @xmath110 ) , they obtained the constraint @xmath111 .
our approach is qualitatively different from theirs , in that we assume @xmath112 but include preferred frame effects in the evaluation of the anomalous magnetic moment .
a similar analysis in our case yields the weaker bounds of @xmath113 for electrons , and @xmath114 for muons . in the latter we used the @xmath0 muon experiments carried at @xmath115 ( @xmath116)@xcite , and @xmath117 ( @xmath118)@xcite .
preferred effects not only modify the anomalous frequency according to ( [ anomal ] ) , but also induce oscillations in the spin component parallel to @xmath97 .
as stated above , this is a qualitatively new consequence of eep violations due solely to radiative corrections in gmqed . searching for such oscillations
therefore provides a new null test of the eep .
we can estimate the magnitude of such effects by taking the temporal average of @xmath119 over the main oscillation given by @xmath98 , which gives = ~v^2 this effect is enhanced in highly relativistic situations , and can be estimated by considering a typical experiment with @xmath120 . for electrons @xmath121 , and so @xmath122 ; for muons @xmath115 , yielding @xmath123 . in both cases
we used the corresponding present constraints for @xmath31 given above .
the novelty of the @xmath119 oscillation suggests the possibility of putting tighter constraints on the non - metric parameter , once appropriate experiments are carried out .
the same goes for the analysis of @xmath98 at different values of @xmath75 ( the angle between the magnetic field and the velocity of the laboratory system with respect to the preferred frame ) .
the rotation of the earth will turn this orientation dependence into a time - dependence of the anomalous magnetic moment , with a period related to that of the sidereal day .
the previous analysis was concerned with effects related to spatial anisotropy .
we turn now to considering possible violations of local position invariance .
the position dependence in the former section was implicit in the redefinitions of charge , mass and fields .
these quantities were rescaled in terms of the functions , which were considered constant throughout the computation .
lpi violating experiments are of two types .
one of these entails the measurement of a given frequency at two different points in a gravitational field ( where differences in the gravitational potential could be significant ) within the same reference system .
the other type involves a comparison of frequencies arising from two different forms of energy ( _ i.e. _ two different clocks ) at the same point in a gravitational potential .
we parameterize the gravitational dependence on a given frequency as : [ para1 ] = ^0+where @xmath124 represents the external gravitational tensor , satisfying @xmath125 , and the ellipsis represents higher order terms ( going as either @xmath126 or velocity times u ) in the gravitational potential or terms independent of it .
the measured redshift parameter related to this frequency may be written as z = u(1-),=^ij where @xmath127 will depend upon the specific frequency measured in the experiments .
note that this tensor is equivalent to the anomalous passive gravitational mass tensor introduced for the study of atomic transitions . in @xmath0 experiments the relevant frequency is @xmath98 , which describes the precession of the longitudinal polarization in the presence of a constant magnetic field .
using the formalism ( see eq.([anomal ] ) ) we obtain [ anor1 ] _
a=+=+where we have omitted terms proportional to velocities , which eventually will contribute as @xmath128 terms at most . in order to carry out the loop calculation ,
the dependence was absorbed into the definition of the parameters under the rescaling [ rescal ] /c_0mm
/ c_0/_0c_0 with @xmath129 as the limiting speed of the massive particles , the subscript ` 0 ' denoting the functions evaluated locally at @xmath130 .
although the product @xmath131 remains invariant under this rescaling , the expression for the constant magnetic field still depends on the parameters once it is written solely in terms of atomic parameters .
this can be seen clearly by considering the magnetic field produced by a long solenoid of length @xmath132 , with @xmath133 turns and carrying a current @xmath134 .
the gravitationally modified maxwell equation to solve is : ( ^-1b)=4j and so we find the non - vanishing magnetic field inside the solenoid to be @xmath135 .
again we assume that the functions are constant throughout the size of the experimental device . in terms of fundamental atomic parameters , @xmath132 is proportional to an integer times the bohr radius ( the interatomic spacing ) , which is known to rescale as @xmath136 .
if we now write @xmath137 , where @xmath138 can be expressed in terms of a density charge @xmath139 in motion ( @xmath140 ) through a volume @xmath95 , and then relate the bohr radius to each spatial dimension along with the limiting particle velocity @xmath141 to the velocity distribution @xmath140 , we can show @xmath142 , and so @xmath143 .
along with ( [ rescal ] ) , this gives the position dependence of ( [ anor1 ] ) to be [ anor2 ] _
a=_a^0(1 + ) with @xmath144 ( recall @xmath145 ) .
note that the functions are evaluated at some representative point of the system , which we have chosen to be the origin @xmath21 . in order to determine how @xmath98 changes as the position of the system varies
, we expand the functions in ( [ anor2 ] ) [ tu ] t(u)=t_0+t_0g_0x+o(g_0x)^2 where @xmath146 , @xmath147 , and @xmath148 .
it is useful to redefine the gravitational potential @xmath23 by [ u ] u - g_0x whose gradient yields the test - body acceleration @xmath149 .
this finally yields [ anor3 ] _
a=_a^0where we have rescaled again according to ( [ rescal ] ) , and omitted terms proportional to @xmath31 , since the main position dependence parameterization is given in terms of : [ gamlam ] _
0=(+ - ) , _ 0=(+ - ) by comparing eq .
( [ anor3 ] ) with ( [ para1 ] ) , we can identify ^g-2=_0-_0 as the lpi - violating parameter .
note that this depends on the anomalous frequency related to the longitudinal polarization of the beam .
it is also species - dependent , with the value of @xmath150 and @xmath61 for the electron differing from that of the muon .
a search for possible position dependence of anomalous spin precession frequencies provides another qualitatively new test of lpi sensitive to radiative corrections .
actually the most precise @xmath0 experiments for electron measure the ratio @xmath151 at non relativistic electron energies ( @xmath152 ) , and so @xmath153 .
this is interesting because by following the former parameterization we can write : [ ciclo ] _
c=_c^0or by taking the ratio of ( [ anor3 ] ) to ( [ ciclo ] ) : a = a^0(1+u^a),^a=_0+_0 and then by identifying @xmath8 with the most precise experimental value @xcite and @xmath154 with the theoretical one@xcite , we can constrain through the resulting theoretical / experimental errors @xmath155 .
this result is sensitive to the absolute value of the total local gravitational potential @xcite , whose magnitude has recently been estimated to be as large as @xmath156 due to the local supercluster @xcite .
hence measurements of this type can provide us with empirical information sensitive to radiative corrections that constrains the allowed regions of @xmath157 parameter space , giving in this case : |_0+_0|<10
^ -3 for muons the analogous constraint is @xmath158 , and so nothing conclusive is obtained .
we pause to compare this result to an analogous result obtained for hyperfine transitions ( maser clocks ) . in this case
the baryonic and leptonic gravitational parameters appear simultaneously .
the atomic hyperfine splitting comes from the interaction between the magnetic moments of the electron and proton ( nucleus ) .
the proton metric appears only in the latter , and so it does not affect the principal and fine structure atomic energy levels .
non - metric effects imply a shift in the hyperfine energy @xmath159 which is @xcite [ hf1 ] e_hf = e_hf(1-u_b)+e_hf u_b^hf with ^hf=3_b-_b+where @xmath160 , @xmath161 and @xmath162 are the baryonic analogues of the parameters appearing in ( [ gamlam ] ) . in ( [ hf1 ] ) we rescaled the atomic parameters to absorb the functions and chose units such that @xmath163 .
the quantity @xmath164 is given by = 2and would vanish under the assumption that the leptonic and baryonic parameters were the same .
the gravity probe a experiment @xcite , employing hydrogen - maser clocks , was able to constrain the corresponding lpi violating parameter related to hyperfine transitions , obtaining @xmath165 for the most stringent bound to date on @xmath166 .
we note that a similar analysis could be carried out for the energy shift defined in ( [ eshift ] ) , which can be used as a frequency test to look for position or frame dependence .
this can be done by following the same procedure as for atomic energy shifts , where the anomalous passive and inertial gravitational tensor are introduced in order to relate non - metric effects to redshift and time dilation parameters .
since radiative corrections are irrelevant in that energy shift , we omit that procedure here .
refined measurements of anomalous magnetic moments can provide an interesting new arena for investigating the validity of the eep in physical systems where radiative corrections are important .
we have considered this possibility explicitly for the class of non - metric theories described by the formalism .
the non - universal character of the gravitational couplings in such theories affects the one loop corrections to the scattering amplitude of a free fermion in an external electromagnetic field in a rather complicated way , giving rise to several novel effects .
an evaluation of the one - loop diagrams reveals that the leg corrections , which in the metric case give no contribution to the total amplitude after a proper renormalization of mass and spinor field , provide contributions which can not be removed after renormalization .
moreover they are essential in ensuring the gauge invariance of the scattering amplitude , which is not fulfilled by the vertex correction alone .
the consistency of the calculation is verified explicitly through the ward identity , which furnishes a cross - check between the fermion self energy and the vertex correction .
the non metric corrections to the scattering amplitude also have an infrared divergence , which could be understood in terms of inelastic soft photon radiation , as in the metric case .
this does not affect the term associated with the anomalous magnetic moment .
the presence of preferred frame effects induces a new type of coupling between the magnetic field and the spin as described by ( [ inter ] ) .
this interaction stems purely from radiative corrections , and generalizes the gyromagnetic ratio of a fermion to a tensorial coupling described by @xmath72 .
we emphasize that qualitatively new information on the validity of the eep will be obtained by setting new empirical bounds on this coupling , as it is associated with purely _ leptonic _ matter .
comparatively little is known about such empirical limits on eep - violation relative to the baryonic sector @xcite , for which previous experiments have set the limit @xcite @xmath167 where @xmath168 is the ratio of the limiting speed of baryonic matter to the speed of light .
we can therefore safely neglect any putative effects of @xmath169 in our analysis .
consequently , discussion of a @xmath0 contribution to the magnetic moment no longer makes sense , and we instead refer to the anomalous frequency as the main connection with experiment . note that this frequency , defined as the relative electron spin precession with respect to its velocity , comes from radiative corrections and it becomes proportional to @xmath0 in the metric case
this frequency shows an explicit dependence on both the preferred frame velocity and its relative direction with respect to the external magnetic field .
there is also a dependence on the electron velocity , which makes the other contributions negligible at relativistic electron energies .
two @xmath0 experiments on the electron ( one at relativistic energies and the other almost at rest ) may then be used to limit the preferred frame parameter to be no larger than @xmath170 , analogous to the work of newman _
et al._. constraining any possible eep violation to be no larger than the present discrepancy between theory and experiment we found the most stringent bounds for @xmath31 yet obtained for leptonic matter , as given in ( [ belectron ] ) and ( [ bmuon ] ) .
we expect that new experiments which probe the anisotropic character ( or angular dependence ) of the frequency could be used to impose stronger limits in different physical regimes .
for example , as the earth rotates , the spatial orientation of the magnetic field changes this should in turn diminish the experimental errors involved in the comparison between two energetically different @xmath0 experiments .
the relativistic generalization of the spin polarization equation ( [ spin5 ] ) , followed the same procedure as for the metric case , where non - metric effects where included in the interaction only ( eq .
( [ inter ] ) ) .
this yields an equation of motion for the spin ( as seen from the rest frame ) which is qualitatively different from that expected from its classical counterpart , where the angular momentum rate is related to the torque applied on the system .
this approach for dealing with violations of lorentz invariance is dynamical ; from a kinematical viewpoint we assume that standard lorentz transformations relate coordinates and fields from one system to another .
perhaps the most remarkable feature of the non - metric effects is that of the oscillations of the component of spin polarization parallel to the magnetic field .
since this component remains constant in the metric case , an experiment which searches for such oscillations is a new null test of the equivalence principle that is uniquely sensitive to radiative corrections in the leptonic sector .
hence an empirical investigation of its behavior will provide qualitatively new information about the validity of eep , and could constrain even further the limits on the preferred frame parameters .
finally , we analyzed the behavior of the anomalous frequency in the context of redshift experiments , which can put constraints on the lpi - violating parameters ( @xmath150 , @xmath61 ) once the corresponding experiments are carried out .
this region of parameter space is qualitatively different from that probed by either lamb - shift or hyperfine effects . in the electron sector
a bound on the magnitude of @xmath171 can be obtained by demanding that it be no larger than the error bounds in the discrepancy between the experimental and theoretical values of the ratio @xmath151 . assuming the local potential to be as large as that estimated from the local supercluster
, we obtain a bound on @xmath172 that is comparable the limit on an analogous quantity in the baryonic sector obtained from redshift experiments @xcite .
however this latter experiment is proportional to changes in the local potential , which are @xmath173 .
more direct limits on @xmath172 must be set by performing a similar sort of redshift experiment on anomalous magnetic moments @xcite .
the logistics and higher precision demanded by such an experiment will be a major challenge to undertake .
this work was supported in part by the natural sciences and engineering research council of canada .
we show the main steps leading to eqs .
( [ vertex - f ] ) , ( [ self - f ] ) , and ( [ pola - f ] ) .
details are given throughout the computation by considering only the first term of the photon propagator ( [ fp ] ) , that is [ fp1 ] g_^=+with the remaining term in ( [ fp ] ) contributing in a similar manner .
we solve for the fermion self energy by replacing ( [ fp1 ] ) in ( [ self ] ) , and using ( [ regu ] ) along with the feynman parameters
@xmath174 ^ 3}.\ ] ] after integrating we obtain ( p)=(p z+m)()+with @xmath175 .
we consider @xmath176 , and expand the above to obtain after some manipulation & & ( p)=\{p(+()^2 ) -(p - m)(()^2 + ) + & & -m(-()^2 ) + 2(p - m)(()^2 - 3 ) } + o((p - m)^2)+where we have kept the leading terms as @xmath52 and @xmath177 , and @xmath178 stands for the terms satisfying @xmath179 we renormalize @xmath39 by subtracting = m+_0pwhere the counterterms respectively account for mass and -parameter renormalization .
choosing the counterterms so that @xmath180 and so m&=&-m(-()^2 ) + [ xi0]_0&=&(+()^2 ) the regularized result is then ( p)&=&(p - m)\{-^2 } + & + & o((p - m)^2 ) . note that the remaining ultraviolet divergence related to this term could be removed after charge renormalization .
we find it convenient to leave it in order to cross - check the calculation , since a similar term from the vertex part should cancel it , thereby removing the divergence from the resulting scattering amplitude .
the evaluation of the vertex function follows a similar procedure , giving the result [ vertex - a ] ^&=&\{^ + & -&x(p^+p^ ) -(p^+p^)(1-x ) + & + & ( p_xp_x^- ^p_x^2 ( - ) -(1-x)p p^-x p p^ ) + & + & p_x(^(p + p)-p_x^ ) } + with @xmath181 .
since ( [ vertex - a ] ) is acting on a free spinor , we can use @xmath182 with @xmath183 , and so expand @xmath184 which after some algebra reduces ( [ vertex - a ] ) to ^(p,p)&=&\{^ + & + & ^(+ ) + ^-^ + & -&^+^+ ^-(+ ) + & -&}+o(q^3)where the vertex function has been renormalized by subtracting a term like ^=_0^with @xmath185 is given by ( [ xi0 ] ) .
we recall that gauge invariance forces this coefficient to be equal to the one participating in the renormalization of the fermion self energy .
in order to describe how self energy effects convert the incident electron from a bare particle to a physical one , it is convenient to introduce a damping function , @xmath186 , which adiabatically switches off the coupling between fields , such that the interaction lagrangian is replaced by _
i = eg(t)(x)a(x)(x ) it is assumed that the time @xmath13 over which @xmath186 varies is very long compared to the duration of the scattering process . in momentum space g(t)=g(_0)(ix)d_0 with @xmath187 , and @xmath188 .
it is supposed that @xmath189 is almost a delta function , being large for @xmath190 in a range of about @xmath191 as @xmath194 , and @xmath195 , @xmath196 , the fermion propagator reduces to s_f=- where we used @xmath197 .
this implies that we can expand @xmath198 up to order @xmath199 only , since higher terms vanish after taking the previous limit .
here we employ the relation = + b+to expand [ sig1 ] ( p-)(p)-^. after renormalization , @xmath198 takes the form [ sig2 ] ( p)=(p - m)(a+b(p)^2)+c(p)\ { ( p - m)+(p - m)}+where the constants @xmath200 , @xmath97 , and @xmath201 can be obtained from eq .
( [ self - f ] ) .
let us introduce @xmath202 , and symmetrize @xmath199 by @xmath203 in ( [ sig1 ] ) , to write s_f(p--)(p-)= ( ( p)-^ ) which after using ( [ sig2 ] ) can be written as [ sig3 ] ( a+b(p)^2)+cpwhere we have used that @xmath204 is acting on a free spinor , and therefore terms of the form @xmath205 vanish .
now , the final evaluation of ( [ leg ] ) follows directly from ( [ sig3 ] ) . | we investigate the possibility of testing of the einstein equivalence principle ( eep ) using measurements of anomalous magnetic moments of elementary particles .
we compute the one loop correction for the @xmath0 anomaly within the class of non metric theories of gravity described by the formalism . we find several novel mechanisms for breaking the eep whose origin is due purely to radiative corrections .
we discuss the possibilities of setting new empirical constraints on these effects . 6.0 in 8.5 in -0.25truein 0.30truein 0.30truein = 0.6 cm * the equivalence principle and anomalous magnetic moment experiments * + c. alvarez and r.b .
mann + department of physics + university of waterloo + waterloo , ont n2l 3g1 , canada + + |
the angular momentum of gas accreting onto massive black holes in active galactic nuclei ( agn ) is likely to change with time , either as a result of mergers funneling fresh gas towards the nucleus ( hernquist & mihos 1995 ) , or because of radiation or disc wind driven warping instabilities in an existing accretion flow ( pringle 1996 , 1997 ; maloney , begelman & pringle 1996 ; maloney , begelman & nowak 1998 ) . in either case , the angular momentum of gas at large radius in the disc will be misaligned with the rotation axis of the black hole , while at small radius the combined action of viscosity and differential precession induced by the lense - thirring effect ( lense & thirring 1918 ) leads to alignment of the disc and hole angular momenta ( bardeen & petterson 1975 ; kumar & pringle 1985 ) . achieving this alignment
requires that the black hole exert a torque on the disc gas , and implies that an equal and opposite torque act on the black hole itself . over time
, this causes a change in the spin axis of the hole towards alignment with the large angular momentum reservoir provided by the disc at large radius . whether the timescale for alignment is long or short compared to the lifetime of an agn has implications for our understanding of the accretion history of these objects . if the black hole is spinning , then jets , regardless of whether they derive power from the aligned inner disc or directly from the hole itself ( blandford & znajek 1977 ; see also ghosh & abramowicz 1997 ; livio , ogilvie & pringle 1999 ) , trace the spin of the black hole .
if the alignment timescale is long , then we expect that the jet direction reflects the initial spin of the black hole established during ( or prior ) to the formation of the host galaxy ( rees 1978 ) .
moreover the jet direction is expected to be stable over time , irrespective of variations in the angular momentum of accreting gas .
conversely , if the timescale is short then jets define the angular momentum of inflowing material , and the observation of systems where they appear to be stable over long time periods of @xmath1 - @xmath2 yr ( alexander & leahy 1987 ; liu , pooley & riley 1992 ; scheuer 1995 ) implies a constancy in the average angular momentum of accreting gas .
the rate of realignment was computed by scheuer & feiler ( 1996 ) , who found an approximate analytic solution to the equations governing the evolution of a warp .
the timescale depends on two poorly known viscosities , @xmath3 and @xmath4 , which correspond to the @xmath5 and @xmath6 components of the shear as defined by papaloizou & pringle ( 1983 ) . if @xmath7 ( i.e. the timescale for warp evolution is comparable to that for the evolution of the surface density ) then the formula of scheuer & feiler ( 1996 ) is similar to that previously assumed by rees ( 1978 ) .
subsequently , natarajan & pringle ( 1998 ) revisited the problem , and pointed out that for the parameters of agn discs the assumption that @xmath7 is at variance with the results of detailed analyses of the hydrodynamics of viscous discs , which suggest instead that @xmath8 . recalculating the alignment timescale with this modification they found that it was _ short _ much less than any reasonable estimate of agn lifetime and used this to argue that the spin of black holes in agn , and the direction of jets produced in the inner regions of the accretion flow , rapidly adjust to trace the angular momentum of gas at large radius .
the analysis in natarajan & pringle ( 1998 ) employed the same torque formula as scheuer & feiler ( 1996 ) , which was derived under the restrictive conditions of small amplitude warps and a disc viscosity that is constant with radius ( or , equivalently , that in a steady - state the disc surface density is independent of radius ) .
since agn may well harbour strongly warped discs , and almost certainly have more complex surface density profiles , our goal in this paper is to generalise their results .
section 2 below sets out the equations for the evolution of a warped disc , which we solve numerically in section 3 for comparison with the approximate analytic solution .
the resulting torque is calculated in section 4 , and the alignment timescale derived for a simple agn disc model in section 5 . section 6 discusses the implications of our results .
the evolution of a thin , warped accretion disc in a keplerian potential was studied by papaloizou & pringle ( 1983 ) . for a disc with a surface density profile @xmath9 , with angular momentum at radius @xmath10 parallel to the unit vector @xmath11
, the evolution can be expressed most compactly in terms of an equation for the angular momentum density @xmath12 . adopting the simplest description ( pringle 1992
; see also papaloizou & pringle 1983 ) , the time evolution is given by , @xmath13\,\nonumber \\ & + & { \frac{1}{r}}\,{\frac{\partial}{\partial r}}\,\left [ ( \nu_2\,r^2\,{\left\vert{\frac{\partial { \it \hat l } } { \partial r}}\right\vert^2}\,-\,{\frac{3}{2}}\,\nu_1)\,{\bf l } \right ] \nonumber \\ & + & { \frac{1}{r}}\,{\frac{\partial}{\partial r}}\,\left ( { \frac{1}{2}}\,\nu_2\,r\,|{\bf l}|\,{\frac{\partial \hat l}{\partial r } } \right ) \,+\,{{{\vec \omega_p } \times { \bf l } } \over r^3}. \label{eq1}\end{aligned}\ ] ] here @xmath3 and @xmath4 are viscosities acting on the @xmath5 and @xmath6 components of the shear , as defined by papaloizou & pringle ( 1983 ) . neither is equal to the shear viscosity ` @xmath14 ' that enters into the navier - stokes equations , and the general relation between these three quantities is extremely complex ( ogilvie 1999 ) . for the parameters of agn discs , the simplest assumptions suggest that @xmath8 ( natarajan & pringle 1998 ) , though for most of this paper we will allow both viscosities to be general power laws with radius , @xmath15 for thin , vertically averaged accretion disc models , the shear viscosity @xmath14 is a power - law in radius in a given opacity regime ( e.g. frank , king & raine 1992 ) .
later , the index @xmath16 will be chosen to correspond to the scaling expected at radii in the disc where the bulk of the realignment torque acts . however , for a general warped disc , the radial run of both @xmath3 and @xmath4 will also depend on the _ shape _ of the warp ( ogilvie 1999 ) .
this effect is not taken into account in this paper .
our assumption that @xmath3 and @xmath4 are simple power - laws in radius , independent of warp shape , is therefore strictly valid in the limit of small warps , and is an approximation for larger amplitude warps .
the range of behaviour allowed by equation ( [ eq1 ] ) is considerably more diverse than simple diffusion of @xmath17 .
nonetheless , we will make frequent use of the characteristic timescales for diffusion of surface density and warp , which we define conventionally as , @xmath18 the final term on the right hand side of equation ( [ eq1 ] ) represents the effect of lense - thirring precession on the disc . for a black hole with mass @xmath19
, we take @xmath20 , where @xmath21 where @xmath22 is the dimensionless angular momentum of the black hole . several assumptions and simplifications
are necessary to derive this equation , and these are set out in detail in papaloizou & pringle ( 1983 ) ( see also pringle 1999 ; ogilvie 1999 ) .
particularly important is the boundary between the diffusive behaviour described by equation ( [ eq1 ] ) and wave - like evolution , which occurs when the usual shakura & sunyaev ( 1973 ) viscosity parameter @xmath23 , where @xmath24 is the disc scale height ( papaloizou & lin 1995 ) .
the observational constraints on @xmath25 in agn discs are weak ( siemiginowska & czerny ( 1989 ) and siemiginowska , czerny & kostyunin ( 1996 ) suggest @xmath26 ) , but in the ionized parts of the disc it is plausible to assume that @xmath27 , as in better studied disc systems ( eg . cannizzo 1993 ) . in any case
, theoretical estimates for @xmath28 at the relevant radii ( collin - souffrin & dumont 1990 ) suggest that in agn @xmath29 , so that we are safely in the diffusive regime . in this respect agn discs
are very different from protostellar discs , for example , where wave - like behaviour is likely to be important ( e.g. larwood et al . 1996 ) .
we also note that although papaloizou & pringle ( 1983 ) considered only small amplitude warps describable using linear perturbation theory , ogilvie ( 1999 ) has shown that strongly warped discs can also be described using a similar approach .
scheuer & feiler ( 1996 ) obtained an approximate steady state analytic solution to equation ( [ eq1 ] ) for the case where @xmath30 , in the limit where the angle of misalignment of the disc and the hole spin was small . in this case
, the surface density obeys the usual relation for a planar disc with accretion rate @xmath31 and zero torque inner boundary condition imposed at @xmath32 , @xmath33 the shape of the disc is given by @xmath34 , where @xmath35 and @xmath36 .
the constant @xmath37 depends on the disc inclination @xmath38 at large radius , at large radius @xmath39 .
the above analytic solution applies only for @xmath40 , and as derived assumes that the disc warp is of small amplitude .
relaxing these assumptions requires a steady state numerical solution of equation ( [ eq1 ] ) . for @xmath41 ,
steady state solutions can be calculated efficiently by solving directly the ordinary differential equations ( odes ) obtained by setting @xmath42 in equation ( [ eq1 ] ) .
we give a full description of our methods in the appendix , but briefly we have found it simplest to use equations expressed in terms of the surface density @xmath43 and unit tilt vector @xmath44 , which can readily be solved iteratively using a finite - difference technique described by pereyra ( 1979 ) .
we use the numerical algorithms group s implementation , and impose boundary conditions of @xmath45 , @xmath46 , @xmath47 , @xmath48 .
these boundary conditions _ assume _ that the inner disc is aligned with the spin axis of the hole , we therefore have to check post facto that this is indeed a valid assumption for each model calculated . for @xmath49
, we have been unable to obtain a converged numerical solution to the steady state odes using these methods . in this regime , we instead evolve the time dependent equation ( [ eq1 ] ) from an arbitrary initial condition ( of a flat , uniformly tilted disc ) until a steady state is obtained , using the numerical method described by pringle ( 1992 ) . this is straightforward but computationally expensive , since many viscous times at the outer edge of the disc are required to obtain a steady state solution .
defining the viscous time of the disc by @xmath50 , we find that more than 10 viscous times of evolution are necessary .
moreover , for either method capturing the torque accurately requires a large range between the inner and outer disc radii , of at least @xmath51 . for these reasons ,
we are only able to obtain solutions using the time dependent code for high values of @xmath16 ( @xmath52 ) , when the viscous time is a weak function of increasing radius .
figure [ f1 ] shows the numerical steady - state solution for the @xmath40 case , and compares it to the approximate analytic solution .
we adopt units in which @xmath53 , and take @xmath54 .
we choose @xmath55 , which gives an inner aligned region extending out to around @xmath56 .
this is roughly appropriate for plausible agn parameters , though we defer until later consideration of detailed disc models .
convergence to the limiting inclination at infinity is slow , so it is necessary to impose the outer boundary condition of @xmath57 at a large radius , @xmath58 . at this radius @xmath59 and @xmath60 are chosen to facilitate comparison with the analytic solution of equation ( [ eq5 ] ) .
evidently , the analytic solution provides an excellent description of the disc shape .
the inner disc is aligned to the rotation axis of the hole out to close to @xmath56 , and warps to around half the limiting inclination by @xmath61 . moving inward from infinity ,
the warp is twisted by an angle of @xmath62 before being flattened effectively into the aligned plane . for an annulus in the disc , the local timescales for precession and for transmission of warp
are given by , @xmath63 we expect that the disc will be aligned with the spin axis of the hole at radii where the precession timescale is much shorter than the timescale over which the disc can communicate warp inwards
. the characteristic radius of the aligned region @xmath64 will then scale with the radius where @xmath65 , which is given by , @xmath66 where @xmath67 is a constant expected to be of the order of unity which we will use later to fit the numerical results .
note that this expression assumes both that the lense - thirring torque is strong enough to ensure that the inner disc _ is _ aligned , and that @xmath68 , so that @xmath69 increases with radius more rapidly than @xmath70 .
figure [ f2 ] shows how the disc shape varies with @xmath16 for a set of models with fixed @xmath71 and @xmath72 . from the numerical solutions ,
we compute @xmath64 as the radius where @xmath38 first exceeds some small inclination @xmath73 . with @xmath74
there is excellent agreement between the computed @xmath64 and the scaling given by equation ( [ eq9 ] ) , i.e. the simple timescale argument given above suffices to give an accurate idea of the scaling of the radius out to which the hole can enforce inner disc alignment .
the torque exerted on the misaligned disc as a result of the lense - thirring precession is given by the integral , @xmath75 of course an oppositely directed torque of the same magnitude is exerted on the black hole .
computing @xmath76 for the disc model shown in fig .
[ f1 ] we find that the torque agrees with that obtained from the scheuer & feiler solution at the level of a few percent ( the small discrepancy being primarily due to the ignored inner boundary condition in the analytic solution , which leads to an error in the surface density in the warp region ) .
the torque scales as @xmath77 , again as found analytically .
we have computed additional models with large inclination angles at infinity ( up to @xmath78 angle of misalignment ) and for those models there is a larger discrepancy between the numerical and analytic results , but still only at the tens of percent level .
figure [ f3 ] shows the tilt angle and surface density profile for three fiducial models , all normalised to the same accretion rate .
the solid line is computed for the parameters used for figure [ f1 ] , namely @xmath30 and @xmath79 .
the short dashed line maintains @xmath30 , but has @xmath80 , @xmath81 .
natarajan & pringle ( 1998 ) find that consideration of the hydrodynamics of a viscous disc suggests that in agn , @xmath82 in which case this ratio of viscosities would be appropriate for a plausible shakura - sunyaev @xmath25 of around 0.25 .
finally the long dashed line illustrates the effect of a declining surface density profile , plotted is a model with @xmath83 and @xmath84 .
all three of these models are chosen to have the same viscous timescale ( both for the surface density and for warp ) at the ( arbitrary ) choice of inner edge radius . in conventional thin disc models ( shakura & sunyaev 1973 ) , fixing the viscous time at a given radius corresponds to a fixed disc thickness @xmath28 at that radius .
we note that in such models the value of @xmath16 is fixed by solving for the vertical disc structure , i.e. it is not a ` parameter ' that can be varied but rather a property of the disc . in section 5
we discuss the appropriate value for @xmath16 in the region of the disc where the dominant torque arises .
also shown for each of the models is where radially the greatest contribution to the torque arises , we plot the components of the integrand in equation ( [ eq6 ] ) ( the extra factor of @xmath10 in the plotted quantity takes account of the logarithmic radial scale to give a true impression of where the strongest torque on the disc is found ) .
for all the models , strong torques arise from a broad region that extends from just outside the alignment radius out to one or two orders of magnitude larger radius .
the large radial range used in these calculations is thus essential to capture the torque accurately .
increasing the ratio of @xmath4 relative to @xmath3 reduces the timescale for diffusion of warp , @xmath85 , and allows the warp to push in closer to the black hole . at fixed accretion rate increasing @xmath4 even by this large factor leaves the surface density profile almost unaltered , and thus the net effect is simply to increase the magnitude of the torque and move it to smaller disc radius .
larger values of @xmath16 likewise decrease the warp timescale at large radius and lead to a shrinkage of the aligned region .
for the parameters used here , the disc shape is in fact rather similar for the two models in which @xmath86 at large radius .
however , for a given accretion rate , models with higher @xmath16 have lower surface density at large radii , which tends to decrease @xmath17 in equation ( [ eq6 ] ) and reduce the integrated torque .
both these effects have a significant impact on the calculation of the total torque .
the numerical results show that the scheuer & feiler ( 1996 ) solution , equation ( [ eq5 ] ) , provides an accurate description of the disc shape and the resultant torque on the black hole . for this solution ,
the torque integral in equation ( [ eq6 ] ) has a magnitude given by , @xmath87 guided by the numerical results , a straightforward extension of this result to non - zero @xmath16 is possible .
we first assume that the radius @xmath88 out to which we need to integrate equation ( [ eq6 ] ) to find the torque scales as does the aligned radius , @xmath89 if we additionally assume that the disc shape is primarily a function of @xmath88 , then the form of the integral in equation ( [ eq6 ] ) suggests the following scaling with @xmath16 , @xmath90 which reduces to the formula in equation ( [ eq7 ] ) if @xmath40 .
we have allowed ourselves a single adjustable parameter in this expression , to best match the numerical results we adopt @xmath91 .
figure [ f4 ] shows how equation ( [ eq10 ] ) compares to the numerical results for choices of parameters testing the scalings implied by the formula .
excellent agreement ( at better than the percent level ) is obtained for two models with @xmath84 and varying @xmath71 , while a single model computed with the time - dependent code at @xmath52 also agrees well with the fitting formula . a model with @xmath92 displays some systematic error , caused by the influence of the increased ratio of @xmath93 on the surface density profile in the warp region , but equation ( [ eq10 ] ) still provides a reasonable approximation to the numerical values for the torque .
the main result is that the torque for a fixed accretion rate ( and fixed viscous time at @xmath32 ) drops by roughly an order of magnitude for disc models with @xmath16 varying between @xmath30 and @xmath52 , while the dependence on other parameters ( @xmath71 , @xmath94 ) remains close to that found in the @xmath30 case .
the torque on the black hole due to the interaction with the warped disc causes it to precess and become aligned with the outer disc s angular momentum vector on a timescale that , to factors of order unity , is just @xmath95 ( e.g. scheuer & feiler 1996 ) .
@xmath96 is specified for a given disc model by equation ( [ eq10 ] ) , while the angular momentum of a hole of mass @xmath19 and spin parameter @xmath22 is @xmath97 .
the resulting expression for @xmath98 is similar , but not equivalent , to the estimate given by natarajan & pringle ( 1998 ) , @xmath99 is the angular momentum of the disc within a warp radius @xmath100 defined similarly to equation ( [ eq9x ] ) .
our expression takes more complete account of the need to integrate the torque for all radii less than @xmath88 . for models of the disc where most of the angular momentum within
the warp radius lies close to that radius , the final inferred timescale for alignment is similar .
obtaining an estimate for @xmath16 and @xmath101 requires a model for the vertical structure of the disc , which serves the purpose of relating the unknown central disc temperature , that controls the viscosity , to the effective temperature fixed by the requirement that viscous dissipation balance radiative losses . at the radii of the warp , gas pressure and an electron scattering opacity @xmath102 are dominant . in this regime ,
collin - souffrin & dumont ( 1990 ) obtain a radial profile of column density @xmath103 ( in units of @xmath104 ) described by , @xmath105 where @xmath106 is the radiative efficiency , @xmath107 is the bolometric luminosity , @xmath108 is the eddington luminosity , and @xmath109 . for this model , @xmath110 .
although this disc model includes the effects of irradiation of the upper layers of the disc by the central source , we note that the scaling @xmath111 is identical to that of non - irradiated models ( e.g. sincell & krolik 1998 ) .
the normalisation is within a factor of two .
other uncertainties thus dominate over that arising from different treatments of the vertical disc structure .
figure [ f5 ] shows the calculated alignment timescale for a @xmath112 black hole accreting at rates between @xmath113 and @xmath114 , where @xmath115 .
we take a radiative efficiency @xmath116 , and assume a maximal kerr hole , @xmath117 . for these parameters , @xmath88 as defined in equation ( [ eq9x ] ) is @xmath118 for reasonable ratios of @xmath93 , so the main warped region of the disc falls self - consistently within the regime where gas pressure and electron scattering opacity are dominant .
the figure shows results calculated assuming that @xmath119 , along with the expectation if @xmath120 for a range of plausible @xmath25 .
the latter models of course predict more rapid realignment , as noted previously ( natarajan & pringle 1998 ) . for holes accreting at rates of the order of the eddington limit , the alignment timescale is found to be @xmath121 - @xmath122 yr .
this is short compared to lifetimes of agn , _ irrespective _ of what one assumes for the ratio of @xmath93 .
producing the giant structures observed in radio galaxies may require an active epoch of the order of @xmath2 yr , while some estimates for the quasar lifetime are similar ( e.g. haehnelt & rees 1993 , though these estimates are subject to considerable uncertainty ) . in these systems
we expect that there was ample time for a rotating black hole , whatever its initial spin axis , to become aligned with the angular momentum of the disc .
thus jets accelerated from the inner disc should be perpendicular to the plane of the outer disc , and constancy in the direction of such jets implies a corresponding stability in the disc plane over time .
figure 6 plots the alignment timescale as a function of black hole mass , for a range of accretion rates .
the alignment timescale is found to be only weakly dependent on the mass of the hole , as expected given the @xmath123 dependence derived by natarajan & pringle ( 1998 ) .
for this figure we have assumed that @xmath84 , and that @xmath124 .
shorter alignment timescales are of course predicted if @xmath92 , as shown in figure 5 . for holes accreting at lower rates ( relative to the eddington limit ) ,
the timescale for alignment grows .
it is not unambiguously shorter than estimates for the active phase of a hole in a galactic nucleus , and the answer depends crucially on differences in the disc s response to shear and warp .
it remains true that @xmath125 , where @xmath126 , for any @xmath31 , so that alignment will always occur if an accretion event lasts long enough to contribute significantly to the hole mass . perhaps more probably , however , low luminosity agn may accrete negligible fractions of the hole mass in many brief episodes . in this scenario
, it remains possible that the hole spin accumulated over the whole accretion history will be sufficient to control the jet direction during subsequent low luminosity outbursts .
the outer disc would then not be expected to be generally perpendicular to the jet direction .
in this paper we have considered the interaction between a misaligned accretion disc and a rotating black hole for the parameters appropriate to active galactic nuclei .
a combination of forced precession and disc viscosity allows the black hole to force the inner disc into a plane perpendicular to the hole s spin axis , but the resulting warped disc exerts a torque that eventually aligns the rotation of the hole with the angular momentum of the outer disc .
we have presented steady - state numerical solutions to the full warped disc equations , verified that the analytic solution of scheuer & feiler ( 1996 ) provides an accurate representation of the disc shape and torque on the hole , and generalised the results for a range of disc models with varying viscosity .
the range in models we consider lead to almost an order of magnitude variation in the integrated torque on the hole .
we have computed the timescale for alignment to occur , and find that for holes accreting at rates of the order of the eddington limit ( for a @xmath112 hole , this implies @xmath127 , so these are luminous agn ) the timescale is short , @xmath121 - @xmath122 yr .
this is the case for a range of assumptions as to the ratio @xmath93 , which characterizes the disc s relative response to azimuthal shear and to warp .
the short derived timescale implies that the spin of black holes in such agn rapidly adjusts to match the angular momentum of accreting gas , even if the hole gains only a small ( @xmath128 10% ) of its mass from accretion . for low luminosity agn
the prediction is more model - dependent , but the timescale for alignment in these systems is sufficiently long that the spin axis of the hole may be able to remain stable if individual accretion episodes are brief . here
the assumed value for @xmath93 _ is _ crucial , and efforts to determine this ratio for specific angular momentum transport mechanisms and disc models would be valuable .
some powerful radio galaxies display jets that appear to have maintained their direction for long periods of time .
our results imply that this can not be due to any intrinsic stability imparted by the spin of the black hole , but instead must reflect a long term constancy in the angular momentum of the outer regions of the accretion disc .
if the gas derives from a reservoir set up by a single accretion event then this constancy would not be surprising .
alternatively , a preferred axis for gas arriving in the nuclear regions might be the consequence of interactions between inflowing gas and the galactic potential .
we have calculated the torque assuming that the disc is able to relax into and maintain a steady - state configuration as realignment occurs .
this assumption is justified in the strongly warped inner region , where @xmath129 and @xmath70 are indeed smaller than the derived @xmath98 , but fails at large radii of @xmath130 - @xmath131 where our solutions show that the disc maintains a small but significant warp .
this will not increase the alignment timescale ( before the disc reaches a steady - state the torque in our time dependent calculations is _ larger _ than the limiting values we solve for ) , but does suggest that following alignment of the hole the outer disc might retain a modest warp for a long period . for a @xmath112 hole a radius of @xmath132 corresponds to @xmath133 , a scale which is observable with vlbi in nearby agn such as ngc4258 ( eg .
miyoshi et al . 1995 ) .
some of the warps in maser discs could then simply reflect the slow decay of the initial conditions of a past accretion event , and be devoid of a persistent forcing mechanism .
of course ngc4258 itself has an accretion rate much lower than the models considered here ( gammie , narayan & blandford 1999 ) , and self - gravity is likely to be important in its disc ( papaloizou , terquem & lin 1998 ) , but the main point that @xmath70 at a few tenths of a parsec is likely to be long remains true .
a related question is whether the disc at @xmath134 is able to attain a steady state , even in the _ absence _ of a warp .
this region of the disc may be thermally unstable ( lin & shields 1986 ; clarke 1989 ; siemiginowska , czerny & kostyunin 1996 ) and prone to a limit cycle involving large changes in the accretion rate .
the interplay between such a cycle and the dynamics of a warped disc would be complex and intriguing , though a more promising setting for exploring such a scenario would be in galactic black hole systems , if those possess warped discs . for agn with rapidly spinning holes ,
the main warped region occurs at radii of @xmath135 . during realignment episodes
there is a substantial deposition of energy , originating in the spin energy of the hole , into this region of the disc .
this is particularly the case if @xmath8 , as we have argued is likely in this paper and previously ( natarajan & pringle 1998 ) . the disc in the vertical structure models we have considered is quite thin at these radii and can easily radiate locally a significantly enhanced luminosity , but substantial changes to the vertical structure might be expected ( note that we have assumed that the vertical structure of a warped disc is just like that of a planar disc , which is wrong on several counts ) .
moreover , the enhanced luminosity would increase the disc flux at optical and infrared wavelengths , which would in any event be raised due solely to the greater covering fraction of the warped disc as seen from the central source .
since the alignment timescale is short , systems of this kind would be rare most agn would harbour quite flat central discs but potentially bright .
we gratefully acknowledge useful discussions with jim pringle , mitch begelman , roger blandford , julian krolik and martin rees , and thank the referee for a very helpful report .
pja thanks space telescope science institute , where part of this paper was completed , for support and hospitality .
alexander p. , leahy j.p .
, 1987 , mnras , 225 , 1 bardeen j.m .
, petterson j.a , 1975 , apj , 195 , l65 blandford r.d . ,
znajek r.l . , 1977 , mnras , 179 , 433 cannizzo j.k . , 1993 ,
apj , 419 , 318 clarke c.j . , 1989 , mnras , 235 , 881 collin - souffrin s. , dumont a.m. , 1990 , a&a , 229 , 292 frank j. , king a.r . , raine d. , 1992 , _ accretion power in astrophysics _ , cambridge university press ( cambridge ) , p. 80
gammie c.f . , narayan r. , blandford r. , 1999 , apj , 516 , 177 ghosh , p. , abramowicz , m. a. , 1997 , mnras , 292 , 887 haehnelt m.g . , rees m.j . , 1993 , mnras , 263 , 168 hernquist l. , mihos j.c . , 1995 , apj , 448 , 41 kumar s. , pringle j.e . , 1985 , mnras , 213 , 435 larwood j.d . , nelson r.p
, papaloizou j.c.b .
, terquem c. , 1996 , mnras , 282 , 597 lense j. , thirring h. , 1918 , phys .
z. , 19 , 156 lin d.n.c . , shields g.a . , 1986 ,
apj , 305 , 28 liu r. , pooley g. , riley j.m . , 1992 ,
mnras , 257 , 545 livio m. , ogilvie g.i . , pringle j.e . , 1999 ,
apj , 512 , 100 maloney p.r . ,
begelman m.c .
, nowak m.a . ,
apj , 504 , 77 maloney p.r . ,
begelman m.c .
, pringle j.e . , apj , 474 , 582 miyoshi m. , moran j. , herrnstein j. , greenhill l. , nakai n. , diamond p. , inoue m. , 1995 , nature , 373 , 127 natarajan p. , pringle j.e . , 1998 , apj , 506 , l97 ogilvie g.i
. , 1999 , mnras , 304 , 557 papaloizou j.c.b . , lin d.n.c . , 1995 ,
apj , 438 , 841 papaloizou j.c.b .
, pringle j.e . , 1983 , mnras , 202 , 1181 papaloizou j.c.b . , terquem c. , lin d.n.c . , 1998 ,
apj , 497 , 212 pereyra v. , 1979 , in _
codes for boundary value problems in ordinary differential equations _ , eds . b. childs , m. scott , j.w .
daniel , e. denman & p. nelson , springer - verlag , lecture notes in comp .
, 76 press w.h . ,
teukolsky s.a .
, vetterling w.t . ,
flannery b.p . , 1992 , in _ numerical recipes in fortran _ , second edition , cambridge university press , chapter 17 pringle j.e . , 1992 , mnras , 258 , 811 pringle j.e . , 1996 , mnras , 281 , 357 pringle j.e . , 1997 , mnras , 292 , 136 pringle j.e . , 1999 , in
astrophysical discs _ , eds . j. a. sellwood and j. goodman , asp conf . ser .
160 , p. 53
rees m.j . , 1978 , nature , 275 , 516 scheuer p.a.g . , 1995 , mnras , 277 , 331 scheuer p.a.g . , feiler r. , 1996 , mnras , 282 , 291 shakura n.i . , sunyaev r.a . , 1973 , a&a , 24 , 337 siemiginowska a. , czerny b. , 1989 , mnras , 239 , 289 siemiginowska a. , czerny b. , kostyunin v. , 1996 , apj , 458 , 491 sincell m.w .
, krolik j.h . , 1998 ,
apj , 496 , 737
the odes for the steady state shape of the disc are most easily solved by expressing equation ( [ eq1 ] ) in terms of separate equations for the tilt vector @xmath11 and surface density @xmath136 . for a keplerian potential
these are ( pringle 1992 ) , @xmath137 { { \partial { \bf \hat{l } } } \over { \partial r } } \nonumber
\\ + { \partial \over { \partial r } } \left ( { 1 \over 2 } \nu_2 { { \partial { \bf \hat{l } } } \over { \partial r } } \right ) + { 1 \over 2 } \nu_2 \left\vert { { \partial { \bf \hat{l } } } \over { \partial r } } \right\vert^2 { \bf \hat{l } } + { { \vec \omega_p } \times { \bf \hat{l } } \over r^3 } , \label{app1}\end{aligned}\ ] ] where the primes denote derivatives with respect to @xmath10 , and @xmath138 \nonumber \\ & + & { 1 \over r } { \partial \over { \partial r } } \left [ \nu_2 \sigma r^2 \left\vert { { \partial { \bf \hat{l } } } \over { \partial r } } \right\vert^2 \right ] .
\label{app2 } \end{aligned}\ ] ] setting the time derivatives to zero these constitute 3 second order equations for @xmath136 and any two components of @xmath11 .
we solve these with the boundary conditions described in section 2.3 using the nag routines d02gaf and d02raf , which implement the finite difference technique with deferred correction described by pereyra ( 1979 ) ( for a general introduction to such methods , see e.g. press et al .
an iterative approach is employed , in which we first solve equation ( [ app2 ] ) assuming zero tilt , and then use the resulting surface density profile in the solution of equation ( [ app1 ] ) . the solution for the tilt vector @xmath11 is then recycled for use in equation ( [ app2 ] ) , and we loop until convergence is achieved . typically only a small number of iterations are required , since the changes to the surface density profile for even quite strong warps vary smoothly with changing disc shape .
a finely space finite difference mesh is required to obtain a solution using this scheme , we use between 8000 and 16,000 mesh points evenly spaced in @xmath139 .
even so , we have found that obtaining solutions with this scheme is still rather difficult , especially when @xmath140 or @xmath141 .
for these cases , we start with the easy @xmath40 , @xmath84 solution , and step towards the desired parameters using the previous solution at each step as the initial guess .
we have been unable to obtain solutions using the above method for large values of @xmath16 . in this regime , we instead evolve directly equation ( [ eq1 ] ) using the numerical code described by pringle ( 1992 ) until a steady - state is obtained .
this code uses an explicit first order finite difference technique , which conserves angular momentum to machine precision .
we have modified the boundary conditions to be as close as possible to those described in section 2.3 , so that the results are directly comparable between the two solution methods .
the time dependent code is vastly more expensive to run , for the runs described here we are restricted to 100 radial grid points , again spaced evenly in @xmath139 .
nevertheless this still provides reasonable accuracy for computing the torque . | warped accretion discs in active galactic nuclei ( agn ) exert a torque on the black hole that tends to align the rotation axis with the angular momentum of the outer disc .
we compute the magnitude of this torque by solving numerically for the steady state shape of the warped disc , and verify that the analytic solution of scheuer & feiler ( 1996 ) provides an excellent approximation .
we generalise these results for discs with strong warps and arbitrary surface density profiles , and calculate the timescale on which the black hole becomes aligned with the angular momentum in the outer disc . for massive black holes and accretion rates of the order of the eddington limit
the alignment timescale is always short ( @xmath0 yr ) , so that jets accelerated from the inner disc region provide a prompt tracer of the angular momentum of gas at large radii in the disc .
longer timescales are predicted for low luminosity systems , depending on the degree of anisotropy in the disc s hydrodynamic response to shear and warp , and for the final decay of modest warps at large radii in the disc that are potentially observable via vlbi .
we discuss the implications of this for the inferred accretion history of those active galactic nuclei whose jet directions appear to be stable over long timescales .
the large energy deposition rate at modest disc radii during rapid realignment episodes should make such objects transiently bright at optical and infrared wavelengths .
psfig accretion , accretion discs black hole physics
galaxies : active galaxies : jets galaxies : nuclei |
at least three of the four currently - known soft gamma repeaters are associated with slowly rotating , extremely magnetized neutron stars located within young supernova remnants ( kouveliotou et al .
1998 , 1999 ) .
they are characterized by the recurrent emission of gamma - ray bursts with relatively soft spectra ( resembling optically - thin thermal bremsstrahlung at @xmath040 kev ) and short durations ( @xmath1 0.1 s ) ( kouveliotou 1995 ) .
thompson and duncan ( 1995 ) suggested that these bursts are due to neutron star crust fractures , driven by the stress of an evolving , ultra - strong magnetic field , @xmath2 gauss .
cheng , epstein , guyer & young ( 1996 ) observed that particular statistical properties of a sample of 111 sgr events from sgr 1806 - 20 are quite similar to those of earthquakes ( eq ) .
these properties include the distribution of event energies , which follow a power law dn @xmath3 e@xmath4 de with an exponent , @xmath5 = 1.6 .
a similar distribution was obtained empirically by gutenberg and richter ( 1956a ; 1965 ) for the distribution of eq energies , with power law index @xmath6=@xmath7 ; and in computer simulations of fractures in a stressed , elastic medium ( katz 1986 ) .
the distribution of time intervals between successive sgr 1806 - 20 events is well described by a log - normal distribution analogous to the waiting times distribution of microglitches seen in the vela pulsar ( see hurley et al .
cheng et al .
( 1996 ) also showed that cumulative waiting time distributions of sgr 1806 - 20 and eq events are similar .
these results support the idea that sgr bursts are caused by starquakes , as expected to occur in the crusts of magnetically - powered neutron stars , or magnetars " ( duncan & thompson 1992 ; thompson and duncan 1995 , 1996 ) . in may 1998 , sgr 1900 + 14 became extremely active after a long period during which only sporadic activity occurred ( kouveliotou 1993 ) . in the period from may 1998 until january 1999 a total of 200 events were detected ( woods et al .
1999b ) with the burst and transient source experiment ( batse ) aboard the compton gamma ray observatory ( cgro ) . out of these 200 events , 63 led to an on - board trigger .
the sudden change in source activity initiated a series of rossi x - ray timing explorer ( rxte ) observations between may 31 and december 21 , 1998 . during these observations , 837 bursts from sgr 1900
+ 14 were detected with the proportional counter array ( pca ) . as noted by kouveliotou et al .
( 1998 ) for sgr 1806 - 20 , the bursts occur in an apparently irregular temporal pattern .
this is also true for sgr 1900 + 14 bursts . in this _ letter _ , we study the statistics of sgr bursting using the new measurements of sgr 1900 + 14 .
our data base of events is larger by @xmath8 than that of previous statistical studies , and extends over a larger dynamic range in burst energy ( or fluence ) by @xmath9 .
the batse instrument is made up of 8 identical detector modules located on each corner of the cgro .
each module contains a large area detector ( lad ) and a spectroscopy detector ( sd ) . in our analysis , we have used discriminator lad ( discla ) data with coarse energy resolution ( 4 channels covering e@xmath1025 kev ) , spectroscopy time - tagged event ( stte ) data and spectroscopy high energy resolution burst ( sherb ) data with fine energy binning ( 256 channels ) . a detailed description of batse instrumentation and data types can be found in fishman et al .
( 1989 ) .
batse was triggered by sgr 1900 + 14 bursts 63 times between may 1998 and january 1999 . for 22 of the brightest events , we obtained stte or sherb data with detailed spectral information .
we fit the background subtracted source spectra to optically - thin thermal bremsstrahlung ( ottb ) and power law models . the ottb model , f(e)@xmath3 e@xmath11@xmath12(@xmath13e / kt ) ,
provides suitable fits ( 0.83 @xmath14 @xmath15 @xmath14 1.32 ) to all of the event spectra with temperatures ranging between 21.0 and 46.9 kev .
the power law model failed to fit most of the spectra .
the mean of the ottb temperatures for this sample of 22 events , appropriately weighted by uncertainties , is @xmath16 kev .
woods et al .
( 1999b ) performed an extensive search for untriggered batse events from sgr 1900 + 14 .
they found , in addition to the 63 triggered events , 137 untriggered burst events between 24 may 1998 and 3 february 1999 . in this study we selected the 187 batse events ( triggered and untriggered ) which had discla data .
this data type is read out continuously ( with the exception of data telemetry gaps ) and therefore is available for the largest sample of events .
we have excluded events which occurred in data gaps , events that were too weak to fit , and four events due to their distinction from typical sgr activity . the events on 1998 october 22 and 1999 january 10 with relatively hard spectra ( woods et al .
1999c ) , and the multi - episodic events on 1998 may 30 and september 1 ( g et al
. 1999b ) will be discussed elsewhere . given the long discla data integration time ( 1.024 s ) relative to typical burst durations ( @xmath1 0.1 s )
, we could only estimate the fluence for each event . in order to determine the fluence of each burst
, we fit the background - subtracted source spectrum to the ottb model with a fixed kt of 25.7 kev , a reasonable choice considering the fairly narrow kt distribution of the triggered bursts .
we find that the fluences of sgr 1900 + 14 bursts observed with batse range between @xmath17 and @xmath18 ergs @xmath19 .
for an estimated distance to sgr 1900 + 14 of 7 kpc ( vasisht et al .
1994 ) , and assuming isotropic emission , the corresponding energy range is @xmath20 @xmath21 ergs .
rxte observed sgr 1900 + 14 for a total exposure time of @xmath1 180 ks between may 1998 and december 1998 . in this work ,
we have analyzed data from 32 pointed observations with the pca .
we performed an automated burst search similar to the one used on batse data described by woods et al .
( 1999b ) .
using standard 1 data ( 2 - 60kev ) for all times where the source was above the earth s horizon by more than 5 @xmath22 , we searched for bursts using the following methodology .
for each 0.125 s bin , a background count rate was estimated by fitting a first order polynomial to 5 s of data before and after each bin with a 3 s gap between the bin searched and the background intervals .
bins with count rates exceeding 1000 counts s@xmath11 were assumed to contain burst emission and were excluded from background intervals . at the beginning(end ) of each continuous stretch of data , extrapolations of background fits after(before ) the bin were used to estimate the background count rate within the bin searched .
a burst was defined as any continuous set of bins with count rates in excess of 5.5 @xmath23 above the estimated background .
the count fluence of each burst was measured by simply integrating the background - subtracted counts over the bins covering the event . in order to compare integrated counts obtained with the pca and batse fluences
, we determined a conversion factor between each pca count and batse fluence .
we assume a constant spectral model ( ottb with kt=25.7 kev ) .
first , we searched for simultaneous bursts observed with both instruments and we found 13 events ( 6 triggered events , 3 in the read - out of triggered events and 4 untriggered events in batse ) .
we then computed the ratio of the batse fluence of each simultaneous event to the pca counts , which ranges over a factor of @xmath1 2 between @xmath24 and @xmath25 ergs @xmath19counts@xmath11 .
the weighted mean of the ratios for sgr 1900 + 14 is @xmath26 ergs @xmath19counts@xmath11 and the standard deviation , @xmath23 = @xmath27 ergs @xmath19counts@xmath11 . invoking this conversion factor , the fluence of the bursts from pca extends from @xmath28 to @xmath29 ergs @xmath19 ( in the batse energy range , e @xmath10 25 kev ) and the burst energies range from @xmath30 to @xmath31 ergs .
_ i ) fluence distributions _ : the fluences of batse bursts were binned in equally spaced logarithmic fluence steps ( @xmath32 ) ( fig.1 ) . using a standard least squares fitting method , we fit a power law model to data between @xmath33 and @xmath34 ergs @xmath19 .
bursts at the low end of the distribution were excluded because of diminished detection efficiency as well as at the high end due to undersampling of the intrinsic distribution .
the power law exponent obtained is @xmath35 ( solid line passing through batse data in fig.1 ) , which corresponds to dn @xmath3 e@xmath36 de .
we also employed a maximum likelihood analysis , instead of the least squares method , to fit a power law model to the unbinned fluence values within the same interval of fluences .
this method yields @xmath5=@xmath37 for the energy exponent which agrees well with the least squares fit . using the conversion factor we derived from rxte counts to batse fluence for sgr 1900 + 14
, we determined the fluence of each rxte burst ( in the batse energy range ) and distributed them over the same logarithmic fluence steps ( fig.1 ) .
we first fit binned rxte fluences between @xmath38 and @xmath29 ergs @xmath19 to a power law model using least squares method which gives an exponent value of @xmath39 ( solid line passing through rxte data in fig.1 ) .
the unbinned fluences were then fit to the same model using the maximum likelihood method obtaining @xmath40 for the power law exponent . combined rxte and batse fluences range from @xmath28 to @xmath41 ergs @xmath19 ( fig.1 ) which demonstrates that power law distribution of energies with an exponent @xmath42 is valid for sgr 1900 + 14 over 4 orders of magnitude .
_ ii ) waiting times statistics _ : we have measured the waiting times ( @xmath43 t ) between successive bursts , uninterupted by earth occultation and data gaps , for 779 events .
fig 2 shows the distribution of waiting times which range from 0.25 to 1421 s. we fit the ( @xmath43t)-distribution to a log - normal function and found a peak at @xmath1 49 s. the solid line in fig.2 shows the interval used to fit and the dashed lines are the extrapolations of log - normal distribution .
we do not include waiting times less than 2 s since these bursts appear to be double peaked events in which the second burst peak appears shortly after the first one , although recorded as two distinct bursts .
we were unable to generate a @xmath43t - distribution for batse bursts due to the much smaller number of events which occurred during a single orbital window . in order to investigate any relations between waiting times till the next burst ( @xmath43@xmath44 ) and the intensity of the bursts
, we divided the 779 events sample into 8 intensity intervals each of which contains approximately 100 events .
we fit the @xmath43@xmath44-distribution to a log - normal distribution and determined the mean-@xmath43@xmath44 ( i.e. where the fitted log - normal distribution peaks ) , and the mean counts for each of the 8 groups .
we show in fig.3-a that there is no correlation between @xmath43@xmath44 and energy of the bursts ( spearman rank - order correlation coefficient , @xmath45 = 0.05 and the probability that this correlation occurs by a random data set , p = 0.91 ) .
we also searched for the relation between the elapsed times since the previous burst ( @xmath43@xmath46 ) and the intensity of the bursts .
similar to the previous case , we sub - divided the events into 8 intensity intervals and determined mean-@xmath43@xmath46 by fitting to a log - normal distribution and mean counts for each group individually .
fig.3-b shows that there appears to be an anti - correlation between mean-@xmath43@xmath46 and burst energy ( @xmath45 = @xmath47 , p = 8@xmath48 ) . _
iii ) burst durations _ : gutenberg and richter ( 1956a ; 1956b ) demonstrated that there is a power law relation between the magnitude , or energy of the eq events and the durations of the strong motion at short distances from an eq region . in order to investigate
if a similar correlation exists for sgr events , we selected all 679 pca bursts from the most active period of sgr 1900 + 14 . in order to determine the durations of the bursts accurately , we used event mode pca data with 1 ms time resolution . for 281 of the bursts
selected , we obtained t@xmath49 durations ( koshut et al .
1996 ) of the bursts .
fig.4 shows that burst energies and durations are correlated ( @xmath45=0.54 , p@xmath1@xmath50 ) , although there is a significant spread of fluences at a given duration .
the power - law size distribution of sgr 1900 + 14 bursts with an index @xmath5 = 1.66 is similar to those found for sgr 1806 - 20 ( cheng et al . 1996 ) and sgr 1627 - 41 ( woods et al .
the lack of a high energy cut - off in the differential size distribution indicates that the highest energy events are not well sampled in our distribution .
the distribution of waiting times between successive sgr 1900 + 14 bursts is characterized by a log - normal function , similar to that of sgr 1806 - 20 ( hurley et al . 1994 ) . waiting times between sgr 1900 + 14 bursts are on average shorter than those of sgr 1806 - 20 since all sgr 1900 + 14 bursts occurred during the most active period of the source .
there is no correlation between the intensity of the burst and the waiting time until the following burst .
this result agrees well with the results of laros et al .
( 1987 ) for sgr 1806 - 20 and distinguishes the physical mechanism of sgr 1900 + 14 bursts from that of type ii x - ray bursts from the rapid burster ( lewin et al . 1976 )
in which the burst energy is proportional to the waiting time till the next burst .
we find an anti - correlation between the intensity of the bursts and the waiting time since the previous bursts .
this is very different from type i x - ray bursts ( thermonuclear flashes , see lewin , van paradijs and taam 1993 ) for which there is a rough positive correlation .
there is evidence of a positive correlation between the energy and the duration of sgr 1900 + 14 bursts .
similar behavior was also observed for eqs ( gutenberg & richter 1956a , 1956b ) and solar flares ( lu et al .
1993 ) .
the eq size distribution appears to be a power law with an exponent between 1.4 - 1.8 independent of geographic location ( gutenberg & richter 1956a,1965 ; lay & wallace 1995 ) . using data taken from the solar maximum mission ( smm ) crosby et al .
( 1993 ) found a power law size distribution for 12000 solar flares with exponents ranging between 1.53 and 1.73 .
the smm results have been confirmed by the results from international cometary explorer ( ice ) for 4350 flares that finds an exponent of 1.6 ( lu et al .
gershberg and shakhovskaya ( 1983 ) found that the size distribution of stellar flares from 23 stars display power law with exponent between 1.5 and 2.1 .
chen et al . (
1991 ) argued that eq dynamics is described by a self - organized critical system .
crosby et al . ( 1993 ) similarly suggested that the size distribution of solar flares reflects an underlying system in a state of self - organized criticality ( see bak et al .
1988 ) which states that many composite systems will self - organize to a critical state in which a small perturbation can trigger a chain reaction that affects any number of elements within the system .
we have been unable to find clear results in the literature on the distribution of waiting times between successive solar flares or eqs .
wheatland et al .
( 1998 ) predicted that the distribution of waiting times of solar flares displays a power law , while biesecker ( 1994 ) proposed that it is consistent with a time - dependent poisson process .
nishenko and bulland ( 1987 ) showed that waiting time distribution of large eqs is well described by a log - normal function . in recent work by nadeau and mcevilly ( 1999 ) there is an evidence of log - normal distribution of waiting times between micro eqs .
the large number of bursts in our samples allow us to stringently test the power law size distribution proposed by cheng et al .
( 1996 ) ; we find that the size distribution of sgr 1900 + 14 bursts follows a power - law of index 1.66 over more than four orders of magnitude in burst fluence .
this behavior , along with a log - normal waiting time distribution and energy - correlated burst durations , are characteristics of self - organized critical systems in general , and earthquakes and solar flares in particular . in the magnetar model ,
the triggering mechanism for sgr bursts is a hybrid of starquakes and magnetically - powered flares ( thompson & duncan 1995 ) . when magnetic stresses induce elastic strains in the crust , the stored potential energy is predominantly magnetic rather than elastic .
in contrast with an eq , this allows much of the energy to be released directly into a propagating disturbance of the external magnetic field of the neutron star ; and in contrast with a solar flare it is the rigidity of the crust that provides a gate or trigger for the energy release .
the extended power - law distribution of burst fluences suggests that the average radiative efficiency does not vary significantly over four orders of magnitude in burst energy , and provides a strong constraint on burst emission models ( thompson et al .
1999 ) .
we thank dr .
robert d. preece for adjusting wingspan software to process untriggered batse events and dr .
markus j. aschwanden for helpful discussions on solar flares .
we acknowledge support from the cooperative agreement ncc 8 - 65 ( eg ) ; nasa grants nag5 - 3674 and nag5 - 7060 ( jvp ) ; texas advanced research project grant arp-028 and nasa grant nag5 - 8381 ( rcd ) .
bak , p. , tang , c. & wiesenfeld , k. , 1988 , , 38 , 364 biesecker , d.a . , 1994 , ph.d .
thesis , univ .
new hampshire chen , k. , bak , p. & obukhov , s.p .
, 1991 , , 43 , 625 cheng , b. , epstein , r.i .
, guyer , r.a . &
young , a.c .
1996 , , 382 , 518 crosby , n.b .
, et al . 1993 , , 143 , 275 duncan , r.c . & thompson c. 1992 , 392 , l9 fishman , g.j . , et al .
1989 , compton observatory science workshop , ed . w.n .
johnson , nasa conference publication , 2 gershberg , r.e .
& shakhovskaya , n.i . , 1983 , , 95 , 235 , e. et al .
, 1999b , in preparation gutenberg , b. & richter , c.f .
1956a , bull .
, 46 , 105 gutenberg , b. & richter , c.f . 1956b , ann . geofis .
, 9 , 1 gutenberg , b. & richter , c.f .
1965 , seismicity of the earth , hafner pub . ,
p.16 hurley , k.j .
, et al . 1994 , , 288 , l49 katz , j.i . , 1986 , j. geophys .
res . , 91 , 10412 kouveliotou , c. , et al .
1993 , , 362 , 728 kouveliotou , c. , 1995 , , 231 , 49 kouveliotou , c. , et al .
1998 , , 393 , 235 kouveliotou , c. , et al .
1999 , , 510 , l115 koshut , t.m . ,
1996 , , 463 , 570 laros , j.g . , et al 1987 , , 320 , l111 lay , t. & wallace , t. , 1995 , modern global seismology , academic press , p.376 lewin , w.h.g , et al . , 1976 , , 207 , l95 lewin , w.h.g , et al
, 1993 , space sci .
, 62 , 223 lu , e.t . , et al . , 1993 , , 412 , 841 nadeau , r.m .
& mcevilly , t.v . , 1999 , science , 285 , 718 nishenko , s.p .
& buland , r. , 1987 , bull .
, 77 , 1382 thompson , c. & duncan , r.c .
1995 , , 275 , 255 thompson , c. & duncan , r.c .
1996 , , 473 , 322 thompson , c. et al . 1999 , in preparation .
vasisht , g. , et al .
1994 , , 431 , l35 wheatland , m.s . ,
1998 , , 509 , 448 woods , p. , et al . 1999a , , 519 , l139 woods , p. , et al .
1999b , , in press woods , p. , et al .
1999c , in preparation | we study the statistics of soft gamma repeater ( sgr ) bursts , using a data base of 187 events detected with batse and 837 events detected with rxte pca , all from sgr 1900 + 14 during its 1998 - 1999 active phase .
we find that the fluence or energy distribution of bursts is consistent with a power law of index 1.66 , over 4 orders of magnitude .
this scale - free distribution resembles the gutenberg - richter law for earthquakes , and gives evidence for self - organized criticality in sgrs .
the distribution of time intervals between successive bursts from sgr 1900 + 14 is consistent with a log - normal distribution .
there is no correlation between burst intensity and the waiting times till the next burst , but there is some evidence for a correlation between burst intensity and the time elapsed since the previous burst .
we also find a correlation between the duration and the energy of the bursts , but with significant scatter .
in all these statistical properties , sgr bursts resemble earthquakes and solar flares more closely than they resemble any known accretion - powered or nuclear - powered phenomena .
thus our analysis lends support to the hypothesis that the energy source for sgr bursts is internal to the neutron star , and plausibly magnetic . |
recent years have witnessed the emergence of the idea of an accelerating universe and due to some observational results @xcite , it is now established that the universe is accelerating .
this signifies a paradigm shift in cosmological research from _ expanding universe _ to _ accelerated expanding universe_. now , the problem lies in detecting an exotic type of unknown repulsive force , termed as dark energy , which is driving this acceleration .
a key factor in dark energy investigation is the equation of state parameter @xmath3 , which relates pressure and density through an equation of state of the form @xmath4 . due to lack of observational evidence in making a distinction between constant and variable @xmath3 , usually the equation of state parameter
is considered as a constant @xcite with phase wise values @xmath5 , @xmath6 , @xmath7 and @xmath8 for dust , radiation , vacuum fluid and stiff fluid dominated universe respectively .
but in general , @xmath3 is a function of time or redshift @xcite .
for instance , quintessence models involving scalar fields give rise to time - dependent @xmath3 @xcite .
there is literature available on models with varying fields , such as cosmological model with a time dependent equation of state in a kaluza - klein metric , cosmological model with a viscous fluid in a kaluza klein metric and wormholes with varying equation of state parameter@xcite .
so , there are enough grounds for considering @xmath3 as time - dependent for a better understanding of the cosmic evolution .
now , various types of physical models including phenomenological ones ( with time - dependent @xmath0 ) are contemplated for unveiling the nature of dark energy .
it may be mentioned here that phenomenological time - varying @xmath0 models were proposed initially even before the emergence of the idea of cosmic acceleration for solving the well known cosmological constant problem @xcite . at present , various kinematical @xmath0 models of phenomenological character are proposed for explaining the present acceleration . in a recent work , ray et al . @xcite
have shown the equivalence of three phenomenological @xmath0 models , viz .
@xmath9 , @xmath10 and @xmath2 in four dimensional space - time for constant @xmath3 while the same three models of @xmath0 were shown to be equivalent in five dimensional kaluza - klein type model by pradhan et al.@xcite treating @xmath3 as a constant quantity .
so , it is quite natural to investigate the behaviour of the above three models when the equation of state parameter is a function of time .
in fact , mukhopadhyay et al .
@xcite have already made an attempt in this direction by investigating two other dynamical @xmath0 models , viz . @xmath11 and @xmath12 with variable equation of state parameter . in the present work , an investigation of the models @xmath1 and
@xmath2 are made under the _ ansatz _ @xmath13 both in four and five dimensional space - time . for a flat universe ,
these two models are shown to be equivalent for variable @xmath3 both in four and five dimensions .
some other features , revealed by letting @xmath3 be a function of time , have also been observed .
the einstein field equations are given by @xmath14,\end{aligned}\ ] ] where the cosmological term @xmath0 is time - dependent , i.e. @xmath15 and @xmath16 , the velocity of light in vacuum , is assumed to be unity .
let us consider the robertson - walker metric @xmath17,\end{aligned}\ ] ] where @xmath18 , the curvature constant , assumes the values @xmath7 , @xmath5 and @xmath8 for open , flat and closed models of the universe respectively and @xmath19 is the scale factor .
for a flat universe ( @xmath20 ) and the spherically symmetric metric ( 2 ) , field equations ( 1 ) yield friedmann and raychaudhuri equations respectively , given by @xmath21 @xmath22 where @xmath23 and @xmath24 are the cosmic matter - energy density and pressure respectively and the hubble parameter @xmath25 is related to the scale factor by @xmath26 .
let us now choose the barotropic equation of state @xmath27 here we assume that the equation of state parameter @xmath3 is time - dependent i.e. @xmath13 such that @xmath28 , where @xmath29 is a constant having dimension of time . from equation ( 3 ) , for flat universe
, we get @xmath30 using equation ( 5 ) and ( 6 ) in ( 4 ) one can get , after some manipulation , the following differential equation @xmath31 let us use the _ ansatz _ @xmath32 ( where @xmath33 is a free parameter ) along with @xmath34 . then equation ( 7 ) reduces to @xmath35 solving equation ( 8) , we get @xmath36 writing @xmath26 in equation ( 9 ) and integrating it further we get our solution set as @xmath37 @xmath38 @xmath39 where @xmath40 is an integration constant .
it is interesting to note that putting @xmath41 ( i.e. @xmath42 ) and * @xmath43 * in equations ( 9 ) , ( 11 ) and ( 12 ) we can recover the same expressions for @xmath44 , @xmath45 and @xmath46 respectively as those obtained by ray et al . @xcite .
but , equation ( 10 ) indicates that @xmath47 can not be equal to zero .
moreover , the expression for @xmath48 in equation ( 10 ) differs from that of ray et al .
so , in general , the constant @xmath3 of ray et al .
@xcite and time - varying @xmath3 of this case present us with different situations . by using equations ( 6 ) and ( 9 )
, we get @xcite @xmath49 where , in absence of any curvature , matter density @xmath50 and dark energy density @xmath51 are related by the equation @xmath52 using the _ ansatz _ @xmath53 ( where @xmath54 is a parameter ) in equation ( 7 ) we get the differential equation @xmath55 solving equation ( 15 ) , we get @xmath56 again , solving equation ( 16 ) we get our solution set as @xmath57 @xmath58 @xmath59 where @xmath60 is an integration constant . in this case also , for @xmath41 and @xmath43 we get back the same expressions for @xmath44 , @xmath45 and @xmath46 as those obtained by ray et al . @xcite . but here again @xmath41 is forbidden via equation ( 17 ) . thus for the @xmath61 case also , variable @xmath3 and constant @xmath3 present us with different situations .
kaluza - klein type robertson - walker metric is given by @xmath62.\end{aligned}\ ] ] for the metric ( 20 ) , the field equations ( 1 ) yield the following two differential equations @xmath63 @xmath64 where @xmath65 . from ( 21 )
, for flat universe ( @xmath20 ) , we can write @xmath66 using equations ( 5 ) and ( 23 ) , for @xmath20 , we obtain @xmath67 let us substitute @xmath68 in equation ( 24 ) , where @xmath29 has the same significance as before . then from equation ( 24 ) , after use of the _ ansatz _ @xmath32 , we get @xmath69 solving equation ( 26 ) , we have @xmath70 again , writing @xmath26 in equation ( 26 ) and solving the resulting differential equation we get our solution set as @xmath71 @xmath72 @xmath73 where @xmath74 is an integration constant . equation ( 27 ) shows that for physical validity @xmath75 .
also , from equation ( 26 ) we find that @xmath76 if @xmath29 is positive .
but , equation ( 29 ) tells us that for a repulsive @xmath0 , @xmath77 . so combining these two cases we can write @xmath78 . using the _ ansatz _ @xmath79 , equation ( 24 ) reduces to @xmath80 solving equation ( 30 ) , we get @xmath81 writing @xmath26 in equation ( 31 ) and solving the resulting differential equation
, we get our solution set as @xmath82 @xmath83 @xmath84 where @xmath85 is an integration constant . equation ( 32 ) shows that , for physical validity @xmath75 . again , by equation ( 31 ) , @xmath86 for a positive @xmath29 , whereas equation ( 34 ) tells us that for repulsive @xmath0 either @xmath87 or @xmath88 .
so , a positive @xmath54 will make all the equations ( 31 ) , ( 33 ) and ( 34 ) physically valid .
using equations ( 6 ) and ( 9 ) it is easy to obtain @xcite @xmath49 where , in the absence of any curvature , matter - energy density @xmath50 ( @xmath89 ) and dark energy density @xmath90 ( @xmath91 ) are related by the equation @xmath52 similarly , using equation ( 16 ) and ( 18 ) , we find that @xmath54 is related to @xmath50 and @xmath51 through the relation @xmath92 now , by the use of equations ( 35 ) and ( 37 ) we find that equations ( 10 ) and ( 17 ) differ by a constant while the equations ( 11 ) and ( 12 ) become identical to equations ( 18 ) and ( 19 ) .
this means that , @xmath93 and @xmath61 models are equivalent . coming to the five dimensional kaluza - klein model , we find that with the help of equations ( 23 ) and ( 26 ) , one can easily obtain @xmath94 for @xmath93 model .
it is to note that equation ( 36 ) still holds in the 5d case .
similarly , using equations ( 31 ) and ( 33 ) , we get @xmath95 for @xmath61 model . with the help of equations ( 38 ) and ( 39 )
it is easy to show that the equations ( 27 ) and ( 32 ) differ by a constant while equations ( 29 ) and ( 30 ) become identical to the equations ( 33 ) and ( 34 ) respectively .
this implies that @xmath93 and @xmath61 models are equivalent for five dimensional space - time also .
apart from being a mere mathematical parameter , @xmath29 bears some deeper physical significance also . in some works
@xcite it has been shown that @xmath29 represents the time - scale of evaporation of bose - einstein condensates which include a time - dependent @xmath0 . in the present work , @xmath29
is linked with the equation of state parameter @xmath3 and cosmic time @xmath96 .
the exact nature of @xmath3 and @xmath96 dependence of @xmath29 may be explored elsewhere after @xmath47 is specified properly .
although in the present work the equation of state parameter @xmath3 is taken as time - dependent , it can be a function of the red - shift @xmath97 or scale factor @xmath98 as well .
the red - shift dependence of @xmath3 can be linear like @xmath99 with @xmath100 @xcite or non - linear as @xmath101 @xcite .
now , in the present work , the structure of @xmath3 suggests that it is of the form @xmath102 which can be regarded as a generalization of the special form @xmath103 @xcite .
now , @xmath29 can be expressed as @xmath104 since @xmath29 has dimensions of time , it can not be negative .
this implies that @xmath105 for integral values of @xmath47 .
sn ia data suggest that @xmath106 @xcite while the limit imposed on @xmath3 by a combination of sn ia data ( with cmb anisotropy ) and galaxy clustering statistics @xcite is @xmath107 .
so , if the present work is compared with the above - mentioned experimental results then , one can conclude that the limit of @xmath3 provided by the equation ( 40 ) may be accommodated within the acceptable range ( figs 1 and 2 ) .
also it is clear that for the present dust - filled universe ( @xmath42 ) , @xmath29 is equal to @xmath96 whereas for vacuum fluid @xmath29 becomes meaningless .
the deceleration parameter @xmath108 is a very important factor for understanding cosmic evolution .
particularly , after the emergence of the idea of an accelerating universe , the role of this parameter has become even more important .
this is because some recent works @xcite have demonstrated that the present acceleration is a phenomenon of the recent past and was preceded by a decelerating phase .
this cosmological picture supports the presently favored @xmath0-cdm universe .
so , during cosmic evolution the deceleration parameter @xmath108 must have undergone a change of sign from a positive to a negative value .
so , calculation of @xmath108 from our model is essential for detecting a possible signature flip of @xmath108 .
now , using equations ( 9 ) and ( 35 ) we can obtain @xmath109 = -[1 - 1.5\omega_m ( t/\tau)^n].\end{aligned}\ ] ] equation ( 42 ) shows that in the four dimensional case the expression for @xmath108 contains a time factor and hence with a suitable choice of @xmath47 , @xmath108 can show its change of sign ( figs 3 and 4 ) .
similarly , in the five dimensional case , using equations ( 26 ) and ( 38 ) the expression for @xmath108 reduces to @xmath110.\end{aligned}\ ] ] equation ( 42 ) shows that @xmath108 is time - dependent for the five dimensional case also and hence may change its sign during cosmic evolution ( figs 5 and 6 ) .
so , for @xmath93 model , @xmath108 shows the possibility of a change of sign in both four and five dimensions .
since @xmath93 and @xmath61 models are equivalent , it is clear that @xmath61 model will also show change of sign of @xmath108 .
again , @xmath108 can be written in terms of @xmath3 as @xmath111,\end{aligned}\ ] ] in the four dimensional case and @xmath112,\end{aligned}\ ] ] in the five dimensional case .
now , considering various experimental and theoretical results , one can safely assume that the present value of the cosmic matter - energy density ( @xmath113)as @xmath114 @xcite ( though seems rather a bit high value as given wmap5 results ) .
so , for the present dust - filled universe ( @xmath42 ) , in the four dimensional case the value of @xmath108 comes out as @xmath115 which is in nice agreement with the present accepted range of @xmath108 for an accelerating universe @xcite .
for the five dimensional case , @xmath116 for the same set of values of @xmath113 and @xmath117 and hence that also supports the idea of an accelerating universe .
although some specific constant values for the equation of state parameter @xmath3 are used for different phases of the cosmic evolution , that technique suffers from some kind of discreteness problem .
the transition from one cosmic era to another can be understood better with a variable @xmath3 .
so , in order to have a continuous picture of cosmic evolution , @xmath3 should be regarded as time - dependent . in the previous sections , by selecting a simple power law expression of @xmath96 for the equation of state parameter @xmath3 , equivalence of the models @xmath118 and @xmath61 has been established in both four and five dimensions . in the four dimensional case ,
the parameters of the two equivalent models bear the same relationship with cosmic matter and vacuum energy densities as was found by ray et al .
@xcite for constant @xmath3 .
but in five dimensions , the relationship of the same two parameters with @xmath50 and @xmath51 differ from that of pradhan et al .
it has also been possible to show that the sought - for signature flipping of the deceleration parameter @xmath108 can be obtained by suitable choice of @xmath47 in both four and five dimensions .
this feature of the work satisfies an important criterion of @xmath0-cdm cosmology .
it should be mentioned that the model @xmath119 which was shown to be equivalent to the present two @xmath0 models by ray et al .
@xcite ( in four dimensions ) as well as by pradhan et al .
@xcite ( in five dimension ) is found not to be equivalent to the two models considered here when @xmath3 is time - dependent . if we now look at the plots , our observations regarding these are as follows : + ( 1 ) the plots ( figs 1 and 2 ) for @xmath3 show that in the early stage , it was positive i.e. in the early stages , the universe was matter - dominated .
the plots for @xmath108 also support this ( i.e. decelerating phase of the universe ) . at late times
it is evolving with negative values ( i.e. at the present time ) .
the earlier real matter later on converted to the dark energy - dominated phase of the universe .
it , finally , ends up at @xmath120 , representing a cosmological constant dominated universe ; + ( 2 ) the plots ( figs 3 - 6 ) for @xmath108 show that in the early stages the universe was in a decelerating phase and at late times , it is in an accelerating one .
finally , the models under consideration are shown to be equivalent in both four and five dimensions and extension to the fifth dimension do not present us any basically new physical feature than the four dimensional case .
so , a natural question can be raised : is there any justification for extension to five dimensions under the framework of general relativity ? in this connection
we would like to mention the works of kapner et al .
@xcite and rahaman et al .
@xcite where it is shown that extension of four dimensional general relativity to higher dimensions becomes mostly ineffective .
one of the authors ( sr ) is thankful to the authority of inter - university centre for astronomy and astrophysics , pune , india for providing visiting associateship programme under which a part of this work was carried out .
j. dunlop _
_ , _ nature _ * 381 * , 581 ( 1996 ) .
h. spinard _
et al . _ , _ astrophys . j. _ * 484 * , 581 ( 1997 ) a. g. riess _
_ , _ astron . j. _ * 116 * , 1009 ( 1998 ) . s. j. perlmutter _
_ , _ astrophys
. j. _ * 517 * , 565 ( 1999 ) .
_ , _ astrophys
. j. _ * 572 * , 1 ( 2002 ) .
m. bartelmann _
_ , _ new astron .
rev . _ * 49 * , 199 ( 2005 ) .
s. v. chevron and v. m. zhuravlev , _ zh . eksp .
fiz . _ * 118 * , 259 ( 2000 ) . v. m. zhuravlev , _ zh . eksp .
* 120 * , 1042 ( 2001 ) .
p. j. e. peebles and b. ratra , _ rev .
phys . _ * 75 * , 559 ( 2003 ) .
r. jimenez , _ new astron .
_ * 47 * 761 ( 2003 ) .
et al . _ , _ phy .
d _ * 72 * , 043528 ( 2005 ) .
b. ratra and p. j. e. peebles , _ phy .
d _ * 37 * , 3406 ( 1988 ) .
m. s. turner and m. white , _ phy .
* 56 * , r4439 ( 1997 ) .
_ , _ phy .
* 80 * , 1582 ( 1998 ) .
a. r. liddle and r. j. scherrer , _ phy .
d _ * 59 * , 023509 ( 1999 ) .
p. j. steinhardt _
_ , _ phy .
d _ * 59 * , 123504 ( 1999 ) .
b. bhui , b. c. bhui and f. rahaman , _ astrophys . space.sc._
* 299 * , 61 ( 2005 ) .
f. rahaman , b. bhui and b. c. bhui , _ astrophys . space.sc._ * 301 * , 47 ( 2006 ) .
f. rahaman , m. kalam and s. chakraborty , _ acta phys .
b _ * 40 * , 25 ( 2009 ) .
k. freese _
_ , _ nucl .
b _ * 287 * , 797 ( 1987 ) .
m. zer and m. o. taha , _ nucl .
b _ * 287 * , 776 ( 1987 ) . w. chen and y. s. yu , _ phy . rev
* 41 * , 695 ( 1990 ) .
j. c. carvalho _
d _ * 46 * , 2404 ( 1992 ) .
j. a. lima and j. m. f. maia , _ phy .
d _ * 49 * , 5597 ( 1994 ) .
s. ray , u. mukhopadhyay and x. -h .
meng , _ grav .
* 13 * , 142 ( 2007 ) .
a. pradhan _
. j. theor .
phys . _ * 47 * 1751 ( 2008 ) .
u. mukhopadhyay , s. ray and s. b. duttachowdhury , _ int . j. mod
d. _ * 17 * , 301 ( 2008 ) .
u. mukhopadhyay and s. ray , _ n. b. u. math .
j. _ * ii * , 51 ( 2009 ) .
i. dymnikova and m. khlopov , _ grav .
suppl . _ * 4 * , 50 ( 1998 ) .
i. dymnikova and m. khlopov , _ mod .
* 15 * , 2305 ( 2000 ) .
i. dymnikova and m. khlopov , _ eur .
j. c _ * 20 * , 139 ( 2001 ) .
d. huterer and m. s. turner , _ phys .
d _ * 64 * , 123527 ( 2001 ) .
j. weller and a. albrecht , _ phys .
d _ * 65 * , 103512 ( 2002 ) .
d. polarski and m.chavellier , _ int .
d _ * 10 * , 213 ( 2001 ) .
e. v. linder , _ phy .
* 90 * , 91301 ( 2003 ) .
u. mukhopadhyay , p. p.
ghosh , m. khlopov and s. ray , gr - qc/07110686 .
r. a. knop _
_ , _ astrophys
. j. _ * 598 * , 102 ( 2003 ) m. tegmark _
_ , _ astrophys
. j. _ * 606 * , 702 ( 2004 ) v. sahni _ et al .
_ , _ pramana _ * 53 * , 937 ( 1999 ) a. g. riess , _ astrophys . j. _ * 560 * , 49 ( 2001 ) .
l. amendola , _ mon . not .
r. astron .
* 342 * , 221 ( 2003 ) .
t. padmanabhan and t. roychowdhury , _ mon . not .
r. astron .
soc . _ * 344 * , 823 ( 2003 ) .
d. j. kapner _
_ , _ phys .
lett . _ * 98 * , 021101 ( 2007 ) .
f. rahaman _
_ , _ int j theor phys _ * 48 * , 3124 ( 2009 ) . | two phenomenological variable @xmath0 models , viz . @xmath1 and @xmath2 have been studied under the assumption that the equation of state parameter @xmath3 is a function of time .
the selected @xmath0 models are found to be equivalent both in four and five dimensions .
the possibility of signature flip of the deceleration parameter is also shown . |
to select weak compact steep spectrum ( css ) sources from the first catalogue ( white et al .
1997 ) we took the following steps : 1 . from the source list based on green bank ( gb ) surveys at 21 and 6 cm ( white & becker 1992 ) we selected those lying within the limits of the very first release of the first survey
it covered the area ra ( j2000 ) @xmath2 , dec ( j2000 ) @xmath3 at that time having steep spectra ( @xmath4 , @xmath5 ) and being stronger than 150mjy at 6 cm .
this flux density limit was chosen in order to produce a sample of a manageable size .
the above declination limits indicate that the overlap between our sample and the b3vla survey - based sample ( fanti et al .
2001 ; dallacasa et al .
2002a , 2002b ) is not large .
thanks to a dramatic difference in the resolution , the majority of sources appearing as single in the gb survey turn out to be double or multiple on first maps .
we rejected all such cases , i.e. we selected only those sources that are single entities in the first catalogue , i.e. more compact than the first beam ( 54 ) and surrounded by an empty field .
we adopted 1arcmin as a radius of that field .
additionally we referred to nvss ( condon et al .
1998 ) to check that , indeed , we picked up compact sources .
such a procedure allows us to make sure that we deal with isolated objects and not parts of larger objects .
we again checked whether our targets fulfill the spectrum steepness criterion : instead of gb survey flux densities at 21 cm we used more accurate values from first .
we rejected candidates with flat spectra ( @xmath6 ) .
we rejected the gps sources .
for this purpose we identified our preliminary candidates with objects listed in the 365mhz texas catalogue ( douglas et al .
we passed only those objects which have non - inverted spectra between 365 and 1400mhz .
in other words the turnover frequencies of our sources lie below 365mhz .
finally we selected 60 candidates for css sources .
the initial survey was performed with merlin at 5ghz .
each of our targets was observed six times in 10 minute scans spread evenly over a 12 hour track .
phase calibrators from the merlin calibrator list were observed twice per target scan for 12 minutes .
six objects having sizes similar to classical css sources , yet being less luminous , were followed up with merlin at 1.6ghz see kunert et al .
( 2002 ) .
twelve targets have either been misdetected or unresolved .
the latter group of nine objects have just been followed up with the global vlbi .
if it turns out that there are compact symmetric objects ( csos ) among them , this will be a very interesting result .
normally csos are gps sources but our sources are not gps by definition so they could be interpreted as ` dying csos ' .
we comment on this group in marecki , spencer , & kunert ( 2003 , hereafter paper ii ) .
the whole set of merlin images is presented by a. marecki et al .
( in preparation ) .
twenty objects have been selected for higher resolution observations . each of them was observed with the vlba at 18 cm and the evn at 6 cm .
comprehensive results of those vlbi campaigns will be presented elsewhere . here
we draw some general conclusions based on the whole observational material and illustrate them with a small subset of images .
1 . since our research was carried out at two frequencies and with three different interferometric arrays it enabled us to see our targets in several frequency / resolution combinations . as a result
we were able to locate cores in many objects and it appears that one - sided structures are _ not _ under - represented in our sample compared to other samples . in other words , it does not seem that css sources in deep samples like ours are predominantly symmetric objects . a single network , single frequency map can be very easily misinterpreted as an image of a symmetric object .
the quasar 1148 + 387 is a good example of such a deceiving object : it appears as a compact ( 02 separation of the components ) and very symmetric double on the 6 cm merlin map but it is actually an asymmetric object according to our dual frequency vlbi surveys the southern component has a flat spectrum so it must be a core ( figure [ 1148 + 387 ] ) .
1343 + 386 ( also a quasar ) , which was observed in parallel at 18 cm with the vlba by dallacasa et al .
( 2002a ) , is a similar case ( figure [ 1343 + 386 ] ) . according to our measurements , the component dallacasa et al .
( 2002a ) denoted as n1 is the strongest one and features a flat spectrum .
2 . as in the case of other samples of css sources those which happen to be medium - sized symmetric objects ( msos )
are frii - like 1709 + 303 is a classical specimen and it is very hard to find a mini - fri ; in our sample only 1601 + 382 resembles an fri ( figure [ frifrii ] ) .
we further comment on this fact in paper ii .
3 . a classical frii object ( fanaroff & riley 1974 ) is very clearly edge brightened thanks to the hot spots located at the source s extremities , and css sources , regardless of whether they are strong or weak
, very often follow this pattern see e.g. 1709 + 303 . however , we found a few sources which , although in principle might be labelled frii just because they are double - lobed but not fris , are _ not _ featured by well - defined hot spots
. these are 1009 + 408 , 1236 + 327 , 1542 + 323 , 1656 + 391 , and 1717 + 315 ( figures [ dying_arcsec ] and [ dying_subarcsec ] ) .
their lobes are diffuse and often ` amoeba - shaped ' 1542 + 323 ( figure [ dying_arcsec ] ) serves as the best example here .
we think these are ` dying ' sources , i.e. agn where the central engine activity has stopped and the lobes are in the so - called ` coasting phase ' .
this subclass may be interpreted as a dead end of the evolution .
we discuss this phenomenon in detail in paper ii .
condon , j.j . , cotton , w.d . ,
greisen , e.w .
, yin , q.f . , perley , r.a . ,
taylor , g.b . , & broderick , j.j .
1998 , aj , 115 , 1693 | we report early results of an observational campaign targeted on a sample of compact steep spectrum sources selected from the first survey which are significantly weaker than those investigated before .
the selection criteria and procedure are given in detail .
we present here an assortment of merlin and vlbi observations and make some general comments based on the morphologies of the sources presented .
= 2em = 15.5 cm = 22.6 cm = -1.0 cm = 0.5 cm = 0.5 cm @xmath0 toru centre for astronomy , nicholas copernicus university , toru , poland + @xmath1 jodrell bank observatory , university of manchester , uk + * keywords : * surveys radio continuum : general quasars : general galaxies : active |
understanding the manner in which protostellar envelopes collapse to stellar and disk systems is an essential problem in star formation . while progress has been made in testing the predictions of the standard theory of protostellar collapse using overall spectral energy distributions ( seds ) and source statistics of protostellar ( class i ) sources in the taurus molecular complex ( e.g. , adams , lada , & shu 1987 ; butner et al .
1991 ; kenyon , calvet , & hartmann 1993 @xmath2 kch ; kenyon et al . 1990 , 1994 ) , many details remain to be understood . in particular , the roles of outflows and non - spherical initial cloud structure in shaping protostellar clouds ( whitney & hartmann 1993 ; kenyon et al .
1993 ; calvet et al .
1994 ; hartmann et al . 1994 , 1996 ; li & shu 1996 , 1997 ) , and the way in which these effects might or might not combine with the angular momentum distribution of protostellar clouds to produce collapse to binary systems and disks ( cf .
terebey , shu , & cassen 1984 ; boss 1995 , and references therein ) , is far from clear .
near - infrared imaging observations can play an important role in help resolving these problems .
while near - ir images are sensitive to very small amounts of material , due to the efficiency of dust scattering and absorption , such observations can be crucial in indicating the positions and properties of the central stars , and in tracing envelope structure on small scales difficult to observe in other spectral regions . here
we report _ hubble space telescope _ near - infrared imaging observations of a taurus protostellar ( class i ) source , iras 04325 + 2402 ( beichman et al .
this is a low - luminosity source ( @xmath3 ) with other properties typical of class i taurus sources ( kch ) .
this source is multiple , and exhibits a complex and confusing morphology , illustrating the limits of sed analysis for understanding protostellar sources .
near - infrared observations of iras 04325 + 2402 were obtained with the _ hst _ nicmos camera 2 on 11/11/97 ut .
images were obtained with the f110w , f160w , f205w , and f212n filters , the latter to search for h@xmath4 emission . a five - position dither pattern with a separation of approximately 6 arcsec was used for the wide filters .
because the source is large and was not well - centered , only four of the 5 dithered images were useful in producing the final images .
dark sky images for f205w were obtained with a single chop position offset by 40 arcsec .
exposure times were 128 sec per frame in f205w and 40 sec in f160w .
the 80 sec f212n exposures provided no hint of excess h@xmath4 emission , and the 40 sec f110w exposures showed only very faint nebulosity , so we do not consider them further .
the data were reduced using the standard `` calnica '' routine in the stsdas packages with updated calibration files and other tasks within iraf .
offsets for the construction of the final image were derived by cross - correlating image features .
final images were created in two ways .
first , we translated the images using fractional pixel shifts using the iraf `` imshift '' task , and constructed a median combined image using the `` imcombine '' task .
second , images were combined using the `` drizzle '' task in iraf / stsdas ( fruchter & hook 1999 ) .
`` drizzle '' employs a variable - pixel linear reconstruction algorithm , for the re - sampling of undersampled , dithered data , while correcting for geometric distortions and preserving photometry . in this method ,
pixels are mapped from an input image , or group of input images , into pixels in a subsampled output image . in this application ,
the input pixels were shrunk to 90% of their original size ( * pixfrac=0.9 * ) before being drizzled onto a grid with pixels half the original pixel size ( * pixscale=0.5 * ) . weighted image masks , taking into account bad pixels and the coronographic spot , were prepared for each image and used in the drizzle .
the images resulting from these two procedures are extremely similar . for basic image display
we show the drizzled versions ; for comparison with models ( 4 ) , we use the imcombined images to maintain the original pixel scale of the observations . supporting images at j , h , k , and an h@xmath4 @xmath5 m filter
were obtained with nsfcam on the nasa irtf telescope on 26 and 27 nov 1998 .
data were obtained on two pixel scales , 0.15 and 0.3 arcsec / pixel , on successive nights .
the data were reduced with standard iraf routines .
figure 1 shows the large - scale k - band nebulosity as observed at irtf with nsfcam .
the large - scale bipolar structure is evident , consistent with typical scattered light structures produced by bipolar outflows in young stellar objects .
the northern scattered light lobe is much brighter ( and perhaps more conical ) than the southern lobe , and exhibits two bright structures toward its apex , in agreement with the k image presented by hodapp ( 1994 ) .
an anomalous bright `` knot '' , which we refer to as object c , is apparent near the western edge of the northern lobe ( as also shown in the image of hodapp ) , and is crossed by a curved bright streak proceeding from se to nw which apparently passes through the knot .
the streaks exhibit a peculiar `` corkscrew '' appearance ; there may be some corresponding structure , though not symmetrically placed , in the southern lobe .
a comparison with the narrow h@xmath4 filter images shows that there is no measurable excess relative to either the broad - band k filter or a neighboring continuum filter , consistent with the f212n images .
thus , the spectrum of c is not dominated by h@xmath4 emission , and must be mostly continuum . the nicmos f205w composite drizzled image is shown in figure 2 . to preserve resolution
the image has not been rotated from the observed position angle of 38@xmath6 for the y axis .
the bright central region of the previous image is now resolved into a bright , extended central object , which we call a / b .
immediately s of a / b there is a bright ring or arc ; to the n there is a peculiar @xmath7-shaped scattered light nebulosity .
object a / b is not a point source ; as shown in the expanded view in figure 3a , there is a double structure , possibly indicating two sources , or possibly indicating a dark absorption lane running roughly ew across the object .
figure 3b shows an enlargement of the f205w image near object c. the object is clearly resolved into a double - lobed emission separated by a dark lane .
this structure is present in the three individual dithered images in which c appears .
this image structure is reminiscent of the hst wfpc2 images of hh30 ( burrows et al .
1996 ) and of the faint companion to hk tau ( stapelfeldt et al . 1998 ) , which have been interpreted as scattered light from the upper surfaces of edge - on , flared disks around t tauri stars , possibly augmented by a dusty , relatively tenuous , envelope ( wood et al . 1998 ) .
figure 4 is a false - color rendition of iras 04325 + 2402 made by combining the f205w ( red ) and the f160w ( green ) images .
the w portion of the knot is relatively brighter in f205w than f160w , as indicated by the color scheme , suggesting greater extinction as would be expected in a disk / envelope model ( 4 ) . in other respects
the f160w image closely parallels the f205w image structure , but with substantially lower signal - to - noise .
table 1 gives positions of components a / b and c. the position given for a / b is actually the position of a , the brightest portion ( s component ) of the extended object ( figure 3a ) ; component b is approximatley 0.074 arcsec east and 0.23 arcsec north of a. determination of fluxes and conversion to magnitude systems is complicated by the difficulty of subtracting spatially - varying background , by the very red colors of these objects , and by the non - standard nature of the nicmos filters . for the nicmos data we used aperture corrections derived from observations of a very red standard star ( oph s-1 ) kindly provided by marcia rieke ; these aperture corrections agreed well with results from the tiny tim program ( krist & hook 1997 ) .
the fluxes reported in table 1 use the standard calibration without any attempt to correct for color . usually color terms are ignored in determining fluxes and magnitudes , as we do here , but we note that color corrections could be significant for such red objects . the corresponding magnitudes reported in table 2 adopt the vega zero point values . at irtf , we measured a k magnitude ( in the ukirt system ) of approximately 11.23 @xmath8 in a 7.2 arcsec diameter aperture centered on the main source ( table 2 ) .
this is consistent with the k = 11.27 reported by kenyon et al .
( 1993 ) in an 8 arcsec diameter aperture ; however , we measured an ( h - k ) = 1.79 ( ukirt ) instead of the ( h - k ) = 2.05 reported by kgwh .
our color results must be viewed with some caution , however , because there is a large color term in the transformation of ( h - k ) between the irtf system and the ukirt or cit systems ( ( h - k)[irtf ] @xmath9 0.8 ( h - k ) [ ukirt , cit ] ) .
this color transformation has been derived for stars which are much bluer than the red systems studied here , and so may not be appropriate to apply to iras 04325 + 2402 .
the brightness of object c is uncertain for reasons of the difficulty in background subtraction .
the f205w magnitude reported in table 2 is considerably fainter than the k magnitude from irtf , but the difference can probably be understood due to the difference in central wavelengths of the two filters , and the difficulty of subtracting out the nebulous background at low spatial resolution .
( we used a larger aperture for the irtf measurement because of seeing limitations . )
the image of object c exhibits similarities to other _
hst _ images of young stellar objects surrounded by circumstellar material in the form of disks and/or envelopes , such as hh30 ( e.g. , burrows et al .
1996 ) . to investigate this possibility
further , we constructed a series of models of disks illuminated by the central star , with and without infalling envelopes .
given the limited observational information available ( for example , the luminosity and mass of the central star are unknown ) , and the number of parameters needed for a disk plus envelope system , no unique model can be constructed . instead we adopt properties and parameters typical of models of class i sources and t tauri disks to investigate the plausibility of our overall interpretation . the envelope is taken to have the density structure of the standard protostellar infall model of terebey , shu , and cassen ( 1984 ) .
the radiative equilibrium temperature of the envelope is calculated as in kch and calvet etal .
the parameters that describe the envelope are @xmath10 , the density the envelope would have at 1 au if it were purely infalling radially ( i.e. , if the material had no angular momentum ) , and the centrifugal radius @xmath11 .
the density parameter is proportional to the mass infall rate @xmath12 , @xmath13 .
the disk is assumed to be physically thin and vertically isothermal , with a radial temperature distribution @xmath14 , where r is the cylindrical radius , and surface density distribution @xmath15 , where @xmath16 is the stellar radius , as suggested by the detailed models of dalessio et al .
( 1998 , 1999 ) .
the disk has a fixed outer radius given by @xmath11 , and its density distribution along the z axis at fixed radius is determined from the equation of hydrostatic equilibrium . to calculate an image of the system for a given inclination to the line of sight , we solve the transfer equation for the specific intensity along rays that cross the system at fixed projected coordinates on the plane of the sky .
the emissivity at each point along the ray is given by the sum of two terms ( 1 ) direct stellar light , attenuated by the intervening material between the star and the point and scattered towards the observer with a henyey and greenstein phase function ( @xmath17 ) ( cf .
calvet et al . 1992 ) , and ( 2 ) isotropic scattering towards the observer of the diffuse radiation field in the envelope .
this diffuse field is obtained directly from the solution of the radial radiative transfer equations in the envelope ( cf .
calvet et al .
the first term corresponds to the single scattering approximation , which describes well the behavior of the photons scattered by the disk ( burrows et al .
the second term represents multiple scattering in the envelope , in the spherically symmetric approximation .
this approximation becomes more accurate at large distances , where the envelope becomes spherical .
draine and lee ( 1984 ) dust properties have been assumed .
the theoretical specific intensity map was binned into pixels of 0.038 `` x 0.038 '' , i.e. , half the original image pixel scale and convolved with the _ hst _ psf function derived from tiny tim ( krist & hook 1997 ) .
after convolution , the images were rebinned back to the original pixel scale to compare with observations . our best envelope @xmath18 disk model for object c
is shown in figure 5 compared with the f205w image .
the disk has a radius of 30 au , and it is immersed in a low density envelope characterized by @xmath19 and @xmath11 = 30 au .
the system is viewed at an inclination angle to the line of sight of 83@xmath6 .
the theoretical image contours agree reasonably well with the observed image contours , though we can not reproduce the observed non - axisymmetric features with an axisymmetric model .
our investigation indicates that neither pure disk models nor pure envelope models of the type considered here can satisfactorily reproduce the observed image .
models which include disks alone ( left panel in figure 6 ) do not have emission extending sufficiently far in the perpendicular direction to the disk plane ( i.e. , the disk image is too `` flat '' ) .
conversely , pure tsc infall models do not exhibit sufficient indication of a dark lane ( middle panel of figure 6 ) .
the situation is reminiscent of that of models for hh30 by wood et al .
( 1998 ) , in which the brightest structures were produced by a disk but the faint , highly - extended scattered light structures were attributed to an envelope .
the model calculations for f160w are in approximate agreement with the observations .
we do not attempt a detailed comparison because of the substantial noise in this image .
the model extinction is ( slightly ) higher toward the southeastern component , in qualitative agreement with the ( modest ) color difference observed . assuming an intrinsic color of the central object of ( h - k)@xmath20 ( see 4.2 ) , the model predicts ( h - k)@xmath21 .
given the uncertainties in the data and the unknown nature of the colors of the central source , this agreement is sufficiently close to demonstrate the general plausibility of the model .
we conclude that a model which includes both a disk and an envelope can qualitatively account for the image of this object .
this model also may explain the scattered light `` streaks '' or structures associated with c. we suggest that these streaks might delineate the edges of an outflow cavity produced by c ( see figure 7 ) .
if we assume a stellar mass of @xmath22 for c ( 4.2 ) , the implied infall rate for the envelope of our model is @xmath23 .
this is below the average infall rate of @xmath24 estimated for typical taurus class i sources , and lower than the overall @xmath25 derived for the main source in iras 04325 + 2402 by kch ( though substantially larger than the @xmath26 adopted by wood et al . [
1998 ] for their hh30 model ) . from the observed source morphology it appears plausible that c is encountering a lower - density environment than typical of the regions near a / b , possibly as a result of orbital motion away from the main central mass
the nature of the central star in c is difficult to determine given the complications involved , but we can make some estimates .
for this purpose we use the ground - based irtf measurements because they are ( approximately ) on a standard magnitude system , though the lower spatial resolution makes extraction of the source with respect to the nebular background more difficult .
the very red ( h - k)@xmath27 color of c indicates that the central star must be heavily reddened .
the intrinsic colors of ( m - type ) classical t tauri stars in taurus ( young stars with inner circumstellar disks ) range from about ( h - k)@xmath28 to @xmath29 ( meyer et al .
1997 ) ( for objects with no contribution from a dusty envelope ) . adopting an intrinsic color ( h - k)@xmath30
, the excess is e(h - k)@xmath31 . using a(k ) @xmath32e(h - k ) from cohen
( 1981 ) , and ignoring the difference in colors between cit and ukirt systems , we find a(k)@xmath33 , so k@xmath34 . using the results of wilking , greene , & meyer ( 1999 ; their figure 1 ) for evolutionary tracks , and adopting an age of @xmath35 yr as typical of class i sources ( kenyon et al . 1990 , 1994 ) , the result is a luminosity of @xmath1 and a mass of @xmath36 , well into the brown dwarf range .
however , this estimate is almost certainly a lower limit to the stellar luminosity ( and therefore to the stellar mass ) because the object is undoubtedly seen entirely in scattered light , and thus the observed colors underestimate the extinction directly toward the central star . for the model shown in figure 6 ,
the observed flux at k is approximately 4% of the true stellar flux .
using this factor to correct the observed k magnitude would result in @xmath37 and a mass of about @xmath38 from the wilking et al .
( 1999 ) estimates .
small changes in envelope parameters easily could increase the true k brightness by substantial factors . at the other extreme , if we assume that c supplies _ all _ of the heating responsible for the iras long - wavelength emission ( cf . kch )
, then the luminosity indicates a central star of @xmath39 from the evolutionary tracks of dantona & mazzitelli ( 1997 ) .
given that the main source a / b seems to be responsible for most of the scattered light in the main nebular lobe , this seems unlikely .
thus the central star in c is probably either a very low mass star or a brown dwarf ; the f205w observations may provide the first image of a protoplanetary disk around a brown dwarf .
infrared spectra should be obtained to better constrain the nature of c. a / b lies near the apex of the main bipolar reflection nebula ( figure 1 ) , and is probably the main luminosity source in the system .
this geometrical relation indicates that a / b is probably the source of bipolar ejection which produces the large - scale reflection nebula structure .
the alignment of these cavities are very different than the assumed alignment of object c ( see figure 7 ) .
thus , it appears that even relatively close , and probably coeval , multiple protostellar systems may have significantly different outflow / disk orientations . the large extinction toward a / b ( as indicated by the colors , and by the extreme faintness of the source at f110w and j ) , indicates that a / b is also a center of infall for the extended dusty envelope , as envisaged by the models of kch and kenyon et al .
( 1993b ) .
since a / b is probably more luminous and thus possibly more massive than c , it would not be surprising if most of the infall is directed toward it rather than toward c. the presence of multiple gravitating centers in this system may account for some of the peculiar nebular structure . whether a / b is itself a multiple stellar system , or simply a complex scattered light structure around a single star , is unclear . while we do not offer a model for the peculiar scattered light structures in the vicinity of this source , we note that they appear to lie along portions of the outflow cones of the main source .
the scattered light images may indicate complicated interactions of a structured infall region with outflow ( figure 7 ) , although the precise way this might happen is far from clear .
the lower ( southern ) lower scattered - light arc is reminiscent of ground - based images of circumbinary dust rings around t tauri stars in taurus , as for example is observed in gg tau ( roddier et al .
1994 ; see also dutrey , guilloteau , & simon 1994 ) and uy aur ( close et al . 1998 ) .
the inner ring radii in these two objects are around 200 au and 500 au , respectively ; for object a , the diameter of the southern ring is about 1.3 arcsec @xmath9 180 au at 140 pc . in this picture ,
the ring could be produced by the tidal action of an orbiting binary or multiple protostellar system .
a possible problem with this picture is that a / b does not seem to lie inside of the ring , but somewhat above it .
object a / b is probably the main gravitating center toward which most of the dusty envelope is presumably collapsing .
the model of kch , which was adjusted to fit the sparsely - observed long - wavelength sed of the object , suggested an infall rate of @xmath40 .
even with a large centrifugal radius , @xmath41 au , to reduce extinction toward the main source , the model did not account for the observed near - infrared emission . the attempt of kenyon et al .
( 1993 ) to explain the near - infrared fluxes and large - scale image including outflow cavities in the collapse model led them to estimate @xmath42 and @xmath43 au .
given the complex scattered light structure seen in our images on scales of 100 au and less , it is clear that any estimate of the centrifugal radii of collapsing envelopes from near - infrared fluxes alone is very uncertain .
the peculiar `` corkscrew '' structure in the northern lobe of the main nebula may be the result of interacting flows of c and a. in this connection we note that the observed ( monopolar ) co outflow ( heyer et al 1987 ) in the region is oriented to the northwest of iras 04325@xmath182402 . with a position angle of about @xmath44 , the outflow is closer to the orientation expected for an outflow from c , and completely inconsistent with the nearly north - south orientation of the main scattered light nebula ( figure 1 ) .
however , the co outflow extends over such a large scale that its connection with iras 04325@xmath182402 is unclear ; moreover , the observed co motions are redshifted , whereas one might expect to observe a blueshifted lobe from object c ( cf .
figure 7 ) . in many respects
iras 04325@xmath182402 resembles another class i taurus protostar with similar bolometric luminosity and near - ir colors , iras 04361@xmath182547 ( kch ) , which has recently been imaged with nicmos by terebey et al .
iras 04361@xmath182547 also shows a complex central structure , scattered light `` streaks '' , and a faint companion .
terebey et al .
interpret their complex central structure to mean that the main source is a binary .
as mentioned above , this may apply to object a as well , though we are hesitant to make a firm interpretation since the possibility also exists that these are simply scattered light structures intersected by absorbing dust lanes . the point - source companion found by terebey et al . in 04361@xmath182547 also lies in a scattered light structure , at a similar distance from the main region ( 1400 au vs. the 1150 au projected distance between objects a / b and c ) .
this companion has roughly similar colors to object c but is about a factor of 10 fainter .
terebey et al .
estimated the companion mass as 2 - 5@xmath45 , and suggested that its position at the end of a scattered light filament might be caused by ejection from the central region over the last 1000 yr .
our object c differs in that it is clearly too bright to be a @xmath46 planet and has its own solar - system sized disk .
another difference is that the streaks or filaments do not point back to a / b , unlike the situation in 04361@xmath182547 .
if object c s disk had been a factor of two or so smaller , it would have appeared to be a point source , raising the possibility that the terebey et al .
object might also appear anomalously faint for its color as a result of disk / envelope extinction .
the class i taurus source iras 04325@xmath182402 has been shown to be a multiple system , with at least one additional protostar lying a projected distance of @xmath47 au from the central source , and that central source may itself be multiple .
the nicmos images of the companion are consistent with the typical model of a low - mass protostar , in which infalling material from a rotating envelope lands upon , and builds up , a disk .
the dust envelope structure around the main source appears more complex , but suggests an independent center of collapse .
the orientation of the main bipolar cavities seen in scattered light , probably driven by outflow from the main source , lies at a substantial angle to the presumed disk plane of the companion object .
this suggests that the orientation of disk systems can be quite different even in relatively close multiple protostellar systems .
we wish to thank bill vacca for his very helpful support at irtf , karla peterson and louis bergeron at stsci , and marcia rieke for providing standards .
this work has been supported in part by nasa through grant number go-07413.01 - 96a from the space telescope science institute .
l c c c c r + + object & @xmath48(j2000 ) & @xmath49(j2000 ) & f110w ( @xmath50jy ) & f160w ( mjy ) & f205w ( mjy ) + a / b & 4 35 35.37 & 24 08 19.5 & 93 @xmath51 & 3.1 @xmath52 & 14.4 @xmath53 + c & 4 35 35.29 & 24 08 27.6 & @xmath54 & 0.14 @xmath55 & 0.47
@xmath8 + l c c c c r + + object & k & h - k & m(f110w ) & m(f160w ) & m(f205w ) + a / b & 11.23 @xmath8 & 1.79 @xmath8 & 18.3 @xmath56 & 13.89 @xmath57 & 11.80 @xmath8 + c & 14.61 @xmath57 & 1.95 @xmath57 & @xmath58 & 17.27 @xmath56 & 15.52 @xmath59 + adams , f.c . ,
lada , c.j . , & shu , f.h .
1987 , apj , 321 , 788 beichman , c.a . , myers , p.c . ,
emerson , j.p . ,
harris , s. , mathieu , r. , benson , p.j .
, & jennings , r.e .
1986 , apj , 307 , 337 boss , a.p .
1995 , in circumstellar disks , outflows , and star formation , revmexaa ( serie de conferencias ) 1 , 165 burrows , c.j . ,
stapelfeldt , k.r . ,
watson , a.m. , krist , j.e . ,
ballester g.e . ,
clarke , j.t .
crisp , d. , gallagher iii , j.s . , griffiths , r.e .
, hester , j.j .
, hoessel , j.g . , holtzman , j.a .
, mould , j.r . , scowen , p.a . ,
trauger , j.t . , & westphal , j.a .
1996 , apj , 473 , 437 butner , h.m . ,
evans , n.j .
ii , lester , d.f . , levreault , r.m .
, & strom , s.e . 1991 , apj , 376 , 636 calvet , n. , canto , j. , binette , l. , & raga , a. 1992 , rmxaa , 24 , 81 calvet , n. , hartmann , l. , kenyon , s.j . , & whitney , b.a .
1994 , apj , 434 , 330 close , l.m .
, dutrey , a. , roddier , f. , guilloteau , s. , roddier , c. , duvert , g. , northcott , m. , mnard , f. , graves , j.e . , & potter , d. 1998 , apj , 499 , 883 cohen , j.g . ,
frogel , j.a . , persson , s.e .
, & elias , j.h .
1981 , apj , 249 , 481 dalessio , p. , canto , j. , calvet , n. , & lizano , s. 1998 , apj , 500 , 411 dalessio , p. , calvet , n. , hartmann , l. , lizano , s. , & canto , j.1999 , apj , ( submitted ) dantona , f. , & mazzitelli , i. 1997 , mem .
soc . astron .
italiana , 68 , 4 draine , b. t. , & lee , h. m. 1984 , apj , 285,89 dutrey , a. , guilloteau , s. , & simon , m. 1994 , a&a , 286 , 149 fruchter , a.s . & hook , r.n .
1999 , pasp , submitted hartmann , l. , boss , a. , calvet , n. , & whitney , b. 1994 , apj , 430 , l49 hartmann , l. , calvet , n. , & boss , a. 1996 , apj , 464 , 387 heyer , m.h .
, snell , r.l . , goldsmith , p.f . , &
myers , p.c .
1987 , apj , 321 , 370 hodapp , k - w . 1994 , apjs , 94 , 615 kenyon , s.j . , hartmann , l. , strom , k. m. & strom , s. e. 1990 , apj , 99 , 869 kenyon , s.j . ,
calvet , n. , & hartmann , l. 1993 , apj , 414 , 676 ( kch ) kenyon , s.j . , whitney , b. , a. , gomez , m. , & hartmann , l. 1993 , apj , 414 , 773 kenyon , s.j . ,
marzke , r. o. & hartmann , l. 1994 , apj , 108 , 251 krist , j.e . , & hook , r.n .
1997 , the tiny tim user s guide version 4.4 ( baltimore : stsci ) li , z .- y . , & shu , f.h .
1996 , apj , 472 , 211 li , z .- y . , & shu , f.h .
1997 , apj , 475 , 251 meyer , m.r . , calvet , n. , & hillenbrand , l.a .
1997 , aj , 114 , 288 roddier , c. , roddier , f. , northcott , m.j . , graves , j.e .
, & jim , k. 1996 , apj , 463 , 326 stapelfeldt , k.r . , krist , j.e . ,
mnard , f. , bouvier , j. , padgett , d.l . , &
burrows , c.j .
1998 , apj , 502 , l65 terebey , s. , shu , f.h . , & cassen , p. 1984 , apj , 286 , 529 terebey , s. , van buren , d. , padgett , d.l . ,
hancock , t. , & brundage , m. 1998 , apj , 507 , l71 whitney , b. & hartmann , l. 1993 , apj , 402 , 605 wilking , b.a .
, greene , t.p .
, & meyer , m.r . 1999 ,
aj , 117 , 469 wood , k. , kenyon , s.j . ,
whitney , b. , & turnbull , m. 1998 , apj , 497 , 404 | we report _ hubble space telescope _ near - infrared nicmos observations of a remarkable low - luminosity class i ( protostellar ) source in the taurus molecular cloud . iras 04325
+ 2402 exhibits a complex bipolar scattered light nebula .
the central continuum source is resolved and may be multiple , or may be crossed by a small dust lane .
complex arcs seen in scattered light surround the central source ; the physical nature of these structures is not clear , but they may reflect perturbations from multiple stellar sources or from time - dependent mass ejection . a second ,
resolved continuum source is found at a projected distance of approximately 1150 au from the central region , near the edge of a nebular lobe probably produced by outflow .
the images indicate that this second source is another low - luminosity young stellar object , seen nearly edge - on through a dusty disk and envelope system with disk diameter @xmath0 au .
we suggest that the scattered light `` streaks '' associated with this second source are limb - brightened outflow cavities in the dusty envelope , possibly perturbed by interaction with the outflow lobes of the main source .
the nature of the companion is uncertain , since it is observed mostly in scattered light , but is most probably a very low mass star or brown dwarf , with a minimum luminosity of @xmath1 .
our results show that protostellar sources may have multiple centers of infall and non - aligned disks and outflows , even on relatively small scales . |
there are three promising theoretical models to explain the formation of the massive stars : monolithic collapse , competitive accretion , and mergers of low mass stars ( see for a review , zinnecker & yorke 2007 ) .
however , their relative large distances ( a few parsecs ) , the clustered formation environments , and their short timescales have made extremely difficult to discard any of these models ( zapata et al .
2015 ) .
massive stars have a substantial impact on their surrounding environments , making important contributions to the chemistry enrichment of the interstellar medium ( e. g. , hernndez - hernndez et al .
thus , searching for complex molecules in massive star formation regions is a crucial building block to understand massive star formation , since these molecules can provide information on the physical conditions and evolutionary phases of massive star formation ( herbst & van dishoeck 2009 ) .
the w51 north is one of well - studied massive star - forming regions within the luminous cluster w51-irs2 ( zapata , tang & leurini 2010 ) .
the distance from w51 north to the sun is approximately 7 kpc ( imai et al .
however , a more recent and accurate estimation from xu et al .
( 2009 ) obtained 5.1@xmath3 kpc .
a large number of h@xmath1o and oh masers , bright dust emission , outflows , and infalling gas were observed toward w51 north region , indicating that this region is forming indeed massive stars ( downes et al .
1979 ; zhang , ho & ohashi 1998 ; zapata et al . 2008 , 2009
; zapata , tang & leurini 2010 ) .
absence of of centimeter emission suggests that w51 north may represent an extremely early stage of the massive star formation ( downes et al .
1979 ; zhang , ho & ohashi 1998 ) . as the submillimeter array ( sma ) has a broad bandwidth ( 4 ghz , at the time of these observations ) one can detect multiple complex molecules simultaneously , allowing to trace the physical and chemical properties of massive star formation environments close to the massive protostars . in this paper
, we present the results from the sma observations of w51 north region .
we have identified 17 species and obtained their physical parameters by the use of xclass program .
we describe the observations in 2 . in
3 we present observational results , followed by data analysis in 4 .
5 discusses individual molecules .
we summarize the results in 6 .
track - sharing observations toward w51 north and w51 main sources were carried out with the sma in 2005 august , using seven antennas in its compact array .
the phase - tracking center of w51 north was placed at r.a.= @xmath4 , decl.=@xmath5 ( j2000.0 ) .
the typical system temperature of 183 k indicates good weather during observations .
the observations covered frequencies from 216.5 to 218.5 ghz ( lower sideband ) , and 226.5 to 228.5 ghz ( upper sideband ) with a frequency resolution of 0.8125 mhz . the calibration and imaging were done in miriad .
bandpass ripples were corrected with the qso 3c454.3 and uranus .
we also corrected baseline - based bandpass errors using the qso 3c454.3 .
qsos 1741 - 038 and 1749 + 096 were used for phase calibration .
uranus was used for flux calibration , and the absolute flux scale is estimated to be accurate to within 20% .
continuum subtraction was made in uv domain .
self - calibration was performed to the continuum data , and the gain solutions from the continuum were applied to the line data .
the synthesized beam sizes of the continuum and line images are approximately [email protected]@xmath7 [email protected] ( p.a.@xmath8 ) .
1 @xmath9 rms noises of the continuum and lines are approximately 0.01 jy beam@xmath10 and 0.1 jy beam@xmath10 , respectively . 1
jy beam@xmath10 corresponds to a main beam brightness temperature of 2.7 k.
continuum image was constructed from line - free channels of lsb and usb uv data , as shown in figure 1 .
the continuum image shows a compact and strong source at the position of w51 north , with an extension to the north , which embraces irs2d source .
kjd3 and w51d2 are located east and west of w51 north , respectively ( zapata et al . 2008 ) .
two dimension gaussian fits were made to the continuum obtaining the peak intensity and total integrated flux of @xmath11 jy beam@xmath10 and @xmath12 jy , respectively . under the assumptions that an average grain radius is 0.1 @xmath13 m , grain density is 3 gr @xmath14 and a gas to dust radio is 100 , the h@xmath1 column density and mass can be calculated by the formula ( lis , carlstrom & keene 1991 ) , @xmath15 and @xmath16 where @xmath17 and @xmath18 are boltzmann constant and planck constant , respectively .
t is the dust temperature , q(@xmath19 ) is the grain emissivity at frequency @xmath19 , @xmath20 is the total integrated flux of the continuum and @xmath21 is the solid angle subtended by the source , d is the distance from the source to sun . the ch@xmath0oh is a grain surface molecule . here , this molecule peaks at the position of the continuum source associated with w51 north ( see figure 3 and discussion is section 5 ) .
probably the dust and gas are well coupled through collisions in w51 north region .
so we assumed that the dust temperature equals the rotation temperature of 138 k derived from the ch@xmath0oh lines .
we adopt @xmath22 at 1.3 mm ( @xmath23 ) ( lis & goldsmith 1990 ; lis , carlstrom & keene 1991 ) .
therefore , we obtained a source - averaged h@xmath1 column density of @xmath24 @xmath25 and an h@xmath1 mass of @xmath26 .
as the continuum peaks at w51 north , then the mass estimated from peak intensity should correspond to gas mass of w51 north source .
the h@xmath1 mass of @xmath27 610 @xmath28 61 @xmath29 is estimated from the continuum peak intensity .
a similar result , @xmath27 400@xmath29 was reported by zhang , ho & ohashi ( 1998 ) , in which they calculated the mass of w51 north based on a dust temperature of 100 k , continuum peak intensity of @xmath27 2.0 jy beam@xmath10 at @xmath30 ghz , d of 7 @xmath27 8 kpc and @xmath31 .
if we take @xmath31 , then q(@xmath19 ) @xmath32 at 1.3 mm and then the h@xmath1 mass of @xmath27 535 @xmath28 54@xmath29 is obtained .
a dust temperature larger than 100 k suggests that dust grains are not coagulated with ice mantles , and thus @xmath33 @xmath27 1.5 is a reasonable guess in w51 north ( ossenkopf & henning 1994 ) .
therefore the difference of mass between ours and @xmath27 400@xmath29 by zhang , ho & ohashi ( 1998 ) may be caused by taking different value of distance , q(@xmath19 ) and uncertain of flux calibration .
the lower sideband and upper sideband spectra extracted from the continuum peak position and related with w51 north , are shown as the black curves in figure 2 .
a few cn transitions in the usb show absorption features indicating gas infalling onto continuum core ( zapata et al .
the sio line in emission is tracing the outflow reported by zapata et al .
( 2009 ) .
we identify all the molecular transitions using the xclass program .
84 molecular transitions from 17 species and their isotopologues are identified ( figure 2 ) . in this image
the corresponding molecular names are labelled , including oxygen - bearing molecules of ch@xmath0oh , h@xmath1co , c@xmath1h@xmath2oh , ch@xmath0ocho , ch@xmath0och@xmath0 , ch@xmath0coch@xmath0 , nitrogen - bearing molecules of cn , ch@xmath0ch@xmath1cn , hc@xmath0n , ch@xmath1chcn , sulfur - bearing molecules of so@xmath1 , h@xmath1s , sis , and deuterated molecules of dcn and ch@xmath0ccd .
the quantum numbers , frequency , @xmath34 and @xmath35 of each transition are summarized in table 1 .
column ( 1 ) lists the species and column ( 2 ) lists quantum number , column ( 3 ) , ( 4 ) and ( 5 ) give frequency , s@xmath36@xmath37 and e@xmath38 .
the source size @xmath39@xmath7@xmath40 , @xmath41v and @xmath42 are given in column ( 6 ) , ( 7 ) and ( 8) , respectively .
figure 3 presents a sample of images from the oxygen- , nitrogen- , and sulfur - bearing molecules , which can provide spatial distribution of specific molecules . in general , all molecular images show a compact structure .
the peak emission of the oxygen - bearing molecules of ch@xmath0oh , ch@xmath0coch@xmath0 and ch@xmath0ocho , and nitrogen - bearing molecules of ch@xmath1chcn , ch@xmath0ch@xmath1cn and dcn are coincident with the continuum peak position .
the oxygen- and nitrogen - bearing molecules are probably well mixed in space , which imply that they may have same chemical origin ( remijan et al .
2004 ) . while the emission peak of the sulfur - bearing molecules of h@xmath1s , and so@xmath1 are located east of the continuum peak .
probably the sulfur - bearing molecules have different chemical routes and form in different environments when compared with the oxygen- and nitrogen - bearing molecules .
in addition to the compact source structure , there is an obvious gas extension to the west of the continuum traced by the dcn , h@xmath1s , and sis is observed .
[ cols= " > , < , < , > , > , > , > , > , < " , ] from table 2 , the rotation temperatures of oxygen - bearing molecules range from 110 to 140 k , while nitrogen - bearing molecules have rotation temperatures ranging from 105 to 140 k , which is consistent with that derived from the transitions of ch@xmath0cn ( zhang , ho & ohashi 1998 ) .
the similar gas temperatures and distributions between oxygen- and nitrogen - bearing molecules ( see 3.3 ) suggest that oxygen - bearing and nitrogen - bearing molecules are well mixed and originate from the same physical environments .
the scenario is inconsistent with the observations toward the massive star - forming regions of orion - kl and g19.61 - 0.23 , in which nitrogen - bearing molecules have higher temperatures than oxygen - bearing molecules and oxygen- and nitrogen - bearing molecules peak at different positions ( blake et al .
1987 ; qin et al .
the sulfur - bearing molecules have gas temperatures ranging from 130 to 180 k. the sulfur - bearing molecules have been proved to be a probe of outflows or shocks .
the outflow motions are identified in w51 north region ( zapata et al .
the higher gas temperature of the sulfur - bearing molecules in w51 north may be caused by outflow heating .
there are obvious difference in the column densities between the oxygen - bearing and nitrogen - bearing molecules ( see table 2 ) .
the column densities of oxygen - bearing molecules range from @xmath43 to @xmath44 @xmath25 . while the column densities of the nitrogen - bearing molecules range from @xmath45 to @xmath46 @xmath25 .
the fractional abundance of the observed molecules relative to h@xmath47 are estimated by use of the expression , @xmath48 , and summarized in column ( 4 ) of table 2 , where @xmath49 is column density of specific molecule and @xmath50 is the column density of h@xmath1 .
the oxygen - bearing molecules have relatively higher abundances at a range of @xmath51 to @xmath52 , while the nitrogen - bearing molecules have abundances ranging from @xmath53 to @xmath54 .
previous observations of massive star - forming regions also showed that the oxygen - bearing molecules have higher abundances but lower temperatures than the nitrogen - bearing molecules ( e. g. , blake et al .
1987 ; wyrowski et al .
1999 ; qin et al .
the chemical model proposed by rodgers & charnley ( 2001 ) suggested that hot - core composition ( o - rich or n - rich ) depends on core temperature and the ammonia / methanol ratio of the evaporated ices , and the timescale of the core formation , in which o - rich cores form first , then become n - rich as temperature rises and evolution . in our case ,
the oxygen- and nitrogen - bearing molecules peak at same position and have similar temperature , which may be caused by the fact that w51 north is at an early evolutionary stage and the hot core is not very developed , as suggested by dynamical analysis ( zhang , ho & ohashi 1998 ; zapata et al .
2008 , 2009 ; zapata , tang & leurini 2010 ) .
the ch@xmath0oh , hc@xmath0n and their isotopologues @xmath55ch@xmath0oh
, hcc@xmath55cn are detected in w51 north .
@xmath56c/@xmath55c ratios of 21.0 and 23.3 are estimated from ch@xmath0oh/@xmath55ch@xmath0oh and hc@xmath0n / hcc@xmath55cn , respectively .
based on the relationship of isotope ratios and distance from the galactic center by willson & rood ( 1994 ) , the @xmath56c/@xmath55c ratio of 22 is obtained if one takes the distance of 1.5 kpc to the galactic center .
the consistent @xmath56c/@xmath55c ratio from the two methods suggest that the parameters derived from the xclass modeling are reasonable .
methanol ( ch@xmath0oh ) is a key molecule in the chemical networks linked with the formation of more complex oxygen - bearing organic molecules ( bottinelli et al . 2007 ;
whittet et al . 2011 ) .
five rotational transitions of methanol have been identified in our observations spanning an upper level energies of @xmath57 k. in addition , we have detected two transitions of ch@xmath0oh from vibrational state at frequencies 217.29920 and 217.64286 ghz with upper level energies of 374 and 746 k , respectively .
the ch@xmath0oh has highest fractional abundance of @xmath58 than other molecules .
higher ch@xmath0oh abundance is also observed in other massive star - forming regions ( e. g. , qin et al .
2010 , 2015 ; neill et al .
2014 ) . the higher ch@xmath0oh abundance of @xmath59 can not be explained by gas - phase chemical model ( lee , et al .
the gas - phase chemical model suggested that ch@xmath0oh can be produced in gas phase via radiative association of ch@xmath60 and h@xmath1o at temperature @xmath61 100 k , giving ch@xmath0oh abundance of @xmath27 @xmath62 ( lee et al .
in contrast , the grain chemical model suggested that ch@xmath0oh is formed on grain surface via hydrogenation of co , followed with evaporating from grain surface into gas phase due to temperature increasing , which can produce higher ch@xmath0oh abundances .
the higher abundance of @xmath63 in our observations favor that this molecule originates from grain surface chemistry ( charnley et al . 2004 ; garrod & herbst 2006 ) .
acetone ( ch@xmath0coch@xmath0 ) had been detected for the first time by combes et al .
( 1987 ) and later confirmed by snyder et al .
so far acetone ( ch@xmath0coch@xmath0 ) was only detected in the hot molecular core sgr b2(n ) ( snyder et al . 2002 ) , the orion kl hot core ( friedel et al . 2005 ; peng et al . 2013 ) , and g24.78 + 0.08 ( codella et al . 2013 ) .
we have identified 13 transitions of ch@xmath0coch@xmath0 in w51 north region , spanning a wide energy range of @xmath64 k. the column density of @xmath65@xmath25 and gas temperature of 140 k are estimated by use of the xclass , and the higher abundance of @xmath66 is obtained .
combes et al .
( 1987 ) proposed that ch@xmath0coch@xmath0 is formed in gas - phase via the radiative association reaction .
however herbst , giles & smith ( 1990 ) argued that this radiative association reaction is very inefficient to produce observed ch@xmath0coch@xmath0 abundance , and this process can only explain the lower abundance of @xmath67 .
recently , the observations of ch@xmath0coch@xmath0 toward the orion - kl have shown that the distribution of ch@xmath0coch@xmath0 is similar to n - bearing molecules concentrated at the hot core rather than o - bearing molecules peaked at the compact ridge , therefore the formation of ch@xmath0coch@xmath0 may invoke n - bearing molecules , and grain surface chemistry or high temperature gas - phase chemistry may play important roles ( friedel et al .
2005 , 2008 ; peng et al .
the well mixed nitrogen- and oxygen - bearing molecules based on gas distributions and temperatures in w51 north indicate that the hot core is not very developed in w51 north and grain surface chemistry and evaporating from grain surface into gas phase may be responsible for the high ch@xmath0coch@xmath0 abundance of @xmath66 , since the gas temperature is not so high in w51 north .
compared to the sgr b2(n ) ( snyder et al . 2002 ) , the orion - kl ( peng et al . 2013 ) , and g24.78 + 0.08 ( codella et al . 2013 ) molecule line cores , w51 north region has higher ch@xmath0coch@xmath0 abundance .
the ch@xmath0coch@xmath0 can be destructed in gas phase by efficient radiative association of h@xmath60 and ch@xmath0coch@xmath0 or the collisions of ch@xmath0coch@xmath0 and c@xmath68 ( combes et al .
1987 ; herbst , giles & smith 1990 ) .
a possible interpretation of higher ch@xmath0coch@xmath0 abundance is that unlike the sgr b2(n ) , the orion - kl , and g24.78 + 0.08 , w51 north is at very early stage of massive star formation , and the ch@xmath0coch@xmath0 is destructed less and converted into the other molecules in a short time .
we have identified 13 transitions of c@xmath1h@xmath2oh in w51 north region .
these transitions have a wide spread of the upper level energies of @xmath69 k. the c@xmath1h@xmath2oh has higher abundance of @xmath70 in our observations . compared to other sources , the c@xmath1h@xmath2oh abundance of @xmath71 and @xmath72/@xmath73 @xmath74 0.32 are much higher than those in other massive star - forming regions ( ohishi et al . 1995 ; qin et al .
2010 ; bisschop et al . 2007 ) .
hence w51 north is a very ethanol - rich source .
the gas - phase chemical reaction via ion - molecule reaction predicted c@xmath1h@xmath2oh abundance of @xmath75 ( leung , herbst & huebenr 1984 ) .
higher abundance of @xmath70 in our observations ca nt be explained by pure gas - phase chemical model .
the grain chemistry suggested that c@xmath1h@xmath2oh can be formed on the grain surface via hydrogenation of ch@xmath0co ( charnley et al .
2004 ; ohishi et al .
1995 ) , and the model predicted high c@xmath1h@xmath2oh abundance of @xmath76 within @xmath27 @xmath77 yr after evaporated from the grain surface into gas phase . our result appears to support that the c@xmath1h@xmath2oh is synthesized on grain surface ( ohishi et al .
1995 ; charnley et al .
1995 ) .
four transitions of ch@xmath0och@xmath0 are detected in w51 north .
we estimated its rotation temperature of 140 k and column density of @xmath78 @xmath25 from the xclass calculation .
the abundance ratio of @xmath79/@xmath72 is approximately 0.24 .
the @xmath79/@xmath72 ratios have a great deal of changes from source to source ( fontani et al . 2007 ;
qin et al . 2010 ; millar et al . 1988 ; schilke et al . 1997 ; bisschop et al . 2007 ) .
although ch@xmath0och@xmath0 is isomer of c@xmath1h@xmath2oh , the abundances of c@xmath1h@xmath2oh and ch@xmath0och@xmath0 are not well correlation , suggesting that they have different chemical route , and do not have same parent molecules .
as one of large oxygen - bearing molecules , ch@xmath0ocho has been reported in many hot molecular cores and corinos ( blake et al .
1987 ; hatchell et al . 1998 ; nummelin et al .
2000 ; cazaux et al .
2003 ; bottinelli et al .
sixteen transitions of ch@xmath0ocho are identified in w51 north , spanning a upper energy level range of @xmath80 k. high fractional abundance of @xmath81 is obtained .
gas temperature , column density and spatial distribution of ch@xmath0ocho are similar to those of ch@xmath0och@xmath0 ( see table 2 and figure 3 ) .
similar case is also seen in orion - kl , in which the distributions of ch@xmath0ocho and ch@xmath0och@xmath0 present a striking similarity , and the two species have comparable gas temperature and column density ( brouillet et al .
2013 ) , suggesting that ch@xmath0ocho and ch@xmath0och@xmath0 may be chemically related and have similar formation mechanism ( brouillet et al .
2013 ) .
charnley , tielens & millar ( 1992 ) proposed that ch@xmath0och@xmath0 and ch@xmath0ocho can be formed in gas phase chemistry via molecule - ion reaction of ch@xmath0oh@xmath82 with ch@xmath0oh and h@xmath1co , respectively .
their model predicted high abundance of @xmath83 for ch@xmath0och@xmath0 and ch@xmath0ocho at age of @xmath84 yr .
our results appear to favor the gas - phase formation routes of ch@xmath0och@xmath0 and ch@xmath0ocho by charnley , tielens & millar ( 1992 ) .
we have identified 9 ch@xmath1chcn transitions with upper level energies e@xmath38 of @xmath85 k. the column density of @xmath86@xmath25 and the rotation temperature of 140 k are obtained .
three ch@xmath0ch@xmath1cn transitions with upper level energies e@xmath38 of @xmath87 k are identified .
the column density of @xmath88 @xmath25 and rotation temperature of 140 k are estimated by use of the xclass .
ch@xmath1chcn and ch@xmath0ch@xmath1cn have same temperature of 140 k , while ch@xmath0ch@xmath1cn has higher column density than ch@xmath1chcn .
previous observations suggested that the abundances of two molecules are strongly correlated ( fontani et al .
@xmath89/@xmath90 in w51 north is approximate 0.35 which agrees well with abundance correlation of ch@xmath0ch@xmath1cn and ch@xmath1chcn in other star - forming regions ( fontani et al .
2007 ) and ch@xmath1chcn may be formed through gas phase reactions involving ch@xmath0ch@xmath1cn ( caselli , hasegawa & herbst 1993 ) .
while ch@xmath0ch@xmath1cn can be formed by successive hydrogenation of hc@xmath0n on dust grains and released into the gas phase as temperature rises ( blake et al .
1987 ; caselli , hasegawa & herbst 1993 ) .
the ratio of @xmath91/@xmath92 @xmath74 10 in our observations is consistent with the model prediction of grain surface chemistry ( blake et al .
1987 ) .
we have detected one transition of the deuterated species of hydrogen cyanide ( dcn ) , in w51 north .
the column density of @xmath93 @xmath25 is estimated , which is much larger than those in other sources associated with uc hii regions ( hatchell , millar & rodgers 1998 ) .
the deuterated species are thought to be synthesized at the early evolutionary stage of star formation ( miettinen , hennemann & linz 2011 ) .
previous studies suggested that dcn can be formed via d - h substitution of the hcn or the reaction of chd with n on the grain mantles and then be released into the gas - phase , while dcn can be destroyed with temperature increasing ( schilke et al .
1992 ; hatchell , millar & rodgers 1998 ) .
hence the larger abundance observed in w51 north is not a surprise and suggests that it may be synthesized at the early evolutionary stages of star formation .
as isotopomer of methyl acetylene ch@xmath1dcch has been reported in the dark cloud tmc-1 cp by gerin et al .
( 1992 ) , and ch@xmath0ccd had been successfully detected in same region by markwick et al .
so far the ch@xmath0ccd has been less reported in other massive star formation regions .
therefore ch@xmath0ccd is an important molecule for investigating the difference of physical and chemical environments between the dark clouds and hot molecular cores .
we have successfully identified 3 transitions of ch@xmath0ccd with e@xmath38 ranging from 78 to 143 k. higher ch@xmath0ccd abundance of @xmath94 and temperature of 130 k compared to dark clouds ( markwick et al . 2005 ) are estimated , which may indicate that ch@xmath0ccd is synthesized on grain surface and release to gas phase as temperature increases .
one transition of sis ( j=1211 ) at 217.81766 ghz is identified , and the estimated column density and abundance are @xmath95 @xmath25 and @xmath96 , respectively .
so far this species is only detected in carbon star irc+10216 , sgr b2(n ) , sgr b2(m ) and sgr b2(oh ) ( bieging & nguyen 1989 ; turner 1991 ; belloche et al .
the column density of @xmath97 @xmath25 is higher than @xmath98@xmath25 in sgr b2 . while the factional abundance of @xmath99 in w51 north is lower than @xmath100 in the carbon star ,
the abundance difference of four order of magnitude may be caused by the different physical and chemical environments in massive star formation regions and the asymptotic giant branch star .
sulfur - bearing molecules of h@xmath1s and so@xmath1 are observed in w51 north , with higher gas temperature than other molecules .
the fractional abundances of so@xmath1 and h@xmath1s are @xmath101 and @xmath102 , respectively .
the ratio of @xmath103/@xmath104 is appropriately 0.05 , which is same order of magnitude as those in massive protostar cores ( herpin et al .
2009 ) , which is consistent with chemical model that h@xmath1s can be formed on the grain surface and h@xmath1s is converted to so first , then the so is converted to so@xmath1 as temperature rises ( charnley 1997 ; wakelam et al .
2004 ; woods et al .
we present the submillimeter array ( sma ) observations of molecular lines in two 2 ghz - wide bands centered at 217.5 and 227.5 ghz , toward massive star forming region w51 north .
we identified 84 transitions from 17 species , including oxygen- , nitrogen- and sulfur - bearing molecules .
our main conclusions are as follows : + 1 .
the gas distributions of both oxygen - bearing and nitrogen - bearing molecules show a compact core concentrated on the strongest continuum source , which indicates that nitrogen- and oxygen - bearing molecules are well - mixed in space in w51 north region .
+ 2 . under the assumption of local thermodynamic equilibrium ,
the molecular rotation temperatures and column densities are estimated by use of the xclass program .
the oxygen - bearing molecules have higher fractional abundance than the nitrogen - bearing molecules .
the rotation temperatures range from 100 to 200 k , suggesting that the molecular emissions originate from warm environments .
thirteen transitions of ch@xmath0coch@xmath0 are identified .
higher fractional abundance of ch@xmath0coch@xmath0 are obtained in w51 north when compared to the other massive star - forming regions , _
e.g. _ sgr b2(n ) , orion kl , g24.78 + 0.08 .
these results seem to indicate that ch@xmath0coch@xmath0 is synthesized on grain surface at the early evolutionary stage of massive star formation .
higher fractional abundance of ch@xmath0oh and c@xmath1h@xmath2oh are obtained , which can not explained by gas phase reactions and the two molecules may be synthesized on grain surface . @xmath105/@xmath106 @xmath74 0.32 are much higher than those in other massive star - forming regions , and w51 north is a very ethanol - rich source .
similar gas distribution , rotation temperature and abundance between ch@xmath0ocho and ch@xmath0och@xmath0 suggest that they may be chemically related and have similar formation mechanism
. the two molecules may originate from gas phase chemistry .
@xmath89/@xmath90 in w51 north agrees well with abundance correlation of ch@xmath107ch@xmath47cn and ch@xmath47chcn in other star - forming regions .
ch@xmath107ch@xmath47cn can be formed by successive hydrogenation of hc@xmath107n on dust grains while ch@xmath47chcn may be formed through gas phase reactions involving ch@xmath107ch@xmath47cn .
ch@xmath0ccd and sis have been less reported in other star - forming regions .
higher fractional abundances of ch@xmath0ccd and sis are estimated in w51 north region .
we thank the anonymous referee , and editor morgan hollis for their constructive comments on the paper . this work has been supported by the national natural science foundation of china under grant nos .
11373026 , 11373009 , 11433004 , 11433008 , u1331116 , and the national basic research program of china ( 973 program ) under grant no .
2012cb821800 , by top talents program of yunnan province and midwest universities comprehensive strength promotion project ( xt412001 , yunnan university ) .
millar t. j. , brown p. d. , olofsson h. , hjalmarson a. , 1988 , a&a , 205 , l5 mller , t. , endres , c. , & schilke , p. 2015
, accepted by a&a mller h. s. p. , schlder f. , stutzki j. , winnewisser g. , 2005 , jmost , 742 , 215 | we present submillimeter array ( sma ) molecular line observations in two 2 ghz - wide bands centered at 217.5 and 227.5 ghz , toward the massive star forming region w51 north .
we identified 84 molecular line transitions from 17 species and their isotopologues .
the molecular gas distribution of these lines mainly peaks in the continuum position of w51 north , and has a small tail extending to the west , probably associated with w51 d2 .
in addition to the commonly detected nitrogen and oxygen - bearing species , we detected a large amount of transitions of the acetone ( ch@xmath0coch@xmath0 ) and methyl formate ( ch@xmath0ocho ) , which may suggest that these molecules are present in an early evolutionary stage of the massive stars .
we also found that w51 north is an ethanol - rich source .
there is no obvious difference in the molecular gas distributions between the oxygen - bearing and nitrogen - bearing molecules . under the assumption of local thermodynamic equilibrium ( lte ) , with the xclass tool , the molecular column densities , and rotation temperatures are estimated .
we found that the oxygen - bearing molecules have considerable higher column densities and fractional abundances than the nitrogen - bearing molecules .
the rotation temperatures range from 100 to 200 k , suggesting that the molecular emission could be originated from a warm environment . finally , based on the gas distributions , fractional abundances and the rotation temperatures ,
we conclude that ch@xmath0oh , c@xmath1h@xmath2oh , ch@xmath0coch@xmath0 and ch@xmath0ch@xmath1cn might be synthesized on the grain surface , while gas phase chemistry is responsible for the production of ch@xmath0och@xmath0 , ch@xmath0ocho and ch@xmath1chcn .
[ firstpage ] ism : abundances ism : individual ( w51 north ) ism : molecules radio lines : ism stars : formation |
the collapse of a dense cloud core into a star happens rapidly @xcite and therefore is rarely observed ; furthermore , most dense cores appear not to be collapsing but to be close to equilibrium ( e.g. , * ? ? ?
if they are not sufficiently supported by gas pressure and turbulence , then a plausible level of magnetic field strength would suffice to support them ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
because the process of collapse is so rarely observed , there is considerable debate about what triggers it and how it proceeds to form a star .
additional examples of dense cores in extreme states that indicate either collapse or a condition just prior to collapse are critical to constrain theories about the first stages of star formation .
mm - wave line observations are the traditional means to search for collapsing cloud cores , such as in the pioneering study by @xcite .
the subsequent discussion ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) illustrates the ambiguities in the interpretation of velocity data in star forming dense cores .
alternative observational approaches are necessary to supplement such line measurements and provide a better understanding of collapsing and nearly - collapsing dense cores . in this paper
we characterize the structure of the extremely dense core l429 , using the infrared shadow technique we have demonstrated on two other globules @xcite .
this core is already known to have an extremely high deuterium fraction , believed to be an indication of its being very close to collapsing into a protostar , and a low internal temperature of only @xmath2 k , reducing pressure support to low levels @xcite .
@xcite detected a 7 shadow from l429 using isocam .
here , we report that this shadow is so opaque that it is detectable at 70 .
the presence of significant absorption at this long wavelength demonstrates that l429 is indeed extremely dense .
our analysis of the shadow also demonstrates that it is very compact with a steep density gradient .
we show how these characteristics put l429 in the collapsing or near collapsing starless core category .
l429 was observed by _
spitzer _ using mips , program i d 30384 , p.i .
tyler l. bourke , and with the irac instrument , program i d 20386 , p.i .
philip c. myers .
the mips observations utilized scan map mode in all three bands .
the data reductions were carried out as described in @xcite .
the mips images are @xmath3 in size , with exposure times of @xmath4 s at 24 , @xmath5 s at 70 , and @xmath6 s at 160 . for the 70 and 24 data respectively the beam size is @xmath7 and @xmath8 , the nominal default pixel scale is @xmath9 and @xmath10 , and the pixel scale in our images is @xmath11 and @xmath12 .
the irac reductions were carried out as described in @xcite .
the irac images are @xmath13 in size , with exposure times of @xmath14 s. the _ spitzer _ data are shown in figure [ fig : img ] .
we also use the 2mass all sky point source catalog ( psc ) @xcite photometry for two well detected sources within @xmath15 of the center of l429reddening estimates for these sources , assuming the @xcite extinction levels , are plotted in figure [ fig : av ] .
the commonly accepted distance to l429 is 200 pc ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , based on the likely association of this cloud with the aquila rift .
@xcite find a distance to the aquila rift based on the co line v@xmath16 km s@xmath17 .
they summarize the results of various studies and conclude that the distance to the aquila rift is consistent with @xmath18 pc .
more recently , @xcite derive a distance to the near side of the rift of @xmath19 pc , while @xcite derive a distance to an aquila cloud complex centered at @xmath20 and @xmath21 of @xmath22 pc , both consistent with previous studies .
the depth of the absorption feature ( see fig . [
fig : img ] ) implies that l429must lie on the near side of , or in front of , the aquila rift because substantial amounts of foreground material would be warmed by the interstellar radiation field and would dilute the shadow .
we assume a distance of 200 pc , consistent with previous distance determinations .
if the distance to l429 is larger the derived mass will increase , making the core less stable and more likely to be collapsing .
l429 is seen in absorption at @xmath23 70 ; at 160 there is large - scale emission but no isolated local peak is observed ( see fig . [
fig : img ] ) .
in addition , no compact emission source is detected at any _ spitzer _ wavelength within the shadow , and so no source of luminosity @xmath24 l@xmath25 is present @xcite .
we derive extinction profiles from the observed shadows at 70 , 25 , 8.0 , and 5.8 using techniques similar to those discussed in detail in @xcite .
the first step in deriving the 70 profile is to subtract from the image a value that represents zodiacal emission and other foreground sources . for l429
the shadow is the darkest part of the image .
for this reason the image zero level can not be determined directly from the image itself . to set an upper limit on the optical depth ,
we zero the image using the lowest pixel value in a @xmath26 box centered on the shadow ; in this case the zero value is @xmath27 mjy sr@xmath17 below the mean pixel flux level immediately outside the shadow .
ideally , one would compute the minimum value based on more pixels .
however , the 70 pixel size is a non - negligible fraction of the size of the core of absorption feature ; taking more pixels would dilute the depth of the shadow .
we use iras data to set the lower limit on the optical depth profile : we measure the flux level outside the shadow to be @xmath28 mjy sr@xmath17 above the foreground component of emission , interpolated to 70 from the 60 and 100 iras images and estimated far away from regions with significant emission ( about @xmath29 away ) . for this value ,
the minimum shadow value is @xmath30 mjy sr@xmath17 above the image zero level .
this zero level is a very conservative estimate because the shadow region is surrounded by high levels of emission which are very likely to raise the foreground contribution above the level we use here .
this lower limit also accounts for the mips 70 calibration error , @xmath31% @xcite , and the iras extended source calibration error , which can be as large as @xmath32% @xcite . after zeroing the image we mask out all pixels with possible emission above the sky level .
the mask threshold is set to the mean value plus @xmath33 measured for pixels in a @xmath34 region @xmath35 north of the shadow .
next we measure the mean pixel values , termed @xmath36 here , in nested , and concentric , circular regions , with radii starting at @xmath37 and increasing in @xmath38 steps , out to @xmath39 , centered at ra = @xmath40 , dec = @xmath41 ( j2000 ) , without accounting for fractions of pixels .
we ignore all pixels in the southern half of the shadow because the shadow has a tail extending to the south and the assumption of spherical symmetry breaks down ( see figure [ fig : img ] ) . finally , to determine the optical depth at 70 , @xmath42 , we average profile values at radii @xmath43 to measure @xmath44 , the intrinsic , unobscured local flux value . the outer radius of the core at 70 , equal to @xmath45 ,
is set by the observed flattening of the measured flux profile at that radius , indicating that there is little to no absorption at larger radii .
following @xcite , we derive the 70 @xmath46 profile shown in figure [ fig : av ] with open triangles : downward pointing triangles represent the upper limit and the upward pointing triangles represent the lower limit .
the innermost region has an optical depth of @xmath47 to @xmath48 ( upper and lower limits ) .
the total mass ( gas plus dust ) implied by the 70 profile is given by @xmath49 where we have assumed a gas to dust mass ratio of 100 and the @xcite model dust opacity for coagulated dust grains of @xmath50 @xmath51 gm@xmath17 .
the value of @xmath52 is the _ upper limit _ on the mass ; the _ lower limit _ mass is @xmath53 . the enclosed mass profile is shown in fig . [
fig : mass ] .
as noted earlier , our limits on @xmath54 are very conservative .
we performed an error analysis to measure the effect of a reasonable perturbation in the assumed zero level and the calculated value for @xmath44 and find that the errors are absorbed into the limits by a wide margin . for a dust opacity of @xmath55 @xmath51 gm@xmath17 , from the @xmath56 model from @xcite
, we calculate a 70 mass of @xmath57 .
the method used to derive the shorter wavelength profiles is essentially the same as described above .
first , we zero the images using the darkest pixel in the shadow .
second , the mask threshold is determined from the pixel value distribution in the sky region .
we measure the mean pixel value in circular regions starting with a radius of @xmath10 and increasing in steps of @xmath10 , out to a radius @xmath39 .
the value of @xmath44 is determined by averaging pixel values between radii of @xmath15 and @xmath58 .
these radii were chosen based on a flattening of the 24 profile in this region . at radii between @xmath58 and @xmath59
the profile continues to rise , likely due to contamination from other nearby sources .
using larger values of @xmath44 would artificially increase @xmath54 .
this choice of @xmath44 is further justified by the 8.0 and 5.8 profiles : they show an increase in the mean pixel value well past @xmath15 but a decreasing value approaching the inner estimate at radii larger than @xmath60 , and by a radius of @xmath61 the profiles match the inner values .
this increase in the profile values at intermediate radii may be caused by limb brightening of the core as it is irradiated by the interstellar radiation field .
if the emission raises the observed shadow flux then it will cause an underestimate of @xmath54 at all wavelengths ; however , if it predominantly increases the unobscured flux levels outside the shadow @xmath54 will be over estimated .
we do not attempt to correct for this effect .
we calculate optical depth limits of @xmath62 and @xmath63 through the center of the profiles .
the implied total masses at these wavelengths is @xmath64 for all three profiles , assuming @xcite model dust opacities .
the discrepancy between this mass and the 70 mass ( between 1.6 and 5.1 ) most likely results from the high optical depth in the core .
if the core is dense enough to be observed in absorption at 70 then it is certainly opaque at shorter wavelengths ; the @xmath46 profile and mass estimates at shorter wavelengths will be lower limits . in the cb190 case
the core was observed in absorption at 24 but not at 70 , and the maximum measured extinction value was @xmath65 mag ( @xmath66 ) , much smaller than the values measured here @xcite .
therefore we reject all 24 , 8.0 , and 5.8 @xmath67 points and we scale the profiles to match the 70 extinction values at the outer two 70 radii . in figure
[ fig : av ] we plot the scaled profiles ; the shapes of the profiles match well .
bonnor - ebert models are often used to characterize starless cores ; these models generally fit the observed density profiles with reasonable fidelity ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
bonnor - ebert models describe a self gravitating pressure confined isothermal sphere in hydrostatic equilibrium .
the density profiles are described by three parameters : temperature , central density @xmath68 ( or equivalently , @xmath69 ) , and the outer radius of the cloud .
a model will be unstable to collapse if @xmath70 .
we fit said models to the extinction profiles discussed above .
we use a 3d grid of models in temperature , radius , and @xmath69 ; see @xcite for a more detailed discussion of the models and the fitting method .
one strong constraint in interpreting the best
fit models is that l429 is dark at @xmath71 , and therefore the temperature can not be greater than @xmath72 k , in agreement with other core temperature measurements ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
fit models for the upper and lower bounds on the 70 profile are shown in figure [ fig : av ] as grey dashed lines .
the 70 profile has a best fit temperature of @xmath73 k , for the upper and lower bounds , respectively ; these temperatures are in clear contradiction with the data .
the best fit outer radius is @xmath74 , for both the upper and the lower limit , and the best
fit @xmath69 = 4.5 and 3.0 , for the upper and lower limit respectively . as the assumed distance
is increased ( in this case to 270 pc ) the best
fit models become even more implausible , with temperatures always @xmath75 k. at a distance of 100 pc the best
fit temperatures are @xmath76 k , still in contradiction with the data .
these high temperatures indicate that l429 requires additional support ( turbulence and magnetic fields ) beyond thermal to prevent collapse .
@xcite observe n@xmath77d@xmath78 and n@xmath77h@xmath78 line widths @xmath79 km s@xmath17 and calculate excitation temperatures @xmath80 k. the particle of mean molecular mass has velocity fwhm @xmath81 km s@xmath17 for temperature 8 k , indicating that the turbulent energy contribution to the support in l429 is only comparable to the thermal contribution @xcite .
magnetic fields are a possible source of support in starless cores . inferred magnetic field strengths in cores range from , e.g. , @xmath82 g to @xmath83 g @xcite . for @xmath84 ,
a distance of 200 pc , and @xmath85 @xmath86 g , the mass that can be supported is @xmath87 @xcite .
furthermore , from equation 1 and the @xcite analysis , the condition of stability requires a minimum magnetic field strength of : @xmath88 where @xmath89 is the angular radius of the core .
the mean column density of the 70 profile is @xmath90 @xmath91 ; given this column , the correlation presented by @xcite implies a line - of - sight magnetic field of @xmath92 @xmath86 g ; we estimate the total field as @xmath93 @xmath86 g , although this relation has large scatter .
although the field required to support l429 is larger than expected , our understanding of the limits on magnetic field strengths does not allow for a strong constraint .
in addition to requiring a field exceeding the likely range of values , supporting the globule entirely with magnetic pressure has other consequences .
one might expect asymmetries to develop , given the asymmetric pressure provided by magnetism .
if such structures can be avoided @xcite , another issue is that the field should leak out of the globule by ambipolar diffusion , on timescales @xmath94 yr @xcite .
thus , either the field is already too small to prevent the collapse of the globule , or the collapse will only be delayed for a time
. if infall is occurring in l429 , the observed line width is very narrow .
@xcite detect a double - peaked n@xmath77h@xmath78 @xmath95 line and @xcite detect a double - peaked cs @xmath96 line , both measurements with roughly equal brightness in the blue and red peaks .
@xcite state that because other thin tracers show single gaussian components , the observed double peak is likely due to self absorption at high optical depth ; these authors also note that l429is one of two sources out of 163 spectra that show small wing components . @xcite
classify l429 as a possible infall candidate based on their observation of cs @xmath97 , which has a double peaked profile with more emission in the blue peak .
@xcite also measure a double peaked profile in hcn @xmath95 ; they note that this source is one of two ( out of 85 cores surveyed ) that has some line strength anomalies in the three hyperfine hcn lines .
recently , @xcite have found that l429 has a large amount of ortho - h@xmath77d@xmath78(1@xmath981@xmath99 ) emission , which they find to be correlated with increased central concentration , co depletion , and deuteration , providing further evidence that l429 is very evolved .
we measure the density profile of l429 from the 70 absorption feature observed with the _ spitzer _ telescope .
we conclude that : \5 .
the magnetic field required to make up the rest of the required pressure for support would be surprisingly large ; even if present , such a large field will leak out of the globule through ambipolar diffusion .
we thank the referee for their helpful comments .
support for this work was provided through nasa contracts issued by caltech / jpl to the university of arizona ( 1255094 ) and to sao ( 1279198 and 1288806 ) .
this work was also supported by the national science foundation grant ast-0708131 to the university of arizona .
basu , s. 2004 , in young local universe , proceedings of xxxixth rencontres de moriond , eds .
a. chalabaev , t. fukui , t. montmerle , and j. tran - thanh - van .
paris : editions frontieres , arxiv astrophysics e - prints , arxiv : astro - ph/0410534 | we present _ spitzer _ infrared observations of the starless core l429 . the ir images of
this core show an absorption feature , caused by the dense core material , at wavelengths @xmath0 70 .
the core has a steep density profile , and reaches @xmath1 mag near the center .
we show that l429 is either collapsing or in a near collapse state . |
0.5 cm quantum computers are actively being sought for the last couple of decades .
basic hope being that quantum mechanics promises several features to help faster computations if quantum features are properly implemented in the hardware architecture of such computers .
traditional architecture of classical computers is logical gate - based ones .
the linear superposition processing of the wave functions in quantum mechanics helps simultaneous ( probabilistic ) processing of the binary bits or qubits ( of information ) .
quantum mechanics also promises major advantages of parallel operation of these gates in appropriate architectures .
the problem of decoherence has not allowed so far any gate - based quantum computer which is able to handle more than a couple of qubits . even in classical computers , in order to solve computationally hard problems , like the traveling salesman problem , one artificially generates stochastic algorithms , like the simulated annealing techniques ( in the so - called ` boltzmann machine ' ) for practical and efficient searches .
the essential problem in such searches is that the system gets locked in some local minimum separated from the other deeper minimum by ( cost function or free energy ) barriers . noting that quantum tunneling feature across such barriers can help @xcite searching for the minimum ( optimal solution ) ,
quantum annealing techniques have been developed recently [ 1 - 29 ] ( see figs
. 1 - 28 ) . in the last couple of years
, such techniques have been efficiently implemented in the computer architectures and such quantum annealing computers have already arrived in the market ( see fig .
[ bbc.pdf ] ) and major successes are being demonstrated @xcite .
these exciting developments are also being captured in several recent notes ( e.g. , @xcite ) , reviews ( e.g. , @xcite ) and books ( e.g. , @xcite ) .
0.5 cm ray et al . @xcite and thirumalai et al .
@xcite suggested ( in 1989 ; see figs .
[ ray-89.pdf ] and [ thirumalai-89.pdf ] ) that quantum tunneling from local trap states across the ( narrow ) free energy barriers of macroscopic height ( of order @xmath0 ; coming from the requirement of the flips of finite fraction of all @xmath0 spins ) in a sherrington - kirppatrick spin glass ( in transverse field ) can help evolving such a complex system towards its ( degenerate ) ground state(s ) .
although the idea was criticized heavily in the subsequent literature , essentially on the ground that the incoherent phase overlaps of the tunneling states ( waves ) will localized the system and will not allow evolution towards the ground state(s ) .
later , theoretical investigations ( @xcite ) and the experimental demonstrations ( @xcite ) lead to a very promising development resulting in the quantum annealing technique @xcite with the hardware implementation by d - wave system @xcite .
successful checking and applications @xcite led to the emergence of a new era in quantum computing ( see e.g. , the comments by bose @xcite , fig .
[ bose-13.pdf ] ) . .
0.5 cm in may this year ( 2013 ) , there have been several news posts and blogs in popular science journals as well as newspapers informing about the successful tests of a quantum computer with about 100 qubits ( order of magnitude higher than those available otherwise ) , based on quantum annealing technique and marketed by the d - wave systems inc .
interestingly , the nasa group of consortium had already placed an order to them for a 512 qubit quantum annealing computer .
[ bbc.pdf ] shows the bbc news blog on the purchase deal with d - wave inc . by the nasa - google consortium . in an attempt to explain in a popular way how quantum tunneling can help such analog computers to get out of the `` local '' solutions and
anneal down quantum mechanically to the `` global '' or `` optimal '' solutions with proper tuning of the tunneling term , a scientific american news blog appeared last may ( see fig .
[ sci-ame.pdf ] ) .
it explains in brief the working principle involved , using the wikipedia entry on quantum annealing ( partly reproduced in fig .
[ wiki.pdf ] ) .
the rugged nature of the ( free ) energy landscape ( energy versus spin configurations ) of a classical ( ising ) spin glass does not allow searches for the global energy minima by simple rolling down the landscape ( say , using an energy dissipative dynamics ) .
essentially the system gets trapped in the local minima , separated from the global minima often by macroscopically high ( free ) energy barriers .
ray et al .
@xcite first pointed out that if such barriers are narrow , quantum mechanical tunneling ( as in a transverse ising spin glass model ) can help such searches ( see fig .
[ ray-89.pdf ] ) .
as mentioned already , this indication was criticized heavily in the following literatures .
however , some crucial features of such tunneling effects were immediately checked , with positive results in some solid - state samples discovered by wu et al .
( see fig .
[ wu-91.pdf ] ) .
the possibility of tuning the quantum tunneling term to achieve the minimization of a multi - variable optimization problem was pointed out by finnila et al .
@xcite ( see fig . [ finnila-94.pdf ] ) .
however , the use of this annealing ( of the tunneling field ) in the well - characterized ground state search problems of ( frustrated ) spin systems were convincingly demonstrated , using numerical techniques , by kadwaki and nishimori @xcite ( see also fig .
[ nishimori-98.pdf ] ) and the reported success in this paper made a major impact on the following developments .
soon , brook et al .
@xcite extended their earlier experimental investigations ( see fig .
[ wu-91.pdf ] ) and with suitable tuning of the tunneling field , observed clear advantages of the quantum annealing in the search for ground state(s ) for such samples ( see fig .
[ aeppli-99.pdf ] ) .
this experimental demonstration of the clear advantages of quantum annealing established the field . soon major investigations by several groups from all over the world started .
farhi et al .
@xcite ( see fig . [ farhi-01.pdf ] ) indicated that such an adiabatic evolution ( zero temperature annealing ) algorithm may help solving the computationally ( np- ) hard problems .
the estimate of the growth of errors in the optimal search , with the decrease in the annealing time , was made first by santoro et al .
@xcite ( see fig . [ santoro-02.pdf ] ) and the extensions ( and some clarifications ) of the experimental investigations by ancona - torres et al .
( see fig .
[ aeppli-prl-2008.pdf ] ) enthused the community to explore further in much more meaningful way . soon
( during 2006 - 2008 ) , some important reviews on quantum annealing techniques were written ( see figs .
[ santoro-06.pdf ] , [ das-rmp.pdf ] and [ morita-08.pdf ] ) : santoro and tosatti @xcite and morita and nishimori @xcite reviewd the mathematical structures of the quantum annealing techniques , while das and chakrabarti @xcite reviewed the physical structure of quantum annealing and discussed its possible implications for analog quantum computations .
in 2011 , d - wave system announced @xcite the arrival of `` world s first commercially available quantum computer , operating on a 128 qubit chip - set using quantum annealing '' ( see figs .
[ d-wave.pdf ] , [ johnson-12.pdf ] ) .
as indicated already , this news created huge enthusiasms as well as a lot of criticisms in the community of scientists .
however , soon some leading research groups came forward to test the performances of these machines with remarkably positive results ( see e.g. , figs .
[ ortiz-12.pdf ] , [ arxiv.pdf ] , [ boixo-12.pdf ] and [ bioxo-13.pdf ] ) .
there were parallel investigations on the possible performances of the quantum annealing technique in the context of various computationally hard problems ( see e.g. , figs .
[ yamamoto.pdf ] , [ somma-12.pdf ] and [ titiloye-12.pdf]).[multiblock footnote omitted ] ; website : http://blogs.scientificamerican.com/guest-blog/2013/05/17/is-it-quantum-computing-or-not/ ) . ,
width=566 ] ) for the possible advantage of quantum tunneling between well characterised ( classically ) localized states in the same long - range transverse ising spin glass model.,width=453 ] .,width=566 ] ) for quantum glasses .
, width=453 ]
approximate solutions of computationally hard problems were obtained more easily in boltzmann - like machines which employ stochastic searches and ( classical or thermal ) annealing rather than employing sequential search methods .
it was noted that for np - hard problems , the effective cost function landscape ( in the solution state or configuration space ) becomes extremely rugged as discussed in the text .
even with classical annealing tricks the probability of escape from a local minimum to another lower one separated by a barrier of height of order @xmath0 ( the problem size ) decreases as @xmath1 , implying that the time to arrive at the solution in not bounded by any polynomial in @xmath0 .
noting that the quantum tunneling probability across such a barrier decreases with width of the barrier ( becoming finite in the delta function barrier limit ) , ray et al .
@xcite proposed in 1989 that quantum tunneling might help solving np - hard problems in polynomial ( in @xmath0 ) time ! subsequently , the researches by finnila et al .
@xcite , kadowaki and nishimori @xcite , brooke et al .
@xcite and ferhi et al .
@xcite led to the robust development of quantum annealing technique , indicating clearly the possible development of analog quantum computers with such tricks ( cf .
santoro and tosatti @xcite , das and chakrabarti @xcite ) .
with the major breakthrough achieved by d - wave computers @xcite ( with indications @xcite of search time @xmath2 ; @xmath3 ) , a new era in quantum computing has started : see e.g. , @xcite etc .
as a few chosen examples only of the rapidly growing publications which ( as shown in figs .
[ ortiz-12.pdf ] , [ arxiv.pdf ] , [ yamamoto.pdf ] , [ boixo-12.pdf ] , [ bioxo-13.pdf ] ) indicate also the role of the original papers in this remarkable development .
we note that the initial contributions by our indian colleagues had indeed been pioneering , though the follow - up researches and contributions have been rather slow .
this documentary note was tentatively planned during the international conference `` diversity and complexity : realm of today s statistical physics '' held in saha institute of nuclear physics , during 14 - 17 january , 2013 to commemorate sixty year of prof .
bikas k. chakrabarti .
we are also grateful to prof .
indrani bose for exciting us with her note on `` breakthrough in quantum computation '' @xcite ( see fig . [ bose-13.pdf ] ) and to dr .
arnab das for many important comments and suggestions .
we are thankful to dr .
masayuki ohzeki for his comments and suggestions and to dr .
diego de falco for sending us some useful documents .
99 p. ray , b. k. chakrabarti and a. chakrabarti , _ sherrington - kirkpatrick model in a transverse field : absence of replica symmetry breaking due to quantum fluctuations _
, physical review b * 39 * , 11828 ( 1989 ) .
b. k. chakrabarti , a. dutta and p. sen , _ quantum ising phases and transitions in transverse ising models _ , ( see chapter 6 on transverse ising spin glass and random field systems , pp .
118 - 161 ) , springer , heidelberg ( 1996 ) .
a. k. chandra , a. das and b. k. chakrabarti ( eds . ) , _ quantum quenching , annealing and computation _ , springer , heidelberg ( 2010 ) . m. w. johnson , m. w. johnson , m. h. s. amin , s. gildert , t .
lanting , f .
hamze , n. dickson , r .
harris , a. j. berkley , j. johansson , p. bunyk , e. m. chapple , c. enderud , j. p. hilton , k. karimi , e. ladizinsky , n. ladizinsky , t. oh , i. perminov , c. rich , m. c. thom , e. tolkacheva , c. j. s. truncik , s. uchaikin , j. wang , b. wilson and g. rose , _
quantum annealing with manufactured spins _ ,
nature * 473 * , 194 ( 2011 ) . v. n. smelyanskiy , e. g. rieffel , s. i. knysh , c. p. williams , m. w. johnson , m. c. thom , and k. l. p.
w. g. macready , _ a near - term quantum computing approach for hard computational problems in space exploration _ , arxiv:1204.2821 ( 2012 ) .
v. bapst , l. foini , f. krzakala , g. semerjian and f. zamponi , _ the quantum adiabatic algorithm applied to random optimization problems : the quantum spin glass perspective _ , physics reports * 523 * , 127 ( 2013 ) .
a. dutta , g. aeppli , b. k. chakrabarti , u. divakaran , t. rosenbaum and d. sen , _ quantum phase transitions in transverse field models : from statical physics to quantum information _ , cambridge university press ( forthcoming book , 2014 ) . | major breakthrough in quantum computation has recently been achieved using quantum annealing to develop analog quantum computers instead of gate based computers . after a short introduction to quantum computation
, we retrace very briefly the history of these developments and discuss the indian researches in this connection and provide some interesting documents ( in the figs . ) obtained from a chosen set of high impact papers ( and also some recent news etc .
blogs appearing in the internet ) .
this note is also designed to supplement an earlier note by bose ( science and culture , * 79 * , pp .
337 - 378 , 2013 ) .
3cm(5cm,-7 cm ) 0.5 cm |
the final state of the gravitational collapse of a massive star is one of the challenges in classical general relativity ( gr ) @xcite .
a significant contribution has been to show that , under reasonable initial conditions , the spacetime describing the collapse process would inevitably admit singularities @xcite .
these singularities , can either be hidden behind an event horizon or visible to distant observers . in the former , a black hole forms as the end product of a continual collapse process , as hypothesized by the cosmic censorship conjecture ( ccc ) @xcite ( see also @xcite for reviews on the conjecture ) .
the latter are classified as naked singularities , whose existence in gr has been established under a variety of specific circumstances and for different models , with matter content of various types , e.g. scalar fields @xcite , perfect fluids @xcite , imperfect fluids @xcite and null strange quark fluids @xcite . the analysis has also been taken to wider gravitational settings , such as @xmath2 theories @xcite , lovelock gravity @xcite ( see also @xcite for some recent reviews ) and hypothesized quantum gravity theories @xcite .
this is an interesting line of research because , the possible discovery of naked singularities may provide us with an opportunity to extract information from physics beyond trans - planckian regimes @xcite , see e.g. ( @xcite for the possibility of observationally detecting naked singularities).[ff ] it is therefore well motivated to consider other realistic gravitational theories whose geometrical attributes ( not present in gr ) may affect the final asymptotic stages of the collapse . more concretely ,
could ingredients mimicking spin effects ( associated with fermions ) potentially influence the final fate of a collapse scenario ?
in fact , if spin effects are explicitly present then gr will no longer be the theory to describe the collapse dynamics . in gr , the energy - momentum couples to the metric . however ,
when the spin of particles is introduced into the framework , it is expected to couple to a geometrical quantity related to the rotational degrees of freedom in the spacetime .
this point of view suggests a spacetime which is non - riemannian , namely generalizations of gr induced from the explicit presence of matter with such spin degrees of freedom @xcite .
one such framework , which will allow non - trivial dynamical consequences to be extracted is the einstein - cartan ( ec ) theory @xcite where the metric and torsion determine the geometrical structure of spacetime .
the torsion can be interpreted as caused by microscopic effects , e.g. , by fermionic fields which are not taken explicitly into account @xcite . within the context of ec theories
, it has been shown that considering the induced repulsive effects extracted from ( averaged ) spin interactions , the big - bang singularity can be replaced by a non - singular bounce , before which the universe was contracting and halts at a minimum but finite radius @xcite . however
, a curvature singularity as the final fate of a gravitational collapse process can still occur even if explicit spin - torsion and spin - spin repulsive interactions @xcite are taken into account .
the argument that has been put forward is that since photons neither produce nor interact with the spacetime torsion , the causal structure of an ec manifold , determined by light signals , is the same as in gr .
hence the singularity theorems in gr can be generalized to the ec theory by taking into account a combined energy - momentum tensor which would include , by means of some suitable averaging procedure , spin contributions @xcite .
the results conveyed within this paper are twofold .
we consider a spherically symmetric configuration in the presence of a negative cosmological constant @xcite whose matter content is assumed to be a homogeneous weyssenhoff fluid @xcite that collapses under its own gravity . on the one hand , the first class of our solutions is shown to evolve towards a spacetime singularity where the role of the negative cosmological constant is to set up the gravitational attraction through a positive pressure term .
then , as the collapse proceeds , a repulsive pressure computed from averaged spin - spin and spin - torsion interactions , balances the inward pressure , preventing trapped surfaces from forming in the later stages . thus , the resulting singularity could be at least locally naked .
moreover , it is pertinent to point out that our analysis shows that , depending on the spin and energy density parameters , trapped surfaces can either be formed or avoided . on the other hand ,
second class of solutions suggest that the spin contributions to the field equations may generate a bounce that averts the formation of a spacetime singularity .
let us furthermore note that , in contrast to some alternative theories of gravity e.g. , the gauss - bonnet theory in which the misner - sharp energy is modified @xcite , our approach will involve only the manipulation of the matter content .
hence the misner - sharp energy which is the key factor that determines the dynamics of the apparent horizon is defined in the same manner as that in gr @xcite .
the organization of this paper is then as follows . in section ( [ eqm ] )
we present a brief review on the background field equations of the ec theory in the presence of a weyssenhoff fluid and a negative cosmological constant .
section ( [ sol ] ) provides a family of solutions , some of which represent a collapse scenario that leads to a spacetime singularity within a finite amount of time . in section ( [ dah ] )
we study the dynamics of apparent horizon and induced spin effects on the formation of trapped surfaces and show that trapped surface avoidance can occur for a subset of collapse settings .
we examine the curvature strength of the naked singularity in [ strength ] and show that the singularity is gravitationally strong in the sense of tipler @xcite . a second class of solutions exhibiting a bounce is presented in ( [ nss ] ) where we show how the presence of a spin fluid could affect the dynamics of the apparent horizon . in section ( [ exterior ] )
we present a suitable solution for an exterior region and discuss therein the matching between interior and exterior regions .
finally , conclusions are drawn in section ( [ con ] ) .
the action for the ec theory can be written in the form @xcite s = d^4x \{(+2)+_m } = d^4x \{+l_m } , [ action ] where @xmath3 is the ricci scalar constructed from the general affine connection @xmath4 and can be expressed , in general , as a function of independent background fields , i.e. , the metric field @xmath5 and the affine connection .
the quantity @xmath6 is the contorsion tensor defined as [ contortion ] k^ _ = ( t^ _ -t _ ^ - t _ ^ ) , with the spacetime torsion tensor @xmath7 being geometrically defined as the antisymmetric part of the general affine connection [ teq ] t^ _ = ^ _ -^ _ . and @xmath8 is the matter lagrangian ; @xmath9 is the cosmological constant .
we take the metric signature and @xmath9 throughout the equations but in plotting the diagrams we set the units so that @xmath10 and @xmath11 . ] as @xmath12 and @xmath13 .
the presence of torsion in the macroscopic structure of spacetime can , theoretically , be attributed to microscopic fermionic matter fields with spin - angular momentum degrees of freedom . in this paper
, we take the matter part of the action to be described by a weyssenhoff fluid @xcite which macroscopically is a continuous medium but also conveys features that can be suitably associated with the ( averaged ) spinor degrees of freedom of microscopic matter fields .
moreover , it has been shown that with the assumption of the frenkel condition ( also known as the weyssenhoff condition ) , the setup may be equivalently described by means of an effective fluid in a plain gr setting where the effective energy momentum tensor contains additional ( spin induced ) terms @xcite .
more concretely , the lagrangian for the matter content can be subsequently decomposed as _
the lagrangian @xmath14 contains the induced effects of a spinning fluid which can be written in terms of a perfect fluid contribution and a characteristic spin part @xcited . ] .
@xmath15 conveys a minimal coupling of a spinor axial current with torsion @xcite can be associated to a chiral interaction that corresponds to the coupling of contorsion to the massless fermion fields due to a massless dirac lagrangian in a curved background . ] .
therefore , we may write [ axial ] _ ac = j^s _ , where @xmath16 and @xmath17 correspond to the ( averaged ) spinor axial current and axial torsion vector , respectively is defined by @xmath18=2g^{\mu\nu}$ ] , @xmath19 is a chiral dirac matrix , @xmath20 is a fermion field , @xmath21 is the conjugate fermion field and @xmath22 is the totally antisymmetric levi - civita tensor . ] .
varying the action with respect to the contorsion together with using ( [ contortion ] ) gives @xcite [ constraint ] t^+2g^=^2 ^ , where @xmath23 and @xmath24 is the spin angular momentum tensor given by ^:=:=^_ac+^_sf .
moreover , considering the decomposition of the spin angular momentum as @xmath25 @xcite , where @xmath26 is the fluid 4-velocity , @xmath27 is the antisymmetric spin density tensor representing the effective source of torsion . in the ec theory , in contrast to the metric
, the torsion is not really a dynamical field ; the left hand side of equation ( [ constraint ] ) contains no derivatives of the torsion tensor and indeed appears as a purely algebraic equation .
torsion can therefore be eliminated by replacing it with the spin density @xmath28 and hence implying a modification to the einstein field equations . using the frenkel condition , @xmath29 , the torsion constraint equation ( [ constraint ] )
may be rewritten in the form [ tconstraint ] t^=-^2(2^j_-s^u^ ) . at this point
it is useful to show how the weyssenhoff fluid can fit the frenkel condition .
more precisely , this condition results in an algebraic relationship between the spin density tensor and torsion as t_=t^ _
= u^s _ , which can also be retrieved directly from the formalism proposed in @xcite .
therefore , by virtue of the frenkel condition , this means that the only remaining degrees of freedom of the torsion are the traceless components of the torsion tensor .
furthermore , the axial torsion vector can be written , with the assistance of equations ( [ contortion ] ) and ( [ tconstraint ] ) , as [ axialvector ] s_=_(2k^)=12 ^ 2 j_+^2_s^u^. it is now straightforward to obtain the dynamical equations of motion , varying the action ( [ action ] ) with respect to the dynamical field @xmath30 which can be written as [ geq ] g_&-&g_-k^ _
k^ _ - t^ _
-t^ _ t _
^ -t_t_^ & + & g_(4k^ _ k^ _ + 2t^t_-t^t _ ) = t_. substituting for the contorsion from equation ( [ contortion ] ) into the above equation and using equation ( [ tconstraint ] ) , we subsequently get [ eom1 ] g_-g_&-&^4=(t_^sf+t_^ac ) , where @xmath31 .
the energy momentum tensor contains two contributions from the axial current , @xmath32 , and the spin fluid , @xmath33 , which can be expressed=0 , but for the spin squared terms @xmath34 . ] after employing a suitable spin averaging as t^ac_&=&-8 ^ 2
g _ j^2 - 4 ^ 2 j_j_,[tac ] + t^sf_&=&-u_u_^2+g_^2 + , [ tsf ] where we have replaced the various spin - averaged quantities with @xcite s_s^&=&2 ^ 2 , + s_^ s_&=&(g_-u_u_)^2 , + s _ j^&=&0 . from a macroscopic point of view , the spin fluid can therefore be considered as a contribution from a conventional perfect fluid with the associated energy density @xmath35 and pressure @xmath36 plus the first two terms in equation ( [ tsf ] ) , which represent characteristic spin contributions and arise from a suitably averaged microscopic treatment of the fluid . inserting equations ( [ tac ] ) and ( [ tsf ] ) into ( [ eom1 ] ) , together with the above spin averaging
, we finally obtain the dynamical field equations as @xcite [ fes ] g_-g_&=&^4 + .
in this section , we will find a class of collapse solutions which lead to the formation of a spacetime singularity . if the spacetime is assumed to have fewer symmetries ( that is , inhomogeneities or anisotropies ) , there is a paucity of physically reasonable exact solutions available owing to the intrinsic difficulties .
we therefore restrict the discussion to a homogeneous and isotropic interior line element , representing the flrw geometry @xcite [ metric ] ds^2=dt^2-a^2(t)dr^2-r^2(t , r)d^2 , where @xmath37 is the physical radius of the collapsing matter , with @xmath38 being the scale factor and @xmath39 the standard line element on the unit 2-sphere . the field equations for the above metric read 3h^2-&=&^4+_sf(t)_eff(t),[fes00 ] + -2 - 3h^2+&=&^4+p_sf(t)p_eff(t),[fes11 ] where @xmath40 is the rate of collapse .
since we are interested in a continual collapse process , @xmath41 must be negative .
notice that we may consider the cosmological constant term as vacuum energy density @xcite .
the continuity equation for the matter fluid is therefore [ ce ] _
sf+3h(_sf+p_sf)=^2 . as we mentioned in section [ eqm ] ,
the geometry and matter content effectively induce a macroscopic perfect fluid contribution with the barotropic equation of state @xmath42 , together with intrinsic spin contributions that are present in the form of averaged quadratic terms which may admit a possible microscopic representation as , e.g. , unpolarized fermions ] .
it is plausible to assume that the fermions participating in the collapse process behave as ultra - relativistic particles .
thus the number density ( @xmath43 ) of a fermionic gas , satisfying the fermi - dirac distribution , can be approximated by @xmath44 , see @xcite for more details .
so the squares of the spin density and the axial current , which are proportional to @xmath45 @xcite , depend on the scale factor as and @xmath46 are the source parameters for the squared spin density and axial current respectively , and should not be confused with their initial values defined as @xmath47 and @xmath48 . ]
[ js ] j^2=j_0 ^ 2 a^-6 , ^2=_0 ^ 2 a^-6 . however , we proceed with a general setup @xcite where @xmath49 and @xmath50 ( @xmath51 ) .
therefore , from ( [ ce ] ) it is easy to obtain is discussed separately at the end of subsection [ nss ] , because it corresponds to a non - singular case . ]
the energy density [ rhonew ] _
sf(a)&=&c_0 a^-3(1+w)+ a^n , where [ c0al ] c_0&=&(_i_sf-)a_i^3(1+w),&= & , and we take @xmath52 . solving equations ( [ fes00 ] ) and ( [ fes11 ] ) with the use of ( [ rhonew ] ) leads to ( ) ^2&=&a^n+c_0 a^-3(1+w)+,[hh ] + 2&=&(n+2)a^n - c_0(1 + 3w)a^-3(1+w)+,[addot]where @xmath53 and @xmath54 are the initial values of the energy density and scale factor respectively , and [ ell ] = ^4 . from equation ( [ hh ] ) , in order to calculate the singularity time we need to evaluate the following integral [ int ] t_i - t=_a_i^a(t ) = _ a_i^a(t)^-a^- .
the initial physical radius of the collapsing volume , @xmath55 , can be chosen so that the scale factor starts at @xmath56 and as the collapse proceeds the scale factor decreases ( @xmath57 ) .
therefore , following theorem 3.1 in @xcite , if we take the interval ( @xmath58 ) to be sufficiently small we can use the binomial expansion to evaluate the term under the square root .
we can then write the integrand in the region @xmath59 and @xmath60 as [ bio ] ^-
= 1+_k=1^^k , leading to [ intbio ] ( t_i - t)=-+_k=1^_a_i^a(t)da a^-f(a ) , where ( a)=^k .
next we proceed to expand @xmath61 which reads [ exsqb ] f(a)=_j=0^k()^ja^-(nk+3j(1+w ) ) . substituting ( [ exsqb ] ) into ( [ intbio ] ) and performing the integration we finally get [ finalsol ] ( t_i - t)&=&-&-&_k=1^_j=0^k()^j\{}. now , if we retain only the terms with @xmath62 in the double summation in ( [ finalsol ] ) we find [ keep ] \{_1+_2+_3 } = -(t - t_i ) , where [ ceff ] a=18^n(n+2(1+w ) ) , _
1=nc_0 ^ 2 , _
2=-36(n+2(1+w ) ) , _ 3=2(n+2(1+w ) ) . as the time of the singularity
is approached ( @xmath63 ) the scale factor must vanish , which is guaranteed if @xmath59 and @xmath64 .
the latter condition , @xmath60 , is sufficient to ensure that @xmath64 .
the time at which the singularity forms is therefore [ ts ] t_st_i+ .
we are now in a position to examine whether the singularity occurring in the collapse setting presented in the previous section is hidden behind a horizon or is visible to external observers .
the singularity is covered within an event horizon if trapped surfaces emerge early enough before the singularity formation and may be visible if the apparent horizon , which is the outermost boundary of trapped surfaces , fails to form or is delayed during the collapse process .
the key factor that determines the dynamics of the apparent horizon is the misner - sharp energy @xcite which describes the mass enclosed within the shell labeled by @xmath65 at the time @xmath66 , and is defined as @xcite [ msmass ] m(t , r)&=&&=&. it is worth mentioning that in our study the effect of the torsion is to add extra spin - dependent terms to the energy momentum tensor , which in turn react on the spacetime geometry .
it thus affects the dynamics of the apparent horizon .
the spacetime is said to be trapped , untrapped and marginally trapped if , respectively > 1 , < 1 , = 1 . the field equations ( [ fes00 ] ) and ( [ fes11 ] ) can then be rewritten as @xcite [ rewritten ] & = & _ eff(t)+___total(t),-&=&p_eff(t)+p_p__total(t ) , from equations ( [ msmass ] ) and ( [ hh ] ) we readily get [ 2mr ] = .so provided that the solution lies in the allowed region , that is @xmath60 , @xmath67 and @xmath68 , then if @xmath69 initially , or equivalently [ regularitycon ] < 1 , ( as required by regularity @xcite ) , the ratio will remain less than @xmath70 until the singularity occurs , and a trapped surface will not form . otherwise
( if @xmath71 or @xmath72 ) , the ratio @xmath73 eventually exceeds @xmath70 as the collapse proceeds , so that a trapped surface forms before the singularity and therefore covers it . from the second part of equation ( [ rewritten ] ) , the total pressure can be obtained as [ peff ] p__total = c_0wa(t)^-3(1+w)-(n+3)a^n- .
the initial data of the collapsing configuration can be chosen so that the effective pressure is positive at initial epoch , the moment at which the matter distribution begins to collapse at rest , @xmath74 .
this can be achieved if [ conpospressure ] c_0(1+w)>na_i^n+3(1+w)>0 . having the above condition satisfied
, we see that the first and second terms in the right hand side of ( [ peff ] ) dominate the third one ( for @xmath75 ) and the total pressure becomes negative for @xmath76 and @xmath67 .
we require @xmath77 so that the singularity time is real .
therefore , we can deduce that at later stages of the collapse the failure of trapped surfaces to form is accompanied by a negative pressure @xcite , which is indeed produced due to the fermion condensation . on the other hand , if @xmath71 and @xmath78 , which satisfies the condition for trapped surface formation , the pressure can be initially set to be positive and remains positive up to the final stages of the collapse .
although the pressure is allowed to take on negative values , the collapse process will be physically reasonable if the weak energy condition is preserved throughout the collapse .
the weak energy condition ( wec ) states that the energy density as measured by any local observer is non - negative .
thus , along any non - spacelike vector , the following conditions have to be satisfied [ wecrhop ] _
_ total0 , _ _ total+p__total0 .
we thus have [ wec ] & & _ _
total=+c_0 a(t)^-3(1+w)+3a(t)^n0,&&__total+p__total = c_0(1+w ) a(t)^-3(1+w)-na(t)^n0.the initial data can be suitably chosen so as to make the second and third terms in the first inequality dominate initially over the first negative term so that the inequality holds initially . then as these terms diverge at later times the whole inequality remains valid .
therefore it is enough to satisfy [ weeek ] + c_0 a_i^-3(1+w)+3a_i^n0 .
the second inequality is automatically satisfied since @xmath79 , @xmath77 and @xmath59 .
therefore , the initial set up which subsequently guarantees the validity of the weak energy condition has to satisfy ( [ weeek ] ) .
we note that the second inequality in ( [ wecrhop ] ) implies the validity of the null energy condition .
the quantities related to spin source parameters , @xmath80 and @xmath46 , together with those related to initial values of energy density and scale factor , the barotropic index , @xmath81 and the exponent of divergence of spin densities , @xmath82 , all feature in a six - dimensional space of free parameters . in view of equation ( [ 2mr ] )
, the quantities @xmath82 and @xmath81 determine the formation of trapped surfaces in the dynamical evolution of the collapse .
therefore , any point chosen from the four - dimensional sub - space ( @xmath83 ) constructed by fixing values of @xmath82 and @xmath81 represents a collapse scenario that may either end in a black hole or a naked singularity . however , not all the points in this sub - space are suitable . for
the sake of physical reasonableness the initial configuration should satisfy the regularity condition , ( [ regularitycon ] ) , the weak energy condition ( [ weeek ] ) and the positivity of total initial pressure ( [ conpospressure ] ) .
that is , the collapse solution starting from the four - dimensional space mentioned above should respect these conditions .
recall that we demand @xmath77 and we choose the initial data so that @xmath84 ( the case with @xmath84 and @xmath85 shall be presented as the bouncing solutions ) . to ensure the physical reasonability of the collapse scenario we require @xmath86 .
figure [ fig1 ] presents numerically the two - dimensional sub - space of the allowed region of @xmath82 and @xmath81 parameters for fixed values of @xmath80 , @xmath46 , @xmath53 and @xmath56 .
we note that the regions of the two - dimensional parameter space for the formation or otherwise of trapped surfaces , as simply given by ( [ 2mr ] ) , get more restricted due to the physical reasonableness of the collapse setting . and @xmath81 for trapped ( brown zone ) and untrapped regions ( yellow zone ) for @xmath87 , @xmath88 , @xmath89 and @xmath90 , @xmath91 and @xmath92 . ] finally , referring back to equation ( [ rhonew ] ) , the only case which we have already excluded is the case for which @xmath93 and @xmath94 .
one can treat this situation by solving equations ( [ fes00]-[ce ] ) for these values to find _
sf(a)&=&_i_sf()^3+(_0 ^ 2 - 48j_0 ^ 2 ) ( ) , + h^2&=&+(_0 ^ 2 - 48 j_0 ^ 2)(a)+[newh ] , where c_2=-(_0 ^ 2 + 48j_0 ^ 2)-(_0 ^ 2 - 48j_0 ^ 2)(a_i ) + a_i^3_i . looking at the form of equation ( [ newh ] ) , it obviously results in the formation of trapped surfaces as the scale factor approaches zero .
taking all discussions about the physical reasonableness and energy conditions in this section as granted , demanding a physical solution will impose the condition @xmath95 on the spin properties of the matter .
in other words , this solution corresponds to a four - dimensional subspace , ( @xmath93 , @xmath94 ) , of the six - dimensional parameter space of the problem which is also constrained by the condition @xmath95 . more interestingly , for the case where @xmath96 corresponding to a macroscopic dust fluid
, the collapse ends in a black hole with the exterior solution , after a suitable spacetime matching , as that of the schwarzchild - anti - de sitter spacetime , see section [ exterior ] for more details . in order to make our discussion in the previous subsection more concrete
we need to investigate the curvature strength of the naked singularity which is an important aspect of its physical nature and geometrical importance .
the main underlying idea is to examine the rate of curvature growth along non - spacelike geodesics ending at the singularity , in the limit of an observer approaching it .
the singularity is said to be gravitationally strong in the sense of tipler @xcite if every collapsing volume element is crushed to zero size at the singularity , otherwise it is known as weak .
it is widely believed that when there is a strong curvature singularity forming , the spacetime can not be extended through it and is geodesically incomplete . while if the singularity is gravitationally weak it may be possible to extend the spacetime through it classically .
in order that the singularity be gravitationally strong there must exist at least one non - spacelike geodesic with tangent vector @xmath97 , along which the following condition holds in the limit mentioned above [ scc ] = _ 0 ^ 2r_^^>0 , where @xmath98 is the ricci tensor and @xmath99 is an affine parameter which vanishes at the singularity .
let us now consider a radial null geodesic with tangent vector @xmath100 that terminates at the singularity at @xmath101 .
we note that since @xmath97 is an affinely parametrized null geodesic , we have [ geo ] ^_=0 , ^_^=0 . from the null condition for @xmath97 , with the help of the spacetime metric ( [ metric ] )
, we obtain [ spm ] = = a(t ) , while the geodesic equation results in the following differential equations [ geq ] ^t^t+a(^r)^2=0 , + ^t^r+2^t^r=0 , which give the vector field tangent to the null geodesics as [ tvf ] ^=(a^-1,a^-2,0,0 ) .
next we proceed to check the quantity given by ( [ scc ] ) which , with the use of field equation ( [ fes ] ) , reads [ sccrw ] _ 0 ^ 2t_^^ = _ 0 ^ 2=_0 ^ 2 , where use has been made of equations ( [ spm ] ) and ( [ tvf ] ) and the null energy condition ( [ wec ] ) .
next , we proceed by noting that [ note ] a^=ha^-1 , a^=a^-2 .
consequently , we find that [ hop ] _ 0\ { -}. substituting for the rate of collapse from equation ( [ hh ] ) and noting that the terms @xmath102 and @xmath103 go to zero when the scale factor vanishes if @xmath59 , @xmath104 and @xmath60 , we finally get in the limit of approach to the singularity [ psi ] > 0 .
therefore , the strong curvature condition is fulfilled along the singular null geodesics and the naked singularity is gravitationally strong in the sense of @xcite .
in subsection [ ss ] we studied the solutions to the field equations that exhibit the formation of spacetime singularity .
however , it is expected that the spin effects , which become more important in the very late stages of the collapse procedure , oppose the pull of gravity to balance it . in such a scenario ,
the collapse changes to expansion at a turn - around point of the scale factor leading to the singularity removal .
such a class of solutions can be found by setting @xmath85 in equation ( [ hh ] ) .
let us consider the process of collapse that begins at an initial epoch @xmath105 with the initial value of the scale factor @xmath56 . as the collapse enters the small scale factor regime at a time , say , @xmath106 with @xmath107 , the third term in equation ( [ hh ] ) is negligible and we can write ( we consider here the dust case with @xmath108 ) [ ssfr ] ^2a(t)^-4+a(t)^-1 , for which the solution reads [ sb ] a(t)=^ , where [ b ] b=^. the above solution exhibits a bounce occurring at the finite time @xmath109 where the collapse halts ( @xmath110 ) at a minimum value of the scale factor given by [ amin ] a_min=()^=^. it is worth noting that the bouncing solutions obtained here stand for @xmath85 and @xmath84 or equivalently @xmath111 and @xmath112 , in contrast to the singular ones . next , we proceed with investigating the dynamics of the apparent horizon , that its radius at each instance of time is given by the condition @xmath113 or correspondingly [ sfah ] r_ah(t)=^- .
it can now be easily checked that the apparent horizon curve has a minimum for [ astar ] a_=^ , whence we can find the minimum radius @xmath114 so that if the boundary of the collapsing volume is chosen as @xmath115 , then no horizon would form during the collapsing and expanding regimes .
correspondingly , from the first part of equation ( [ rewritten ] ) we can define a threshold mass , setting @xmath116 [ thrm ] m_=m(a_,r_min)= , in such a way that if @xmath117 then the formation of apparent horizon is avoided . in figure [ sftb ]
, we present numerically the trajectory of the scale factor featuring the occurrence of a bounce , for different values of the space parameters .
as it is shown in the left panel of figure [ sftb ] for @xmath118 , the collapse progresses till the singularity formation ( dashed curve ) while for @xmath85 ( solid curve ) the collapsing matter bounces back at a finite value of the scale factor .
the dotted curve represents a case , with @xmath77 case in which the singularity happens sooner than the case where the effects are totally excluded ( @xmath118 ) .
the right panel emphasizes the role of spin contributions in the time behavior of the scale factor .
for larger values of @xmath82 , it takes longer for spin effects to become strong enough to prevent the collapse , which consequently happens at smaller radii .
it is also seen , from equation ( [ amin ] ) that for a fixed value of @xmath82 the larger the initial energy density the smaller the minimum value of the scale factor at which the bounce occurs . , @xmath108 , @xmath93 and @xmath119 , @xmath120 , @xmath121 ( solid curve ) , @xmath122 , @xmath123 ( dashed curve ) and @xmath124 and @xmath125 ( dotted curve ) .
the first and second inflection points occur at the times @xmath126 and @xmath127 , respectively .
the scale factor at this point takes the value @xmath128 .
the bounce occurs at the time @xmath129 for which the scale factor takes the value @xmath130 .
right panel : time behavior of the scale factor for different values of spin density divergence parameter.,title="fig : " ] , @xmath108 , @xmath93 and @xmath119 , @xmath120 , @xmath121 ( solid curve ) , @xmath122 , @xmath123 ( dashed curve ) and @xmath124 and @xmath125 ( dotted curve ) .
the first and second inflection points occur at the times @xmath126 and @xmath127 , respectively .
the scale factor at this point takes the value @xmath128 .
the bounce occurs at the time @xmath129 for which the scale factor takes the value @xmath130 .
right panel : time behavior of the scale factor for different values of spin density divergence parameter.,title="fig : " ] the left panel in figure [ adadd ] further illustrates that the dynamical evolution can be divided to four regimes , two of which during the contracting and the rest during the expanding phases .
the matter volume begins to collapse from rest ( @xmath74 and @xmath131 ) , immediately entering an accelerated contracting phase , called regime , until the small scale factor regime is reached where a decelerated contracting phase starts as a result of spin domination , that is a regime .
the point at which transition between these two phases occurs corresponds to the first inflection point where @xmath132 and the collapse velocity reaches its maximum value @xmath133 .
the collapse then ceases to proceed at the bounce where @xmath110 and the scale factor approaches its minimum value @xmath134 . at this moment
, the acceleration reaches its absolute maximum value @xmath135 .
the early stages of the post - bounce evolution are controlled by an inflationary expanding phase ( or a regime ( of expansion ) ) until the small scale factor regime ends where the acceleration curve reaches its second inflection point ( @xmath136 ) and the velocity reaches its maximum with the same absolute value as in the contracting phase ( but actually positive ) . afterward
a decelerated expanding phase governs the scenario ( an expansionary regime ) . in the right panel of figure [ adadd ] , we plot the hamiltonian constraint ( [ hh ] ) throughout the dynamical evolution of the collapsing object as governed by equation ( [ addot ] ) . we see that this constraint is numerically satisfied with the accuracy of the order of @xmath137 or less . , @xmath121 , @xmath108 , @xmath93 , @xmath119 and @xmath120 .
right panel : the hamiltonian constraint during the whole evolution of the system.,title="fig : " ] , @xmath121 , @xmath108 , @xmath93 , @xmath119 and @xmath120 .
right panel : the hamiltonian constraint during the whole evolution of the system.,title="fig : " ] the left panel in figure [ rhowec ] shows the behavior of total energy density as a function of time . at the beginning of the collapse , the spin contribution is insignificant , while as the collapse advances , the gravitational attraction succumbs to the spin density correction term which comes into play and behaves as a negative energy density . , @xmath121 , @xmath108 , @xmath93 and @xmath119 , @xmath120 ( solid curve ) , @xmath122 ( dashed curve ) and @xmath124 and @xmath125 ( dotted curve ) .
right panel : time behavior of the kretschmann scalar for the same parameters as above.,title="fig : " ] , @xmath121 , @xmath108 , @xmath93 and @xmath119 , @xmath120 ( solid curve ) , @xmath122 ( dashed curve ) and @xmath124 and @xmath125 ( dotted curve ) .
right panel : time behavior of the kretschmann scalar for the same parameters as above.,title="fig : " ] the solid curve shows that the total energy density increases up to its first maximum , then it decreases suddenly , owing to the dominating negative energy density coming from spin correction term , and tends to zero at the bounce time at which @xmath138 .
this tells us that the matter content within the collapsing volume becomes incompressible . at the post - bounce regime
the spin density term becomes diluted as a consequence of the inflationary expanding phase causing the total energy density to increase up to its second maximum and then falls off to finite values .
thus , the total energy density is finite for @xmath85 during the dynamical evolution of the collapse scenario while the dashed ( @xmath118 ) and dotted ( @xmath77 ) curves signal the occurrence of a spacetime singularity where the kretschmann scalar ( right panel ) and energy density diverge .
the left panel in figure [ pwec ] shows the behavior of total pressure during the whole dynamical evolution of the collapse .
the solid curve ( @xmath85 ) shows that the total pressure is positive in the early stages of the collapse where the spin contribution is weak .
as the collapse proceeds the pressure becomes negative and reaches a maximum value in negative direction , where the contracting phase turns to an expanding one .
it is the appearance of such a negative pressure , as produced by a spin correction term , which causes the bounce . whereas the dashed curve ( @xmath77 ) shows that the pressure begins from a positive value and remains positive up to the singularity formation .
the right panel shows that the weak energy condition is satisfied in the absence of spin effects ( dashed curve ) and also for the case in which @xmath77 . for @xmath85
, wec holds in the weak field regime while it is violated in the spin dominated regime .
such a violation of wec can be compared to the models in which the effects of quantum gravity are taken into account @xcite . ,
@xmath121 , @xmath108 , @xmath93 and @xmath119 , @xmath120 ( solid curve ) , @xmath124 and @xmath125 ( dashed curve ) .
right panel : the weak energy condition for the same parameters as above.,title="fig : " ] , @xmath121 , @xmath108 , @xmath93 and @xmath119 , @xmath120 ( solid curve ) , @xmath124 and @xmath125 ( dashed curve ) .
right panel : the weak energy condition for the same parameters as above.,title="fig : " ] we also need to check if a dynamical horizon is formed during the whole contracting and expanding phases .
firstly , as we stated in subsection [ ss ] , the regularity condition has to be respected at the time at which the collapse commences .
let us define the maximum radius @xmath139 in such a way that if @xmath140 , then the regularity condition would be violated .
therefore , if the boundary is chosen so that @xmath141 , four horizons can form .
the left panel in figure [ abra ] shows the behavior of apparent horizon curve ( [ sfah ] ) as compared to the case in which spin effects are absent .
as the solid curve shows ( @xmath85 ) , the apparent horizon curve decreases for a while in the contracting phase , and increases just before the bounce occurs .
the first horizon appears in the accelerated contracting phase and , after passing through the first inflection point , the second one forms in the decelerated contracting phase . as the collapse process turns to an expanding regime , the apparent horizon decreases again , in the accelerated expanding phase , to meet the boundary for the third time but at the same radius .
after the second inflection point is reached , the scenario is entering a decelerated expanding phase where the fourth horizon intersects the boundary ( see the horizontal dotted line labeled as c ) .
the next possibility for the horizon formation happens if we take @xmath142 . in this case
two horizons could form .
the first one appears at the moment of transition between accelerated and decelerated contracting regimes , i.e. , the first inflection point .
the second one appears at the same radius but at the time at which the accelerated expanding regime transits to the decelerated expanding one , i.e. , the second inflection point ( see the horizontal dotted line labeled as b ) . finally , if we take @xmath115 no horizon would form indicating the existence of a minimum value for the size of the collapsing object so that the formation of the apparent horizon is prevented ( see the horizontal dotted line labeled as a ) . however
, the apparent horizon in the absence of spin effects ( dashed curve ) propagates inward to finally cover the singularity .
there can be found no minimum for the surface boundary of the collapsing matter in order to avoid the formation of the apparent horizon and the collapse scenario inevitably results in black hole formation . , @xmath121 , @xmath108 , @xmath93 and @xmath119 , @xmath120 ( solid curve ) and @xmath122 , @xmath123 ( dashed curve ) .
the horizontal dotted lines show the times at which four horizons form ( c ) two horizons form ( b ) and no horizon forms ( a ) .
the times at which the horizons form at the four phases of evolution are @xmath143 , @xmath144 , @xmath145 and @xmath146 , respectively .
there can be found a minimum value for the size of the collapsing object so that the formation of the apparent horizon is prevented ( solid curve ) .
this situation can not be happened when the apparent horizon radius vanishes at a finite amount of time ( dashed curve ) .
the blue arrows lying on the solid curve display the direction of collapse process and the red arrows show the direction of expansion process .
right panel : the absolute value of the collapse velocity against the scale factor for the same parameters as above .
the horizontal dotted lines intersecting the solid curve show that in case in which the collapse velocity is bounded , four horizons ( c ) , two horizons ( b ) and no horizon ( a ) can be formed during the dynamical evolution of the collapse process .
the absolute maximum value of the collapse velocity at the inflection points has the value @xmath147 .
the blue arrows lying on the solid curve show the direction of collapse regime and the red ones show the direction of expanding regime .
the formation of an apparent horizon is always guaranteed as the dashed curve shows.,title="fig : " ] , @xmath121 , @xmath108 , @xmath93 and @xmath119 , @xmath120 ( solid curve ) and @xmath122 , @xmath123 ( dashed curve ) .
the horizontal dotted lines show the times at which four horizons form ( c ) two horizons form ( b ) and no horizon forms ( a ) .
the times at which the horizons form at the four phases of evolution are @xmath143 , @xmath144 , @xmath145 and @xmath146 , respectively .
there can be found a minimum value for the size of the collapsing object so that the formation of the apparent horizon is prevented ( solid curve ) .
this situation can not be happened when the apparent horizon radius vanishes at a finite amount of time ( dashed curve ) .
the blue arrows lying on the solid curve display the direction of collapse process and the red arrows show the direction of expansion process .
right panel : the absolute value of the collapse velocity against the scale factor for the same parameters as above .
the horizontal dotted lines intersecting the solid curve show that in case in which the collapse velocity is bounded , four horizons ( c ) , two horizons ( b ) and no horizon ( a ) can be formed during the dynamical evolution of the collapse process .
the absolute maximum value of the collapse velocity at the inflection points has the value @xmath147 .
the blue arrows lying on the solid curve show the direction of collapse regime and the red ones show the direction of expanding regime .
the formation of an apparent horizon is always guaranteed as the dashed curve shows.,title="fig : " ] the existence of a minimum value for the size of the collapsing object can be translated as saying that the speed of collapse has to be limited .
as the solid curve in the right panel in figure [ abra ] shows , in the early stages of the collapse , the trajectory of the system in @xmath148 plane follows the dashed curve in which the spin effects are neglected . at later times
, it deviates from this curve to reach the maximum value for the speed of collapse , i.e. , at the first inflection point .
after this time , the collapse progresses with a decreasing speed reaching the minimum value of the scale factor , after which the collapsing cloud turns into an expansion .
the absolute value of the collapse velocity is bounded throughout the contracting and expanding phases . in this sense ,
there can be found a maximum value for the collapse velocity ( as related to a minimum value for the surface boundary or a threshold mass ) so that for @xmath149 the horizon equation is never satisfied ( see the horizontal dotted line labeled as a ) .
however , if @xmath150 , two horizons could still appear , first one at the contracting and the second one at the expanding regimes . both of these horizons form at the same value of the scale factor at inflection points ( see the horizontal dotted line labeled as b )
the third possibility is @xmath151 , for which four horizons could appear at the four phases of dynamical evolution of the scenario ( see the horizontal dotted line labeled as c ) . on the other hand , in contrast to these cases ,
the collapse velocity diverges when the spin effects are absent and the horizon equation is always satisfied , the dashed curve .
finally , regarding the case @xmath152 , equation ( [ ce ] ) leads to the following expression for the energy density _ sf=_i_sf+(a^n - a_i^n ) , whereby from equations ( [ fes00 ] ) and ( [ fes11 ] ) we can solve for the collapse rate as h=^. from the above equation we see that the collapse begins with a contraction phase ( choosing the minus sign ) which proceeds until the scale factor reaches a critical value , @xmath153 , a_c=^ , for which @xmath154 .
this occurs at the time t_c = t_i- , and the kretchmann invariant and the energy density behave regularly , thus no singularity forms and @xmath155 are finite at @xmath156 .
a type iv singularity does not occur either since the higher derivatives of @xmath157 do not diverge at @xmath156 @xcite . ] . to our knowledge
this unprecedented situation is a specific feature of the present model .
it can be interpreted as a stationary state .
also there is a chance that this unorthodox behavior of the fluid will be followed by a transition to an expansion phase corresponding to the situation , the effective energy density of the collapsing object would be negative and , as a result , the weak energy condition will be violated .
the gravitational collapse of regions with negative energy density has been discussed in the literature , mainly in the context of topological black holes @xcite .
it has been claimed that topology changing processes , due to quantum fluctuations of spacetime , would be a possible mechanism for such behavior .
however this discussion is beyond the scope of this paper . ] where @xmath157 jumps from a @xmath158 to a @xmath159 branch @xcite .
the gravitational collapse setting studied so far deals with the interior of the collapsing object .
we found two classes of solutions , where for the singular ones , depending on the spin source parameters and initial energy density the apparent horizon can be avoided .
however , the absence of apparent horizon in the dynamical process of collapse does not necessarily imply that the singularity is naked @xcite .
in fact , the singularity is naked if there exist future pointing null geodesics terminating in the past at the singularity .
these geodesics have to satisfy @xmath160 in the interior space - time so that the area radius must increase along these geodesics .
as discussed in @xcite , this situation can not happen since the singularity occurs at the same time for all collapsing shells .
however , this process , to be completely discussed , may require a suitable matching to an exterior region whose boundary @xmath161 is the surface of the collapsing matter that becomes singular at @xmath162 , into which null geodesics can escape .
employing the junction conditions @xcite , our aim here is to complete the full spacetime geometry presented for the spherically symmetric gravitational collapse via matching the homogeneous interior spacetime to a suitable exterior spacetime .
let us therefore consider a timelike three - dimensional hypersurface @xmath163 resulting from the isometric pasting of two hypersurfaces @xmath164 and @xmath165 , which , respectively , bound the four - dimensional exterior ( @xmath166 ) and interior ( @xmath167 ) spacetimes .
for the interior region we take the line element ( [ metric ] ) in the frw form as [ int ] ds_-^2=dt^2-a^2(t)dr^2-a^2(t)(r^2d^2+r^2 ^ 2d^2 ) , where the interior coordinates are labeled as @xmath168 .
the line element for the exterior region in retarded ( exploding ) null coordinates is taken as [ out ] ds_+^2=f(v , r)dv^2 + 2dvdr - r^2(d^2+^2d^2 ) , where @xmath169 with @xmath170 being the exterior mass function and the exterior coordinates are labeled as @xmath171 .
we assume that @xmath163 is endowed with an intrinsic line element given by [ intrinsicmetric ] ds_^2=d^2-y^2()(d^2+^2d^2 ) . here
@xmath172 , @xmath173 are the coordinates of @xmath163 with @xmath174 being the time coordinate defined on it and we have chosen the angular coordinates @xmath175 and @xmath176 to be continuous .
the governing equations of hypersurface @xmath163 in the coordinates @xmath177 are given by [ eqsigma ] r - r_=0 v^+ , r - r_=0 v^- .
using the above two equations , the interior and exterior induced metrics on @xmath165 and @xmath164 respectively take the form [ inmin ] ds_^-^2=dt^2-a^2(t)r_^2(d^2+^2d^2 ) , and [ inmex ] ds_^+^2=dt^2 -r^2(t)(d^2+^2d^2 ) .
we assume that there is no surface stress - energy or surface tension at the boundary ( see e.g. @xcite for the study of junction conditions for boundary surfaces and surface layers ) .
then the junction conditions require [ fff ] ds_^2=ds_^-^2=ds_^+^2 , which gives [ fff1 ] f(v(t),r(t))^2 + 2=1 , r(t)=r_a(t)=y ( ) , where an overdot denotes @xmath178 .
next , we need to compute the components of the extrinsic curvature of the interior and exterior hypersurfaces .
the unit spacelike normal vector fields to these hypersurfaces are given by [ nvf ] n^-_= , n^+_=. let us take @xmath179 as the parametric equations of @xmath163 .
the extrinsic curvature or second fundamental form of the hypersurface @xmath163 is a three - tensor defined as @xcite [ excdefine ] k_abe_a^e_b^_n _ , where @xmath180 are the basis vectors tangent to the hypersurafce @xmath163 and the covariant derivative is taken with respect to the christoffel symbols @xmath181 .
the above expression can be re - written in the following form [ excrew ] k_ab&&e_a^e_b^ = e_a^e_b^-\{^ _ } e_a^e_b^n _ + & = & ( n_e^_a)e^_b-()n _ -\{^ _ } n _ + & = & -n_-\{^ _ } n _ + & = & -n_. now , if we take @xmath182 and @xmath183 as parametric relations for the hypersurfaces @xmath164 and @xmath165 on the exterior and interior regions , we get , respectively [ exc ] k^_ab =- n_^. here we should pay attention to the second term containing the asymmetric affine connection defined in ( [ teq ] ) . since in ec theory torsion can not propagate outside the spin matter distribution ( it is non - vanishing only inside the matter ) @xcite , the connection for the exterior region is precisely the christoffel symbol whose non - vanishing components are [ chrex ] \{^v _ vv}^+=- , \{^r _ vv}^+= ( f_,v+ff_,r ) , \{^r _ vr}^+= , where @xmath184 , @xmath185 and @xmath186 . in order to calculate the nonzero components of ( [ exc ] ) we proceed by noting that [ notep ] & = & , & = & . substituting the above expressions into equation ( [ exc ] ) and noting that @xmath187 , we finally get [ excex ] k_tt^+=- , k^+_&=&k^+_=. in the process of obtaining the extrinsic curvature of @xmath163 , proceeding from the interior region
, we should note that the connection through which the extrinsic curvature tensor is calculated is no longer symmetric owing to the presence of the spin matter .
therefore , we need to begin with [ acon ] ^- _ = ^- _ ( ) + ^- _ [ ] = \{^_}^-+ k^ _ = \{^ _ } ^-+(t^ _ -t _ ^ -t _ ^ ) .
substituting for the torsion tensor from equation ( [ tconstraint ] ) and rearranging the terms we find [ conspin ] ^- _ = \{^ _ } ^-+ .
a subsequently suitable spacetime averaging reveals that the second term in square brackets vanishes since the connection is linear with respect to the spin density tensor .
however , the first term may not generally become zero since the axial current is a timelike vector field .
substituting the averaged affine connection in the minus sign of ( [ exc ] ) we have [ excaff ] k^-_ab =- n_^- , from which we readily find that the third term in parentheses vanishes due to antisymmetrization property of the levi - civita tensor and partial derivatives .
after a straightforward calculation we find [ excin ] k_tt^-=0 , k^-_=k^-_=. however , due to the presence of the third term in ( [ excaff ] ) there may remain other components of the extrinsic curvature tensor though the spacetime is spherically symmetric .
let us calculate them to show that these terms vanish too .
the @xmath188 component reads [ ttexc ] k^-_t =- a(t)\{+\{^ r _
} ^-+^2 ( _ ^ r + _
^ r ) j_}. since @xmath189 has only a time component , the levi - civita tensor vanishes and we get [ kttheta ] k^-_t = k^-_t=0 .
the remaining components can be calculated in the same way as [ kthphi ] k^-_=k^-_=-^2a(t ) ( _ ^ r t+ _
^ r t)j_t=0 , since both the levi - civita tensors are equal but with opposite signs . using ( [ fff1 ] ) , the continuity of the extrinsic curvatures across @xmath163 implies the following relations f+&=&1,[a15 ] + ^2&+&2(-)=0.[a16 ] taking derivatives of ( [ a15 ] ) and the first part of ( [ fff1 ] ) we have [ a17 ] 2=^2 , 2=-(2+f ) , hence we can construct the following relation as [ rec ] = -(2+f ) .
substituting ( [ rec ] ) and the first part of ( [ a17 ] ) into ( [ a16 ] ) , we finally find [ kttpf ] k_tt^+=-=0 , which clearly shows that @xmath190 must be a function of @xmath191 only . solving equations ( [ a15 ] ) and the first part of ( [ fff1 ] ) we get the four - velocity of the boundary , as seen from an exterior observer [ 4v ] v^=(,,0,0)= , where a minus sign for @xmath192 has been chosen since we are dealing with a collapse setting . from the second component of the above vector field and the interior solution ( [ hh ] ) , we find for a smooth matching of the interior and exterior spacetimes that , @xmath193 . we thus have [ inmoutm ] 2m(r , v)=r^3 + 2m_0r^-3w+2s_0r^3+n , where @xmath194 and @xmath195 . here the new term @xmath196 is treated as a correction introduced by the spin contribution . therefore , the line element for exterior spacetime reads [ exsp ] ds_+^2=dv^2 + 2dvdr - r^2(d^2+^2d^2 ) .
the location of the apparent horizon is marked by requiring that @xmath197 which lies on the boundary surface @xmath161 if @xmath198 , or simply from equation ( [ 2mr ] ) , @xmath199 .
thus , we find that once the collapse velocity satisfies the following equation [ heq ] collapse velocity is bounded the boundary surface can be chosen so that no horizon forms @xcite .
now , from the first part of equation ( [ hh ] ) , we see that for specific values of @xmath82 and @xmath81 , taken from the brown region of figure [ fig1 ] , the collapse velocity tends to infinity .
thus , there is no minimum value for @xmath200 ( or correspondingly a minimum mass for the collapsing volume ) so that the horizon can be avoided .
in contrast , for @xmath82 and @xmath81 taken from the yellow region of figure [ fig1 ] , the speed of collapse stays bounded until the singularity time at which the scale factor vanishes .
this means that to satisfy the horizon condition in the limit of approach to the singularity , the boundary of the volume must be taken at infinity which is physically irrelevant .
thus we can always take the surface boundary so that the apparent horizon is avoided .
therefore , if the collapse velocity is bounded we can take the boundary surface to be sufficiently small so that the formation of horizon is avoided during the entire phase of contraction .
furthermore , the null geodesic that has just escaped from the outermost layer of the mass distribution of the cloud ( @xmath201 ) can be extended to the exterior region exposing the singularity to external observers . for bouncing solutions , as the right panel of figure [ abra ] shows
, @xmath202 remains finite throughout the collapsing and expanding phases , thus by a suitable choice of the boundary surface , the apparent horizon is failed to cover the bounce . for @xmath94 , @xmath93 and @xmath203 the exterior spacetime metric is written as [ schads ] ds_+^2=dv^2 + 2dvdr - r^2(d^2+^2d^2 ) , where @xmath204 and @xmath205 , clearly exhibiting a schwarzchild - anti - de sitter metric in retarded null coordinates @xcite with a constant mass @xmath206 with corrections due to spin contributions .
the study of the end - state of matter gravitationally collapsing becomes quite interesting when averaged spin degrees of freedom and torsion are taken into account . to our knowledge ,
the literature concerning this line of research is somewhat scarce , see e.g. , @xcite .
torsion is perhaps one of the important consequences of coupling gravity to fermions . in general
, this leads to non - riemannian spacetimes where departures from the dynamics of gr would be expected and should be explored .
the well known and established csk @xcite theories can also be a starting point .
nevertheless , the explicit presence of fermionic fields may not provide a simple enough set - up to investigate the final outcome of a gravitational collapse .
there are , however , other , perhaps more manageable scenarios .
they employ torsion just to mimic the effects of matter with spin degrees of freedom on gravitational systems .
it was in that precise context that we have therefore considered the approach presented in this paper .
more precisely , we studied the gravitational collapse of a cloud whose matter content was taken as a weyssenhoff fluid @xcite in the context of the ec theory @xcite , i.e. , with torsion .
a negative @xmath9 was included to provide an initially positive pressure , so that a collapse process could initially be set up .
the torsion is not , however , a dynamical field , allowing it to be eliminated in favor of algebraic expressions .
in addition , we have restricted ourselves to a special but manageable spacetime model where the interior region line element is a flrw metric , allowing a particularly manageable framework to investigate . the corresponding effective energy momentum from a macroscopic perspective
has a perfect fluid contribution plus those induced from averaged spin interactions .
a relevant feature is that this effective matter can , within specific conditions , convey a negative pressure effect . as a consequence
, this may induce the avoidance of the formation of trapped surfaces , from one hand , and the possibility of singularity removal from the other hand . in a compact manner ,
our main results are as follows : * for singular solutions ( @xmath77 ) , the formation or otherwise of trapped surfaces not only depend on the equation of state parameter but also on the spin density divergence term ( @xmath82 ) .
therefore , from determining the initial setting subject to ( @xmath207 ) the regularity condition on the absence of trapped surfaces at an initial epoch , ( @xmath208 ) the validity of the energy conditions and ( @xmath209 ) the positivity of the effective pressure at an initial time , trapped surfaces can either develop ( for @xmath210 ) or be avoided ( @xmath211 ) throughout the collapse . *
a special case in which the equation of state of spin fluid is @xmath212 was considered separately and it was found that no singularity occurs .
this very unorthodox case can be thought of as a stationary state . *
the set of collapse solutions can be categorized through the six - dimensional space of the parameters ( @xmath213 ) so that the first two are related to initial values for spin source parameters ( note that @xmath214 and @xmath215 ) .
the next two parameters are the rate of divergence of spin density and barotropic index and the last two are the initial values of the scale factor and energy density .
each point from this space represents a collapse process that can be either led to a spacetime singularity or a non - singular bounce . determining the suitable ranges for this set of initial data is not straightforward and so , for the sake of clarity , we have to deal with the two - dimensional subspaces by fixing four of the above parameters .
however , we could infer that among the allowed sets of the initial data we can always pick up those for which trapped surfaces are prevented ( in singular solutions ) during the collapse scenario , ( see the regions in figure [ fig1 ] ) , where we have fixed the same initial values for energy density and scale factor . * depending on the initial value of energy density and the source parameters related to spin - spin contact interaction and axial current , singular ( @xmath77 and @xmath84 ) and non - singular ( @xmath85 and @xmath84 ) solutions can be found . in the former
the singularity occurs sooner than the case in which the spin correction term is neglected ( see the left panel in figure [ sftb ] ) . for the non - singular scenario , the collapse process halts at a finite value of the scale factor and then turns to expansion .
it is worth mentioning that besides the model presented here , non - singular scenarios have been reported within models of @xmath2 theories of gravity in palatini @xcite and metric @xcite formalisms , generalized teleparallel theories of gravity @xcite , bouncing in brane models @xcite and modified gauss - bonnet gravity @xcite ( see also @xcite for recent review ) . in @xcite
, it is shown that a quantized neutral scalar field minimally coupled to classical gravitational field may avoid the singularity .
it is also interesting to note that , beside the frenkel condition we employed here , we could consider the possibility of relaxing it , therefore allowing to take a more general matter content .
if such a modification is employed , then the number of degrees of freedom of the torsion tensor would increase , seemingly bringing a more complicated setting to deal with .
finally , we would like to present a few possible additional subsequent lines of exploration .
although being a wider set - up with respect to gr , it could be fruitful to generalize action ( [ action ] ) . more concretely , replacing the cosmological constant by some scalar matter .
this would allow for the establishment of limits for the dominance of any matter component ( and associated intrinsic effects ) towards a concrete gravitational collapse outcome where , for example , bosonic and fermionic matter would be competing . perhaps more challenging would be to employ a weyssenhoff fluid description that could have different features whether we use @xmath216 fermion or a rarita - schwinger field with @xmath217 spin angular momentum .
the gravitational theory of such latter particles in the presence of torsion has been discussed in @xcite .
the authors are grateful to d. malafarina and r. goswami for useful discussions and to f. w. hehl for helpful correspondence .
a. ranjbar also would like to thank f. canfora and j. zanelli for their interesting comments .
the riemann tensor in ec theory reads [ ecri ] ^ _ = _ ^ _ - _ ^ _ + ^ _ ^ _ -^ _ ^ _ . the general affine connection can be written in terms of christoffel symbols and contorsion ( we note that in ec theory @xmath218 ) [ gafc ] ^ _ = \{^ _ } + k^ _ . substituting the above expression into ( [ ecri ] ) we get [ ecri1 ] ^ _ & = & r^ _ ( \{})+_k^ _ -_k^ _ + \{^ _ } k^
_ + \{^ _ } k^ _ + k^ _ k^ _ & -&\{^ _ } k^ _ -\{^ _ } k^ _
-k^ _ k^ _ , whereby contracting twice gives the ricci scalar as [ ricci ] & = & r(\{})+g^_k^ _
-g^_k^ _ + g^\{^ _ } k^ _ + \{^ _ } k^ _
+ k^ _ k^ _ -g^\{^ _ } k^ _ & -&g^\{^ _ } k^ _
-g^k^ _ k^ _ .
the metricity condition ( @xmath219 ) leaves us with the following expressions [ pmet ] _ g^=-g^\{^ _ } -g^\{^ _ } , _ g^=-g^\{^ _ } -g^\{^ _ } , by the virtue of which we can simplify ( [ ricci ] ) to finally get [ riccif ] = r(\{})+_k^ _ -_k^ _ + k^ _ k^ _ -k^ _
k^ _ . performing the integration and neglecting the total derivatives we arrive at the integrand given in the expression ( [ action ] ) .
a. krlak , prog .
. suppl . * 136 * ( 1999 ) 45 , [ arxiv : gr - qc/9910108 ] ; + p. r. brady , i. g. moss and r. c. myers , phys .
* 80 * ( 1998 ) 3432 , [ arxiv : gr - qc/9801032 ] .
r. p. a. c. newman `` _ topological properties and global structure of spacetime _ '' , ed v. de sabbata ( new york : plenum ) ( 1986 ) .
r. penrose , riv .
nuovo cimento * 1 * ( 1969 ) 252 [ gen
. rel . grav .
* 34 * ( 2002 ) 1141 ] ; + r. penrose , in _ general relativity , an einstein centenary survey _ , edited by s. w. hawking and w. israel ( cambridge university press , cambridge , england , 1973 ) ; + r. penrose , in gravitational radiation and gravitational collapse , iau symposium * 64 * , ed . c. dewitt - morettee , reidel , drdrecht ( 1974 ) . c. j. s. clarke , class .
quantum grav .
* 11 * ( 1994 ) 1375 ; r. m. wald , [ arxiv : gr - qc/9710068 ] ; s. jhingan and g. magli , [ arxiv : gr - qc/9903103 ] ; t. p. singh , j. astrophys . astron . *
20 * ( 1999 ) 221 [ arxiv : gr - qc/9805066 ] .
r. giamb , class .
quantum grav .
* 22 * ( 2005 ) 2295 [ arxiv : gr - qc/0501013 ] .
s. bhattacharya , r. goswami and p. s. joshi , int . j. mod
d * 20 * ( 2011 ) 1123 [ arxiv:1010.1757[gr - qc ] ] ; s. bhattacharya , [ arxiv:1107.4112[gr - qc ] ] ; y. tavakoli , j. marto , a. h. ziaie and p. v. moniz , gen .
* 45 * ( 2013 ) 819 [ arxiv:1105.0445[gr - qc ] ] .
t. harada , phys .
d * 58 * ( 1998 ) 104015 [ arxiv : gr - qc/9807038 ] ; t. harada and h. maeda , phys .
d * 63 * ( 2001 ) 084022 [ arxiv : gr - qc/0101064 ] ; r. goswami and p. s. joshi , class . quantum grav . * 19 * ( 2002 ) 5229 [ arxiv : gr - qc/0206086 ] ; r. giamb , f. giannoni , g. magli and p. piccione , gen . rel . grav .
* 36 * ( 2004 )
1279 [ arxiv : gr - qc/0303043 ] ; j. f. villas da rocha and a. wang , class . qaunt . grav .
* 17 * ( 2000 ) 2589 [ arxiv : gr - qc/0007004 ] ; r. giamb , f. giannoni , g. magli and p. piccione , class .
quantum grav . * 20 * ( 2003 ) 4943 .
p. szekeres and v. iyer , phys .
d * 47 * ( 1993 ) 4362 ; s. barve , t. p. singh , and l. witten , gen .
* 32 * ( 2000 ) 697 [ arxiv : gr - qc/9901080 ] ; a. a. coley and b. o. j. tupper , phys .
d * 29 * ( 1984 ) 2701 ; k. lake , phys .
d * 26 * ( 1982 ) 518 .
p. s. joshi , pramana * 55 * ( 2000 ) 529 [ arxiv : gr - qc/0006101 ] ; p. s. joshi and d. malafarina , int . j. mod
d * 20 * ( 2011 ) 2641 [ arxiv:1201.3660 [ gr - qc ] ] ; m. celerier , and p. szekeres , phys .
d * 65 * ( 2002 ) 123516 ; r. giamb , f. giannoni , g. magli and p. piccione , commun .
* 235 * ( 2003 ) 546 [ arxiv : gr - qc/0204030 ] ; t. harada , h. iguchi and k. nakao , prog .
phys * 107 * ( 2002 ) 449 [ arxiv : gr - qc/0204008 ] ; r. penrose _ black holes and relativistic stars _ , edited by r. m. wald ( chicago , il : university of chicago press 1998 ) .
v. p. frolov and g. a. vilkovisky , phys .
b * 106 * ( 1981 ) 307 ; p. singh and a. toporensky , phys .
d * 69 * 104008 ( 2004 ) ; m. bojowald , r. goswami , r. maartens , and p. singh , phys . rev . lett .
* 95 * ( 2005 ) 091302 ; v. husain and o. winkler , class .
quantum grav .
* 22 * ( 2005 ) l127 .
goswami , p. s. joshi and p. singh , phys .
* 96 * ( 2006 ) 031302 ; e. greenwood , d. stojkovic , jhep 0806 042 ( 2008 ) ; m. bojowald and g. m. paily , class .
quantum grav .
* 29 * ( 2012 ) 242002 ; a. kreienbuehl , t. pawlowski , phys .
d * 88 * ( 2013 ) 043504 ; y. tavakoli , j. marto , and a. dapor , arxiv:1303.6157 ; c. bambi , d. malafarina and l. modesto , eur .
phys . j. c * 74 * ( 2014 ) 2767 ; c. rovelli and f. vidotto , arxiv:1401.6562 [ gr - qc ] ; y. zhang , y. zhu , l. modesto and c. bambi , arxiv:1404.4770 [ gr - qc ] ; y. liu , d. malafarina , l. modesto and c. bambi , arxiv:1405.7249 [ gr - qc ] .
m. bojowald , phys .
lett , * 86 * ( 2001 ) 5227 ; m. bojowald , r. maartens and p. singh , phys .
d * 70 * ( 2004 ) 083517 ; p. singh , k. vandersloot , phys .
d * 72 * ( 2005 ) 084004 ; d.h .
coule , class .
quantum grav .
* 22 * ( 2005 ) r125 ; p. singh , k. vandersloot , g. v. vereshchagin , phys .
d * 74 * ( 2006 ) 043510 ; a. ashtekar , t. pawlowski , p. singh , k. vandersloot , phys
d * 75 * ( 2007 ) 024035 ; t. cailleteau , a. cardoso , k. vandersloot and d. wands , phys .
* 101 * ( 2008 ) 251302;t cailleteau , p. singh , k. vandersloot , phys .
d * 80 * ( 2009 ) 124013 ; j. amors , j. de haro and s. d. odintsov , arxiv:1402.3071 [ gr - qc ] ; a. ashtekar , d. sloan , phys .
b * 694 * ( 2010 ) 108 ; b. gupt and p. singh , class .
quantum grav .
* 30 * ( 2013 ) 145013 ; phys .
d * 86 * ( 2012 ) 024034 ; e. ranke and p. singh , phys .
d * 85 * ( 2012 ) 104002 ; e. wilson - ewing , jcap 1303:026 ( 2013 ) ; m. artymowski , a. dapor , t. pawlowski , arxiv:1207.4353 [ gr - qc ] .
p. s. joshi , d. malafarina , int . j. mod .
d , * 20 * 2641 ( 2011 ) ; p. s. joshi , d. malafarina , and r. narayan , class .
quantum grav .
* 28 * 235018 ( 2011 ) ; k. s. virbhadra , d. narasimha , and s. m. chitre , astron . astrophys .
* 337 * 1 ( 1998 ) ; k. s. virbhadra , and c. r. keeton , phys .
d * 77 * 124014 ( 2008 ) ; k. s. virbhadra , and g. f. r. ellis , phys .
d * 65 * 103004 ( 2002 ) ; phys .
d * 62 * 084003 ( 2000 ) ; c - m claudel , k. s. virbhadra , and g. f. r. ellis , j. math .
* 42 * 818 ( 2001 ) .
m. gasperini , gen .
. grav . * 30 * ( 1998 ) 1703 ; j. tafel , phys .
a * 45 * 341 ( 1973 ) ; g. d. kerlick , annals of physics * 99 * 127 ( 1976 ) ; n. j. poplawski , phys .
d * 85 * ( 2012 ) 107502 ; n. j. poplawski , gen .
* 44 * ( 2012 ) 1007 ; a trautman , nature , ( phys .
sci . ) , * 242 * ( 1973 ) 7 ; w. kopczynski , phys .
a * 39 * ( 1972 ) 219 ; phys .
a * 43 * ( 1973 ) 63 ; a. v. minkevich , phys .
a , * 80 * ( 1980 ) 232 ; m. demianski , r. de ritis , g. platania , p. scudellaro and c. stornaiolo , * 116 * ( 1986 ) 13 ; c. wolf , gen .
* 27 * ( 1995 ) 1031 ; b. vakili , s. jalalzadeh , phys .
b * 726 * 28 ( 2013 ) ; m. hashemi , s. jalalzadeh and a. h. ziaie , arxiv:1407.4103 [ gr - qc ] .
s. s. deshingkar , s. jhingan , a. chamorro , and p. s. joshi , phys .
d * 63 * ( 2001 ) 124005 [ arxiv : gr - qc/0010027 ] ; m. cissoko , j. c. fabris , j. gariel , g. l. denmat , and n. o. santos , [ arxiv : gr - qc/9809057 ] ; a. ilha , a. kleber , and j. p. s. lemos , j. math .
* 40 * ( 1999 ) 3509 ; j. p. s. lemos , phys . rev .
d * 57 * ( 1998 ) 4600 ; d. markovic and s. l. shapiro , phys .
d * 61 * ( 2000 ) 084029 ; k. lake , phys .
d * 62 * ( 2000 ) 027301 ] ; a. ilha and j. p. s. lemos , phys .
d * 55 * ( 1997 ) ; w. l. smith and r. b. mann , phys .
d * 58 * ( 1998 ) 124021 [ arxiv : gr - qc/9806013 ] ; _ ibid _ , * 56 * ( 1997 ) .
u. brauer and e. malec , phys .
d * 45 * ( 1992 ) r1836 ; t. shiromizu , k. nakao , h. kodama , and k. maeda , phys .
d * 47 * ( 1993 ) r3099 ; s. m. hayward , t. shiromizu , and k. nakao , phys .
d * 49 * ( 1994 ) 5080 .
j. weyssenhoff and a. raabe , acta phys . polon .
* 9 * ( 1947 ) 7 ; j. weyssenhoff , in : max - planck - festschrift-1958 .
kockel et al .
( deutscher verlap wissensch .
1958 ) p. 155 ; f. halbwachs , thorie relativiste des fluides spin , ( gauthier - villars , paris , 1960 ) ; g. a. maugin , sur les fluides relativistes spin , ann . inst .
henri poincar * 20 * ( 1974 ) 41 ; j. r. ray and l. l. smalley , phys .
d , * 27 * ( 1983 ) 1383 .
f. j. tipler , phys .
a * 67 * ( 1977 ) 8 ; + f. j. tipler , c. j. s. clarke and g. f. r. ellis , in `` _ general relativity and gravitation _ '' , edited by a. held ( plenum , new york , 1980 ) , vol .
2 , p. 97 ; c. j. s. clarke , `` _ the analysis of space - time singularities _ '' , cambridge university press ( cambridge , 1993 ) .
b. kuchowicz , acta cosmologica * 4 * , ( 1976 ) 67 ; d. tsoubelis , phys . rev .
d * 23 * , ( 1981 ) 823 ; v. de sabbata and m. gasperini , lett .
nouvo cimento , * 27 * , ( 1980 ) 133 ; _ ibid _ , lett .
nouvo cimento , * 27 * , ( 1980 ) 289 .
k. lake , in vth _ brazilian school of cosmology and gravitation _ edited b m. novello ( world scientific , singapore , 1987 ) ; p. musgrave and k. lake , class .
quantum grav .
* 13 * ( 1996 ) 1885 ; _ ibid _ , class .
quantum grav .
* 14 * ( 1997 ) 1285 .
t. a. madhav , r. goswami , and p. s. joshi , phys .
d * 72 * ( 2005 ) 084029 [ arxiv : gr - qc/0502081 ] ; a. wang and y. wu , gen .
* 31 * ( 1999 ) 107 [ arxiv : gr - qc/9803038 ] ; s. g. ghosh , phys . rev .
d * 62 * ( 2000 ) 127505 [ arxiv : gr - qc/0106060 ] .
cai , s .- h .
chen , j. b. dent , s. dutta , and e. n. saridakis , classical quantum gravity * 28 * ( 2011 ) 215011 ; a. v. astashenok , arxiv:1308.0581 [ gr - qc ] ; j. amoros , j. de haro , s. d. odintsov , phys .
d , * 87 * ( 2013 ) 104037 ; j. haro , arxiv:1309.0352 [ gr - qc ] .
s. mukherji and m. peloso , phys .
b * 547 * ( 2002 ) 297 ; e. m. prodanov , phys .
b * 530 * ( 2002 ) 210 ; y. shtanov and v. sahni , phys .
b * 557 * ( 2003 ) 1 ; r. maier , i. d. soares and e. v. tonini , phys .
d * 79 * ( 2009 ) 023522 ; r. maier , n. pinto - neto and i. d. soares , phys .
d * 87 * ( 2013 ) 043528 . | we consider the gravitational collapse of a spherically symmetric homogeneous matter distribution consisting of a weyssenhoff fluid in the presence of a negative cosmological constant . our aim is to investigate the effects of torsion and spin averaged terms on the final outcome of the collapse . for a specific interior spacetime setup , namely the homogeneous and isotropic flrw metric ,
we obtain two classes of solutions to the field equations where depending on the relation between spin source parameters , @xmath0 the collapse procedure culminates in a spacetime singularity or @xmath1 it is replaced by a non - singular bounce .
we show that , under certain conditions , for a specific subset of the former solutions , the formation of trapped surfaces is prevented and thus the resulted singularity could be naked .
the curvature singularity that forms could be gravitationally strong in the sense of tipler .
our numerical analysis for the latter solutions shows that the collapsing dynamical process experiences four phases , so that two of which occur at the pre - bounce and the other two at post - bounce regimes .
we further observe that there can be found a minimum radius for the apparent horizon curve , such that the main outcome of which is that there exists an upper bound for the size of the collapsing body , below which no horizon forms throughout the whole scenario . |
the phase behavior of polymer - solvent systems has important application in the industry for the production and processing of many kinds of plastic materials @xcite .
an example is the formation of solid polystyrene foams @xcite .
in addition , both the understanding of the equilibrium phase diagram of these systems and the kinetic mechanism of phase separation are challenging problems of statistical mechanics . on the one hand ,
variation of the molecular weight of the polymer ( i.e. the `` chain length '' of the flexible linear macromolecule , n ) offers a control parameter that leaves intermolecular forces invariant , and hence allows a more stringent test of the theory than would be possible for small molecule systems . on the other hand , it is crucial to take the compressibility of the system fully into account , since for practical reasons one normally wishes to use supercritical fluids ( e.g. , co@xmath0 ) as a solvent @xcite
. then small changes of the pressure result in a large variation of the solvent density , and this fact obviously facilitates applications . while for an incompressible polymer solution ( as it is modelled , for instance , by the well - known flory - huggins lattice theory @xcite ) the early stages of phase separation can only be studied if one realizes a rapid temperature quench from a state in the one phase region into the miscibility gap ( cf .
[ fig1]a ) , for compressible polymer solutions it is possible to induce phase separation by an ( experimentally much more convenient ) pressure jump ( cf . fig .
[ fig1]b ) . of course , in fig .
[ fig1 ] we have made the tacit assumption that including the pressure does not change the qualitative form of the phase diagram , i.e. , the point ( @xmath1 ) in @xmath2-space becomes a line @xmath3 or , expressed in the appropriate inverse functions , @xmath4 , respectively ( fig . [
fig1]b ) . in reality , the situation is more complicated , since one must treat both the densities of co@xmath0 and hexadecane as two coupled order parameters , which both play a crucial role in the possible phase transitions ( remember that for the pure solvent at its critical point gas - liquid phase separation starts ) . as a result , it can happen that both gas - liquid and liquid - liquid phase separation are present in the phase diagram , leading to the occurrence of a triple line ( fig .
[ fig2 ] ) @xcite .
as will be discussed below , the presence of the triple line has a profound influence on the interface properties and hence the nucleation behavior in the system .
the present paper is devoted to a study of the early stages of phase separation kinetics in such systems by computer simulation .
we use a coarse - grained model for mixtures of hexadecane and carbondioxide as an archetypical example .
the reason for the choice of this particular system is the fact that recently extensive experiments on nucleation in this system have been performed @xcite . in sec . 2
, we shall briefly describe the model and the simulation techniques .
presents the relevant static equilibrium properties , in particular , the phase diagram and the properties of planar interfaces separating coexisting phases .
4 describes simulations of quenching experiments under different conditions , that lead either to nucleation and growth or spinodal decomposition . sec .
5 summarizes our conclusions .
a fully atomistic simulation of mixtures of hexadecane ( c@xmath5 h@xmath6 ) and carbondioxide ( co@xmath0 ) , which aims at both establishing phase diagrams as a function of the three variables @xmath7 and @xmath8 , and the study of phase transition kinetics under various conditions , would be an extremely formidable problem . the potentials for the length of the covalent bonds in these molecules ( as well as the potentials for bond angles ) are very stiff . hence , an extremely short time step ( of the order of about 1 fs ) would be required in a molecular dynamics ( md ) simulation . in a corresponding monte carlo ( mc )
simulation only very small random displacements of atoms would be admissible .
thus , for liquid alkanes it is a quite common and well - established practice to integrate ch@xmath0-groups ( as well as the ch@xmath9 end groups ) into `` united atoms '' @xcite , and also to work with a fixed c - c bond length .
these simplifications already reduce the necessary computer time by a factor of 100 .
however , we have estimated that even for such a simplified model computer resources as they are available today still are not yet sufficient . thus it was chosen to simplify the model even further , representing co@xmath0 by a single pseudo - atom , and representing c@xmath5 h@xmath6 by a flexible chain of 5 subsequent effective segments , each of which then contains roughly 3 successive c - c bonds .
all such effective monomeric units interact with a truncated and shifted lennard jones ( lj ) potential , @xmath10,\hspace{0.5 cm } r \leq r_c=2\cdot2^{1/6 } \sigma \ ; , \ ] ] while @xmath11 .
subsequent effective monomers along a chain are in addition exposed to a finitely extensible nonlinear elastic ( fene ) potential @xcite @xmath12\ ] ] the constants in this potential are chosen such that the most favorable distance between bonded neighbors is 0.96 @xmath13 , while the preferred distance between non - bonded effective monomers is @xmath14 .
this mismatch is desirable to prevent crystallization at high densities @xcite , which is appropriate for glass - forming polymers @xcite .
the parameters @xmath13 , @xmath15 set the scales for energy and length of our hexadecane model .
since we wish that our model fits the thermodynamic properties of this material in the fluid phase as faithfully as possible , we have adjusted them such that the critical temperature @xmath16 k and the critical density @xmath17 g/@xmath18 of hexadecane are correctly reproduced . in simulations ,
a critical point can be computed by a finite size scaling study of the liquid - gas transition along the lines of the techniques proposed by wilding et al .
a comparison of critical temperature and density in lennard - jones units with the experimental values yields @xmath19 j , @xmath20 m @xcite .
since the description of hexadecane is thus already reduced to a crudely coarse - grained model , it would not make sense to keep all atomistic detail for carbon dioxide .
thus the co@xmath0 molecule is also coarse - grained into a point particle , and we require an interaction potential between co@xmath0 molecules of exactly the same lj form as in eq .
( [ eq1 ] ) , but with parameters @xmath21 , @xmath22 .
requiring once more that the critical temperature and density of real co@xmath0 are correctly reproduced we obtained @xcite @xmath23 j , and @xmath24 m. it is clear that this procedure ignores some physical effects , e.g. , the fact that the co@xmath0 molecules carry electrical multipole moments is completely neglected .
nevertheless our model not only describes the gas - liquid coexistence curve of hexadecane and co@xmath0 over a reasonable temperature range @xcite , but also other data ( like the critical pressure , or the temperature dependence of the surface tension near @xmath25 @xcite ) . of course , the choice of interaction parameters between co@xmath0 and hexadecane is more subtle .
we model the interaction between the pseudoatoms representing co@xmath0 and the pseudoatoms representing three subsequent ch@xmath0 groups again by a lj potential of the form eq .
( [ eq1 ] ) , but now with parameters @xmath26 , @xmath27 .
it then remains to find an optimal choice for these parameters .
the simplest choice would be the well - known lorentz - berthelot mixing rule @xcite @xmath28 however , when one tries the choice eq .
( [ eq3 ] ) the resulting phase diagram of our model system would be of type i in the classification scheme of van konynenburg and scott @xcite ( i.e. , the critical points of pure hexadecane and co@xmath0 are connected by a critical line of the mixture system .
liquid - liquid phase seperation does not exist and thus there is no three phase coexistence ) .
however , experimentally it is known @xcite that alkane
co@xmath0 mixtures exhibit a type i phase behavior only for very short alkanes , while for hexadecane
co@xmath0 mixtures the phase diagram is of type iii . instead of
a connecting line we rather observe a topology as shown in fig .
[ fig3 ] @xcite .
in the ( p , t ) projection of the phase diagram , a critical line emerges from the critical point of hexadecane and does not end at the critical point of co@xmath0 .
in addition we observe liquid - liquid immisciblity and a three - phase line .
when we want to obtain this behavior from our coarse - grained model we must allow for a deviation from the first equation of eq .
( [ eq3 ] ) , assuming rather @xmath29 it turns out that a suitable choice for the parameter @xmath30 which characterizes the deviation from the lorentz - berthelot mixing rule is @xcite @xmath31 .
( [ eq4 ] ) with this choice of @xmath30 was in fact used to compute the phase diagram shown in fig . [ fig3 ] . however , experimental measurements of the critical line vary considerably and a small modification in @xmath30 rises or lowers the critical line .
details about how such phase diagrams are in fact estimated from the simulation are given elsewhere @xcite .
for a study of the phase diagram ( fig .
[ fig3 ] ) as well as for a study of the interfacial free energy between coexisting gas and liquid phases @xcite the grandcanonical ensemble is used , where volume @xmath32 , temperature @xmath33 , and the chemical potentials @xmath34 , @xmath35 are fixed .
both the particle numbers @xmath36 and @xmath37 of co@xmath0 and hexadecane and the pressure @xmath38 are then ( fluctuating ) observables `` measured '' in the simulation @xcite . fig .
[ fig4 ] shows an isothermal slice through the phase diagram at @xmath39 k. this temperature is higher than the temperature of the critical end point where the critical line ending at the critical point of co@xmath0 and the line of triple points ( @xmath40 ) in the ( @xmath41 ) phase diagram ( fig .
[ fig3 ] ) meet . therefore unlike fig .
[ fig2 ] there is no triple point in the phase diagram , and the qualitative features are the same as in fig .
[ fig1]b . at this temperature
we shall examine the kinetics of the phase separation in sec.4 . in order to observe the kinetics of bubble nucleation in the vicinity of the binodal we control the undersaturation of the system by fixing the chemical potential of both species . as a starting configuration we use a homogeneous state
equilibrated at a higher temperature .
this situation corresponds to a simulation cell in contact with a much larger reservoir which is held at constant undersaturation . as the number of particles in the simulation cell is not conserved , and the particles do not move according to a realistic dynamics , we do not obtain information about the time scale of bubble nucleation .
the kinetics of phase separation in the vicinity of the binodal is , however , chiefly determined by the free energy barrier the system encounters on its path towards the equilibrium state .
hence , we expect the relaxation path to be similar to a simulation with realistic dynamics
. the most natural choice would be to apply molecular dynamics methods @xcite .
however , if the number of particles was conserved we would need prohibitively large simulation cells for the undersaturation not to decrease substantially even in the very early stages of nucleation .
the simulation of spinodal decomposition took place in the canonical ensemble .
again , a homogeneous starting configuration was equilibrated at a higher temperature and quenched to a state below a critical point of the mixture system .
we only allowed for local monte carlo displacements which should yield kinetics comparable to those obtained in molecular dynamics .
first we return to the phase diagrams in figs .
[ fig3 ] , [ fig4 ] and discuss their properties more closely .
note that at the temperature chosen in fig .
[ fig4 ] ( @xmath39 k ) we are very far below the critical point of pure hexadecane ( which is @xmath16 k ) . therefore the hexadecane melt ( with some dissolved carbon dioxide ) coexists with almost pure co@xmath0 the composition of which gradually changes from gas- to liquid - like with increasing pressure . in both cases there is almost no hexadecane present .
this is reflected in the coexistence curve which rapidly approaches the co@xmath0 molar fraction @xmath42 .
we have also included the results of an analytical approximation for the phase diagram of our model , obtained from a thermodynamic perturbation theory @xcite called tpt1 , developed along the lines of wertheim et al .
it is seen that this theory describes the coexistence curve in fig.[fig4 ] at pressures that are much smaller than the critical pressure @xmath43 almost quantitatively , while the critical pressure itself is clearly overestimated .
of course , such an overestimation of the critical parameters @xmath43 , @xmath44 is quite typical for all theories that invoke a mean field description of the critical behavior , as tpt1 does .
a similar analytical study for more alcane - co@xmath45 systems has been performed recently @xcite .
note also that the tpt1 theory readily yields a spinodal curve , which has the standard meaning of separating `` metastable states '' from `` unstable states '' in the phase diagram @xcite .
the naive ( mean - field ) description implies that in the metastable region of the phase diagram the initial stages of phase separation kinetics are described by nucleation and growth @xcite , while in the unstable region the decay mechanism is spinodal decomposition @xcite . in systems with short range forces
there is no well - defined sharp spinodal line , the actual transition between both decay mechanisms occurs rather gradually in a broad transition region , and this region is not centered around the mean field spindodal but occurs close to the coexistence curve @xcite .
however , for the quenching experiment performed at @xmath39 k and @xmath46 , where the final state corresponds to a pressure of about @xmath47130 bar , we are so close to the coexistence curve and so far from the mean field spinodal , that classical nucleation - type behavior @xcite should be observable . in order to demonstrate the mechanism of spinodal decomposition , we quenched into the unstable region as defined by the mean - field spinodal ( compare with fig .
[ fig4 ] ) .
classical nucleation theory assumes ( in our case ) the formation of spherical gas bubbles of essentially pure co@xmath0 within a hexadecane matrix . for a quantitative understanding of the free energy barrier against homogeneous nucleation ,
clearly the understanding of the interface between macroscopic coexisting phases is a prerequisite . as discussed elsewhere
@xcite , such interfaces are most conveniently studied in the grandcanonical ensemble in conjunction with a multicanonical preweighting scheme that generates mixed - phase configurations . choosing a simulation box of rectangular shape @xmath48 with a constant @xmath49 or larger and periodic boundary conditions , one generates system configurations where the two coexisting phases are separated by two parallel @xmath50 interfaces ( compare with fig .
[ fig5]a ) .
`` measuring '' the canonical probability of these two - phase configurations relative to the probability of the pure phases is a standard method for the estimation of interfacial free energies @xcite .
the results for t=486 k are shown in the inlet of fig .
in addition , this method can be used to generate well - equilibrated configurations of interfaces allowing a study of the interfacial structure .
[ fig5 ] shows , as an example , two snapshot pictures of such states containing interfaces , one at @xmath33=486 k and the other at @xmath33=243 k. the size of the particles is enlarged ( radii shown are @xmath22 , @xmath13 rather than @xmath51 , @xmath52 ) , for the sake of a clearer view .
the temperature @xmath33=243 k is slightly lower than the temperature of the critical point of pure co@xmath0 , and thus the system possesses a triple point where a polymer - rich phase coexists with two co@xmath0-rich phases , one being liquid co@xmath0 and the other the gas .
consequently , when we study the interface between the polymer - rich phase and gaseous co@xmath0 , a layer of the third phase ( liquid co@xmath0 ) intrudes at this interface .
the thickness of this layer is expected to diverge when the pressure approaches the value of the triple point pressure . for higher pressures
, we then have a coexistence between the polymer rich - phase and the liquid - like dense co@xmath0 , similar to the situation that occurs at the higher temperature ( fig . [ fig5]a ) . at low enough temperatures , where a triple point occurs
, one expects a strong decrease of the interfacial tension already when the triple point pressure is approached @xcite : the interfacial tension between the polymer - rich phase and liquid - like co@xmath0 is much smaller than the interfacial tension between the polymer - rich phase and the gas .
the barrier against nucleation also decreases when one approaches the spinodal from the metastable region .
this fact facilitates nucleation of relatively large droplets observable on the time - scales of our monte carlo simulation .
the animation in fig . [ fig6 ] visualizes how phase separation in polymer solutions proceeds via nucleation . in the beginning
the system fluctuates around a metastable free energy minimum .
one observes the `` birth and death '' of small density fluctuations in the hexadecane matrix ( displayed as red spheres ) which become visible whenever the blue background shines through the slice .
these irregular voids are too small , however , to lead to an immediate decay of the initial metastable state .
usually the voids also contain a few co@xmath0 molecules ( displayed as small yellow spheres ) . only after some time
lag a void manages to grow to critical size and overcome the free energy barrier which seperates the metastable from the homogeneous equilibrium state . from now on the bubble will grow until it fills the whole simulation box . as expected , the critical void gets also filled with co@xmath0 molecules , thus decreasing the supersaturation of the remaining polymer - rich phase . remarkably , this filling of the bubble does not occur homogeneously : at the interface between the gas and the polymer , i.e. , at the surface of the bubble , the co@xmath0 density is clearly higher than it is in the center of the bubble .
this surface enrichment effect may be interpreted as a precursor to interfacial wetting of the fluid co@xmath0 predicted when one approaches the triple point @xcite .
a quantitative analysis of this phenomena is in preparation @xcite .
of course , the observation of a nucleation event displayed in fig . [ fig6 ] is by no means the first observation of nucleation phenomena by computer simulations : e.g. , already ref .
@xcite contains a series of snapshot pictures illustrating nucleation in the two - dimensional lattice gas model . since then many more elaborate studies have appeared
however , the present work deals with nucleation in a system of industrial relevance , namely bubble nucleation in metastable polymer melts supersaturated with supercritical co@xmath0 . in addition , the nucleation mechanism here is quite nontrivial , since the `` critical bubble '' ( which just has the size to be at the nucleation barrier , where there is a 50% chance of growth or decay ) is characterized by two variables , its size and the number of co@xmath0 molecules it contains .
similarly the interface is described by nontrivial profiles of two variables , the total density and the relative concentration of co@xmath0 .
such a nontrivial coupling between two order parameters has also been predicted by recent self - consistent field calculations based on a qualitatively similar model @xcite . since these mean field descriptions do not take into account the effects of fluctuations , a check by computer simulation methods is clearly warranted . when we quench to a position below the critical point ( fig .
[ fig7 ] ) , the behavior is rather different : fluctuations in the relative concentration of co@xmath0 get gradually more pronounced everywhere .
these fluctuations are not localized , however , in the form of identifiable bubbles , but rather form an irregular percolating network .
this is the hallmark of spinodal decomposition @xcite .
since the linear theory of spinodal decomposition only holds for systems with long range forces in the very early stages @xcite , we do not attempt a more quantitative analysis of this data .
note also that the build - up of these concentration fluctuations is relatively faster than the nucleation of bubbles .
hence the separation of time scales between the structural relaxation and the rate constant of spinodal decomposition is less well established . in the intermediate stages of spinodal decomposition
, we also expect hydrodynamic mechanisms to prevail @xcite , which are not captured by monte carlo moves .
returning now to the simulation of nucleation phenomena , we ask how one can quantify these observations
. a straightforward but tedious method would be to repeat dynamic simulations as shown in fig .
[ fig6 ] many times in order to extract an estimate of the nucleation rate from the average time it takes to form a bubble or a droplet .
of course , since the nucleation rate is expected to decrease dramatically if the quench depth is decreased slightly , such a direct method works hardly in practice . an indirect method to obtain information on the surface free energy of clusters as a function of cluster size simply uses the final equilibrium states of simulations .
an estimate for the free energy as a function of density can be obtained directly from the probability distribution . in this context special care
has to be taken of finite - size effects to identify regions of densities which correspond to a single cluster @xcite .
if we assume that bubbles ( or droplets ) are spherical , filled with gas and surrounded by a homogeneous liquid of coexistence density ( or vice versa ) , we can assign a radius to each density and hence a free energy estimate to each radius .
[ fig8 ] indicates that the surface free energy obtained this way is somewhat smaller than expected from the so - called `` capillarity approximation '' .
a simple estimate of the free energy is @xmath53 , where @xmath54 is the radius of the cluster and @xmath55 the interfacial tension of flat interfaces as shown in fig .
[ fig5 ] and derived in fig . [ fig4 ] .
as a consequence , the nucleation barrier is lower than the classical theory of homogeneous nucleation @xcite would predict .
differences in @xmath55 also decrease with increasing cluster size and distance to the critical point of the mixture . at this point , however , the findings are still somewhat preliminary , and more work will be necessary to obtain sufficiently accurate results @xcite .
in summary , we have presented a simple model and computational tools which allow us to examine equilibrium properties and kinetics of phase separation in systems of industrial relevance .
we were able to produce a two - component mixture phase diagram with a complete critical line which is in qualitative agreement with experiments . for a system at t=486
k we determined the isotherm and interfacial tension of a flat surface as a function of pressure . in a small animations and a series of snapshots we visualized bubble nucleation and spinodal decomposition for this specific system .
for the first case we found a precursor of interfacial wetting which was also observed on a flat surface in the vicinity of the three - phase coexistence line . finally , we found evidence that for small bubbles the interfacial tension is smaller than expected by classical theory of homogeneous nucleation .
financial support from the basf ag ( p.v . ) and from the deutsche forschungsgemeinschaft ( dfg ) via a heisenberg fellowship ( m.m . ) and by grant no bt 314/17 - 3 ( l.g.m ) is gratefully acknowledged .
lgm would like to thank ministerio de ciencia y tecnologia ( mcyt ) and universidad complutense ( ucm ) for the award of a ramon y cajal fellowship and for financial support under contract bfm-2001 - 1420-c02 - 01 .
we are grateful to e. hdicke , b. rathke , r. strey , and d.w .
heermann for stimulating discussions , and to nic jlich , hlr stuttgart , and zdv mainz for generous grants of computer time .
b. krause , h .- j .
p. sijbesma , p. mnkl , n.f.a .
van der vegt , and w. wessling , macromolecules * 34 * , 8792 ( 2001 ) ; b. krause , r. mettinkhof , n.f.a van der vegt , and m. wessling , _ ibid _ , * 34 * , 874 ( 2001 ) n.b . wilding , in _ annual reviews of computational physics
by d. stauffer ) p.37 ( world scientific , singapore , 1996 ) ; n.b .
wilding , m. mller , and k. binder , j. chem .
phys . * 105 * , 802 ( 1996 ) ; n.b .
wilding , f. schmid , and p. nielaba , phys .
rev . e*58 * , 2201 ( 1998 ) m.s .
wertheim , j. stat .
phys . * 35 * , 19 , 35 ( 1984 ) ; _ ibid _ , * 42 * , 459 , 477 ( 1986 ) ; j. chem . phys . *
85 * , 2929 ( 1986 ) ; _ ibid _ , * 87 * , 7323 ( 1987 ) ; w.g .
chapman , g. jackson , and k.e .
gubbins , mol .
phys . * 65 * , 1057 ( 1988 ) | a coarse - grained model for solutions of polymers in supercritical fluids is introduced and applied to the system of hexadecane and carbon dioxide as a representative example .
fitting parameters of the model to the gas - liquid critical point properties of the pure systems , and allowing for a suitably chosen parameter that describes the deviation from the lorentz - berthelot mixing rule , we model the liquid - gas and fluid - fluid unmixing transitions of this system over a wide range of temperatures and pressures in reasonable agreement with experiment .
interfaces between the polymer - rich phase and the gas can be studied both at temperatures above and below the end point of the triple line where liquid and vapor carbon dioxide and the polymer rich phase coexist . in the first case
interfacial adsorption of fluid carbon dioxide can be demonstrated .
our model can also be used to simulate quenches from the one - phase to the two - phase region . a short animation and a series of snapshots help to visualize the early stages of bubble nucleation and spinodal decomposition .
furthermore we discuss deviations from classical nucleation theory for small nuclei .
+ + pacs : 64.60.qb , 82.20.wt , 05.70.fh , 61.41.+e |
very hot jupiters are gas - giant exoplanets with orbital periods less than about 3 days . the close proximity of vhjs to their host stars enhances the influence of irradiation , tidal forces , and stellar activity on their structure and evolution .
corot-2b @xcite is a vhj of particular interest because of lingering questions about the structure of its atmosphere , which can be studied with observations of its secondary eclipse .
@xcite announced the first secondary eclipse observations of corot-2 in the _ corot _ optical waveband , followed by the mid - infrared _ spitzer _ secondary eclipse measurements of @xcite , re - analyzed and expanded with warm _ spitzer _ eclipses by @xcite .
@xcite added a secondary eclipse point in the k@xmath4 band .
the analysis of @xcite favored a poor day - night - side heat distribution in corot-2b s atmosphere .
@xcite found a high 4.5@xmath0 m flux as the only disagreement with a solar - composition , equilibrium chemistry model of the atmospheric temperature structure .
@xcite considered possible emission in the 4.5@xmath0 m band from co mass loss .
both works question , but do not rule out , the presence of a temperature inversion in the atmosphere caused by an upper atmosphere absorber .
@xcite finds that either a carbon - rich or solar abundance non - inverted model fits the data available in the literature . these widely varied , competing explanations for this planet demonstrate the importance of spectroscopic observations .
corot-2b clearly does not fit the standard solar - composition , equilibrium chemistry model that satisfactorily describes many planets in its class , and we explore the anomalous spectral shape . for a clear illustration of corot-2b s standing as an outlier among vhjs ,
see @xcite .
corot-2 is a very active star , a young solar analog , and yet a temperature inversion can not be ruled out and the planet does not fit clearly into the otherwise well - defined inverted / non - inverted planet classifications .
this curious state of the planet is perhaps due to a magnetic interaction between the planet @xcite and corot-2 .
any further understanding would require more measurements of the planet in new wave bands .
cccc[t ! ] a & 10 - 18 - 2010 11:12 - 16:45 & 271 & 80.4@xmath5 + b & 9 - 16 - 2011 09:37 - 15:07 & 276 & 93.9@xmath5 + c & 9 - 23 - 2011 07:41 - 13:11 & 275 & 90.7@xmath5 [ tab : corot2obs ] in this paper , we use the g141 infrared grism on the hubble space telescope s wide - field camera 3 ( hst s wfc3 ) to detect the day - side thermal emission spectrum of corot-2b from 1.1@xmath0 m to 1.7@xmath0 m .
the corot-2 system is part of an hst cycle-18 program that observed a wide range of hjs / vhjs in transit and secondary eclipse , and gives us the basis for new insights into the instrumental effects of wfc3 @xcite . in what follows
, we describe the observations of the corot-2 system in [ sec : observations ] and the initial stages of data analysis in [ sec : red ] . in [ sec : systematics ] we place our observations in the larger context of other hst programs with wfc3 in order to provide a comprehensive systematic description of the instrumental effects encountered in these observations .
we then present our methods of obtaining the band - integrated secondary eclipse curve ( [ sec : whitelight ] ) and derivation of the spectrum ( [ sec : spectrum ] ) of corot-2b .
finally , we use our results to constrain models for the atmosphere of the planet in [
sec : models ] , and we summarize in 8 .
we observed corot-2 using the g141 grism of wfc3 ( 1.1 - 1.7@xmath0 m ) , in three separate visits , each comprising four orbits of hst and hereafter called visits a , b , and c. we used the 128@xmath6128-pixel subarray of the 1024@xmath61024-pixel detector . at the beginning of each visit
, we acquired a single direct image of the system with the f139 m filter , a medium - band filter centered at 1.39@xmath0 m ; the location of the target in this direct image defines the initial wavelength solution for the grism spectra .
a summary of the observations is in table [ tab : corot2obs ] .
most of our observations in program 12181 , including those of corot-2 , were executed before the advent of spatial scan mode @xcite .
lacking the spatial scan , wfc3 observations of relatively bright stars can be inefficient , because the time required to transfer the data greatly exceeds the exposure time for bright exoplanet host stars .
we maximize the efficiency by using subarrays and by exposing the detector to fluence levels approaching or equaling saturation . even at a saturated exposure level
, an unsaturated signal is available because the detector is sampled ` up the ramp ' multiple times within each exposure , and all the samples are saved in the data . isolating less than the full number of samples ,
a linear signal can be obtained even in the saturated case .
our corot-2 grism data are exposed so that the brightest pixel contains about 70,000 electrons in a full exposure , which is approximately the level of 5% non - linearity .
in order to explore whether our results are sensitive to details of the data analysis , we use two parallel but independent methods to process the data . to avoid confusion with the visit terminology ( a , b , c ) , we denote the two methods as @xmath7 and @xmath8 .
method @xmath7 makes more explicit corrections and manipulations of the data than does method @xmath8 .
exoplanet signals are subtle , and the more the data are processed , the more the potential for adding numerical noise that may mask the small exoplanet signal or even fooling oneself into detecting a false signal .
our dual - track analysis allows us to evaluate the trade - off between the most ` complete ' method versus the potential for degrading the results by over - processing of the data .
it also allows us to evaluate what corrections are necessary , and what corrections can be neglected . upon measurement of the eclipse curve ,
the methods yield consistent results .
method @xmath7 uses flt " fits image files retrieved from the mikulski archive at space telescope ( mast ) server , located at the space telescope science institute ( stsci ) .
flt " files were calibrated through the wfc3 pipeline s high - level task , calwf3 , which includes two low - level tasks , wf3ir and wf3rej , that apply to the infrared channel .
wf3ir performs standard calibrations , including corrections for bias , non - linearity , dark current , and bad pixels due to energetic particle hits , while wf3rej completes more bad pixel rejection and combines images .
@xcite gives details of this pipeline .
we multiply the resultant signal rates ( electrons per second ) by the integration time to infer the accumulated signal on each pixel , in electrons .
method @xmath8 begins with
i m a " fits files from the mast server .
these files give the ` sample - up - the ramp ' values of each pixel at 4 times during each 22-second exposure , and are processed to correct for non - linearity , but not to reject energetic particle hits .
we process these files ( minimally ) by fitting a linear slope to the four samples as a function of time for each pixel , to determine the rate at which electrons are accumulating in the pixel .
our linear fit weights each sample of a given pixel by the square - root of the signal level , as appropriate for poisson errors . multiplying the fitted slope by the 22-second integration time yields the accumulated signal in electrons .
this process does not include any correction for energetic particle hits .
rather , we correct those at later stages of the @xmath8-analysis , and we also evaluate the success of the non - linearity corrections by repeating the @xmath8-analysis and restricting the linear fit versus sample time to only the first three samples . using
the smaller subarray means the grism data consists of the central 128 pixel columns of the first - order spectrum out of the 150 on a larger ( sub)array . nevertheless , using the 128 subarray
increased the efficiency of the observations ( i.e. , minimizing data transfer time on the spacecraft ) , more than justifying the loss of points at the edges of the grism response .
to extract the spectrum of the star+planet system , we sum the pixels after background subtraction , using a box defining a range in rows .
we adopt a box size of height 61 pixels ( a central pixel , plus 30 above and below it ) .
the box length is the full 128 pixel length of the subarray , but we later trim the spectrum in wavelength .
we sum the box over rows to produce spectra , and we further sum over wavelength to produce a ` white light ' photometric time series .
the spectra are very stable in position ( jitter less than several hundredths of a pixel ) , and the intensity level falls by 2.5 orders of magnitude over the 30-pixel half - height of the box .
therefore we use fixed integral coordinates for the box in each visit , and we weight each pixel equally when performing the sum .
this spectral extraction is the same for both the @xmath7 and @xmath8 analyses . in the following ,
we discuss the various sub - elements of the data analysis ( [ sec : bp_cr],[sec : bg ] ) including the wavelength calibration ( [ sec : wave ] ) and flat - fielding ( flux calibration , [ sec : flux ] ) , while the more extensive task of characterizing the instrumental systematics is discussed in
[ sec : systematics ] .
bad pixel correction due to energetic particle hits is part of the calwf3 processing used for our @xmath7 analysis .
additional pixels not identified by calwf3 may still be erroneously high or low in value and need correction . for both @xmath7 and @xmath8 analyses , we identify and correct bad pixels immediately prior to the spectral extraction ( i.e. , before summing the box ) .
our @xmath7 analysis inspects pixels in each column of the spectral box ( i.e. , a single wavelength ) that deviate significantly from a gaussian profile of the spectral trace .
such deviations are virtually always characterized by much higher intensity levels .
those pixels that are more than 10 times greater than the fitted gaussian value are replaced by a 7-pixel median in the vertical direction ( perpendicular to the dispersion ) at that wavelength .
our @xmath8 analysis must be more sophisticated as regards bad pixels , since these data have not been processed by calwf3 .
we examine the ratio of a given pixel to the total of all pixels in that row , i.e. , the ratio of a single pixel to the sum over wavelength at each spatial position . because of spatial pointing jitter
, pixel intensities can vary with time in an absolute sense , but their relative variation should be similar at all wavelengths .
we examine the ratio as a function of time ( i.e. , for each exposure ) and we identify instances where a given pixel does not scale with its row sum .
we identify @xmath9 outliers , and correct them using a 5-frame median value of the ratio at that time . for both the @xmath7 and @xmath8 analyses ,
we calculate the background individually for each exposure by using pixels outside of the spectral box .
specifically , the pixels used are those that lie directly below the spectrum on the subarray , which is the section of the subarray corresponding to the width of the spectrum and extending from the bottom edge of the spectral box to the bottom edge of the subarray .
we construct a histogram of intensity values in these pixels and fit a gaussian to the histogram .
the adopted background value is the intensity corresponding to the central value of the fitted gaussian , and we do assume that it is independent of wavelength to the limit of our precision .
this is typically a few tens of electrons per pixel , several orders of magnitude less than the signal in the stellar spectrum , and the sum is thus also significantly lower when calculating the white light curve and its corresponding background .
background subtraction therefore has a relatively minor effect on our analysis .
wavelength calibration utilizes both the direct image and the spectral image , as the wavelength of a given pixel depends upon its location on the detector relative to the direct image .
@xcite outline the procedure for wavelength calibration in an stsci calibration report .
the equations governing the wavelength for a pixel at a given x - position in the first - order spectrum are : @xmath10 where @xmath11 @xmath12 @xmath13 @xmath14 the terms @xmath15 and @xmath16 are the central coordinates of the direct image .
the coefficients ( @xmath17 , @xmath18 , etc . ) are calculated in @xcite . in performing this calibration
, we found that the calibrated grism response ( sensitivity ) curve did not line up precisely in wavelength space with the observed response ( see figure [ fig1 ] ) .
we therefore adjusted the coefficients empirically to obtain optimal agreement with the observed grism response curve and with the wavelengths of two stellar hydrogen lines ( pa-@xmath8 at 1.282@xmath0 m and br-12 at 1.646@xmath0 m ) .
these adjustments yielded : @xmath19 @xmath20 @xmath21 we used these adjusted values in the calibration presented in this work and also successfully applied them to other data sets in this hst program .
therefore , this empirical correction is not specific to this target or these visits , and we in fact used another object in the program ( tres-2 ) to find the correction , as it was observed on the larger subarray and thus the observations include all 150 pixels of the spectrum .
the flat field and sensitivity curve of the g141 grism on the wfc3 detector are the two components of flux calibration , and both are wavelength - dependent . for imaging observations , calwf3 applies the flat field to the data , but flat - fielding of grism data must be done by the observer .
stsci provides a flat - field cube for the g141 grism .
this cube is a four - extension fits file , and each extension is the size of the full wfc3 ir array . for a given pixel on the data image with a given wavelength ,
the flat - field value for that pixel is given by a polynomial function with coefficients defined by the values of the flat - field cube extensions at the pixel s location .
this method is described in the axe handbook @xcite and laid out with the equations that follow . for a pixel at position ( i , j ) , they define a normalized wavelength coordinate , @xmath22 : @xmath23 the parameters @xmath24 and @xmath25 are constants found in the flat - field cube header .
the flat field value of a pixel ( i , j ) with normalized wavelength coordinate @xmath22 is then a polynomial function in @xmath22 : @xmath26 where @xmath27 , @xmath28 , @xmath29 , and @xmath30 are the values at ( i , j ) in the zeroth , first , second , and third extension arrays in the flat - field cube file , respectively . for both our @xmath7 and @xmath8 analyses
, we apply the flat - field correction to the spectral box by dividing by the corresponding flat - field box , " generated pixel - by - pixel from the method above .
stsci also provides the wavelength - dependent sensitivity of the g141 grism . in figure
[ fig1 ] , we plot the scaled sensitivity curve over a flat - fielded spectrum from a single exposure of tres-2 , along with the two hydrogen lines used as reference points in adjusting the wavelength calibration coefficients .
corot-2 has a companion star , so our analysis must remove or correct for this second source .
the direct image of corot-2 appears in figure [ fig2 ] , where the second , fainter source is evident .
the proximity of the second source in the image depends on the orientation angle of the telescope , and varies between the three visits , but it is close enough to be of concern for source contamination .
the spectra overlap minimally in visits a and c , but there is significant overlap in visit b , which has the lowest orientation angle and thus the smallest distance between the two spectra of the three observations .
the orientation angles , which only vary a few degrees from each other , are reported in table [ tab : corot2obs ] .
this second source is an infrared source , 2mass j19270636 + 0122577 , but is just barely spatially resolved by the 2mass observations . in the planet s discovery paper , @xcite suggest it may be a late - k or m - type star , and @xcite identify it as a late k - type star . both works posit that it may be gravitationally bound to corot-2 .
we here address how to remove , or correct the effect , of this second source from the flux of the corot-2 system .
we have explored two approaches .
our @xmath7 analysis removes the second source prior to extracting the grism spectrum from the 2-d frames .
our @xmath8 analysis includes the second source in the extracted grism spectra , and corrects the derived exoplanetary spectrum after deriving that stellar spectrum .
the location of the second source allows us to generate its spectrum , albeit in a limited wavelength range .
its spatial offset results in losing the long - wavelength end of its spectrum . comparing the partial spectrum to a grid of kurucz models
shows general agreement with the findings of @xcite ; a temperature of 4000 k and surface gravity log(g ) = 4.0 , produces the best agreement with the observed partial spectrum .
that corresponds to a late k- or early m - type main sequence star . in our @xmath7 analysis ,
the strategy for removing the second source from visits a and c is to determine the average spatial shape of the source s signal , and scale and subtract it from each column of the spectral box .
we fit a gaussian plus a second - order polynomial baseline to the spatial profile at each column of the data . averaging that fit over all exposures
then approximates the signal from the second source for a given column , after scaling the average to represent the amplitude of the second source for each column .
the original spectral trace the plot of wavelength - integrated flux versus spatial pixel appears in figure [ fig2 ] , as well as the corrected spectral trace ( overplotted ) , showing significant improvement .
for visit b , the task is more difficult .
the peaks of the two sources are separated by just four pixels , compared to twelve and ten pixels for visits a and c , respectively .
the overlap leaves too few points to use a fitting procedure to isolate and approximate the source .
instead , we use the descent of the psf on the opposite side of corot-2 from the overlap of the second source , and mirror the psf column - by - column onto the side with overlap .
we subtract the mirrored psf , and fit a gaussian column - by - column to the difference . averaging that fit over all exposures , we approximate and remove the second source from each column for visit b.
our data exhibit trends in the measured stellar intensity that are not manifestations of physical stellar or planetary phenomena . instead , they represent tendencies of the detector to report signal counts that are different from what actually fell on a given pixel .
we note that instrument - related systematic errors in wfc3 exoplanetary spectroscopy are believed to be less severe than in nicmos observations @xcite .
nevertheless , instrument effects do exist in the wfc3 data , especially for observations taken before the advent of spatial scan mode @xcite such as ours .
some aspects of these instrumental effects were discussed by @xcite .
@xcite reached nearly the photon limit in their analysis of wfc3 g141 transit spectra of the super - earth gj1214b , as did @xcite for two giant transiting exoplanets with the implementation of the spatial scan mode .
we will discuss the analysis of the @xcite work and how we modified it for more general purposes in [ sec : whitelight ] .
we identify three primary manifestations of systematic error , and all are patterns in intensity as a function of time .
the first is a continuous trend of the source s white light curve lasting the entire length of a visit , during which the intensity gradually decreases with time ( or increases , in one case ) .
this visit - long ramp `` is linear in nature ( to within the errors ) , and continuous between orbits .
its slope varies widely between observations , not only among the corot-2 visits , but among all in our hst program .
its strict linearity and variation even when separately observing the same star places it clearly in the category of instrumental effects rather than stellar modulations , but the exact cause is an open question .
the second systematic error feature is a decrease of intensity as a function of time , which repeats for every orbit .
this effect is apparent in the pixels _ not _ illuminated by the source spectrum including the pixels we use for the background subtraction and is shown for the examples of corot-2 and others in figure [ fig3 ] .
for most objects in the program , the effect shows a smooth , exponential decrease in the signal for these pixels over the course of an orbit , though some observations show an effect in more of an ' ' s "- shape instead .
removal of the background , as per the method described in [ sec : bg ] , removes any discernible presence of this effect , which allows us to conclude first that the effect is isolated to the lowest - valued pixels , and second that we need not perform further tasks to eliminate this orbit - long feature , as the problem is solved by careful background definition and subtraction .
we find no definitive cause , though we suspect it may be due to scattered light from the earth s limb .
the third example of systematic error is an increase in intensity of the source s white light curve which occurs on a shorter time scale , over the course of several exposures , and which repeats three or more times in every orbit .
we call this the `` hook '' @xcite because of its characteristic shape , which is a steep jump for the first one to three exposures and then a flattening .
the hook appears to a varying degree in all of the observations , and produces a significant distortion in the data .
examples of the hook within a single orbit of observations for four different objects are shown in figure [ fig4 ] .
the reset of this pattern corresponds with the time when the data stored in the wfc3 buffer are sent to the solid - state recorder on the spacecraft .
this causes a short break in observations , and also resets the detector .
neither of the other two primary systematic effects appear to have any dependence on times of data transfer .
the hook pattern is of similar shape in all sets of observations in the program , but the parameters of its manifestation , e.g. , length of time , number of exposures , number of iterations , vary from object to object .
figure [ fig4 ] shows examples of the pattern in four different objects ; the shape of the pattern is similar , but the amplitude of the hook and the time between buffer dumps ( and thus the number of exposures and total time of each hook ) varies .
@xcite concluded that it was most significant for the @xmath31 subarray .
we concur that it is often prominent at @xmath31 , and it is considerably steeper for longer - duration patterns on @xmath31 , but we detect it in other subarrays also .
the prominence of the pattern correlates with brightness of the star .
while the visit - long slope appears to be linear , both the orbit - long and hook effects are exponential in shape , and therefore each begins as a very strong effect and then becomes nearly indiscernible in the final exposures of each hook pattern .
there are further apparent systematic effects seen in the first orbit of every observation ; they are most likely due to telescope settling and readjusting to a new pointing , and do not have a consistent pattern .
therefore we discard the first orbit once we begin applying corrections to the systematic effects for the purpose of calculating the wavelength - integrated transit depth , and the spectrum of the planet .
since the eclipse of corot-2b is covered by three visits , loss of the first orbit is only a minor perturbation for our analysis .
one potential cause of the hook effect is detector persistence , the phenomenon in which trapped charge in an exposure is slowly released in following exposure(s ) to produce a falsely increased signal detection @xcite .
stsci publishes persistence models and even predictions of persistence for a given exposure based on the exposures prior to it .
the predictions are for an additive effect , and the data product for a given exposure is an image array the size of the original exposure , but with each pixel value equal to the the predicted persistence , so the correction is simply to subtract the corresponding pixel values .
the persistence is low for the first exposure , but jumps up quickly and remains at a higher value until the time of the data transfer , when it , too , resets .
the additive correction as given by stsci do decrease the severity of the hook , but they do _ not _ entirely remove the hook , and we conclude that the hook is a combination of a additive and _ multiplicative _ effect .
this will justify our methods of correction outlined and examined in the sections that follow .
we have made the stsci persistence correction in our @xmath7 analysis .
our @xmath8 analysis ignores additive persistence , as do most wfc3 exoplanetary investigations published to date .
@xcite demonstrated that the hook is more prominent at high exposure levels .
we have investigated the amplitude of the hook as a function of the per - pixel exposure level , and other parameters , and we seek quantitative relationships . for each pixel
, we average the change in signal level over the multiple iterations of the pattern within one orbit , and then examine the change as a function of time within the pattern , flux of the pixel , and location of the pixel on the detector .
the average shape of the hook for two objects in the program can be seen in figure [ fig5 ] .
the normalized signal is shown against the exposure number within the pattern .
for each visit , the pixels have been split between those with flux below the mean and those with flux above the mean .
this is done to confirm that the existence of the hook does indeed depend upon the flux of the pixel .
figure [ fig6 ] shows this dependence of the additive change on the flux of the pixels , where every pixel has been plotted by its initial flux and `` jump '' in electrons between the first exposure and the last exposure in the pattern .
the jump is statistically insignificant below a certain original pixel value , but shows a reliable parabolic rise starting around 30,000 electrons .
the scatter is nevertheless remarkably large , which ultimately means that we can not depend on a unique quantitative relation to correct this effect . in principle , the hook could be removed by using figure [ fig6 ] to predict the magnitude of the jump for a pixel given its initial flux in the first exposure of the pattern , and thereby correct each pixel in each image .
we attempted such a correction , and it does remove the obvious appearance of the hook pattern , but it leaves the data with much more scatter than is acceptable , due to the wide variations seen in figure [ fig6 ] .
we also examined the amplitude of the hook as a function of position on the detector .
we find no correlation in column ( wavelength ) space , but some correlation with the slope of the hook pattern and the row on the detector , i.e. , how far a given row is from the spatial center of the spectrum .
this correlation does seem to strongly depend on which subarray we used . especially in the case of the 128@xmath6128 subarray ,
the slope of the trend is more positive for the rows of pixels below the central peak of the spectrum ( in the direction perpendicular to dispersion ) , while the slope is less positive for those rows above the central peak .
this correlation is weaker for the 512@xmath6512 subarray , but still discernible .
figure [ fig7 ] shows the correlation for the smaller subarray by demonstrating the shift in the spatial center of the spectrum between the starting and ending frames of the hook .
our finding that the nature of the hook depends on the row of the spectrum may be a significant clue to the nature of this effect .
reading the detector involves addressing the pixels by row , and it is conceivable that the hook is related to the manner in which the detector is addressed and sampled .
we conclude that the effect in figure [ fig7 ] can not be explained by anything like telescope drift .
the trend featured in figure [ fig7 ] is correlated with the hook and therefore the transfer of the detector buffer , a task performed with no relation to telescope motion .
128 subarray .
we show here that the location of the maximum point of the spectrum ( the peak of the gaussian ) typically moves , and typically moves in the same direction , over the course of the hook .
this indicates that the hook pattern has a row dependence . ]
we wish to produce a time series of the wavelength - integrated ( ` white light ' ) signal measured from corot-2 in order to determine the amplitude and central phase of the secondary eclipse .
this will yield the total signal from the planet over the g141 bandpass , while the spectrum that we calculate in sec .
[ sec : spectrum ] will distribute that signal as a function of wavelength .
we begin with the light curves for the three visits shown in figure [ fig8 ] , then we correct these light curves to remove the instrumental systematic effects , and we combine the three visits to form a single eclipse curve as a function of orbital phase .
@xcite successfully removed systematic effects from their dataset .
the steps of their * divide - oot * method for correcting a transit / eclipse curve are as follows , assuming a five - orbit set of observations , with orbits three and four in transit : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ \1 .
ensure that all orbits have the same number of exposures .
the fifth orbit usually has fewer exposures than orbits two , three , and four , so simply repeat the last element to make up the difference .
since the hook pattern is flatter at its end , this is a reasonable approximation .
+ 2.create an average out - of - transit orbit by simply averaging orbits two and five .
divide each orbit ( two , three , four , and five ) by the average orbit .
+ 4.remove the artificial elements that were added in the first step .
+ 5 . fit a line to the second and fifth orbits , as there is still usually a hint of the visit - long ramp .
divide by the linear fit to normalize the data in units of the stellar flux .
+ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ this method should yield an acceptable eclipse curve with out - of - transit flux normalized to unity .
application of * divide - oot * to objects in our hst program 12181 proved successful only in some cases @xcite . a modification of the method will be explained below .
we observe corot-2 in four orbits per visit , but each visit contains at most one orbit that is completely in - eclipse ( when the planet does not contribute ) , and each visit contains the virtually unusable first orbit . for this reason and due to our significantly lower signal - to - noise ratio than for the @xcite planet s observations
the gj1214b transit depth is two orders of magnitude larger than the depth of the corot-2b secondary eclipse in the same waveband and on the same grism
our corot-2 data are not well - suited for the * divide - oot * method _ per se_. another issue with corot-2 is the severity of the visit - long ramp , which causes trouble when trying to average pattern shapes before removing the ramp .
therefore , instead of dividing by an average _ orbit _ , we elect to divide by an average _ pattern _ , defined both by the occurrence of a buffer dump and through visual assessment , and we proceed as follows : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ \1 .
identify the patterns that are out - of - eclipse . divide the entire white - light curve by the median of the out - of - eclipse exposures from a single , early orbit ( usually orbit 2 ) .
this normalizes the curve to unity .
+ 2 . fit a line to the out - of - eclipse patterns , but exclude all points below intensity level 0.997 .
these outliers are due to the hook effect , and would bias the visit - long slope correction .
divide by the fitted curve to re - normalize to unity .
create an average pattern by averaging the out - of - eclipse patterns .
divide each occurrence of the pattern in the entire white - light curve by the average pattern .
+ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ this creates a vast improvement in the data , with a significant reduction in the presence of systematic effects .
we are also able to utilize the later patterns of the first orbit , rather than discarding it completely , as the problems presented by settling or other effects of unknown origin diminish significantly after one to two iterations of the pattern .
an average pattern is plotted in the inset of figure [ fig8 ] , and the corrected data are shown in comparison to our best - fit eclipse curve in figure [ fig9 ] . with the corrected data in hand after applying our modified * divide - oot * procedure , we fit an eclipse curve using the data from our @xmath7 analysis .
we calculate the shape of the theoretical eclipse curve using expressions from @xcite , with orbital parameters from @xcite , except for the orbital period where we adopt the updated value from @xcite . in fitting the data ,
we vary only the central phase and amplitude of the eclipse , the latter by scaling the amplitude of the theoretical curve .
we perform the fit using two @xmath32-minimization methods .
first , we implement a levenberg - marquardt algorithm to vary the eclipse amplitude and central phase simultaneously , to find the global minimum in @xmath32 .
second , we vary the central phase incrementally from 0.49 to 0.51 in steps of @xmath33 . at each step
, we calculate the best - fit eclipse amplitude at that phase in closed form , using linear least - squares .
cycling through the range of trial central phases , we again find the global minimum @xmath32 .
results from the two methods were in excellent agreement .
we find a best - fit eclipse depth of @xmath1ppm ( parts per million ) ; the fit is shown in figure [ fig9 ] . the reduced @xmath34 ;
as it was calculated estimating the error to be poissonian , the ideal scenario , this @xmath35 value indicates that the achieved per - point scatter is 2.6 times the photon noise .
the error level , and the appearance of figure [ fig9 ] , suggests that red noise remains in the data , in spite of our modified * divide - oot * procedure . to verify the presence of red noise , we binned the residuals from the best - fit eclipse over @xmath36 points per bin , and calculated the standard deviation of the binned points , @xmath37
. we solve for the slope of the relation between @xmath38 and @xmath39 using linear least - squares .
poisson noise will produce a slope of @xmath40 , whereas we find a slope of @xmath41 for the figure [ fig9 ] data , confirming the presence of red noise .
given the presence of red noise in the white light eclipse data , we assign errors to the best - fit eclipse parameters ( eclipse amplitude and central phase ) using the residual permutation ( prayer - bead " ) method @xcite .
figure [ fig10 ] shows histograms of the results for the best - fit amplitude and central phase , based on the residual permutation fits . for reference , we fit gaussians to these histograms .
a gaussian is a reasonable approximation to the central phase histogram , but the eclipse amplitude histogram has a higher central peak , and lower wings , than does a gaussian .
our adopted errors are equivalent to the @xmath42 points in the histograms , in the sense that @xmath43 of the histogram area lies beyond each quoted @xmath44 value ( @xmath45 considering both ends of the range )
. our best - fit eclipse is centered at a phase of @xmath46 .
the light - travel time across the orbit is 28 seconds .
the central phase for a circular orbit would be 0.50019 , consistent with our result , within our errors .
@xcite found the eclipse to occur slightly earlier than expected for a circular orbit , at phase @xmath47 .
@xcite found a central phase of @xmath48 , weakly supporting the result from @xcite .
the low signal - to - noise due to the shallower secondary eclipse at shorter wavelengths of the eclipse in the wfc3 bandpass contributes to a relatively large error level for the central phase ( approximately 4 to 8 times larger than the spitzer errors ) .
although we find good agreement with a circular orbit , we can not exclude the result of @xcite who concluded that the orbit is slightly eccentric .
@xcite used his * divide - oot * method for gj1214b to derive the depth of transit as a function of wavelength , i.e. , the transmission spectrum . in principle that method is applicable to exoplanetary spectra at secondary eclipse , but we use an alternate technique .
we have at most one in - eclipse reference orbit ( when the planet does not contribute ) per visit .
moreover , corot-2 is a relatively faint star ( v=12.6 , h=10.4 ) . in the faint - source limit , dividing single - wavelength data by a single reference orbit would increase the random noise in the quotient to an unacceptable degree , because we are photon - starved . to obtain the spectrum of the planet
, we utilize the differential method described by @xcite and explained below .
we apply this method to data from both our @xmath7 and @xmath8 data analyses , finding consistent results .
a by - product of this method is a time - dependent scaling factor obtained by fitting a template spectrum ( see below ) .
this scaling factor is an excellent proxy for the white light eclipse , and we find consistent results between the modified * divide - oot * and differential methods when calculating that white light eclipse .
that comparison also served to verify that our @xmath7 and @xmath8 analyses produce consistent values for the white light eclipse depth .
the differential method is intended to exploit the characteristics of the systematic hook pattern in order to cancel it , while also correcting for the effects of jitter in wavelength over time .
the amplitude of the hook is a function of the flux level in the affected pixels ( [ sec : systematics ] ) .
the procedure of the differential method , in its simplest form , is to therefore extract the intensity in each column of each grism image , and divide that intensity by the wavelength - integrated intensity in the entire spectrum observed at that time . in other words , ratio the intensity in a given column on the detector ( after subtracting the background , and integrating over rows ) to the sum of all columns , and we repeat this process for the grism image at each orbital phase @xmath49 .
this ratio adds minimal noise , because the precision of the wavelength - integrated spectrum is much better than the precision of a single wavelength .
moreover , the ratio should be effective in removing the hook , as long as the wavelength used in the numerator is not too close to the edges of the grism response , where the intensity rolls - off to much smaller values , as does the hook ( [ sec : systematics ] ) .
the observed grism spectral intensity varies only modestly ( figure [ fig1 ] ) over the 1.1 - 1.7@xmath0 m range of our analysis .
thus , dividing a single wavelength by the sum of all wavelengths is a comparison of similar intensity levels , so we expect much of the hook pattern to cancel , and this expectation is met by the actual data ( see below ) .
the differential method also removes the white - light eclipse .
specifically , the eclipse shown on figure [ fig9 ] , by summing over wavelengths , will identically cancel .
however , wavelength - to - wavelength _ variations _ in the eclipse depth will be preserved .
we call these differences _ differential depths _ and we derive them either positive or negative , by fitting to the wavelength - ratioed data
. we then add the depth of the white - light eclipse , reconstructing the full emergent spectrum of the planet at eclipse . in actual practice , the implementation of this differential method is more complex than the simple division implied above .
we do not explicitly divide by a wavelength integral ; we use an equivalent but more subtle procedure that we now describe .
we must account for possible wavelength shifts in each grism spectrum .
wavelength shifts have two effects .
first , a shift of the spectrum changes the intensity in a given column because the grism response varies with wavelength .
second , a shift in the spectrum changes the range of wavelengths sampled by a given column of the detector .
we find that the wavelength shifts are of order 0.02-pixels , and they vary within an orbit , but tend to reset and exhibit a similar pattern in subsequent orbits . given this magnitude of shifts , the second effect mentioned above - a perturbation to the wavelength assigned to a given column - has negligible effect .
we therefore ignore the wavelength perturbations per se , and we use the wavelength scale from the calibration described in [ sec : wave ] .
however , the first effect ( changes in grism response with wavelength ) is important , and we deal with it as follows : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ \1 . for each visit ,
form a template " spectrum of the star alone by summing the in - eclipse ( planet hidden ) spectra .
denote this spectrum by @xmath50 , where @xmath22 is the column coordinate on the detector .
fit the template to each individual spectrum by re - sampling , shifting ( in steps of @xmath51 pixels ) , and scaling @xmath50 in intensity using linear least - squares .
perform this least - squares fit over a large range of shift values ( @xmath52-pixels ) and choose the shift that exhibits the best fit as judged by the standard deviation of the ratio .
each individual spectrum , @xmath53 at orbital phase @xmath49 , matches a version of @xmath50 with a scaling factor @xmath54 : @xmath55 .
the prime marks the change in intensity due to the shift in @xmath22 , and the zero - point constant @xmath56 is negligibly small .
form the ratio @xmath57 .
+ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ an example of this basic process of shifting and fitting the template spectrum , for a randomly selected spectrum in visit c , is illustrated in figure [ fig11 ] .
however , our actual analysis adds an additional step in order to deal with the undersampling of the stellar spectrum as discussed by @xcite . between steps 3 and 4 above : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 3.5 . smooth all of the spectra using a gaussian kernel with fwhm = 4 pixels .
+ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the choice of pixels ( columns in wavelength ) is dictated by the tradeoff between suppressing the undersampling , and preserving the spectral resolution .
[ t ! ] the wavelength integrals of @xmath53 and @xmath55 are closely equal because of the fitting process that matches them .
moreover , the shape of @xmath50 is constant over a visit , i.e. , its value at any single wavelength , relative to its wavelength integral , is constant .
hence the point - by - point division described above is conceptually equivalent to dividing a single wavelength ( equivalently , @xmath22-value ) in @xmath53 by the wavelength integral of @xmath53 .
however , our procedure has the advantage that we do not have to re - sample any spectra wherein the potential signal is present , or where the reference stellar spectrum is changing .
hence we introduce no extra noise in this process , while also correcting for wavelength jitter in the spectrum . performing the procedure described above yields
a set of ratio values @xmath58 for each visit .
we now combine visits as follows : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ \1 . for each column of the detector @xmath22 ,
fit a straight line to the @xmath58 , where the independent variable in the linear fit is phase @xmath49 , and then divide by that line .
+ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ dividing each visit by the linear fit removes any slight slopes that are present in each visit ( as described by @xcite and [ sec : systematics ] ) and places all three visits on a common scale . _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ \2 .
fit an eclipse curve to the combined @xmath58 at each @xmath22 , holding the central phase fixed at @xmath59 for the eclipse fit , solving only for the depth .
+ 3 . use the wavelength calibration to associate a wavelength with each column @xmath22 ; @xmath58 becomes @xmath60 .
+ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the wavelength scale is sufficiently similar for each of the visits that we associate visit - averaged wavelengths with each @xmath22 .
the upper panel of figure [ fig12 ] shows the result of fitting an eclipse curve to the visit - combined @xmath60 at a randomly - selected wavelength .
because the white - light eclipse has been removed by the process used to generate the @xmath60 , the differential eclipse depth at individual wavelengths can be either positive or negative depending on whether the intensity of the exoplanetary spectrum is greater or less at that wavelength compared to the average over the band defined by the grism response .
note that the scatter in the individual points on figure [ fig12 ] is large compared to these differential eclipse depths .
however , the precision of the differential eclipse depths is much better than the single - point scatter in @xmath60 , and we also average adjacent wavelengths to derive spectral structure in the exoplanetary spectrum ( see below ) . [
t ! ] as the final step , _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ \4 .
we add the white - light eclipse depth ( 0.000495 , [ sec : eclipsedepth ] ) to the differential eclipse depths , and thereby derive the planet - to - star contrast versus wavelength .
+ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ this emergent spectrum of the planet is illustrated on figure [ fig13 ] , from both our @xmath7- and @xmath8-analyses .
the upper panel shows the values for individual wavelengths , i.e. , single columns of the detector , and the lower panel bins the results in bins of width 0.05@xmath0 m ( 4 columns ) . we have estimated the errors on the differential eclipse depths using two methods . for both methods , we remove the fitted differential eclipse and examine the properties of the point - to - point scatter ( figure [ fig12 ] , top ) for each wavelength .
first , we bin these points using bin widths of 2 , 4 , 8 , 16 , 32 and 64 points , and we calculate the scatter in those binned values . for poisson noise , we expect that the scatter as a function of bin size @xmath61 will decrease as @xmath62 .
an example of the measured relation at a single randomly - chosen wavelength is shown in the lower panel of figure [ fig12 ] , where the dashed line is an extrapolation from the single - point scatter using an exponent of @xmath40 , and the solid line is what we calculate from the actual data .
these differential data are nearly photon - limited at almost all wavelengths , and @xmath61 decreases very close to @xmath62 .
we write @xmath63 and we solve for @xmath54 and @xmath56 .
we then use that relation to calculate the expected precision for the aggregate in - eclipse points and the aggregate out - of - eclipse points , and we propagate those errors to calculate the error on the differential eclipse depths . as a check on the above error calculation , we also derive the precision of the differential eclipse directly using the residual - permutation method @xcite . removing the best - fitting differential eclipse , we permute the residuals sequentially and add them back to the best - fit eclipse curve to make new data .
fitting to these re - cast data for all possible permutations ( 580 of them ) , we calculate the dispersion in the resultant differential eclipse depths . on average , we find that this produces excellent agreement with the first method described above . for our final spectrum and errors , we bin the results - and propagate the errors - to the same resolution ( 4 columns , 0.05@xmath0 m ) that we used as a smoothing kernel in the wavelength jitter correction .
figure [ fig13 ] shows the exoplanetary spectrum from our analyses at single - column resolution ( top panel , only @xmath7 results for illustrative purposes ) , and binned to a wavelength resolution of 0.05@xmath0 m ( bottom panel ) .
the error bars on the @xmath7 binned spectrum in figure [ fig13 ] are 77 ppm on average , which is 25% greater than the photon noise . from our @xmath8-analysis ,
the binned spectrum is similar , and the errors average to 73 ppm ( 18% greater than the photon noise ) .
the values of our binned spectra , and errors , are listed in table [ table2 ] .
ccccc 1.125 & 334.6 & 67.4 & 248.6 & 86.0 + 1.169 & 272.4 & 83.7 & 366.7 & 109.6 + 1.218 & 339.4 & 119.3 & 309.0 & 83.2 + 1.278 & 344.2 & 72.0 & 313.5 & 60.5 + 1.324 & 338.9 & 64.7 & 279.9 & 56.8 + 1.369 & 403.9 & 77.1 & 376.2 & 60.1 + 1.424 & 454.5 & 59.8 & 480.9 & 65.5 + 1.475 & 320.3 & 93.5 & 304.8 & 80.3 + 1.525 & 438.3 & 62.6 & 454.6 & 63.4 + 1.574 & 548.7 & 61.3 & 632.1 & 61.9 + 1.619 & 382.2 & 82.2 & 414.0 & 73.8 + [ table2 ]
no single model for the atmosphere of the planet fits all of the available data to within the errors .
the observed properties of the planet s atmosphere include : 1 ) the optical eclipse observed by the _ corot _ mission @xcite , 2 ) a ground - based eclipse near 2@xmath0 m @xcite , 3 ) the overall level , general slope with wavelength , and lack of obvious or known spectral features seen in our wfc3 data , and 4 ) eclipses in 3 spitzer bands @xcite .
figure [ fig14 ] shows these data in comparison to several modeled spectra : a best - fit blackbody , conventional solar abundance models @xcite , and a carbon - rich model @xcite .
although none of these are ideal fits to the data , each model has characteristics that account for some observed properties of the planet , as we now discuss .
the lower panel of figure [ fig13 ] includes the contrast produced by a best - fit blackbody for the planet compared to the results from our @xmath7 and @xmath8 analyses , and figure [ fig14 ] plots that blackbody in comparison to the totality of existing eclipse data .
we adopt a kurucz model for the star ( teff=5750 , log(g)=4.5 ) , yielding a best - fit blackbody temperature of @xmath2k for the planet in our wfc3 band , from our @xmath7-analysis .
this blackbody temperature gives acceptable agreement with the infrared eclipse results at longer wavelength ( figure [ fig14 ] ) .
the 1788k blackbody - derived from the wfc3 data alone - misses the ground+spitzer eclipse amplitudes by an average of about @xmath64 .
however , a blackbody spectrum for the planet does not produce the best _
slope _ over the wfc3 band , as we now discuss .
our observed wfc3 spectrum for corot-2b has two striking features : 1 ) it slopes slightly upward with increasing wavelength , and 2 ) it shows little to no evidence for water absorption or emission in the 1.4@xmath0 m band .
statistically , the first question to resolve is whether the simplest possible fitting function can account for our spectrum .
the simplest function is a single value in contrast , i.e. a flat line at the average contrast level . for our @xmath7 analysis spectrum ( red points on figure [ fig13 ] ) the @xmath32 of the best - fit flat line is 12.8 for 10 degrees of freedom , so our @xmath7 analysis accepts a flat line as representing the planet s contrast across the wfc3 band . for our @xmath8 analysis ( blue points on figure [ fig12 ] ) ,
the flat line @xmath32 is 28.6 , rejecting the flat line at @xmath65 confidence .
so our @xmath8 analysis indicates a stronger and more significant upward slope than does our @xmath7 analysis .
that is the single largest difference between our @xmath7 and @xmath8 analyses , that are otherwise very consistent , with all points overlapping within their error bars ( figure [ fig13 ] ) .
both of our wfc3 analyses reject the best - fit blackbody slope for the planet , but only at about the 93% confidence level .
the @xmath32 values are 17.0 and 17.8 ( 10 degrees of freedom ) for our @xmath7 and @xmath8 spectra , respectively . on the other hand ,
the blackbody is obviously consistent with the weakness of water absorption in the wfc3 band .
we checked that our results are not affected by inadequate corrections for detector non - linearity at the high fluence levels of our data .
we repeated the @xmath8 analysis omitting the last ( fourth ) sample of the exposure , and using only the first 3 samples , where the fluence level ( @xmath6647,000 electrons ) is well within the linear regime . that modified version of the @xmath8 analysis shows little difference from the @xmath8 spectrum shown on figs .
[ fig13 ] & [ fig14 ] ( but , with larger errors due to the lower fluence levels ) .
the slope of the planet s spectrum across the wfc3 band is relevant to the interpretation of the eclipse amplitude observed in the optical by corot @xcite . if a 1788k blackbody agreed with the slope of our observed spectrum , it would be reasonable to extrapolate that blackbody to judge the magnitude of the thermal emission from the planet at optical wavelengths .
a blackbody of 1788k would produce negligible thermal emission in the optical , and we would conclude that the optical eclipses are due to reflected light .
however , given that the observed slope across the wfc3 band does not decline as strongly as a 1788k blackbody , it remains possible that the optical eclipses are due to thermal emission . that could happen , for example , if temperatures on the star - facing hemisphere of the planet were spatially inhomogenous .
hotter regions having a small filling factor , combined with cooler regions of larger filling factor , could in principle produce the observed slope across the wfc3 band , and account for the optical eclipses , while still remaining consistent with the observed contrast at wavelengths exceeding 2@xmath0 m . in order to probe the viability of our speculation concerning temperature inhomogeneities
, we performed exploratory fits ( not illustrated ) using two different blackbody temperatures and filling factors on the star - facing hemisphere of the planet . we find a good fit to our wfc3 and the corot data using @xmath67k and @xmath68k , with filling factors of @xmath69 and @xmath70 , respectively .
this combination matches the level of the contrast in the corot bands as well as the contrast level and wavelength dependence of our wfc3 results , but it significantly underestimates the contrast in the spitzer bands ( by about 0.001 ) .
recent hydrodynamic simulations of hot jupiter atmospheres show brightness temperature variations as large as a factor of two on the star - facing hemisphere of hd189733b @xcite .
since that planet is less strongly irradiated than corot-2 , the temperatures found by our exploratory fits seem plausible . nevertheless , we do not here attempt to model the atmosphere of corot-2 using a self - consistent 3-dimensional approach ( temperature varying with depth and with horizontal coordinate ) .
higher quality data , such as we anticipate from the james webb space telescope , may justify such an approach in the future .
both our @xmath7 and @xmath8 spectra agree that a straight line ( contrast increasing linearly with wavelength ) gives a good account of our results across the wfc3 band : the @xmath32 values for a linear fit ( 9 degrees of freedom ) are 6.1 and 13.7 for our @xmath7 and @xmath8 spectra , respectively .
these values leave little room for absorption or emission by water vapor at 1.4@xmath0 m . in order to specify a limit on the degree of water absorption or emission , we scale and fit a burrows model to the data , using the model shown in blue on figure [ fig14 ] . in order to make the limit responsive to the modulation caused by the actual water absorption ( as opposed to the slope of the continuum ) , we allow for a linear baseline difference as a function of wavelength .
we construct 10,000 trial data sets , adopting the error at each binned wavelength from our @xmath8-analysis , and we fit the model plus a linear baseline to each trial data set using linear regression . based on the distribution of fitted amplitudes , we find an 85 ppm 3@xmath71 limit on the amplitude of water absorption or emission , measured at the bandhead at 1.38@xmath0 m .
this limit assumes that the shape of the water absorption is the same as in the burrows model .
the @xmath3 limit of 85 ppm is significantly less than the already weak water absorption seen during transmission spectroscopy of the giant planets xo-1b and hd209458b @xcite , wasp-19 @xcite , and hat - p-1b @xcite .
this conclusion is significant , as can been seen by reference to one conventional solar abundance burrows model @xcite illustrated as the dark blue line on figure [ fig14 ] .
this model is not intended as a fit to the wfc3 data , but it was invoked by @xcite in an attempt to account for the spitzer observations .
although it misses the 4.5@xmath0 m spitzer point , @xcite discussed the possibility of circumplanetary carbon monoxide emission in that band , due to tidal stripping by the star .
however , this model produces a much larger spectral modulation in the wfc3 band than is seen in our observed spectra .
corot-2b is an unusual planet , and the spitzer data have been particularly difficult to understand , as discussed by @xcite ( but , see @xcite ) .
the relatively high contrast at 3.6- and 4.5@xmath0 m seems to require a hot continuum , allowing little if any molecular ( principally water ) absorption .
simultaneously , the lower contrast at 8@xmath0 m requires absorption to a significant degree .
we here explore the potential for conventional solar abundance model planetary atmospheres to account for the totality of the corot-2b eclipse data .
the weakness of absorption features can be produced in a solar abundance model by adding continuous opacity by small particle scattering and/or absorption .
that could dampen features in the emergent spectrum at short wavelengths , but a reduced scattering cross - section with increasing wavelength could allow greater spectral contrast at 8@xmath0 m ( mentioned by @xcite ) . if the temperature remains nearly constant as a function of pressure / altitude in the planet s atmosphere , that would also suppress any absorption or emission features in the emergent spectrum .
figure [ fig14 ] shows the contrast from a burrows model @xcite having three additional sources of opacity not present in a clear atmosphere .
this model is shown in green on figure [ fig14 ] , and has redistribution parameter @xmath72 @xcite .
the additional opacity sources are first , a high altitude optical ( 0.4 - 1.0@xmath0 m ) absorber of opacity @xmath73 .
second , an absorbing haze opacity of @xmath74 uniformly distributed at all pressures and wavelengths , and third , a scattering opacity of @xmath75 also uniformly distributed at all pressures and wavelengths .
the scattering opacity acts to increase the reflected light , but not increasing the thermal emission .
note that in principle we could include a wavelength dependence to the opacity of the broadly distributed hazes , but we prefer to keep this _
ad hoc _ opacity as simple as possible .
the uniformly distributed hazes damp the spectral modulation in the wfc3 bandpass to an acceptable degree , but the model misses the overall wfc3 contrast level , being too high by 161 ppm .
like all single - spatial - component models , it s slope across the wfc3 band is larger than our data
. given the error level of our white light eclipse ( @xmath1ppm ) , the overall contrast difference is significant at @xmath76 , which is the single largest problem with this model . on the other hand ,
the scattering opacity increases the contrast in the optical to the point where it underestimates the _ corot _ eclipse amplitude by less than @xmath77 . also , among the models we ve tested , it does the best job of reproducing the long wavelength eclipse amplitudes ( @xmath78 on average ) .
the aggregate eclipse data are ambiguous concerning the possibility of a thermal inversion in the dayside atmosphere of corot-2b .
as discussed in @xcite , the lower brightness temperature in the 8@xmath0 m spitzer bandpass compared to the brightness temperatures in the shorter wavelength channels ( except 4.5@xmath0 m ) suggests a temperature profile decreasing outward in the atmosphere .
if that gradient is flatter than radiative equilibrium models predict , it will help to account for the lack of strong spectral features . on the other hand , the solar abundance radiative equilibrium model ( green line on figure [ fig14 ] ) achieves good agreement with the 4.5@xmath0 m spitzer eclipse depth by incorporating @xmath73 of extra optical - wavelength opacity at low pressures ( @xmath661 mbar ) close to where radiation in the 4.5@xmath0 m band is formed @xcite .
indeed , that model shows a temperature rise of about 75kelvins , near 0.2 mbars .
but due to the more widely distributed absorbing haze , the temperatures in this model at high altitude are already hundreds of kelvins over the values they would have in a clear atmosphere . to the extent that this model is preferred , or that a flattened temperature gradient counts as a weakly - inverted atmosphere , then corot-2b could be claimed to have a temperature inversion . however , this evidence for an inversion is weaker than for hd209458b @xcite , and is ambiguous in the sense that the atmosphere could be heated without satisfying a strict definition of inversion ( temperature increasing with height ) .
@xcite hypothesized that planets orbiting active stars will not have strong atmospheric temperature inversions , because the absorbing species that causes the inversion ( e.g. , @xcite ) may be destroyed by the enhanced uv flux from stellar activity .
corot-2a is an active star @xcite , and lack of a strong thermal inversion in corot-2b would support the @xcite hypothesis .
an alternate way to reduce the spectral modulation by water vapor in the wfc3 bandpass is to reduce the equilibrium water vapor mixing ratio , for example by increasing the carbon abundance relative to oxygen .
this also helps to decrease absorption in the 3.6- and 4.5@xmath0 m bands ( although methane does contribute some absorption at 3.6-@xmath0 m ) , while preserving absorption at 8@xmath0 m via the 7.8@xmath0 m methane band .
we used the methodology described by @xcite and @xcite to find a possible carbon - rich match to the aggregate data for this planet ( except for the optical eclipses ) .
@xcite discussed corot-2b and was able to fit the pre - wfc3 data by varying the c / o ratio to various degrees .
the magenta line on figure [ fig14 ] is a model with an enhanced carbon abundance ( c / o=1 ) , and having a non - inverted atmosphere with modest thermal contrast ( 700k increase in temperature from upper boundary to the optically thick photosphere ) . the enhanced carbon abundance weakens the water absorption , but allows sufficient absorption near 8@xmath0 m to account for that spitzer point to within @xmath79 .
the average agreement with the ground+spitzer eclipses is @xmath80 , not quite as good as the blackbody and the solar abundance model . on the other hand ,
the carbon - rich model does the best job of the atmospheric models , i.e. , beyond just a linear fit of reproducing the wfc3 spectrum ( @xmath81 for 10 dof ) , and in particular it agrees essentially perfectly with the amplitude of the wfc3 white - light eclipse .
it does not require additional haze opacity .
we here summarize the main conclusions from comparing the aggregate eclipse data for this planet with emergent spectra from different models .
we tested a blackbody , as well as more sophisticated solar abundance and carbon - rich models .
no model fits all of the data .
the limit on spectral modulation due to water absorption in the wfc3 band is our main observational result .
given the lack of clear and unequivocal molecular absorption features in the wfc3 and other bands , emergent spectra more sophisticated than a blackbody are unproven .
a blackbody spectrum gives an acceptable fit to the data except for the optical eclipses as seen by _
a blackbody spectrum fits the slope over the wfc3 band poorly , but multi - component blackbodies due to spatial inhomgeneities on the star - facing hemisphere have the potential to help account for the observations , including the optical eclipses as seen by _ corot _ , especially if extra scattering opacity increases toward short wavelengths .
note that the absorbing and scattering hazes invoked in our solar abundance model are qualitatively similar to extra absorption and scattering opacity inferred for the archetype planet hd189733b @xcite .
although a blackbody spectrum reasonably accounts for the infrared eclipse data , it is not a model of the planet s atmosphere _ per se_. instead , the planetary atmosphere can mimic a blackbody via the presence of extra continuous opacity that damps the observed thermal contrast , or due to a high carbon abundance that weakens the bands of the principal molecular absorber ( water vapor ) . in either case ,
extra scattering opacity at optical wavelengths could help to account for the amplitude of the optical eclipses .
we find only weak evidence for a strong temperature inversion , but extra absorbing opacity in the solar abundance model would perturb the temperature profile in a manner similar to a temperature inversion , but less extreme .
we observed the very hot jupiter corot-2b in secondary eclipse using three visits by the wfc3 g141 grism on hst . even without utilizing the new spatial scan mode , we obtained spectra with errors approaching the photon noise limit .
we characterized the instrument - related systematic effects present in the data .
we find a time - dependent variation in the background intensity , a visit - long slope , slopes associated with each orbit , and we investigate the ` hook ' effect that occurs after data transfers .
we explored the behavior , dependencies , and how best to account for these effects in data analyses .
in particular , we defined the amplitude of the hook effect versus the exposure level in electrons ( figure 6 ) .
we measure the thermal emission from the planet in the 1.1 - 1.7@xmath0 m band , but we find no spectral features to a @xmath3 limit of 85 ppm .
we used a differential method to derive the spectrum and cancel the systematic errors @xcite , obtaining results close to photon - limited .
no model fits all available eclipse data for this planet to within the errors .
we consider solar abundance and carbon - rich spectral models , as well as a simple blackbody spectrum , to account for the eclipse data .
the spectral models do not clearly surpass the blackbody spectrum in terms of the quality of the fit .
the slope of the data within the wfc3 bandpass is less than given by all of the models , including the best - fit blackbody .
there is weak and ambiguous evidence that the atmospheric temperature structure is inverted , but a reduced temperature gradient may be present , and may help to mimic the quasi - blackbody nature of the emergent spectrum .
extra atmospheric continuous opacity is a strong possibility to account for the lack of spectral features in the wfc3 band .
if that opacity has a scattering component , it can help to account for the optical eclipse amplitude of this planet as observed by corot .
spatial inhomogeneities in temperature on the star - facing hemisphere may also help to account for the optical eclipse and the slope of the spectrum in the wfc3 band . | we have used hubble / wfc3 and the g141 grism to measure the secondary eclipse of the transiting very hot jupiter corot-2b in the 1.1 - 1.7@xmath0 m spectral region .
we find an eclipse depth averaged over this band equal to @xmath1parts per million , equivalent to a blackbody temperature of @xmath2k .
we study and characterize several wfc3 instrumental effects , especially the `` hook '' phenomenon described by @xcite .
we use data from several transiting exoplanet systems to find a quantitative relation between the amplitude of the hook and the exposure level of a given pixel . although the uncertainties in this relation are too large to allow us to develop an empirical correction for our data , our study provides a useful guide for optimizing exposure levels in future wfc3 observations .
we derive the planet s spectrum using a differential method .
the planet - to - star contrast increases to longer wavelength within the wfc3 bandpass , but without water absorption or emission to a @xmath3 limit of 85 ppm .
the slope of the wfc3 spectrum is significantly less than the slope of the best - fit blackbody .
we compare all existing eclipse data for this planet to a blackbody spectrum , and to spectra from both solar abundance and carbon - rich ( c / o=1 ) models .
a blackbody spectrum is an acceptable fit to the full dataset .
extra continuous opacity due to clouds or haze , and flattened temperature profiles , are strong candidates to produce quasi - blackbody spectra , and to account for the amplitude of the optical eclipses .
our results show ambiguous evidence for a temperature inversion in this planet . |
the maser ( microwave amplification by stimulated emission of radiation ) was conceived based on thermodynamic considerations @xcite . in its micromaser implementation , coherent radiation is generated by inverted two - level ( rydberg ) atoms that are randomly injected into a microwave cavity one by one @xcite . for years
, the focus of micromaser studies had been on its quantum - electrodynamics features @xcite , including its extensions to the cooperative regime of multiatom clusters that are simultaneously present in the cavity @xcite , until scully et al .
@xcite revived the interest in the thermodynamics of such devices .
they treated the atomic beam as a thermodynamic resource , since randomly injected atoms , which are discarded ( traced out ) after they exit the cavity , constitute an effective reservoir ( bath ) for the cavity field mode ( in the markovian approximation ) .
the surprising finding of scully et al .
@xcite was that a beam of three - level atoms with coherence between two of its levels may be viewed as a nonthermal , quantum - coherent ( `` phaseonium '' ) bath that , given an appropriate phase @xmath0 of the interlevel coherence , can thermalize the cavity field to a temperature @xmath1 , where @xmath2 is the atoms temperature without coherence .
the dramatic consequence of the higher temperature attainable by the cavity field owing to the phaseonium coherence is a transgression of the nominal carnot efficiency bound in a heat engine , with the cavity field in the role of a working fluid ( wf ) : if the wf undergoes a cycle where it is coupled to the phaseonium bath in one stroke and to a cold bath at temperature @xmath3 in another , then the efficiency bound of the engine satisfies @xmath4 , instead of @xmath5 .
this landmark proposal has triggered a variety of proposals for engine schemes based on nonthermal baths that are capable of `` super - carnot '' operation @xcite , among them engines fuelled by a squeezed nonthermal bath @xcite . some of us have recently asserted @xcite that machines fuelled by nonthermal baths may be divided into two categories according to their operation paradigm : * _ machines of the first kind _ are those fuelled by a nonthermal bath , such as a squeezed - thermal or coherently - displaced thermal bath , that render the wf steady - state _ nonpassive _ @xcite .
such baths change the machine into a _
thermo - mechanical engine _ that , unlike a heat engine , is fuelled by both mechanical work and heat imparted by the bath to the wf
. the carnot bound may be transgressed in such machines at the expense of work supplied by the bath .
however , their efficiency bound can not be properly compared with the carnot bound , since the latter is a restriction imposed by the second law on heat @xcite but not on work imparted by the bath . *
_ machines of the second kind _ are those where the wf is thermalized by the nonthermal bath , as is the case of an engine fuelled by a phaseonium bath .
such a machine is a proper _ heat engine _ but the ability of the phaseonium bath to thermalize the wf to a temperature @xmath1 elevates its carnot bound above that associated with an incoherent bath at temperature @xmath2 . intriguingly , in micromaser setups , a beam of multiatom clusters has been shown to thermalize the cavity - field wf in some cases @xcite , but also coherently displace @xcite or squeeze it @xcite .
this implies that the wf may receive both work ( and thus become nonpassive ) and heat from the bath
. the cavity field may thus be the key ingredient in machines of the two kinds surveyed above . however , the criteria whereby atoms in a micromaser can fuel machines of either the first or the second kind are generally unknown , notwithstanding several recent results obtained along this line @xcite . here
we pose the question : to what extent is the quantum state preparation of multiatom clusters ( before they are injected into the cavity ) instrumental for determining not only the kind of machine we may operate but also the quantitative bounds of its performance ? figuratively speaking ,
if the multiatom cluster is the `` crude oil '' , the question is : which preparation of the cluster is the refining process that can deliver a `` gasoline '' with a `` specific octane '' ? to answer this question , we first derive ( in sec . [ sec : model ] ) a master equation that governs the cavity field under the standard assumption of a short interaction of each atom with the cavity field , @xmath6 , where @xmath7 is the atom field coupling strength and @xmath8 is the interaction time . in this regime , the steady - state density matrix of the cavity field may only be a gaussian state : thermal , coherently - displaced or squeezed @xcite .
this restriction is shown to imply ( sec .
[ sec : classification ] ) that two- or three - atom clusters suffice for the preparation of all gaussian states of the field , thus making larger clusters qualitatively redundant .
[ sec_squeezing ] and [ sec : results ] are devoted , respectively , to the classification of coherences or quantum correlations among the atoms that may be associated with ( i ) fuel for machines of the first kind that correspond to states whose coherences displace or squeeze the cavity field , as well as cause its heating ; and ( ii ) fuel for machines of the second kind which is purely `` combustible '' , i.e. , corresponds to atomic states that only allow for heat and entropy exchange with the field . in both sec . [ sec_squeezing ] and [ sec : results ] we identify highly promising multiatom states for each kind of fuel and infer the best parameters relevant to the machine operation . in sec .
[ sec : conclusions ] we discuss the results and propose viable experimental protocols for their implementation .
we consider a micromaser - type setup wherein the cavity - field mode is the working fluid ( wf ) that is energized ( fuelled ) by a beam of two - level atoms which are injected into the cavity at random , poisson - distributed , times @xcite . by contrast to the standard micromaser scenario @xcite , the atomic beam is here assumed to be composed of @xmath9-atom clusters that are prepared in a controlled quantum - correlated ( entangled ) state prior to their injection into the cavity , where they interact with the cavity field simultaneously ( see fig .
[ sec2fig1 ] ) . yet
, the replacement of single atoms by @xmath9-atom clusters does not change the basic premise of micromaser theory whereby their random injections allow to treat the atom clusters as an ergodic `` bath '' that is continuously coupled to the cavity mode , so that the latter is governed by a master equation ( upon tracing out this `` bath '' ) @xcite .
nevertheless , we will show that quantum coherence or interatomic quantum correlations ( entanglement ) in the cluster may crucially influence the dynamics of the cavity field .
experimentally feasible schemes for the present scenario will be discussed in sec .
[ sec : conclusions ] . in keeping with the standard assumptions of micromaser theory @xcite
, we take the transition time of the atoms through the cavity to be short enough to neglect atomic relaxation and dephasing as well as cavity loss , and to assume that there can be at most one cluster present in the cavity at a given time . under these standard assumptions
, we may derive a master equation for the dynamics of the cavity field .
the interaction of the atomic cluster with the cavity is described by the tavis
cummings model @xcite @xmath10 where the atom , cavity and interaction hamiltonians are respectively given by @xmath11 here @xmath12 are the annihilation and creation operators for the cavity field and @xmath13 are the @xmath14 , raising and lowering pauli operators for the @xmath15th atom with @xmath16 .
the atomic transition frequency @xmath17 is resonant with the cavity frequency @xmath18 .
the interaction between the atoms and the cavity is assumed to be spatially homogeneous with strength @xmath7 . under the foregoing assumptions ,
the combined system of the atomic cluster and the cavity field evolves unitarily during the short interaction time @xmath8 .
the unitary propagator @xmath19 in the interaction picture can be analytically computed to second order in @xmath20 ( see appendices [ app_one_atom][app_three_atoms ] ) . denoting the injection time of the @xmath15th cluster into the cavity by @xmath21 , the evolution of the reduced density operator of the field mode , which is obtained upon tracing out the atoms , reads @xcite @xmath22 \equiv s(\tau ) \rho(t_j ) .
\label{sec2eq4}\ ] ] here @xmath23 is the initial density operator of the atomic cluster and @xmath24 is a superoperator that propagates the cavity state @xmath25 to @xmath26 .
the atomic clusters arrive randomly at a rate @xmath27 and pass through the cavity within a time interval of @xmath28 with a probability of @xmath29 .
the field changes according to @xmath24 when a cluster is present and otherwise does not change at all , so that the overall change of the field state is @xmath30 for @xmath31 we obtain the master equation @xmath32 \rho(t ) \label{sec2eq6}\ ] ] which describes the markovian dynamics of the single - mode cavity @xcite .
here we do not include the usual cavity decay term , in order to clearly identify the role of coherences for the field evolution , particularly whether it thermalizes or not , but it is straightforward to do so .
the master equation can be rewritten as @xmath33^{\dagger}- \rho(t ) \right ] , \label{sec2eq9}\ ] ] where @xmath34 denote the matrix elements of @xmath35 . in the standard basis of energy - state products the diagonal elements @xmath36 are the populations and the off - diagonal elements @xmath34 with @xmath37 are coherences or correlations , respectively .
this standard basis is @xmath38 for one atom and shown in fig .
[ sec2fig2 ] for two ( @xmath39 ) and three atoms , respectively . using the explicit forms of the respective propagators @xmath40 for one- , two- and three - atom clusters [ eqs . , and in the appendix ]
, the master equation can be expressed in the illuminating form @xmath41 + \mathbb{l}_s \rho + \mathbb{l } \rho .
\label{sec : single : eq1}\ ] ] here the first term corresponds to the effect of a coherent drive applied to the cavity , which is described by the effective hamiltonian @xmath42 the lindbladian @xmath43 in eq .
describes a squeezing process and is given by @xmath44 where @xmath45 is an effective coupling rate .
the squeezing excitation and de - excitation lindbladians are @xmath46 respectively @xcite .
the lindbladian @xmath47 is given by @xmath48 where @xmath49 are the lindbladians for incoherent de - excitation and excitation , respectively .
the coefficients for different cluster sizes are shown in table [ table_coefficients ] . [
cols="^,^,^,^",options="header " , ] the master equation allows for the generation of arbitrary ( gaussian ) field states , i.e. , thermal- , displaced- and squeezed states .
higher - order ( i.e. , non - gaussian ) processes can not be induced by a second - order master equation .
the case @xmath50 is the minimal cluster generating these processes : adding another particle ( @xmath51 ) does not make a qualitative difference compared to eq . .
the key observation we infer from eq .
is that coherences or correlations in the multiatom cluster may be classified according to the disjoint blocks in the density matrix @xmath35 that are associated with qualitatively different terms in the master equation , each term giving rise to a different kind of field dynamics .
figuratively , the different coherences are different types of fuel for the cavity - field wf ( fig . [ sec2fig4 ] ) : * the blocks adjacent to the main diagonal of the @xmath35 matrix in the standard basis of fig . [ sec2fig2 ] contain coherences that can only induce absorption- and emission processes in the field ( wf ) , as they are associated to @xmath52 in the master equation .
we shall refer to these elements as _ heat - exchange coherences_. they have a caloric ( `` flammable '' ) value , i.e. , they may contribute to the thermalization of the cavity field .
heat - exchange coherences do not arise in the single - atom case , as they correlate states of the same energy , e.g. , @xmath53 in two - atom clusters and @xmath54 in three - atom clusters . * _ displacement coherences _ associated with the @xmath55 $ ] term in the master equation arise for all cluster sizes as they correlate states differing by one excitation , i.e. , @xmath56 and @xmath57 in single atoms , @xmath53 and its hermitian conjugate in two - atom clusters and , say , @xmath58 in three - atom clusters . * _ squeezing coherences _ correspond to an exchange of two excitations and may exist in two - atom clusters in the form of @xmath59 and its hermitian conjugate , or in three - atom clusters in , say , the form @xmath60 .
those matrix elements of @xmath35 that do not contribute to the field evolution shall be called _
ineffective coherences_. larger cluster sizes will not change the qualitative features of the master equation , which holds to second order in @xmath20 and thus may only induce the same second - order ( gaussian ) processes as listed above .
the different types of coherences and their relation to the number of excitations is illustrated in the tree diagram of fig .
[ sec2fig3 ] ( which should not be confused with the same term in graph theory ) .
the beam of atomic clusters interacting with a cavity mode can realize one of two operation paradigms @xcite : * if displacing or squeezing coherences are present in the bath , the cavity state becomes _ nonpassive _ ( displaced or squeezed , respectively ) , which implies that not only heat but also work has been transferred from the bath to the cavity mode .
consequently , a machine fuelled by such a bath is a _ machine of the first kind _ that operates _ thermo - mechanically_. * if the atomic state _ only _ contains heat - exchange coherences , the mode is _ thermalized _ by the bath and only heat is exchanged .
such a setup is thus a viable implementation of a _ heat engine _ powered by a nonthermal bath , which has been dubbed a _ machine of the second kind_. under what conditions
does the master equation possess a nonpassive ( nonthermal ) steady state of the cavity mode that is required for a machine of the first kind ( a thermo - mechanical machine ) ?
what is the nature of such a state ? to obtain an insight into these questions , we write the ( ehrenfest ) equations of motions of the field mean value , variance and mean intensity ( photon number ) in terms of the @xmath35 matrix elements that are grouped into the coefficients @xmath61 , @xmath62 , @xmath63 and @xmath64 as detailed in table [ table_coefficients ] , @xmath65 whose steady - state solutions read @xmath66 we see that a nonzero @xmath63 increases the thermal mean photon number by coherently displacing the cavity field to a nonzero expectation value [ eqs . and ] .
accordingly , the cavity field attains thermal - coherent character .
by contrast , a nonzero @xmath64 introduces quadrature squeezing to the cavity field and hence the cavity field acquires a squeezed - thermal character . as an example , consider the two - atom state @xmath67 for @xmath68 ( for simplicity we restrict the angle to @xmath69 ) this state gives rise , according to eq .
, to the master equation ( cf .
[ fig_states ] ) @xmath70,\ ] ] which may be cast into the standard form that yields thermal - squeezed solutions @xcite @xmath71,\ ] ] with the coefficients [ eq_master_squeezing_standard_coefficients ] @xmath72 here @xmath73 denotes the squeezing parameter and @xmath74 the ambient photon number . upon comparing eqs . ,
and we find @xmath75 in conventional experimental squeezing schemes , squeezing parameters range from @xmath76 @xcite up to @xmath77 @xcite .
remarkably , the squeezing parameter may greatly surpass existing values if we choose @xmath78 ( see fig . [ fig_squeezing ] ) .
of the two - atom state . ]
let us note that the existence of a steady - state solution ( whether coherent or not ) requires @xmath79 , implying @xmath68 .
we will discuss this point in more detail in what follows .
it is thus clear that thermal equilibrium of the cavity field requires @xmath80 so that all coherent processes vanish .
this can be achieved by either setting the displacement- and squeezing coherences to zero in fig .
[ sec2fig4 ] or making their respective contributions to the coefficients @xmath63 and @xmath64 ( see table [ table_coefficients ] ) to cancel each other . under these conditions that are required for thermalization of the cavity field
, the master equation reads @xmath81 where @xmath82 is the rate of emission of quanta into the bath and @xmath83 is the absorption rate of quanta from the bath . by virtue of the kubo
schwinger ( kms ) detailed - balance condition @xcite , it is then possible to attribute a temperature @xmath2 to the effective bath through @xmath84 this temperature @xmath2 is only positive and finite if @xmath79 .
let us pretend that this effective bath is composed of fictitious oscillators . then eq .
yields @xmath85 where @xmath86 is the effective thermal excitation of the bath ( mean number of fictitious quanta ) at the cavity frequency @xmath18 .
the denominator in eq . then corresponds to the sum of stimulated and spontaneous emission into the effective heat bath .
this effective description of the effective bath is vindicated by the steady - state solution of the master equation which is the thermal ( gibbs ) state @xcite @xmath87 of the cavity mode , where @xmath88 denotes the partition function . due to the unbounded character of the hamiltonian @xmath89 , this steady - state solution only exists if @xmath90 , i.e. , if @xmath79 .
the case @xmath91 , formally resulting in an infinite bath temperature , can be identified as the maser threshold @xcite ( see also appendix [ app_maser_threshold ] ) .
we have thus arrived at an important conclusion : a nonthermal beam of atoms interacting with a single cavity mode may act as an effective heat bath for the latter , thereby thermalizing it to a finite temperature @xmath2 , although the quantum state of the atoms may be distinctly _ nonthermal _ ,
i.e. , the _ atomic - cluster state is not associated with the notion of temperature_. nevertheless , it will drive the cavity field mode into a gibbs state with a finite and positive temperature provided the cavity mode is below the maser threshold .
this conclusion is consistent with the well - known fact that the regime below the micromaser threshold is thermal radiation with a thermodynamic equilibrium temperature @xcite . here , however , this temperature @xmath2 of the cavity field depends explicitly on the coherences and correlations of the atoms . in table
[ table_values ] we present the explicit dependence of the temperature @xmath2 , the steady - state photon number @xmath92 , and the micromaser threshold on the multiatomic density - matrix parameters from table [ table_coefficients ] . in
what follows , we focus on the relation between the correlations in distinctly entangled states of the cluster on the temperature and threshold conditions . ' '' '' @xmath93^{-1}$ ] & @xmath94^{-1}$ ] & @xmath95^{-1}$ ] & @xmath96^{-1}$ ] + ' '' '' @xmath97 & @xmath98 & @xmath99 & @xmath100 + ' '' '' valid for ( @xmath79 ) & @xmath101 & @xmath102 & @xmath103 + and @xmath104 , when @xmath105 and @xmath106 .
the symmetric w state yields the largest mean photon number in equilibrium and hence can be imagined as the `` hottest '' effective three atom reservoir among the w class states . ]
we may parameterize the singly - exited entangled states of three atoms via @xmath107 the mean photon number in terms of the angles then reads ( cf .
table [ table_values ] and note that @xmath108 and @xmath109 vanish ) @xmath110 the maximum photon number corresponds to @xmath105 and @xmath106 .
the variation of the mean number of photons with respect to the remaining parameters @xmath111 and @xmath104 is shown in fig .
[ sec3.1fig1 ] .
it is seen that the maximum number corresponds to the symmetric w state @xmath112 which is known for its robust entanglement @xcite . this state yields @xmath113 , which means that coherences ( correlations ) in the w state increase the photon number in thermal equilibrium from the value @xmath114 , which would be the case for a phase - averaged w state .
this amplification of photon population in the cavity is here due to dicke superradiance @xcite : the quantum interference in the w state enhances the processes described in the master equation , which equilibrate the cavity field to a canonical thermal state .
although each cluster is in a pure state , the entropy of the cavity increases via the partial - trace operation after each interaction , which removes the information about the atomic state .
the crucial contribution of the heat - exchange coherences present in the w state can be traced to the effective temperature @xmath115 that these coherences induce as to compared to the temperature obtained for its phase - averaged ( classically - correlated ) counterpart ( i.e. , without any heat - exchange coherences ) , @xmath116 .
this temperature @xmath117 is solely determined by the populations of the computational - basis states and may hence be thought of a `` classical '' effect . by contrast , the augmented temperature @xmath118 stems from the heat - exchange coherences ( that here lead to constructive quantum interference ) .
the deviation from @xmath117 is thus of quantum - mechanical origin .
therefore , we conclude that the symmetric w state provides the highest equilibrium temperature to the cavity field among the entangled singly - excited states .
this comes about since in the symmetric w state all contributions @xmath119 in eq .
add up coherently , allowing for cooperatively enhanced interaction in this three - particle dicke state @xcite .
equation only possesses a steady - state solution [ eq .
] if @xmath120 ( cf .
table [ table_values ] ) . this condition can be fulfilled for a generalized ghz state parameterized by @xmath121 , @xmath122 for which the mean photon number in thermal equilibrium becomes @xmath123 we have plotted the steady - state photon number as a function of @xmath121 in fig .
[ ghz_fig ] .
the figure shows that as @xmath124 one reaches the micromaser threshold , where the mean photon number diverges , signifying an infinite effective temperature of the bath .
let us now consider the case @xmath125 that yields the ghz state @xmath126 the only nonzero coherences are then @xmath127 and @xmath128 , which according to fig .
[ sec2fig4 ] are ineffective . indeed , the only nonvanishing parameters of the master equation are @xmath129 .
such parameters in the thermal lindblad equation @xmath130 correspond , according to the kms detailed - balance condition , to an _ infinite temperature _ of the bath .
a beam of ghz states thus can not thermalize a cavity mode .
indeed , the mean photon number in the cavity grows in an unbounded fashion , @xmath131 according to eq . .
similarly to the symmetric w state , a nearly symmetric ghz - type state ( cf .
[ fig_states ] ) is the optimal choice for reaching high cavity temperatures .
however , the mechanism is entirely different in the two cases .
the symmetry in the w state allows for constructive quantum interference ( superradiance ) , so that the enhancement is purely quantum - mechanical .
by contrast , the nearly - symmetric entangled ghz state allows us to approach the maser threshold .
the photon number only depends on the populations since , according to fig .
[ sec2fig4 ] , all coherences in the state are ineffective . hence , the phase - averaged counterpart of eq .
results in the same effective temperature as that of the state , i.e. , @xmath132 .
incoherently _ mix w states with states that we denote as e states that belong to the general class of three - atom w states with two excitations in the upper blue triangle in fig .
[ sec2fig3 ] ( which are also dicke states @xcite ) , @xmath133 according to table [ table_values ] they contribute to @xmath108 and @xmath109 and correspond to @xmath134 and @xmath135 , namely their rate of absorption surpasses the emission rate , leading [ by the kms detailed balance condition ] to a _ negative temperature _ , which is outside the scope of this paper .
the chosen nearly equal mixture of w- and e - states has the form @xmath136 where @xmath137 is a small positive number .
this state is a mixture of the two coherence triangles indicated by solid blue lines in fig .
[ sec2fig3 ] .
the corresponding nonvanishing parameters of the master equation @xmath138 and @xmath139 ( cf .
table [ table_coefficients ] ) imply that @xmath79 , so that this mixed state corresponds to a _ positive and finite _ effective temperature , @xmath140 .
the corresponding classical - like phase - averaged counterpart of eq . would thermalize the cavity to @xmath141 . here
, two enhancement factors are involved .
the first factor is @xmath142 ( cf . table [ table_values ] ) , enhancing @xmath117 to higher temperatures by the quantum interferences due to the coherences in the @xmath143 and @xmath144 states .
the second factor is the classical enhancement of @xmath117 due to the operation near the maser threshold .
we have studied the thermodynamic implications of a generalized micromaser model wherein the cavity mode interacts with a beam of quantum - coherent or quantum - correlated multiatom clusters .
our central goal has been to classify the states of such clusters prior to their injection into the cavity according to their ability to fuel the cavity field as `` working fluid '' in a machine of either the first kind ( thermo - mechanical engine ) or the second kind ( heat engine ) . to this end
we have derived a lindblad master equation for the cavity field mode that describes absorption- and emission of the field , its coherent displacement and squeezing caused by the atoms that may act , respectively , as a thermal , displaced - thermal or squeezed - thermal bath .
these distinct gaussian processes that the field may undergo are determined by the prefactors of the respective terms in the master equation that are , in turn , determined by disjoint blocks ( coherences ) of the multiatom density matrix .
the main results of our analysis are as follows : * an important insight that we have obtained is that two- and three - atom clusters are capable of acting as fuel for both kinds of machines in a highly effective fashion , so that there is no need to involve larger clusters .
still , a larger number of coherences as the cluster grows in size may further enhance the work output . * for machines of the first kind , our analysis has revealed a particularly promising , simple , fuel in the form of two - atom clusters whose state is a nearly equal superposition of doubly - excited and doubly - ground states .
such a state is expected to give rise to very large squeezing of the cavity field .
it may thus present a far superior alternative to existing squeezing schemes of cavity fields @xcite .
such a strong squeezing may have fascinating applications @xcite also outside of quantum thermodynamics .
our interest here is that this strong squeezing source may fuel a cavity field in a hybrid thermo - mechanical machine @xcite with nearly 100% efficiency , at the expense of mechanical work supplied by the two - atom clusters . * for machines of the second kind ,
we have found w states of three - atom clusters to act as conventional heat - baths fuel at a positive finite temperature that is controllable by the w state .
by contrast , three - atom ghz- and e - states have been found to correspond to effective baths at infinite or negative temperatures , respectively , that do not allow for a thermal steady - state solution for the cavity field . on the other hand ,
nearly - equal mixtures of w and e states have been identified as fuel capable of thermalizing the cavity field to an _
ultrahigh temperature_. to conclude , our results are potentially useful for the design of thermal and nonthermal machines based on micromaser setups .
the availability of all gaussian processes via preparation of two- and three - atom clusters allows to implement heat engines ( wherein the cavity field is thermalized ) but also thermo - mechanical engines ( wherein the cavity mode is coherently displaced or squeezed ) .
we wish to stress the feasibility of the diverse forms of state preparation of multiatom clusters ( prior to their injection into the cavity ) we have employed in our analysis : * the arsenal of quantum gate operations @xcite can in principle prepare two or three trapped atoms in an entangled state on demand , but such preparation may require single - atom addressability .
* alternatively , w states can be generated via quantum feedback control @xcite or at fusion - based light
matter interfaces @xcite .
multipartite entangled states may also be generated via photon - mediated interactions , as recently discussed in @xcite .
* another alternative is an optimized probabilistic scheme for multiatom entangled - state preparation in a cavity @xcite . * for two - atom entangled - state preparation we may resort to controlled diatomic dissociation @xcite , collisions in a cavity @xcite or long - range dipole
dipole interactions @xcite . on the fundamental side ,
our results provide clues to the thermalization or nonthermalization of a system ( here the cavity field ) via its contact with quantum - correlated multipartite clusters that act as nonthermal baths .
such processes reflect the subtle rapport between quantum correlations in the bath and thermalization @xcite .
c. b. d. is thankful for intuitive discussions with a. levent subasi and a. mit hardal .
. e. m. acknowledges support from ko university and lockheed martin corporation university research agreement .
g. k. acknowledges the isf and bsf for support .
. e. m. conceived the idea .
c. b. d. , . e. m. , and w. n. equally contributed to the analytical and numerical calculations .
all authors revised and wrote the paper in collaboration .
all authors have read and approved the final manuscript .
for one atom the time - evolution operator @xmath145 to second order in @xmath20 readily evaluates to @xmath146
the time - evolution operator @xmath147 of the joint cavity
atoms system can be computed to second order in @xmath20 using the collective angular - momentum operators @xmath148 such that @xmath149 the latter can be decomposed into irreducible subspaces by changing from the computational basis ( spanned by products of single - atom @xmath150 and @xmath151 states , cf .
[ sec2fig2 ] ) to the basis of dicke states @xcite by means of the transformation matrix @xmath152 the addition of two spin-@xmath153 gives rise to a triplet and a singlet , @xmath154 . as a consequence , @xmath155 and
the propagator is given by the direct sum @xmath156 where to second order in @xmath20 @xmath157 here @xmath158 denotes the unit matrix of the same dimensionality as @xmath159 .
explicitly , we find @xmath160 and @xmath161 transforming back to the computational basis @xmath39 yields @xmath162
for three particles one proceeds exactly like in the preceding section .
the transformation matrix now reads @xcite @xmath163 the addition of three spin-@xmath153 gives rise to a quadruplet and two doublets , @xmath164 . as a consequence , @xmath165 and
the propagator is given by the direct sum @xmath166 where @xmath167 and @xmath168 transforming back to the computational basis ( cf .
[ sec2fig2 ] ) yields the matrix elements @xmath169 of the time - evolution operator .
the remaining elements evaluate to zero .
it is illuminating to derive the threshold condition from the mean photon number , following ref .
we first take @xmath80 and thereby eliminate the coherent - displacement and squeezing terms . according to eq .
( [ sec2eq4 ] ) , the cavity density matrix will change to @xmath170 after the passage of the @xmath15th atom during the short interaction time @xmath8 .
the mean photon number can be calculated to be @xmath171 = k \braket{\hat{n}(t_j ) } + ( g\tau)^2 \frac{r_e}{2 } .
\label{seccc : eq3.3}\ ] ] the change of the mean photon number between consecutive injections of two atom clusters is determined by the increment ratio @xmath172 , which is given by @xmath173 assuming that the cavity is initially in the vacuum state , the last term in eq .
( [ seccc : eq3.3 ] ) yields the mean number of photons after the first - cluster passage , @xmath174 .
after the @xmath15th cluster passage the mean photon number rises to @xmath175 the summation in eq .
( [ seccc : eq3.7 ] ) is convergent if @xmath176 , which is equivalent to the threshold condition @xmath79 . as @xmath177 ,
the summation converges to eq . .
61ifxundefined [ 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/revmodphys.71.s263 [ * * , ( ) ] link:\doibase 10.1103/physrev.112.1940 [ * * , ( ) ] link:\doibase 10.1070/pu1996v039n07abeh000158 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.54.551 [ * * , ( ) ] link:\doibase 10.1103/physreva.34.3077 [ * * , ( ) ] link:\doibase 10.1038/35001526 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.58.353 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.82.3795 [ * * , ( ) ] link:\doibase 10.1103/physreva.71.053818 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.59.1899 [ * * , ( ) ] link:\doibase 10.1103/physreva.49.2933 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.74.900 [ * * , ( ) ] link:\doibase 10.1103/physreva.34.2032 [ * * , ( ) ] link:\doibase 10.1103/physreva.65.023809 [ * * , ( ) ] link:\doibase 10.1103/physreva.46.5913 [ * * , ( ) ] @noop _ _ ( , , ) link:\doibase 10.1063/1.1523786 [ * * , ( ) ] link:\doibase 10.1126/science.1078955 [ * * , ( ) ] http://stacks.iop.org/0295-5075/88/i=5/a=50003 [ * * , ( ) ] link:\doibase 10.1103/physreve.89.052132 [ * * , ( ) ] link:\doibase 10.1103/physreva.81.052121 [ * * , ( ) ] link:\doibase 10.1103/physreve.73.036122 [ * * , ( ) ] link:\doibase 10.1103/physreve.84.051122 [ * * , ( ) ] link:\doibase 10.1103/physreve.86.051105 [ * * , ( ) ] http://stacks.iop.org/0295-5075/106/i=2/a=20001 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.112.030602 [ * * , ( ) ] link:\doibase 10.1038/srep12953 [ * * , ( ) ] link:\doibase 10.1103/physreve.93.012145 [ * * , ( ) ] http://arxiv.org/abs/1508.06519 [ ( ) ] link:\doibase 10.1007/bf01614224 [ * * , ( ) ] link:\doibase 10.1007/bf01011769 [ * * , ( ) ] http://stacks.iop.org/0295-5075/67/i=4/a=565 [ * * , ( ) ] link:\doibase 10.1103/physreve.91.032119 [ * * , ( ) ] http://stacks.iop.org/0295-5075/103/i=6/a=60005 [ * * , ( ) ] link:\doibase 10.1103/physreve.90.022102 [ * * , ( ) ] link:\doibase 10.1016/bs.aamop.2015.07.002 [ * * , ( ) ] @noop _ _ , ed .
( , , ) @noop * * , ( ) @noop _ _ , ed .
( , ) link:\doibase 10.1088/1751 - 8113/40/28/s01 [ * * , ( ) ] @noop _ _ ( , , ) link:\doibase 10.1103/physrev.170.379 [ * * , ( ) ] @noop _ _ ( , , ) @noop _ _ ( , , ) @noop _ _ ( , ) link:\doibase 10.1103/physrevlett.92.153601 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.98.070502 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.104.251102 [ * * , ( ) ] link:\doibase 10.1103/physrev.159.208 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.68.127 [ * * , ( ) ] @noop _ _ ( , , ) link:\doibase 10.1103/physrev.93.99 [ * * , ( ) ] @noop _ _ ( , , ) @noop _ _ ( , , ) link:\doibase 10.1103/physreva.88.062311 [ * * , ( ) ] link:\doibase 10.1038/srep16245 [ * * ( ) , 10.1038/srep16245 ] link:\doibase 10.1103/physrevx.6.011032 [ * * , ( ) ] link:\doibase 10.1103/physreva.75.022327 [ * * , ( ) ] link:\doibase 10.1103/physreva.36.90 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.83.714 [ * * , ( ) ] link:\doibase 10.1103/physreva.87.033831 [ * * , ( ) ] | in this paper we address the question : to what extent is the quantum state preparation of multiatom clusters ( before they are injected into the microwave cavity ) instrumental for determining not only the kind of machine we may operate but also the quantitative bounds of its performance ? figuratively speaking , if the multiatom cluster is the `` crude oil '' , the question is : which preparation of the cluster is the refining process that can deliver a `` gasoline '' with a `` specific octane '' ?
we classify coherences or quantum correlations among the atoms according to their ability to serve as ( i ) fuel for nonthermal machines corresponding to atomic states whose coherences displace or squeeze the cavity field , as well as cause its heating ; and ( ii ) fuel which is purely `` combustible '' , i.e. , corresponds to atomic states that only allow for heat and entropy exchange with the field and can energize a proper heat engine .
we identify highly promising multiatom states for each kind of fuel and propose viable experimental schemes for their implementation . |
gamma ray bursts ( grbs ) have been deeply studied in the kev
mev energy range , but only little information in the gev range has been provided in the past decade by egret measurements .
only 3 bursts have been detected at energies @xmath21 gev , with one photon reaching 18 gev @xcite .
the study of the high energy emission of grbs could provide extremely useful data able to constrain the emission models and the value of the ambient parameters . at these high energies
the detection from space is hampered by the very low fluxes , requiring large collection areas . from ground
the last generation cherenkov telescope magic @xcite , designed also to point at the detected grb direction in a very short time , is still making efforts to lower the threshold at energies @xmath3100 gev .
its duty cycle and field of view are however very small : a wide field detector , able to cover simultaneously and continuously a significant ( @xmath01 steradian ) fraction of the sky , is thus necessary . with such a detector
the milagro collaboration reported evidence for tev emission from grb 970417a with a significance slightly greater than 3@xmath4 @xcite .
the study of the high energy spectrum of grbs is perhaps the strongest motivation for an all - sky vhe detector .
the study of transient phenomena can be successfully performed at energies down to 1 gev by air showers arrays working in `` single particle mode '' @xcite , i.e. , counting all the particles hitting the individual detectors during fixed time intervals .
the observation of an excess in coincidence with a grb detected by satellites would be an unambiguous signature of the nature of the signal . both the sensitivity and the energy threshold improve with larger detection areas and higher observation levels , making air shower detectors at very high altitude the most suitable . the argo - ybj experiment , located at the yangbajing cosmic ray laboratory ( 4300 m a.s.l . ) with a detection area of @xmath0 6700 @xmath5 , is an air shower array exploiting the full coverage approach at very high altitude , with the aim of studying the cosmic radiation with a low energy threshold . in this paper we present results on the search for grbs in coincidence with satellite detections performed in the december 2004 - may 2006 period with the argo - ybj experiment .
the argo - ybj detector is constituted by a single layer of resistive plate chambers ( rpcs ) with @xmath093@xmath6 of active area .
this carpet has a modular structure , the basic module being a cluster ( [email protected] m@xmath1 ) , divided into 12 rpcs ( [email protected] m@xmath1 each ) .
each chamber is read by 80 strips of 67.5@xmath7618 mm@xmath1 , logically organized in 10 independent pads of [email protected] cm@xmath1 which are individually acquired and represent the high granularity pixel of the detector @xcite .
the carpet is composed by 154 clusters for a total surface of @xmath06700 m@xmath1 .
the detector is connected to two different daq systems , which work independently : in shower mode , for each event which fulfill the trigger conditions the position and time of each detected particle is recorded , allowing the reconstruction of the lateral distribution and of the arrival direction @xcite ; in scaler mode , where there is no trigger , the counting rate of each cluster is measured every 0.5 s , with no measurement of the space distribution and arrival direction of the detected particles . in the scaler mode daq , for each cluster the signal coming from the 120 pads is added up and put in coincidence in a narrow time window ( 150 ns ) , giving the rate of counts @xmath91 , @xmath92 , @xmath93 , @xmath94 , read by four independent scaler channels .
the corresponding measured rates are , respectively , @xmath0 40 khz , @xmath0 2 khz , @xmath0 300 hz and @xmath0 120 hz for each cluster .
the counting rates for a given multiplicity are then obtained with the relation n@xmath10 = n@xmath11 - n@xmath12 for i = 1 , 2 , 3 .
the use of four different scalers may give an indication of the source spectrum in case of signal detection . in order to correctly handle the data
, it is very important to evaluate the response to particles hitting the detector . for scaler mode operations ,
the most important effect is the strip cross - talk , i.e. , the probability of having more than one strip fired by a single particle , giving fake coincident counts . due to the front - end logic
, this can happen only for strips belonging to different pads , since the maximum number of counts for each pad is 1 independently of the number of particles hitting simultaneously the pad .
an analytical calculation based on the measured `` occupancy '' , i.e. , the mean number of strips fired by 1 particle , has been made and checked experimentally . from the experimental point of view , it is important to take into account the background counting rate variations due to changes in environmental parameters such as the atmospheric temperature and pressure ( which modify the shower development in the atmosphere ) and the detector temperature ( instrumental effect ) .
more troublesome are other possible instrumental effects , such as the electronic noise , that could simulate narrow signals in time , producing spurious increases in the background rate . working in single particle mode
requires very stable detectors , and a very careful and continuous monitoring of the experimental conditions . by comparing the counting rate of the single detectors and requiring simultaneous and consistent variations in all of them , it is possible to identify and reject most of the fake excesses due to instrumental effects .
short variations in the single particle counting rate have been measured in coincidence with strong thunderstorms and have been ascribed to the effects of atmospheric electric fields on the secondary particles flux @xcite . the static electric field is measured on the roof of the argo - ybj building with an efm100 boltek atmospheric field monitor .
anyway , we note that the occurrences of these events are very rare and even in this case the observed time scales ( @xmath010 - 15 minutes ) are longer than the typical grb duration .
the study of the counting distribution for each cluster is important in order to monitor the stability of the detector and its statistical ( poissonian ) behaviour . in fig .
[ fig : csum ] the total counting rate of a typical cluster , added up on the 4 multiplicity channels during a period of 30 minutes , follows a poissonian distribution with a @xmath13 given by : @xmath14
in order to study the detector response to extensive air showers ( eass ) , a detailed mc simulation has been carried out for both protons and photons with fixed energies in the range 1 gev - 1 tev and zenith angles @xmath15 = 0@xmath16 , 10@xmath16 , 20@xmath16 , 30@xmath16 , 40@xmath16 .
the corsika / qgsjet code 6.204 @xcite has been used with a full electromagnetic component development down to e@xmath17= 0.05 mev for both electrons and photons and 50 mev for muons and hadrons .
a detailed description of the detector has been carried out to correctly simulate the `` cluster size '' , i.e. , the correlation between the number of particles hitting the detector and the number of signals generated in the different multiplicity channels .
since the actual efficiency depends essentially on the shower particle lateral distribution , a huge quantity of showers must be simulated over a very large area to completely contain it . to save the computing time
the shower sampling can be performed by means of the _ `` reciprocity technique '' _ @xcite . the sampling area a@xmath18 ( @xmath0 5000@xmath75000 m@xmath1 ) is uniformly filled with replicas of the same argo - ybj carpet , one adjacent to the other .
following the reciprocity concept , we sample the shower axis only over the area covered by the carpet located at the center of the array , with the prescription of considering the response of all the detector replicas . on an event - by - event basis
we calculate the number of clusters which contain more than 1 , 2 , 3 , 4 fired pads , summed on the entire grid .
fig.[fig : aeff_20 g ] shows the effective areas for primary photons and protons with zenith angle @xmath19 in the four multiplicity channels for the complete argo - ybj detector constituted by 154 clusters ( @xmath06700 m@xmath1 sensitive area ) .
we note that for a multiplicity @xmath20 the detector sensitivity does not depend on its geometrical features , like the area of the single counters or their relative positions , but only on the total sensitive area @xcite .
therefore , the effective areas for any carpet dimension can be scaled from the plotted values .
the effective areas for primary protons are then convoluted with the following spectrum : @xmath21 with @xmath22 @xcite and taking into account the local geomagnetic cutoff @xcite .
the resulting counting rates , considering an opening angle of 60 degrees around the zenith , are the following : 21 khz for n = 1 , 1.7 khz for n = 2 , 180 hz for n = 3 and 80 hz for n @xmath23 4 .
comparison with the measured rates , i.e. , 38 khz for n = 1 , 1.7 khz for n = 2 , 180 hz for n = 3 and 120 hz for n @xmath23 4 , shows that the values obtained by our simulations are lower in the multiplicity channels n = 1 and n @xmath23 4 . the discrepancy for n = 1
is expected because of dark counting and natural radioactivity . since from both of them
we expect mostly single counts , these effects are expected to influence only the @xmath91 scaler channel .
the search for emission from grbs started with the first grb detection by the swift satellite on december 17 , 2004 , when only 16 clusters ( @xmath0693 m@xmath1 of sensitive area ) out of the total 154 were in data taking . up to may 2006 ,
28 grbs detected by satellites were within the argo - ybj field of view ( for this search , @xmath24 ) .
because of detector installation and debugging operations , the duty cycle of data taking has been reduced and reliable data are available only for 16 of these grbs ( see table [ tab : grb ] ) . for every grb ,
the number of counts @xmath25 , recorded in each of the four multiplicity channels during the duration time @xmath26 measured by the satellites , is compared with the number @xmath27 expected from the background ( obtained from the average counting rate in @xmath28 around the burst ) . the difference @xmath29 in units of standard deviations , i.e. , @xmath30 , gives the statistical significance @xmath31 of the excess , which we report in column 8 of table [ tab : grb ] for @xmath20 .
the data analysis of 3 grbs ( grb051114 , grb060105 and grb060510a ) gives 2.8 , 3.6 and 3.7 as the statistical significance of the signal , respectively .
taking into account that we considered a sample of 16 grbs , these values correspond to a post - trial probability p(@xmath32 ) , p(@xmath33 ) and p(@xmath34 ) , respectively , of being a background fluctuation . as a consequence ,
no convincing excess in the scaler counts was observed in the duration time measured by the satellites .
therefore @xmath35 upper limits to the fluence of these events were calculated in the 1 100 gev energy range using the spectral indices determined at lower energies by satellites .
our results are reported in the last column of table [ tab : grb ] .
for those grbs whose redshift has been also determined , the upper limit was calculated including a model for @xmath36 absorption by the extragalactic background light ( ebl ) @xcite and the corresponding values printed in bold . for the other grbs z = 0 was assumed ( below 300 gev the @xmath36 absorption is almost negligible for [email protected] ) .
llccccccc grb & sat . & t90/dur . & @xmath38 & redshift & spectral & carpet & @xmath39 & ul@xmath40 + & & ( s ) & ( deg ) & & index & area ( m@xmath1 ) & & ( fluence ) + 041228 & swift & 62 & 28.1 & & 1.56 & 693 & -1.3 & 3.3@xmath4110@xmath42 + 050408 & hete & 15 & 20.4 & 1.24 & 1.98 & 1820 & -2.2 & * 9.6@xmath4110@xmath43 * + 050509a & swift & 12 & 34.0 & & 2.1 & 1820 & 0.29 & 1.6@xmath4110@xmath42 + 050528 & swift & 11 & 37.8 & & 2.3 & 1820 & -0.012 & 6.5@xmath4110@xmath42 + 050802 & swift & 20 & 22.5 & 1.71 & 1.55 & 1820 & 0.74 & * 1.0@xmath4110@xmath42 * + 051105a & swift & 0.3 & 28.5 & & 1.33 & 3379 & 0.90 & 1.4@xmath4110@xmath43 + 051114 & swift & 2 & 32.8 & & 1.22 & 3379 & 2.8 & 1.9@xmath4110@xmath43 + 051227 & swift & 8 & 22.8 & & 1.31 & 3379 & 0.93 & 2.5@xmath4110@xmath43 + 060105 & swift & 55 & 16.3 & & 1.11 & 3379 & 3.6 & 5.9@xmath4110@xmath43 + 060111 & swift & 13 & 10.8 & & 1.63 & 3379 & 0.82 & 2.5@xmath4110@xmath43 + 060115 & swift & 142 & 16.6 & 3.53 & 1.76 & 4505 & -2.2 & * 2.3@xmath4110@xmath42 * + 060421 & swift & 11 & 39.3 & & 1.53 & 4505 & -0.46 & 1.6@xmath4110@xmath42 + 060424 & swift & 37 & 6.7 & & 1.72 & 4505 & 1.9 & 4.1@xmath4110@xmath43 + 060427 & swift & 64 & 32.6 & & 1.87 & 4505 & -1.8 & 1.8@xmath4110@xmath42 + 060510a & swift & 21 & 37.4 & & 1.55 & 4505 & 3.7 & 2.3@xmath4110@xmath42 + 060526 & swift & 14 & 31.7 & 3.21 & 1.66 & 4505 & 0.75 & * 1.2@xmath4110@xmath42 * + @xmath44 zenith angle .
+ @xmath45 significance of the signal for the single event .
+ @xmath46 upper limit on the fluence ( 1 100 gev ) in erg @xmath47 .
the numbers in bold take into account absorption by the ebl .
a search for vhe emission from grbs has been performed with an increasing detector area of the argo - ybj experiment .
a total of 16 satellite - triggered grbs in the field of view ( @xmath48 ) of argo - ybj in the dec .
2004 - may 2006 period has been analyzed .
no significant emission was detected and typical fluence upper limits of @xmath49 erg @xmath47 in the 1 100 gev energy range were obtained using the measured counting rates and grb parameters determined by the satellite observations .
we expect to increase the sensitivity by a factor @xmath02 converting the secondary photons with a 0.5 cm thick layer of lead .
catelli , j.r .
et al . : in : c.a .
meegan ( ed . ) , aip conf . proc .
428 , p. 309
( aip , new york , 1998 ) .
albert , j. et al .
: apj lett . *
641 * l9 ( 2006 ) .
atkins , r. et al .
: apj lett . * 553 * l119 ( 2000 ) .
vernetto , s. : astrop .
* 13 * 75 ( 2000 ) .
aielli , g. et al .
: nim * a562 * 92 ( 2006 ) .
di sciascio , g. et al . : 29th icrc , pune , * 2 * , 33 - 36 ( 2005 ) .
aglietta , m. et al . : 26th icrc , salt lake city , * 7 * , 351 - 354 ( 1999 ) .
heck , d. et al .
: report * fzka 6019 * forschungszentrum karlsruhe , 1998 .
battistoni , g. et al .
: astrop . phys .
* 7 * 101 ( 1997 ) .
gaisser , t.k . , honda m. : ann .
nucl . part .
sci . * 52 * 153 ( 2002 ) .
storini m. , smart d.f .
, shea m.a . : in : 27th icrc , hamburg , * 10 * 4106 - 4109 ( 2001 ) .
kneiske , t.m .
a@xmath50a * 413 * 807 ( 2004 ) . | the argo - ybj experiment is almost completely installed at the yangbajing cosmic ray laboratory ( 4300 m a.s.l . ,
tibet , p.r .
china )
. the lower energy limit of the detector ( e @xmath0 1 gev ) is reached with the scaler mode , i.e. , recording the single particle rate at fixed time intervals . in this technique , due to its high altitude location and large area ( @xmath0 6700 m@xmath1 )
, this experiment is the most sensitive among all present and past ground - based detectors . in the energy range under investigation
, signals due to local ( e.g. solar gles ) and cosmological ( e.g. grbs ) phenomena are expected as significant enhancements of the counting rate over the background .
results on the search for grbs in coincidence with satellite detections are presented .
example.eps gsave newpath 20 20 moveto 20 220 lineto 220 220 lineto 220 20 lineto closepath 2 setlinewidth gsave .4 setgray fill grestore stroke grestore |
over the past few decades , a considerable number of studies have been performed on 3@xmath1 transition - metal oxides which have a considerably narrow 3@xmath1 band . in particular , a metal - to - insulator transition caused by a strong electron correlation@xcite ( mott transition ) as well as anomalous electronic properties in the metallic phase near the mott transition have attracted the interest of many researchers . since the discovery of the high-@xmath2 cuprate superconductors , the importance of two types of experimental approaches to the mott transition
have been discussed intensively : a filling control and a band - width control .
the former consists in doping holes or electrons to the system , and the latter in varying the strength of the electron correlation @xmath3 , where @xmath4 is the electron correlation due to coulomb repulsion and @xmath5 is the one - electron band - width . in recent years , the systematic evolutions of optical conductivity spectra in going from a correlated metallic phase to the mott - hubbard insulating phase have been reported on both the filling controlled and the band - width controlled mott - hubbard systems.@xcite the smallest energy gap for charge excitations of the mott - hubbard insulator is the excitation energy of the charge fluctuation @xmath6 , so - called a mott - hubbard gap.@xcite the optical conductivity of the mott - hubbard insulator is considered to show a gap feature due to the above charge excitation from the lower hubbard band to the upper hubbard band .
v@xmath7o@xmath8 and related compounds have been extensively studied as typical materials which show the mott transition with varying the strength of the @xmath3 ratio .
the temperature dependent optical conductivity of v@xmath7o@xmath8 was reported by thomas _
et al_. through the metal - to - insulator transition.@xcite the optical conductivity of v@xmath7o@xmath8 in the insulating phase shows a gap feature , which is attributed to the mott - hubbard gap excitation . on the other hand , in the metallic state , a low - energy contribution to
the optical conductivity shows an anomalous feature , which is reproduced by two lorentzians .
moreover , the optical conductivity shows an anomalous enhancement of the spectral weight as a function of temperature .
et al_. reported that these experimental results are in good agreement with the theoretical prediction obtained by the infinite - dimension hubbard model within the mean - field approach.@xcite since the formation of the mott - hubbard bands are predicted even in the metallic state for this kind of strongly correlated system , the optical conductivity spectra should be affected by the precursor features.@xcite it is very interesting to see how the spectral weight varies with the electron correlation in the correlated metallic state near the mott transition .
for a detailed discussion , however , we need to control the strength of the electron correlation more precisely .
the perovskite - type early 3@xmath1 transition - metal oxides are ideal mott - hubbard systems for controlling the band filling and band - width by chemical substitutions .
it has been reported that , for the filling controlled systems la@xmath9sr@xmath10tio@xmath8 and @xmath11ca@xmath10tio@xmath8 ( @xmath12=rare earth ) , the spectral weight of the optical conductivity transfers from the higher energy feature corresponding to an excitation through the mott - hubbard gap , to the mid - infrared inner - gap region corresponding to the drude - like absorption extending from @xmath13.@xcite the rate of the spectral weight transfer by doping increases systematically with the increase of the one - electron band - width @xmath5 . on the other hand , the ti
bond angle can be decreased as we decrease the ionic radius of the @xmath12 site .
the decrease of the ti o ti bond angle gives rise to the decrease of the value of @xmath5 .
a systematic change of the optical conductivity spectra was reported on @xmath12tio@xmath8 ( @xmath12=la , ce , pr , nd , sm , and gd).@xcite the lowest gap - like feature systematically increases as the ionic radius of the @xmath12 site decreases , namely , as the value of @xmath5 decreases .
a similar change was also observed on an alloy system , la@xmath9y@xmath10tio@xmath8.@xcite these systematic variations of the optical conductivity are interpreted as the successive increase of the mott - hubbard gap with the increase of the strength of the @xmath3 ratio . in these materials
, however the system remains an insulator even for the least correlated latio@xmath8 , therefore one can not study the evolution of the metallic properties under the band - width control in this system .
the purpose of this paper is to clarify the evolution of the optical conductivity spectrum in the metallic phase near the mott transition , as we control the strength of the electron correlation @xmath3 without changing the band filling .
a perovskite - type @xmath14 vanadate is considered to be a strongly correlated metal close to the mott transition.@xcite we can control the strength of the @xmath3 ratio , by chemical substitution of a sr ion for a ca ion of the same valence without varying the nominal @xmath15-electron number per vanadium ion.@xcite we report the optical conductivity spectra in this strongly correlated alloy system .
the effective mass estimated from the optical measurements are shown in sec .
[ subsec : mass ] .
the evolution of the optical conductivity is discussed in sec .
[ subsec : spectral ] .
single crystals of ( @xmath16=0 , 0.25 , 0.5 , 1 ) were grown by a floating - zone method using an infrared image furnace with double halogen lamps .
details on the preparation of these samples are described in the preceding paper.@xcite since as - grown samples are slightly oxygen deficient , all the samples were annealed in air at 200 @xmath17c for 24 hours in order to make the oxygen concentration stoichiometric.@xcite raman scattering spectra were measured at room temperature in back - scattering geometry using a triple spectrometer system ( jasco trs-600 ) equipped with a charge coupled device ( ccd ; photometrics tk512cb ) cooled by liquid nitrogen .
the samples were excited by the 514.5 nm ar ion laser line .
polarization of the incident light was taken to be parallel to that of the scattered light .
optical reflectivity measurements were carried out at room temperature ( @xmath18300 k ) in the energy range between 0.05 ev and 40 ev using a michelson - type fourier - transform infrared spectrometer ( 0.050.6 ev ) , a grating monochrometer ( 0.55.6 ev ) , and a seya - namioka - type grating for the synchrotron radiation ( 540 ev ) at the beamline bl-11d of photon factory , tsukuba . the surfaces of the samples were mechanically polished with diamond paste for the optical measurements .
the absolute reflectivity was determined by referring to the reflectivity of an al or ag film which was measured at same optical alignment .
we have calculated a complex dielectric function @xmath19 by the kramers - kronig ( k - k ) transformation of the measured reflectivity @xmath20 , where @xmath21 is the photon energy . the real part of the complex optical conductivity re@xmath22 $ ] is related to the imaginary part of the dielectric function @xmath23 by re@xmath22= ( \omega/4\pi)\epsilon_{2}(\omega)$ ] .
since the k - k analysis requires @xmath20 for @xmath24 , two assumptions must be made to extrapolate the observed data beyond the upper and lower bounds of the measurements . in the present study ,
the reflectivity data were first extrapolated from the lowest measured energy down to @xmath13 with the hagen - rubens formula which is an approximation for @xmath20 of conventional metals .
then , beyond the highest measured energy , the reflectivity data were extrapolated up to @xmath25 with an asymptotic function of @xmath26 .
powder x - ray diffraction measurements were carried out to characterize the samples and to determine the lattice constants . in fig .
[ f.lattice ] , we present the lattice constants , @xmath27 , @xmath28 , and @xmath29 against the sr content @xmath16 .
the crystal structure of belongs to the perovskite - type structure with orthorhombic distortion ( gdfeo@xmath8-type).@xcite the amount of the distortion is almost proportional to the amount of the ca content ; _ i.e. _ , is a cubic perovskite .
however , we assumed the crystal structure of all samples ( @xmath30 ) orthorhombic and deduced the lattice constants .
the lattice constants increase monotonously with the increase of @xmath16 , ensuring the appropriate formation of the solid solution over the whole composition range .
raman scattering measurements give indirect information about the crystal symmetry , because the appearance of some raman - active phonon lines depends crucially on the crystal symmetry of the system . in fig .
[ f.raman ] , we show the raman spectra of at room temperature in the wave - number range of 80400@xmath31 . in this wave - number range , four raman active @xmath32 phonon lines are observed in orthorhombic ( @xmath33 ) .
the energies of the phonon lines shift to the lower energy side , and the width of the peaks become broader , as we increase the sr content @xmath16 .
the raman active phonon lines disappear completely in . according to the group theory analysis,@xcite
it is predicted that several raman active phonon modes @xmath34 can exist in the orthorhombically distorted perovskite which belongs to the point group symmetry of @xmath35 . on the other hand ,
the cubic perovskite which belongs to the point group symmetry of @xmath36 is `` raman forbidden '' , namely it has no raman - active phonon modes .
therefore , the experimental results tell us that the crystal symmetry actually changes from the orthorhombic distorted perovskite ( ) to the cubic perovskite ( ) . the orthorhombic to cubic transition is considered to occur between @xmath37 and 0.75 .
as discussed in the preceding paper,@xcite the orthorhombic distortion implies that the v o v bond angle is deviated from 180@xmath38 , _
i.e. _ , there is an alternately tilting network of the vo@xmath39 octahedra .
the v o v bond angle of is 180@xmath17 same as an ideal perovskite structure , while that of is @xmath18160@xmath17 .
the bond angle deviation from 180@xmath38 reduces the overlap between the neighboring v 3@xmath1 orbital mediated by the o 2@xmath40 orbital .
therefore , the one - electron band - width @xmath5 of v 3@xmath1 band decreases with decreasing the v o v bond angle .
accordingly , we can control the value of @xmath5 by chemical substitution of the ca@xmath41 ion for the sr@xmath41 ion of the same valence without varying the nominal @xmath15-electron number per vanadium ion . since the electron correlation energy @xmath4 are almost the same in and , we can thus control the strength of the @xmath3 ratio by the chemical substitution .
the v o v bond angle of , in addition , is almost equal to that of latio@xmath8 which is a mott - hubbard type insulator , so that it seems reasonable to consider that is close to the mott transition . moreover , there are many other evidences of the strong electron correlation in this system discussed so far.@xcite thus , system is ideal for the study of the metallic states near the mott transition . in fig .
[ f.reflection ] , we show optical reflectivity spectra of measured at room temperature ( @xmath18300 k ) , for four single crystals with different @xmath16 ( @xmath16=0 , 0.25 , 0.50 , 1 ) . the chemical substitution of sr@xmath41 for ca@xmath41 seems to make no remarkable change at the lower - energy region ( below @xmath185 ev ) in the optical reflectivity spectra .
all the samples exhibit high reflectivity from far - infrared to near - infrared region , and we can recognize a sharp reflectivity edge appearing at @xmath181.3 ev .
since is metallic over the whole composition range , the optical reflectivity is dominated by the signal of conduction electrons in this photon energy region .
systematic spectral changes with @xmath16 are observed in the energy range of the ultra - violet and vacuum - ultra - violet light .
the changes are partly due to the differences in the conduction bands of the ca@xmath41 and sr@xmath41 cations . but
this is irrelevant for the discussion of the main subject .
first of all , we will concentrate on the low energy response of the itinerant carriers .
the contribution of the conduction electrons to the complex dielectric function @xmath42 is well described by the drude model . according to the generalized drude model,@xcite @xmath42
is expressed as @xmath43 where @xmath44 is the high - energy dielectric constant , which is a high - energy contribution of the interband transitions , @xmath45 is the complex conductivity , @xmath46 is the energy - dependent scattering rate , and @xmath47 is the plasma frequency . the plasma frequency @xmath48 is defined as @xmath49 where @xmath50 is the total density of conduction electrons , and @xmath51 is the energy - dependent effective mass .
to begin , let us confine our attention to the plasma frequency .
if we assume that the nominal electron number per vanadium ion is exactly 1 for the whole composition range , we can deduce the carrier density @xmath50 from the unit - cell volume .
then , we can estimate a variation of the effective mass @xmath51 from the value of @xmath47 .
the energy - loss function im(@xmath52 ) is obtained by the kramers - kronig analysis of the measured reflectivity spectra @xmath20 .
provided that @xmath46 and @xmath51 do not depend on @xmath21 strongly , we can estimate @xmath53 from the energy - loss function , because , the energy - loss function peaks at the energy of @xmath54 ( @xmath55 ) . accordingly , we can obtain the energy - independent plasma frequency @xmath56 from the peak position of the energy - loss function . in fig .
[ f.im1e ] , the spectra of im(@xmath52 ) of are shown in the photon energy range from 0.6 to 2.0 ev to focus on the peak near the reflectivity edge ( around 1.3 ev ) .
the peak position of the energy - loss function @xmath54 systematically shifts to higher energy with increasing @xmath16 . at first ,
we have estimated @xmath54 from the peak energy , and deduced @xmath57 .
if we consider only the response of the conduction electrons , @xmath44 is the contribution from the high - energy interband transitions .
since the interband transition appeared above @xmath582.5 ev , the value of @xmath44 can be taken from the real part of the dielectric function @xmath59 at around 2.5 ev.@xcite since the value of @xmath60 is nearly independent of @xmath16 , we have used @xmath61 over the whole composition range . then , we can deduce the value of using the lattice constants and the value of @xmath56 . in fig .
[ f.masswp ] , the ratios of the deduced effective mass @xmath62 to the bare electron mass @xmath63 are plotted as a function of the sr content @xmath16 .
the value of @xmath64 systematically increases as varying @xmath16 from to .
this carrier - mass enhancement , however , is not so large , even though the system is near the mott transition .
this result is consistent with the value of estimated from the results of the specific heat measurements.@xcite it is instructive to compare the measured low frequency @xmath45 with the simple drude model , in which @xmath46 and @xmath51 do not depend on @xmath21 . according to eq .
( [ epsilon ] ) , the real part of the optical conductivity re@xmath22 \equiv \sigma(\omega)$ ] is given by the formula @xmath65 where @xmath66 is the dc conductivity .
the dc conductivity is expressed by the scattering rate @xmath67 and the plasma frequency @xmath56 by the following relation : @xmath68 here , we have used the value of @xmath66 obtained by electric resistivity measurements at room temperature .
then the value of @xmath67 can be deduced from the above relation .
[ f.nondrude ] shows the comparison of the experimentally obtained @xmath69 for to the optical conductivity calculated by the simple drude model . as shown in fig .
[ f.nondrude ] , the contribution to the optical conductivity below @xmath581 ev , which is considered to be a response of the itinerant carriers , is not properly reproduced with the simple drude model especially above 0.2 ev .
as we increase the photon energy , the experimentally obtained @xmath69 deviates from that of the simple drude model . the observed @xmath69 has a tail decaying slower than the drude - type @xmath70 dependence .
we consider that the discrepancy between the simple drude model and the experimental results is attributed to the energy - dependence of the scattering rate @xmath46 and the effective mass @xmath51 .
the @xmath46 corresponds to the renormalized scattering rate @xmath71 , which is @xmath72 .
the quantity @xmath73 is closer to the microscopic intrinsic quasiparticle lifetime at @xmath74 .
we can determine @xmath46 and @xmath75 from eq .
( [ epsilon ] ) ; _ i.e. _ , when we define @xmath19 , @xmath76 @xmath77 \\ % & = & \frac{4{\pi}ne^{2}}{\omega^{2}}\ , \frac{\epsilon_{\infty}-\epsilon_{1}(\omega ) } { ( \epsilon_{\infty}-\epsilon_{1}(\omega))^{2}+\epsilon_{2}^{2}(\omega)}.\end{aligned}\ ] ] figs .
[ f.exdrudemg](a ) and [ f.exdrudemg](b ) show @xmath46 and @xmath51 of as a function of photon energy . in case of the simple drude model , the scattering rate @xmath67 is provided to be independent on the photon energy .
but , in this system , @xmath46 actually increases as we increase the photon energy , as shown in fig .
[ f.exdrudemg](a ) .
the energy - dependent scattering rate is generally based on the electron - phonon scattering and the electron - electron scattering . since an extremely large t@xmath78-dependence of the dc conductivity observed in the system
can be well ascribed to the electron - electron scattering,@xcite it is reasonable to consider that the electron - electron scattering governs the behavior of @xmath46 . according to the fermi liquid theory ,
the electron - electron scattering rate is proportional to @xmath79 .
[ f.exdrudemg](a ) , however , indicates that @xmath46 looks more proportional to @xmath21 rather than @xmath79 . on the contrary , since the electron - phonon scattering is proportional to @xmath80 up to the debye frequency , it is necessary to elucidate the scattering process which contributes to @xmath46 .
this is still an open question . on the other hand ,
the energy dependence of @xmath51 is not so large . except for the low energy region ( @xmath81 ev ) , the value of @xmath82 increases with the decrease of sr content @xmath16 . in the sr@xmath9la@xmath83tio@xmath84 system , which is a typical doping system ,
as one approaches @xmath85 , a large energy dependence of as well as a critical enhancement at the low - energy region are observed.@xcite however , in system , does not exhibit such a critical enhancement with varying @xmath16 in going from to , although there is a difference between the filling control and the band - width control . in the low energy limit ( @xmath13 )
, @xmath62 should correspond to the effective mass estimated by the specific heat measurement .
we have interpolated @xmath75 down to @xmath13 with two kinds of tangential lines drawn from 0.4 ev and 0.15 ev . as shown in fig .
[ f.massexdrude](a ) , the intercepts , at which the two tangential lines from 0.4 ev and 0.15 ev cut the vertical axis , are defined as @xmath86 and @xmath87 . fig .
[ f.massexdrude](b ) indicates @xmath16-dependence of the values of @xmath86 and @xmath87 . the value of @xmath87 does not show a systematic behavior , because phonons , randomness , or other extrinsic contributions possibly cause this non - systematic change
however , the value of @xmath86 increases systematically with decreasing @xmath16 ; moreover , the values are almost equal to the value of @xmath62 deduced from plasma frequency .
we regard @xmath86 as a good measure of @xmath64 for this system .
the effective mass estimated from the plasma frequencies and the generalized drude analysis ( @xmath86 ) appear in table [ t.mass ] . .
effective mass @xmath88 deduced from the plasma frequencies @xmath56 and the generalized drude model ( @xmath86 ) .
[ t.mass ] [ cols="^,^,^,^,^",options="header " , ] it is expected that we should observe , near the mott transition , a critical enhancement of the effective mass of the 3@xmath1 conduction electrons . if we substitute the ca@xmath41 ion for the sr@xmath41 ion in the system , the 3@xmath1 band - width successively decreases .
then , the value of @xmath88 is expected to increase drastically reflecting the change of the @xmath3 ratio .
but we have observed that such a large mass enhancement does not actually take place in this system .
this is consistent with the result shown in the preceding paper.@xcite the density of states ( dos ) of orthorhombic and cubic calculated using the full - potential augmented plane - wave method with the local density approximation ( lda ) are shown in the top of fig .
[ f.dosandoc ] .
the band - calculation shows that the dos near the fermi level @xmath89 is dominated by the v 3@xmath1 electrons .
the v 3@xmath1 band crosses the fermi level , and the dos below 4 ev is mainly the o 2@xmath40 band . in the metallic states , @xmath69 is expected to consist of two basic components : intraband transitions within the v 3@xmath1 conduction band , _
i.e. _ , the drude part extending from @xmath13 , and interband transitions appearing at much higher energy .
the latter is regarded from the calculated dos as the charge - transfer contribution ( an excitation from the o 2@xmath40 band to the unoccupied part of the v 3@xmath1 band above @xmath89 ) .
a corresponding schematic picture of the optical conductivity is shown in the bottom of fig .
[ f.dosandoc ] .
as seen in the picture , the charge - transfer contribution is expected to appear above @xmath184 ev , and the absorption edge of the charge - transfer transition in is considered to shift slightly to lower energy than that of , reflecting the shift of the o 2@xmath40 band . based on this picture , let us now look at the experimental results , fig .
[ f.opticon ] shows the real part of the optical conductivity , @xmath69 , of the single crystals ( @xmath16=0 , 0.25 , 0.50 , 1 ) .
the optical conductivity spectra are different from our naive schematic picture ( fig .
[ f.dosandoc ] bottom ) ; they show the presence of two anomalous features in the intraband transition part below 4 ev besides the drude - like absorption ( discussed above ) : a small peak which appears at @xmath181.7 ev and a large peak at @xmath183.5 ev .
it must be noted that the two peak - like structures below 4 ev have no naive origin as far as we can infer from the calculated dos ( fig .
[ f.dosandoc ] ) .
this large spectral weight redistribution is generally believed to be a manifestation of the strong electron correlation in this system .
[ f.ctback ] shows a comparison of the optical conductivity spectra of to those of other perovskite oxides , ( lightly doped 3@xmath90 metal),@xcite and ( 3@xmath0 insulator ) reported by okimoto _
et al_.@xcite in the optical conductivity of , the most prominent low - energy feature , that distinctly rises around 4 ev , can be interpreted as originating in a transition from the o 2@xmath40 band to the ti 3@xmath1 band , which corresponds to the optical gap of the parent insulator srtio@xmath8.@xcite the doped 3@xmath1 electrons contribute to @xmath69 with a small spectral weight extending from @xmath13 . on the other hand ,
is considered to be a mott - hubbard insulator .
two electronic gap - like features are observed around 1 ev and 4 ev . these features have been respectively interpreted as originating in excitations through the mott - hubbard gap , namely , from the lower - hubbard band ( lhb ) to the upper - hubbard band ( uhb ) , and an excitations through the charge - transfer gap , _
i.e. _ , from the o 2@xmath40 band to the uhb.@xcite recently , bouarab _
et al_. have reported interband optical conductivities obtained by the energy - bands calculation of the y_m_o@xmath8 ( _ m_=ti - cu ) system with a local spin - density approximation.@xcite their calculated results of interband optical conductivity in is shown in the bottom of fig .
[ f.ctback ] as a shaded portion .
we find out from this comparison that the peak at around 5 ev can not be explained by the transition between the o 2@xmath40 band and the ti 3@xmath1 band alone . in
, photoemission spectroscopy@xcite has revealed that the o 2@xmath40 band is located at a binding energy which is almost the same as that of metallic ; hence , the absorption edge of the charge - transfer excitation of should be approximately equal to that of .
therefore , it is reasonable to consider that the shaded portions of @xmath69 in fig .
[ f.ctback ] correspond to the charge - transfer type transitions as well as the other interband transitions with much higher energies , on the analogy of the band - calculation in .@xcite accordingly , the remaining white portions correspond solely to the intraband transition within the v 3@xmath1 band . in order to focus on the spectral weight of the optical conductivity arising from intra-3@xmath1-band transitions , we have subtracted the shaded portion in the middle of fig .
[ f.ctback ] as backgrounds , assuming an appropriate function of @xmath91 , where @xmath92 has been obtained by fitting the lower energy tail of the o 2@xmath40 band in photoemission spectroscopy spectra of single crystals.@xcite a quantitative measure of the spectral weight has been obtained by deducing the effective electron number per vanadium ion defined by the following relation @xmath93 where @xmath94 is the bare electronic charge and @xmath95 is the bare band mass of a non interacting bloch electron in the conduction band .
@xmath96 is the cell volume for one formula unit ( one v atom in this system ) .
the significance of will be appreciated by considering the sum rule of the conductivity @xmath97 where @xmath98 corresponds to the total number of electrons in the unit formula .
that is , ( @xmath21 ) is proportional to the number of electrons involved in the optical excitations up to @xmath21 . in fig .
[ f.neff ] , we show of the system , after subtracting the higher - energy background . since , in fig [ f.neff ] , we have assumed @xmath99 , where @xmath63 is the bare electronic mass , the total number @xmath100 results smaller than 1 , reflecting the difference between @xmath95 and @xmath63 ( @xmath101 ) .
if we rather use the value @xmath95 obtained from lda , @xmath102 for the v 3@xmath1 band , we find @xmath103 .
thus , we conclude that the assumed background ( shaded area in fig .
[ f.ctback ] ) is reasonable to deduce the intrinsic contributions of the interband transition .
the initial steep rise of is due to the drude - like contributions extending from @xmath13 .
the drude - like contribution can be distinguished below @xmath181.5 ev , where exhibits a flat region .
therefore , at @xmath181.5 ev is considered to be a good measure for the effective mass of the carries .
the values of @xmath64 estimated from ( @xmath104 ev ) are 3.1(6 ) , 3.0(5 ) , 3.0 , 2.7 for @xmath16 = 0 , 0.25 , 0.50 , 1 , which are almost equivalent to the values of @xmath105 discussed in sec .
[ subsec : mass ] . fig .
[ f.peakab ] shows the optical conductivity spectra @xmath69 of the single crystals ( @xmath16=0 , 0.25 , 0.50 , 1 ) in the photon energy range of 0@xmath185 ev .
the high - energy background corresponding to the interband transition is subtracted . as discussed above , @xmath69 ( @xmath106 )
reflects only the intraband transition of the v 3@xmath1 electrons . in the spectrum
, there is a small feature at @xmath181.7 ev , which we call a peak `` a '' [ fig .
[ f.peakab](a ) ] and also a large feature at @xmath183.5 ev , which we call a peak `` b '' [ fig . [ f.peakab](b ) ] . with the increase of @xmath16 , the excitation energy of the peak `` b '' shifts slightly to lower energy , and
its spectral weight decreases ; whereas , the excitation energy of the peak `` a '' shifts to higher energy , and its spectral weight increases .
in addition , the width of the peak `` a '' broadens with the increase of @xmath16 .
[ f.abshift ] shows the excitation energy , the full - width at half maximum ( fwhm ) , and the spectral weight of the peak `` a '' and the peak `` b '' as functions of @xmath16.@xcite in the valence band photoemission spectra of the system , two features have been observed : one is a peak at @xmath181.5 ev below @xmath89 and the other is the emission from a broad quasiparticle band which lies on @xmath89.@xcite the former is assigned to an incoherent emission associated with the formation of the lower hubbard band and the latter corresponds to a renormalized 3@xmath1 band at @xmath89 .
et al_. reported that , upon increasing the strength of @xmath3 in system , the spectral weight is systematically transferred from the quasiparticle band to the incoherent part.@xcite in the inverse - photoemission spectra of and , morikawa _
et al_. have found a prominent peak at 2.5@xmath183 ev above @xmath89 and a shoulder within around 1 ev of @xmath89.@xcite these features have been also assigned to the incoherent and coherent parts of the spectral function of the v 3@xmath1 electron .
these results lead us into consideration that the two features ( the peak `` a '' and `` b '' ) observed in the optical conductivity spectra should originate in possible combinations of the transitions among the incoherent and coherent features of v 3@xmath1 electron around the fermi level .
the experimental results of the optical conductivity can be compared to the theoretical prediction obtained by the self - consistent local - impurity approximation of the infinite - dimension hubbard model.@xcite the theory seems to give us a clue to understand the origin of the two features : the peak `` a '' and the peak `` b '' . according to the prediction , the optical response is composed of basically three contributions except the drude part : a broad part centered at a frequency @xmath107 , a few narrow features near @xmath108 , and an `` anomalous '' part that is present in the range @xmath13 to 1 ev approximately .
the contribution at @xmath4 corresponds to direct excitations between the hubbard bands , the features at @xmath109 corresponds to excitations from the lhb to the empty part of the quasiparticle band and from the filled part of the quasiparticle band to the uhb , and finally , the `` anomalous '' part corresponds to excitations from the filled to the empty part of the quasiparticle band . in our previous paper,@xcite
we analyzed the spectrum of in the light of these predictions .
the parameters @xmath4 and @xmath5 , which were used in the model calculation , were taken from the results of photoemission spectroscopy@xcite . although it was expected that the parameter @xmath5 would systematically change with composition ,
we chose to vary @xmath4 for the sake of simplicity , given that fits of equivalent quality could be obtained for the photoemission spectra . in , which has the narrowest 3@xmath1 band in the system , the peaks `` a '' and `` b '' have been well described by the features at @xmath109 and @xmath4 , respectively , so that the infinite - dimension hubbard model seems to reproduce the experimentally obtained optical conductivity reasonably well.@xcite we find that our new experimental results as summarized in fig .
[ f.abshift ] : with the increase of @xmath16 , the peak energy of the peak `` b '' shifts slightly to lower energy side and its spectral weight gradually increases .
the spectral weight of the peak `` a '' increases with @xmath16 .
the peak `` a '' , however , shifts slightly to higher energy .
we shall now like to emphasize an important point .
our present systematic study of the compound gives conclusive evidence that we can tune the band - width of the system by controlling @xmath16 .
the position of the peak `` b '' gives a direct measure of the value of @xmath4 , and the fact that it remains almost a constant is a clear evidence that the ratio @xmath3 is controlled by a change of the band - width @xmath5 .
this situation is in sharp contrast with our previous analysis@xcite based on photoemission data which did not allow us to resolve which parameter was actually controlling the @xmath3 ratio .
a crucial ingredient that makes the study of the optical response so valuable for this analysis is that , unlike photoemission , it probes also the unoccupied part of the spectra , therefore , it is sensitive to the relative position of the hubbard bands . in order to gain some further insight in the qualitative behavior of the systematic evolution of our experimental data ,
we have used our initial estimates for @xmath4 and @xmath5 as input parameters in a calculation of the optical response of the hubbard model and changed the value of @xmath5 instead of @xmath4 .
we shall consider the model within the dynamical mean field theory which becomes exact in the limit of large lattice connectivity ( or large dimensionality ) .
for convenience we have computed the optical response using the iterated perturbation theory ( ipt ) method which allows for a simple evaluation of this quantity at @xmath74 and near the mott - hubbard transition.@xcite in fig .
[ f.largedoc ] we show the theoretical prediction using the value of @xmath110 ev for the local repulsion and for the half - bandwidth @xmath111 and 0.95 ev for and , respectively .
note that the spectra do not display the drude contribution as it corresponds to a delta - function at @xmath13 since our model does not contain disorder and the calculation is performed at @xmath74 .
the particular lineshape that we obtain is originated in the behavior of the spectral density of states that is obtained within the ipt method as shown in fig .
[ f.largeddos ] . we should point out that while the details of the line shape may not be correctly given by this method , the main distribution of the spectral weight of the various contributions and their systematic evolution are very reliably captured.@xcite we observe that the theoretical results for the systematic dependence of the various contributions to the optical response by controlling the value of @xmath5 are in a better qualitative agreement with the experimental data of fig . [ f.abshift ] than the previous calculation where we changed the value of @xmath4 .
one of the most notable improvements consists in that the unexpected systematic evolution of the feature at @xmath109 , which shifts upward with increasing @xmath5 , is well qualitatively captured .
this peculiar effect can be interpreted as a `` band repulsion '' between the hubbard band and the quasiparticle band .
as we increase @xmath5 , the latter becomes broader and `` pushes '' the hubbard band further out . however , there are still some discrepancies .
such a large spectral weight redistribution is predicted to be concomitant with a large effective mass in the mean - field treatment of the hubbard model.@xcite this is , however , inconsistent with the observed effective mass in this system .
moreover , in the optical conductivity spectra , there is apparently a notable discrepancy in respect to the relative spectral weight of the peak `` a '' to the peak `` b '' .
it is remarkably suppressed in the experimental data , compared with the theoretical data .
other discrepancies between some experimental results and the prediction of the mean - field approach for the electron correlation were also reported in the photoemission spectroscopy measurements in this system.@xcite in the mean - field fermi liquid approach , the renormalized quasiparticle band at @xmath89 should be narrowed with increasing the value of @xmath3,@xcite but , in those experiments , the quasiparticle band - width remains broad , even if the system approaches to the mott transition .
since the peak `` a '' has been assigned to the transitions associated with the quasiparticle band , the conspicuous suppression in spectral intensity of the peak `` a '' reflects the broadness of the quasiparticle band .
the broad quasiparticle band also accounts for the lack of the strong mass enhancement in this system .
as discussed in the preceding paper,@xcite the momentum - dependent self - energy becomes significant near the mott transition , resulting in a reduction of the mass enhancement .
although our measurement can not clarify the validity of introducing a momentum - dependent self - energy , we conclude that there must be other interactions not present in the mean - field treatment of the electron correlation in the metallic regime close to the mott transition .
finally , the presence of the `` anomalous '' contribution at low frequencies that extends down to @xmath112 in the theoretical data sheds a different light for the interpretation of the drude - like response discussed in sec .
[ subsec : mass ] .
it may be possible to say that the deviation from the simple drude model would be partly due to this `` anomalous '' contribution .
the origin of this effect is again traced to the presence of the incoherent contribution coming from the low energy tails of the hubbard bands that is observed in the theoretical density of states . however ,
it is experimentally very difficult to disentangle unambiguously the contribution of the coherent optical response of carriers and that of the incoherent process in optical conductivity , so this issue remains an open question .
this study has aimed at elucidating the electronic structure of the correlated metallic vanadate by means of the optical spectroscopy measurements . we have synthesized the system to control solely the 3@xmath1 band - width without varying the band filling .
we have found that the low energy contribution to the optical conductivity spectra can not be reproduced by the simple drude model with the energy - independent scattering rate and effective mass .
the energy - dependent @xmath46 determined by the generalized drude model shows relatively large energy dependence .
however , @xmath46 is proportional to @xmath21 rather than that of the electron - electron scattering @xmath79 .
the effective mass of the v 3@xmath1 electron has been evaluated from the plasma frequency .
the value of @xmath64 gradually increases with decreasing the band - width @xmath5 .
however , any symptom of the critical mass enhancement has not been observed , even though the system is close to the mott transition .
we observed two anomalous peaks in the optical conductivity spectra around 1.7 ev and 3.5 ev . these features can be assigned to the possible combinations of transitions between the incoherent and coherent bands of quasiparticles around the fermi level .
this large spectral weight redistribution substantiates the strong electron correlation in this system , which is , however , not concomitant with a large effective - mass enhancement .
we would like to thank the staff of the photon factory for technical support .
thanks are also given to y. nishihara and f. iga for helpful suggestions and experimental supports .
99 present address : electrotechnical laboratory , tsukuba 305 - 8568 , japan . to whom correspondence should be addressed .
( electronic address : )
. n. f. mott , _ metal - insulator transitions _
( taylor and francis , london , 1974 ) . y. fujishima , y. tokura , t. arima , and s. uchida , phys . rev .
b * 46 * , 11167 ( 1992 ) .
y. taguchi , y. tokura , t. arima , and f. inaba , phys .
b * 48 * , 511 ( 1993 ) .
m. kasuya , y. tokura , t. arima , h. eisaki , and s. uchida , phys . rev .
b * 47 * , 6197 ( 1993 ) .
d. a. crandles , t. timusk , j. d. garrett , and j. e. greedan , phys .
b * 49 * , 16207 ( 1994 ) .
t. katsufuji , y. okimoto , and y. tokura , phys .
lett . * 75 * , 3497 ( 1995 ) .
d. a. crandles , t. timusk , and j. e. greedan , phys .
b * 44 * , 13250 ( 1991 ) ; d. a. crandles , t. timusk , j. d. garrett , and j. e. greedan , physica c * 201 * , 407 ( 1992 ) . y. okimoto , t. katsufuji , y. okada , t. arima , and y. tokura , phys . rev .
b * 51 * , 9581 ( 1995 ) .
g. a. thomas , d. h. rapkine , s. a. carter , a. j. millis , t. f. rosenbaum , p. metcalf , and j. m. honig , phys .
* 73 * , 1529 ( 1994 ) ; g. a. thomas , d. h. rapkine , s. a. carter , t. f. rosenbaum , p. metcalf , and d. f. honig , j. low temp . phys . * 95 * , 33 ( 1994 ) .
m. j. rozenberg , g. kotliar , h. kajueter , g. a. thomas , d. h. rapkine , j. m. honig , and p. metcalf , phys .
lett . * 75 * , 105 ( 1995 ) .
j. zaanen , g. a. sawatzky , and j. w. allen , phys .
55 * , 418 ( 1985 ) .
a. georges , g. kotliar , w. krauth , and m. j. rozenberg , rev .
mod . phys . * 68 * , 13 ( 1996 ) .
i. h. inoue , k. morikawa , h. fukuchi , t. tsujii , f. iga , and y. nishihara , jpn .
. phys . * 32 * , 451 ( 1993 ) .
i. h. inoue , o. goto , h. makino , n. e. hussey , and m. ishikawa , preceding paper , phys .
b ( submitted ) .
i. h. inoue , i. hase , y. aiura , a. fujimori , y. haruyama , t. maruyama , and y. nishihara , phys .
lett . * 74 * , 2539 ( 1995 ) .
a. fukushima , f. iga , i. h. inoue , k. murata , and y. nishihara , j. phys .
. jpn . * 63 * , 409 ( 1994 ) .
n. shirakawa , k. murata , h. maikino , f. iga , and y. nishihara , j. phys .
soc . jpn . * 64 * , 4824 ( 1995 ) .
b. l. chamberland and p. s. danielson , j. solid state chem . * 3 * , 243 ( 1971 ) .
m. couzi and p. v. huong , j. chim . phys . *
69 * , 1339 ( 1972 ) .
j. w. allen and j. c. mikkelsen , phys .
b * 15 * , 2952 ( 1977 ) .
b. c. webb , a. j. sievers , and t. mihalisin , phys .
* 57 * , 1951 ( 1986 ) .
in fact , we should take into account a small contribution of a peak around 1.7 ev in the @xmath59 spectra .
the peak corresponds to the `` a '' peak in the optical conductivity . but here , we have disregarded the small peak .
h. makino _ et al .
_ , unpublished
. m. cardona , phys .
rev . * 140 * , a651 ( 1965 ) . t. arima , y. tokura , and j. b. torrance , phys .
b * 48 * , 17006 ( 1993 ) .
s. bouarab , a. vega , and m. a. khan , j. phys . : condens .
matter * 9 * , 6267 ( 1997 ) .
k. morikawa , t. mizokawa , k. kobayashi , a. fujimori , h. eisaki , s. uchida , f. iga , and y. nishihara , phys .
b * 52 * , 13711 ( 1995 ) .
i. h. inoue , h. makino , i. hase , y. aiura , y. haruyama , and y. nishihara , physica b * 230 - 232 * , 780 ( 1997 ) .
we have divided the peak `` a '' and `` b '' by fitting the @xmath23 data with two gaussian . and the spectral weight have defined as an amount of the contributions to @xmath113 from the peak `` a '' and `` b '' . m. j. rozenberg , i. h. inoue , h. makino , f. iga , and y. nishihara , phys .
lett . * 76 * , 4781 ( 1996 ) .
a. georges and g. kotliar , phys .
b * 45 * , 6479 ( 1992 ) .
x. y. zhang , m. j. rozenberg and g. kotliar , phys .
. lett . * 70 * , 1666 ( 1993 ) .
w. f. brinkman and t. m. rice , phys .
b * 2 * , 4302 ( 1970 ) . | optical conductivity spectra of single crystals of the perovskite - type 3@xmath0 metallic alloy system have been studied to elucidate how the electronic behavior depends on the strength of the electron correlation without changing the nominal number of electrons .
the reflectivity measurements were made at room temperature between 0.05 ev and 40 ev .
the effective mass deduced by the analysis of the drude - like contribution to the optical conductivity and the plasma frequency do not show critical enhancement , even though the system is close to the mott transition .
besides the drude - like contribution , two anomalous features were observed in the optical conductivity spectra of the intraband transition within the 3@xmath1 band .
these features can be assigned to transitions involving the incoherent and coherent bands near the fermi level .
the large spectral weight redistribution in this system , however , does not involve a large mass enhancement . |
diluted magnetic semiconductors ( dms ) of iii - v or ii - vi types have been obtained by doping semiconductors with magnetic impurities ( mn for example)@xcite .
these materials are very interesting due to their potential applications for spintronics@xcite .
however , the low curie temperature ( @xmath5 ) has limited their interest@xcite ( for example ga@xmath0mn@xmath1as , with @xmath6 has a @xmath7@xcite ) .
based on the theoretical works of dietl _
et al._@xcite , several groups @xcite have studied the growth of co - doped zno films@xcite which is a good candidate having a high @xmath8 @xcite .
using pulsed laser depositions ( pld ) , ueda _ et al .
_ reported ferromagnetism ( fm ) above room temperature@xcite , while jin _
et al . _ found no indication of fm by utilizing laser molecular beam epitaxy@xcite .
this controversy between research teams may result from the growth method used and/or from the growth conditions ( oxygen pressure , deposition temperature , etc ... ) . in the particular case of the pld technique
, it may also arise from the targets preparation and this parameter has never been considered up to now .
one of the reason is that the control of the dopant incorporation would be quite difficult to obtain using a pre - doped ceramic oxide target@xcite .
this is a crucial point since the properties of the dms are very sensitive to the percentage of dopant@xcite .
the homogeneity of dopant incorporation as well as the precise control of the growth might be responsible for the changes in the physical properties of the films obtained by the different groups .
thus , we have recently developed an accurate method to grow the zn@xmath0co@xmath1o films with a precise doping by using an alternate deposition starting from zn and co targets . using this procedure@xcite ,
we have been able to observe ferromagnetism at room temperature in one sample having the composition zn@xmath9co@xmath10o confirming previous results@xcite but in contrast to others@xcite . in this paper , we have grown a series of zn@xmath0co@xmath1o films with very low doping of co. we have studied the structural and magnetic properties of the films and our results are reported in this communication .
the zn@xmath0co@xmath1o films were grown using the pulsed laser deposition technique .
metal zinc ( @xmath11 ) and cobalt ( @xmath11 ) targets were used as purchased ( neyco , france ) without further preparations .
the films are deposited using a krf laser ( @xmath12)@xcite on ( 0001 ) al@xmath3o@xmath13 substrates ( in this paper , we used the three - index notation instead of the four - index one@xcite ) .
the substrates were kept at a constant temperature in the range 500@xmath14c-750@xmath14c during the deposition which was carried out a pressure around @xmath15 of pure oxygen . after deposition , the samples
were slowly cooled to room temperature at a pressure of @xmath16 of o@xmath3 .
the deposition rate is @xmath17 and the energy density is close to @xmath18 .
the composition of the film was checked by energy dispersive scattering and rutherford backscattering spectometry ( rbs ) .
the structural study was done by x - ray diffraction ( xrd ) using a seifert xrd 3000p for the @xmath19 scans and the @xmath20scan to evaluate the full - width - at - maximum ( fwhm ) .
an xpert phillips was utilized for the _ in - plane _ measurements obtained from the ( @xmath21 ) reflection .
ion channeling technique ( with 2 mev he@xmath22 ions ) was used to study the epitaxial nature of the films .
magnetization ( @xmath23 ) was recorded as a function of the temperature ( @xmath24 ) and the magnetic field ( @xmath25 ) in a squid magnetometer .
zno thin film was deposited using the optimal conditions found previously@xcite .
the films are highly crystallized as seen from the sharp diffraction peaks ( fig.1 ) and the fwhm of the rocking curve close to 0.25@xmath26 ( fig.4 ) .
the two diffractions peaks observed around @xmath27 and @xmath28 are characteristic of the hexagonal zno wurtzite , the @xmath29-axis being perpendicular to the substrate plane .
the out - of - plane lattice parameter is calculated to be @xmath30 which corresponds to the theoretical bulk one@xcite .
the epitaxial relationships between zno films and al@xmath3o@xmath4 substrates are determined using asymmetrical xrd .
the inset of fig.1 displays the @xmath31scan both of the zno films obtained from the ( @xmath21 ) planes and the al@xmath3o@xmath4 substrate obtained from the ( @xmath32 ) planes .
the peaks belonging to the film are separated by 60@xmath14 , indicating a six - fold symmetry in agreement with the hexagonal structure of zno whereas the peaks of the sapphire substrate are separated of 120@xmath14 indicating a three - fold symmetry in agreement with the rhombohedral symmetry of al@xmath3o@xmath4 .
note that 30@xmath14 spacing between the diffraction peaks of the film and those of the substrate indicate a rotation of 30@xmath14 between the in - plane axes . in order to obtain additional information , on the structural properties of the zno films , we have determined the strains of the optimized film .
this technique used the distance between atomic plane of a crystalline specimen as an internal strain gage@xcite . using this model , we minimized the strains of the zno film by changing the growth conditions @xcite . at a temperature of 600@xmath33 under @xmath15 of o@xmath3
, we found@xcite that the value of residual stress ( @xmath34 ) along the in_-plane _ direction ( @xmath35 ) is about the same value along the _ out of - plane _ direction ( @xmath36 ) .
such values are in agreement with previous report @xcite .
the study of the ( @xmath37 ) planes indicates a concentration gradient along the _ out - of - plane _ direction . in other words , the strains is larger near the interface and decreases along the _ out - of - plane _ direction as the thickness increases .
indeed , when the thickness of the film , the substrate - induced strain becomes lower and the lattice parameters of films become closer to the value of the bulk . moreover
, the positive slope@xcite indicates an extensive stress _ in - the - plane _ of the substrate which is in agreement@xcite with the decrease of the _ out - of - plane _ lattice parameter as compared to the bulk value@xcite .
these conditions have been used to grow the zn@xmath0co@xmath1o films with various doping x. an example of the ion backscattering channeling data for zn@xmath0co@xmath1o ( x=4% ) is given in fig.2 .
the observed minimum yield @xmath38 of @xmath395% for channeling reflects good quality of the film , thereby indicating optimum growth of zn@xmath0co@xmath1o films .
note also that a surface peak of co is seen which reveals that the top layer ( about 400@xmath40 ) has large amount of co ( @xmath41 ) .
the channeling spectrum also shows the co signal .
this indicates that the co in this layer has not gone substitutional .
fig.3 shows the @xmath19 scans recorded around the @xmath42 reflection in the the range @xmath43 .
as the co content is increasing there is a shift toward the high angle , indicating a decrease of the _ out - of - plane _ lattice parameter up to @xmath44 . above this value ,
the position of the diffraction peak is moving toward the low values of 2@xmath45 , leading to an increase of the lattice parameter up to @xmath46 .
then , the _ out - of - plane _ lattice parameter becomes constant@xcite ( not show in this graph ) .
to explain this results , we consider the zno bulk wurtzite structure , where zinc ( zn@xmath47 ) is in tetrahedral site with a ionic radius of @xmath48@xcite .
the substitution of the zn@xmath47 with co@xmath47 ( ionic radius : @xmath49@xcite ) leads to a compression of the bulk structure and a decrease of the lattice parameters.on the contrary , a higher coordination number of co@xmath47 leads to an ionic radius of co@xmath47 larger than the ionic radius of zn@xmath47 . in this configuration
, co@xmath47 would be in an interstitial site inducing an expansion of the structure ( and an increase of the lattice parameters ) and thus , an increase of defects . in the thin film , due to the substrate - induced strains , this will lead to a decrease of the _ out - of - plane _ lattice parameter ( up to @xmath44 of co ) and then to an increase of this parameter ( @xmath50 ) .
this explanation ( substitution and then interstitial co ) is also confirmed by looking at the evolution of the fwhm of the rocking curve as a function of the co content ( see fig.4 ) where the same tendancy is observed .
for very low doping of co ( @xmath51 ) , the fwhm is slightly decreasing up to @xmath52 of co. above this value ( @xmath53 ) , the fwhm increases when the co content is also increasing indicating that for these values of doping , the films are less oriented and/or contained more defects .
we investigated the magnetic properties of these thin film samples .
fig.5 shows the @xmath54 recorded for various zn@xmath0co@xmath1o film .
the films at low doping of co ( @xmath44 and @xmath55 ) clearly evidence a ferromagnetic state with a curie temperature respectively around @xmath56 and @xmath57 . for the other concentration of co ,
the situation is more confusing .
we have performed @xmath58 at different temperatures but we did not observed any hysteresis .
we also measure the magnetization in very low doping ( @xmath59 ) but the signal of the substrate is too high .
we discuss now the possibility of the co clusters in our films .
the transition from the ferromagnetic state to the paramagnetic state is clearly seen , suggesting that the metallic co clusters ( the @xmath8 of the metal co clusters is above 1000k ) are not responsible for the effect observed at 300k @xcite . moreover , the saturation moment ( 0.7 @xmath60/mole co ) is very weak compared to @xmath61 of metallic co@xmath62}$ ] , suggesting that the co state should be close to co@xmath47 . we believe that this is due to the technique used in the study where not only the conditions of the deposition minimize the strains but also the alternate deposition from the two targets favors the homogeneity of the doped films .
moreover , it has been seen that the low temperature ( 600@xmath33 is our case ) of deposition leads to homogenous films@xcite .
in conclusion , we have developed an alternative method for the growth of pulsed laser deposited oxide thin films .
this method permits an accurate control of the dopant in the matrix .
firstly using this procedure , we have deposited high quality zn@xmath0co@xmath1o thin films on al@xmath3o@xmath4 ( 0001 ) substrates with a low doping of co. secondly , we have also studied the evolution of the structure and the magnetic properties of the films for various co concentration .
we suggest that the incorporation of co inside the wurtzite structure leads to an increase of interstitial defects while the epitaxial defects decrease .
finally , the growth of these ferromagnetic films opens the route for the fabrication of spin - based electronics since this original method can be used to grow various oxide thin films .
we acknowledge the centre franco - indien pour la promotion de la recherche avancee / indo - french centre for the promotion of advance research ( cefipra / ifcpar ) for financial support ( project n@xmath142808 - 1 ) .
a.f . also thanks the cnrs and the conseil regional de basse normandie for his bdi fellowship .
a family of planes in the four - index notation is represented by @xmath63 where @xmath64 and becomes in the three - index notation @xmath65 @xmath66 @xmath67 .
see for example : _ fundamentals of crystallography _ , c.giacavazzo , oxford university press ( 2002 ) .
figure 4 : rocking curve ( @xmath71-scan ) recorded around the @xmath42 reflection for a series of zn@xmath0co@xmath1o film.the percentage of co is indicating .
the fwhm is increasing from 0.23@xmath72for pure zno to 0.43@xmath73 for zn@xmath0co@xmath1o with x=5% . | we report on the synthesis of low doping zn@xmath0co@xmath1o ( @xmath2 ) thin films on ( 0001)-al@xmath3o@xmath4 substrates .
the films were prepared in an oxidizing atmosphere , using the pulsed laser deposition technique starting from zn and co metallic targets .
we first studied the influence of the strains of zno and their stuctural properties .
second , we have investigated the structural and the magnetic properties of the zn@xmath0co@xmath1o films .
we show that at low doping , the lattice parameters and the magnetization of the zn@xmath0co@xmath1o films depend strongly on the co concentration . |
extragalactic jets at the parsec scale are present in numerous active galactic nuclei ( agn ; see a review by zensus 1995 ) .
impressive progress has been made by the very large baseline interferometry ( vlbi ) , and details still closer to the central engine are expected with the advent of millimeter vlbi .
already , a lot of information can be gained from the structure of these jets with typically 1 parsec resolution .
it has been possible to detect motions within a few years in about 100 sources ( vermeulen 1995 ) , a lot of them being superluminal with apparent speeds up to @xmath0 .
the motions detected are those of blobs moving on curved trajectories . generally speaking , these paths are wiggling , reminiscent of more or less helical lines seen in projection , and apparently different from one blob to the other , and the blob velocities vary along the trajectory ( zensus 1995 ; qian et al .
1996 ) .
not much information is available on the nature of the blobs themselves .
they are very generally believed to be shock fronts , because i ) shock waves are expected in these jets and ii ) they are an excellent means of accelerating particles through the first - order fermi acceleration process as has been worked out in the kpc scale jets . recently , in a series of papers , gmez et al .
( 1993 , 1994a , 1994b ) performed numerical simulations of a vlbi jet where the blobs are shock fronts traveling along a helical relativistic jet . nevertheless , the reality of these shock fronts is far from established .
the hypothesis of a relativistic jet is also debatable .
firstly , due to the compton drag close to the black hole , it is very difficult to extract a jet with lorentz factors higher than 2 or 3 ( phinney 1987 , henri & pelletier 1991 ) .
secondly , at the kpc scale , jets are probably non- or only mildly - relativistic ( e.g. parma et al .
1987 , fraix - burnet 1992 ) .
some authors conclude that the jets should decelerate ( bowman et al .
1996 ) from super- to subluminal speeds , but obviously , the lost energy should be observed in a manner or in another .
an interesting alternative to relativistic shocked jets is the two - fluid concept , in which the bulk of the jet ( electrons and protons ejected from the accretion disk in the form of a collimated wind ) is non- or mildly - relativistic at all scales , and synchrotron radiation is produced by a beam of relativistic electrons / positrons .
this idea has been worked out theoretically by sol , pelletier and asso ( 1989 ) and applied to kpc jets ( pelletier & roland 1986 , 1988 ; fraix - burnet & pelletier 1991 ; fraix - burnet 1992 ) for the particle acceleration problem . at small scales , observed relativistic phenomena can be produced by the relativistic electrons / positrons , and pelletier & roland ( 1989 ) found a very interesting application for cosmology using superluminal radio sources . in this series of papers ,
we propose to apply this two - fluid concept to vlbi jets .
the idea is based on the correlation between outbursts of agns and the subsequent appearance of vlbi blobs . if these bursts are explained by bursts of high - energy particles ( as in marcowith et al .
1996 ) , then it is probable that these particles propagate on a few parsecs away .
a relativistic beam propagating within the jet plasma has been shown ( sol et al .
1989 ; achatz , lesch & schlickeiser 1990 ; pelletier & sol 1992 ; achatz & schlickeiser 1992 ) to be stable relatively to the excitation of langmuir , alfvn and whistler waves , on scales up to several hundreds of parsecs .
hanasz & sol ( 1996 ) recently showed that large scale fluid ( kelvin - helmoltz ) stability is also possible .
hence , we suggest that the blobs seen in vlbi jets are these ` clouds ' of relativistic electron - positron pairs propagating along helical trajectories wrapped around a non - relativistic jet .
the term _ cloud _ is defined in this work as an ensemble of relativistic particles occupying a limited region of the jet , but these particles and the jet plasma are fully interpenetrated , making a two - component plasma .
_ cloud _ should _ not _ be understood in the fluid sense of an isolated component with a well defined boundary .
we thus consider that the jet itself does not radiate .
its magnetohydrodynamics determines the structure of the trajectories ( magnetic field lines ? ) that the radiating clouds will follow .
the emphasis is on the physics of the clouds , because in a later paper , the properties of these clouds will be taken from high - energy emission models from agns ( marcowith et al .
this two - fluid concept will thus build a coherent picture of extragalactic jets from their extraction in the agn to the largest scale up to the extended lobes . in this first paper ,
the basic model is presented in a simple configuration where the magnetic field is supposed to be uniform and oriented along the helix .
the characteristics of the cloud are constant in time ( stationary case ) .
synthetic maps are presented as well as the evolution of apparent speed and brightness of the clouds along a period of the helix . in a subsequent paper , a turbulent component of the magnetic field
will be added , and polarization maps will be computed . then , in a third paper , the temporal evolution of the cloud will be considered together with the self - compton radiation . the model is presented in sect . 2 while the numerical method is described in sect .
results are shown in sect . 4 and a discussion is given in sect .
the description of the model in this section is divided in three parts .
the geometrical aspects deal with the shapes of the jet and the cloud ( see introduction ) , the description of the helical trajectory and the definition of the different reference frames .
the physical aspects of the model include the magnetic field characteristics and properties of the particles within the cloud .
the synchrotron radiation is then computed through the stokes parameters .
psfig.sty we consider a cylindrical jet of radius @xmath1 making an angle @xmath2 with the line of sight .
the trajectory of the cloud is defined by a helix wrapped around the jet with the same axis ( fig .
[ croquis ] ) .
the ratio of the pitch @xmath3 to the radius is given by : @xmath4 .
the shape of the cloud is taken to be an ellipsoid because we have in mind the study , in a future paper , of the temporal evolution of a spherical cloud of radius @xmath5 propagating along a magnetic field line .
we intuitively expect a stretch of the cloud in the direction of propagation to a half large axis @xmath6 . in a reference frame
_ r _ linked to the cloud in which the @xmath7 axis is along the trajectory , the equation of the ellipsoid writes : @xmath8 the coordinates of the ellipsoid center @xmath9 define the helix considered above and are parametrized in a frame _ r _ linked to the jet : @xmath10 the @xmath11 axis is parallel to the jet axis , the @xmath12 axis lies in the plane of the sky ( fig .
[ croquis ] ) and @xmath13 is the angular speed if @xmath14 is interpreted as the time .
we make the further asumption : @xmath15 so that the curvature of the helix along the cloud is negligible , or in other words , the magnetic field is uniform across the cloud . finally , the observer frame _ r _ has its @xmath16 axis along the line of sight and its @xmath17 axis parallel to the @xmath12 axis ( fig .
[ croquis ] ) .
we assume that the jet and the parent agn are at rest with respect to the observer .
relativistic effects only concern the cloud moving at a speed @xmath18 along the helix .
the @xmath7 axis of the cloud reference frame _ r _ is defined by this velocity vector which makes an angle @xmath19 with the line of sight .
naturally , @xmath19 varies along the trajectory .
the doppler factor @xmath20 is then : @xmath21 where @xmath22 is the lorentz factor of the cloud .
the magnetic field is split in two components : @xmath23 , where @xmath24 is uniform throughout the jet and always tangent with the helical trajectory , and @xmath25 is a non - uniform component . in this first paper , we take : @xmath26 . since we do not consider here the origin of the magnetic field and its structure ,
there is no need to precise further the physics of the jet .
the relativistic cloud is made of electron - positron pairs .
the energy distribution per unit volume of radiating particles is assumed to be a power law : @xmath27 , and the velocity distribution is isotropic in the cloud reference frame .
in contrast with gmez et al .
( 1993 , 1994a , 1994b ) , we take into account the upper cutoff energy @xmath28 because it plays a role in high energy spectra of agns we will consider in a later paper . the global particle density ( @xmath29 ) is thus : @xmath30\ ] ] where @xmath31 is the lower cutoff energy .
the synchrotron radiation from the relativistic cloud is computed through the stokes parameters @xmath32 and @xmath33 .
all the necessary background and formulae for an uniform density distribution of particles with isotropic velocity distribution can be found in pacholczyk ( 1970 ) and can also be found in gmez et al .
we neglect the elliptical polarisation ( @xmath34 ) , and focalize only on the intensity @xmath35 in this paper since polarization will be the subject of a forthcoming paper .
the magnetic field is here assumed to be uniform across the cloud ( see sect .
2.1 ) which is supposed to be homogeneous , so that we are allowed to use the analytical resolution of the full transfer equations described by pacholczyk ( 1970 ) .
this of course saves us considerable cpu time for this first stage , but resolution of the transfer equation via numerical techniques will be necessary in the next paper with an additional non - uniform magnetic field .
the observed frequency @xmath36 and the rest frequency @xmath37 in the cloud frame are related by : @xmath38 .
likewise , the emission and absorption coefficients are computed in the cloud frame _ r _ ( respectively @xmath39 and @xmath40 ) but the transfer equations are solved in the observer frame _ r _ with : @xmath41 cosmological corrections would imply the doppler factor @xmath20 to be replaced by @xmath42 , with @xmath43 the redshift of the source . in this work
we take @xmath44 .
our model of a vlbi jet considered in this first paper requires 11 parameters to be defined : _ jet : _
@xmath1 , radius of the jet ; \2 .
@xmath45 , ratio of pitch to jet radius ; \3 .
@xmath2 , angle of the jet to the line of sight ; \4 .
@xmath46 , magnetic field ; _ cloud : _ \5 .
@xmath47 , half small and large axes of ellipsoidal cloud ; \6 .
@xmath18 , cloud speed ; \7 .
@xmath48 , particle density ( @xmath29 ) ; \8 .
@xmath49 , lower and upper cutoff energy ; \9 .
@xmath50 , spectral index of the particle energy distribution ; _ observer : _ \10 .
@xmath36 , frequency of the observations ; \11 .
@xmath51 , distance to the source .
as described in sect . 2.1.1 , the helical trajectory is parametrized in the reference frame _ r _ where the @xmath11 is the jet axis .
then , a simple rotation by the angle @xmath2 around the @xmath17 or @xmath12 axis defines the trajectory of the cloud in the observer reference frame , especially the projection onto the plane of the sky .
the tangent of the trajectory gives the direction of the cloud velocity vector and @xmath24 .
the center of the cloud moves along the trajectory defined above . at each position ,
a cloud reference frame _ r _ is defined where the @xmath7 axis is tangent to the helix and makes an angle @xmath19 to the line of sight .
the large axis of the ellipsoid is parallel to this @xmath7 axis . in this frame ,
the ellipsoid is given by eq . 1 ,
and a simple transformation entirely defines the 3-d ellipsoid in the observer reference frame . at this stage
, the sky plane is discretized into 2-d cells ( pixels ) .
each cell is associated with the depth @xmath52 of the cloud along the line of sight and is given a size of @xmath53 pc .
since the cloud is homogeneous and the magnetic field uniform across the cloud , @xmath52 is the only quantity varying from a cell to the other . the transfer equation in this simple case
is then solved analytically for each cell .
doppler effects are the same for all cells , but , for a given configuration of the jet ( i.e. @xmath2 , @xmath45 , @xmath18 ) , vary depending on the position of the cloud on the trajectory ( because the angle @xmath19 varies ) .
given the relatively important number of parameters of the model , many types of jet can be produced . in this section ,
only two geometrical configurations are studied .
the emphasis is put on observational diagnostics as well as on the understanding of the effect of the different parameters .
some of the parameters listed in sect .
2.4 , are kept constant in all the results presented in this paper : @xmath54mpc , @xmath55 pc , @xmath56 , @xmath57 , @xmath58 , @xmath59 , @xmath60 , @xmath61 .
all distances have been fixed because they are `` morphological '' and are more or less constrained by the observations .
we think the chosen values are typical for close extragalactic vlbi jets ( i.e. m87 ) .
the value for @xmath50 is also typical for these objects .
the parameter @xmath62 is kept constant because it is coupled to @xmath63 through eq .
( 2 ) , whereas @xmath64 has no influence on the results of this paper since we are not concerned with high - energy radiation . changing all these parameters
would not affect very much the results presented here .
the synchrotron intensity would be modified if a different cloud size is chosen , but the particle density or the magnetic field intensity have about the same effect .
the variable parameters considered in the following are thus : @xmath65 . for clarity ,
results are shown for one cloud moving over one period of the helix , although real jets have several clouds propagating at the same time , possibly on different trajectories . the resulting jet from our model with @xmath66 g , @xmath67 , @xmath68 @xmath29 is shown in fig . [ map - large ] ( @xmath69 and @xmath70 hz ) and fig . [ map - small ] ( @xmath71 and @xmath72 hz ) .
in the first case , the cloud is optically thick .
each figure is a set of 6 maps corresponding to 6 positions of the cloud along one period of the helix . a motionless object of constant arbitrary intensity
is added to reproduce the core of agn .
this object has no means in our model and is placed on the axis of the jet , hence not on the trajectory . to mimic realistic observations ,
all maps have been convolved with a gaussian of fwhm=@xmath1 .
the resemblance with some observed vlbi jets is obvious .
one interesting point to note here is that the cloud initially appears to move in a direction nearly perpendicular to the axis of the jet . also , on fig .
[ map - small ] , the intensity of the cloud changes dramatically along the trajectory , in contrast with the optically thick case of fig . [ map - large ] .
this flux variation of the cloud is illustrated on fig .
[ figmap2 ] and fig .
[ figmap3 ] .
the apparent speed of the cloud is also plotted in these figures .
it is always superluminal here ( up to 7@xmath73 in the case of fig .
[ map - small ] ) , but more importantly it greatly varies along the trajectory . the variation of the cloud flux along the helix for different speeds is shown in fig . [ flux - large ] ( @xmath69 ) and fig .
[ flux - small ] ( @xmath71 ) . two phenomena are competing in the flux variation along the helix : the doppler effect and the orientation of the magnetic field .
the doppler factor depends on the cosine of the angle @xmath19 between the velocity vector and the line of sight , while the synchrotron flux depends on the sine of this same angle ( because magnetic field and cloud velocity vector are parallel and both tangent to the helix , and the synchrotron intensity depends on the magnetic field component which is perpendicular to the line of sight ) .
hence , at low speeds , the intrinsic flux is maximum where this angle is the largest ( middle of the curves in our examples ) , whereas the doppler factor creates the opposite behaviour at very high speeds . at intermediate speeds
, two maxima can appear due to the two competing effects . in the case of a large angle to the line of sight ( fig .
[ flux - large ] ) , the doppler factor is smaller than 1 ( flux dimming ) and decreases with increasing cloud speed ( for @xmath74 ) . here , the effect of the magnetic field is dominated by the doppler effect at speeds as low as @xmath75 .
this is because the variation of the angle @xmath19 between the magnetic field and the line of sight is small . in the opposite case ( fig .
[ flux - small ] ) , the doppler factor @xmath20 is always larger than 1 ( flux amplification ) and increases with the cloud speed .
the effect of the magnetic field is dominant at speeds as high as @xmath76 where two maxima are present .
the consequence of the competition between these two phenomena is that the flux does not simply increase with @xmath18 .
this is true only for a limited range of speeds and at some locations along the trajectory . the ratio @xmath77/@xmath78 ( that we call contrast ) of the maximum to the minimum fluxes of the cloud over one period of the helix , is plotted vs frequency in fig .
[ contrast ] for several speeds .
the angle of the jet to the line of sight is set to @xmath79 for this figure .
the contrast depends on the optical thickness of the cloud : it is higher in the optically thin regime at high frequencies .
the competition between the two effects discussed in sect .
4.2 is illustrated by the fact that the difference in this contrast between the two regimes has a minimum for @xmath80 .
the constrast increases with speed in the optically thick regime , but it first decreases and then increases with increasing speed in the optically thin regime .
synchrotron spectra of the cloud at two positions distant by half a period of the helix for @xmath69 ( corresponding to the first and fourth frame in fig .
[ map - large ] ) are presented in fig .
[ spectra ] . at small frequencies
the slope is @xmath81 ( the cloud is optically thick ) , and at larger frequencies , the slope is @xmath82 , as given by synchrotron radiation theory for a particle energy distribution spectral index of 2 in the optically thin regime .
as the cloud moves along the helix , these two slopes naturally remain the same .
but the transition frequency @xmath83 , where the flux of the cloud is maximum , increases with the projected magnetic field and the doppler factor . the same two competing effects discussed in sect .
4.2 are again in play here . in the case of fig . 9 ,
it has been shown in sect . 4.2
that this is the doppler effect that dominates . in general , the variation of @xmath83 along the trajectory could imply an apparent transition between the two regimes of optical thickness if the source is observed at a fixed frequency .
the influence of the particle density @xmath63 and the magnetic field @xmath46 on the synchrotron spectra is shown in fig .
[ spectra - parameters ] . increasing the particle density or the magnetic field shifts upward the optically thin part of the spectrum .
the optically thick part is not sensitive to the particle density , while it is shifted downward with increasing magnetic field .
the previous section shows that it is possible to explain observed vlbi jets with the two - fluid concept , even with a jet at rest .
the presence of a relativistic ` cloud ' ( see introduction ) propagating inside the jet is the key ingredient in our model .
we think that the idea of non- or mildly - relativistic jets in agn and radiosources is now fully viable at all scales .
it reconciles observed relativistic phenomena at scales smaller than the parsec and/or at vlbi scales , with non - relativistic jets both at large scale ( observations ) and at the central part of agns ( theories of jet extraction ) .
the helical trajectory , observationally suggested , relaxes the constraint on the angle between the jet and the line of sight .
the consequences of curved paths of vlbi blobs have not been fully appreciated , but agn `` unification '' models would certainly benefit from such considerations .
the helical trajectory also yields the observed behaviour that the initial direction of propagation of a blob can be nearly perpendicular to the jet axis .
this is observed in quite a few sources ( e.g. mrk 501 , conway & wrobel 1995 ) .
the case of a small angle to the line of sight shown in fig .
[ map - small ] is rather reminiscent of the bl lacertae object 0235 + 164 ( chu et al 1996 ) . from the synthesized maps ,
different observational quantities are presented in sect .
this helps in understanding the origin of flux variation along the trajectory .
these are also observational curves that could bring some information on the different parameters of the model .
even if it requires multiepoch and multifrequency data , our model can probably be already applied in some cases .
for instance , qian et al .
( 1996 ) used a helical model to interpret the intrinsic evolution of the vlbi blobs in 3c345 .
as has been seen in sect .
4.2 , the orientation of the magnetic field also yields a variation of the flux along the trajectory .
this has not been taken into account by these authors , but it could lead to different results . naturally , the present work is very simplistic , but undoubtly justifies sophistication of the simulations .
such simulations are necessarily limited because the reality encompasses so many physical phenomena .
the originality of our work is that no ad - hoc assumption is made in the sense that the physics of the radiating cloud can be entirely derived from theories of jet extraction and high - energy radiation . in the same way
, the trajectory could also be precised from physical calculations .
our goal here is to build a fully physically coherent picture of agns from the accretion disk up to the vlbi jet , under observational constraints from the radio to the high energy radiation .
the next step will be the complete simulation of the stationary jet , by including the polarization with the addition of a turbulent magnetic field . in a later stage ,
the evolution of the cloud along its way from the region where it produces @xmath84-rays to the vlbi scale will be theoretically studied and implemented in the numerical simulations .
achatz u. , lesch h. , schlickeiser r. , 1990 , a&a 233 , 391 achatz u. , schlickeiser r. , 1992 , in:_extragalactic radio sources : from beams to jets _ , eds . j. roland , h. sol and g. pelletier ( cambridge university press ) , p. 256 bowman m. , leahy j.p . , komissarov s.s . , 1996 ,
mnras 279 , 899 chu h.s . ,
bth l.b .
, rantakyr f.t . , zhang f.j . , nicholson g. , 1996 , a & a 307 , 15 conway j.e . , wrobel j.m . , 1995 , apj 439 , 98 fraix - burnet d. , 1992 , a & a 259 , 445 fraix - burnet d. , pelletier g. , 1991 , apj 367 , 86 gmez j.l . , alberdi a. , marcaide j.m . , 1993 , a & a 274 , 55 gmez j.l . , alberdi a. , marcaide j.m .
, 1994a , a & a 284 , 51 gmez j.l . , alberdi a. , marcaide j.m . ,
marscher a.p . ,
travis j.p . , 1994b , a & a 292 , 33 hanasz m. , sol h. , 196 , mnras in press henri g. , pelletier g. , 1991 , apj 383 , l7 marcowith a. , henri g. , pelletier g. , 1995 , mnras 277 , 681 pelletier g. , roland j. , 1986 , a & a 163 , 9 pelletier g. , roland j. , 1988 , a & a 196 , 71 pelletier g. , roland j. , 1989 , a & a 224 , 24 pelletier g. , sol h. , 1992 , mnras 254 , 635 phinney e.s .
, 1987 , in:_superluminal radio sources _ , eds .
zensus and t.j .
pearson ( cambridge university press ) , p. 301
qian s.j . ,
krichbaum t.p .
, zensus j.a . , steffen w. , witzel a. , 1996 , a & a 308 , 395 sol h. , pelletier g. , asso e. , 1989 , mnras 237 , 411 vermeulen r. , 1995 , in:_iau symposium 175 , extragalactic radio sources _ , ( dordrecht : kluwer ) zensus j.a . , 1995 ,
in:_iau symposium 175 , extragalactic radio sources _ , ( dordrecht : kluwer ) | in this series of papers , we develop a two - fluid model for vlbi jets .
the idea is that the jet itself is non- or mildly - relativistic ( electrons and protons ) , while the radiating blobs are relativistic electron - positron ` clouds ' moving on helical paths wrapped around the jet . in this work ,
the emphasis is on the physical description of the clouds , and not on the structure or origin of the trajectory . in the simple case where the magnetic field is uniform and homogeneous accross the cloud , and
the properties of the cloud are constant , the present paper shows synthetic maps of vlbi jets in different configurations , as well as the variation of different observational parameters along the trajectory . |
the pepsi and sphinx programs are monte carlo event generators for longitudinally polarized deep inelastic lepton - nucleon scattering ( dis ) and longitudinally polarized high energy hadron - hadron collisions , respectively .
they both use probabilistic methods to generate events with unit weight .
the generation process can be split up into three parts . at first
, the kinetic variables of the partonic reaction are determined according to the complete cross section yielding a two parton initial state and a two , three or four parton final state .
the partons include quarks , leptons and gauge bosons here .
the next stage consists of an iterative generation of initial and final state showers .
finally , the parton cascades fragment into hadrons which may further decay . until now
, polarization is only taken into account by the partonic cross section and the initial state shower .
a polarized treatment of the final state shower is in principle straight forward for longitudinal polarization , whereas to our knowledge no model of polarized fragmentation exists so far . in the following we report recent developments concerning sphinx and pepsi . for further information
please refer to @xcite and the www pages and .
high energy scattering of a ( quasi-)real photon on a hadron is very similar to hadron - hadron collisions .
the hadronic component of the photon is described by unpolarized and polarized photonic parton distribution functions much in the same fashion as for hadrons . to leading order in the strong coupling constant one can
thus clearly distinguish between a direct cross section and a resolved cross section . in the first case , the photon couples completely to a parton of the hadron , in the second case it fluctuates into quarks ( and possibly further into gluons ) right before the hard interaction with the hadron occurs . to next - to - leading order this distinction breaks down , but
a new pragmatic definition can be found @xcite .
the unpolarized parton distributions of the photon have already been explored at hera during the last years and a resolved component of the total cross section was established .
polarized photo production experiments have not been conducted yet and therefore it is `` unknown territory '' .
we sought to develop a tool for experimentalists which enables them to study the feasibility of such experiments .
pythia 5.6 @xcite , the basis of the sphinx program , already includes unpolarized photo production which made an extension of sphinx to longitudinally polarized photo production quite simple . in principle , only the leading order polarized direct cross sections @xmath0 , @xmath1 and @xmath2 had to be added and the polarized parton distribution functions of the photon had to be included . in the following we want to discuss some points of interest : * the polarized parton distributions of the photon are completely unknown .
we included the minimally and maximally saturated anstze by glck and vogelsang @xcite .
these can be used to determine the sensitivity of the asymmetries to the polarized parton distributions of the photon . *
the current experimental setup at desy is such , that electrons and hadrons collide and react via the exchange of a virtual photon .
the program therefore allows not only to generate photon - hadron events at fixed energies but also electron - hadron collisions at fixed energies .
this is done by calculating parton distributions of the electron via a convolution of the polarized weizscker - williams spectrum with the parton distributions of the photon . *
the leading order cross sections for the production of qcd jets with high transverse momentum and the production of direct photons are implemented . *
double spin asymmetries for photo production are expected to be of the order of one percent and below .
consequently , one needs at least @xmath3 _ events per bin _ to achieve decent accuracy and program runs become somewhat lengthy .
this is illustrated by figure [ jets ] which shows the inclusive jet asymmetry for photo production at polarized hera using the maximally saturated set of polarized photon distribution function from @xcite .
we chose the same kinematic cuts as the authors of @xcite and the showering routines were switched on .
@xmath4 events were simulated yielding a total execution time of approximately 3 days on a pentium 200 personal computer .
the possibility of measurements of cc events by missing momentum in deep inelastic scattering is one of the most interesting perspectives of a polarized hera . for this purpose on twist-2 level within the framework of the naive parton model the polarized contribution to cc events together with the @xmath5 electroweak corrections is now included into the mc - generator pepsi .
unpolarized cc events are already contained in the lepto code @xcite , which forms the basis of pepsi , and the polarized extension can be provided on twist-2 level by a simple substitution of the parton distributions . for @xmath6 exchange : @xmath7 and for @xmath8 exchange : @xmath9 @xmath10 denotes the u - like flavor of each generation , @xmath11 the corresponding d - like flavor of each generation and @xmath12 are the corresponding anti particles .
@xmath13 is the sign of the relative polarization configuration of the two beams , i.e. , @xmath14 for anti parallel and @xmath15 for parallel spin configuration .
@xmath16 is the usual bjorken @xmath16 and @xmath17 the virtuality of the exchanged w boson .
the very important feature of the cc events is that the polarization of the quarks can be expressed by their helicity .
consequently , for the electroweak corrections on the twist-2 level we can use the corresponding unpolarized but helicity dependent formula @xcite , and include those electroweak corrections again by a substitution of the parton distributions : @xmath18 \quad,\ ] ] with @xmath19 for particles and anti particles , and @xmath20 being the charge factor .
@xmath21 denotes the quark flavor and @xmath22 the incoming lepton .
the genuine electroweak corrections are included in the form factor @xmath23 while the functions @xmath24 contain the one photon bremsstrahlung originating from the quark leg ( @xmath21 ) , from the lepton leg ( @xmath22 ) and from the interference of both ( @xmath25 ) .
( 0 ) denotes the tree contribution .
the functions @xmath24 are in general convolutions of the parton distributions , to be more precise @xmath26 \right\ } \nonumber \\
\bar r_b & = & \frac{\alpha_{\rm em}}{\pi}\left\ { \bar s_b q(x , q^2 ) + \int_1^{1/x } dz \left [ \bar t_b \frac{\bar r_b q(zx , q^2 ) - q(x , q^2)}{z-1 } + \bar u_b \frac{1}{z } q(zx , q^2 ) \right ] \right\ } \quad.\end{aligned}\ ] ] the exact definition of the functions @xmath27 and @xmath28 is to be found in @xcite .
these formulae are available in fortran in the program hector @xcite , and we took them partially from this code . from the four contributions the single photon bremsstrahlung radiation from the lepton leg ( @xmath29 ) is dominant , while the others give only minor corrections . in fig .
[ corr ] the electroweak ( ew ) corrections are shown for the spin dependent charged current cross sections for @xmath6 and @xmath8 exchange .
the parton distributions used are from the polarized / unpolarized package grsv lo std described in @xcite . in the kinematic range used here ,
i.e. , @xmath30 , @xmath31 .
the electroweak corrections are rather small . only in the case of 1000 pb @xmath32
there might be a chance in the @xmath6 case that the error bars could become smaller than the effect due to the electroweak corrections .
the new upgrade of the program sphinx including polarized photo production allows the simulation of various models for the photon structure .
the upgrade on pepsi allows the analyses of polarized cc processes at a polarized hera including electroweak corrections .
the effect of the latter on the spin dependent part of the cross section is rather small in the kinematic range covered by the polarized hera .
m. glck and w. vogelsang , _ z. phys .
_ * c55 * ( 1992 ) 353 .
+ m. glck and w. vogelsang , _ z. phys . _ * c57 * ( 1993 ) 309 .
+ m. glck , m. stratmann and w. vogelsang , _ phys .
lett . _ * b337 * ( 1994 ) 373 .
+ m. stratmann and w. vogelsang , _ z. phys . _
* c74 * ( 1997 ) 641 .
d. y. bardin , p. ch .
christova , o. m. fedorenko , _ nucl .
* b175 * ( 1980 ) 435 ; + d. y. bardin , p. ch .
christova , o. m. fedorenko , _ nucl .
phys . _ * b197 * ( 1982 ) 1 ; + d. y. bardin , c. burdik , p. ch .
christova , t. riemann , _ zeit .
phys . _ * c42 * ( 1989 ) 679 , _ zeit . phys . _ * c44 * ( 1989 ) 149 .
a. arbuzov , d. bardin , j. bluemlein , l. kalinovskaya , t. riemann , _ hector 1.00 a program for the calculation of qed , qcd and electroweak corrections to e - p and lepton + - n dep inelastic neutral and chareged currend scattering _ desy-95 - 185 , nov 1995 .
e - print archive : hep - ph/9511434 | recent developments of the two mc - generators sphinx and pepsi are discussed . in sphinx
the process of polarized photo production was included to simulate the potential of a polarized hera to measure the photon structure function . in pepsi
the cc polarized cross sections on tree level have been added together with the electroweak corrections to simulate their effect on cc signals proposed for a polarized hera .
* recent developments of pepsi and sphinx * + |
@xmath1-decay is the every - day reflection of weak - interaction in nuclei . as such
, it provides an experimental window to the properties of the weak interaction at nuclear density .
in particular , theoretical studies of @xmath1-decay rates of nuclei have argued for a suppression of the axial coupling constant @xmath3 , from its vacuum value , as extracted from the lifetime of the neutron @xmath4 @xcite , to unity , i.e. @xmath5 @xcite . according to a recent study , this suppression occurs gradually , as the mass of the nucleus grows , and fully utilized for @xmath6 @xcite .
the ramifications of this suppression are numerous .
for example , to the understanding of astrophysical phenomena , such as neutron - star cooling and core collapse supernovae , whose dynamics is controlled by the weak interactions .
it is of no surprise that the source of this suppression has been the target of many theoretical works , which have associated it with a partial restoration of chiral symmetry in finite densities , deficiencies in the inclusion of correlations between nucleons , loop - corrections to the axial current originating in nucleonic excitations and mesonic currents , or combination of the three @xcite .
an important assumption is hidden in these suppression mechanism : if a full calculation of the weak interaction inside nuclei was possible from first principles , then the calculated decay - rates should agree with the experimental ones .
that is , if one could describe correctly the correlations between nucleons , and the weak interaction of an external probe with a nucleus , then one should recover the physical value of the axial constant . in order to do that , pertinent
is to solve the nuclear problem from first - principles . due to the strong correlations involved in the problem ,
a calculation of nuclear wave functions from the nucleonic degrees of freedom is at reach only for very light nuclei . the lightest nucleus that undergoes a @xmath1-decay is the triton .
however , the theory can not be checked in the triton since its half - life is used to remove some freedom in the weak interaction of a lepton with a nucleus , as will be explained explictly later .
the lightest nucleus that can provide a test to the theory is thus @xmath0he .
@xmath0he ( @xmath7 ) is an unstable nucleus , which undergoes a @xmath1 decay with a half - life @xmath8 to the ground state of @xmath0li ( @xmath9 @xcite .
however , a microscopic calculation of @xmath0he from its nucleonic degrees of freedom , failed to reproduce the @xmath1-decay rate .
this study , accomplished by schiavilla and wiringa @xcite , has used the realistic argonne @xmath10 ( av18 ) nucleon - nucleon potential , combined with the urbana - ix ( uix ) three - nucleon - force ( 3nf ) , to derive the nuclear wave functions , through the variational monte - carlo approach . the model used for the nuclear weak axial current includes one- and two - body operators .
the two - body currents are phenomenological , with the strength of the leading two - body term associated with @xmath11-isobar excitation of the nucleon adjusted to reproduce the gamow - teller matrix element in tritium @xmath1-decay . the calculated half - life of @xmath0he overpredicts the measured one by about 9%
. an unexpected result of the calculation , was that two - body currents lead to a @xmath12 increase in the value of the gamow - teller matrix element of @xmath0he , thus worsening the comparison with experiment .
the authors of this paper have presumed that the origin of this discrepancy is either in the approximate character of the vmc wave functions , or in the discrepancies of the nuclear model of the weak interaction .
pervin _ et _
@xcite , have used the gfmc approach to evolve the vmc wave functions ansatz .
they showed that this brings the single nucleon gamow - teller matrix element to about 0 - 3% deviation from the experimental value .
however , the mec are still expected to increase the deviation from the experiment to about 2 - 5% from the experimental value , leaving this problem intact . in the current
_ paper _ , we argue that the origin of the discrepancy is indeed in the model of the weak interaction inside the nucleus .
the foundation of such an argument has to be in the underlying theory , i.e. quantum chromodynamics ( qcd ) .
thus , we describe the weak currents within the nucleus , using an effective theory of qcd , namely chiral perturbation theory ( @xmath13pt ) , applicable at low energies , relevant to @xmath1-decay processes @xcite .
we use the triton @xmath1-decay to calibrate the strength of the contact interaction part of the meson - exchange currents , thus the calculation is without any free - parameters .
the six - body nuclear problem is solved in a fully _ ab - initio _ approach , expanded in hyperspherical - harmonics function , from its nucleonic degrees of freedom @xcite .
the nuclear wave functions are derived from j - matrix inverse scattering nucleon - nucleon potential ( jisp ) , describing two - nucleon scattering data and bound and resonant states of light nuclei to high accuracy @xcite . using this approach not only brings the calculated @xmath1-decay rate to within @xmath2 of the measured data , but also changes qualitatively the contribution of the two - body meson exchange currents ( mec ) compared to the work of schiavilla and wiringa .
@xmath13pt based mec are found to decrease the gamow - teller matrix element , compared with the increase found by schiavilla and wiringa .
we argue that this qualitatively different behavior originates in the use of @xmath13pt based mec , rather than the specific choice of the potential .
we start with a brief reminder of @xmath1-decay process , and the formalism used in the calculation .
the decay is a weak process , in which an unstable nucleus of charge @xmath14 emits an electron and anti - electron - neutrino , leaving a nucleus of charge @xmath15 .
the interaction is mediated through the exchange of heavy w@xmath16 boson .
as the momentum transfer in the process is much smaller than the mass of the @xmath17 boson , the weak interaction hamiltonian is given by @xmath18 , where @xmath19 is the fermi coupling constant @xcite , @xmath20 is the ckm matrix element mixing @xmath21 and @xmath22 quarks involved in the process @xcite , @xmath23 is the lepton charge lowering current , and @xmath24 is the nuclear charge raising current .
the decay rate can be calculated using fermi s golden rule , and it is proportional to the squared matrix element of this weak hamiltonian @xmath25 , where @xmath26 ( @xmath27 ) is the initial ( final ) state .
the lepton current is well - approximated as a current of charged dirac particles , thus results in kinematical factors to the decay - rate .
the weak nuclear current can be written as : @xmath28 , where @xmath29 is a pauli matrix .
@xmath30 ( @xmath31 ) has a polar- ( axial- ) vector symmetry . here
, we will discuss either a triton decay , or @xmath0he decay , hence the transitions are constrained by a selection rule on the angular - momentum change in the transition : @xmath32 .
thus , a multipole decomposition of the nuclear current is helpful . due to the small momentum - transfer
only the lowest multipoles contribute , i.e. the @xmath33 electric multipole of axial - vector symmetry @xmath34 , and in the case of triton also the @xmath35 coulomb multipole of polar - vector symmetry @xmath36 .
we explicitly checked that indeed the contribution of neglected multipoles to the decay - rate of @xmath0he can be bounded by @xmath37 @xcite .
the leading order contribution to the @xmath34 and @xmath36 operators are proportional to the gamow - teller and fermi operators , respectively .
thus , it is customary , when discussing the experimental rates , to talk about the empirical fermi and gamow - teller matrix elements , instead of @xmath34 and @xmath36 , using the relations @xmath38 and @xmath39 here , @xmath40 , and similarly for @xmath41 , @xmath42 is the total angular momentum of the initial nucleus , and @xmath43 is the axial constant @xcite . as discussed by simpson @xcite , and later revisited by schiavilla _
et al_. @xcite , the `` comparative '' half - life is related to the
empirical " gt and f operators thorough @xmath44 here , @xmath45 ( such that @xmath46 ) , and @xmath47 @xcite accounts for the small difference in the statistical rate function between vector and axial - vector transitions . putting the measured @xmath0he comparative half - life @xmath48 @xcite , one extracts @xmath49 . for triton , @xmath50 @xcite , thus @xmath51 .
@xcite . ] in order to complete a calculation , we have to specify the detailed structure of the weak - current , and to calculate the nuclear wave functions
. these will combine to produce the theoretical @xmath34 , which will be compared to the experimental ones above .
the main difference between the current work and previous ones , is the physical origin of the currents .
the last two decades of theoretical developments have provided us with an effective theory of qcd , in the form of @xmath13pt @xcite .
the @xmath13pt lagrangian is constructed by integrating out qcd degrees of freedom of the order of @xmath52 and higher .
it retains all assumed symmetry principles , particularly the approximate chiral symmetry of the underlying theory .
this su(2)@xmath53su(2)@xmath54 symmetry is based on the small up- and down - quarks masses ( compared to the qcd breaking scale ) .
the lack of parity doublets in the qcd scale is interpreted as an indication that this symmetry is spontaneously broken , with the pions as the goldstone - nambu bosons .
their finite , albeit small , mass is due to the finite quark masses , explicitly breaking the chiral symmetry . furthermore , the chiral lagrangian can be organized in terms of a perturbative expansion in positive powers of @xmath55 where @xmath56 is the generic momentum in the nuclear process , i.e. the @xmath1-decay or the pion mass @xcite .
the chiral symmetry dictates the operator structure of each term of the effective lagrangian , however not the coupling constants . a theoretical evaluation of these coefficients , or low - energy constants ( lecs ) , is equivalent to solving qcd at low - energy , and it is not yet feasible to obtain them from lattice calculations because of computational limitations .
alternatively , these undetermined constants can be constrained by low - energy experiments .
as the chiral symmetry is a gauging of the electro - weak interaction , the weak currents are the nther currents of this symmetry .
the weak axial current adopted in this work is the nther current derived from the axial symmetry of the chiral lagrangian up to n@xmath57lo @xcite . at leading order ( lo )
this current consists of the standard single - nucleon part , which , as mentioned above , at low momentum transfer is proportional to the gamow - teller ( gt ) operator , @xmath58 where @xmath59 , @xmath60 are spin and isospin - raising operators of the @xmath26th nucleon . corrections to the single - nucleon current appear at n@xmath61lo in the form of relativistic terms .
it is easily verified @xcite that the single nucleon current achieved in the @xmath13pt formalism , is identical to that achieved in the standard nuclear physics approach ( snpa ) . at n@xmath57lo
, additional corrections appear in the form of axial mec . while the relativistic corrections are negligible for the half life , the mec have a substantial influence on this @xmath1-decay rate .
this is a reflection of the fact that @xmath34 is a chirally unprotected operator @xcite .
the mec , to this order , include two topologies : a one-(charged)-pion exchange , and a contact term ( that represents , for example , two - pion exchange or the exchange of heavier mesons ) . in configuration space the one - pion exchange part of the axial mec is given by : @xmath62 where @xmath63 is the pion - decay constant , @xmath64 is the mass of the nucleon , @xmath65 is the charged - pion mass @xcite , and the low - energy constants @xmath66 and @xmath67 are calibrated in the @xmath68-n sector @xcite .
the operators used here are defined as , @xmath69 and @xmath70 .
in addition , the yukawa - like functions are : @xmath71 @xmath72 is a cutoff function , which we take as a gaussian .
apart from this , the mec include a contact term , that has the form : @xmath73 where the `` smeared '' delta function is @xmath74 the lec @xmath75 is the only lec up to n@xmath57lo that can not be calibrated in the single nucleon sector , as it originates in the contact interaction @xmath68-nn in the chiral lagrangian . as a result , in order to determine @xmath75 , one has to resort to a larger nuclear system .
we will use the triton half - life as an experimental datum to determine this lec , sec .
[ subsec:3h3he ] .
the difference between the one - body contribution to the @xmath0he-@xmath0li gt matrix element and the experimental value is of the order of few percent . a result which on the one hand is very satisfying , but on the other hand implies that numerical accuracy at a _ per mil _ level is required if we to regard the @xmath0he @xmath1-decay as a test of the mec model . in view of this required level of convergence
we use the jisp16 potential @xcite to model the interaction between the nucleons .
the jisp16 nn potential utilizes the j - matrix inverse scattering technique to construct a soft nuclear potential , formulated in the harmonic oscillator basis , that by construction reproduces the nn phase shifts up to pion threshold and the binding energies of the light nuclei with @xmath76 .
we use the hyperspherical - harmonics ( hh ) expansion to solve the schrdinger equation .
the hh functions constitute a general basis for expanding the wave functions of an @xmath77-body system @xcite . in the hh method ,
the translational invariant wave - function is written as @xmath78 } c_{n[k]}r_{n}(\rho){\cal y}_{[k]}(\omega , s_i , t_i)\ ] ] where @xmath79 is the hyperradius , and @xmath80 are a complete set of basis functions .
the hyperangle , @xmath81 , is a set of @xmath82 angles , and @xmath83}(\omega , s_i , t_i)$ ] are a complete set of antisymmetric basis functions in the hilbert space of spin , isospin and hyperangles .
the hyperradius @xmath79 is symmetric under particle permutations since @xmath84 .
the functions @xmath83}(\omega , s_i , t_i)$ ] are characterized by a set of quantum numbers @xmath85 $ ] @xcite and possess definite angular momentum , isospin , and parity quantum numbers .
they are the eigenfunctions of the hyperspherical , or generalized angular momentum operator @xmath86 , @xmath87}(\omega , s_i , t_i)=k(k+3a-5){\cal y}_{[k]}(\omega , s_i , t_i)$ ] .
the details of our method are explained thoroughly in ref .
our results for the ground state properties of the @xmath88 nuclei , @xmath57h and @xmath57he , are presented in table [ taba3 ] . in the table , we present the energies , matter radii , and the leading order gt matrix element ( see eq .
( [ eq : e1a_gt_lo ] ) ) as a function of @xmath89 , the limiting value of the hyperspherical angular momentum @xmath90 in the hh expansion .
as we are using the bare interaction our results are variational .
.the jisp16 nn interaction @xmath91he ,
@xmath91h binding energies , rms matter radius , and the leading order gt matrix element as a function of @xmath89 . [ taba3 ] [ cols=">,^,^,^,^,^ " , ] summarizing , the predicted gt of @xmath0he is : @xmath92 the first error is the cutoff variation dependence , the second is numerical , the third is due to uncertainties in the triton half - life , and the last is due to uncertainties in @xmath3 .
this should be compared to the experimental matrix - element @xmath93 .
thus , the theory overpredicts gt by about @xmath12 .
the use of phenomenologically based potential , jisp , combined with a @xmath13pt based mec , is an inconsistency inherent to our calculation .
clearly , the chiral lagrangian can be used to derive the nuclear forces as well .
this inconsistency , however , allows us to overcome limited computational resources , as well as theoretical difficulties ( the n@xmath57lo nuclear potential has not been fully developed yet ) , and to accomplish the task of a microscopic calculation of a six - body problem .
the use of a hybrid approach , sometimes coined eft@xmath94 , has had great success in the literature @xcite . in all these checks ,
the phenomenological nuclear forces included realistic potentials , the av18+uix force model .
this potential , though different than the @xmath13pt force models in the short - range character of the force , has the correct long - range behavior , due to the pion exchange .
the jisp potential , however , is different in this respect , as it is built in an _ ab - exitu _ approach , and does not have an asymptotic long pion behavior , thus not consistent with chiral symmetry even at long - distances .
in addition , the jisp potential does not include a three - body force .
it is hard to estimate the effect of these approximations .
however , in a recent work @xcite , the triton @xmath1-decay process was calculated using force model and current derived consistently from the same @xmath13pt n@xmath57lo lagrangian .
one of the conclusions of this work has been that the short - range correlations of the force and the short - range correlations of the weak current are not correlated , thus the effect of the three - body - force is negligible for gt - type operators .
in addition , the jisp potential successfully reproduces nucleon - nucleon scattering data , and the binding energies of @xmath95 mass nuclei .
however , the most convincing reason to believe the stability of the current results , is the minimal dependence of the half - life in the cutoff .
we thus believe that even in the current calculation , the effect of the approximation will not change qualitatively the results , and the effect of the mec .
the qualitative difference originates in the different structure of the snpa and @xmath13pt based mec .
a careful analysis of the difference between the mec originating in @xmath13pt and those used in snpa , has been accomplished by park _
et al _ @xcite .
they have shown that one - pion exchange term exists in both models . of particular importance is the part of this term in the snpa based mec that represents the exchange of a pion due to a delta excitation of the nucleon , which is found to correspond roughly to the @xmath96 term in the @xmath13pt based mec .
the coupling constant of this term @xmath97 has been fixed by schiavilla and wiringa @xcite , so that the theory would reproduce the triton half - life .
however , differences between the approaches arise in their short range character . in the snpa approach , these correspond to the exchange of a @xmath79-meson
. such a term does not exist in the @xmath13pt approach as it arises only at n@xmath98lo @xcite .
moreover , a contact interaction of the form of eq . ( [ eq : contact ] ) does not appear in the snpa approach .
it is this contact interaction that creates the qualitative difference between the current work and that of schiavilla and wiringa @xcite . in order to acknowledge that , we plot in fig .
[ fig : relative_importance ] the relative contribution of each the terms , i.e. one - body , one - pion - exchange and full calculation , to the gt matrix element .
one first recognizes that the one - pion contribution to the matrix element has a positive sign in both @xmath57h and @xmath0he , and that the contact interaction has a negative contribution to the matrix element . in the case of @xmath57h
this is only a partial cancellation , as it is calibrated to increase the 1-body matrix element and to bring the calculation into the experimental value . in view of the fact that the one - body calculation in the case of the jisp potential almost exhausts the total gt strength
, one might suggest that the negative sign of the contact interaction , as well as the partial cancellation is an artifact of the potential .
however , the same partial cancellation is found also in a consistent n@xmath57lo calculation of @xmath57h decay , thus it is not a result of the use of the jisp potential @xcite .
in contrast to @xmath57h , when examining the case of @xmath0he , one observes that the negative contribution of the contact term is bigger ( in absolute value ) than the one - pion - exchange contribution , thus leading to a total negative contribution of the mec .
this negative contribution is needed as the single - nucleon gt is bigger than the experimental gt .
recalling the fact that the snpa approach does not contain a contact interaction , we understand the origin of the positive contribution of the mec in that approach , which increases the difference between the calculated and measured decay - rates .
in this work we have used the @xmath0he @xmath1-decay as a testing ground for the nuclear weak - current derived from @xmath13pt . a precondition for such
a task is an accurate evaluation of the @xmath0he-@xmath0li weak transition matrix - element at the per mil level . to this end
we have used the soft nn potential jisp16 to describe the nuclear dynamics and the hh expansion method to solve the schrdinger equation . the weak interaction in the nucleus is completely determined by fixing the short range behavior of the scattering operator to reproduce the experimental @xmath57h half - life , resulting in a parameter - free prediction of the @xmath0he @xmath1-decay rate .
we have found that at the 1-body , impulse approximation , level the @xmath99he-@xmath99li gt matrix - element is over predicted by roughly @xmath2 .
this observation for the jisp16 potential is in agreement with previous findings for other potential models .
adding 2-body , meson - exchange , currents derived within @xmath13pt , we have found that in contrast with the previous work of schiavilla and wiringa @xcite , the 2-body mec contribution to the @xmath99he-@xmath99li transition matrix element is negative .
we argue that this difference originates in the different short - range character of the mec derived in the two approaches .
we find that both for @xmath57h and @xmath0he , there is a sign difference between the positive contribution of the long - range one - pion - exchange current , and the negative contribution of the contact interaction in @xmath13pt , representing higher degrees of freedom which were integrated out in the development of the effective theory . in the case of the 6-body transition
, however , the contact interaction has a bigger value than the one - pion exchange contribution .
thus , it provides the origin to the sign difference between the mec contribution in @xmath57h and @xmath0he .
this contact interaction does not exist in the standard nuclear physics approach , adopted by schiavilla and wiringa .
therefore , the reconciliation between the theoretical and the experimental @xmath99he half - life is due to the use of the @xmath13pt formalism .
our calculation points to an agreement at the level of about 1.7% between the measured and calculated gt matrix elements .
this result should be contrasted with the difference of 5.4% obtained by schiavilla _ et _
more importantly , it shows that dominant contributions that arise naturally in the @xmath13pt formalism , and do not appear in the standard nuclear physics approach , are essential to a successful prediction of this weak observable . in order to pin - point
this argument , a use of a consistent approach , in which both the weak currents and the nuclear forces are derived from the same microscopic theory , is called for .
the agreement between the calculated and measured decay rates of @xmath0he indicates that there is no signature in this observable for an additional suppression of the axial constant .
it appears that all the needed suppression originates in correlations between nucleons in the nucleus , revealing itself in the form of exchange currents .
we thank rocco schiavilla for helpful discussions . the work of s.v .
. was supported by the israel science foundation ( grant no .
d.g . acknowledges support from u. s. doe grant de - fg02 - 00er41132 . | we present a microscopic calculation of the @xmath0he @xmath1-decay into the ground state of @xmath0li . to this end , we use chiral perturbation theory at next - to - next - to - next - to - leading order to describe the nuclear weak - currents . the nuclear wave functions are derived from the j - matrix inverse scattering nucleon - nucleon potential ( jisp ) , and the schrdinger equation is solved using the hyperspherical - harmonics expansion .
our calculation brings the theoretical decay - rate within @xmath2 of the measured one .
this success is attributed to the use of chiral - perturbation - theory based mesonic currents , whose contribution is qualitatively different compared to standard nuclear physics approach , where the use of meson exchange currents worsens the comparison to experiment .
the inherent inconsistency in the use of the jisp potential together with chiral - perturbation - theory based is argued not to affect this conclusion , though a more detailed investigation is called for .
we conclude that any suppression of the axial constant in nuclear matter is included in this description of the weak interaction in the nucleus . |
deep inelastic scattering ( dis ) offers the unique opportunity to study the production mechanism of heavy quarks via the strong interaction in a particularly clean environment where a point - like projectile , a virtual photon with a virtuality @xmath3 , collides with a proton .
a first measurement of the total inclusive @xmath4-quark cross section in the dis region has been released by the h1 collaboration @xcite .
recently , the zeus collaboration has measured the production cross section of @xmath4-quarks in dis at hera , in the reaction with at least one hard jet in the breit frame , where @xmath5 is the momentum of the exchanged photon , @xmath6 is the bjorken scaling variable and @xmath7 is the proton momentum , a purely space - like photon and a proton collide head - on . ] and a muon in the final state ( @xmath8 ) @xcite .
due to the large @xmath4-quark mass , muons from semi - leptonic @xmath4 decays usually have high values of @xmath9 , which is the transverse momentum of the muon with respect to the axis of the closest jet .
this allows a statistical separation of the signal from the background .
the data used in the measurement were collected during the 1999 - 2000 hera running period , corresponding to a total integrated luminosity of 60 pb@xmath10 and at a center - of - mass energy of 318 gev .
a detailed description of the zeus detector can be found elsewhere @xcite .
the event sample was selected in four steps : * dis events were selected by requiring a well reconstructed outgoing positron with energy greater than 10 gev , @xmath0 gev@xmath1 and inelasticity @xmath11 , where @xmath12 .
* muons were identified by requiring a reconstructed segment both in the inner and outer part of barrel and rear muon chambers , where the rear direction is defined by the incoming positron .
the reconstructed muons are matched in space and momentum with a track found in the central tracker .
in addition , a cut on the muon momentum @xmath13 gev is applied together with a cut on the muon polar angle @xmath14 , @xmath15 .
* hadronic final - state objects , reconstructed from tracks and energy deposit in the calorimeter , were boosted to the breit frame and clustered into jets using the @xmath16 clustering algorithm @xcite .
selected events had at least one jet with @xmath17 gev and within the detector acceptance @xmath18 , where @xmath19 is the pseudorapidity in the laboratory frame is defined as @xmath20 , where @xmath21 is the polar angle measured in with respect to the proton beam direction ] . * the muons in the sample
were associated with a jet using the @xmath16 algorithm information , where the associated jet was not necessarily the jet satisfying the cuts above . to ensure a good reconstruction of the associated jet , it was required to have transverse energy in the breit frame @xmath22 gev .
after all selection cuts the final data sample contains 836 events . to correct the results for detector effects and to extract the fraction of events from @xmath4 decays the rapgap monte carlo program @xcite was used .
the light - flavour and charm - quark rapgap samples were mixed according to the relative luminosities and the @xmath4-quark sample was added according to the beauty fraction determined from the analysis of the @xmath9 distribution . to determine the fraction of events from @xmath4 decays in the data , the contribution from light - plus - charm flavours and beauty were allowed to vary , and the best mixture was extracted using a binned maximum - likelihood method .
the measured fraction of @xmath4 decays is @xmath23 , where the error is statistical .
this mixture describes well the shape of all event kinematic variables .
the total visible cross section , @xmath24 , was determined in the kinematic range @xmath0 gev@xmath1 , @xmath11 with a muon with @xmath25 gev and @xmath15 and at least one jet in the breit frame with @xmath17 gev and @xmath18 .
the cross section is @xmath26 the nlo qcd predictions were evaluated using the hvqdis program @xcite .
fragmentation of @xmath4-quarks into hadrons was performed using the peterson function @xcite with @xmath27 as suggested in ref .
@xcite . the semi - leptonic decay of @xmath4-hadrons was modeled using a parameterisation of the muon momentum spectrum extracted from rapgap .
the @xmath4-quark mass was set to @xmath28 gev and the renormalisation and factorisation scales to @xmath29 .
the cteq5f4 proton parton densities @xcite were used .
the nlo qcd prediction is @xmath30 pb , where the error was estimated by varying the scale @xmath31 by a factor of 2 and the mass @xmath32 between 4.5 and 5.0 gev and adding the respective contributions in quadrature . a more detailed discussion of the qcd prediction and the related uncertainties for this process can be found in ref .
the measured total cross section is somewhat above but agrees with the nlo prediction within the experimental and theoretical uncertainties .
the program cascade @xcite , based on the ccfm evolution equations @xcite and on gluon densities depending on the transverse parton momenta emitted along the cascade , gives a prediction of 35 pb , which is in good agreement with the data .
the differential cross sections were calculated in the same kinematic range as the total cross section by repeating the fit of the @xmath9 distribution in each bin .
figure [ fig : datanlo ] shows the differential cross section as functions of @xmath3 and @xmath6 compared to the nlo .
the nlo predictions agree reasonably well with the data .
figure [ fig : datalo ] shows the same differential cross sections compared with the lo monte carlo programs .
the lo qcd prediction folded with the dglap qcd @xcite evolution equations ( rapgap ) underestimates the measured cross section .
cascade is in good agreement with the data .
the production of beauty quarks was measured in the process @xmath33 with the zeus detector at hera .
the nlo qcd predictions agree with the measured total cross section and differential cross sections as functions of @xmath3 and the bjorken variable @xmath6 within the experimental and theoretical uncertainties .
the cascade monte carlo program gives a good description of the measured cross sections .
i would like to thank my collegues of the zeus collaboration for their work and their suggestions to this talk .
zeus collaboration , _
`` measurement of beauty production in deep inelastic scattering at hera''_. abstract 783 , international conference on high energy physics , amsterdam , the netherlands , july 24 - 31 , 2002 .
m. derrick _ et al . _
[ zeus collaboration ] , phys .
b * 293 * ( 1992 ) 465 .
s. catani , y. l. dokshitzer , m. h. seymour and b. r. webber , nucl .
b * 406 * ( 1993 ) 187 .
h. jung , comput .
commun .
* 86 * ( 1995 ) 147 .
b. w. harris and j. smith , phys .
d * 57 * ( 1998 ) 2806 .
+ b. w. harris and j. smith , nucl .
b * 452 * ( 1995 ) 109 .
+ b. w. harris and j. smith , phys .
b * 353 * ( 1995 ) 535 [ erratum - ibid .
b * 359 * ( 1995 ) 423 ] . c. peterson _
phys . rev .
d * 27 * ( 1983 ) 105 .
m. cacciari and p. nason , phys .
* 89 * ( 2002 ) 122003 .
h. l. lai _ et al . _
[ cteq collaboration ] , eur .
j. c * 12 * ( 2000 ) 375 .
t. carli , v. chiochia and k. klimek , preprint hep - ph/0305103 , 2003 .
h. jung and g. p. salam , eur .
j. c * 19 * ( 2001 ) 351 .
+ h. jung , comput .
commun .
* 143 * ( 2002 ) 100 .
m. ciafaloni , nucl .
b * 296 * ( 1988 ) 49 .
+ s. catani , f. fiorani and g. marchesini , phys .
b * 234 * ( 1990 ) 339 .
+ g. marchesini , nucl .
b * 445 * ( 1995 ) 49 .
v. n. gribov and l. n. lipatov , yad .
* 15 * ( 1972 ) 781 .
+ g. altarelli and g. parisi , nucl .
b * 126 * ( 1977 ) 298 .
+ y. l. dokshitzer , sov .
jetp * 46 * ( 1977 ) 641 . | the measurement of the beauty quark production cross section in deep inelastic scattering with the zeus detector at hera is presented . the cross section for beauty quarks , measured in the kinematic
range @xmath0 gev@xmath1 and @xmath2 , and for events with at least one muon and one jet in the breit frame , is somewhat above the next - to - leading order ( nlo ) qcd prediction , but in agreement within the experimental and theoretical uncertainties . |
m - theory is expected to unify all of our understandings on the nonperturbative effects of string theory . despite its importance , many aspects on this interesting theory await to be understood .
one of the most mysterious facts on m - theory is the degrees of freedom of the branes . from the ads / cft correspondence , it is known @xcite that @xmath5 coincident m2-branes ( or m5-branes )
have the degrees of freedom of order @xmath6 ( or @xmath7 , respectively ) . in the case of d
- branes in string theory , the worldvolume theory can be described by the super yang - mills theory , which has the degrees of freedom @xmath8 .
compared with such a case , the field theoretical realization on the branes in m - theory is still unclear .
recently , there is a breakthrough @xcite ( along with its extension with fractional branes @xcite ) in understanding branes in m - theory , where it was proposed that the worldvolume theory of m2-branes is described by the supersymmetric extension of the chern - simons theory . concretely speaking ,
the proposal states that @xmath5 coincident m2-branes on the geometry @xmath9 is described by the @xmath10 chern - simons theory with gauge groups @xmath11 and bifundamental matters @xmath12 forming the superpotential @xmath13 where the coupling constant is given by @xmath14 .
the geometry @xmath9 in a special case with @xmath2 is nothing but the flat space , hence it is natural to expect that the supersymmetry is enhanced to @xmath15 in that case .
it was then proposed that assuming the existence of the monopole operator we can construct the @xmath15 supersymmetry algebra when @xmath16 @xcite . subsequently , it was shown in @xcite that the localization technique @xcite , which reduces the partition function defined by infinite - dimensional path integral to a finite - dimensional matrix model , is applicable to supersymmetric chern - simons theories .
this matrix model has the measure of the chern - simons type and many techniques from the chern - simons matrix model such as @xcite was utilized @xcite to obtain not only the @xmath6 behavior of the free energy in the planar limit but also the shift of t hooft coupling @xmath17 as expected from the euler coupling in the gravity dual @xcite . besides
, it was also noticed that this matrix model is closely related to the topological string theory on a local calabi - yau over the hirzebruch surface @xmath18 . using this relation
, we can study the higher genus contribution by borrowing a celebrated recursive relation , the holomorphic anomaly equation @xcite , from the topological string theory .
following interesting observations by @xcite , it was found in @xcite that we obtain the airy function after summing up the contributions from all genus .
this result on the airy function suggests some interesting relations to previous understanding on m - theory . in @xcite
it was noticed that the airy function also appears in other context of m - theory as the wave - function of the universe .
further , the integral expression of the airy function @xmath19 , which is a contour integral of an exponentiated cubic term , looks similar to the partition function @xmath20)$ ] of the cubic chern - simons action @xmath21 $ ] .
therefore , we are led to expect a deep relation among m - theory , the chern - simons theory and the airy function .
another expectation from this result is its implication in understanding some of hidden structures in the membrane theory .
string theory originates from the veneziano amplitude in the dual resonance model , out of which the string worldsheet conformal symmetry was realized .
similarly , one might expect to find some hidden symmetry out of this airy function and realize some structure for m - theory .
an interesting hidden structure of this airy function was beautifully realized by @xcite .
it was shown that the partition function of m2-branes ( as well as its generalization with @xmath22 supersymmetry ) can be rewritten exactly into the partition function of a fermi gas system with the density matrix ( related to hamiltonian by @xmath23 ) @xmath24 and , with the standard analysis of the statistical mechanics , we can reproduce the behavior of @xmath6 and airy function easily . very recently
, it was further proposed in @xcite that we can generalize this formalism to the @xmath25 case by introducing interactions to the fermi particles and in @xcite that we can extend to the cases with wilson loops .
this simple rederivation of the airy function implies the importance of this fermi gas formulation . since we have already a good control on the perturbative terms using the statistical mechanical method , as a next step , we would like to proceed to the study of this quantum mechanics extensively beyond the perturbation theory and to obtain some non - perturbative information out of it . aside from its possible important message on m - theory , the mathematical structure in this maximally supersymmetric theory seems very interesting .
the supergravity analysis @xcite suggests the bps condition is expressed by a continuous toda equation and the hamiltonian of this fermi gas system takes the form similar to the relativistic toda theory @xcite .
therefore , it is natural to expect some kind of the exactly solvable structure in this theory
. some of the exact results was found in @xcite .
we would like to make a first step towards the goal to compute the exact partition function of the abjm theory for arbitrary @xmath5 and arbitrary chern - simons level @xmath26 , and to solve the eigenvalue equation of the corresponding fermi gas system . heading for this goal
, we have found some interesting exact results on this theory as we will report in this paper .
as we will see in the next section , the ( grand ) partition function of this fermi gas system is computed by the trace of the power of the density matrix .
the main difficulty resides in the computation of convoluting the matrix element successively , @xmath27 following the previous work on the thermodynamic bethe ansatz ( tba ) equations @xcite , here we are able to express the matrix elements in terms of more basic vector quantities and reducing the computation of convoluting matrices into that of convoluting vectors .
in addition , we have also attempted to reduce the integration by commuting the coordinate @xmath28 and the momentum @xmath29 .
we have found that , although the naive commutation relation is anomalous , interestingly the additional term to the naive commutation relation can be expressed in terms of a projection operator @xmath30 . using this relation
, we can express @xmath31 for odd @xmath32 by some combinations of @xmath33 .
for even @xmath32 , we have found an interesting novel relation which separates the density matrix into two parities .
we will mostly restrict ourselves to the maximally supersymmetric case @xmath2 .
most of our following analysis goes straightforwardly to other values of @xmath26 though a key relation will only look simple for @xmath2 .
the case @xmath2 is particularly interesting from the viewpoint of holography since the abjm theory at @xmath2 is conjectured to be dual to the m - theory on @xmath34 which is in the truly m - theoretic regime .
the study of the abjm matrix model in the large @xmath5 , finite @xmath26 regime was initiated by @xcite and further studied in @xcite from the fermi gas approach .
however , the approach of @xcite is restricted to the perturbation around @xmath35 .
we would like to study the @xmath2 case exactly .
as a result , we have computed the exact partition function @xmath0 up to @xmath1 .
we shall summarize our results in the following , @xmath36 surprisingly , all of these results are written in terms of polynomials of @xmath3 with rational coefficients .
this paper is organized as follows . in section 2
, we present our general procedure to compute the grand partition function . applying the lemma of tracy - widom @xcite
, we first show that the power of density matrix @xmath37 is written in terms of a set of functions of one variable .
we also show that the eigenvalue problem of @xmath38 is reduced to the diagonalization of a certain hankel matrix . in section 3
, we find that the commutation relation of @xmath39 and @xmath40 appearing in @xmath38 contains an anomalous term , which can be expressed by a insertion of projection operator .
then we show that such anomalous commutation relation leads to a nontrivial relation of the grand partition function . in section 4 ,
using our general method , we compute exactly the first few terms of the taylor expansion of the grand partition function at @xmath2 .
these are nothing but the partition functions @xmath0 . along the way
, we find a non - trivial equation relating the even part and odd part of density matrix . in section 5 , some numerical results are given .
we consider the tba - type integral equation proposed in @xcite to compute @xmath31 .
these integral equations are powerful in the numerical analysis .
we also study the numerical solutions of the eigenvalue problem of the fermi gas system . in section 6
, we compare our results with the known perturbative results of the abjm matrix model and find a good agreement with the airy function even for small @xmath5 .
we study the non - perturbative corrections to the free energy , and find numerical evidence that the leading correction scales as @xmath41 .
we also consider the expectation value of @xmath5 in the grand canonical ensemble .
we find that the numerical computation using tba - like equations smoothly interpolates our exact results and the perturbative results .
finally , we conclude in section 7 with discussions and future directions . appendices
a , b and c contain some details of the computations skipped in the text .
the aim of this section is to present a general prescription which is helpful for computing the grand partition function of the abjm theory .
the concrete computation based on the method given here will be presented in section [ sec : exact ] .
let us start by reviewing the fermi gas formalism developed in @xcite . by using the localization ,
the partition function of the abjm theory on @xmath42 is written as the matrix integral @xmath43 ^ 2 [ 2\sinh(\frac{\nu_i-\nu_j}{2})]^2 } { \prod_{i , j } [ 2\cosh(\frac{\mu_i-\nu_j}{2})]^2 } \exp\left[\frac{ik}{4\pi}\sum_{i=1}^n(\mu_i^2-\nu_i^2)\right ] .
\label{eq : z_abjm}\end{aligned}\ ] ] applying the cauchy identity to the measure factor and performing the gaussian integral over @xmath44 , this partition function is recast into the `` mirror '' expression @xcite @xmath45 the key idea of @xcite is to rewrite it as the partition function of an ideal fermi gas @xmath46 where @xmath47 is understood as the density matrix of this fermi gas system . since the sum over permutations in can be also written as a sum over conjugacy classes of the permutation group , it is convenient to consider the grand partition function @xmath48 where @xmath49 is the fugacity and @xmath50 is the chemical potential . in our analysis , it is important to notice that the grand partition function has the following form @xmath51 , \label{eq : gpf}\end{aligned}\ ] ] where @xmath31 is defined by @xmath52 it is easy to see that @xmath53 can be expressed as the fredholm determinant of the kernel @xmath38 @xmath54 once the grand partition function is known , one can obtain the partition function by performing the contour integral @xmath55 for later analysis , it is convenient to divide the density matrix into two parts . since the density matrix @xmath56 is parity - preserving : @xmath57 , we decompose it into the parity even and odd parts @xmath58 by definition , @xmath59 , and thus all the matrix elements of the product @xmath60 vanish : @xmath61 this means that @xmath31 splits into two parts , @xmath62 and the grand partition function is factorized into @xmath63 where @xmath64 as we will see later , such a decomposition of the density matrix plays an important role in the analysis of the grand partition function .
now let us consider the computation of the grand partition function .
as can be seen from , we can compute the grand partition function if we know @xmath31 . to compute the partition function @xmath0 , in particular , we need @xmath31 ( @xmath65 ) .
however it is technically difficult to directly compute @xmath31 ( or the matrix element @xmath66 ) because it is given by the multi - variable integral as in ( or ) . to overcome this difficulty
, we here use the technique proposed in @xcite , which enables us to compute @xmath66 from a series of functions of one variable .
this technique can be applicable to a wide class of the kernels with the form @xmath67 here we apply it to the kernels @xmath68 .
below we concentrate our attention to @xmath2 for simplicity .
it is straightforward to generalize the method here to general @xmath69 .
the important point is that two matrices @xmath70 have the following forms ] itself is also of this class . ]
@xmath71 where @xmath72 we can apply lemma 1 in @xcite to .
it is shown that ( the generating functions of ) the density matrices are given by @xmath73 ( q_1,q_2 ) = \frac{q_\pm(q_1)q_\pm(q_2)-p_\pm(q_1)p_\pm(q_2 ) } { \cosh q_1+\cosh q_2 } , \label{eq : rpm}\\ \biggl[\frac{z^2 \rho_\pm^2}{1-z^2 \rho_\pm^2}\biggr ] ( q_1,q_2 ) = \frac{q_\pm(q_1)p_\pm(q_2)-p_\pm(q_1)q_\pm(q_2 ) } { \cosh q_1-\cosh q_2 } , \label{eq : spm}\end{aligned}\ ] ] where @xmath74 ( q , q ' ) \sqrt{z}e_{\pm}(q'),\quad q_\pm(q)=\int \ ! \frac{dq'}{2\pi } \biggl[\frac{1}{1-z^2 \rho_\pm^2}\biggr ] ( q , q ' ) \sqrt{z}e_{\pm}(q').\end{aligned}\ ] ] if we expand @xmath75 and @xmath76 as @xmath77 where @xmath78 then and indicate that the density matrices are given by @xmath79 these are the key relations in our analysis later .
these results show that the matrix elements @xmath80 are determined algebraically by the functions @xmath81 ( @xmath82 ) of one variable respectively
. therefore our problem reduces to finding @xmath81 , which is much simpler than the original one of finding @xmath80 themselves . for the determination of @xmath81 , as discussed in appendix [ app : rec - phi ]
, we can show that @xmath81 satisfy the following recurrence relations @xmath83 throughout this paper , we often use the notations @xmath84 from with the initial condition @xmath85 , one can recursively fix @xmath81 . in section
[ sec : exact ] , we will indeed determine the first few @xmath81 by using these recurrence relations .
let us summarize our general prescription to compute the grand partition function .
we first determine @xmath81 by using .
then from and , we find the matrix elements @xmath80 .
computing @xmath86 from such expressions , we obtain the grand partition function through and . in the previous subsection
, we considered @xmath31 to know the grand partition function .
there is another approach to analyze it .
the grand partition function can be expressed in terms of the eigenvalues of the density matrix @xmath38 , @xmath87 where @xmath88 is the @xmath32-th eigenvalue of @xmath38 with decreasing order @xmath89 .
thus if we can solve the eigenvalue problem of @xmath38 , we obtain @xmath53 .
the eigenvalue equation is given by the homogeneous fredholm integral equation of the second kind @xmath90 it is , of course , hard to solve this equation , but we can rewrite it as the eigenvalue equation for an infinite dimensional hankel matrix @xmath91 whose matrix elements are given by @xmath92 where @xmath93 is the @xmath26-th chebyshev polynomial of the first kind .
this representation is useful in our later analysis . to rewrite the eigenvalue equation ,
let us first expand the kernel into @xmath94 where @xmath95 since @xmath96 , the decomposition indicates that @xmath68 are respectively expressed as @xmath97 plugging into , we obtain @xmath98 where @xmath99 substituting back to , we find @xmath100 where the matrix elements of @xmath91 are given by .
the equation formally shows that all the eigenvalues of @xmath38 are also the eigenvalues of the infinite dimensional matrix @xmath91 .
the matrix elements @xmath101 depend only on @xmath102 , and such matrices are called hankel matrices . for the present case , it is obvious that @xmath103 if @xmath102 is odd .
namely , this hankel matrix @xmath91 has the following form @xmath104 we shall call such a matrix a parity - preserving hankel matrix .
as can be seen easily , the parity - preserving hankel matrix can be decomposed into two blocks of hankel matrices with opposite parity @xmath105 these two hankel matrices are related by @xmath106 with the shift matrix @xmath107 .
the eigen - spaces of @xmath91 also decompose into the direct product of the eigen - spaces of @xmath108 . for @xmath2
, @xmath101 has the following very simple form ( @xmath102 : even ) @xmath109 where @xmath110 is the catalan number @xmath111 for @xmath112 , we find ( @xmath102 : even ) @xmath113,\end{aligned}\ ] ] where @xmath114 is the digamma function . after seen the simple hankel matrix representation , before closing this subsection ,
now let us express various quantities appearing in the previous subsection in terms of this hankel matrix @xmath91 .
since the matrix @xmath91 and the density matrix @xmath38 share the same set of eigenvalues assuming a certain regularity condition , it is obvious that @xmath115 is equal to @xmath31 for any @xmath32 , thus the grand partition function is given by @xmath116 repeating the same argument above , one finds that the eigenvalue problems for @xmath117 are mapped to those for the hankel matrices @xmath118 , respectively .
let @xmath119 ( @xmath120 ) be the @xmath32-th eigenvalue of the density matrix @xmath117 .
then , we find @xmath121 we observe from the numerical evaluations ( see section [ sec : num_eigen ] ) that the eigenvalues of @xmath117 satisfy the inequality @xmath122 which means the following relations @xmath123 also , @xmath124 and its fourier transformation @xmath125 @xmath126 can be expressed in terms of @xmath91 as @xmath127 where @xmath128 .
the derivation of these representations will be given in appendix [ app : phi - hankel ] .
in the previous subsection , we have found an interesting hankel matrix representation for the density matrix @xmath56 .
it is natural to wonder whether there is an orthonormal basis in which the density matrix @xmath56 is represented by this hankel matrix .
we will show in this subsection that the hankel matrix @xmath91 is closely related to the density matrix in a basis constructed by the chebyshev polynomials .
motivated by the expansion , let us introduce the following function @xmath129 where @xmath130 is the chebyshev polynomial of the second kind and its appearance originates from the orthonomality of the basis @xmath131 , @xmath132 then , the density matrix in this basis , @xmath133 , is given by @xmath134 where we have used the expansion .
let @xmath135 be the matrix for the basis change from chebyshev polynomials to monomials @xmath136 the explicit form of the non - vanishing matrix element @xmath137 is given by @xmath138 when @xmath102 is even and @xmath139 . since one can check @xmath140 the matrix element @xmath141 is given by @xmath142 on the other hand , the hankel matrix @xmath91 for @xmath2 is written as @xmath143 in summary , the hankel matrix @xmath91 and the density matrix @xmath144 in the chebyshev basis
are written as @xmath145 these two matrices are
isospectral , _
i.e. _ , they have the same set of eigenvalues ( assuming a certain regularity condition ) @xmath146 note that @xmath144 is not a hankel matrix while @xmath91 is so .
here we consider an algebraic treatment of our fermi gas system . note that the matrix element in the position representation @xmath147 comes from a one - dimensional quantum mechanical system with the operator @xmath148 expressed in terms of the coordinate @xmath149 and the momentum @xmath150 as @xmath151 below we focus on the @xmath2 case .
let us summarize our convention for the phase space used in this paper , which is especially clean for @xmath2 .
the commutation relation of @xmath149 and @xmath150 is given by @xmath152=2\pi i.\end{aligned}\ ] ] the normalization of the bases @xmath153 and @xmath154 is as follows : @xmath155 therefore the projectors are normalized as @xmath156 hereafter , we suppress the hats on operators for simplicity .
note that the matrix elements of @xmath157 and @xmath158 in the position representation are @xmath159 using the bra - ket notation , the matrix elements of @xmath160 are given by @xmath161 since @xmath162=\operatorname{tr}[(c_qc_p)^n]$ ] from this equation , the grand partition function is also written as @xmath163 as explained around , the difficulty in computing @xmath31 consists in the multi - dimensional integral . in the operator representation , it can also be stated as the difficulty in the integration over the multi - dimensional phase space .
therefore , we would like to reduce the dimension .
if we know the commutation relation between @xmath157 and @xmath164 , we can reduce the dimension of the phase space involved . as is seen here
, one should be careful to derive the commutation relation . from the operator identity @xmath165
one would naively expect @xmath166 however this is not the case . there is an anomalous term in this commutation relation . to see this ,
let us consider the matrix element of the difference between both sides of @xmath167 if we introduce a projection operator @xmath168 satisfying @xmath169 , we can express the relation as @xmath170 at the operator level .
similarly , we find @xmath171 using the commutation relations and , we can calculate @xmath31 correctly .
let us see the computation of @xmath172 as a simple example : @xmath173 where we have used @xmath174 . since @xmath175 , is further rewritten as @xmath176 the first term vanishes because the integrand is an odd function .
the second term is evaluated as @xmath177 where @xmath178 and @xmath179 .
therefore we obtain @xmath180 one can easily check that the same result is reproduced by a brute force integration using . in the above computation , it is obvious that the anomalous term in the commutation relation is crucial to obtain the correct answer . in principle
, we can compute @xmath31 by using the relations and repeatedly instead of doing multi - integral . in practice
, however , the computation becomes very complicated as @xmath32 grows .
one advantage to consider the algebraic treatment is to give us some non - trivial implications for the grand partition function as we see in the next subsection . in the previous subsection
, we originally intended to reduce the dimensions of the phase space integration by commuting the coordinate @xmath28 and the momentum @xmath29 .
though this was not successful because of an anomaly term , the anomaly term is controlable and gives us an algebraic method to compute @xmath31 . in this subsection , we would like to discuss a more important implication . using and
, it is not difficult to show @xmath181 since @xmath182 is real , this should be equal to its conjugate @xmath183 .
the observation shows that , if the power @xmath32 is odd , we have @xmath184 which implies that , by expanding @xmath175 , @xmath185 can be expressed in terms of a certain combination of @xmath186 .
the easiest way to incorporate the combinatorics correctly is as follows .
let us consider the determinant @xmath187 .
this determinant is rewritten as @xmath188 where we have used @xmath189 .
this determinant is formally written as @xmath190 the first factor is just the grand partition function @xmath53 , and ( the logarithm of ) the second factor is evaluated as @xmath191 due to the insertion of the projection operator , the last equation is drastically simplified @xmath192 therefore the relation becomes @xmath193 where we have introduced the new function @xmath194 by @xmath195 let us show that @xmath196 is invariant under @xmath197 . using the commutation relations and , we have @xmath198 since @xmath196 is a real function for @xmath199 , this implies @xmath200 plugging into , we find a relation @xmath201 now note that the left hand side is nothing but ( the square of ) the grand partition function with odd power terms . more explicitly , let us divide @xmath53 into two parts @xmath202 where @xmath203,\quad \xi_{\rm even}(z)=\exp \left [ - \sum_{n=1}^\infty \frac{(-z)^{2n}}{2n } \operatorname{tr}\rho^{2n } \right ] .
\label{eq : xi_odd / even } \end{aligned}\ ] ] by construction , one easily finds @xmath204 then , shows @xmath205 using this relation , we can compute @xmath206 from @xmath194 . as we will see in the next section
, the function @xmath194 also has an important meaning for the even power terms , which connects the two determinants @xmath207 and @xmath208 defined in the previous subsection . after introducing the bra - kets such as @xmath209
, we can simplify various definitions and relations appearing previously . using the bra - ket notations , the functions @xmath124 and @xmath210
are expressed as @xmath211 it is easily shown that @xmath124 satisfies the recurrence relation @xmath212 similarly , @xmath210 also satisfies the recurrence relation @xmath213 finally let us comment on some useful algebraic relations steming from and by setting one of the coordinate to zero .
first , comparing with , one finds the relation between @xmath214 and @xmath215 @xmath216 since @xmath217 for @xmath218 , @xmath219 holds .
secondly , by definition of @xmath194 , the expansion of @xmath194 is given by @xmath220 from and , one immediately finds @xmath221 thus , with and , and imply novel relations between @xmath124 and @xmath210 @xmath222 recall that @xmath124 and @xmath210 are related through the fourier transformation originally .
in this section , we give some exact results for the grand partition function at @xmath2 . after the study of the algebraic treatment in the previous section
, our strategy to compute @xmath31 explained in section [ sec : gpf ] is improved .
there are three steps to know @xmath31 .
we first fix @xmath81 by using the recurrence relations .
we then compute @xmath31 separately for odd @xmath32 and for even @xmath32 .
the computation of @xmath206 reduces to the problem to compute the expansion coefficients @xmath223 of @xmath194 , which are fixed by @xmath124 . for the computation of @xmath224 ,
on the other hand , we use the formula . from this prescription , we have determined the exact partition functions of the abjm theory at @xmath2 up to @xmath1 .
let us start by determining @xmath81 .
since the integral equations have the convolution integral , the fourier transformation is useful to solve .
let us consider @xmath124 .
the initial condition is @xmath225 .
one can trivially find @xmath226 . using the fourier transformation ,
the convolution between @xmath164 and @xmath227 can be evaluated as @xmath228 where we have used the formula @xmath229 thus we obtain @xmath230 . iterating this method , we can fix @xmath124 recursively .
we list the explicit forms of the first seven functions below @xmath231 similarly @xmath232 is also fixed .
the explicit forms for @xmath233 are given by @xmath234 now we proceed to computing @xmath206 . for this purpose ,
let us start by computing @xmath235 .
it is easy to see that @xmath223 has the following integral form @xmath236 note that @xmath223 obviously does not depend on the insertion position @xmath237 ( @xmath238 ) , thus we can choose it handily . if @xmath239 is chosen , we obtain because of @xmath240 . using the results
, we find the first fourteen values of @xmath223 @xmath241 these lead to the exact value of @xmath206 from @xmath242 finally let us see the computation of @xmath224 .
we have already determined @xmath81 . using these results
, we can know the matrix elements @xmath243 .
the explicit forms of @xmath243 are given in appendix [ app : explicit_rhopm ] .
these matrix elements enable us to compute @xmath86 . here
we give the exact values of @xmath244 ( @xmath245 ) @xmath246 and @xmath247 note that @xmath31 is computed by using .
collecting the above results , the grand partition function is expanded as @xmath248 let us recall that the coefficient of @xmath249 in @xmath53 is just the partition function @xmath0 , thus we obtain the exact abjm partition function @xmath0 at @xmath2 up to @xmath1 .
the partition functions for higher @xmath5 can also be computed in this way . from the results in the previous subsection
, we find a novel relation @xmath251 in the expansion around @xmath252 .
as we will see below , it is shown that the above relation holds exactly for all @xmath253 , that is , @xmath254 before proceeding to the proof , let us comment on a consequence led by the above equality .
using , the grand partition function is written as @xmath255 since the expansion coefficients of @xmath194 are just given by @xmath256 ( see ) , the function @xmath194 is fixed by @xmath256 . on the other hand , the matrix elements @xmath257
are fixed by @xmath124 through .
therefore @xmath258 is also determined by @xmath124 .
we conclude that the total grand partition function is completely fixed by @xmath124 ( or @xmath259 ) only !
we do not need any information about @xmath232 ( or @xmath260 ) to know @xmath53 .
the novel relation we have found in this section suggests that we have an alternative strategy for computing the grand partition function .
previously we need to compute @xmath261 .
in particular , to compute @xmath224 , we need both of @xmath244 and sum up them : @xmath262 . after knowing the relation , the ingredients we need are the coefficients @xmath223 of @xmath194 and @xmath263 .
@xmath264 is automatically fixed by the above relation .
the rest of this subsection is devoted to the proof of by using the properties of the parity - preserving hankel matrix @xmath91 . using the relation
, one can check that the odd power part of holds for all - order in @xmath253 .
thus we concentrate ourselves to the even power part .
the statement to be shown is now rephrased as @xmath265 let us recall that @xmath91 decomposes into two hankel matrices @xmath108 . from , we find @xmath266 where @xmath267 and @xmath268 is the projector to the zero - th component .
this relation implies @xmath269 the right hand side looks similar to and therefore suggests that a similar manipulation is possible .
all we have to do is to rewrite the right hand side by expanding @xmath270 .
again , this is most easily done by studying the determinants .
let us rewrite the right hand side of .
we use the fact @xmath271 , then @xmath272 from the similar argument to derive , this determinant becomes simple @xmath273 expanding this in @xmath253 , we obtain @xmath274 now we want to write the right hand side of in terms of the coefficients @xmath223 of @xmath194 .
to do this , we use and . setting @xmath275 in the second equation of ,
we obtain @xmath276 on the other hand , if setting @xmath275 in , we also obtain @xmath277 where we have used . comparing these relations
, we find @xmath278 substituting into , one can confirm that really holds .
in the previous section , we compute the ( grand ) partition function analytically . here
we discuss another approach to analyze the partition function of the abjm theory .
the key idea is to use a link between the fredholm theory for the kernel like and tba system .
this novel link was first considered by zamolodchikov to study two - dimensional polymers in @xcite .
the kernel appearing there is similar to .
thus we can use some of the results in @xcite .
particularly , we can write down the tba - type integral equations that give @xmath31 . though the tba equations can not be solved analytically in general , it is easy to solve them at least numerically . in this section
, we numerically solve the tba - like equations and compute @xmath31 .
we compare them with the exact results in the previous section .
our numerical results show the very good agreement with the exact ones , and further give a prediction of @xmath31 ( and also the partition function ) , of which we have not found the exact values , with a very high degree of accuracy .
we also study numerically the eigenvalue spectrum of the density matrix . in order to see the relation to the tba - like equations , it is convenient to rewrite the kernel as the form @xmath279 where @xmath280 .
\label{eq : u}\end{aligned}\ ] ] in @xcite , zamolodchikov considered the following integral @xmath281 one can easily check that if @xmath282 is chosen , the integral of @xmath283 is related to @xmath284 thorough @xmath285 let us introduce the generating functionals of @xmath283 , @xmath286 from and , we obtain ( @xmath49 ) @xmath287 where @xmath288 is the grand potential . thus if we know the function @xmath289 , we can obtain the grand partition function .
one of the important results in @xcite is that the function @xmath290 is conjectured to be the solution of the following tba - type integral equations @xmath291 where @xmath292 is an auxiliary function , and we have introduced the following functions @xmath293 these two equations determine the functions @xmath290 and @xmath292 uniquely for given @xmath253 .
once @xmath290 and @xmath292 are determined , the function @xmath294 is given by @xmath295 where @xmath296 therefore we can obtain the whole function @xmath289 from these integral equations . in our case ,
the driving term in takes the form @xmath297 , which is not a standard form @xmath298 in ordinary tba systems for 2d relativistic quantum field theories . note that this conjecture has been proved in @xcite .
now we want to solve the tba - like equations , and and to compute @xmath31 . for this purpose ,
we expand @xmath292 as @xmath299 substituting and into , we obtain the following first few equations , @xmath300 where the convolution is defined by @xmath301 similarly , from , we obtain @xmath302 for @xmath303 . starting with @xmath304
, one can determine @xmath305 from , then @xmath306 is determined by the second equation in .
the flow of the determined functions is as follows , @xmath307 in the similar way , @xmath283 for odd @xmath32 is fixed from such as @xmath308 the integral equations , , etc .
are solved recursively . in principle
, we can solve them analytically by evaluating the convolutions of the known functions at each step .
technically , however , it is very messy to do that . alternatively , here we solve these sequential integral equations numerically at @xmath2 .
up to @xmath309 . ]
if @xmath283 is determined , we can compute @xmath284 from .
we have indeed computed the numerical values of @xmath261 from such integral equations . in table
[ tab : zn ] , we summarize the numerical values for @xmath310 .
the numerical results show the very good agreement with the exact values obtained in the previous section .
we would like to emphasize that the tba - like equations are valid for general @xmath26 ( even for non - integer values ) .
the only difference is the initial condition @xmath304 .
we can predict the high - accurate values of the partition functions for various @xmath5 and @xmath26 in the same manner . in this subsection , we expand all the unknown functions @xmath290 , @xmath294 and @xmath292 in @xmath253 to compute @xmath31 .
we note that the tba - like equations , and can be also solved for given @xmath253 by iteration as we will see in the next section .
.numerical values of @xmath311 at @xmath2 .
these numerical values are computed by using the tba - like equations , and .
our numerical results nicely agree with the exact values in the previous section . [
cols="^,^,^",options="header " , ] let us consider the non - perturbative corrections to the free energy @xmath312 where @xmath313 . where @xmath314 .
] we find @xmath315 is negative for all @xmath316 .
in @xcite it was noticed that there are two kinds of non - perturbative instanton effects contributing to the free energy .
one of them is the worldsheet instanton @xmath317 , while the other is the membrane instanton @xmath318 which comes from d2-branes wrapping on @xmath319 in @xmath320 . since in the case @xmath2 there is no @xmath319 to wrap , we expect that the membrane instanton effect vanishes and the non - perturbative corrections start from @xmath321 . in figure
[ fig : free_energy ] ( b ) , we tried to fit our exact data of @xmath322 by the function @xmath323 and found the best values of the fitting @xmath324 note especially that the value of @xmath325 is very close to @xmath326 .
this result strongly suggests that the membrane instanton @xmath327 is absent when @xmath2 , and the leading non - perturbative correction behaves as @xmath328 at large @xmath5 .
finally , let us briefly comment on the pre - factor of the instanton correction @xmath41 . from the general argument in @xcite ,
one expects that @xmath315 scales as @xmath329 with some coefficients @xmath330 and @xmath331 .
our analysis seems to indicate that the leading term of @xmath315 behaves as @xmath332 rather than @xmath333 .
however , our result up to @xmath1 is insufficient to tell .
one possibility is that @xmath315 may actually scale as for much larger @xmath5 than @xmath1
. it would be interesting to study the behavior of the non - perturbative correction @xmath315 in more detail . in the previous section
, we solved the tba - like equations by expanding all the unknown functions in @xmath253 in order to compute @xmath31 .
it is possible to solve them for given @xmath253 by iteration .
let us compare the results from such integral equations with the perturbative ones as functions of @xmath49 .
for this purpose , it is convenient to focus on the derivative of the grand potential @xmath334 because this quantity naturally appears in the tba approach as in . from the standard argument of statistical mechanics ,
the derivative @xmath335 corresponds to the expectation value of @xmath5 in the grand canonical ensemble @xmath336 where has been used in the last equality . using , the grand potential is also expressed in terms of @xmath31 .
the expectation value @xmath337 is thus written as the expansion around @xmath338 @xmath339 let us consider the perturbative contribution of @xmath334 . as shown in @xcite , the grand potential of the abjm theory is given by @xmath340 if non - perturbative corrections are neglected .
thus the derivative of @xmath334 is given by @xmath341 note that , in , the contribution from @xmath342 is dropped .
we numerically solve the tba - like equations , and at @xmath2 for various values of @xmath49 , and compute the expectation value of @xmath5 by using as a function of @xmath50 .
the results are shown in figure [ fig : mu - n ] . in the figure
, the dots represent our numerical results from the tba - like equations .
the red solid line is the plot of the truncated expansion up to @xmath343 where we have used the exact values of @xmath31 .
the green dashed line represents the perturbative result .
we find that our numerical results are close to the perturbative result in the large @xmath50 limit while in the limit @xmath344 the expansion agrees with the numerical results as expected .
the tba interpolates both results smoothly .
this is not surprising because the tba results contain all the non - perturbative corrections , which are neglected in . as a function of @xmath50 .
the dots show the results obtained from tba - like equations .
the red solid line represents the finite - truncated expansion up to @xmath343 with the exact @xmath31 while the green dashed line represents the perturbative result .,height=302 ] finally , let us give a comment on an underlying technical issue in this tba system . as was mentioned in @xcite , the iterative solution of the tba equations describing the fredholm determinant converges only when @xmath253 is in the regime @xmath345 where @xmath346 is the largest eigenvalue of the kernel @xmath38 .
if @xmath347 crosses over this value , the iteration method does not work any more . in our interested case @xmath2 , the bound of the convergence exists at @xmath348 ( @xmath349 ) .
in this paper , we have studied the fermi gas quantum mechanics associated to the abjm matrix model .
we have computed @xmath31 , which is the ingredient of the grand partition function , and obtained the partition function @xmath0 up to @xmath1 as a result . in reducing the number of integrations by commuting coordinates and momenta ,
we have found an exact relation concerning the grand partition function , which is interesting by itself and very helpful for determining the partition function .
we have also performed various numerical analysis .
all of these results are consistent with our exact results and also with the perturbative analysis in @xcite .
interestingly , all the partition functions we obtained are written in terms of polynomials of @xmath3 with rational coefficients .
similarly , the functions @xmath81 which lead to the determination of @xmath31 are also expressed in terms of polynomials of @xmath3 , @xmath28 , @xmath350 , @xmath164 , @xmath351 and @xmath352 with rational coefficients .
since we have computed the partition function exactly and integer numbers ( or rational numbers ) usually counts some information of the moduli space , we would like to know what kind of the information we are counting right now .
we would like to understand better the origin of the anomaly in commuting the coordinate @xmath28 and the momentum @xmath29 . from the analogy of string field theory ,
typically these kinds of anomaly appear because of the singularity of the moduli space we are considering and give some information to the physical problem .
for example , the associativity anomaly @xcite in cubic string field theory @xcite comes from the singularity at the boundary of the moduli space and relates to the translational invariance in string field theory .
also , the twist anomaly @xcite in vacuum string field theory @xcite originates from the divergence at the midpoint interaction point and contains the information of the tachyon mass and the brane tension .
we have shown that finding the energy spectrum of our fermi gas system boils down to the diagonalization of the hankel matrix @xmath91 defined by .
as in the case of the diagonalization of neumann matrices in open string field theory @xcite , the first step would be to find a ( hidden ) symmetry @xmath353 commuting with @xmath91 : @xmath354=0 $ ]
. it would be very interesting to find the exact spectrum of @xmath91 and write down the exact grand partition function .
we showed in that the grand partition function of the abjm theory is factorized into two parts .
a similar structure has appeared in the study of holographic dual of bps black holes @xcite , and it was argued that going to the fixed charge sector leads to the entanglement of two sectors .
such entangled state has an interpretation of branched tree of baby universes and the catalan number appears as the number of branchings @xcite
. it would be interesting to see if the factorized grand partition function of the abjm theory has some interpretation on the bulk gravity side .
we also would like to remark on an analogy with the so - called ode / im correspondence @xcite . in @xcite
, it was found that spectral determinants for certain schrdinger equations satisfy functional relations . in @xcite ,
these spectral determinants were identified with baxter s @xmath355-functions of the corresponding integrable system , and the functional relations of the spectral determinants just correspond to the quantum wronskian relations @xcite .
our analysis implies that there is a non - trivial relation between the two determinants @xmath356 because the total determinant @xmath53 is fixed by the information of either of the two sectors @xmath68
. it would be interesting to explore whether @xmath356 satisfy a functional relation like the quantum wronskian relation or not .
recently , the expectation values of some bps wilson loops of the abjm theory were analyzed using the fermi gas formalism @xcite .
it was found that the semiclassical expansion of such wilson loop expectation values is divergent for @xmath16 .
we hope that our formalism will give some improvements for the computation of wilson loops .
* acknowledgements * we are grateful to hiroyuki hata , masazumi honda , kazuo hosomichi , hiroaki kanno , tomoki nosaka , satoru odake , ryu sasaki , naoki sasakura for useful discussions .
. would like to thank junji suzuki for telling him the mathematica code to solve tba equations by using fft algorithm .
y.h . is grateful to the hospitality of the ipht , cea / saclay and s.m .
is grateful to the hospitality of yitp at kyoto university . the work of y.h .
is supported in part by the jsps research fellowship for young scientists , while the work of k.o .
is supported in part by jsps grant - in - aid for young scientists ( b ) # 23740178 .
let us derive the recurrence relations . using , we find @xmath357 using the identity @xmath358 we find that the kernel is written as @xmath359 since @xmath360 s are even functions of @xmath28 , we get @xmath361 which are the recurrence relations .
let us derive . we expand @xmath124 and @xmath210 in @xmath128 @xmath362 we would like to show @xmath363 since @xmath225 and @xmath364
, one immediately finds @xmath365 let us substitute into the recurrence relation . using the expansion @xmath366 we obtain @xmath367 the integral over @xmath368 causes the hankel matrix because of .
thus we find @xmath369 comparing this equation with , we obtain @xmath370 from this relation and the initial condition , one can show that @xmath371 is given by .
the derivation of @xmath372 is straightforward .
let us summarize the explicit forms of @xmath80 up to @xmath373 : @xmath374 and @xmath375 these matrix elements are directly derived from , by using and . for @xmath376 . since it can be obtained without difficulty , we shall omit their expressions here due to the complexity . ]
using these matrix elements , we can also compute @xmath377 up to @xmath378 in addition to @xmath244 as given in and .
the results are given by @xmath379 and @xmath380 interestingly , in the above results the terms with @xmath381 appear .
such terms are canceled out in @xmath206 , and @xmath206 do not contain them .
i. r. klebanov and a. a. tseytlin , `` entropy of near extremal black p - branes , '' nucl .
b * 475 * , 164 ( 1996 ) [ hep - th/9604089 ] .
o. aharony , o. bergman , d. l. jafferis , j. maldacena , `` n=6 superconformal chern - simons - matter theories , m2-branes and their gravity duals , '' jhep * 0810 * , 091 ( 2008 ) .
[ arxiv:0806.1218 [ hep - th ] ] .
o. aharony , o. bergman and d. l. jafferis , `` fractional m2-branes , '' jhep * 0811 * , 043 ( 2008 ) [ arxiv:0807.4924 [ hep - th ] ] .
m. k. benna , i. r. klebanov and t. klose , `` charges of monopole operators in chern - simons yang - mills theory , '' jhep * 1001 * , 110 ( 2010 ) [ arxiv:0906.3008 [ hep - th ] ] .
a. gustavsson and s. -j .
rey , `` enhanced n=8 supersymmetry of abjm theory on r**8 and r**8/z(2 ) , '' arxiv:0906.3568 [ hep - th ] .
kwon , p. oh and j. sohn , `` notes on supersymmetry enhancement of abjm theory , '' jhep * 0908 * , 093 ( 2009 ) [ arxiv:0906.4333 [ hep - th ] ] .
d. bashkirov and a. kapustin , `` supersymmetry enhancement by monopole operators , '' jhep * 1105 * , 015 ( 2011 ) [ arxiv:1007.4861 [ hep - th ] ] .
a. kapustin , b. willett , i. yaakov , `` exact results for wilson loops in superconformal chern - simons theories with matter , '' jhep * 1003 * , 089 ( 2010 ) .
[ arxiv:0909.4559 [ hep - th ] ] .
n. halmagyi and v. yasnov , `` the spectral curve of the lens space matrix model , '' jhep * 0911 * , 104 ( 2009 ) [ hep - th/0311117 ] .
n. drukker , m. marino , p. putrov , `` from weak to strong coupling in abjm theory , '' commun .
phys . * 306 * , 511 - 563 ( 2011 ) .
[ arxiv:1007.3837 [ hep - th ] ] . o. bergman and s. hirano , `` anomalous radius shift in ads(4)/cft(3 ) , '' jhep * 0907 * , 016 ( 2009 ) [ arxiv:0902.1743 [ hep - th ] ] .
m. bershadsky , s. cecotti , h. ooguri and c. vafa , `` kodaira - spencer theory of gravity and exact results for quantum string amplitudes , '' commun .
phys . * 165 * , 311 ( 1994 ) [ hep - th/9309140 ] .
n. drukker , m. marino , p. putrov , `` nonperturbative aspects of abjm theory , '' [ arxiv:1103.4844 [ hep - th ] ] .
h. ooguri , c. vafa and e. p. verlinde , `` hartle - hawking wave - function for flux compactifications , '' lett .
phys . * 74 * , 311 ( 2005 ) [ hep - th/0502211 ] .
m. marino and p. putrov , `` abjm theory as a fermi gas , '' j. stat . mech . * 1203 * , p03001 ( 2012 ) [ arxiv:1110.4066 [ hep - th ] ] .
m. marino and p. putrov , `` interacting fermions and n=2 chern - simons - matter theories , '' arxiv:1206.6346 [ hep - th ] .
a. klemm , m. marino , m. schiereck and m. soroush , `` abjm wilson loops in the fermi gas approach , '' arxiv:1207.0611 [ hep - th ] .
h. lin , o. lunin and j. m. maldacena , `` bubbling ads space and 1/2 bps geometries , '' jhep * 0410 * , 025 ( 2004 ) [ hep - th/0409174 ] . s. n. m. ruijsenaars and h. schneider , `` a new class of integrable systems and its relation to solitons , '' annals phys . *
170 * , 370 ( 1986 ) .
s. odake and r. sasaki , `` discrete quantum mechanics , '' j. phys .
a a * 44 * , 353001 ( 2011 ) [ arxiv:1104.0473 [ math - ph ] ] . k. okuyama , `` a note on the partition function of abjm theory on @xmath42 , '' prog .
phys . * 127 * , 229 ( 2012 ) [ arxiv:1110.3555 [ hep - th ] ] . c. a. tracy and h. widom , `` proofs of two conjectures related to the thermodynamic bethe ansatz '' commun.math.phys . 179
( 1996 ) 667 - 680 [ solv - int/9509003 ] .
a. b. zamolodchikov , `` painleve iii and 2-d polymers , '' nucl .
b * 432 * , 427 ( 1994 ) [ hep - th/9409108 ] .
c. p. herzog , i. r. klebanov , s. s. pufu and t. tesileanu , `` multi - matrix models and tri - sasaki einstein spaces , '' phys .
d * 83 * , 046001 ( 2011 ) [ arxiv:1011.5487 [ hep - th ] ] .
a. kapustin , b. willett , i. yaakov , `` nonperturbative tests of three - dimensional dualities , '' jhep * 1010 * , 013 ( 2010 ) .
[ arxiv:1003.5694 [ hep - th ] ] .
i. k. kostov , `` solvable statistical models on a random lattice , '' nucl .
suppl . * 45a * , 13 ( 1996 ) [ hep - th/9509124 ] .
v. a. kazakov , i. k. kostov , n. a. nekrasov , `` d particles , matrix integrals and kp hierarchy , '' nucl .
phys . * b557 * , 413 - 442 ( 1999 ) .
[ hep - th/9810035 ] .
j. hoppe , v. kazakov and i. k. kostov , `` dimensionally reduced sym(4 ) as solvable matrix quantum mechanics , '' nucl .
b * 571 * , 479 ( 2000 ) [ hep - th/9907058 ] .
m. hanada , m. honda , y. honma , j. nishimura , s. shiba and y. yoshida , `` numerical studies of the abjm theory for arbitrary n at arbitrary coupling constant , '' jhep * 1205 * , 121 ( 2012 ) [ arxiv:1202.5300 [ hep - th ] ] .
g. t. horowitz and a. strominger , `` translations as inner derivations and associativity anomalies in open string field theory , '' phys .
b * 185 * , 45 ( 1987 ) .
e. witten , `` noncommutative geometry and string field theory , '' nucl .
b * 268 * , 253 ( 1986 ) .
h. hata and s. moriyama , `` observables as twist anomaly in vacuum string field theory , '' jhep * 0201 * , 042 ( 2002 ) [ hep - th/0111034 ] .
l. rastelli , a. sen and b. zwiebach , `` vacuum string field theory , '' hep - th/0106010 . l. rastelli , a. sen and b. zwiebach , `` star algebra spectroscopy , '' jhep * 0203 * , 029 ( 2002 ) [ hep - th/0111281 ] .
r. dijkgraaf , r. gopakumar , h. ooguri and c. vafa , `` baby universes in string theory , '' phys .
d * 73 * , 066002 ( 2006 ) [ hep - th/0504221 ] .
p. dorey and r. tateo , `` anharmonic oscillators , the thermodynamic bethe ansatz , and nonlinear integral equations , '' j. phys . a a * 32 * , l419 ( 1999 ) [ hep - th/9812211 ] .
v. v. bazhanov , s. l. lukyanov and a. b. zamolodchikov , `` spectral determinants for schrodinger equation and q operators of conformal field theory , '' j. statist .
* 102 * , 567 ( 2001 ) [ hep - th/9812247 ] .
a. voros , `` spectral zeta functions , '' adv .
pure math .
* 21 * ( 1992 ) 327 - 358 . | 2ex = 3.5ex we study the fermi gas quantum mechanics associated to the abjm matrix model .
we develop the method to compute the grand partition function of the abjm theory , and compute exactly the partition function @xmath0 up to @xmath1 when the chern - simons level @xmath2 .
we find that the eigenvalue problem of this quantum mechanical system is reduced to the diagonalization of a certain hankel matrix . in reducing the number of integrations by commuting coordinates and momenta
, we find an exact relation concerning the grand partition function , which is interesting on its own right and very helpful for determining the partition function .
we also study the tba - type integral equations that allow us to compute the grand partition function numerically .
surprisingly , all of our exact results of the partition functions are written in terms of polynomials of @xmath3 with rational coefficients .
= 0.7 cm tit / hep-618 * exact results on the abjm fermi gas * .6em 1.2 cm yasuyuki hatsuda , sanefumi moriyama and kazumi okuyama 2em @xmath4 _ department of physics , tokyo institute of technology + tokyo 152 - 8551 , japan _ 1 em @xmath4 _ kobayashi maskawa institute & graduate school of mathematics , nagoya university + nagoya 464 - 8602 , japan _ 1 em @xmath4 _ department of physics , shinshu university + matsumoto 390 - 8621 , japan _ 1 em 1.5em july 2012 = 3.5ex |
in this paper , we reconsider the classical topic of the stability of visco - elastic poiseuille flow in the zero reynolds number limit . from a _ weakly nonlinear expansion _ , we find that this flow is nonlinearly unstable for high enough flow rates . the first linear stability analysis of the flow of a so - called oldroyd - b fluid one of the simplest continuum models for a visco - elastic polymeric fluid with nonzero normal stress differences , characterized by a single relaxation time @xmath0 @xcite was already performed almost thirty years ago @xcite . since the subsequent careful linear stability analysis of ho and denn @xcite it is generally accepted that _ poiseuille flow of an oldroyd - b fluid is linearly stable _ , even though the stability is weak for large values of the weissenberg number , the dimensionless quantity which measures the strength of polymer relaxation effects .
the definition used in this paper for the case of cylindrical coordinates relevant for pipe flow is @xmath1 where the term in the numerator is the normal stress difference of the flow . for the planar geometry the index @xmath2 has to be replaced by @xmath3 , with
the @xmath3-axis taken normal to the plates .
the term in the denominator of ( [ widef ] ) is the shear stress at the wall . for an oldroyd - b or upper convected maxwell ( ucm ) fluid ,
the unperturbed flow field @xmath4 is simply parabolic , and we get @xmath5 here @xmath6 is the maximum velocity of the unperturbed profile , @xmath0 is the aforementioned relaxation time characterizing the oldroyd - b or ucm fluid , and @xmath7 is the radius of the pipe . for the planar case , r has to be replaced by @xmath8 , half the spacing between the plates .
it is well known that visco - elastic poiseuille flow is linearly stable for well - established models like the ucm , oldroyd - b @xcite and the giesekus model @xcite . however , there are good reasons to reconsider the _ nonlinear instability _ of this flow configuration .
first of all , atalik and keunings @xcite have recently presented strong numerical evidence tht visco - elastic poiseuille flow is indeed _ nonlinearly unstable _ for the ucm , oldroyd - b and giesekus model : when they injected the laminar flow field with a perturbation of sufficiently large amplitude , instead of dying out ( like one would expect for a strictly stable flow ) the perturbation grew and saturated at a finite value , resulting in a finite amplitude oscillatory flow .
such behavior , in which the flow is always linearly stable but nonlinearly unstable , is also called a _
subcritical instability_. a recent experiment @xcite also supports this nonlinear instability scenario . secondly , as we will discuss in detail in section [ motivation ] below , there are actually a lot of indirect indications from well - established results on visco - elastic taylor - couette flow that planar couette or poiseuille flow between plates or in a tube might have a subcritical instability .
last but not least is the following observation : if visco - elastic poiseuille flow is unstable , the nonlinear spatially and temporally oscillatory flow pattern it will give rise to will inevitably result in distortions of the flow after exiting the pipe or slit .
in other words , if visco - elastic poiseuille flow is unstable for large enough @xmath9 , it automatically leads to an intrinsic route to `` melt - fracture '' type phenomena , a generic name for the fact that a polymer extrudate normally develops strong undulations or irregularities beyond some critical flow rate @xcite .
it is well - established that there are various mechanisms that can lead to such type of behavior ( e.g. , stick - slip behavior or exit instabilities leading to sharkskin type patterns @xcite ) .
most of these can be ( partially ) suppressed by the proper choice of material or of the extruder shape and coating .
our main conclusion ( in line with @xcite ) that polymer poiseuille flow is nonlinearly unstable for large enough @xmath9 implies that melt fracture type behavior is an unavoidable consequence of a nonlinear instability in the extruder . in this paper , we focus on the subcritical instability of visco - elastic poiseuille flow itself .
our arguments and experimental support for this scenario are summarized in two recent letters @xcite , but the issue is complex and deserves further experimental and theoretical study .
we will come back to it in the future .
as stated , the calculations that we report in this paper give the details of the explicit nonlinear amplitude expansion for the stability of visco - elastic poiseuille flow in the case of fixed average pressure gradient ( we consider periodic modulations of the pressure , so there are periodic pressure modulations but the average pressure drop remains unchanged ) .
we show explicitly that for large values of @xmath10 there is indeed subcritical behavior , in other words that the flow exhibits a weakly nonlinear instability
. moreover , the subcritical behavior sets in quite abruptly around a value @xmath11 of order 5 for flow in a tube .
this value is consistent with the value where melt fracture type behavior is normally reported to set in @xcite , but somewhat larger than the critical value of 0.1 found numerically for for the related oldroyd - b model with viscosity contrast @xmath12 and a reynolds number of 0.1 @xcite .
we will discuss the possible origins of these discrepancies at the end of this paper . since the nonlinear amplitude expansion that we will employ has been used mostly in other fields of physics , and since it involves some unusual subtleties , we present our results in some detail . for the benefit of the reader not interested in the derivation ,
we summarize our main results in section [ introsummary ] of this introduction . before doing so
, we first discuss the relation of this work with the taylor - couette problem .
just over a decade ago , larson _
@xcite investigated the stability of the flow of a visco - elastic polymer solution in a taylor - couette cell , a cell consisting of two concentric rotating cylinders .
they found that in this system the flow does exhibit a well - defined linear instability for some value of the deborah number @xmath13 , which is analogous to the weissenberg number .
their calculations were done for the oldroyd - b polymer model . in the same paper the predictions
were confirmed experimentally by a series of measurements on a polymer solution which is well described by this model .
this work was later extended by joo and shaqfeh @xcite , who considered the more general case of flow in a curved channel . in the limit of boundary
driven flow this reduces to the taylor - couette case , while in the limit of static curved walls the flow is driven by a pressure gradient ( so - called dean flow ) . in all these cases , as well as in the experimentally relevant cone and plate geometry @xcite ,
the flow is linearly unstable at large enough flow velocities .
the main conclusion from this line of research has therefore been that a linear instability occurs in visco - elastic fluids due to `` hoop stresses '' if the stream lines are curved @xcite .
this is confirmed by the observation that the various stability calculations show that the instability threshold goes to infinity if the curvature of the walls goes to zero .
for the taylor - couette system this is the limit in which the radius of the cylinders goes to infinity ; for the dean flow problem investigated by joo and shaqfeh @xcite this result is consistent with the weak stability of poiseuille flow between two parallel planes . in the last few years , groisman and steinberg @xcite
have experimentally investigated the visco - elastic instability in the taylor - couette system in detail .
they find very good agreement with the theory for the onset of the instability when the polymer flow velocity is increased ; however , an important result of their careful study is that the bifurcation at the critical weissenberg / deborah number is _ subcritical _ or `` backward '' @xcite .
groisman and steinberg @xcite have also presented intuitive arguments why the instability in visco - elastic taylor - couette flow at low reynolds numbers is subcritical . in their arguments
the fact that the flow is occurring between curved walls is playing essentially no role : the curvature of the walls is necessary to make the flow linearly unstable , but the nonlinear positive feedback mechanism that in their picture gives rise to the subcritical behavior does not rely on the curvature .
this suggests the scenario sketched in fig .
[ figconjecture ] : the stable nonlinear flow behavior , indicated in the figure by the full lines , is quite independent of the precise radius or curvature of the cylinders , assuming the `` gap '' ( distance ) between the cylinders to be constant .
the points where the curves touch the horizontal axis correspond to the linear instability threshold .
these points therefore are strongly dependent on the curvature of the cylinders : as indicated in the figure , the results of the theoretical analysis @xcite imply that they shift to the right as the radius @xmath7 of the inner cylinder increases , and that there is no linear instability in the limit @xmath14 .
the continuity found by joo and shaqfeh @xcite in going from the taylor - couette flow to the case of poiseuille - like dean flow between fixed curved planes , as well as the intuitive arguments of groisman and steinberg @xcite strongly suggest that the scenario for poiseuille - flow through a channel will be very close to the one sketched in fig .
[ figconjecture ] for the limit @xmath15 : the unperturbed flow is linearly stable for any @xmath10 , but nonlinearly unstable as if there is a subcritical bifurcation at @xmath16 .
moreover , just like the rightmost curve for @xmath15 approaches the horizontal axis rapidly for sufficiently large @xmath10 , implying that the threshold for the nonlinear instability is small at large @xmath10 , we expect that the threshold for the subcritical - like nonlinear instability of visco - elastic poiseuille flow is small for sufficiently large @xmath10 .
the work done in this paper , summarized in the next subsection , fully confirms this expectation .
it is of interest to note that the scenario we propose here for visco - elastic fluids in the zero reynolds number limit has strong similarities for large reynold number planar poiseuille flow of newtonian fluids : planar poiseuille flow of ordinary fluids becomes linearly unstable at a reynolds number of 5772 , but in practice the flow becomes unstable at much lower reynolds numbers of @xcite .
thus the transition is _
subcritical_. the scenario that has emerged is that the nonlinear branch extends down to @xmath17 for two - dimensional perturbations and to @xmath18 for three - dimensional flows , and in fact already over thirty years ago @xcite , an amplitude expansion for planar poiseuille flow in the spirit of ours was instrumental in showing that the instability at @xmath19 is subcritical .
the scenario we propose for visco - elastic flow is even closer to the one that is found for the transition to turbulence in poiseuille flow of newtonian fluids in a pipe : although the flow is also linearly stable for any @xmath20 , the flow is nonlinearly unstable for @xmath21 with a threshold which decreases as @xmath22 for @xmath23 @xcite , very much like we suggest in fig . [ figconjecture ] with the dashed line labeled @xmath24 .
the transition to turbulence of newtonian fluids therefore shows that whether or not there is a true linear instability is not very relevant in practice , and we believe that this is true for visco - elastic flows too .
let us conclude this subsection with the following observation .
as noted in the preceding paragraphs , it is generally accepted that when the stream lines are curved , visco - elastic flow becomes _ linearly unstable _ at sufficiently large flow rates . in accord with this visco - elastic poiseuille flow
is linearly stable since its stream lines are straight : infinitesimal perturbations must and do decay . however , following this same line of reasoning it is also very natural for visco - elastic poiseuille flow or planar couette to be _ nonlinearly unstable _ : in the presence of a _
finite _ perturbation , the stream lines _ are _ curved , and consequently we expect the flow to be unstable to any finite perturbation for sufficiently large flow rates . this is the essence of the subcritical instability scenario sketched above . in this paper
we will consider the limit that the reynolds number @xmath20 @xmath25 is negligible . here
@xmath7 is a characteristic length scale ( @xmath8 for the planar case , the radius @xmath7 for the case of a cylindrical tube ) , @xmath26 a characteristic velocity , @xmath27 the density and @xmath28 the viscosity of the fluid .
as @xmath20 measures the strength of the inertial terms with respect to the viscous terms , in the limit @xmath29 we can ignore the nonlinear convective terms in the momentum ( navier - stokes ) equation , and make a quasistationary approximation in which the temporal derivatives in the momentum equation are neglected . as stated before ,
as the constitutive equation for the polymer fluid we take the so - called ucm ( upper convected maxwell ) model @xcite , which expresses the stress tensor @xmath30 of the polymer fluid in terms of the shear tensor @xmath31 through @xmath32 where `` _ the upper convected derivative _ ''
@xmath33 is given explicitly in eq .
( [ deft1 ] ) below , and where @xmath0 is the parameter with the dimension of time that characterizes the ucm model . in this paper , we consider a perturbation of the velocity and stress fields with single wavenumber @xmath34 along the direction of the flow , and with amplitude @xmath35 , i.e. @xmath36 where @xmath37 means complex conjugate .
then , in an expansion in powers of @xmath38 , we determine to lowest nonlinear order the equation for @xmath38 , @xmath39 to linear order in @xmath38 this equation simply reproduces the term @xmath40 of the dispersion relation of a single mode @xmath41 ; although we have to redo the linear stability analysis in order to proceed to the nonlinear term , in principle this term is already contained in the analysis by ho and denn @xcite . in particular , since we know that every mode @xmath34 is linearly stable , @xmath42 for all @xmath34 .
the essence of our analysis consists of calculating the coefficient @xmath43 explicitly .
in particular the real part of @xmath43 is of importance : if the real part @xmath44 , then the nonlinear terms increase the damping of the amplitude and the unperturbed state is , within this approximation , not only linearly , but also nonlinearly stable . on the other hand ,
if @xmath45 , then the nonlinear term promotes the growth of the amplitude , and in particular amplitudes satisfying @xmath46 grow without bound . hence , in this approximation @xmath47 defined above constitutes the _
critical amplitude of the perturbation beyond which the flow in nonlinearly unstable_. in our analysis , we do find that indeed for sufficiently large weissenberg number @xmath48 ; we then determine the value of @xmath34 for which @xmath47 is smallest , and take this as the critical amplitude for the nonlinear flow instability .
the value of @xmath47 obtained this way from our analysis is plotted as a function of @xmath10 in fig .
[ plaatje4 ] for the planar case and in fig .
[ plaatje5 ] for the cylindrical case .
our normalization is such that @xmath38 is the ratio of the maximum perturbation in the shear rate at the wall over one wavelength , divided by the unperturbed shear rate , @xmath49}{\partial v^{\text{unp}}_z/ \partial y } \right|_{\text{wall } } .\label{amplitude}\ ] ] as we see from figs .
[ plaatje4 ] and [ plaatje5 ] that the overall behavior of the critical amplitude required to trigger the instability is in accordance with the picture as suggested in fig . 1 :
for @xmath50 , the planar limit , the threshold amplitude is expected to get increasingly small as one increases @xmath10 , as indeed it does . in fig .
[ velocityplotcyl ] we show the velocity profile @xmath51 corresponding with the linear eigenmode with wavelength @xmath52 in the planar case .
this wavelength is close to the one with the lowest instability threshold .
the roll - type structure of the flow - profile ( which we also find in the planar case , see section iii ) has very much the same structure as the one in the taylor - couette cell that according to the arguments of groisman and steinberg @xcite underlies the subcritical instability in that case this confirms that essentially the same mechanism is responsible for the subcritical instability in poiseuille flow .
an important feature of our results which is of practical importance is that the nonlinear instability sets in quite abruptly : below some critical value @xmath53 of the weissenberg number , the flow in the cylindrical case is within our approximation also nonlinearly stable as @xmath44 , while above @xmath53 the critical amplitude , especially the critical amplitude for the shear stress at the wall , drops rapidly to a small value ( as we shall see , the case of planar poiseuille flow is slightly different ) .
we can make this more precise as follows . in the general scenario , the nonlinear branch where the flow - profile is nontrivial and of the form sketched in fig .
[ velocityplotcyl ] , ends at a so - called saddle - node bifurcation point this is the point marked with a dot in fig .
[ figconjecture ] where the unstable dashed branch and the stable branch , indicated with a full line , meet .
for control parameters below the value corresponding to the saddle - node bifurcation , the nontrivial flow pattern is not dynamically stable anymore .
now , as we extend our expansion only up to the first nontrivial order in the amplitude , we are unable , from our expansion , to precisely locate the saddle - node bifurcation point .
however , it is reasonable to associate the approximate location with the point where @xmath54 . in this approximation ,
our estimate for the saddle - node bifurcation , and hence for the possible onset of the modulations in the flow profile and hence in the distortions of the extrudate , is @xmath55 .
note also that the threshold value for the nonlinear instability drops quite sharply when @xmath10 increases beyond this value : assuming even careful experiments can not avoid perturbations of the order of a few percent , we would expect to actually see the nonlinear state at @xmath10 values somewhat larger than @xmath56 in the cylindrical tube .
this is quite consistent with the weissenberg value @xmath11 where according to the literature @xcite extrudates typically exhibit undulations and deformations . in the planar case
the value of @xmath57 is somewhat less sharply defined , according to our results , but for all practical purposes it appears to be somewhat smaller than @xmath58 .
we will discuss the competition between the unperturbed laminar flow profile and the nonlinear profile further in the concluding section [ secdisc ] .
corresponding to the linear eigenmode of the cylindrical geometry with wavelength @xmath52 at @xmath59 .
the basic flow profile is in the horizontal direction . ]
a few remarks concerning figs .
[ plaatje4 ] and [ plaatje5 ] , which constitute the main result of our analysis , are in order : + of course , our analysis is only based on an expansion to lowest nontrivial order in @xmath38 ; one may therefore wonder how much higher order nonlinear terms affect the result for @xmath47 .
clearly , for small @xmath10 where @xmath47 becomes of order unity , our results should _ not _ be trusted quantitatively : higher order ( quintic ) terms will change the answer ( for the planare case they may even give rise to the existence of a true saddle - node bifurcation point ) . however , for large enough @xmath10 , our results _ can _ be trusted quantitatively .
this is because the linear damping term @xmath60 becomes arbitrarily small at large @xmath10 ( it varies roughly as @xmath61 ) , so that then the amplitude expansion for @xmath47 becomes nicely ordered .
+ since we have only determined the cubic nonlinearity in the equation , we can not say anything about the finite amplitude at which the instability will saturate .
it is in fact questionable whether the saturation amplitude can be determined from _ any _ perturbative method .
one hint that this might be possible to a reasonable approximation comes from the experiments of bertola _ et al _ @xcite , which indicate that the amplitude of the perturbations of the extrudate just above @xmath53 is rather small .
+ the scale on the vertical axis is _ not _ arbitrary . in figs .
[ plaatje4 ] and [ plaatje5 ] we have plotted the size of the shear rate perturbation normalized to the shear rate at the wall in the unperturbed case . using the equation @xmath62 } { \tau^{\text{unp}}_{rz } } \right|_{\text{wall } } = |c(1 ) a_{\text{c}}| , \label{stressamp}\ ] ] which holds both for the cylindrical case and the planar case(with @xmath2 replaced by @xmath3 ) , with @xmath63 a numerical constant defined in appendix [ appendixa ] , we show in figs .
[ plaatje4 ] and [ plaatje5 ] the ratio of the perturbed shear stress at the wall over the unperturbed shear stress at the wall , beyond which the flow is unstable .
note also the steep drop of the curve for @xmath10 just above @xmath53 : _ for all practical purposes the transition is quite sharp_. + our analysis also yields an idea of the value of @xmath34 of the mode with the smallest critical amplitude . for large enough @xmath10 that our analysis can be trusted , we find typical values about twice the diameter of the slit or the tube both in the planar case and in the cylindrical case .
a precise comparison has to be based on analyzing the frequency of the flow distortions measured at a fixed position , however , since the flow distorts upon exiting the die see section [ secdisc ] .
we present our nonlinear analysis in some detail for various reasons .
first of all , an amplitude analysis is normally used for problems with a true linear instability ; the use of the method for cases like this one without a true instability involves some unusual nontrivial subtleties .
second , such an expansion makes use of left eigenvectors of the linear operator , whose behavior and boundary conditions are quite intricate and worth discussing .
third , the analysis we introduce may be of relevant to other visco - elastic flow problems as well . in order to perform our weakly nonlinear analysis
, we first have to reanalyze the linear stability problem .
the essential results are summarized in section [ secla ] . in order to facilitate comparison with the earlier work by denn and coworkers @xcite
, we write the equation for the stability eigenmodes in terms of the stream function , which satisfies a fourth order linear differential equation .
the coefficients that we have ( re)derived for this differential equation are given in appendices , and are the same as those given in the appendix of @xcite .
a feature not discussed in the earlier work , however , is that there are various eigenmodes with different symmetries in the vertical direction for the planar case . in section [ secna ]
we first rewrite the linear eigenvalue problem in a form which is closer to the one usually found in derivations of an amplitude expansion @xcite .
this is then followed by a discussion of the derivation of the cubic nonlinearity in the amplitude equation .
a somewhat special feature of our approach that should be kept in mind is that normally amplitude expansions are used for an expansion around a true bifurcation point where a particular mode loses stability . here
the relevant linear modes are always weakly damped .
this gives rise to some slight differences in the formulation .
after having formulated the problem for the planar case in the first subsection , we will summarize the main steps of the linear stability analysis in the second subsection .
detailed expressions for the coefficients are relegated to to appendices .
the numerical results for the dispersion relation of the linear modes are presented in the last subsection .
we would like to investigate the linear stability of polymeric flow between two plates , separate by a distance of @xmath64 , and through a cylinder of radius @xmath7 .
the direction orthogonal to the planes will be our @xmath3-direction , the direction of the unperturbed flow will be the @xmath65-direction the advantage of this convention is that both in the planar and in the cylindrical geometry the mean flow is in the @xmath65-direction . from now on we take the flow two - dimensional by putting @xmath66 in the planar case and @xmath67 in the cylindrical case .
the first step is the derivation of an expression for the unperturbed flow : @xmath68 the next step is the perturbation of the unperturbed flow : @xmath69 we will keep the terms linear in @xmath70 and @xmath71 and write the @xmath70 and @xmath72 in terms of @xmath73 ( planar ) or @xmath74 ( cylindrical ) and its derivatives .
this will give us a fourth order equations for @xmath73 or @xmath74 .
with boundary conditions . solving this equation
yields the dispersion relation @xmath75 ; the imaginary part of @xmath76 determines the stability of the flow . the flow is taken to be incompressible , so that the conservation of mass equation becomes the incompressibility condition @xmath77 moreover , as was already explained in the introduction , since we are interested in the zero reynolds number limit , the navier - stokes equation , expressing conservation of momentum , reduces to the linear equation @xmath78 we will take the curl of equation ( [ defpro ] ) to eliminate the pressure .
this leaves us with an equation for the components of the stress tensor .
the ucm model describes the stresses in the polymeric fluid @xcite : @xmath79 where the stress tensor @xmath80 is defined in the following way @xcite @xmath81 eq .
( [ obeq ] ) illustrates that the ucm model is characterized by one time constant @xmath0 , which models the polymer relaxation time .
the `` upper convected derivative '' @xmath80 is the simplest frame - independent formulation that implements this @xcite .
the results in the cylindrical case are : @xmath87 , \ ] ] and @xmath88 the first step in a linear stability analysis is to linearize equation ( [ obeq ] ) to obtain expressions for the @xmath89 which are linear in @xmath71 . at this stage , it is useful to introduce the stream function @xmath90 . generally , the stream function is introduced by writing ( for the planar case ) @xmath91 so that the incompressibility condition ( [ diveq ] ) is satisfied automatically . for linear perturbations of the form ( [ lineqv ] ) we simply have @xmath92 , so that @xmath93 the stream function is slightly more complicated in the cylindrical case : @xmath94 it is also convenient to introduce dimensionless variables @xmath95 for the planar die , and @xmath96 for the cylindrical die . for notational simplicity , we will rename @xmath97 , @xmath98 and @xmath99 and @xmath100 ; it just means that we have to remember that all lengths are henceforth measured in units of @xmath8 or @xmath7 , inverse lengths in units of @xmath101 or @xmath102 , and times in unites of @xmath103 or @xmath104 .
the imaginary part of @xmath105 determines the linear stability of the flow ; the flow is linearly stable if @xmath106 . in order to facilitate comparison of our explicit expressions for the linear equation , given in appendix [ appendixa ] , with the expressions of rothenberger _
@xcite , we also introduce their dimensionless number @xmath107 @xcite , @xmath108 using definition ( [ deffi ] ) we obtain equations for the @xmath89 in terms of @xmath109 .
equation ( [ diveq ] ) is always satisfied because of definition ( [ deffi ] ) , equation ( [ defpro ] ) will give us a fourth order equation for @xmath109 of the following form : @xmath110 where @xmath111 we used the symbolic manipulation program maple to find the explicit expressions for the @xmath112 : these expressions are the same as those given by rothenberger _
_ @xcite and can be found in the appendix .
the aim of the linear stability calculation is to determine @xmath113 for fixed @xmath34 and @xmath107 , such that @xmath114 satisfies the usual stick boundary conditions : @xmath115 for @xmath114 these boundary conditions translate into @xmath116 note that we have four boundary conditions , and a fourth order equation . at first sight
, one might think that therefore the equation might have unique solutions for any @xmath117 . however , because of the linearity of the problem , if we find a solution @xmath118 , @xmath119 with @xmath120 an arbitrary complex constant is a solution of equation ( [ deffieq ] ) as well .
we eliminate this arbitrary degree of freedom by setting @xmath121 since this eliminates two trivial degrees of freedom which do not affect the solution , it is now clear that for a given @xmath34 and @xmath107 , one or more unique branches of the complex quantity @xmath113 will be fixed by the differential equation for @xmath114 . because of the vertical symmetry of the problem , eigenfunctions will either be asymmetric or symmetric . in the first case
the boundary conditions on the centerline are @xmath122 which implies for @xmath109 @xmath123 and in the second case the conditions are @xmath124 which implies @xmath125 we will explore the stability of both profiles below .
we will take the usual stick boundary conditions @xmath126 , @xmath127 at @xmath128 .
in the center we require @xmath129 , @xmath130 finite . in terms of @xmath114
we have @xmath131 , @xmath132 at @xmath133 , and we choose @xmath134 .
we have to be careful in the origin , because of the @xmath135 terms in the equations for the @xmath112 .
the boundary conditions on @xmath136 imply @xmath137 , @xmath132 at @xmath138 . in order to avoid numerical problems
, we use a polynomial @xmath139 for @xmath140 .
at @xmath141 we match both solutions by imposing equality of @xmath114 and the first three derivatives . in summary , we have to solve equation ( [ deffieq ] ) with the condition ( [ bc2 ] ) and the boundary conditions ( [ bc1 ] ) and ( [ bc3 ] ) for asymmetric modes or with the boundary conditions ( [ bc1 ] ) and ( [ bc4 ] ) in the case of symmetric modes .
this means that we have to find the two complex parameters @xmath142 , @xmath143 , such that the conditions on the centerline are satisfied . in the cylindrical case
we have to solve equation ( [ deffieq ] ) with the boundary conditions discussed above .
this means that we have to find the four complex parameters @xmath142 , @xmath144 , @xmath145 and @xmath146 such that both solutions can be matched together .
we have used a shooting program @xcite to find the right parameters and to construct the eigenmodes corresponding to the two classes of boundary conditions .
we present the results below for the range of @xmath34-values for which our program converged essentially to arbitrary precision . for larger values of @xmath34
the convergence becomes poorer , but since these larger values do not appear to be relevant for the nonlinear analysis of section [ secna ] , we content ourselves with reporting simply the range where sufficient precision could be reached . of the weakest damped asymmetric mode for @xmath147 and @xmath148 in the planar case .
[ plaatje1 ] shows the magnitude of the imaginary part of the eigenvalue @xmath144 of the _ asymmetric mode _ as a function of @xmath34 and for various @xmath10 .
for all values of @xmath10 , @xmath149 . as the temporal behavior of the modes is as @xmath150 , this confirms that the flow is _
linearly stable_. especially for large @xmath10 , @xmath151 is very small , which means that the flow is only weakly stable .
this is the reason that we also introduce the dimensionless temporal decay rate @xmath152 , so that the positive quantity @xmath151 is a small quantity for sufficiently large @xmath10 .
this is very important , as the smallness of @xmath151 will allow us to use an amplitude expansion in the next section : this expansion is based on an adiabatic approximation for the growth of the amplitude relative to the intrinsic oscillation of the waves with frequency @xmath153 . since the intrinsic frequency is of the order of unity in our dimensionless units , the condition for the amplitude expansion to work is that @xmath154 .
we found a second asymmetric mode close to the first one reported in fig .
[ plaatje1 ] .
this mode is slightly more damped than the first one .
the two asymmetric modes are shown in fig .
[ plaatje3 ] .
the symmetric mode lies between the two asymmetric modes , for , quite surprisingly , we have been able to show analytically that the symmetric mode obeys @xmath155 so that for the the weissenberg number @xmath59 in fig . [ plaatje3 ]
the damping rate of the symmetric mode is @xmath156 .
for this mode the phase velocity @xmath157 is extremely close to 1 , but the precise value has to be determined numerically . since the damping rates of all the three modes are very close , the above analytical result nicely shows that the damping rate of the modes becomes arbitrarily small for sufficiently large @xmath158 .
therefore , our amplitude expansion becomes self - consistent for sufficiently large @xmath10 . in the planar case , and the symmetric mode
whose damping rate is given exactly by eq .
( [ exact ] ) . ]
once the eigenmode has been obtained numerically , one has the velocity and shear stress fields as a function of the coordinates .
[ velocityplot ] of the introduction illustrates perturbed velocity field in the planar case for @xmath59 and @xmath159 .
our nonlinear analysis in the next section will be based on using the asymmetric mode which is the weakest damped .
the reason for not basing our expansion on the symmetric mode is twofold : + the symmetry of the velocity @xmath160 of this mode is such that it corresponds to a type of undulation mode between the plates , which does not generalize easily to the case of a cylindrical tube .
+ the result that the eigenvalue @xmath161 for this mode , with @xmath162 a very small real quantity , implies that the factor @xmath163 which appears at various places in the equations for the @xmath164 ( see appendix [ appendixa ] ) , becomes very large near the center .
consequently , the components of @xmath165 become very large , almost singular , near the center .
+ the asymmetric mode _ is _ the one least damped .
we now turn to the numerical results for the cylindrical case .
[ plaatjecyl1 ] shows the magnitude of the imaginary part of the eigenvalue @xmath144 as a function of @xmath34 and for various @xmath10 . as in the planar case ,
all values of @xmath10 , @xmath166 .
as the temporal behavior of the modes is as @xmath167 , this again confirms that the flow is _ linearly stable _ ; the decrease of @xmath151 with increasing @xmath10 again confirms that the linear stability becomes less and less with increasing flow velocity , roughly as @xmath168 . and @xmath169 .
] corresponding to the linear eigenmode mode of the planar slit geometry with wavelength @xmath170 at @xmath59 .
the basic flow profile is in the horizontal direction . ] plotted differently . here
the component @xmath171 is shown in a three - dimensional plot ; it illustrates how the perturbation is largest at the center of the tube . ] in fig .
[ velocityplotcyl ] we already showed the velocity profile @xmath51 corresponding with the linear eigenmode with wavelength @xmath172 in the cylindrical case .
[ velocityplotcyl ] confirms that the same type of behavior is found in the planar case ; the main difference is that in the cylindrical case , the mode is more confined close to the center of the tube ( see also fig .
[ velocityplotcyl2 ] ) than in the planar case .
this is consistent with the fact that , as we will see in section iv , the nonlinearly most unstable mode has a shorter wavelength in the cylindrical tube than in the case of the planar slit .
in fact , the roll - type structure of the planar flow - profile is more evenly distributed over the whole cross - section , its qualitative appearance is closer than the cylindrical one to the flow pattern in the taylor - couette cell that according to the arguments of groisman and steinberg @xcite underlies the subcritical instability in that case .
we will start with an outline of the method that we use in the first section .
as we shall see , in important ingredient that we need to obtain to determine certain solvability conditions is the solution of the linear adjoint operator .
we will discuss the derivation of the adjoint problem in the sections [ secadj ] and [ secsym ] .
we finally discuss the numerical results in section [ secnum ] . in order to perform the nonlinear analysis
it is convenient to rewrite the equations in a vector notation .
we shall develop the framework for the planar case , and then indicate the changes in the cylindrical case at the end of this section . since we only consider perturbations which are independent of the coordinate @xmath173 in the planar case , the stress components @xmath174 , @xmath175 , and @xmath176 are all zero , also in the nonlinear regime .
we can therefore capture the nonzero fields in a five - component vector @xmath177 whose components are @xmath178 and @xmath179 .
furthermore , we rewrite equations ( [ diveq]),([defpro ] ) and ( [ obeq ] ) in the following form : @xmath180 where @xmath181 is the linear operator associated with the linear problem of section [ secla ] , and @xmath182 contains all the nonlinear terms . thus the linear problem is simply @xmath183 , and @xmath182 contains only terms from the left hand side of equation ( [ obeq ] ) ; for the analysis below it is convenient to take @xmath184 symmetrized , so that @xmath185 .
explicit expressions for @xmath182 and @xmath181 can be found in appendix [ appendixb ] .
normally , an amplitude expansion is performed about the critical point where one of the modes is marginal ( i.e. , neither grows nor decays ) for a particular value of the wavenumber @xmath34 @xcite .
this critical wavenumber corresponds to a maximum of the linear dispersion relation . in the present case ,
the situation is not quite like this : the dispersion relation does not have a clear maximum , and the linear modes are always weakly damped .
this difference is in practice not a great problem .
first of all , we will not be interested in spatial variations of the envelope ; instead , we just pick a particular mode @xmath34 , and determine the important one for our analysis at a later stage ( through the requirement that the threshold amplitude be minimal ) .
it is therefore not necessary to expand about a maximum of the dispersion relation .
secondly , the amplitude expansion is based on an adiabatic decoupling of the fast and slow scales . in this problem ,
the fast scale is the period of the modes and the slow time scale is associated with the linear decay time of the modes .
we saw in section [ secla ] that in dimensionless units , the frequency of the modes is of order unity , while the damping rate @xmath186 becomes much smaller than 1 for large enough @xmath10 .
thus , there is indeed a separation of times scales , and this is essentially all that is needed for the analysis below .
in fact , for readers experienced with amplitude expansions , we could essentially go ahead pretending we are expanding about a true critical point where the growth rate of one of the modes is zero , and then at the end add the linear damping term by hand .
nevertheless , we prefer to formulate the analysis more carefully by keeping the damping term as it stands .
a linear eigenmode of the equations is of the form @xmath187 where we have introduced @xmath188 .
of course , all components of @xmath189 can just be obtained directly from the results of the linear stability analysis of section [ secla ] . in an amplitude equation formalism , one can in general also allow for spatial variations of the amplitude on a slow scale . as we already remarked above , here we confine the analysis to the temporal evolution of a single mode @xmath34 and its harmonics .
we do so for two reasons : first of all , it simplifies the analysis , secondly , our main goal is to determine whether there is a weakly nonlinear instability . anticipating that we will find that there is one , there is then not much to be gained in allowing for slow spatial variations .
our aim thus is to study the weakly nonlinear evolution in time of the amplitude @xmath38 of a single mode of the form @xmath190 of course , such a single mode is not really a solution of the ( weak ) nonlinear equation , but the corrections are automatically accounted for in the amplitude equation formalism and are slaved to ( [ amode ] ) .
we also note that because of the normalization ( [ bc2 ] ) , @xmath134 , we have @xmath191 = 2 |a|\ ] ] which , because in dimensionless units the shear rate at the wall in the unperturbed flow field equals 2 , leads immediately to the the relation ( [ amplitude ] ) between the maximum relative shear rate distortion at the wall and the amplitude @xmath38 .
likewise the relation ( [ stressamp ] ) between the amplitude @xmath38 and the maximum shear stress distortion at the wall follows from the explicit linearized dynamical equations .
these identifications are important in translating our results to real values , as was already noted in section i.c and fig .
[ plaatje4 ] and [ plaatje5 ] .
we want to know in particular whether the amplitude @xmath38 will grow or decay in time when the nonlinearities are taken into account . in the amplitude equation formalism , @xmath35 is expected to depend only on the slow timescale @xmath192 , but we have made this explicit , for reasons that become clear below .
we have to keep track of the fact that the true zero mode of our linear operator @xmath181 is @xmath189 as given in ( [ v0eq ] ) .
this term includes the weakly damped exponential factor @xmath193 , and this term is not explicit in ( [ amode ] ) .
the amplitude equation now proceeds by constructing the weakly nonlinear solution by writing @xcite @xmath194 where @xmath195 is the real quantity defined as the sum @xmath196 and where in writing @xmath197 we have anticipated that the amplitude @xmath198 varies on the slow timescale @xmath199 .
note that the real combination @xmath200 is needed because the vector @xmath177 refers to real physical fields .
once we have derived the equation for @xmath198 , comparison with ( [ amode ] ) shows that we should make the association @xmath201 the other terms in ( [ sub8 ] ) are then precisely the corrections to the dominant mode which are necessary to construct a weakly nonlinear solution . in the subsequent analysis , we have to keep in mind that @xmath189 , with the temporal factor included , is the true zero eigenmode of the linear operator @xmath202 : the temporal derivatives in the linear operator in fact work explicitly on both temporal exponential factors in the expression ( [ v0eq ] ) for @xmath189 .
when substituting the form ( [ sub8 ] ) in the equations , we then also have to account for the time derivatives of @xmath197 . for a product term of the form
@xmath203 we then have @xmath204 which we can simply summarize by making the substitution @xmath205 in all the time derivatives in the linear operator . since only first order time derivatives enter the linear operator @xmath181 , we get an expansion of @xmath181 of the form @xmath206 .
the amplitude expansion now proceeds by substituting the expansions for @xmath181 and ( [ sub8 ] ) for @xmath177 in the nonlinear equation ( [ fulleq ] ) and collecting the terms order by order in @xmath207 . for the first three orders in @xmath207 we get : @xmath208 equation ( [ sub12 ] ) is satisfied identically , as it should , because @xmath200 is the sum of two terms with wavenumber @xmath34 and @xmath209 which are both zero eigenmodes of @xmath181 .
the components of the @xmath210 follow from the numerical results for @xmath114 of section [ secla ] .
we can now solve equation ( [ sub14 ] ) to find @xmath211 .
this gives us @xmath212 where we have to find the vector @xmath213 numerically , because n(v ) contains only quadratic terms .
the second term is allowed with arbitrary @xmath120 , since @xmath189 is a zero mode of the linear operator @xmath214 ; it will not be needed in the subsequent analysis . in principle
, there would also have to be an additional @xmath215 term , but this term is identically zero for the case of the asymmetric boundary conditions of interest here .
once @xmath211 is known , we can proceed to eq .
( [ sub14 ] ) ; in this equation , the operator @xmath216 works on @xmath197 and its complex conjugate only : we can write the equation as @xmath217 this equation determines the time derivative of @xmath198 through a solvability condition : since the operator @xmath214 has a right zero mode , it can be solved if and only if the other two terms in the equation are orthogonal to the left zero mode of @xmath214 @xcite .
this requirement gives us the desired equation of @xmath218 . to make this explicit
, we first have to define the adjoint problem and an inner product .
we define a space of 5-dimensional vector functions of the three variables @xmath219 and @xmath220 , @xmath221 the components of the functions @xmath222 satisfy the physical boundary conditions discussed in section [ secla ] .
furthermore , an inner product @xmath223 on @xmath224 is defined : @xmath225 on this space of functions , an adjoint operator @xmath226 is defined such that @xmath227 for every function @xmath222 in @xmath224 . because the adjoint operator is obtained through partial integrations with respect to @xmath228 , the requirement that ( [ sub33 ] ) does not yield any boundary terms from these partial integrations yields the appropriate boundary conditions for the functions @xmath229 in the adjoint space .
we will state these explicitly for our case in section [ secadj ] .
let @xmath230 be the zero mode of @xmath226 .
the solvability condition applied to ( [ sub16 ] ) then becomes @xmath231 let us focus on the term with @xmath198 ; if we write @xmath232 we see that we only need the term with @xmath233 , since the inner product in ( [ sub30 ] ) is seen to vanish for all other terms after the @xmath65-integration is performed . for the same reason ,
we only need the terms which are proportional to @xmath234 from @xmath235 . using this
we obtain the following equation for the derivative of @xmath198 : @xmath236 the complex conjugate of this equation is obtained from analyzing the term with @xmath237 in ( [ sub30 ] ) . in ( [ sub30 ] )
, we have used the approximation that @xmath207 is sufficiently small that we are allowed to write @xmath238 this is nothing but the usual adiabaticity assumption . as we have shown in section [ secla ] , this approximation is justified for large @xmath10 .
we can summarize the eq .
( [ sub32 ] ) in the form @xmath239 where @xmath43 is a complex quantity which is obtained from working out the two inner product terms .
we can now translate this result back into the lowest order nonlinear equation for the amplitude @xmath38 . upon combining ( [ atob ] ) , and ( [ bfinal ] ) , we finally get @xmath240 which is nothing but eq . ( [ atriple ] ) of the introduction in dimensionless units .
as discussed there , the sign of the real part of @xmath43 determines whether or not we are dealing with a subcritical bifurcation . in the following sections the boundary conditions of the adjoint problem are derived , and we then proceed to solve adjoint problem and to determine @xmath43 .
the framework laid out above can be extended rather easily to the case of the cylindrical tube . in that case
we consider axially symmetric perturbations only , so that both @xmath241 and @xmath242 vanish ; note , however , that @xmath243 is nonvanishing in this case . the vector @xmath177
therefore now has six components , @xmath244 and @xmath245 .
apart from this change , the structure of the expansion is essentially the same , except for trivial changes like the fact that the first integration in ( [ iexpression ] ) should now be taken over the two - dimensional scaled radial coordinate @xmath228 . in this section
we will calculate the components of the adjoint operator @xmath226 using the defining equation ( [ sub33 ] ) .
we will follow again the planar case , and indicate the major changes for the cylindrical case at the end an in the appendices .
writing @xmath246 and @xmath247 we have @xmath248 the various functions and coefficients in this expression are given in appendix [ appendixb ] .
we will illustrate the structure of the calculation by analyzing two terms of @xmath249 , and simply state the results for the other terms .
one term we get is @xmath250 : @xmath251 both integrals on the last line vanish , the first one because the integrand is a partial @xmath65-derivative of a term which is periodic in @xmath65 , the second one because the integrand is a partial @xmath220-derivative of a term which is periodic in @xmath220 .
thus we simply obtain @xmath252 in short : we pick up a minus sign for every partial integration and in performing the partial integrations we get , in general , boundary terms which have to vanish . in the above example
, the boundary terms trivially vanished because of the periodicity of the terms with respect to @xmath220 and @xmath65 , but the partial integrations with respect to @xmath228 do not automatically vanish . in particular , from the various terms
we get the following boundary conditions : @xmath253 and @xmath254 the first set of conditions ( [ amp5 ] ) is always satisfied , because @xmath177 satisfies the boundary conditions of the original problem : @xmath255 and @xmath256 . by setting @xmath257 at @xmath258
, we have only one condition left of the second set : @xmath259 in order to understand how many boundary conditions we need to impose , it is good to realize that the operator @xmath260 actually works on physical , and hence real functions ; thus , we need the real part of this combination to be zero .
thus , we can still choose the phases of both amplitudes : we use this freedom to set the phase difference between the adjoint solution @xmath230 and the solution @xmath177 such that the combination @xmath261 becomes purely imaginary .
this shows us that if we impose @xmath262 all conditions are satisfied for physical functions .
we can give all @xmath263 in terms of @xmath264 and its derivatives .
this gives a fourth order equation for @xmath264 which we solve using a shooting method .
we will discuss the appropriate symmetry conditions in the next section . in the cylindrical case ,
we need @xmath264 to be bounded . in the present case ,
this can be achieved by putting @xmath265 , @xmath266 .
these conditions are satisfied if we expand @xmath267 , for @xmath268
. furthermore we can set @xmath269 to eliminate the degree of freedom we have .
this leaves us with one shooting parameter , @xmath270 .
we again have the boundary condition that @xmath264 vanishes on the wall , @xmath271 .
we choose @xmath270 such that this condition is satisfied . as we discussed in section section [ secla ] , the linear planar mode that we investigate is asymmetric , which means that its symmetry is @xmath272 which can be verified , see appendix [ appendixa ] . for the adjoint solution
we have also have two zero modes with different symmetry , @xmath273 and @xmath274 it turns out that for our choice of the linear mode ( [ v0symmetry ] ) , the second mode @xmath275 does give a solvability condition .
this can be seen as follows .
only the last three components of the vector @xmath177 contain a derivative @xmath276 .
thus only the overlap of the third , fourth and fifth component of @xmath230 and @xmath189 come in ; because of the symmetry of the opposite symmetry of the components , all these integral vanish .
the same holds for the @xmath184-term , and hence the solvability condition is trivially satisfied . as a result , @xmath277 is the adjoint zero mode which gives the nontrivial boundary condition .
the oddness of the component @xmath264 is then guaranteed by taking @xmath278 these conditions , together with the boundary condition at @xmath128 completely fixes the adjoint zero mode . as with the linear problem discussed in section [ secla ] , we solve the differential equation together with the boundary conditions with a shooting method . in the cylindrical case , we have just one single mode , also for the adjoint problem ; the boundary conditions that we already derived above for the cylindrical case are completely analogous to the above boundary conditions ( [ askbernard ] ) . in summary , the nonlinear term @xmath43 whose real part governs the weak nonlinear stability is obtained numerically as follows .
we first solve the fourth order equation for the stream function @xmath279 , derived in section [ secla ] .
this gives us the components of the vector @xmath210 .
a second routine of this program generates the term @xmath280 too .
a second program calculates the stream function @xmath281 of the inhomogeneous equation ( [ sub13 ] ) from which we obtain the components of the vector @xmath282 .
a third program solves the adjoint problem and gives the vector @xmath230 .
because we have @xmath283 and @xmath284 we have trivial @xmath65 and @xmath220 integrations in the inner product ( [ sub32 ] ) .
the @xmath228-integration is done numerically from @xmath285 to @xmath286 because of symmetry .
this then gives us the coefficient @xmath43 .
we first discuss the cylindrical case . in fig .
[ f10 ] we plot the value of the critical amplitude @xmath287 , determined from @xmath43 by eq .
( [ criticalamp ] ) , as a function of the wavenumber @xmath34 for three different values of the weissenberg number .
the curves illustrate that for @xmath288 there is a band of wavenumbers where @xmath289 and that hence there is , in our approximation , a critical amplitude beyond which the flow is unstable .
the sharp rise of the critical amplitude at the edges of the bands shown in fig .
[ f10 ] are close to the edges of the band where @xmath290 .
note also that for @xmath291 , the critical value still has a rather sharp minimum near @xmath292 , but that for increasing values of @xmath10 the bottom of the band flattens rapidly . for decreasing values of @xmath10 , especially when @xmath10 approaches about 5 , the band likewise sharpens ; we will see this from a different perspective below . as a function of the wavenumber @xmath34 for three different values of the weissenberg number @xmath10 for the case of the cylinder . ]
[ f10 ] also shows the behavior of the critical amplitude @xmath287 as a function of @xmath34 shows rather complicated structure .
for @xmath293 there is a plateau in the critical value of @xmath287 around @xmath294 ; upon increasing @xmath10 , this plateau shifts and disappears , whereas the absolute minimum of the curve shifts to smaller @xmath34 values while a new minimum develops at larger @xmath34 . already at @xmath295 , the two minimum almost correspond to the same values of @xmath47 , but for slightly larger @xmath10
the minimum at larger @xmath34 value becomes the absolute minimum .
this is further illustrated in fig .
[ f11 ] , where we plot the values of @xmath34 corresponding to the absolute minima of the curves , as well as the values corresponding to an amplitude 1.1 times the minimum value , as a function of @xmath10 .
the figure illustrates that between @xmath296 and @xmath297 , the minimum of the curve shifts to smaller @xmath34 values , but that around @xmath298 a local minimum at higher @xmath34-values becomes the absolute minimum .
corresponding to the minimum value of the curves in fig .
[ f10 ] as a function of @xmath10 , and the values of @xmath34 at which the amplitude @xmath38 is 1.1 times the minimum value .
the jumps of the curves are due to fact that a curve of @xmath47 as a function of @xmath34 has several minima . ]
as explained in section [ summary ] , our normalization is such that @xmath38 is the maximum perturbation in the shear rate at the wall over one wavelength , divided by the shear rate at the wall of the unperturbed parabolic profile [ see eq .
( [ amplitude ] ) ] . upon increasing @xmath10 , the minimum values of the curves , which were already plotted in fig .
[ plaatje5 ] , quickly decrease .
as we already mentioned in the introduction , we have also analyzed the critical value for the relative shear stress perturbation at the wall beyond which the flow is unstable [ see eq .
( [ stressamp ] ) ] . the data for this ratio as a function of @xmath34 for the same values of the weissenberg number as in fig .
[ f10 ] are shown in fig .
there are several important things to note about this figure .
first of all , the curves of the critical shear stress perturbation have just one minimum , contrary to those for the critical shear perturbation , and secondly , the values for the critical shear stress perturbation are typically a factor ten smaller than those for the critical shear . as fig .
[ f12 ] shows , for @xmath299 a perturbation of about 1% in the wall shear stress is sufficient to render the flow unstable .
this is why in fig .
[ plaatje5 ] the values of the critical shear stress amplitude ( the values corresponding to the minima of the curves in fig .
[ f12 ] ) are multiplied by 10 to draw them on the same scale as @xmath47 .
finally , we note that the edge of the band of unstable shear stress perturbations @xmath165 is the same as the edge of the band of unstable shear perturbations this is simply due to the fact that the edge of the band is marked by the point where @xmath54 .
beyond which the flow is unstable in the cylinder , as a function of @xmath34 for the same three values of @xmath10 as in fig .
[ f10 ] . ] in fig .
[ f13 ] we plot the band in which modes are nonlinearly unstable beyond some critical value as a function of @xmath10 .
this figure clearly illustrates that the width of the band vanishes at @xmath300 , since below this values @xmath301 for all @xmath34 .
this figure is somewhat reminiscent of the so - called `` busse balloon '' of rayleigh - bnard convection @xcite , but one should keep in mind that the interpretation is slightly different as we are dealing with a subcritical ( inverted ) bifurcation .
thus , the regions marked `` stable '' indicate the regions in the diagram where the perturbations with a wavenumber @xmath34 in that region are nonlinearly stable in our approximation .
the basic poisseuille flow profile is , however , nonlinearly unstable at all @xmath53 to modes whose wavenumber lies in the band marked `` unstable '' . in reality
, we expect that the flow pattern will settle to some kind of stable nonlinear behavior dominated by modes in this band .
-band where the corresponding modes render the basic cylindrical poiseuille flow nonlinearly unstable .
the dashed line indicates the @xmath34-value of the mode which is most unstable to shear perturbations , i.e. the minimum of the curves plotted in fig .
[ f10 ] . ]
although the band of unstable modes widens toward small @xmath34-values for @xmath10 between 6 and 7.5 a feature which is related to the plateau in fig . [ f10 ] for @xmath293
at flow rates ( weissenberg numbers ) about 50% beyond the critical value , the band of unstable wavenumbers ranges from just over 2 to a value somewhat larger than 4 . although with our expansion to cubic order in the amplitude we can not probe the stable nonlinear patterned flow regime nor the nonlinear selection of the wavenumber ( it need _ not _ correspond to the most unstable mode ) , it seems reasonable to assume that the wavenumbers of the pattern in the tube will lie in the range identified by fig .
[ f13 ] .
because near @xmath53 the critical amplitude is relatively large , and again because we have only expanded up to cubic order , one has to interpret our results near @xmath53 with caution .
nevertheless , our results appear to identify the location of the saddle - node bifurcation to nontrivial flow patterns near @xmath302 .
since the band of unstable modes in this limit is very small , it appears likely that if one follows the nontrivial flow pattern down to this value , its wavenumber is should be close to the value where the band in fig .
[ f13 ] closes , @xmath303 since the basic profile is then nonlinearly stable to perturbations with wavenumber that differs significantly from this value . in dimensional units , this implies for the wavelength @xmath304 of the pattern @xmath305 it is important to keep in mind that @xmath34 is the wavenumber _ inside _ the cylindrical tube .
if the ( near)periodic flow and stress patterns inside the tube indeed cause the surface undulations that are the first stage leading to melt fracture outside the tube at larger flow rates , then one should keep in mind that @xmath34 is _ not _ the wavenumber of these surface undulations . for
, the polymer extrudate swells upon leaving the tube , and the flow velocity profile of the polymer becomes essentially constant after leaving the tube . a better way to compare our theory with measurements on the extrudate
is therefore to compare the ( dominant ) frequency of the undulations : since no oscillations will disappear at the outlet of the tube , the frequency measured inside the tube must be equal to the frequency with which the the extrudate width oscillates after flowing out of the tube @xcite . of course , the frequency of the oscillations on the nonlinear flow branch corresponding to the solid line in fig .
[ figconjecture ] can not be obtained precisely from our expansion ; assuming that the nonlinearities do not change this frequency too much , we estimate it from our results for the linear modes . from our analysis of the linear eigenmodes in section [ secla ]
we find that the dimensionless frequency @xmath306 of the modes is to a very good approximation ( better than to a percent or so ) given by @xmath307 this result implies that for large weissenberg numbers , the phase velocity @xmath308 approaches 1 .
since according to ( [ dimensionlesscyl ] ) we measure velocities in units of @xmath6 , this means that the periodic stress pattern moves essentially with the maximum velocity of the flow in the large weissenberg limit .
this indicates that the linear mode more and more concentrated near the axis of the tube for large @xmath10 .
the above result gives as an estimate for the frequency @xmath222 in dimensional units @xmath309 finally , using that we expect the transition to occur at @xmath302 and @xmath310 , and that for the unperturbed ( parabolic ) profile of a ucm fluid @xmath311 with @xmath312 the average flow velocity in the tube , our estimate for the frequency at the transition becomes @xmath313 it is important to realize that this estimate is based on the assumption that the frequency ( or phase velocity ) does not get renormalized significantly by the nonlinearities .
this is maybe not unreasonable near threshold , where the amplitude of the periodic modulation of the pattern will be smallest ( keep in mind that since we are dealing with a subcritical bifurcation , the amplitude near threshold remains finite ) . especially further above threshold
, we expect the renormalization of the frequency to be significant .
although this can not be calculated from our first nontrivial nonlinear term in the amplitude expansion , we can reasonably estimate the frequency at values of @xmath314 of the order of 50 to 100% above threshold , say , as follows .
as we noted above , in this range , the band of unstable wavenumbers @xmath34 ranges from just over 2 to about 4.5 ( see fig .
[ f14 ] ) .
it is unlikely that the range of wavenumbers will change drastically due to nonlinear interactions , since in according to our calculations outside this range both the linear and the cubic term in the amplitude expansion are stabilizing . on the other hand , the frequency of oscillations , if one studies the pattern in a fixed lab frame , will get strongly renormalized : once the pattern gets well developed , we expect ( but have no proof ) that the nonlinearly modulated profile moves with the _ average speed _ @xmath312 rather than with the maximum speed of the unperturbed parabolic profile ( keep in mind that in linear order , the periodic linear eigenmode does not affect the average speed , as it averages out to zero over one wavelength , but that this does not remain true in nonlinear order ) .
which particular wavenumber will be selected nonlinearly ( or whether in fact a well - defined sharp wavenumber will be selected nonlinearly in a carefully controlled experiment ) we do not know , but with the assumption that it lies in the unstable band @xmath315 and that it moves with the average speed , we get the following estimate for the dimensional frequency@xcite @xmath316 which gives @xmath317 we finally turn to a brief discussion of our results for the case of a planar slit .
the main difference between the cylindrical and planar case is that in the latter case , the coefficient @xmath318 is always found to be positive .
this is also illustrated by figs .
[ f14 ] and [ f15 ] , which show that in the case of the planar slit there is no finite band of nonlinearly unstable wavenumbers for the critical shear rate ( fig .
[ f14 ] ) and critical shear stress ( fig .
[ f15 ] ) as a function of @xmath34 .
note also that in both cases these curves just have a single minimum , contrary to what we found for the cylindrical tube . as a result of this ,
the wavenumber corresponding the minimum of the amplitude shear rate critical value @xmath47 now decreases smoothly with @xmath10 , see fig .
[ f16 ] . as a function of @xmath34 in the planar case for three different values of @xmath10 . ] as a function of @xmath34 in the planar case for three different values of @xmath10 . ]
the absence of a finite band of wavenumbers and of a value of @xmath10 below which @xmath44 for all @xmath34 ( as in the cylindrical case ) , appears to have fewer practical implications than one might expect at first sight .
after all , the overall behavior of the minimal value of the critical amplitudes as a function of @xmath10 shows quite the same behavior as in the case of the cylinder , compare figs .
[ plaatje4 ] and [ plaatje5 ] : upon decreasing @xmath10 , the critical values rapidly decrease below @xmath319 . as we have stressed before , our expansion ceases to be valid in this regime where @xmath47 becomes of order unity .
we expect that in reality there is also a true saddle - node bifurcation point where the branch ends in the case of planar poiseuille flow ; possibly , our results indicate that in the planar case the corresponding critical value @xmath53 is lower than in the cylindrical case .
however , our results do imply that it is difficult to make a prediction for the wavenumber near this ( presumed ) critical value than in the case of the cylinder .
nevertheless , since the wavenumber of the mode whose threshold to nonlinear instability is smallest is about 1.8 , it seems reasonable to expect for the dimensional wavelength near threshold @xmath320 where the width of the slit equals @xmath64 . like for the case of flow in a pipe , according to our linear results the phase velocity is very close to the maximum velocity at the center line between the plates , so the above result for the most likely wavelength near threshold yields a frequency of about @xmath321 .
however , since we do not find a finite band of unstable modes above threshold , it is difficult to determine the range of frequencies expected further above threshold , even though we do expect here too that the patterns move roughly with the average velocity . of the wavenumber corresponding to the minimum of @xmath47 as a function of @xmath34 for the case of planar poiseuille flow . ]
in this article we have shown that for the simplest model of a polymer fluid with normal stress effects , the ucm model , poiseuille flow through a planar channel or cylindrical tube becomes weakly nonlinearly unstable for weissenberg numbers somewhat larger than unity
. stated differently , since the ucm model _ only _ includes the essential normal stress effect , we find that the nonlinear flow instability is characterized _ by the weissenberg number only _ , and the phenomenon appears to be very robust in that almost any more complicated polymer fluid model that includes normal stress effects will exhibit the same instability in the same range of weissenberg numbers .
we presented evidence in @xcite that this instability yields an intrinsic route to melt fracture behavior in the absence of other mechanisms such as stick - slip phenomena .
one should also keep in mind that our expansion is only carried out to lowest order in the nonlinearity , so one may wonder about the robustness of these results as long as higher order terms in the expansion are unknown .
investigations of these issues for couette flow and poiseuille flow are underway and will be reported in due course .
the critical weissenberg number we find is a factor 10 larger than the values for which atalik and keunings @xcite observed self - sustained quasi - periodic oscillations in their numerical simulations of poiseuille flow .
the two results are not necessarily inconsistent however : while all of our calculations apply to the ucm model , their simulations were done for the oldroyd - b model , with a viscosity ratio of @xmath12 and reynolds number @xmath322 . moreover ,
atalik and keunings added a stress diffusion term to their equations to improve numerical stability .
further analysis is clearly necessary to investigate the effects of these differences a detailed comparison of the simulations with the amplitude expansion results for one and the same model , is called for . just above the onset of a well - defined linear ( or `` supercritical '' ) instability threshold , the wavelength of the pattern is well - defined : upon approach of the threshold , the wavenumber of the pattern approaches the wavenumber @xmath323 at which the instability sets in . here , however , it is more difficult to draw sharp conclusions about the wavelength of the pattern close to onset .
first of all , our expansion can not be fully trusted quantitatively for weissenberg numbers of the order of @xmath53 , as the damping rate is large there , not small as is needed in our method .
this technical caveat aside , one should keep in mind that once the instability sets in the nonlinear flow behavior which will develop can not be addressed by our expansion method .
hence it is difficult to rule out the possibility of nonlinear interaction terms changing the velocity of the pattern in the die to a value different from the one we have assumed , namely the average flow velocity @xmath324 .
if on the other hand the flow pattern stabilizes in some weakly nonlinear regime , then one would expect its wavenumber to be close to the onset value we calculate and the frequency to be close to the one we have estimated in eq .
( [ transfreq ] ) . as a rule of thumb
, the wavelength of the undulations of the extrudate is typically about twice the diameter of the die . in conclusion , our nonlinear analysis establishes the nonlinear flow instability essentially beyond reasonable doubt and predicts onset values which are consistent with those reported experimentally @xcite .
moreover , the hypothesis of highly similar subcritical behavior in taylor - couette geometries @xcite and poiseuille flow ( and hence possibly melt fracture ) is fully confirmed by our calculations .
indeed , the similarity between the flow field perturbation shown in fig . [ velocityplot ] and the roll - type pattern which groisman and steinberg have argued gives rise to the subcritical nature of the instability @xcite in taylor - couette flow is striking .
from this perspective , the only difference between the two cases is that the general mechanism @xcite that in visco - elastic flows the curvature of the streamlines makes the flow _ linearly _ unstable _ is _ operative in taylor - couette cells but obviously _ not _ in poiseuille flow .
nonlinearly , the two flows appear to be much more closely connected . from this perspective
, the evidence for `` turbulence without inertia '' @xcite as a result of these elastic effects is intriguing .
wvs would like to thank daniel bonn for stimulating his interest in the polymer flow problem and daniel bonn for numerous discussions .
moreover , he is grateful to the lps of the ens in paris for its hospitality and to the french ministre des affaires trangres and french ministre de lducation nationale , de la recherche et de la technologie , for a descartes - huygens award which made frequent visits to paris possible .
in this section we give the expressions for the operators @xmath181 and @xmath184 from the equation @xmath325 in the planar case and the cylindrical case in dimensional form .
for the planar case we have @xmath326 where the operators @xmath327 are defined in the following way : @xmath328 and @xmath329 for the cylindrical case we get @xmath330 where the operators @xmath331 are defined in the following way : @xmath332 and the vector @xmath184 has components @xmath333
the coefficients of the equation for the streamfunction in the planar case : there is an important lesson from the earlier work on planar poiseuille flow : expansions to higher order in the amplitude do not appear to help to capture the nonlinear branch in the case of newtonian fluids .
herbert , aiaa journal * 18 * , 243 ( 1980 ) has extended the expansion of @xcite through terms of order 15 in the amplitude expansion , but the series does not appear to capture the physically relevant regime of reynolds numbers and amplitudes . | it is well known that the poiseuille flow of a visco - elastic polymer fluid between plates or through a tube is linearly stable in the zero reynolds number limit , although the stability is weak for large weissenberg numbers . in this paper
we argue that recent experimental and theoretical work on the instability of visco - elastic fluids in taylor - couette cells and numerical work on channel flows suggest a scenario in which poiseuille flow of visco - elastic polymer fluids exhibits a _ nonlinear `` subcritical '' instability _ due to normal stress effects , with a threshold which decreases for increasing weissenberg number .
this proposal is confirmed by an explicit weakly nonlinear stability analysis for poiseuille flow of an ucm fluid .
our analysis yields explicit predictions for the critical amplitude of velocity perturbations beyond which the flow is nonlinearly unstable , and for the wavelength of the mode whose critical amplitude is smallest .
the nonlinear instability sets in quite abruptly at weissenberg numbers around 4 in the planar case and about 5.2 in the cylindrical case , so that for weissenberg numbers somewhat larger than these values perturbations of the order of a few percent in the wall shear stress suffice to make the flow unstable . we have suggested elsewhere that this nonlinear instability could be an important intrinsic route to melt fracture and that preliminary experiments are both qualitatively and quantitatively in good agreement with these predictions . |
we would like to thank s. pratt for providing the crab program and en - guang zhao for helpful discussions and encouragements .
we acknowledge support by the frankfurt center for scientific computing ( csc ) . the work is supported in part by the key project of the ministry of education of china under grant no .
209053 and the national natural science foundation of china under grant no .
10675046 .
chen l w , ko c m and li b a. studying stiffness of nuclear symmetry energy through isospin diffusion in heavy - ion collisions .
phys rev lett , 2005 , 94:032701 li b a and chen l w. nucleon - nucleon cross sections in neutron - rich matter and isospin transport in heavy - ion reactions at intermediate energies .
phys rev c , 2005 , 72:064611 kowalski s , et al .
experimental determination of the symmetry energy of a low density nuclear gas .
phys rev c , 2007 , 75:014601 tsang m b , zhang y , danielewicz p , famiano m , li z , lynch w g and steiner a w. constraints on the density dependence of the symmetry energy .
phys rev lett , 2009 , 102:122701 li q. the influence of reconstruction criteria on the sensitive probes of the symmetry potential .
mod phys lett a , 2009 , 24:41 hanbury - brown r and twiss r q. philos mag , 1954 , 45:663 ; nature ( london ) , 1956 , 178:1046 bauer w , gelbke c k and pratt s. hadronic interferometry in heavy ion collisions .
ann rev nucl part sci , 1992 ,
42:77 chen l w , greco v , ko c m and li b a. isospin effects on two - nucleon correlation functions in heavy - ion collisions at intermediate energies .
phys rev c , 2003 , 68:014605 chen l w , ko c m and li b a. effects of momentum - dependent nuclear potential on two - nucleon correlation functions and light cluster production in intermediate energy heavy - ion collisions .
phys rev c , 2004 , 69:054606 li q , li z , soff s , bleicher m and stcker h. medium modifications of the nucleon - nucleon elastic cross section in neutron - rich intermediate energy hics .
j phys g : nucl part phys , 2006 , 32:407 verde g , danielewicz p , lynch w g , brown d a , gelbke c k and tsang m b. probing transport theories via two - proton source imaging .
phys rev c , 2003 , 67:034606 | based on the updated urqmd transport model , the effect of the symmetry potential energy on the two - nucleon hbt correlation is investigated with the help of the coalescence program for constructing clusters , and the crab analyzing program of the two - particle hbt correlation .
an obvious non - linear dependence of the neutron - proton ( or neutron - neutron ) hbt correlation function ( @xmath0 ) at small relative momenta on the stiffness factor @xmath1 of the symmetry potential energy is found : when @xmath2 , the @xmath0 increases rapidly with increasing @xmath1 , while it starts to saturate if @xmath3 .
it is also found that both the symmetry potential energy at low densities and the conditions of constructing clusters at the late stage of the whole process influence the two - nucleon hbt correlation with the same power .
keywords : density dependence of the symmetry energy , sensitive observables , hbt correlation in recent years , the isospin dependent equation of state ( eos ) of both infinite and finite nuclear matter has been investigated in both deeper and broader range . theoretically , a large amount of probes have been brought out to be sensitively affected by the isovector part of the eos , i.e. , the symmetry energy term . based on different theories
, the symmetry energy has been parametrized in different ways ( please see @xcite for details ) . among these parametrizations ,
the @xmath4 is often used ( @xmath5 being the symmetry energy at the normal density , and @xmath6 the reduced density , while @xmath1 is the stiffness factor of the density dependence of the symmetry energy ) . especially , the density dependence of the symmetry potential energy is paid much more attention since the uncertainty of the symmetry energy mainly comes from this term .
recently , with some comparisons between the results of experimental data and dynamical model ibuu , the uncertainty in the density dependence of the symmetry energy at subnormal densities is reduced into the range of @xmath7 , depending on the probable medium modifications of the nucleon transport @xcite . nevertheless , recently a quite soft symmetry energy ( the corresponding @xmath1 factor is only about @xmath8 ) was implied experimentally at very low nuclear densities ( @xmath9 is @xmath10 times normal density ) @xcite .
furthermore , in the most recent analyses by tsang s group @xcite , although they found that several different observables can provide consistent constraints on the density dependence of the symmetry energy , the constraints are still relatively loose . to bring forward more sensitive probes for symmetry energy at subnormal densities is still urgent and prerequisite . however , we think that one should firstly check more carefully the difficulties of detecting the existing observables for forthcoming experiments . therefore , the currently crucial question becomes : can the sensitive observable suggested by theoretical physicists be _ experimentally _ taken as a sensitive candidate for detecting the symmetry energy at subnormal densities ? in the previous work @xcite this question has been concerned for the free neutron / proton ( @xmath11 ) ratio as well as the @xmath12 ratio . in this paper , we would like to continue this topic to further check the two - nucleon correlation , which is based on the hanbury - brown - twiss interferometry ( hbt ) @xcite technique .
it was pointed out that it is a good candidate to probe the symmetry energy since it is quite unsensitive to the iso - scalar part of eos as well as the in - medium nn cross section @xcite . in their initiated investigations ,
the isospin effect is found to occur only in the late stage of the heavy ion collisions ( hics ) where the nuclear medium becomes dilute .
therefore , the two - nucleon correlation has the advantage in probing the density dependence of the symmetry potential at low densities .
it was further claimed that the sensitivity of the correlation function of the nucleon pair to the density dependence of the symmetry energy is reduced after considering a ( isospin - dependent ) momentum dependent nuclear potential @xcite .
however , in their studies , only several parameter sets of the stiffness of the symmetry energy were employed .
it is believed to be enough in the studies of the density dependent symmetry energy by adopting the conventional single - particle observables since they behave monotonously with the increase of the stiffness factor @xmath1 .
but , is that still true in the symmetry - energy dependence of the two - nucleon hbt correlation ?
here we would like to vary the @xmath1 factor with a larger range to see the variance of the correlation function .
besides the symmetry potential energy , the freeze - out conditions of clusters should also heavily influence the final two - nucleon correlation function . in the transport model calculations , when the transport is stopped at a certain time , such as @xmath13 @xmath14 or so , one cluster constructing program is then used . generally speaking , two methods ( i.e. , buu - like and qmd - like ) for constructing clusters exist in the transport model calculations @xcite . in our urqmd calculations below , the coalescence model , which is often used for analyzing qmd - like model outputs ,
is employed , in which the nucleons with relative momenta smaller than @xmath15 and relative distances smaller than @xmath16 are considered to belong to one cluster .
it is comparable to the buu - like coordinate density cut @xmath17 when we only adopt the relative distance @xmath16 but not the relative momentum @xmath15 in the qmd - like analysis .
the freeze - out condition certainly influences the two - nucleon correlation function since it reflects the spatial ( and momentum ) separation of two nucleons .
it is also interesting to see if the small uncertainty in the multiplicity of the observed particle changes the hbt correlation function accordingly . in this paper
, we use the recently updated urqmd transport model @xcite for studies of intermediate energy hics to investigate the effects of the density dependent symmetry potential energy and the freeze - out conditions on the neutron - proton and the neutron - neutron hbt correlations .
the proton - proton correlation will not be shown in this paper since it has similar isospin - dependent effect to the neutron - proton and neutron - neutron ones , although the largest effect is shown at the relative momentum of pair @xmath18 mev@xmath19 but not at @xmath20 mev@xmath19 . as a default , @xmath15 and @xmath16 used in the coalescence model
are set to be @xmath8 gev@xmath19 and @xmath21 fm , respectively ( except otherwise stated ) .
the central collisions ( @xmath22 of total cross section @xmath23 ) of neutron - rich intermediate - mass and heavy systems @xmath24ca@xmath25ca and @xmath26au@xmath27au are adopted in the urqmd calculations . as stated above
, the correlation function of the nucleon pair from hics with the super - intermediate - mass system at the intermediate energy is claimed to be unsensitive to the isospin - scalar part of eos and the nn cross section @xcite .
it is also reported that by using the imaging method the shape of the source of the two - proton pairs from the hics with medium sized system at beam energies around @xmath28a mev is sensitive to the medium modification of the nn cross section @xcite .
however , the corresponding hbt correlation function is seen to be less and less affected by the medium modification with the increase of the beam energy . in this work
, we focus our investigation on the correlation functions at beam energies @xmath28a @xmath29a mev . and
, we randomly adopt several parameter sets of eos used in previous works @xcite , for checking .
those are , s - eos , sm - eos , and hm - eos , without considering the medium modifications on nn cross sections . to analyze the two - nucleon hbt interferometry , the correlation after burner ( crab v3.0@xmath30 ) program is employed @xcite .
the correlation function is expressed as @xmath31 , where @xmath32 is the inclusive two - particle emission probability in the same event , @xmath33 is the product of two single - particle emission probability in different events .
the squared relative two - particle wave function @xmath34 serves as a weight , while @xmath35 is the probability for emitting a particle with momentum @xmath36 from the space - time point @xmath37 .
the @xmath38 and the @xmath39 are the total and relative momenta of the particle pair , respectively . ) within the relative momentum @xmath40 mev@xmath19 and the rapidity region @xmath41 ( top plot ) , and free neutron - proton ratio in the rapidity region @xmath42 ( bottom plot ) as a function of the stiffness factor ( @xmath1 ) of the symmetry potential energy . the s - eos and sm - eos
are adopted for @xmath24ca@xmath25ca central collisions ( @xmath22 of total cross section @xmath23 ) at @xmath43a mev .
the marked area illustrates the value @xmath44 .
the open squares in the top plot show the @xmath45 values when the symmetry potential energy is switched off at the reduced densities @xmath46.,scaledwidth=60.0% ] figure [ fig1 ] shows the neutron - proton correlation function ( @xmath45 ) within the relative momentum bin @xmath40 mev@xmath19 and the rapidity region @xmath41 ( @xmath47 , @xmath48 and @xmath49 are the energy and longitudinal momentum of the nucleon in the center - of - mass system ) ( top plot ) , and the free @xmath11 ratio in the rapidity region @xmath42 ( bottom plot ) as a function of the stiffness factor @xmath1 of the symmetry potential energy .
the @xmath24ca@xmath25ca central collisions ( @xmath22 of total cross section @xmath23 ) at @xmath43a mev are adopted .
it is known that the @xmath11 ratio of free nucleons is obviously rapidity dependent . but
conclusions of the two - nucleon hbt correlation drawn in this paper is not changed by the small rapidity - cut difference used in both observables @xcite .
the s - eos and the sm - eos are selected for comparison in each plot .
it is seen that the iso - scalar part of the eos only affects slightly the sensitivity of both the @xmath45 and the @xmath11 ratio to the stiffness of the symmetry potential energy .
thus , in the following calculations , we will randomly use various eos .
it is also known that the @xmath11 value decreases monotonously with increasing @xmath1 factor , which is again shown in the bottom plot of this figure .
it is interesting to see that , however , the @xmath45 moves up rapidly with increasing @xmath1 when @xmath1 is less than @xmath50 , and then saturates with the further increase of the @xmath1 value , the saturation value is @xmath44 , which is shown in a marked area .
it is known that the symmetry potential plays obvious role on the two - nucleon hbt correlation at the late stage ( @xmath51 @xmath14 ) of the whole process @xcite , where the density surrounding the emitted protons is very small .
therefore , we further calculate the @xmath45 with @xmath52 and @xmath53 ( open squares ) but without considering the symmetry potential energy at very low densities ( the reduced density @xmath54 ) . it is obvious that the @xmath45 values move back to the marked area . we know that the stiffer the symmetry energy is , the smaller the value of the symmetry energy at subnormal density becomes . at densities
@xmath46 the stiff symmetry energy is too weak to influence the two - particle correlation , while it plays strong role with a very soft symmetry energy @xcite .
it is concluded that , firstly , the correlation function can hardly distinguish the stiffness of the symmetry energy when @xmath55 .
secondly , the soft symmetry energy plays more obvious role on the neutron - proton correlation function at the late stage .
therefore , the neutron - proton correlation function is suitable to explore the symmetry energy in the dilute nuclear medium , especially for a very soft symmetry - energy assumption . at beam energy around @xmath28a mev
, it is found that the isospin effect on the correlation function is reduced in heavy - system hics such as sn+sn .
it is also confirmed in our calculations with the au+au reaction which is shown in the upper - left plot of figure[fig2 ] , even when a different stop time @xmath56 @xmath14 is chosen .
further , we calculate the @xmath45 values at higher beam energies , i.e. , at @xmath57a , @xmath58a , @xmath59a , @xmath60a , and @xmath61a mev , which are shown in other plots of figure [ fig2 ] .
it is interesting to see that with the increase of the beam energy , the effect of the density dependence of symmetry energy becomes visible
. it might be understandable when considering the fact that with the increase of beam energy , more unstable light fragments emit and decay with a longer time and in the dilute medium so as to be influenced by the soft symmetry energy at the late stage . a , @xmath62a , @xmath58a , @xmath59a , @xmath60a , and @xmath61a mev .
the urqmd calculations are stopped at @xmath13 @xmath14 . at @xmath43a mev ,
the results at @xmath56 @xmath14 are also shown .
the hm - eos and the symmetry energies with @xmath63 and @xmath64 are adopted in calculations.,scaledwidth=80.0% ] for the neutron - neutron correlation , the phenomena shown in figs.[fig1 ] and [ fig2 ] are even more obvious . in figure[fig3 ] we show the @xmath1-dependence of the correlation functions of both neutron - proton and neutron - neutron pairs from central au+au collisions at @xmath65a mev .
the strong non - linear dependence of the neutron - neutron correlation function on the @xmath1 factor of the symmetry potential energy is due to the long time evolution of the neutron - rich unstable light fragments at the late stage .
mev@xmath19 and the rapidity region @xmath41 as a function of the stiffness factor ( @xmath1 ) of the symmetry potential energy .
the hm - eos is adopted for au+au central collisions at @xmath65a mev .
, scaledwidth=70.0% ] in order to deeper understand the isospin effect on the correlation function , it is necessary to check the influence of the freeze - out conditions on the hbt correlator .
figure + [ fig4 ] illustrates the neutron - proton correlator as a function of the relative momentum @xmath66 of the nucleon pair .
central @xmath24ca@xmath25ca collisions at @xmath43a mev are calculated with urqmd and sets of relative momentum ( @xmath15 ) and relative distance ( @xmath16 ) are chosen for the following cluster construction in the coalescence model .
the final correlators by the crab analyzing program are shown with different lines in the figure .
the multiplicities of the light clusters ( @xmath67 ) with different ( @xmath15,@xmath16 ) sets are shown in the top - right plot .
the stiffness of the symmetry potential energy @xmath68 is chosen in current calculations .
when the values of ( @xmath15 , @xmath16 ) are adjusted from ( @xmath8 gev@xmath19 , @xmath69 fm ) to ( @xmath70 gev@xmath19 , @xmath21 fm ) , the multiplicities of clusters keep almost unchanged , and so does the correlator .
however , when the values of ( @xmath15 , @xmath16 ) are changed from ( @xmath8 gev@xmath19 , @xmath69 fm ) to ( @xmath8 gev@xmath19 , @xmath21 fm ) , the multiplicities of free nucleons alter slightly , and so does the correlator .
this vivid phenomenon reveals that the freeze - out condition influences the value of the correlator visibly and should be paid more attention . in order to constrain the density dependence of the symmetry energy by using the two - nucleon hbt correlation , the uncertainty of the multiplicity of nucleons , with whatever experimental cuts ,
should be largely reduced firstly .
ca@xmath25ca collisions at @xmath43a mev are illustrated with different relative momenta @xmath15 and relative distances @xmath16 in the coalescence model , and in the top - right plot the multiplicities of light clusters ( @xmath67 ) are shown , correspondingly . the s - eos and the symmetry potential energy with @xmath68 are used .
, scaledwidth=70.0% ] to summarize , based on the updated urqmd transport model , the coalescence program for constructing clusters , and the crab analyzing program of the two - particle hbt correlation , a more detailed analysis of the effect of the symmetry potential energy on the two - nucleon hbt correlation is investigated for hics with super medium - sized system at intermediate energies .
it is found that , for hics with medium - sized neutron - rich system at @xmath28a mev , there exists a non - linear dependence of the neutron - proton ( or neutron - neutron ) hbt correlation function ( @xmath0 ) at small relative momenta @xmath66 on the stiffness factor @xmath1 of the symmetry potential energy : when @xmath71 , the @xmath0 at small @xmath66 increases rapidly with increasing @xmath1 , while it starts to saturate if @xmath3 .
this phenomenon is also seen for the hics with the heavy system at higher beam energies . at @xmath65a mev ,
the non - linear dependence of the @xmath72 is even more obvious than the neutron - proton case . with the upgrading of the land facility at gsi to the `` newland ''
, it is expected to detect the stiffness of the symmetry energy with more neutron - related observables in the near future . in order to understand the non - linear dependence of the hbt correlator on the @xmath1 factor ,
the different freeze - out conditions , such as the stop times and the ( @xmath15,@xmath16 ) parameter sets for the coalescence model , are taken into account .
it is found that both the symmetry potential energy at low densities and conditions of constructing clusters at the late stage of the whole process influence the two - nucleon hbt correlation with the same power and should be studied more carefully . |
polymers , whether linear or branched , typically display a two phase structure in solvent with a @xmath0-transition @xcite , whose position depends on temperature and solvent composition , intervening between an extended and a collapsed phase .
the case of linear polymers has proved amenable to various analytical treatments such as conformal field theory and coulomb gas methods @xcite , but branched polymers have been rather more stubborn . a discretized model of branched polymers is provided by lattice animals , in essence graphs of connected sites on some lattice .
contact interactions exist between nearest neighbour sites that are not directly linked by a bond of the animal and solvent interactions exist between occupied sites and unoccupied nearest - neighbour sites .
all the ingredients of the continuum polymer in solvent problem are thus present .
it is known that in @xmath3 dimensions the lattice animal exponents are related to those of the yang - lee edge singularity in @xmath4 dimensions which allows the calculation of the bulk entropic exponent for the so - called strongly embedded lattice animal @xcite .
it is also known that the model corresponds exactly to the @xmath1 limit of a certain extended @xmath5-state potts model @xcite , which has led to the conjecture that a percolation critical point separates two different branches of the @xmath0 line @xcite .
more controversially , exact enumeration techniques have suggested that _ two _ collapsed phases might exist @xcite whereas transfer matrix investigations have found no sign of this @xcite .
this conflict motivated henkel and seno @xcite to conduct a two - pronged investigation of lattice animals using the extended potts model formalism .
they attacked the infinite dimensional , mean - field problem by solving the model on a bethe lattice and the low - dimensional problem by using migdal - kadanoff @xcite recursion relations . in both cases they found only one collapsed phase , and observed that it was unlikely that the topology of the phase diagram would change in intermediate dimensions . in this paper
we also use the extended potts model formalism , this time on @xmath2 random graphs of arbitrary topology , to show that only one collapsed phase exists for these .
a couple of interesting technical points emerge along the way : firstly , ratios of the saddle point equations in the random graph model are isomorphic to the fixed points of the recursion relations that solve the model on the bethe lattice , so the phase structure on @xmath2 graphs in the large @xmath6 limit is identical in all respects to the bethe lattice ; secondly , the symmetry breaking pattern that emerges at the @xmath0-transition is not that which one might have naively expected given earlier results for potts models and percolation on random graphs .
the former phenomenon also occurs in other models on random graphs @xcite and is heuristically explained by noting that the loops in the random graph model are predominantly large in the thermodynamic limit @xcite , so the random graphs look locally like the corresponding bethe lattice and one generically obtains mean field behaviour . the partition function for lattice animals
may be written in several equivalent ways , but we restrict ourselves to that in equ.([an1 ] ) to facilitate easy comparison with @xcite .
if an animal contains @xmath6 sites , @xmath7 bonds and @xmath8 contact interactions we can write @xmath9 where the site , bond and contact fugacities are given by @xmath10 respectively and the @xmath11 s give the number of different animals with the specified numbers of sites , bonds and contacts .
there is an exact correspondence between the lattice animal partition function of equ.([an1 ] ) and the @xmath1 limit of an extended @xmath5 state potts model @xcite with hamiltonian @xmath12 where the @xmath13 sums are over nearest neighbour vertices and the spins @xmath14 take on @xmath5 values .
the @xmath15 term is non - standard , in its absence we would have the usual potts hamiltonian ( with an external field ) .
the potts model parameters are related to @xmath16 and @xmath17 in equ.([an1 ] ) by @xmath18 where @xmath19 is the coordination number of the lattice or graph on which the animals live , and the partition functions are related by @xmath20 where the potts partition function is @xmath21 the idea behind the correspondence is essentially identical to the use of the standard potts model hamiltonian in the @xmath1 limit to describe percolation @xcite , which is based fortuin and kastelyn @xcite cluster transcription of potts and related models .
having discussed the lattice animal model in general we now move on to formulate it on random graphs of arbitrary topology .
a particularly economical way of writing down statistical mechanical models on such random graphs was introduced in @xcite where it was observed that taking the @xmath22 limit in the hermitian matrix models familiar from studies of 2d gravity generated random graphs of arbitrary topology , as opposed to the planar graphs picked out by the @xmath23 limit that is used in 2d gravity .
the random graphs appear as feynman diagrams in the perturbative expansion of the partition function .
the matrix indices vanish in the @xmath22 limit , leaving `` thin '' graphs rather than the fat , or ribbon , graphs of the matrix model proper .
we shall denote the non - planar , arbitrary topology random graphs as thin graphs and the animals which live on them as thin animals throughout the rest of the paper for brevity .
in general the partition function of a statistical mechanical model on an ensemble of thin random graphs with @xmath24 vertices is given by @xmath25 where the contour integral over the vertex coupling @xmath26 picks out the graphs with @xmath24 vertices , @xmath27 is an appropriate action , @xmath28 is the inverse of the quadratic form in this action , and the @xmath29 are the various matter variables in the action .
as the number of thin random graphs increases factorially with @xmath30 the factor @xmath31 gives the number of undecorated ( i.e. without matter ) graphs in the class of interest and disentangles this growth from any non - analyticity due to phase transitions .
for the @xmath2 ( 3-regular ) random graphs we discuss here @xmath32 it is perhaps worth emphasising that the @xmath29 appearing in equ.([part ] ) are _ scalar _ variables , so the evaluation of thin graph partition functions is inherently easier than dealing with the matrix integrals of the planar limit . in the large @xmath30 , thermodynamic ,
limit saddle point methods may be used to evaluate equ.([part ] ) . the saddle point equation for @xmath26
may be trivialised by scaling it out of the action as an overall factor , leaving any critical behaviour to be revealed by the behaviour of the saddle point equations for the matter fields @xmath29 .
phase transitions are manifested by an exchange of dominant saddle point , either continuously or discontinuously .
we can now write down the action which gives the boltzmann factors appropriate to the hamiltonian of equ.([ham ] ) for a @xmath5-state potts model on thin @xmath2 random graphs @xmath33 where @xmath34 and we have pre - emptively scaled out the cubic vertex coupling @xmath26 as it plays no role in the determination of the phase structure .
the inverse of the quadratic form in equ.([action ] ) has the diagonal terms @xmath35 with all off - diagonal terms 1 , which is precisely what is required .
we can also see that when @xmath36 we have @xmath37 , @xmath38 , and we recover ( up to a trivial rescaling ) the action for the standard potts model used in @xcite .
the saddle point equations for the action in equ.([action ] ) may be solved for increasing @xmath5 until patience , or algebraic computing power , run out . `` generic '' behaviour sets in at @xmath39 , as opposed to @xmath40 for the standard action with @xmath36 .
one finds a disordered phase where @xmath41 is distinct and the @xmath42 @xmath29s are equal to @xmath43 , which we denote as having @xmath44 symmetry .
there are two patterns of symmetry breaking from this phase to ordered states : in the first we have @xmath41 , one distinct @xmath45 and the other @xmath46 equal to @xmath43 , which we denote as @xmath47 ; in the second we find @xmath41 , @xmath48 and @xmath49 which we denote by @xmath50 .
we now follow the path used in the analysis of @xcite by writing down effective actions that are sufficient to capture these symmetry breaking patterns . in effect , most of the @xmath5 potts variables are redundant as @xmath51 or @xmath52 of them are equal in the ordered phases and @xmath42 in the disordered phase . imposing the appropriate symmetries we find the effective actions @xmath53 and @xmath54 solving the saddle point equations @xmath55 for either effective action to get @xmath56 involves only a quadratic equation in the disordered phase , but one is left with a messy quartic to solve in the ordered phase . a direct attack is not very illuminating , but we can use the same legerdemain exercised in @xcite for bethe lattice to simplify the problem greatly .
the useful observation made there was that the lattice animal partition function could be written as @xmath57 where @xmath58 was the canonical ensemble ( fixed @xmath6 ) free energy .
the sum in equ.([gc ] ) can be seen to diverge at a critical value of @xmath59 , which we denote as @xmath60 , which in turn gives the critical canonical ensemble free energy @xmath61 of the infinite lattice animal .
therefore , if we can calculate @xmath62 as a function of @xmath63 we obtain the canonical free energy @xmath64 of the infinite lattice animal as a function of these variables . to see how this helps , consider the solutions of the saddle point equations obtained from equs.([subdom],[dom ] ) when @xmath65 , which are given by @xmath66 for both in the disordered phase , with symmetry @xmath67 , where @xmath68 and where we have not solved the quadratics explicitly for reasons that will become apparent below .
the ordered phase solutions are obtained from , for equ.([subdom ] ) with breaking pattern @xmath69 @xmath70 \tilde \phi^3 + \left [ 4 ( \tau - 1)^2 - 6 y ( 1 - \tau ) + y^2 ( 1 + x y ) \right ] \tilde \phi^2 \\ \nonumber & - & \left [ 4 ( 1 - \tau)^2 + x y^2 ( 2 \tau - y ) + 2 y ( \tau - 1 ) \right ] \tilde \phi + \left [ ( \tau - 1 ) ^2 + \tau x y^2 \right ] \\ \nonumber & & \\
\nonumber \phi & = & 1 - \tilde \phi \\ \nonumber & & \\ \nonumber \psi & = & - { \tau^2 \over y } \left [ y ( \tilde \phi^2 - \tilde \phi ) - 2 ( \tau - 1 ) \tilde \phi + ( \tau - 1 ) \right ] \end{aligned}\ ] ] and , for equ.([dom ] ) with breaking pattern @xmath50 @xmath71 \tilde \phi^3 + \left [ 16 ( \tau - 1)^2 + 12 y ( 1 - \tau ) + y^2 ( 1 + x y ) \right ] \tilde \phi^2 \\ \nonumber & - & \left [ 16 ( 1 - \tau)^2 - x y^2 ( 4 \tau - y ) - 4 y ( \tau - 1 ) \right ] \tilde \phi + \left [ 4 ( \tau - 1 ) ^2 - 2 \tau x y^2 \right ] \\ \nonumber & & \\ \nonumber \phi & = & 1 - \tilde \phi \\ \nonumber & & \\ \nonumber \psi & = & - { \tau^2 \over y } \left [ y ( \tilde \phi^2 - \tilde \phi ) + 4 ( \tau - 1 ) \tilde \phi - 2 ( \tau - 1 ) \right].\end{aligned}\ ] ] to avoid the onerous task of solving for @xmath72 in all of these we now , as advertised , treat equ.([solhot ] ) and the first of equs.([solsub],[soldom ] ) as equations for @xmath73 , in which variable they are only linear .
if we make the substitution @xmath74 to simplify the resulting expressions we find from equs.([solhot],[solsub],[soldom ] ) @xmath75 where we have labelled the solutions by their symmetry breaking pattern .
we see here the first hint of a close parallel with the bethe lattice solution of @xcite since our two ordered solutions are related by @xmath76 which mirrors a similar symmetry in @xcite . the physical solutions @xmath77 are obtained by maximising these expressions with respect to @xmath78 , which plays the role of an order parameter .
the appropriate solution for a given @xmath79 is then the largest of these .
bearing in mind that we must have the fugacity @xmath59 positive to make sense physically and that the natural ranges of @xmath80 and @xmath17 are @xmath81 , @xmath82 , we see that the physical ranges for @xmath78 are @xmath83 for @xmath84 , @xmath85 for @xmath86 and @xmath87 for @xmath88 . at this point
it might appear that we have succeeded in showing precisely the reverse of what was announced in the introduction in that we have found _ two _ candidate ordered phases , which might be identified with the two contentious collapsed polymer phases .
however , an extensive numerical investigation throughout the @xmath79 range reveals that the one of the solutions , @xmath89 , is always larger than the other , @xmath90 , for a given @xmath79 so it is the former which is always dominant . for the disordered solution we clearly have a maximum at @xmath91 @xmath92
so the free energy is @xmath93 and we also find @xmath94 we are now in a position to map out the phase structure of thin animals .
we have found two phases , whose free energies are given by @xmath95 which we can identify with the extended phase and @xmath96 which we identify with the ( single ) collapsed phase .
the transition line between these two phases may be pinned down by noting that @xmath97 iff @xmath98 .
we then observe that @xmath99 so @xmath100 has a maximum at @xmath91 if @xmath101 , a minimum if @xmath102 and a turning point if @xmath103 . a glance at figs.1,2 where we plot @xmath104 and @xmath105 for @xmath106 and @xmath107 respectively clarifies what is going on . in fig.1
we see that @xmath104 and @xmath105 share a common maximum at @xmath91 when @xmath106 ( which is the case for all @xmath108 ) , whereas for @xmath107 the maximum of @xmath105 lies at some @xmath109 .
this means that we see a continuous transition as @xmath17 is varied across @xmath98 when @xmath110 between extended animals at small @xmath17 and collapsed animals at large @xmath17 . for @xmath102 the position of the critical line in the @xmath79 plane must be determined numerically and the transition is first order .
the physical solution jumps discontinuously from the maximum of @xmath88 at @xmath91 to the maximum of @xmath86 at @xmath109 .
the appropriate equations for pinpointing the first order line are thus @xmath111 which give the conditions for the maximum in @xmath112 at @xmath109 to be equal to the maximum in @xmath113 at @xmath91 . a triple point at @xmath114
separates the first and second order transition lines .
all of these features are indicated in the schematic phase diagram of fig.3 .
the presence of a jump across the first order line is confirmed explicitly by the sample values in figs.4,5 where we see that as we cross the line by varying @xmath17 at fixed @xmath107 we move from @xmath88 dominant in fig.4 to the situation in fig.5 where @xmath86 is just equal to the maximum of @xmath88 ( @xmath115 at the origin ) for a _ non - zero _ @xmath78 value ( @xmath116 ) at @xmath117 . at this point @xmath78
jumps discontinuously and we move into the collapsed phase as @xmath17 increases .
when moving across the second order portion of the line however , the maximum of @xmath86 is at the origin and we see no jump in @xmath78 , giving a second order transition .
this can be seen by referring back to fig.1 and comparing it with fig.6 . in fig.6
we are _ below _ the second order line at @xmath118 and the two curves look much like fig.4 .
as we move up to the line at @xmath119 in fig.1 we can see that the maximum of @xmath86 now lies at the origin and increasing @xmath17 further takes us into the collapsed phase via a second order transition .
in closing this section it should be remarked that although the broad features of the random graph phase diagram appear to be in accordance with two - dimensional models , there is one notable difference .
the triple point separating the first and second order portions of the @xmath0 line is the percolation point in @xmath120 .
that is _ not _ the case here the percolation point in the random graph model is at @xmath121 and @xmath98 , which in turn implies @xmath122 . model . ]
this presumably can be ascribed to the mean - field nature of the model , a comment which also applies to the bethe lattice results of @xcite , whose similarities with the work here we now move on to discuss .
anyone familiar with the results of @xcite for bethe lattice animals may at this point be experiencing a strong sense of dj vu . the thin animal phase diagram has proved to be identical to that on the bethe lattice , down to non - universal quantities such as the position of the @xmath0 line and triple point . as we have already noted , even before extracting the physical solutions @xmath60 parallels are apparent since the two ordered solutions on both random graphs and the bethe lattice possess a particular @xmath123 symmetry .
our criteria for extracting the @xmath0 line are also identical to those used on the bethe lattice , although the equations for @xmath59 on the bethe lattice are quadratic rather than linear , as above .
all this clearly begs an explanation .
our saddle point equations for @xmath86 at @xmath65 , which we have so far not written explicitly , are given by @xmath124 at first sight these bear little relation to the fixed points of the recursion relations used in @xcite to obtain the dominant solution to the model on the bethe lattice .
the method of derivation is certainly completely different . on the bethe lattice
one considers the behaviour of the central spin @xmath125 @xcite , writing @xmath126 where @xmath127 is the spin on the @xmath128th sub - branch and @xmath129 and the primed sums are over the sub - branch with first spin @xmath130 .
the solution then follows by defining @xmath131 where @xmath6 labels the `` shell '' of the bethe lattice , and noting that it satisfies the recursion relation @xmath132 defining @xmath133 and taking the limit @xmath134 allows one to write the fixed point of the recursion relations in equ.([curse ] ) as @xmath135 where @xmath136 , @xmath137 .
given the identical phase diagram we have found on random graphs to the bethe lattice results of @xcite it is clear that the two sets of equations , equs.([ours ] , [ his ] ) , must somehow be equivalent .
this can be made explicit by a simple transformation . equs.([ours ] ) may be symbolically written as @xmath138 , where @xmath139 is the column vector @xmath140 and @xmath141 is @xmath142 .
we are at liberty to rewrite these as @xmath143 as @xmath144 is invertible in general @xmath145 if we now divide @xmath146 and @xmath147 @xmath148 and make the identifications @xmath149 we obtain exactly the two recursion equation fixed points of equ.([his ] ) .
although we have two distinct physical systems , closed random graphs of arbitrary topology which _ do _ contain loops , and the loop - less bethe lattice the preceding calculation shows that the thermodynamic behaviour of lattice animals on both is identical . in a precise sense the loops do not matter , as the saddle point equations that govern the random graph thermodynamic limit are isomorphic to the recursion relations that arise in solving the loop - less bethe lattice .
the identity between the bethe lattice recursion relations and ratios of the random graph saddle point equations also occurs for other systems ( standard ising , potts spins etc . )
@xcite , where the consequences are the same , namely the random graph solutions are identical to those on the bethe lattice .
the astute reader might remark that the solution of the lattice animal problem on random graphs presented in this paper does not really provide any further evidence against the existence of a second collapsed phase , as we have also shown that the solutions to the model have the same content as the bethe lattice solution in @xcite , though this is an interesting result in itself .
one could counter the objection by stating that we have found a different physical system , and solved it by very different means , in order to arrive at these conclusions .
the thin graph solution also clarifies one lacuna in the bethe lattice approach .
as we have noted , solving the full set of saddle point equations @xmath150 ( i.e. _ before _ going to the effective actions ) shows that only one disordered and two ordered phases are present for arbitrary @xmath5 . on the bethe lattice
enough order parameters are introduced by hand to distinguish the two putative ordered phases , which is then shown to be self - consistent .
the possibility of a more exotic phase diagram is not , however , explicitly excluded .
the random graph saddle point equations show that this really is enough , as no other patterns of symmetry breaking are observed for any @xmath5 .
the observed symmetry breaking pattern itself on random graphs is unexpected , namely @xmath151 .
for the standard @xmath5-state potts model one has @xmath152 , so the pattern observed for the sub - dominant solution @xmath153 might have been a more obvious first guess .
we have also remarked that the critical percolation point , @xmath154 on thin graphs , is not equal to the tricritical point @xmath155 that separates the two portions of @xmath0 line , in contrast to the @xmath120 model , although the general topology of the phase diagram is similar .
although the work presented in this paper is fairly self - contained one obvious generalisation suggests itself .
the action of equ.([action ] ) will also serve to represent lattice animals on _ planar _ random graphs if one takes the @xmath156 to be @xmath157 hermitian matrices and performs the limit @xmath23 .
the @xmath121 case ( percolation ) has already been solved by kazakov @xcite , which suggest that the animal problem might also be tractable .
the interest in doing this is that it could shed light on a another vexed question for animals the ( lack of ) conformal invariance in the 2d model . as a technical point , sidestepping the solution of difficult saddle point equations by considering the canonical free energy and solving for the fugacity associated with the external field is an idea which might also be fruitfully applied to the solution of the standard potts ( or even ising ) models in an external field on random graphs , where similarly complicated equations arise .
cardy , `` conformal invariance and statistical mechanics '' in _ fields , strings and critical phenomena _ , les houches xlix , e. brzin and j. zinn - justin eds , north holland amsterdam ( 1990 ) ; + p. christe and m. henkel , _ introduction to conformal invariance and its applications to critical phenomena _ , springer verlag heidelberg ( 1993 ) ; + b. nienhuis , `` coulomb gas formulation of two dimensional phase transitions '' in _ phase transitions and critical phenomena _
vol 11 , c. domb and j.l .
lebowitz eds , academic press ( 1987 ) ; + b. duplantier and h. saleur , phys . rev . lett . * 59 * , 539 ( 1987 ) ; + c. vanderzande , a.l .
stella and f. seno , phys .
lett . * 67 * , 2757 ( 1991 ) g. parisi and n. sourlas , phys .
lett . * 46 * , 871 ( 1981 ) m. giri , m.j .
stephen and g.s .
grest , phys .
b16 , 4971 ( 1977 ) ; + f.y .
wu , j. stat . phys . * 18 * , 115 ( 1978 ) ; rev .
phys . * 54 * , 235 ( 1982 ) ; + a. coniglio , j. phys . *
a16 * , l187 ( 1983 ) s. flesia , d.s .
gaunt , c.e .
soteros and s.g .
whittington , j .
phys . * a25 * , l1169 ( 1992 ) ; + c. vanderzande , phys . rev . lett .
* 70 * , 3595 ( 1993 ) f. seno and c. vanderzande , j. phys .
* a27 * , 5813 ( 1994 ) ; corr .
* a27 * , 7937 ( 1994 ) n. madras , c.e .
soteros , s.g .
whittington , j.l .
martin , m.f .
sykes , s. flesia and d.s .
gaunt , j. phys .
* a23 * , 5327 ( 1990 ) ; + i.s . chang and y. shapir ,
phys . rev .
* b38 * , 6736 ( 1988 ) b. derrida and h.j .
herrmann , j. physique * 44 * , 1365 ( 1983 ) ; + b. derrida and l. de seze , j. physique * 43 * , 475 ( 1982 ) m. henkel and f. seno , phys .
* e53 * , 3662 ( 1996 ) t.w .
burkhardt and j.m.j .
van leeuwen , `` progress and problems in real - space renormalization '' in _ real space renormalization _
burkhardt and j.m.j .
van leeuwen editors , springer - verlag ( 1982 ) , p. 1 ; + t.w .
burkhardt , `` bond - moving and variational methods in real - space renormalization '' , _ ibid .
d. johnston and p. plech ' a , j. phys .
* a30 * ( 1997 ) 7349 . c. baillie , d. johnston and j - p .
kownacki , nucl . phys .
* b432 * ( 1994 ) 551 ; + c. baillie , w. janke , d. johnston and p. plech ' a , nucl . phys .
* b450 * ( 1995 ) 730 ; + c. baillie and d. johnston , nucl .
b47 * ( proc
. suppl . )
( 1996 ) 649 ; + c. baillie , n. dorey , w. janke and d. johnston , phys .
lett * b369 * ( 1996 ) 123 .
d. johnston and p. plech ' a , j. phys .
* a31 * ( 1998 ) 475 .
b. bollobs , `` random graphs '' , academic press , 1985 .
f. peruggi , j. phys .
* a16 * ( 1983 ) l713 .
+ f. peruggi , f. di liberto and g. monroy , j. phys .
* a16 * ( 1983 ) 811 ; + f. peruggi , physica * 141a * ( 1987 ) 140 ; + f. peruggi , f. di liberto and g. monroy , physica * 141a * ( 1987 ) 151 ; + j. essam , j - c lin and p. taylor , phys . rev .
* e52 * ( 1995 ) 44 ; + f. peruggi , f. di liberto and g. monroy , z. phys . *
b66 * , ( 1987 ) 379 .
c. fortuin and p. kasteleyn , physica * 57 * ( 1972 ) 536 ; + p. kasteleyn and c. fortuin , j. phys .
jpn ( suppl . ) * 26 * ( 1969 ) 11 .
c. bachas , c. de calan and p. petropoulos , j. phys .
* a27 * ( 1994 ) 6121 .
v. kazakov , mod .
* a4 * ( 1989 ) 1691 . | lattice animals provide a discretized model for the @xmath0 transition displayed by branched polymers in solvent .
exact graph enumeration studies have given some indications that the phase diagram of such lattice animals may contain two collapsed phases as well as an extended phase .
this has not been confirmed by studies using other means .
we use the exact correspondence between the @xmath1 limit of an extended potts model and lattice animals to investigate the phase diagram of lattice animals on @xmath2 random graphs of arbitrary topology ( `` thin '' random graphs ) .
we find that only a two phase structure exists
there is no sign of a second collapsed phase .
the random graph model is solved in the thermodynamic limit by saddle point methods .
we observe that the ratio of these saddle point equations give precisely the fixed points of the recursion relations that appear in the solution of the model on the bethe lattice by henkel and seno .
this explains the equality of non - universal quantities such as the critical lines for the bethe lattice and random graph ensembles
. 22.5 cm 0.05 cm 0.2 cm 0.2 cm |
ion and photon beam irradiation are widely used to generate nanoscale surface features for a variety of applications : optoelectronic , plasmonic , magnetic , bio- and chemical sensing , and catalytic devices @xcite .
understanding interfacial transport mechanisms under fast thermal processing or energetic beam irradiation is essential to accurately predicting the spatio - temporal dynamics of film evolution that imparts the desired nanoscale surface features .
the contributions from various solid state transport processes that cause surface changes such as flattening or grooving in _ bulk materials _ were studied by mullins @xcite . in his formulation ,
smoothening of bulk surfaces is driven by curvature effects that manifest in solid state surface diffusion , volume diffusion , or evaporation - condensation mechanisms@xcite .
mullins theory has been widely accepted and used extensively in the analyses of surface processes @xcite . however , mullins theoretical framework is inadequate to describe solid state mass transport in _ ultrathin films _ in which one has to account for the effects of other driving forces on pattern formation , e.g. dispersion forces and epitaxial stress .
for instance , it is well - established that epitaxial stress has a significant effect on surface diffusion .
there is extensive literature on pattern formation in thin films in the presence of epitaxial stresses @xcite .
in contrast , the literature on the evolution of ultrathin films on amorphous substrates , a scenario in which epitaxial stress effects are negligible , is sparse . in this article
, we develop a generally applicable mathematical model for describing such processes and apply the model to study solid state transport in thin films exposed to time periodic laser thermal stimulus .
recently , we have observed that thermal processing of ultrathin metal films by multiple instances of a @xmath5 pulsed laser ( with each pulse separated in time by @xmath6 ) results in a variety of spatially ordered patterns @xcite .
one of the important findings , relating to the processing conditions , was that in order for pattern formation to occur , it was necessary to irradiate the films at a laser fluency at or above a critical value , which was well - correlated with the threshold energy required to melt the films @xcite . consequently , mass transport was considered to occur primarily in the liquid phase and pattern formation was attributed to a thin film hydrodynamic ( tfh ) dewetting instability , which arises in spinodally unstable films when attractive long - range intermolecular dispersion forces amplify surface perturbations @xcite .
thermal modeling of the film / substrate bilayer showed that during one laser pulse , the film experienced a nearly uniform , but transient temperature field for a time scale of @xmath7 , of which the time to reach the melt temperature during the heating phase was @xmath8 , the liquid lifetime , @xmath9 was typically between @xmath10 and cooling time was @xmath11 @xcite .
therefore , approximately @xmath12 of the thermal cycle following each laser pulse was spent in the cooling phase .
this is important in the context of mullins analysis , which suggests that solid state diffusion and evaporation - condensation processes could influence surface deformation .
hence , it is necessary to analyze both liquid and solid phase mass transport processes to develop a comprehensive understanding of the physics of interface evolution in thin films processed under pulsed laser irradiation . in this work ,
we have suitably modified mullins original formulation @xcite to describe the evolution of infinitesimally small surface perturbation of ultrathin solid films on amorphous substrates .
in particular , we have derived new results for volume diffusion transport and have employed the approach adopted by spencer et al .
@xcite originally for describing diffusion in epitaxially stressed films , in order to account for dispersion forces that are influential only in thin films .
for pulsed laser experiments described in @xcite , we determined that there is negligible contribution to surface deformation from solid state material redistribution by evaluating solid state mass transport rates via diffusion and evaporation - condensation processes .
these results suggest that film deformation occurring under each melting cycle in the liquid phase is essentially _ quenched - in _ during rapid resolidification .
subsequent laser pulses cause progressively more interface deformation until the film ruptures .
this mechanism is consistent with experimental observations @xcite .
local curvature differences provide the driving force for mass rearrangement through various mechanisms in the solid state including surface diffusion , volume diffusion , and evaporation - condensation . in a series of papers
, mullins outlined a theory to predict the mass transfer rates by the aforementioned processes for surface deformation of bulk materials @xcite . for thin films ,
additional driving forces need to be accounted for , such as the free energy contributions from dispersion ( van der waals ) forces and thermal strain . for films supported by amorphous substrates ,
epitaxial strain is irrelevant .
however , mismatch in thermal expansion coefficients of the film and the substrate can result in thermoelastic strain . in this work ,
we first develop the theory by incorporating dispersion forces , in the absence of thermal stresses .
the effect of thermal stresses is only discussed qualitatively in sec .
[ sec : thermal - stress ] . following spencer et al .
@xcite , the evolution of the film height _ h _ is described by : @xmath13 where @xmath14 , @xmath15 is the surface tension , @xmath16 is the local curvature , @xmath17 is the surface diffusion coefficient , @xmath18 is the atomic volume , @xmath19 is the number of atoms per unit area , @xmath20 is boltzmann s constant , @xmath21 is the absolute temperature , @xmath22 is a free energy per volume term , and @xmath23 is an additional velocity term such as a material deposition term @xcite . in the limit that the height of any surface fluctuations are small , the curvature can be approximated as @xmath24 .
it has already been established that dispersion forces play a key role in the evolution of a thin metal film @xcite , especially upon laser - induced melting @xcite . by including the effect of the disjoining pressure to the free energy ,
the height evolution equation becomes : @xmath25+v\label{eq : dh / dtsd}\ ] ] where @xmath26 is the hamaker coefficient and the free energy per volume contribution due to dispersion forces , @xmath27 is strictly valid for a flat film .
the effects of evaporation - condensation and volume diffusion processes can be included into the height evolution equation as additional velocity terms , in eq .
[ eq : davis ] . by relating local pressure differences between the flat and perturbed states to the local curvature
though the gibbs - thompson formula , mullins has shown that the evolution of a rough surface , due to evaporation - condensation is given by : @xmath28 where @xmath29)@xmath30 , @xmath31 is the equilibrium vapor pressure and @xmath32 is the mass of the metal atom @xcite .
as in the case of evaporation - condensation , volume diffusion can be considered and included in the height evolution equation as another @xmath23 term . in the derivation below , _ x _ and _ z _ denote the lateral and normal coordinates respectively such that _ z _ = 0 and _ h _ correspond to the film - substrate interface and local film height respectively .
the substrate is much thicker than the film and hence considered as a semi - infinite domain .
the decay rate due to volume diffusion can be related to the local density , @xmath33 as @xmath34 , where @xmath35 is the coefficient of self - diffusion and _ z _ denotes the coordinate normal to the substrate plane @xcite . for a bulk surface ( @xmath36 ) , assuming quasi steady - state diffusion @xmath37 and solving the laplace equation with boundary conditions @xmath38 at @xmath39 and @xmath40 at @xmath41 , where @xmath42 and @xmath2 are the amplitude and wavevector of the surface perturbation @xmath43 respectively , mullins derived @xmath44 , where @xmath45 .
however , for a film of finite thickness the boundary conditions need to be modified , specifically : @xmath38 at @xmath41 and @xmath40 at @xmath46 . assuming quasi - steady state diffusion and @xmath47 where @xmath48 the expression for volume diffusion becomes @xmath49 : @xmath50 where @xmath3 is the average film thickness .
note that the quasi - steady state ( qss ) assumption is only valid if @xmath51 , and even for sub - nanometer thick metal films , the time to reach qss , @xmath52 .
however , since @xmath53 ( the density gradient at the surface of the film ) will have its largest magnitude for short times and decrease to its steady - state value , the decay rate represented by a quasi steady - state approximation is a limiting case . for systems with a timescale @xmath54 the decay rate will be faster than that predicted above . for macroscopic systems @xmath55 , eq .
[ eq : dh / dtvol ] reduces to the mullins equation for volume diffusion , @xmath56 @xcite .
it is interesting to note that in the opposite limit , i.e. , for thin films with a relatively large perturbation wavelength ( @xmath57 ) and after linearizing the exponential about @xmath3 , the expression for decay due to volume diffusion simplifies to:@xmath58 this limit applicable for thin films is qualitatively different from that for a bulk material , _ with the rate of decay varying as @xmath59 _ _ as compared to @xmath60 for the latter .
further , the thickness dependence , i.e. , the decay rate varying as @xmath61 , is absent in the case of bulk materials . _
by combining eqs .
[ eq : dh / dtsd ] , [ eq : dh / dtevap ] and [ eq : dh / dtvol ] , the evolution equation for height perturbations @xmath43@xmath62 can be expressed as : @xmath63-b_{vol}k^{3}\frac{1+exp(-2kh_{o})}{1-exp(-2kh_{o})}h'+b_{ec}\nabla^{2}h'.\label{eq : dh / dtsolid}\ ] ] letting @xmath48 , the growth / decay rate , @xmath64 of surface perturbations in the solid state is given by : @xmath65-b_{vol}k^{3}\frac{1+exp(-2kh_{o})}{1-exp(-2kh_{o})}-b_{ec}k^{2}\label{eq : sigmasolidstate}\ ] ]
in this section we summarize the results for the growth / decay rate of thin liquid films with the purpose of comparing the mass transport rates in the liquid state to that in the solid state . for a detailed description of liquid state mass transport and dewetting in thin films , see @xcite .
when the navier - stokes equations , describing mass transport in thin liquid films , are simplified via the lubrication approximation , a film - thickness dependent relationship between the material flow rate and pressure gradient can be derived @xcite .
the pressure is a combination of the laplace pressure from capillarity and a disjoining pressure that arises from long - range dispersion or van der waals forces @xcite . when the disjoining pressure , which tends to amplify perturbations , exceeds the stabilizing effect of the surface tension , the film is in an unstable state and perturbations will grow , eventually causing the film to rupture . the thin film hydrodynamic equation describing this behavior is given by the following : @xmath66 where @xmath67 is the dynamic viscosity .
once again letting @xmath68 , the growth / decay rate @xmath69 in the liquid phase is given by:@xmath70 yielding the largest unstable wavevector , @xmath71 and a fastest growing perturbation wavelength , @xmath72 .
as mentioned previously ( sec . [ sec : hydrodynamic - flow - in ] ) , analytical models and numerical simulations have shown that for @xmath4 pulsed laser irradiation of thin metal films , the cooling time @xmath73 of the film is approximately @xmath74 times as large as the time that the film is in the molten state , i.e. , @xmath75 @xcite . at the beginning of the cooling phase ,
the film is at its melting temperature .
the high temperatures suggest that solid state material transport mechanisms could be active . to estimate the relative magnitude of the effect that solid state mass transport could have on deformation of the film during the cooling period , we employed eqs .
[ eq : sigmasolidstate ] and [ eq : sigmaliquid ] . for the values of @xmath17 and @xmath35
, we used the models of flynn for surface and volume diffusion in which @xmath76 and @xmath77 @xcite . in order to gain quantitative insight on the effect of mass transport in each phase ,
we have compared the growth / decay rate of a perturbation multiplied by the timescale during which it is applicable .
since diffusion and evaporation processes are disproportionally influenced by the time spent at the peak temperature , we have used an overestimate of the time at the peak temperature of @xmath78 in order to find an upper limit of the contribution from the solid state transport mechanisms . in fig .
[ fig : codispersionplots](a - b ) we have plotted the magnitude of @xmath79 and @xmath80 as a function of perturbation wavevector @xmath2 for material redistribution in both phases , where @xmath81 is the typical liquid lifetime . due to the thickness dependency of several of these mechanisms , we have compared transport mechanisms in a @xmath82 and a @xmath83 co film . in the case of a @xmath82 film ,
the stability of the solid film is strongly dependent on the choice of the hamaker coefficient @xmath26 , which dictates the magnitude of the disjoining pressure . when using the experimentally determined value of @xmath84 @xcite in eq .
[ eq : sigmasolidstate ] , the film is unstable and there exists a characteristic wavelength that will grow the fastest . the dispersion curve for growth / decay rates of perturbations in the solid state has been shifted to larger wavelengths , as compared to that for the tfh instability , as shown in figure [ fig : codispersionplots](a ) . in this case
the cut - off and the fastest growing wavelengths have been increased due to the stabilizing effects of volume diffusion and evaporation - condensation . in the absence of of these effects @xmath85 and @xmath86
are the same for both the solid and liquid states .
the importance of the magnitude of the dispersion term on mass transport can be seen if one uses the lower limit of the hamaker coefficient for metals @xmath87 @xcite .
when a theoretically estimated @xmath88 was employed , the disjoining pressure was not large enough to destabilize the solid film ; but rather , the film was seen to be stable against perturbations of all wavelengths . for a @xmath83 film ,
the dispersion forces with @xmath84 are not large enough to instigate a dewetting instability in the solid film , as shown if figure [ fig : codispersionplots](b ) .
for the length scales stemming from a tfh instability , @xmath89 , thereby confirming that hydrodynamic flow is the more dominant mass transport mechanism during @xmath4 pulsed laser melting .
we obtained similar results for a variety of metals suggesting that this is a general result , applicable to most thin metal films .
most metals investigated have similar diffusion rates , but the evaporation rates vary greatly . figs . [
fig : alandcrdispersionplots](a - b ) show plots of dispersion curves for two extreme scenarios , one metal ( al ) with a low evaporation rate and the other ( cr ) with a large rate . in both cases , films with @xmath90
are seen to be stable to small perturbation of all wavelengths .
there are several implications manifested in these results .
first , the timescales of solid state process are much larger than the timescales associated with hydrodynamic flow .
therefore , the _ quenched - in _ structures from the pulsed laser melting and resolidification are likely unchanged during the cooling period .
this conclusion is corroborated by experimental observations of pattern formation in ultrathin metal films by using multiple pulses of ( @xmath91 wavelength , @xmath5 pulsed ) ultraviolet ( uv ) laser irradiation : see @xcite for details .
a variety of metal thin films , including co , fe , cu , ag , ni and pt , were deposited under ultrahigh vacuum ( @xmath92 torr ) by e - beam evaporation or pulsed laser deposition onto commercially obtained sio@xmath93/si substrates with the thermally grown oxide layer of 400 nm thickness .
film thicknesses ranged from @xmath94 nm .
the initial films had an average surface roughness @xmath95 as evaluated by atomic force microscopy .
following deposition , the films were irradiated at various laser energies above a threshold energy density @xmath96 such that appreciable surface deformation was observed for various irradiation times , as measured by the number of pulses @xmath97 ranging from @xmath98 .
following such irradiation , the surface morphology and concentration of metal was evaluated by imaging and energy dispersive x - ray spectrometry in a scanning electron microscope , respectively .
it was found that all the metals possessed a thickness - dependent threshold energy density @xmath96 above which substantial surface changes were observed .
however , for energies below this threshold , no surface roughening was visible even after the longest irradiation times of 10,500 pulses .
this energy threshold has been attributed to the energy required to melt the films @xcite . by irradiating films with a laser energy density above the melt threshold , we performed detailed studies of the dynamics of the pattern evolution for co films as a function @xmath97 and found that the dewetting patterns showed characteristic film - thickness dependent length scales and spatial order consistent with the predictions of eq .
[ eq : sigmaliquid ] . hence , despite the much larger time spent in the solid phase ( i.e. cooling period ) , the overall contribution from solid state transport was inferred to be negligible .
since we focus on the dynamics of perturbed thin films on amorphous substrates , the prevalent stresses that are present in the film are thermally generated . due to the mismatch in the thermal expansion coefficients between the film and the substrate ,
the film is under compressive stress while it is heated and a tensile stress when it is cooling .
this stress is given by the following expression : @xmath99}\label{eq : thermalstress}\ ] ] where @xmath100 is the thermal expansion coefficient , @xmath19 is poisson s ratio , @xmath101 is young s modulus , @xmath102 is the thickness of the layer , @xmath103 is the temperature change , and the subscripts @xmath104 and @xmath105 denote the film and substrate quantities , respectively @xcite . in the limiting case where @xmath106 , the preceding expression reduces to : @xmath107 stresses related to thermal expansion for most thin metal films on a silica substrate are on the order of a few @xmath108 . however , they are still lower than the critical value required to initiate a buckling instability in the film , which is given by the following equation:@xmath109 where @xmath110 is a constant .
typical critical stress values range from @xmath111 .
stress can also result from an anisotropic temperature distribution in the film @xcite .
this could arise in pulsed laser processing .
as described above , the thickness variations caused by the tfh instability are quenched in during the rapid resolidification phase .
hence , in the solidified deformed film , a non - uniform temperature profile could prevail for times comparable to the diffusion time scale @xcite . within the context of linear theory of elasticity
, the thermoelastic stress will also decay at the same rate as the the temperature profile in the film @xcite .
hence a characteristic time scale of temperature / stress diffusion can be estimated by solving the transient heat equation for a 2-dimensional film:@xmath112 where @xmath113 , @xmath114 , and @xmath115 are the density , heat capacity and thermal conductivity of the film , respectively .
assuming an initial sinusoidal temperature profile in the lateral ( _ x _ ) direction and a uniform temperature distribution throughout the thickness ( _ z _ direction ) of the film , one finds the following relationship for the time dependent component of the temperature , @xmath116 , in which the temperature , @xmath117:@xmath118\label{eq : tempprofile}\ ] ] @xmath119 is the thermal diffusivity of the film , @xmath2 is the wavevector of the perturbation , @xmath120 is a constant between @xmath121 and @xmath122 and @xmath123 , @xmath124 , and @xmath125 are the density , heat capacity and thermal conductivity of the substrate , respectively .
since the length scale of the periodic thickness variations is being imprinted on the film in the liquid state , one can take the expected value of @xmath2 for a corresponding @xmath3 from tfh theory . using these values @xmath126 ,
the timescale during which this stress could influence solid state mass transport is on the order of a picosecond for a perturbation wavelength , @xmath127 and on the order of a nanosecond for @xmath128 .
these timescales are too short to appreciably effect any solid state process in modifying the surface of the film .
we have developed a mathematical model to describe thin film dynamics on amorphous surfaces by incorporating the effects of surface / volume diffusion , evaporation - condensation and dispersion forces .
new results are presented for volume diffusion in thin films .
specifically , for @xmath129 < < 1 , the decay rate of surface perturbations by this mechanisms is proportional to @xmath130 as compared to @xmath60 for bulk materials .
it is shown that depending on the magnitude of the hamaker coefficient , ultrathin solid films with thickness < 2 nm could be prone to pattern forming instabilities while films with thickness > 2 nm are predicted to be stable for a variety of metals .
the model was applied to analyze the role of solid state mass transport mechanisms on the dynamics of thin films processed by pulsed laser irradiation .
the upper limits for the time constants associated with the decay of surface perturbations in the solid state was found to be much smaller compared to the time interval between subsequent pulses .
further , a qualitative analysis of thermoelastic stress dissipation in the film - substrate bilayer system showed that the time scale of stress relaxation is much smaller than the pulse duration .
these findings suggest that contributions to the changes in surface deformation from solid state processes are minimal and the pattern formation occurs through a hydrodynamic instability that is followed by laser melting of the film , an observation corroborated by experimental observations .
rk and rs acknowledges support by the national science foundation through grant career dmi-0449258 and cts-0335348 , respectively .
cf acknowledges valuable contributions from hare krishna . | a mathematical model is developed to analyze the growth / decay rate of surface perturbations of an ultrathin metal film on an amorphous substrate ( @xmath0 ) .
the formulation combines the approach of mullins [ j. appl .
phys .
v30 , 77 , 1959 ] for bulk substrates , in which curvature - driven mass transport and surface deformation can occur by surface / volume diffusion and evaporation - condensation processes , with that of spencer et al .
[ phys .
rev .
lett .
v67 , 26 , 1991 ] to describe solid state transport in thin films under epitaxial strain .
modifications of mullins model to account for thin film boundary conditions result in qualitatively different dispersion relationships especially in the limit as _
@xmath1 _ where _ @xmath2 _ is the wavenumber of the perturbation and _ @xmath3 _ is the unperturbed film height .
the model is applied to study the relative rate of solid state mass transport as compared to that of liquid phase dewetting in a thin film subjected to fast a thermal pulse .
specifically , we have recently shown that multiple cycles of nanosecond ( @xmath4 ) pulsed laser melting and resolidification of ultrathin metal films on amorphous substrates can lead to the formation of various types of spatially ordered nanostructures [ phys .
rev .
b , v75 , 235439 ( 2007 ) ] .
the pattern formation has been attributed to the dewetting of the thin film by a hydrodynamic instability . in such experiments
the film is in the solid state during a substantial fraction of each thermal cycle .
however , results of a linear stability analysis based on the aforementioned model suggest that solid state mass transport has negligible effect on morphology changes of the surface .
further , a qualitative analysis of the effect of thermoelastic stress , induced by the rapid temperature changes in the film - substrate bilayer , suggests that stress relaxation does not appreciably contribute to surface deformation .
hence , surface deformation caused by liquid phase instabilities is rapidly quenched - in during the cooling phase .
this deformed state is further evolved by subsequent laser pulses .
these results have implications to developing accurate computer simulations of thin film dewetting by energetic beams aimed at the manufacturing of optically active nanoscale materials for applications including information processing , optical devices and solar energy harvesting . |
observations of sbs1542 + 541 using the international ultraviolet explorer ( porter _ et al .
_ unpublished ) show that there is significant continuum flux down to below @xmath19 , corresponding to @xmath20 in the rest frame .
this made the object a good candidate for possibly observing ly@xmath10@xmath21304 absorption in the intergalactic medium with the hopkins ultraviolet telescope ( hut ) during the astro-2 mission .
observations of hs1700 + 64 ( @xmath22 ) on the same mission successfully measured the opacity ( davidsen , kriss , & zheng 1996 ) .
however , during the observations of sbs1542 + 541 on 1995 march 16 we discovered that there was a cutoff in flux slightly redshifted from the expected position of the ly@xmath10 absorption , causing us to suspect that the observed feature was a lyman limit from a low - redshift intervening absorber . to further investigate this possibility we performed observations on 1995 september 3 with the g130h , g190h , and g270h gratings of the hst fos .
the presence of the intervening system is verified by a strong @xmath232796 , 2804 doublet in the g270h spectrum at the same redshift ( @xmath5 ) as the lyman edge visible in the hut spectrum .
fortuitously , the fos spectra also reveal broad absorption lines from several very highly - ionized species , many of which have never before been observed in a broad absorption line qso .
it is the study of this broad absorption system that is the focus of our present interest .
roughly 10% of radio - quiet qsos exhibit broad ( @xmath24 @xmath25 ) blueshifted absorption features associated with the qso , known as broad absorption lines ( bals ) , with outflow velocities generally @xmath26 ( turnshek 1984 ) .
these objects are collectively referred to as broad absorption line qsos , or balqsos .
the absorption features most commonly observed are ly@xmath10 @xmath211216 , @xmath211549 , @xmath211397 , and @xmath211240 , though low - ionization lines , usually @xmath212798 and @xmath211857 , are observed in some objects , usually called low - ionization balqsos .
higher - ionization lines are sometimes reported , up through @xmath21774 and @xmath21615 ( korista & arav 1997 ) .
when studying a bal system , one of the primary goals is to determine column densities of the various observed ions , which are used in turn to derive the ionization state of the gas and its chemical composition . to determine column densities
, it is necessary to measure the optical depth of the absorption as a function of outflow velocity . in deriving the optical depth from the relative depression of the continuum flux ,
one typically makes the assumptions that ( 1 ) there are no unresolved narrow lines in the bal troughs , ( 2 ) scattered radiation does not fill in the absorption troughs , and ( 3 ) the bal gas completely and homogeneously covers the continuum source on the observed line of sight .
if any of these assumptions are invalid , then the true column densities could be much greater than those inferred from the observations . based on high - resolution observations of bal troughs (
barlow & junkkarinen 1994 ) , assumption ( 1 ) appears to be a good one .
studies of bal geometry ( turnshek 1988 ; hamann , korista , & morris 1993 ) suggest that the global bal region covering factor , which is the fractional solid angle subtended by the bal gas as seen from the continuum source , is generally @xmath27 ( except possibly in low - ionization balqsos ; boroson & meyers 1992 ; turnshek _ et al . _
1997 ) . if the global bal covering factor is small , and especially if the outflow velocities are large , then ( 2 ) should be adequately satisfied , although spectropolarimetry suggests that some scattered light is present in the depths of the bal troughs ( goodrich & miller 1995 ; cohen _ et al . _ 1995 ) .
however , there is evidence that assumption ( 3 ) is invalid : in some instances absorption doublets appear to show saturation of the components but not at zero intensity ( korista _ et al . _ 1992
) , implying that the absorbing gas only partially covers the source .
more definite evidence for partial covering has been observed in the so - called associated or @xmath28 absorption lines ( hamann _ et al . _ 1997 ; hamann , barlow , & junkkarinen 1997 ) , which are distinguished from bals primarily by being much narrower , usually having widths @xmath29 a few hundred @xmath25 .
the fraction of the continuum source occulted by the bal region along the line of sight is called the covered fraction , or for clarity the _ continuum source covered fraction _ , to distinguish it from the global bal region covering factor . for the remainder of this paper
, we will use the term _ covered fraction _ to refer to the continuum source covered fraction .
we begin with a presentation of the observations used for this research ( [ sec : obs ] ) , followed by a discussion of the spectra and how we fit the bals in order to estimate minimum column densities and covered fractions of the absorbing ions ( [ sec : analysis ] ) .
we then discuss some properties of the system that can be inferred in the context of a photoionization model , including constraints on the ionization state and column density and the resulting implications for the geometry and physical structure of the gas ( [ sec : discuss ] ) . finally we summarize the characteristics of this particular balqso that make it so interesting and potentially relevant to the study of balqsos as a whole ( [ sec : summary ] ) .
our observations with the hst fos were performed on 1995 september 3 .
they include 14880 s with the g130h grating , 5470 s with g190h , and 1600 s with g270h .
these three spectra are shown in figure [ fig : uvspec ] .
the g130h data had to be corrected for a flickering diode not recognized by the fos pipeline at the time of processing .
this diode ( # 22 affecting wavelengths @xmath30 ) was added to the list of dead diodes and the spectrum was reprocessed .
these spectra have a resolution of 1 ( g130h ) , 1.44 ( g190h ) , and 2 ( g270h ) , corresponding to a range of velocity resolution of @xmath31 @xmath25 for the fos data . at optical wavelengths we use two observations .
the first was performed on 1995 april 1 with the mmt and covers @xmath32 with a resolution of 3.2 , corresponding to a velocity resolution @xmath33 @xmath25 . the other observation was performed on 1995 june 3 with the steward observatory 2.3 m telescope , including the wavelength range @xmath34 with a 5.54 resolution , corresponding to a velocity resolution of @xmath35 @xmath25 .
the mmt spectrum is of excellent quality in the wavelength region below @xmath36 for a ground - based observation , which is fortunate since this reveals well the broad absorption feature centered at @xmath37 .
the optical spectra are shown in figure [ fig : optspec ] .
together , these five spectra provide us with continuous wavelength coverage from @xmath2 to @xmath3 , or @xmath4 in the qso rest frame . for the purpose of identifying the @xmath5 intervening absorption system
, we also refer to the hut spectrum of sbs1542 + 541 obtained on 1995 march 16 during the astro-2 mission , shown in figure [ fig : hutspec ] . a normalized template created from blank - field spectra
was used to remove airglow .
this removes the broad wings of ly@xmath10 , but some narrow residuals remain for ly@xmath10 and a few other strong lines which were removed by hand .
the spectrum is of low s / n due to the short integration time ( 4320 s ) , so we do not use it for the analysis of the bal system .
we have also analyzed x - ray data from a @xmath38 exposure of sbs1542 + 541 obtained with the _ rosat _ position sensitive proportional counter ( pspc ) on 1993 august 14 .
the observed count rate was @xmath39 .
the s / n is too low to derive meaningful information about the spectral shape , so we use the data only to estimate the x - ray flux .
this is discussed further in [ sec : photoion ] .
we corrected the spectra for galactic extinction based on a column density of neutral hydrogen along the line of sight of @xmath40 @xmath12 ( stark _ et al . _
1992 ) . using a conversion of @xmath41 ( shull & van steenberg 1985 )
, this corresponds to an extinction @xmath42 .
we then performed the extinction correction using the formulation of cardelli , clayton , & mathis ( 1989 ) , assuming @xmath43 . in order to estimate the errors in the optical data , we fit small bands of apparently featureless continuum in each spectrum . near the @xmath231548 , 1551 feature in the steward data , we fit the region @xmath44 .
we find that the error per pixel is around @xmath45 @xmath46 , corresponding to a s / n per pixel of @xmath47 .
we used this value to produce a constant error array that was used in the fit of . in the mmt spectrum ,
we fit @xmath48 and found an error per pixel of @xmath49 @xmath46 , corresponding to a s / n per pixel of @xmath50 .
again we simply produced a constant error array that was used in the fit of ly@xmath10 , @xmath231239 , 1243 , and @xmath231032 , 1038 .
however , given that the observation includes short wavelengths so near the atmospheric limit , the s / n is lower at the shorter wavelengths .
hence , we realize that this is probably an overestimate of the errors in the region of ly@xmath10 and and an underestimate of the errors near , but since there are no spectral regions near these lines that can be well fit to estimate the errors , we will use this constant value .
many of the problems associated with studying bals are minimized in the spectrum of sbs1542 + 541 . due to the relatively small velocity width of the lines as well as the apparent lack of low - ionization species ,
most of the interesting lines are not blended and therefore the continuum is generally well - defined in the vicinity of the broad absorption features .
also , the line of sight to the object contains an unusual paucity of intervening lyman - limit absorption systems .
we identify only three at @xmath51 , @xmath52 , and @xmath5 . in the next section
we briefly discuss each narrow - line system .
a more comprehensive study of the intervening narrow - line systems ( _ e.g. _ vogel & reimers 1995 ) would require data of higher s / n
. the @xmath51 system is clearly established by definite lyman - series lines through ly@xmath53 as well as lines from ly@xmath54 that appear to be blended with other narrow lines , presumably from the ly@xmath10 forest .
in addition , there are strong lines corresponding to @xmath21977 and @xmath21903 along with possible lines from @xmath21788 and @xmath21833 .
this system overlaps the broad absorption in velocity space , precluding an accurate estimate of the column density from the lyman - edge opacity , since the relative contribution of each system to the flux depression around @xmath55 is not known .
however , given that we find in [ sec : cover ] that most of the neutral hydrogen from the bal system covers only @xmath5620% of the continuum source , we can conclude that at least some of this flux depression must be due to the lyman limit of the @xmath51 system , since the full depression is @xmath5640% of the total continuum flux .
this enables us to estimate that the total column of neutral hydrogen in this system probably lies in the range @xmath57
. the system could be intrinsic rather than intervening ; however , given that the observed metal lines suggest a lower ionization level than the broad lines , we assume in our analysis that it is physically distinct from the highly - ionized bal system .
the system at @xmath52 is identified mainly on the basis of the very strong lines of ly@xmath10 ( @xmath58 ) and ly@xmath59 ( @xmath60 ) that have unusually broad profiles , the ly@xmath10 line having a full width at zero absorption ( fwza ) @xmath61900
@xmath25 and ly@xmath59 having a fwza@xmath61750 @xmath25 .
a weak ly@xmath8 feature can be seen , as well as a possible feature from @xmath21977 blended with the broad absorption of @xmath21521 .
it is difficult to confirm the presence of any lyman - limit feature , which should lie at @xmath62 , due to the uncertain continuum shape in this region .
there does appear to be some small decrease in the flux , but the situation is complicated by the fact that qsos can only rarely be observed so far into the rest - frame uv , and consequently knowledge of the broad emission line properties in this region is extremely limited .
we estimate @xmath63 .
the redshift of the @xmath5 system is calculated from the strong @xmath232796 , 2804 doublet that occurs around @xmath64 .
absorption from ly@xmath10 and possibly @xmath231548 , 1551 are also detected .
none of the higher - order lyman lines can be seen since they lie below the wavelength coverage of our data . however , the system is firmly established by the doublet and by the presence of the strong lyman - limit feature at @xmath65 visible in the hut spectrum of this object ( figure [ fig : hutspec ] ) .
we have identified broad absorption features corresponding to resonance lines of , , , , through , , , and .
each feature is primarily characterized by a single trough that is blue - shifted at line center by @xmath66 @xmath25 with respect to the rest frame and has a width of @xmath67 @xmath25 .
the troughs are fairly flat across the bottom with no obvious consistent structure and have sharp velocity cutoffs on both the red and blue end . because of the high ionization levels and small widths of the lines ,
many of the doublets are resolved in velocity space , including @xmath23770 , 780 , @xmath23610 , 625 , and @xmath23499 , 521 . the lyman - series lines through ly@xmath53 are also resolved , with the exception of ly@xmath59 which is strongly blended with @xmath231032 , 1038 .
there are several other possible lines that are either part of a strong blend or occur in portions of the spectrum where the continuum level is not well - defined and hence the existence of the line is difficult to determine : 1 .
absorption from @xmath21788 could be present around @xmath68 , but the continuum shape is uncertain due to the broad emission line and a blend of narrow absorption lines .
there appears to be some slight evidence for an @xmath21630 feature around @xmath69 , but the absorption is either weak or the covered fraction is small ( @xmath27 ) .
also , this lies in between the apparent broad emission line and the @xmath21625 bal , again making the continuum shape uncertain .
3 . there is a dip in the flux shortward of an apparent broad emission line around @xmath70 , possibly @xmath21559 + @xmath21568 .
the dip extends from @xmath71 to @xmath72 and could be a blend of several possible bals including @xmath23553 , 554 , @xmath21568 , @xmath21559 , and @xmath23550 , 568 . for the reasons stated above and the uncertainties in the profiles , we make no attempt to fit or deblend any of the above lines .
we did measure the equivalent width of the broad feature around @xmath73 since this feature appears to be well - defined .
this is likely a blend of @xmath21433 , @xmath21430 , and possibly @xmath21429 .
table [ ta : atomicdata ] lists atomic data for all of the resonance lines relevant to this paper .
lccrl & 0.0 & 13.6 & 1215.67 & 0.416 & & & 1025.72 & 0.0790 & & & 972.54 & 0.0290 & & & 949.74 & 0.0139 & & & 937.80 & 0.00780 & 47.9 & 64.5 & 1548.20 & 0.190 & & & 1550.77 & 0.0952 & 77.5 & 97.9 & 1238.82 & 0.157 & & & 1242.80 & 0.0780 & 77.4 & 113.9 & 629.73 & 0.515 & 113.9 & 138.1 & 1031.93 & 0.133 & & & 1037.62 & 0.0661 & 97.1 & 126.2 & 480.41 & 0.110 & & & 568.42 & 0.0928 & 126.2 & 157.9 & 399.82 & 0.0844 & & & 401.14 & 0.168 & & & 433.18 & 0.0505 & & & 558.59 & 0.0907 & 157.9 & 207.3 & 465.22 & 0.389 & 207.3 & 239.1 & 770.41 & 0.103 & & & 780.32 & 0.0505 & 186.5 & 225.0 & 429.13 & 0.0801 & 225.0 & 266.0 & 430.47 & 0.0763 & 328.2 & 367.5 & 609.79 & 0.0842 & & & 624.95 & 0.0410 & 399.4 & 442.1 & 550.03 & 0.0773 & & & 568.15 & 0.0375 & 33.5 & 45.1 & 1393.76 & 0.524 & & & 1402.77 & 0.260 & 476.1 & 523.5 & 499.40 & 0.0719 & & & 520.67 & 0.0345 & 91.0 & 124.3 & 585.75 & 1.24 although the absorption system of sbs1542 + 541 qualifies it as a balqso according to the canonical definition of the class ( weymann , carswell , & smith 1981 ) , it certainly has characteristics that distinguish it from `` classical '' balqsos .
the line widths of @xmath67 @xmath25 are relatively small absorption over a range of velocities exceeding @xmath74 @xmath25 is common in balqsos .
in addition , the features from , , and ly@xmath10 are quite weak relative to `` classical '' systems , and absorption from @xmath231394 , 1403 , a feature that is virtually omnipresent in bal systems , is absent completely .
thus objects such as sbs1542 + 541 would be excluded from typical searches for balqsos which look for these features to be strong .
we also find no evidence for absorption by other commonly observed low - ionization species such as , , , and .
absorption from @xmath231118 , 1128 , sometimes reported in broad absorption , is also not observed .
there is a feature with the right profile shape and at the right wavelength to be @xmath211123 , but the feature is weak and the existence of would require a state of ionization much lower than that inferred from the other lines ( [ sec : photoion ] ) , so we dismiss the feature as a probable cluster of ly@xmath10-forest lines .
the absence of low - ionization species , particularly , is useful for constraining the ionization state of the gas ( [ sec : discuss ] ) . in figure
[ fig : speclabel ] we show for reference a combined spectrum of the fos and optical data , corrected for galactic extinction and rebinned for higher s / n .
all of the identified broad absorption features are labeled , as are the prominent broad emission lines and several intervening absorption features . for fitting purposes we attempted to develop optical depth templates from various observed absorption profiles such as @xmath231548 , 1551 and @xmath23770 , 780 ( see korista
_ et al . _
however , we found that no single derived optical depth template was able to fit all of the observed lines satisfactorily , due mainly to slight variations in the widths of the profiles . instead of a rigid template ,
it is necessary to use a more flexible absorption profile .
we use two such profiles , one a simple gaussian and the other a profile that we will refer to as the `` flat profile '' , an absorption profile that we synthesized in order to match the primary characteristics of the observed troughs , specifically the relatively flat bottoms and sharp velocity cutoffs .
the flat profile is a feature of constant optical depth smoothed at each edge by a semi - gaussian with fwhm fixed at the instrumental resolution .
each of these profiles provides a simple way to characterize the absorption features by three free parameters : the velocity centroid , the fwhm , and the equivalent width .
table [ ta : flatfitparams ] shows the resulting fit parameters for the flat template , and table [ ta : gaussfitparams ] shows the parameters for the gaussian fits .
all of the fits were performed with the iraf task _ specfit _ ( kriss 1994 ) .
we fit a separate profile to each doublet component , even when the components are strongly blended as is the case for and .
we constrained the velocity centroids and widths of the individual components of the multiplet to be the same during the fits .
whenever two lines are blended , the equivalent width of the entire blend is given .
the quoted errors on all quantities are purely @xmath75 confidence limits derived from the error matrix of the fit .
lcccc & 1215.67 & @xmath76 & @xmath77 & @xmath78 & 972.54 & @xmath79 & @xmath80 & @xmath81 & 949.74 & & & @xmath82 & 937.80 & & & @xmath83 & 1548.20 , 1550.77 & @xmath84 & @xmath85 & @xmath86 & 1238.82 , 1242.80 & @xmath87 & @xmath88 & @xmath89 & 1031.93 , 1037.62 & @xmath90 & @xmath91 & @xmath92 & 480.41 & @xmath93 & @xmath94 & @xmath95 & 399.82 , 401.14 & @xmath96 & @xmath97 & @xmath98 & 465.22 & @xmath99 & @xmath100 & @xmath101 & 770.41 & @xmath102 & @xmath103 & @xmath104 & 780.32 & & & @xmath105 & 609.79 & @xmath106 & @xmath107 & @xmath108 & 624.95 & & & @xmath109 & 499.40 & @xmath110 & @xmath111 & @xmath112 & 520.67 & & & @xmath113 & 585.75 & @xmath114 & @xmath115 & @xmath116 + & 433.18 , 430.47 & & & @xmath117 lcccc & 1215.67 & @xmath118 & @xmath119 & @xmath120 & 972.54 & @xmath121 & @xmath122 & @xmath123 & 949.74 & & & @xmath124 & 937.80 & & & @xmath125 & 1548.20 , 1550.77 & @xmath126 & @xmath127 & @xmath128 & 1238.82 , 1242.80 & @xmath129 & @xmath130 & @xmath131 & 1031.93 , 1037.62 & @xmath132 & @xmath133 & @xmath134 & 480.41 & @xmath135 & @xmath136 & @xmath137 & 399.82 , 401.14 & & & & 465.22 & @xmath138 & @xmath139 & @xmath140 & 770.41 & @xmath141 & @xmath142 & @xmath143 & 780.32 & & & @xmath144 & 609.79 & @xmath145 & @xmath146 & @xmath147 & 624.95 & & & @xmath148 & 499.40 & @xmath149 & @xmath150 & @xmath151 & 520.67 & & & @xmath152 & 585.75 & @xmath153 & @xmath154 & @xmath155 + & 433.18 , 430.47 & & & @xmath156 only the most obvious narrow lines ( fwhm@xmath56instrumental resolution ,
significance@xmath157 ) were deblended from the broad lines , with all other features in the wavelength range of the bals being treated as part of the bal profiles .
less - significant narrow lines were sometimes fit outside the bal profiles to improve the continuum fit . the presence of other narrow lines in the bal profiles could cause additional uncertainty in the fit parameters .
this is particularly true for ly@xmath10 , , , and ly@xmath8 through ly@xmath53 , which lie in the portion of the spectrum where the ly@xmath10 forest is most dense .
the fits for and to a lesser extent show some deviation from the observed profiles ( see figure [ fig : profile ] ) indicating that these also are probably somewhat corrupted by narrow lines .
the possible effects of blending on the derived column densities and covered fractions are discussed in [ sec : colden ] and [ sec : cover ] , respectively . for the fit of ,
the ly@xmath59 bal corresponding to the best fit of the higher - order lyman lines , including a partial covered fraction as discussed in [ sec : cover ] , was also deblended from the absorption profile .
if there is additional ly@xmath59 corresponding to the stronger ly@xmath10 line , we could still be overestimating the strength of .
both profiles are capable of producing reasonable fits to the data , which is why we fit every feature with each profile . because we created the flat profile specifically to fit the shapes of the observed absorption features
, it achieves more satisfying fits than the gaussian by eye , but only slightly better in @xmath158 because of the generally low s / n .
the tails of the gaussians extend out slightly beyond where there is any apparent absorption , causing the equivalent widths of the gaussian fits to be generally somewhat larger than those of the flat profile . because the flat profile provides better fits
, we believe that it parameterizes better the basic characteristics of the lines .
however , using both profiles helps to characterize the uncertainties of our results associated with the uncertainty of the profile shape . as we discuss in [ sec : cover ] , we believe that the covered fractions of the different absorbing ions are significantly less than unity , and that the absorption may in many cases be optically thick . without knowing the covered fraction , it is not possible to determine accurately the optical depth of the lines , and hence to determine the column densities .
we can , however , put a good lower limit on the absorbing columns by calculating the column densities under the assumption of a covered fraction of unity . rather than estimate the columns from the linear portion of the curve of growth , we derive values from direct integration of the optical depth of the best - fit absorption profiles .
we produce the best fit possible for the lines from each ion assuming a covered fraction of unity , with the constraint that the ratio of the optical depths of the individual components is fixed by the oscillator strengths . to facilitate this , for
the gaussian fits we use a gaussian defined in terms of optical depth rather than flux .
the column densities with statistical errors as calculated from these fits are shown in table [ ta : colden ] . since none of the lines are optically thick under the assumption of total covering
, the uncertainty in the measured column densities due to the possible existence of narrow - line blends in the bals should be relatively small .
lcc & @xmath159 & @xmath160 & @xmath161 & @xmath162 & @xmath163 & @xmath164 & @xmath165 & @xmath166 & @xmath167 & @xmath168 & @xmath169 & & @xmath170 & @xmath171 & @xmath172 & @xmath173 & @xmath174 & @xmath175 & @xmath176 & @xmath177 & @xmath178 & @xmath179 some of the ions , particularly , , and , exhibit multiple lines with similar apparent optical depths despite different oscillator strengths , suggesting that the lines are likely saturated and therefore that the continuum source is not fully occulted by the absorber . as a result ,
the fits to these features under the assumption of a covered fraction of unity are poor , particularly for the higher - order lyman lines , for which no reasonable fit could be achieved . also , none of the bal troughs depress the flux more than @xmath180% , consistent with partial covering .
in the next section we discuss the necessity of including partial covered fractions for fitting and interpreting the bal system . in principle , given a fully - resolved absorption multiplet , one can uniquely determine both the optical depth and covered fraction as a function of velocity ( hamann _ et al . _
however , given that the quality of the data does not permit such a detailed analysis , we proceed under the assumption that the covered fraction is a constant function of velocity , but may vary from ion to ion . allowing the covered fraction to vary ,
we re - fit the doublets of , , and .
figure [ fig : profile ] shows the absorption profiles and best - fit flat profiles of these ions with a partial covered fraction after the removal of the narrow lines and the known ly@xmath59 in .
the fits for the strongly - blended doublets of the other identified lithium - like ions and are also shown for completeness , assuming a covered fraction of unity .
the best fit of the higher - order lyman lines , showing only the contributions from the continuum and broad absorption features , is plotted in figure [ fig : lyman ] .
we not only fit the resolved features ly@xmath8 through ly@xmath53 but also included lines through ly29 to fit the spectrum down to the bal lyman limit . around @xmath181
the continuum flux begins to be depressed significantly beyond the 20% level , presumably due to the high - order lyman - line and lyman - limit absorption from the @xmath51 narrow - line system .
the resulting best - fit values for the covered fractions are listed in table [ ta : cover ] , along with minimum covered fractions for the singlet features derived from the penetration depths of the troughs . as is evident from comparing the values derived from the different absorption profiles , the uncertainty in constraining the covered fractions due to the choice of absorption profile is larger in most cases than that derived from the statistics of the fits . in figure
[ fig : cover ] we plot the covered fractions from the flat profile fits against the creation ionization potential for each ion .
the closed circles are best - fit values derived from multiple components , while the open circles represent lower limits .
it is possible that many or most of the singlet absorption features are saturated .
if this is the case , then the open circles in figure [ fig : cover ] closely represent the true covered fractions and not just lower limits .
lcc & @xmath182 & @xmath183 & @xmath184 & @xmath185 & @xmath186 & @xmath187 & @xmath188 & @xmath189 & @xmath190 & @xmath191 & @xmath192 & @xmath193 & @xmath194 & & @xmath195 & @xmath196 & @xmath197 & @xmath198 & @xmath199 & @xmath200 & @xmath201 & @xmath202 & @xmath203 & @xmath204 blending of the observed bals with weak broad lines or with narrow lines could be important in the determination of the covered fractions because a fairly small change in the profile of one of the components of a doublet could cause a significant change in the inferred covered fraction . as mentioned previously
, it appears that such contamination does exist in the and doublets , and it is likely that this is true to some extent for the and features as well .
hence the true uncertainties in the covered fractions for these ions are probably somewhat larger than the listed statistical errors .
this is not so much the case for , for which the covered fraction is determined from several lines and is therefore more reliable . despite this complication ,
it is clear that the covered fraction for the dominant absorbing source is not the same for all ions . the most convincing evidence for this is provided by the higher - order lyman lines , for which we can reliably constrain the covered fraction to be less than 0.2 .
this covered fraction is less than that of most of the other lines , including the doublets of , , and , as well as the observed singlet features , all of which depress the flux more that 20%
. therefore there must be multiple absorbing regions contributing to the bal system , providing direct evidence for inhomogeneity in the ionization structure of the bal gas .
there appears to be a general trend for the very high - ionization species ( creation ip @xmath205 ) to occult more of the continuum source than the lower - ionization species ( creation ip @xmath206 ) .
since we can constrain well the covered fraction of , it is useful to estimate a better lower limit on the column density of , as this provides a very useful constraint for our photoionization models ( [ sec : discuss ] ) . for the fit of the lyman lines with a partial covered fraction
, a large column density of , @xmath207 , is required to obtain a good fit . however , as mentioned previously , there is probably some narrow - line blending , which could cause us to slightly overestimate the strength of the higher - order lyman lines . to get a more conservative lower limit on the column density of
, we allowed for the contribution of additional narrow lines below the @xmath208 significance level that appear to be blended with the broad lines . with these lines included , we find that with 90% confidence a column density of @xmath209 is necessary to provide a good fit using a flat profile , and @xmath210 using a gaussian profile .
an important point is that we can make these estimates because we can clearly see the higher - order lyman lines usually not visible in bal spectra .
the fact that the ly@xmath10 feature indicates a larger covered fraction than the higher - order lyman lines suggests that there are at least two different regions causing absorption in a region with a small covered fraction but a high column density of that contributes the absorption for the higher - order lyman lines , and a region with a lower column density of but significantly higher covered fraction that dominates the observed ly@xmath10 absorption profile . as a result , if we were to estimate the total column of in this bal system from ly@xmath10 alone and assume a covered fraction of unity , we would underestimate the true value by orders of magnitude .
in order to produce self - consistent photoionization models we construct an ionizing continuum based on the observed continuum . however , to accurately represent the euv continuum , particularly for a high - redshift qso , we need to first correct for the accumulated lyman - line and lyman - continuum absorption from intervening clouds , the primary characteristic of which is a broad trough centered around @xmath211 known as the lyman valley ( mller & jakobsen 1990 ) .
we use the empirical formula for the statistics of the forest - line distribution @xmath212 ( press & rybicki 1993 ) . for the parameter values
we adopt @xmath213 , @xmath214 , and @xmath215 , since these yield a line number in agreement with observations of high - redshift qsos ( hu _ et al . _
1995 ) . using this distribution we applied the statistical correction for @xmath1 , including intervening absorbers with a total column density between @xmath216 @xmath12 and @xmath217 @xmath12 .
any absorption system with a column density greater than @xmath218 @xmath12 produces a significant lyman limit feature for which we can correct directly , which we did for the @xmath52 and @xmath51 systems .
the resulting corrected spectrum is plotted in figure [ fig : corspec ] , shifted into the qso rest frame . also shown for comparison , with arbitrary normalization , is the composite spectrum of radio - quiet qsos produced by zheng _
et al . _ ( 1997 ) .
the lines below the spectrum indicate the wavelength regions used to fit the continuum of sbs1542 + 541 .
the spectrum shows a break in the power - law index @xmath10 ( @xmath219 ) around @xmath220 .
such a break is consistent with zheng _
et al . _ ( 1997 ) , who find that a break in the uv power - law index around @xmath221 is a general feature of qso continua . from the steward spectrum
we derive a near - uv power - law index of @xmath222 by fitting the region @xmath223 , longward of the break .
shortward of the break , we use the fos data to derive an extreme - uv power - law index of @xmath224 by fitting the continuum range @xmath225 . to estimate the soft x - ray flux from the _ rosat _ pspc data
, we use the total count rate of @xmath39 in the observed energy band @xmath226 ( @xmath227 in the rest frame ) and the galactic neutral hydrogen column density of @xmath228 . assuming a power - law index of @xmath229 , a typical value for the soft x - ray ( @xmath230 ) continuum of radio - quiet qsos as derived from pspc observations ( schartel _ et al .
_ 1995 ; laor _ et al . _
1997 ) , the observed flux at 1 kev is @xmath231 @xmath232 , which corresponds to a flux of @xmath233 @xmath232 at 1 kev in the qso rest frame .
this is an order of magnitude less flux than expected based on an extrapolation of the extreme - uv continuum .
we illustrate this in figure [ fig : broadband ] by plotting the extrapolated extreme - uv continuum and the derived x - ray continuum .
we also show the continuum derived from the _ rosat _ data assuming @xmath234 , a more typical value for harder x - rays ( @xmath235 ; comastri _ et al . _
1992 ; williams _ et al . _
1992 ) . the relative luminosity between the optical and the x - ray is commonly represented by the parameter @xmath236 , where @xmath237 is the spectral flux at 2500 and @xmath238 is the flux at 2 kev .
an extrapolation of the near - uv continuum yields a rest - frame flux at 2500 of @xmath239 @xmath232 , and from the x - ray data we infer a rest - frame flux at 2 kev of @xmath240 @xmath232 , yielding @xmath241 .
such a high value for @xmath242 is consistent with the results of green & mathur ( 1996 ) , who find values of @xmath243 in a sample of balqsos observed with the _ rosat _ pspc .
the authors attribute the lack of x - ray flux to absorption by the bal gas and derive bal column densities @xmath244 . it is therefore reasonable to suspect that the x - ray continuum of sbs1542 + 541 may be absorbed by the bal gas .
hence , to choose an intrinsic ionizing continuum for our photoionization models , we defer to the uv observations and use the extrapolated power law with index @xmath245 shown in figure [ fig : broadband ] .
the possibility of x - ray absorption is further discussed in
[ sec : xray ] . in
[ sec : altcont ] we discuss the effect of choosing an alternative continuum shape .
all of our calculations were performed with the photoionization code cloudy ( version 90.01 ; ferland 1996 ) .
we assume that the bal gas is in photoionization equilibrium with the intrinsic qso continuum and has solar chemical abundances .
we use a constant density of @xmath246 , although it has been previously established ( _ e.g. _ hamann _ et al . _ 1995 ) in similar models that the ionization state of the gas is not sensitive to the absolute density ( up to a limit of @xmath247 ) .
hence our results are valid for any reasonable space density , so long as it is constant .
the continuum normalization is determined by the ionization parameter @xmath248 , defined as the ratio of the number density of ionizing photons ( energy above the ionization threshold of hydrogen ) to hydrogen nuclei , @xmath249 where @xmath250 is the frequency at the hydrogen ionization threshold .
we do not assume that the gas is optically thin to the ionizing continuum .
in fact , given the necessarily large total column densities that we infer , continuum opacity must be an important factor and will alter the spectral energy distribution of the ionizing continuum as it propagates through the bal gas , resulting in changes in the ionization state .
the simplest model of the bal gas is a single slab characterized by its total column density @xmath251 and the ionization parameter @xmath248 at the face of the slab .
we use our measured limits on the ionic column densities to constrain the total column density and ionization parameter for such a model .
we illustrate this in figure [ fig : nuplaneone ] by plotting in the @xmath252 parameter space contours of constant ionic column density equal to a few of our lower limits , including those which provide the most restrictive constraints on the parameters .
the contour for corresponds to a column density of @xmath253 @xmath12 .
models consistent with our lower limits on the column densities must lie above the solid lines .
we also include a constraint from the lack of absorption . by fitting the data assuming a flat absorption profile with a centroid at @xmath254 @xmath25 , a width of @xmath255 @xmath25 , and a covered fraction of @xmath9
, we derive an upper limit on the column density of of @xmath256 @xmath12 with 90% confidence .
this upper limit is represented by the dashed line in figure [ fig : nuplaneone ] .
models consistent with the upper limit must lie below the dashed line to avoid producing an observable amount of .
thus the shading indicates the region of parameter space consistent with our column density limits in this simple model .
the existence of a region in the @xmath252 parameter space that satisfies all of our column density constraints means that , based on the column density limits alone , we could explain the observations by a model of a single , marginally optically thick , bal region of constant density .
in such a model the large range of ionization states is possible because the very high - ionization states would exist near the face of the slab , whereas the lower - ionization states exist deeper in the slab ( _ i.e. _ farther from the continuum source ) as a result of x - ray self - shielding .
however , to understand the bal system we would like to be able to explain not only the ionization state of the gas but also the covered fractions , and in particular the correlation of a higher covered fraction with a higher state of ionization . this would be impossible to achieve in a single - zone model .
we now discuss how we can modify this simple single - slab picture to arrive at a more plausible explanation of the covered fractions .
multiple zones are necessary to explain the range in covered fractions that we observe .
empirically , what is required is that the high - ionization gas have a larger covered fraction than the lower - ionization gas .
here we divide the bal region into two zones , one with a low covered fraction contributing the absorption from the higher - order lyman lines and perhaps also and , and another with a higher covered fraction providing the dominant absorption source of the higher - ionization species , up through . we will refer to these as the `` smaller '' and `` larger '' zones , respectively .
figure [ fig : nuplanetwo ] shows the constraints for each zone provided by our observed column density limits as in the single - slab case . for the larger zone we apply an additional constraint provided by the fact that absorption clearly can not be optically thick in this zone . in the same manner as described in the previous section for , but here assuming a covered fraction of @xmath257 , we place an upper limit of @xmath258 @xmath12 on the column density of . placing upper limits on the column densities of and results in constraints that are very similar but slightly less - restrictive , so they are not shown .
further plausible physical assumptions can place more restrictions on the possible range of parameter space covered by the two zones .
for example , we could assume that the clouds that make up the bal gas are equidistant from the central radiation source , but have small , dense cores surrounded by lower - density halos .
the cores would naturally have a lower ionization parameter , and they would cover a smaller fraction of the source .
more generally , one could envision a density - fluctuated medium containing small , dense clouds of radius @xmath259 , and larger , lower - density clouds of radius @xmath260 satisfying the constraint @xmath261 , where we use the subscripts @xmath262 and @xmath263 to denote the smaller and larger zones , respectively .
we can relate the cloud radii to their column densities as @xmath264 . since @xmath265
, we have @xmath266 . assuming the two zones are at roughly the same distance @xmath267 from the continuum source
, we can therefore write the size constraint as @xmath268 .
this reduces the range of acceptable two - slab models in the four - dimensional parameter space spanned by @xmath269 , @xmath270 , @xmath271 , and @xmath272 . by considering the effect of this constraint on the solutions shown in figure [ fig : nuplanetwo ] ,
we conclude that , in this density fluctuation model , the ionization parameter in the smaller zone must be a factor @xmath18 lower than in the larger zone , which would correspond to a larger density in the smaller zone by the same factor .
density fluctuations in the bal gas and the resulting correlation between covered fraction and ionization state could have implications for the determination of chemical abundances in bal systems .
studies of bal systems typically find that a large range in ionization parameters ( more than an order of magnitude ) and large abundance modifications relative to solar ( as much as an order of magnitude , or even two for nitrogen ) are necessary to explain the observed column densities ( _ e.g. _ turnshek _ et al . _ 1996 ) .
the large metallicity enhancements are required primarily because of the observed weakness of the ly@xmath10 bal , which is used to derive the column density .
however , in the bal system of sbs1542 + 541 , the small equivalent width of ly@xmath10 despite a large optical depth ( @xmath273 ) is explained by the fact that most of the covers only a small fraction of the continuum source . as a result
, the contribution from most of the is `` hidden '' beneath the absorption from more highly - ionized gas with a much lower column density but higher covered fraction .
this situation is revealed by the direct observation of the saturated lyman lines of order ly@xmath8 and higher , lines that in a more typical bal system would not be visible due to blending with other lines , particularly @xmath21977 . if inhomogeneity of the bal gas is a generic feature of bal systems , particularly that which would result in large columns of hidden , then this could affect ( and make more difficult ) the measurement and interpretation of bal column densities and chemical abundances . in this model , the source of the absorption is constrained to the small region of parameter space shown in figure [ fig : nuplanetwo ] by the tight upper limit on absorption . however , it is possible that the absorption arises in a very weakly - ionized zone ( @xmath274 ) with a low enough column density ( @xmath275 @xmath12 ) that no strong low - ionization metal absorption lines are produced .
if such a zone exists , then the and lines require an additional higher - ionization zone .
this arrangement is illustrated in figure [ fig : nuplanealt ] , as calculated using continuum shape ( b ) from figure [ fig : broadband ] ( see [ sec : altcont ] ) .
alternatively , they could be produced in the larger zone if the lines are assumed to be not optically thick .
this is unlikely , however , since the profiles , particularly , strongly suggest saturation .
although a two - zone model is the simplest one that can explain the basic properties of the system , including the observed line strengths and the general trend of the covered fractions , there are complications . in particular , the observed range in covered fractions for the highly - ionized species , especially the apparently high covered fraction of , can not be explained by a simple model .
if this is a real effect , then the physical structure of the bal gas could be extremely complex , and much more detailed modeling would be necessary to reproduce this peculiar behavior .
however , it is entirely possible that this effect is simply due to the uncertainties induced by blending and the choice of fitting profile .
the correlation of covered fraction with ionization state can also be explained by more complex physical models .
it is plausible that the bal region has an overall disk - like geometry as suggested by the high continuum and even higher trough polarization observed in balqsos ( goodrich & miller 1995 ; cohen _ et al . _ 1995
this has been explored theoretically in the disk - wind models of agn broad line regions ( murray _ et al . _ 1995 ; murray & chiang 1995 ) .
one can imagine that we are looking along a particular line of sight , just above ( or below ) the disk , as illustrated schematically in figure [ fig : disk ] .
the states of lower ionization , being farther from the continuum source , would naturally have a smaller covered fraction .
an appealing consequence of this interpretation is that the bal region that we observe in sbs1542 + 541 would not be required to be inherently exceptional in its lack of absorption from states of lower ionization such as , , and .
these ions could be part of the bal region but be far enough from the continuum source that they lie outside our line of sight .
the existence of high column densities of highly - ionized gas suggests that the uv absorption system of sbs1542 + 541 is associated with a `` warm absorber '' as seen in the x - ray spectra of many low - redshift agn .
models of x - ray warm absorbers ( netzer 1993 ; krolik & kriss 1995 ; netzer 1997 ) predict that the strongest features should be the k ionization edges of and as well as k and l edges of ionization stages of iron @xmath56 . according to our photoionization calculations ,
given the likely range of ionization parameters for the bal gas in sbs1542 + 541 , the dominant stages of oxygen should be and .
the predicted large column densities of these ions should produce observable ionization edges at @xmath276 kev ( ) and @xmath277 kev ( ) , each with an optical depth of @xmath278 .
observations of these x - ray features would enable us to form a much more complete picture of the ionization state of the system , making this object a good candidate for further study of the unification of x - ray / uv absorbers ( mathur _ et al . _ 1994
; mathur , elvis , & wilkes 1995 ; mathur , wilkes , & aldcroft 1997 ) .
we now consider the possibility that the weak x - ray flux is an intrinsic characteristic of the continuum being intercepted by the bal region ; _ i.e. _ it is not due primarily to x - ray absorption by the bal gas itself .
for this purpose we have repeated our photoionization calculations using the ionizing continuum shape labeled ( b ) in figure [ fig : broadband ] . here
we connect the x - ray continuum inferred from the _ rosat _ flux assuming @xmath279 with the euv continuum by arbitrarily introducing a break in the euv power - law index at the location of the ionization edge ( 54 ev ) .
the break point is chosen to be consistent with the uv observations ( below our wavelength coverage ) , but at a low enough energy that it should maximize the effects on the results .
naturally , due to the softer continuum , much higher ionization parameters are necessary to produce the highest observed ionization states , @xmath280 for .
more interestingly , the contours of constraint from and shift relative to one another such that the solution for the smaller zone shown in figure [ fig : nuplanetwo ] does not exist for this model .
that is , one can not construct a simple model of the smaller zone that produces the necessary columns of , , and that does not also produce a large column density of . the only explanation in this model for the absorption , therefore , is that there is a very weakly - ionized , low column density zone as mentioned in [ sec : twoslab ] .
figure [ fig : nuplanealt ] shows the location of such a zone in the @xmath252 parameter space , using upper limits of @xmath258 @xmath12 on ( from @xmath211335 ) and @xmath281 @xmath12on ( from @xmath21977 ) , again derived in the same manner as for .
this would require an enormous spread in ionization parameters ( 5 orders of magnitude ) between the different zones .
since the existence of a zone with such a different ionization state , yet having a very similar distribution in velocity space , is not particularly appealing , this indirectly suggests that the ionizing continuum extrapolated from the euv ( labeled ( a ) in figure [ fig : broadband ] ) is a better representation of the true continuum and that the weak x - ray flux is indeed a consequence of x - ray absorption .
the bal system of sbs1542 + 541 has several unusual properties .
the most interesting are : 1 .
the state of ionization is extremely high , higher than what is typically inferred for `` classical '' bal systems from the study of the uv absorption lines .
strong absorption is present from the highly - ionized lithium - like ions of , , and even , which has a creation ionization potential of 476 ev .
absorption from , nearly always present in bal spectra , is undetected .
it should be noted that high ionization similar to this has been suggested in connection with bals , but this is based on x - ray rather than uv absorption properties ( mathur , elvis , & singh 1995 ) .
the high state of ionization and large column density ( @xmath282 @xmath12 ) suggest that the bal system should be associated with an x - ray `` warm absorber '' .
2 . the covered fractions for the various absorbing ions are in at least some cases significantly less than unity .
even more interesting is that the covered fraction is apparently ion - dependent with a general trend towards higher covered fractions with higher states of ionization .
this could be caused by either a special line of sight through a bal disk or by the existence of multiple ionization zones in the bal gas , possibly as a result of density fluctuations .
these properties make the bal system of sbs1542 + 541 observationally unique and certainly intriguing .
it begs the question is this system extraordinary , or is it telling us something important about the character of balqsos in general ?
the weakness of the standard bals of ly@xmath10 , , , and is clearly unusual compared to `` classical '' balqsos .
however , the lack of observed and other low - ionization species itself is perhaps not so surprising given the weakness of , particularly if complex covered - fraction effects are involved .
the very high - ionization species we observe in sbs1542 + 541 occur in the rest - frame euv , a spectral region not usually observable due to intervening lyman - limit absorbers .
hence the existence of these ions is not inconsistent with other balqsos , for which it is not possible to confirm or deny their presence .
it therefore seems plausible that the ionization state of this system is similar to that of normal balqsos . if the existence of large column densities of very highly - ionized gas and/or non - trivial ionization structure as we find in sbs1542 + 541 are common properties of bal systems , then this object could indeed be providing important clues to the nature of the bal phenomenon . this work is based on observations with the nasa / esa hubble space telescope , obtained at the space telescope science institute , which is operated by the association of universities for research in astronomy , inc .
, under nasa contract nas5 - 26555 .
support for this research was provided by nasa grant nag5 - 1630 for the faint object spectrograph team , and nasa contract nas5 - 27000 for the hopkins ultraviolet telescope .
barlow , t. a. , & junkkarinen , v. t. 1994 , , 25 , 1339 boroson , t. a. , & meyers , k. a. 1992 , , 397 , 442 cardelli , j. a. , clayton , g. c. , & mathis , j. s. 1989 , , 345 , 245 cohen , m. h. _ et al .
_ 1995 , , 448 , 77 comastri , a. _ et al . _
1992 , , 384 , 62 davidsen , a. f. , kriss , g. a. , & zheng , w. 1996 , , 380 , 47 ferland , g. j. 1996 , `` hazy a brief introduction to cloudy , '' univ .
kentucky , dept .
physics & astron .
, internal rep .
goodrich , r. w. , & miller , j. s. 1995 , , 448 , 73 green , p. j. , & mathur , s. 1996 , , 462 , 637 hamann , f. _ et al . _
1995 , , 443 , 606 hamann , f. , barlow , t. b. , & junkkarinen , v. 1997 , , 478 , 87 hamann , f. , barlow , t. b. , junkkarinen , v. , & burbidge , e. m. 1997 , , 478 , 80 hamann , f. , korista , k. t. , & morris , s. l. 1993 , , 415 , 541 hu , e. m. , kim , t .- s . , cowie , l. l. , songaila , a. , & rauch , m. 1995 , , 110 , 1526 korista , k. t. _ et al . _
1992 , , 401 , 529 korista , k. t. , & arav , n. 1997 , in asp conf .
128 , mass ejection from agn , ed .
n. arav , i. shlosman , & r. j. weynmann ( san francisco : asp ) , 201 kriss , g. a. 1994 , in asp conf .
61 , astronomical data analysis software and systems iii , ed .
d. r. crabtree , r. j. hanisch , & j. barnes ( san francisco : asp ) , 437 krolik , j. h. , & kriss , g. a. 1995 , , 447 , 512 laor , a. , fiore , f. , elvis , m. , wilkes , b. j. , & mcdowell , j. c. 1997 , , 477 , 93 mathur , s. , elvis , m. , & singh , k. p. 1995
, , 455 , 9 mathur , s. , elvis , m. , & wilkes , b. 1995 , , 452 , 230 mathur , s. , wilkes , b. j. , & aldcroft , t. 1997 , , 478 , 182 mathur , s. , wilkes , b. , elvis , m. , & fiore , f. 1994 , , 434 , 493 mller , p. , & jakobsen , p. 1990 , .
228 , 299 murray , n. , chiang , j. , grossman , s. a. , & voit , g. m. 1995 , , 451 , 498 murray , n. , & chiang , j. 1995 , , 454 , 105 netzer , h. 1993 , , 411 , 594 netzer , h. 1997 , , 473 , 781 press , w. h. , & rybicki , g. b. 1993 , , 418 , 585 schartel , n. _ et al . _
1996 , , 283 , 1015 shull , j. m. , & van steenberg , m. e. 1985 , , 294 , 599 stark , a. a. _ et al . _
1992 , , 79 , 77 turnshek , d. a. 1984 , , 280 , 51 turnshek , d. a. 1988 , in qso absorption lines : probing the universe , ed . s. c. blades , c. a. norman , & d. a. turnshek ( cambridge : cambridge univ . press ) , 17 turnshek , d. a. , monier , e. m. , sirola , c. j. , & espey , b. r. , 1997 , , 476 , 40 turnshek , d. a. _ et al . _
1996 , , 463 , 110 verner , d. a. , barthel , p. d. , & tytler , d. 1994 , , 108 , 287 vogel , s. & reimers , d. 1995 , , 294 , 377 weymann , r. j. , carswell , r. f. , & smith , m. g. 1981 , , 19 , 41 williams , o. r. _ et al . _
1992 , , 389 , 157 zheng , w. , kriss , g. a. , telfer , r. c. , grimes , j. p. , & davidsen , a. f. 1997 , , 475 , 469 | we have analyzed the broad absorption line system of the bright ( @xmath0 ) high - redshift ( @xmath1 ) qso sbs1542 + 541 using uv spectra from the hst fos along with optical data from the mmt and the steward observatory 2.3 m telescope .
these spectra offer continuous wavelength coverage from @xmath2 to @xmath3 , corresponding to @xmath4 in the qso rest frame .
the line of sight to the object contains only three identified intervening lyman - limit absorption systems .
only one of these is optically thick at the lyman edge , a low - redshift ( @xmath5 ) system with a strong lyman edge observed at @xmath6 ( @xmath7 in the rest frame ) in a hopkins ultraviolet telescope spectrum from the astro-2 mission .
the spectra therefore offer a rare opportunity to study broad absorption lines in the rest - frame extreme uv .
we find that the broad absorption line system is lacking in species of relatively low ionization often seen in broad absorption systems , such as , , and . instead , the system is dominated by very high - ionization species .
the strongest features correspond to , , and .
in addition to other high - ionization lines , we identify apparently saturated broad lyman - series lines of order ly@xmath8 and higher .
there is strong evidence for partial occultation of the qso emission source , particularly from the higher - order lyman lines which indicate a covered fraction less than @xmath9 . with the exception of and , which are low - ionization species in the context of this system ,
all of the other lines depress the flux by more than 20% .
absorption from ly@xmath10 also depresses the flux more than 20% , indicating that there are at least two different regions contributing to absorption .
overall , the data suggest a correlation between a larger covered fraction and a higher state of ionization .
these observations reveal inhomogeneity in the ionization structure of the broad absorption line gas .
we have used photoionization models to constrain the total column density and ionization state of the system .
a single - slab model consistent with our observational limits on the column densities requires @xmath11 @xmath12 and an incident ionization parameter @xmath13 .
since the observed covered fractions suggest multiple zones , we also produced a two - slab model and find @xmath14 and @xmath15 for the smaller zone , and @xmath16 @xmath12 and @xmath17 for the larger zone .
we suggest that the different covered fractions can be explained by either a special line of sight through a disk - like geometry or by the existence of density fluctuations of a factor @xmath18 in the bal gas .
the large column density and high state of ionization suggest that the system is likely associated with an x - ray `` warm absorber '' . |
_ ab initio _ treatments of electronic systems become unworkable for sufficiently complex systems . on the other hand , the kohn - sham formulation @xcite of density functional theory ( dft ) @xcite incorporates many - body correlations ( beyond hartree - fock ) , while only single - particle ( sp ) equations must be solved . due to this simplicity dft
is the only feasible approach in some modern applications of electronic structure theory .
there is therefore a continuing interest both in developing new and more accurate functionals and in studying conceptual improvements and extensions to the dft framework .
in particular it is found that dft can handle short - range interelectronic correlations quite well , while there is room for improvements in the description of long - range ( van der waals ) forces and dissociation processes .
microscopic theories offer some guidance in the development of extensions to dft .
orbital dependent functionals can be constructed using many - body perturbation theory ( mbpt ) @xcite .
more recently , the development of general _ ab initio _ dft @xcite addressed the lack of a systematic improvement in dft methods . in this approach
one considers an expansion of the exact ground - state wave function ( e.g. , mbpt or coupled cluster ) from a chosen reference determinant . requiring that the correction to the density vanishes at a certain level of perturbation theory allows one to construct the corresponding approximation to the kohn - sham potential .
a different route has been proposed in ref .
@xcite by developing a quasi - particle ( qp)-dft formalism . in the qp - dft approach the full spectral function
is decomposed in the contribution of the qp excitations , and a remainder or background part
. the latter part is complicated , but does not need to be known accurately : it is sufficient to have a functional model for the energy - averaged background part to set up a single - electron selfconsistency problem that generates the qp excitations .
such an approach is appealing since it contains the well - developed standard kohn - sham formulation of dft as a special case , while at the same time emphasis is put on the correct description of qps , in the landau - migdal sense @xcite .
hence , it can provide an improved description of the dynamics at the fermi surface .
given the close relation between qp - dft and the green s function ( gf ) formulation of many - body theory @xcite , it is natural to employ _ ab initio _ calculations in the latter formalism to investigate the structure of possible qp - dft functionals . in this respect
it is imperative to identify which classes of diagrams are responsible for the correct description of the qp physics .
some previous calculations , based on gf theory , have focused on a self - consistent treatment of the self - energy at the second order @xcite for simple atoms and molecules .
for the atomic binding energies it was found that the bulk of correlations , beyond hartree - fock , are accounted for while significant disagreement with experiment persists for qp properties like ionization energies and electron affinities .
the formalism beyond the second - order approximation was taken up in ref .
@xcite by employing a self - energy of the @xmath0 type @xcite . in this approach
, the random phase approximation ( rpa ) in the particle - hole ( _ ph _ ) channel is adopted to allow for possible collective effects on the atomic excited states .
the latter are coupled to the sp states by means of diagrams like the last two in fig .
[ fig : sigr](c ) .
two variants of the @xmath1 formalism were employed in ref .
@xcite ( where the subscript `` @xmath2 '' indicates that non - dressed propagators are used ) .
in the first only the direct terms of the interelectron coulomb potential are taken into account . in the second version ,
also the exchange terms are included when diagonalizing the _ ph _ space [ generalized rpa ( grpa ) ] and in constructing the self - energy [ generalized @xmath0 ( @xmath3 ) ] .
although the exchange terms are known to be crucial in order to reproduce the experimentally observed rydberg sequence in the excitation spectrum of neutral atoms , they were found to worsen the agreement between the theoretical and experimental ionization energies @xcite . in the @xmath0 approach the sp states
are directly coupled with the two - particle one - hole ( _ 2p1h _ ) and the two - hole
one - particle ( _ 2h1p _ ) spaces . however , only partial diagonalizations ( namely , in the _ ph _ subspaces ) are performed .
this procedure unavoidably neglects pauli correlations with the third particle ( or hole ) outside the subspace . in the case of the @xmath3 approach
, this leads to a double counting of the second order self - energy which must be corrected for explicitly @xcite .
we note that simply subtracting the double counted diagram is not completely satisfactory here , since it introduces poles with negative residues in the self - energy . more important , the interaction between electrons in the two - particle ( _ pp _ ) and two - hole ( _ hh _ ) subspaces are neglected altogether in ( @xmath4)@xmath0 . clearly , it is necessary to identify which contributions , beyond @xmath3 , are needed to correctly reproduce the qp spectrum . in this respect , it is known that highly accurate descriptions of the qp properties in finite systems can be obtained with the algebraic diagrammatic construction ( adc ) method of schirmer and co - workers @xcite .
the most widely used third - order version [ adc(3 ) ] is equivalent to the so - called extended @xmath5 tamm - dancoff ( tda ) method @xcite and allows to predict ionization energies with an accuracy of 10 - 20 mh in atoms and small molecules . upon inspection of its diagrammatic content , the adc(3 ) self - energy
is seen to contain all diagrams where tda excitations are exchanged between the three propagator lines of the intermediate @xmath5 or @xmath6 propagation .
the tda excitations are constructed through a diagonalization in either @xmath5 or @xmath6 space , and neglect ground - state correlations .
however , it is clear that use of tda leads to difficulties for extended systems . in the high - density electron gas _
e.g. _ , the correct plasmon spectrum requires the rpa in the @xmath7 channel , rather than tda . in order to bridge the gap between the qp description in finite and extended systems
, it seems therefore necessary to develop a formalism where the intermediate excitations in the @xmath8 propagator are described at the rpa level .
this can be achieved by a formalism based on employing a set of faddeev equations , as proposed in ref .
@xcite and subsequently applied to nuclear structure problems @xcite . in this approach the grpa equations are solved separately in the _ ph _ and _ pp / hh _ subspaces .
the resulting polarization and two - particle propagators are then coupled through an all - order summation that accounts completely for pauli exchanges in the _
2p1h_/_2h1p _ spaces .
this faddeev - rpa ( f - rpa ) formalism is required if one wants to couple propagators at the rpa level or beyond .
apart from correctly incorporating pauli exchange , f - rpa takes the explicit inclusion of ground - state correlations into account , and can therefore be expected to apply to both finite and extended systems .
the adc(3 ) formalism is recovered as an approximation by neglecting ground - state correlations in the intermediate excitations ( _ i.e. _ replacing rpa with tda phonons ) .
in this work we consider the neon atom and apply the f - rpa method to a nonrelativistic electronic problem for the first time .
the relevant features of the f - rpa formalism ( also extensively treated in ref .
@xcite ) , are introduced in sect .
[ theory ] .
the application to the neon atom is discussed in sec .
[ results ] , where we also investigate the separate effects of the ladder and ring series on the self - energy , as well as the differences between including tda and rpa phonons .
our findings are summarized in sec .
[ conclusions ] .
some more technical aspects are relegated to the appendix , where the interested reader can find the derivation of the faddeev expansion for the _ 2p1h_/_2h1p _ propagator , adapted from ref .
in particular , the approach used to avoid the multiple - frequency dependence of the green s functions is discussed in app .
[ app_time ] , along with its basic assumptions .
the explicit expressions of the faddeev kernels are given in app .
[ app_kernel ] . together with ref .
@xcite , the appendix provides sufficient information for an interested reader to apply the formalism .
the theoretical framework of the present study is that of propagator theory , where the object of interest is the sp propagator , instead of the many - body wave function . in this paper
we will employ the convention of summing over repeated indices , unless specified otherwise .
given a complete orthonormal basis set of sp states , labeled by @xmath9,@xmath10 , ... , the sp propagator can be written in its lehmann representation as @xcite @xmath11 where @xmath12 ( @xmath13 ) are the spectroscopic amplitudes , @xmath14 ( @xmath15 ) are the second quantization destruction ( creation ) operators and @xmath16 ( @xmath17 ) . in these definitions ,
@xmath18 , @xmath19 are the eigenstates , and @xmath20 , @xmath21 the eigenenergies of the ( @xmath22)-electron system .
therefore , the poles of the propagator reflect the electron affinities and ionization energies .
the sp propagator solves the dyson equation @xmath23 which depends on the irreducible self - energy @xmath24 .
the latter can be written as the sum of two terms @xmath25 where @xmath26 represents the hartree - fock diagram for the self - energy . in eqs .
( [ eq : dys ] ) and ( [ eq : sigma1 ] ) , @xmath27 is the sp propagator for the system of noninteracting electrons , whose hamiltonian contains only the kinetic energy and the electron - nucleus attraction .
the @xmath28 represent the antisymmetrized matrix elements of the interelectron ( coulomb ) repulsion .
note that in this work we only consider antisymmetrized elements of the interaction , hence , our result for the ring summation always compare to the generalized @xmath0 approach .
equation ( [ eq : sigma1 ] ) introduces the _
2p1h_/_2h1p_-irreducible propagator @xmath29 , which carries the information concerning the coupling of sp states to more complex configurations . both @xmath24 and @xmath29
have a perturbative expansion as a power series in the interelectron interaction @xmath30 .
some of the diagrams appearing in the expansion of @xmath29 are depicted in fig .
[ fig : sigr ] , together with the corresponding contributions to the self - energy .
note that already at zero order in @xmath29 ( three free lines with no mutual interaction ) the second order self - energy is generated . in terms of the ( antisymmetrized ) coulomb interaction and
undressed propagators .
b ) @xmath29 is related to the self - energy according to eq .
( [ eq : sigma1 ] ) .
c ) by substituting the diagrams a ) in the latter equation , one finds the perturbative expansion of the self - energy . ]
different approximations to the self - energy can be constructed by summing particular classes of diagrams . in this work
we are interested in the summation of rings and ladders , through the ( g)rpa equations . in order to include such effects in @xmath29
, we first consider the polarization propagator describing excited states in the @xmath31-electron system @xmath32 and the two - particle propagator , that describes the addition / removal of two electrons @xmath33 we note that the expansion of @xmath29 arises from applying the equations of motion to the sp propagator ( [ eq : g1 ] ) , which is associated to the ground state @xmath34 .
hence , all the green s functions appearing in this expansion will also be ground state based , including eqs .
( [ eq : pi ] ) and ( [ eq : g2 ] ) .
however the latter contain , in their lehmann representations , all the relevant information regarding the excitation of _ ph _ and _ pp / hh _ collective modes .
the approach of ref .
@xcite consists in computing these quantities by solving the ring - grpa and the ladder - rpa equations @xcite , which are depicted for propagators in fig .
[ fig : rpaeq ] . in the more general case of a self - consistent calculation
, a fragmented input propagator can be used and the corresponding dressed ( g)rpa [ d(g)rpa ] equations @xcite solved [ see eqs .
( [ eq : rpa_ph ] ) and ( [ eq : rpa_ii ] ) ] .
since the propagators ( [ eq : pi ] ) and ( [ eq : g2 ] ) reflect two - body correlations , they still have to be coupled to an additional sp propagator in order to obtain the corresponding approximation for the _ 2p1h _ and _ 2h1p _ components of @xmath29 .
this is achieved by solving two separate sets of faddeev equations .
taking the _ 2p1h _ case as an example , one can split @xmath35 in three different components @xmath36 ( @xmath37 ) that differ from each other by the last pair of lines that interact in their diagrammatic expansion , @xmath38 + \sum_{i=1,2,3 } \bar{r}^{(i)}_{\alpha \beta \gamma ,
\mu \nu \lambda}(\omega ) \ ; , \label{eq : faddfullr } \ ] ] where @xmath39 is the _ 2p1h _ propagator for three freely propagating lines .
these components are solutions of the following set faddeev equations @xcite @xmath40 \ ; , ~ ~ i=1,2,3 ~ \nonumber\end{aligned}\ ] ] where ( @xmath41 ) are cyclic permutations of ( @xmath42 ) . the interaction vertices @xmath43 contain the couplings of a _ ph _ or _ pp / hh _ collective excitation and a freely propagating line .
these are given in the appendix in terms of the polarization ( [ eq : pi ] ) and two - particle ( [ eq : g2 ] ) propagators .
equations .
( [ eq : faddtda ] ) include rpa - like phonons and fully describe the resulting energy dependence of @xmath29
. however , they still neglect energy - independent contributions even at low order in the interaction that also correspond to relevant ground - state correlations .
the latter can be systematically inserted according to @xmath44 where @xmath29 is the propagator we employ in eq .
( [ eq : sigma1 ] ) , @xmath45 is the one obtained by solving eqs .
( [ eq : faddtda ] ) , @xmath46 , and @xmath47 is the identity matrix . following the algebraic diagrammatic construction method @xcite ,
the energy independent term @xmath48 was determined by expanding eq .
( [ eq : uru ] ) in terms of the interaction and imposing that it fulfills perturbation theory up to first order ( corresponding to third order in the self - energy ) .
the resulting @xmath48 , employed in this work , is the same as in ref .
@xcite and is reported in app .
[ app_kernel ] for completeness .
it has been shown that the additional diagrams introduced by this correction are required to obtain accurate qp properties .
equations .
( [ eq : faddtda ] ) and ( [ eq : uru ] ) are valid only in the case in which a mean - field propagator is used to expand @xmath29 .
this is the case of the present work , which employs hartree - fock sp propagators as input .
the derivation of these equations for the general case of a fragmented propagator is given in the appendix .
more details about the actual implementation of the faddeev formalism to _
2p1h_/_2h1p _ propagation have been presented in ref .
the calculation of the _
2h1p _ component of @xmath29 follows completely analogous steps . ) ( left ) . the corresponding contribution to the self - energy , obtained upon insertion into eq .
( [ eq : sigma1 ] ) , is also shown ( right ) . ]
it is important to note that the present formalism includes the effects of _ ph _ and _ pp / hh _ motion to be included simultaneously , while allowing interferences between these modes .
these excitations are evaluated here at the rpa level and are then coupled to each other by solving eqs .
( [ eq : faddtda ] ) .
this generates diagrams as the one displayed in fig .
[ fig : faddex ] , with the caveat that two phonons are not allowed to propagate at the same time .
equations .
( [ eq : faddtda ] ) also assure that pauli correlations are properly taken into account at the _
2p1h_/_2h1p _ level .
in addition , one can in principle employ dressed sp propagators in these equations to generate a self - consistent solution .
if we neglect the ladder propagator @xmath49 ( [ eq : g2 ] ) in this expansion , we are left with the ring series alone and the analogous physics ingredients as for the generalized @xmath0 approach .
however , this differs from @xmath3 due to the fact that no double counting of the second - order self - energy occurs , since the pauli exchanges between the polarization propagator and the third line are properly accounted for ( see fig .
[ fig : faddex ] ) .
alternatively , one can suppress the polarization propagator to investigate the effects of _ pp / hh _ ladders alone .
it is instructive to replace in the above equations all rpa phonons with tda ones ; this amounts to allowing only forward - propagating diagrams in fig .
[ fig : rpaeq ] , and is equivalent to separate diagonalisations in the spaces of @xmath7 , @xmath50 and @xmath51 configurations , relative to the hf ground state .
it can be shown that using these tda phonons to sum all diagrams of the type in fig .
[ fig : faddex ] reduces to one single diagonalization in the @xmath5 or @xmath6 spaces .
therefore , eqs .
( [ eq : faddtda ] ) and ( [ eq : uru ] ) with tda phonons lead directly to the `` extended '' @xmath5 tda of ref .
@xcite , which was later shown to be equivalent to adc(3 ) in the general adc framework @xcite .
the faddeev expansion formalism of ref .
@xcite creates the possibility to go beyond adc(3 ) by including rpa phonons .
this is more satisfactory in the limit of large systems . at the same time
, the computational cost remains modest since only diagonalizations in the @xmath8 spaces are required .
note that complete self - consistency requires the use of fragmented ( or dressed ) propagators in the evaluation of all ingredients leading to the self - energy .
this is outside the scope of the present paper , but we included partial selfconsistency by taking into account the modifications to the hf diagram by employing the correlated one - body density matrix and iterating to convergence .
this is relatively simple to achieve , since the @xmath8 propagator is only evaluated once with the input hf propagators . below we will give results with and without this partial selfconsistency at the hf level .
.parameters that define the sp basis : radius of the confining wall @xmath52 ( in atomic units ) and number of orbits @xmath53 used for different partial waves @xmath54 .
the value of @xmath55 is always set to 5 a.u .. [ cols="<,^,^,^,^,^,^,^,^",options="header " , ] .
[ tab:4 ] as discussed in sec
. [ intro ] , the the f - rpa self - energy contains rpa excitations of both @xmath7 type ( ring diagrams ) and @xmath56 type ( ladder diagrams ) .
it is instructive to analyze their separate contributions to the final ionization energies , in order to understand how the f - rpa self - energy is related to the standard @xmath57 self - energy .
table [ tab:4 ] compares the results for the ionization energies , obtained with the second - order self - energy , to different approximations for including the ring summations .
as one can see , the second - order self - energy generates an @xmath54=1 sp energy of -0.747 mh , which is 46 mh above the empirical @xmath58 ionization energy .
the @xmath1 self - energy , which includes the ring summation with only direct coulomb matrix elements , improves this result and brings it close to experiment .
the @xmath59 behaves in a similar way .
unfortunately , including the exchange terms of the interelectron repulsion in the @xmath60 method turns out to have the opposite effect ( the @xmath58 ionization energy becomes -0.712 h @xcite and the @xmath60 results of ref .
@xcite were obtained by retaining only the diagonal part of the electron self - energy , @xmath61 in the hf+continuum basis .
this approximation was _ not _ made in the present work .
the error in the ionization energies by retaining the diagonal approximation is quite small ( about @xmath622 mh @xcite for the ne atom ) , but larger effects are possible for the total binding energy . ] ) and the agreement with experiment is lost .
obviously , @xmath60 is too simplistic to account for exchange in the @xmath7 channel . with the f - rpa(ring ) self - energy one
can go one step further and employ the faddeev expansion to also force proper pauli exchange correlations in the _
2p1h_/_2h1p _ spaces . as shown in table [ tab:4 ] , this enhances the screening due to the exchange interaction terms , leading to even less binding for the @xmath59 and @xmath58 . the corrections relative to the second - order self - energy can be large ( 100 mh for the @xmath59 state ) and in the direction away from the experimental value .
we also note that the larger shift , in the 2@xmath63 orbit , is accompanied by an increase of the fragmentation ( see fig .
[ fig : sh ] and tab .
[ tab:4 ] ) .
similar observations were also made in ref .
@xcite for other atoms : in general ring summations in the direct channel alone bring the quasihole peaks close to the experiment .
this agreement is then spoiled as soon as one includes proper exchange terms in the self - energy . on the other hand , exchange in the _ ph _ channel
is required to reproduce the correct rydberg sequence in the excitation spectrum of neutral atoms .
so further corrections must arise from other diagrams , and obviously the summation of ladder diagrams can play a relevant role , since these contribute to the expansion of the self - energy at the same level as that of the ring diagrams .
the result when only including ladder - type rpa phonons in the f - rpa self - energy is also shown in table [ tab:4 ] .
one can see that _
pp / hh _ ladders do actually work in the opposite way as the _ ph _ channel ring diagrams , and have the same order of magnitude with , _ e.g. _ , a shift of 66 mh for the @xmath59 relative to the second - order result . when combined with the ring diagrams in the full f - rpa self - energy , the agreement with experiment
is restored again .
note that the final result can not be obtained by adding the contributions of rings and ladders , but depends nontrivially on the interplay between these classes of diagrams thereby pointing to significant interference effects . with the f - rpa(ring ) self - energy , where only the contributions of the @xmath7 channel are included , the main peaks listed in table [ tab:4 ] are not only considerably shifted but also strongly depleted , _
e.g. _ a strength of only 0.56 for the main @xmath59 peak .
the complete spectral function for the @xmath64 strength in fig .
[ fig : sh ] shows that the depletion of the main fragment is accompanied by strong fragmentation over several states . while correlation effects are overestimated in f - rpa(ring ) , they are suppressed in f - rpa(ladder ) , where only the @xmath56 ladders are included in the self - energy . in this case one
finds a spectral distribution closer to the hf one , with a main @xmath59 fragment of strength 0.95 and less fragmentation than the the second - order self - energy .
the spectral distribution generated by the complete f - rpa self - energy is again a combination of the above effects .
the strength of the deeply bound @xmath65 orbital behaves in an analogous way .
the strength of the main peak is reduced but several satellite levels appear due to the mixing with @xmath6 configurations . in all the calculations reported in fig . [
fig : sh ] we found a summed @xmath64 strength exceeding 0.98 in the interval [ -40 h , -30 h ] which can be associated with the @xmath65 orbital , and this remains true even in the presence of strong correlations using the f - rpa(ring ) self - energy .
of course , the mixing with @xmath66 configurations , not included in this work , may further contribute to the fragmentation pattern in this energy region .
in conclusion , the electronic self - energy for the ne atom was computed by the f - rpa method which includes simultaneously the effects of both ring and ladder diagrams .
this was accomplished by employing an expansion of the self - energy based on a set of faddeev equations .
this technique was originally proposed for nuclear structure applications @xcite and is described in the appendix . at the level of the self - energy one
sums all diagrams where the three propagator lines of the intermediate @xmath5 or @xmath6 propagation are connected by repeated exchange of rpa excitations in both the @xmath7 and the @xmath56 channel .
this differs from the adc(3 ) formalism in the fact that the exchanged excitations are of the rpa type , rather than the tda type , and therefore take ground - state correlations effects into account .
the coupling to the external points of the self - energy uses the same modified vertex as in adc(3 ) , which must be introduced to include consistently all third - order perturbative contributions .
the resulting main ionization energies in the neon atom are at least of the same quality , and even somewhat improved , compared to the adc(3 ) result .
note that , numerically , f - rpa can be implemented as a diagonalization in @xmath8 space implying about the same cost as adc(3 ) .
the present study also shows that in localized electronic systems subtle cancellations occur between the ring and ladder series .
in particular , only a combination of the ring and ladder series leads to sensible results , as the separate ring series tends to correct the second - order result in the wrong direction .
since the limit to extended systems requires an rpa treatment of excitations , the f - rpa method holds promise to bridge the gap between an accurate description of quasiparticles in both finite and extended systems .
in particular , the @xmath0 treatment of the electron gas has been shown to yield excellent binding energies , but poor quasiparticle properties @xcite .
further progress beyond @xmath0 theory requires a consistent incorporation of exchange in the @xmath7 channel .
the f - rpa technique may be highly relevant in this respect .
a common framework for calculating accurate qp properties in both finite and extended systems , is also important for constraining functionals in quasiparticle density functional theory ( qp - dft ) @xcite .
finally , complete self - consistency requires sizable computational efforts for bases as large as the hf+continuum basis used here .
it would nevertheless represent an important extension of the present work , since it is related to the fulfillment of conservation laws @xcite .
these issues will be addressed in future work .
this work was supported by the u.s .
national science foundation under grant phy-0652900 .
although only the one - energy ( or two - time ) part of the _ 2p1h_/_2h1p _ propagator enters the definition of the self energy , eq .
( [ eq : sigma1 ] ) , a full resummation of all its diagrammatic contributions would require to treat explicitly the dependence on three separate frequencies , corresponding to the three final lines in the expansion of @xmath29 .
for example , inserting the rpa ring ( ladder ) series in @xmath29 implies the propagation of a _ ph _ ( _ pp / hh _ ) pair of lines both forward and backward in time , while the third line remains unaffected .
a way out of this situation is to solve the bethe - salpeter - like equations for the polarization and ladder propagators separately and then to couple them to the additional line .
if it is assumed that different phonons do not overlap in time , the three lines in between phonon structures will propagate only in one time direction [ see figures ( [ fig : faddex ] ) and ( [ fig : int_ex ] ) ] .
in this situation the integration over several frequencies can be circumvented following the prescription detailed in the next subsection .
this approach will be discussed in the following for the general case of a fully fragmented propgator , in order to derive a set of faddeev equations capable of dressing the sp propagator self - consistently . since the forward ( _ 2p1h _ ) and the backward ( _ 2h1p _ ) parts of @xmath29 decouple in two analogous sets of equations , it is sufficient to focus on the first case alone . ) .
double lines represent fully dressed sp green s funcions which , however , are restricted to propagate only in one time direction [ i.e. , only one of the two terms on the r.h.s .
( [ eq : g1 ] ) is retained ] .
the faddeev eqs .
( [ eq : fadd_1w ] ) and ( [ eq : faddtda ] ) allow for both forward and backward propagation of the phonons @xmath67 and @xmath68 as long as these do not overlap in time . for the propagators
, time ordereing is asumed with forwad propagation in the upward direction . ]
we start by considering the effective interactions in the _ ph _ and _ pp_/_hh _ channels that correspond to eqs .
( [ eq : pi ] ) and ( [ eq : g2 ] ) stripped of the external legs . in the present work , these are the following two - time objects : where the residues and poles for the ring series are @xmath70 and @xmath71 . for the ladders , @xmath72 and @xmath73 , with poles @xmath74 and @xmath75 . equations .
( [ eq : dg ] ) solve the ring and ladder rpa equations , respectively to display how the phonons ( [ eq : dg_ph ] ) and ( [ eq : dg_ii ] ) enter the expansion of @xmath29 , we perform explicitly the frequency integrals for the diagram of fig .
[ fig : int_ex ] . since it is assumed that the separate propagators lines evolve only in one time direction , only the forwardgoing ( @xmath77 ) or the backwardgoing ( @xmath78 ) part of eq .
( [ eq : g1 ] ) must be included for particles and holes , respectively .
after some algebra , one obtains @xmath79 \ ; \omega^{n'_\pi}_{\gamma_1 \beta_1 } \left ( \omega^{n'_\pi}_{\lambda_1 \sigma_1 } \right)^ * } { [ -\varepsilon^{\pi}_{n'_\pi } - \varepsilon^+_{n_2 } + \varepsilon^-_{k_3 } ] [ -\varepsilon^{\pi}_{n'_\pi } - \varepsilon^+_{n_4 } + \varepsilon^-_{k_7 } ] } \right\ } \nonumber \\ & & \times ~ \frac { \left ( { \cal x}^{n_4}_{\sigma_1 } \right)^ * \ ; { \cal x}^{n_4}_{\sigma_2 } } { \omega - ( \varepsilon^+_{n_1 } + \varepsilon^+_{n_4 } - \varepsilon^-_{k_7 } ) + i \eta } \ ; \frac{1}{2 } \left\ { v_{\alpha_1 \sigma_2 , \mu_1 \nu_1 } ~+~ \frac{\left ( \delta^{+,n_{ii}}_{\alpha_1 \sigma_2 } \right)^ * \delta^{+,n_{ii}}_{\mu_1 \nu_1 } } { \omega - ( \varepsilon^{\gamma+}_{n_{ii } } - \varepsilon^-_{k_7 } ) + i \eta } ~+~ \frac { [ \omega + \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_1 } - \varepsilon^+_{n_4 } - \varepsilon^+_{n_5 } - \varepsilon^+_{n_6 } + \varepsilon^-_{k_7 } ] \ ; \delta^{-,k_{ii}}_{\alpha_1 \sigma_2 } \left ( \delta^{-,k_{ii}}_{\mu_1 \nu_1 } \right)^ * } { [ \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_1 } - \varepsilon^+_{n_4 } ] [ \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_5 } - \varepsilon^+_{n_6 } ] } \right\ } \nonumber \\ & & ~ \times ~ \frac { \left ( { \cal x}^{n_5}_{\mu_1 } { \cal x}^{n_6}_{\nu_1 } { \cal y}^{k_7}_{\lambda_1 } \right)^ * \ ; { \cal x}^{n_5}_{\mu } { \cal x}^{n_6}_{\nu } { \cal y}^{k_7}_{\lambda } } { \omega - ( \varepsilon^+_{n_5 } + \varepsilon^+_{n_6 } - \varepsilon^-_{k_7 } ) + i \eta } \nonumber \\ & - & \frac{1}{\omega - ( \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_4 } - \varepsilon^{\pi}_{n'_\pi } ) - i \eta } \frac { \left ( { \cal x}^{n_1}_{\alpha } { \cal x}^{n_2}_{\beta } { \cal y}^{k_3}_{\gamma } \right)^ * \ ; { \cal x}^{n_1}_{\alpha_1 } { \cal x}^{n_2}_{\beta_1 } { \cal y}^{k_3}_{\gamma_1 } ~ \omega^{n'_\pi}_{\gamma_1 \beta_1 } \left ( \omega^{n'_\pi}_{\lambda_1 \sigma_1 } \right)^ * ~ \left ( { \cal x}^{n_4}_{\sigma_1 } \right)^ * \ ; { \cal x}^{n_4}_{\sigma_2 } ~ \delta^{-,k_{ii}}_{\alpha_1 \sigma_2 } \left ( \delta^{-,k_{ii}}_{\mu_1 \nu_1 } \right)^ * ~ \left ( { \cal x}^{n_5}_{\mu_1 } { \cal x}^{n_6}_{\nu_1 } { \cal y}^{k_7}_{\lambda_1 } \right)^ * \ ; { \cal x}^{n_5}_{\mu } { \cal x}^{n_6}_{\nu } { \cal y}^{k_7}_{\lambda } } { [ -\varepsilon^{\pi}_{n'_\pi } - \varepsilon^+_{n_2 } + \varepsilon^-_{k_3 } ] \ ; [ -\varepsilon^{\pi}_{n'_\pi } - \varepsilon^+_{n_4 } + \varepsilon^-_{k_7 } ] \ ; [ \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_1 } - \varepsilon^+_{n_2 } ] \ ; [ \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_5 } - \varepsilon^+_{n_6 } ] } \nonumber\end{aligned}\ ] ] the last term in this expression contains an energy denominator that involves the simultaneous propagation of two phonons .
thus , it will be discarded in accordance with our assumptions .
it must be stressed that similar terms , with overlapping phonons , imply the explicit contribution of at least _
3p2h_/_3h2p_. a proper treatment of these would require a non trivial externsion of the present formalism , which is beyond the scope of this paper . the remaining part in eq .
( [ eq : int_ex ] ) is the relevant contribution for our purposes .
this has the correct energy dependence of a product of denominators that correspond to the intermediate steps of propagation .
all of these involve configurations that have at most _
although , ground state correlations are implicitely included by having already resummed the rpa series .
still , this term does not factorize in a product of separate green s functions due to the summations over the fragmentation indices @xmath80 and @xmath81 [ labeling the eigenstates of the ( n@xmath821)-electron systems ] .
this is overcome if one defines the matrices @xmath83 , @xmath84 and @xmath85 , with elements ( no implicit summation used ) [ eq : g0gmforward ] @xmath86 \ ; \omega^{n'_\pi}_{\gamma \beta } \left ( \omega^{n'_\pi}_{\lambda \nu } \right)^ * } { [ -\varepsilon^{\pi}_{n'_\pi } - \varepsilon^+_{n_\beta } + \varepsilon^-_{k_\gamma } ] [ -\varepsilon^{\pi}_{n'_\pi } - \varepsilon^+_{n_\nu } + \varepsilon^-_{k_\lambda } ] } \right\ } \ ; , \label{eq : gm12f } \\ \gamma^{(3)>}_{\alpha n_\alpha \beta n_\beta \gamma k_\gamma ; \ ; \mu n_\mu \nu n_\nu \lambda k_\lambda}(\omega ) & = & \nonumber \\ & & \hspace{-3 cm } = ~ \frac{\delta_{\gamma , \lambda } \ ; \delta_{k_\gamma , k_\lambda } } { 2 \ ; \sum_\sigma \left| { \cal y}^{k_\gamma}_\sigma \right|^2 } \left\ { v_{\alpha \beta , \mu \nu } ~+~ \sum_{n_{ii } } \frac{\left ( \delta^{+,n_{ii}}_{\alpha \beta } \right)^ * \delta^{+,n_{ii}}_{\mu \nu } } { \omega - ( \varepsilon^{\gamma+}_{n_{ii } } - \varepsilon^-_{k_\gamma } ) + i \eta } ~+~ \sum_{k_{ii } } \frac { [ \omega + \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_\alpha } - \varepsilon^+_{n_\beta } - \varepsilon^+_{n_\mu } - \varepsilon^+_{n_\nu } + \varepsilon^-_{k_\gamma } ] \ ; \delta^{-,k_{ii}}_{\alpha \beta } \left ( \delta^{-,k_{ii}}_{\mu \nu } \right)^ * } { [ \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_\alpha } - \varepsilon^+_{n_\beta } ] [ \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_\mu } - \varepsilon^+_{n_\nu } ] } \right\ } \ ; .
\hspace{1.5 cm } \label{eq : gm3f}\end{aligned}\ ] ] in these definitions , the row and column indices are ordered to represent at first two quasiparticle lines and then a quasihole .
the index ` @xmath87 ' in @xmath88 refer to the line that propagates independently along with the phonon .
using eqs.([eq : g0gmforward ] ) , the first term on the r.h.s .
( [ eq : int_ex ] ) can be written as @xmath89_{\alpha n_\alpha \beta n_\beta \gamma k_\gamma ; \ ; \mu n_\mu \nu n_\nu \lambda k_\lambda } \nonumber\end{aligned}\ ] ] eq .
( [ eq : factor ] ) generalizes to diagrams involving any number of phonon insertions , as long as the terms involving two or more simultaneous phonons are dropped . based on this relation ,
we use the following prescription to avoid performing integrals over frequencies .
one extends all the green s functions to objects depending not only on the sp basis indices ( @xmath90 ) but also on the indices labeling quasi - particles and holes ( @xmath80 and @xmath81 ) .
whether a given argument represents a particle or an hole depends on the type of line being propagated . at this point
one can perform calculations working with only two - time quantities .
the standard propagator is recovered at the end by summing the `` extended '' one over the quasi - particle / hole indices .
the 2p1h/2h1p propagator that includes the full resummation of both the ladder and ring diagrams at the ( g)rpa level is the solution of the following bethe - salpeter - like equation , @xmath91 \ ; g_{\lambda \gamma}(-\omega_3 ) ~ + \nonumber \\ & & \hspace{-0.8 cm } \left\ { g_{\beta \beta_1}(\omega_2 ) g_{\gamma_1 \gamma}(-\omega_3 ) v_{\beta_1 \sigma , \gamma_1 \rho } \int \frac{d s}{2\pi i } r_{\alpha \rho \sigma , \mu \nu \lambda}(\omega_1 , s , \omega_2 + \omega_3 - s ) \right .
\nonumber \\ & & \hspace{-0.8 cm } + \left .
g_{\alpha \alpha_1}(\omega_1 ) g_{\gamma_1
\gamma}(-\omega_3 ) v_{\alpha_1 \sigma , \gamma_1 \rho } \int \frac{d s}{2\pi i } r_{\rho \beta \sigma , \mu \nu \lambda}(s , \omega_2 , \omega_1 + \omega_3 - s ) \right . \nonumber \\ & & \hspace{-1 cm } + \left .
\frac{1}{2 } g_{\alpha \alpha_1}(\omega_1 ) g_{\beta \beta_1}(\omega_2 ) v_{\alpha_1 \beta_1 , \rho \sigma }
\int \frac{d s}{-2\pi i } r_{\rho \sigma\gamma , \mu \nu \lambda}(s , \omega_1 + \omega_2 - s , \omega_3 ) \right\ } .
\nonumber\end{aligned}\ ] ] if this equation is solved , a double integration of @xmath92 would yield the two - time propagator @xmath29 contributing to eq .
( [ eq : sigma1 ] ) . however , the numerical solution of eq .
( [ eq : r_bse ] ) appears beyond reach of the present day computers and one needs to avoid dealing directly with multiple frequencies integrals .
the strategy used is to first solve the rpa equations ( [ eq : rpa_ph ] ) and ( [ eq : rpa_ii ] ) separately .
once this is done it is necessary to rearrange the series ( [ eq : r_bse ] ) in such a way that only the resummed phonons appear . following the formalism introduced by faddeev @xcite
, we identify the components @xmath93 with the three terms between curly brakets in eq .
( [ eq : r_bse ] ) . by employing eqs .
( [ eq : rpa_ph ] ) and ( [ eq : rpa_ii ] ) one is lead to the following set of equations differ from the ones of ref .
@xcite which contain the additional term @xmath94 $ ] .
the two different forms of the faddeev equations that result can be easily related into each other and are completely equivalent .
the definition used here agrees with the standard literature on the subject @xcite .
] , @xmath95 \ ; g_{\lambda \lambda_1}(-s_3 ) \right .
\nonumber \\ & & \left .
+ r^{(j)}_{\mu_1 \nu_1 \lambda_1 , \mu
\nu \lambda}(s_1,s_2 , s_3 ) + r^{(k)}_{\mu_1 \nu_1 \lambda_1 , \mu \nu \lambda}(s_1,s_2 , s_3 ) \right\ } \ ; , \ ; i=1,2,3 \ ; , \nonumber \end{aligned}\ ] ] where ( i , j , k ) are cyclic permutations of ( 1,2,3 ) and the interaction vertices @xmath96 are given by finally , we apply the prescription of sec .
[ app_time ] and substitute @xmath92 with its extended but two - time version @xmath29 .
this leads to the following set of faddeev equations which propagate 2p1h forward in time , @xmath98 \ ; , \nonumber \\ & & \hspace{4 cm } i=1,2,3 \ ; .
\end{aligned}\ ] ] since the full energy dependence is retained in eq .
( [ eq : fadd_3w ] ) , the self - energy corresponding to its solution , @xmath92 , is complete up to third order [ see eq .
( [ eq : sigma1 ] ) ] .
this is no longer the case after the reduction to a two - time propagator . in particular , the approximation that only forward @xmath5 propagation is allowed between different phonons
implies that all diagrams with different time propagation of their external lines are neglected in eqs .
( [ eq : fadd_1w ] ) .
however , these terms are not energy dependent and can be can be reinserted in a systematic way a posteriori as in eq .
( [ eq : uru ] ) . in the general case , @xmath99 and @xmath100 where the correction @xmath101 can be determined by comparison with perturbation theory .
the vertices ( [ eq : g0gmforward ] ) , that appear in eqs .
( [ eq : fadd_1w ] ) , and @xmath102 are expressed in terms of the fully fragmented propagator .
therefore , this approach allows to obtain self - consistent solutions of the sp green s function @xcite .
whenever , like in this work , only a mean - field propagator is employed as input there exist a one - to - one correspondence between the fragmentation indices and the sp basis .
this is expressed by the relations @xmath103 and @xmath104 , where @xmath105 represents the set of occupied orbits . in this case , it is possible to drop one set of indices so that eqs .
( [ eq : fadd_1w ] ) and ( [ eq : urudressed ] ) simplify into the form ( [ eq : faddtda ] ) and ( [ eq : uru ] ) . in practical applications ,
it is worth to note that the poles of the free propagator @xmath106 , eq .
( [ eq : g0f ] ) , do not contribute to the kernel of eqs .
( [ eq : fadd_1w ] ) . this can be proven by employing the closure relations for the rpa problem , in the form obtained by extracting the free poles in eqs .
( [ eq : rpa ] ) . as an example , for the forward poles of the ladder propagator these are @xmath107 } ~ ~\longrightarrow } & & \\ & ~ & \left ( { \cal x}^{n_1}_{\alpha } { \cal x}^{n_2}_{\beta } \right)^ * { \cal x}^{n_1}_{\mu } { \cal x}^{n_2}_{\nu } \ ; \gamma^{(ii)}_{\mu \nu , \gamma \delta}(\omega = \varepsilon^+_{n_1 } + \varepsilon^+_{n_2 } ) ~=~ 0 \ ; , ~ ~ \forall n_1 , n_2 \ ; , \nonumber\end{aligned}\ ] ] and similarly for other cases .
making use of these relations one can derive the following working expression of the kernels of the _
2p1h _ faddeev equations ( no implicit summations used ) [ eq : vert2p1h ] @xmath108_{\alpha n_\alpha \beta n_\beta \gamma k_\gamma ; \ ; \mu n_\mu \nu n_\nu \lambda k_\lambda } & = & \left [ { \bf g^{0>}}(\omega){\bf \gamma^{(2)>}}(\omega ) \right]_{\beta n_\beta \alpha n_\alpha \gamma k_\gamma ; \ ; \nu n_\nu \mu n_\mu \lambda k_\lambda } ~=~ \\ & & \hspace{-5 cm } = ~ \delta_{n_\alpha , n_\mu } \ ; \frac{\left ( { \cal x}^{n_\alpha}_{\alpha } { \cal x}^{n_\beta}_{\beta } { \cal y}^{k_\gamma}_{\gamma } \right)^ * { \cal x}^{n_\alpha}_{\mu } } { \sum_\sigma \left| { \cal x}^{n_\alpha}_\sigma \right|^2 } \left\ { \sum_{n_\pi } \frac { \sum_{\beta_1 \gamma_1 } { \cal x}^{n_\beta}_{\beta_1 } { \cal y}^{k_\gamma}_{\gamma_1 } \left ( \omega^{n_\pi}_{\beta_1 \gamma_1 } \right)^ * \omega^{n_\pi}_{\nu \lambda } } { [ \varepsilon^{\pi}_{n_\pi } - \varepsilon^+_{n_\beta } + \varepsilon^-_{k_\gamma } ] [ \omega - ( \varepsilon^+_{n_\alpha } + \varepsilon^{\pi}_{n_\pi } ) + i \eta ] } ~+~ \sum_{n'_\pi } \frac { \sum_{\beta_1 \gamma_1 } { \cal x}^{n_\beta}_{\beta_1 } { \cal y}^{k_\gamma}_{\gamma_1 } \omega^{n'_\pi}_{\gamma_1 \beta_1 } \left ( \omega^{n'_\pi}_{\lambda \nu } \right)^ * } { [ -\varepsilon^{\pi}_{n'_\pi } - \varepsilon^+_{n_\beta } + \varepsilon^-_{k_\gamma } ] [ -\varepsilon^{\pi}_{n'_\pi } - \varepsilon^+_{n_\nu } + \varepsilon^-_{k_\lambda } ] } \right\ } \ ; , \nonumber \\ \nonumber \\
\label{eq : vert2p1h_3 } \left [ { \bf g^{0>}}(\omega){\bf \gamma^{(3)>}}(\omega ) \right]_{\alpha n_\alpha \beta n_\beta \gamma k_\gamma ; \ ; \mu n_\mu \nu n_\nu \lambda k_\lambda } & = & \\ & & \hspace{-5 cm } = ~ \delta_{k_\gamma , k_\lambda } \ ; \frac{\left ( { \cal x}^{n_\alpha}_{\alpha } { \cal x}^{n_\beta}_{\beta } { \cal y}^{k_\gamma}_{\gamma } \right)^ * { \cal y}^{k_\gamma}_{\lambda}\ ; } { 2 \ ; \sum_\sigma \left| { \cal y}^{k_\gamma}_\sigma \right|^2 } \left\ { \sum_{n_{ii } } \frac{\sum_{\alpha_1 \beta_1 } { \cal x}^{n_\alpha}_{\alpha_1 } { \cal x}^{n_\beta}_{\beta_1 } \left ( \delta^{+,n_{ii}}_{\alpha_1 \beta_1 } \right)^ * \delta^{+,n_{ii}}_{\mu \nu } } { [ \varepsilon^{\gamma+}_{n_{ii } } - \varepsilon^+_{n_\alpha } - \varepsilon^+_{n_\beta } ] [ \omega - ( \varepsilon^{\gamma+}_{n_{ii } } - \varepsilon^-_{k_\gamma } ) + i \eta ] } ~+~ \sum_{k_{ii } } \frac{\sum_{\alpha_1 \beta_1 } { \cal x}^{n_\alpha}_{\alpha_1 } { \cal x}^{n_\beta}_{\beta_1 } \delta^{-,k_{ii}}_{\alpha_1 \beta_1 } \left ( \delta^{-,k_{ii}}_{\mu \nu } \right)^ * } { [ \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_\alpha } - \varepsilon^+_{n_\beta } ] [ \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_\mu } - \varepsilon^+_{n_\nu } ] } \right\ } \ ; . \nonumber\end{aligned}\ ] ] after substituting eq . ( [ eq : urudressed ] ) into ( [ eq : sigma1 ] ) , one needs the working expression for the matrix product @xmath109 ( where @xmath110 is the interelectron interaction ) .
the minimum correction that guaranties to reproduce all third order self - energy diagrams is @xmath111_{\alpha ; \ ; \mu n_\mu \nu n_\nu \lambda k_\lambda } & = & v_{\alpha \lambda , \mu \nu } ~+~ \frac { v_{\alpha \lambda , \gamma_1 \delta_1 } \ ; { \cal y}^{k_\gamma}_{\gamma_1 } { \cal y}^{k_\delta}_{\delta_1 } \left ( { \cal y}^{k_\gamma}_{\gamma_2 } { \cal y}^{k_\delta}_{\delta_2 } \right)^ * \ ; v_{\gamma_2 \delta_2 , \mu \nu } } { 2 \ ; [ \varepsilon^-_{k_\gamma } + \varepsilon^-_{k_\delta } - \varepsilon^+_{n_\mu } - \varepsilon^+_{n_\nu } ] } \\ & & ~~+~ \frac { v_{\alpha \delta_1 , \mu \gamma_1 } \ ; { \cal y}^{k_\gamma}_{\gamma_1 } { \cal x}^{n_\delta}_{\delta_1 } \left ( { \cal y}^{k_\gamma}_{\gamma_2 } { \cal x}^{n_\delta}_{\delta_2 } \right)^ * \ ; v_{\gamma_2 \lambda , \delta_2 \nu } } { [ \varepsilon^-_{k_\gamma } + \varepsilon^-_{k_\lambda } - \varepsilon^+_{n_\delta } - \varepsilon^+_{n_\nu } ] } ~ - ~ \frac { v_{\alpha \delta_1 , \nu \gamma_1 } \ ; { \cal y}^{k_\gamma}_{\gamma_1 } { \cal x}^{n_\delta}_{\delta_1 } \left ( { \cal y}^{k_\gamma}_{\gamma_2 } { \cal x}^{n_\delta}_{\delta_2 } \right)^ * v_{\gamma_2 \lambda , \delta_2 \mu } } { [ \varepsilon^-_{k_\gamma } + \varepsilon^-_{k_\lambda } - \varepsilon^+_{n_\delta } - \varepsilon^+_{n_\mu } ] } \ ; . \nonumber\end{aligned}\ ] ] the case of _
2h1p _ is handled in a completely analogous way along the steps of secs .
( [ app_time ] ) and ( [ app_faddeqs ] ) . after extending @xmath92 to depend on the fragmentation indices ( @xmath112,@xmath113,@xmath114 ) , the _
2h1p _ equivalent of eq .
( [ eq : fadd_1w ] ) is obtained with the following definitions of the kernels , [ eq : vert2h1p ] @xmath115_{\alpha k_\alpha \beta k_\beta \gamma n_\gamma ; \ ; \mu k_\mu \nu k_\nu \lambda n_\lambda } & = & \left [ { \bf g^{0>}}(\omega){\bf \gamma^{(2)>}}(\omega ) \right]_{\beta k_\beta \alpha k_\alpha \gamma n_\gamma ; \ ; \nu k_\nu \mu k_\mu \lambda n_\lambda } ~=~ \\ & & \hspace{-5 cm } = ~ \delta_{k_\alpha , k_\mu } \ ; \frac { { \cal y}^{k_\alpha}_{\alpha } { \cal y}^{k_\beta}_{\beta } { \cal x}^{n_\gamma}_{\gamma } \left ( { \cal y}^{k_\alpha}_{\mu } \right)^ * } { \sum_\sigma \left| { \cal y}^{k_\alpha}_\sigma \right|^2 } \left\ { \sum_{n'_\pi } \frac { \sum_{\beta_1 \gamma_1 } \left ( { \cal y}^{k_\beta}_{\beta_1 } { \cal x}^{n_\gamma}_{\gamma_1 } \right)^ * \omega^{n'_\pi}_{\gamma_1 \beta_1 } \left ( \omega^{n'_\pi}_{\lambda \nu } \right)^ * } { [ -\varepsilon^{\pi}_{n'_\pi } - \varepsilon^-_{k_\beta } + \varepsilon^+_{n_\gamma } ] [ \omega - ( \varepsilon^-_{k_\alpha } - \varepsilon^{\pi}_{n'_\pi } ) - i \eta ] } ~+~ \sum_{n_\pi } \frac { \sum_{\beta_1 \gamma_1 } \left ( { \cal y}^{k_\beta}_{\beta_1 } { \cal x}^{n_\gamma}_{\gamma_1 } \omega^{n_\pi}_{\beta_1 \gamma_1 } \right)^ * \omega^{n_\pi}_{\nu \lambda } } { [ \varepsilon^{\pi}_{n_\pi } - \varepsilon^-_{k_\beta } + \varepsilon^+_{n_\gamma } ] [ \varepsilon^{\pi}_{n_\pi } - \varepsilon^-_{k_\nu } + \varepsilon^+_{n_\lambda } ] } \right\ } \ ; , \nonumber \\ \nonumber \\
\label{eq : vert2h1p_3 } \left [ { \bf g^{0>}}(\omega){\bf \gamma^{(3)>}}(\omega ) \right]_{\alpha k_\alpha \beta k_\beta \gamma n_\gamma ; \ ; \mu k_\mu \nu k_\nu \lambda n_\lambda } & = & \\ & & \hspace{-5 cm } = ~ \delta_{n_\gamma , n_\lambda } \ ; \frac { { \cal y}^{k_\alpha}_{\alpha } { \cal y}^{k_\beta}_{\beta } { \cal x}^{n_\gamma}_{\gamma } \left ( { \cal x}^{n_\gamma}_{\lambda}\ ; \right)^ * } { 2 \ ; \sum_\sigma \left| { \cal x}^{n_\gamma}_\sigma \right|^2 } \left\ { \sum_{k_{ii } } \frac{\sum_{\alpha_1 \beta_1 } \left ( { \cal y}^{k_\alpha}_{\alpha_1 } { \cal y}^{k_\beta}_{\beta_1 } \right)^ * \delta^{-,n_{ii}}_{\alpha_1 \beta_1 } \left ( \delta^{-,n_{ii}}_{\mu \nu } \right)^ * } { [ \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^-_{k_\alpha } - \varepsilon^-_{k_\beta } ] [ \omega - ( \varepsilon^{\gamma-}_{k_{ii } } - \varepsilon^+_{n_\gamma } ) - i \eta ] } ~+~ \sum_{n_{ii } } \frac{\sum_{\alpha_1 \beta_1 } \left ( { \cal y}^{k_\alpha}_{\alpha_1 } { \cal y}^{k_\beta}_{\beta_1 } \delta^{+,k_{ii}}_{\alpha_1 \beta_1 } \right)^ * \delta^{+,k_{ii}}_{\mu \nu } } { [ \varepsilon^{\gamma+}_{n_{ii } } - \varepsilon^-_{k_\alpha } - \varepsilon^-_{k_\beta } ] [ \varepsilon^{\gamma+}_{n_{ii } } - \varepsilon^-_{k_\mu } - \varepsilon^-_{k_\nu } ] } \right\ } \ ; , \nonumber\end{aligned}\ ] ] and correction to the external legs , @xmath116_{\alpha ; \ ; \mu k_\mu \nu k_\nu \lambda n_\lambda } & = & v_{\alpha \lambda , \mu \nu } ~+~ \frac { v_{\alpha \lambda , \gamma_1 \delta_1 } \ ; { \cal x}^{n_\gamma}_{\gamma_1 } { \cal x}^{n_\delta}_{\delta_1 } \left ( { \cal x}^{n_\gamma}_{\gamma_2 } { \cal x}^{n_\delta}_{\delta_2 } \right)^ * \ ; v_{\gamma_2 \delta_2 , \mu \nu } } { 2 \ ; [ \varepsilon^-_{k_\mu } + \varepsilon^-_{k_\nu } - \varepsilon^+_{n_\gamma } - \varepsilon^+_{n_\delta } ] } \\ & & ~~+~ \frac { v_{\alpha \delta_1 , \mu \gamma_1 } \ ; { \cal x}^{n_\gamma}_{\gamma_1 } { \cal y}^{k_\delta}_{\delta_1 } \left ( { \cal x}^{n_\gamma}_{\gamma_2 } { \cal y}^{k_\delta}_{\delta_2 } \right)^ * \ ; v_{\gamma_2 \lambda , \delta_2 \nu } } { [ \varepsilon^-_{k_\delta } + \varepsilon^-_{k_\nu } - \varepsilon^+_{n_\gamma } - \varepsilon^+_{n_\lambda } ] } ~ - ~ \frac { v_{\alpha \delta_1 , \nu \gamma_1 } \ ; { \cal x}^{n_\gamma}_{\gamma_1 } { \cal y}^{k_\delta}_{\delta_1 } \left ( { \cal x}^{n_\gamma}_{\gamma_2 } { \cal y}^{k_\delta}_{\delta_2 } \right)^ * v_{\gamma_2 \lambda ,
\delta_2 \mu } } { [ \varepsilon^-_{k_\delta } + \varepsilon^-_{k_\mu } - \varepsilon^+_{n_\gamma } - \varepsilon^+_{n_\lambda } ] } \ ; . \nonumber\end{aligned}\ ] ] it should be pointed out that while the prescription of sec .
[ app_time ] allows sp lines to propagate only in one time direction , it allows for backward propagation of the phonons .
these contributions translate directly into the energy independent terms of eqs .
( [ eq : vert2p1h ] ) and ( [ eq : vert2h1p ] ) and are a direct consequence of the inversion pattern typical of rpa theory .
these terms have normally a weaker impact than the direct ones on the solutions of eqs .
( [ eq : fadd_1w ] ) .
however , it is show in ref .
@xcite that they are crucial to guarantee the exact separation of the spurious solutions always introduced by the faddeev formalism @xcite if rpa phonons are used .
for the same reasons , the last terms in curly brackets of eqs .
( [ eq : vert2p1h ] ) and ( [ eq : vert2h1p ] ) should be dropped whenever tamm - dancoff ( tda ) phonons are propagated .
the approach followed in this work for solving eqs .
( [ eq : fadd_1w ] ) is to transform them into a matrix representations @xcite .
once this is done , one is left with an eigenvalue problem that depends only on the _ 2p1h _ ( _ 2h1p _ ) configurations ( @xmath117 ) [ ( @xmath118 ) ] .
the spurious states are known exactly @xcite and can be projected out analytically to reduce the computational load . in any case , they would give vanishing contributions to eq .
( [ eq : sigma1 ] ) .
w. kohn and l. j. sham , phys .
140 , a1133 ( 1965 ) .
p. hohenberg and w. kohn , phys .
136 , b864 ( 1964 ) .
a. grling and m. levy , phys .
a 50 , 196 ( 1994 ) .
a. grling , j. chem .
123 , 062203 ( 2005 ) .
r. j. bartlett _ et al .
_ , j. chem .
122 , 034104 ( 2005 ) ; j. chem .
123 , 062205 ( 2005 ) .
p. mori - sanchez , q. wu , and w. t. yang , j. chem .
123 , 062204 ( 2005 ) .
d. van neck , s. verdonck , g. bonny , p. w. ayers , and m. waroquier , phys .
a 74 , 042501(2006 ) .
a. b. migdal , _ theory of finite fermi systems and applications to atomic nuclei _ ( john wiley and sons , new york , 1967 ) . a. l. fetter and j. d. walecka , _ quantum theory of many - particle physics _
( mcgraw - hill , new york , 1971 ) . w. h. dickhoff and d. van neck , _ many - body theory exposed ! _ ( world scientific , singapore , 2005 ) .
d. van neck , k. peirs , and m. waroquier , j. chem .
115 , 15 ( 2001 ) .
k. peirs , d. van neck , and m. waroquier , j. chem .
117 , 4095 ( 2002 ) .
n. e. dahlen and r. van leeuwen , j. chem .
122 , 164102 ( 2005 ) .
s. verdonck , d. van neck , p. w. ayers , m. waroquier , phys .
a 74 , 062503 ( 2006 ) .
shirley and r.m .
martin , phys .
b 47 , 15404 ( 1993 ) .
dahlen , r. van leeuwen , and u. von barth , phys .
a 73 , 012511 ( 2006 ) .
n.e . dahlen and u. von barth , phys . rev .
b 69 , 195102 ( 2004 ) .
a. stan , n.e .
dahlen and r. van leeuwen , europhys .
76 , 298 ( 2006 ) .
l. hedin , phys .
139 , a796 ( 1965 ) .
n. fukuda , f. iwamoto , and k. sawada , phys .
135 , a932 ( 1964 ) . c. barbieri , n. paar , r. roth , and p. papakostantinou , arxiv : nucl - th/0608011v1 .
j. schirmer , l.s .
cederbaum , and o. walter , phys .
a * 28 * , 1237 ( 1983 ) o. walter and j. schirmer , j. phys .
b : at . mol .
14 , 3805 ( 1981 ) . c. barbieri and w. h. dickhoff , phys .
c 63 , 034313 ( 2001 ) . c. barbieri and w. h. dickhoff , phys . rev .
c 65 , 064313 ( 2002 ) .
w. h. dickhoff and c. barbieri , prog .
52 , 337 ( 2004 ) . c. barbieri , phys .
b 643 , 268 ( 2006 ) .
w. j. w. geurts , k. allaart , and w. h. dickhoff , phys . rev .
c 50 , 514 ( 1994 ) .
l. d. faddeev , zh .
. fiz . * 39 * 1459 ( 1961 ) [ sov .
jetp * 12 * , 1014 ( 1961 ) ] .
t.h . dunning jr . ,
j. chem . phys .
90 , 1007 ( 1989 ) ; d.e .
woon and t.h .
dunning jr , j. chem .
98 , 1358 ( 1993 ) a.b .
trofimov and j. schirmer , j. chem .
123 , 144115 ( 2005 ) nist atomic spectra database , nist standard reference database # 78 , http://physics.nist.gov/physrefdata/asd/in-dex.html a. thompson _
_ , _ x - ray data booklet _
( lawrence berkeley national laboratory , berkeley , ca , 2001 ) , and references cited therein .
u. von barth and b. holm , phys .
b * 54 * , 8411 ( 1996 ) ; b. holm and u. von barth , phys .
b * 57 * , 2108 ( 1998 ) ; b. holm , phys . rev . lett . *
83 * , 788 ( 1999 ) .
p. garcia - gonzales and r.w .
godby , phys .
b * 63 * , 075112 ( 2001 ) g. baym and l.p .
kadanoff , phys .
124 , 287 ( 1961 ) .
g. baym , phys .
127 , 1391 ( 1962 ) . c. j. joachain , _ quantum collision theory _ , ( north - holland , amsterdam , 1975 ) .
s. k. adhikari and w. glckle , phys .
c 19 , 616 ( 1979 ) .
j. w. evans and d. k. hoffman , j. math .
* 22 * , 2858 ( 1981 ) . | the spectral function of the closed - shell neon atom is computed by expanding the electron self - energy through a set of faddeev equations .
this method describes the coupling of single - particle degrees of freedom with correlated two - electron , two - hole , and electron - hole pairs .
the excitation spectra are obtained using the random phase approximation , rather than the tamm - dancoff framework employed in the third - order algebraic diagrammatic contruction [ adc(3 ) ] method .
the difference between these two approaches is studied , as well as the interplay between ladder and ring diagrams in the self - energy .
satisfactory results are obtained for the ionization energies as well as the energy of the ground state with the faddeev - rpa scheme that is also appropriate for the high - density electron gas . |
ensuring accountability for security violations is essential in a wide range of settings .
for example , protocols for authentication and key exchange @xcite , electronic voting @xcite , auctions @xcite , and secure multiparty computation ( in the semi - honest model ) @xcite ensure desirable security properties if protocol parties follow their prescribed programs .
however , if they deviate from their prescribed programs and a security property is violated , determining which agents should be held accountable and appropriately punished is important to deter agents from committing future violations .
indeed the importance of accountability in information systems has been recognized in prior work @xcite .
our thesis is that _ actual causation _
( i.e. , identifying which agents actions caused a specific violation ) is a useful building block for accountability in decentralized multi - agent systems , including but not limited to security protocols and ceremonies @xcite .
causation has been of interest to philosophers and ideas from philosophical literature have been introduced into computer science by the seminal work of halpern and pearl @xcite . in particular ,
counterfactual reasoning is appealing as a basis for study of causation .
much of the definitional activity has centered around the question of what it means for event @xmath0 to be an actual cause of event @xmath1 .
an answer to this question is useful to arrive at causal judgments for specific scenarios such as `` john s smoking causes john s cancer '' rather than general inferences such as `` smoking causes cancer '' ( the latter form of judgments are studied in the related topic of type causation @xcite ) .
notably , hume @xcite identified actual causation with counterfactual dependence
the idea that @xmath0 is an actual cause of @xmath1 if had @xmath0 not occurred then @xmath1 would not have occurred .
while this simple idea does not work if there are independent causes , the counterfactual interpretation of actual causation has been developed further and formalized in a number of influential works ( see , for example , @xcite ) .
even though applications of counterfactual causal analysis are starting to emerge in the fields of ai , model - checking , and programming languages , causation has not yet been studied in connection with security protocols and violations thereof . on the other hand
, causal analysis seems to be an intuitive building block for answering some very natural questions that have direct relevance to accountability such as ( i ) _ why _ a particular violation occurred , ( ii ) _ what _ component in the protocol is blameworthy for the violation and ( iii ) _ how _ the protocol could have been designed differently to preempt violations of this sort .
answering these questions requires an in - depth study of , respectively , explanations , blame - assignment , and protocol design , which are interesting problems in their own right , but are not the explicit focus of this paper .
instead , we focus on a formal definition of causation that we believe formal studies of these problems will need .
roughly speaking , explanations can be used to provide an _ account _ of the violation , _ blame assignment _ can be used to hold agents _ accountable _ for the violation , and protocol design informed by these would lead to protocols with better accountability guarantees .
we further elaborate on explanations and blame - assignment in section [ sec : domains ] . formalizing actual causes as a building block for accountability in decentralized multi - agent systems raises new conceptual and technical challenges beyond those addressed in the literature on events as actual causes .
in particular , prior work does not account for the program dynamics that arise in such settings .
let us consider a simple protocol example . in the movie _ flight _
@xcite , a pilot drinks and snorts cocaine before flying a commercial plane , and the plane goes into a locked dive in mid - flight .
while the pilot s behavior is found to be deviant in this case he does not follow the prescribed protocol ( program ) for pilots it is found to not be an actual cause of the plane s dive .
the actual cause was a deviant behavior by the maintenance staff they did not replace a mechanical component that should have been replaced .
ideally , the maintenance staff should have inspected the plane prior to take - off according to their prescribed protocol .
this example is useful to illustrate several key ideas that influence the formal development in this paper .
first , it illustrates the importance of capturing the _ actual interactions _ among agents in a decentralized multi - agent system with non - deterministic execution semantics .
the events in the movie could have unfolded in a different order but it is clear that the actual cause determination needs to be done based on the sequence of events that happened in reality .
for example , had the maintenance staff replaced the faulty component _ before _ the take - off the plane would not have gone into a dive .
second , the example motivates us to hold accountable agents who exercise their choice to execute a deviant _ program _ that actually caused a violation .
the maintenance staff had the choice to replace the faulty component or not where the task of replacing the component could consist of multiple steps .
it is important to identify which of those steps were crucial for the occurrence of the dive .
thus , we focus on formalizing _ program actions _ executed in sequence ( by agents ) as actual causes of violations rather than individual , independent events as formalized in prior work .
finally , the example highlights the difference between deviance and actual causes a difference also noted in prior work on actual causation .
this difference is important from the standpoint of accountability . in particular
, the punishment for deviating from the prescribed protocol could be suspension or license revocation whereas the punishment for actually causing a plane crash in which people died could be significantly higher ( e.g. , imprisonment for manslaughter ) .
the first and second ideas , reflecting our program - based treatment , are the most significant points of difference from prior work on actual causation @xcite while the third idea is a significant point of difference from prior work in accountability @xcite .
the central contribution of this paper is a formal definition of _ program actions as actual causes_. specifically , we define what it means for a set of program actions to be an actual cause of a violation .
the definition considers a set of interacting programs whose concurrent execution , as recorded in a log , violates a trace property .
it identifies a subset of actions ( program steps ) of these programs as an actual cause of the violation .
the definition applies in two phases .
the first phase identifies what we call _ lamport causes_. a lamport cause is a minimal prefix of the log of a violating trace that can account for the violation . in the second phase ,
we refine the actions on this log by removing the actions which are merely _ progress enablers _ and obtain _ actual action causes_. the former contribute only indirectly to the cause by enabling the actual action causes to make progress ; the exact values returned by progress enabling actions are irrelevant .
we demonstrate the value of this formalism in two ways .
first , we prove that violations of a precisely defined class of safety properties always have an actual cause .
thus , our definition applies to relevant security properties .
second , we provide a cause analysis of a representative protocol designed to address weaknesses in the current public key certification infrastructure .
moreover , our example illustrates that our definition cleanly handles the separation between joint and independent causes a recognized challenge for actual cause definitions @xcite .
in addition , we discuss how this formalism can serve as a building block for causal explanations and exoneration ( i.e. , soundly identifying agents who should not be blamed for a violation ) .
we leave the technical development of these concepts for future work .
the rest of the paper is organized as follows .
section [ sec : example ] describes a representative example which we use throughout the paper to explain important concepts .
section [ sec : definitions ] gives formal definitions for program actions as actual causes of security violations .
we apply the causal analysis to the running example in section [ sec : application ] .
we discuss the use of our causal analysis techniques for providing explanations and assigning blame in section [ sec : domains ] .
we survey additional related work in section [ sec : related ] and conclude in section [ sec : conclusion ] .
in this section we describe an example protocol designed to increase accountability in the current public key infrastructure .
we use the protocol later to illustrate key concepts in defining causality .
[ [ security - protocol ] ] security protocol + + + + + + + + + + + + + + + + + consider an authentication protocol in which a user ( @xmath21 ) authenticates to a server ( @xmath31 ) using a pre - shared password over an adversarial network .
@xmath21 sends its user - id to @xmath31 and obtains a public key signed by @xmath31 .
however , @xmath21 would need inputs from additional sources when @xmath31 sends its public key for the first time in a protocol session to verify that the key is indeed bound to @xmath31 s identity .
in particular , @xmath21 can verify the key by contacting multiple notaries in the spirit of _ perspectives _ @xcite .
for simplicity , we assume @xmath21 verifies @xmath31 s public key with three authorized notaries@xmath41 , @xmath42 , @xmath43and accepts the key if and only if the majority of the notaries say that the key is legitimate . to illustrate some of our ideas
, we also consider a parallel protocol where two parties ( @xmath22 and @xmath23 ) communicate with each other .
we assume that the prescribed programs for @xmath31 , @xmath21 , @xmath41 , @xmath42 , @xmath43 , @xmath22 and @xmath23 impose the following requirements on their behavior : ( i ) @xmath31stores @xmath21 s password in a hashed form in a secure private memory location .
( ii ) @xmath21 requests access to the account by sending an encryption of the password ( along with its identity and a timestamp ) to @xmath31 after verifying @xmath31 s public key with a majority of the notaries .
( iii ) the notaries retrieve the key from their databases and attest the key correctly .
( iv ) @xmath31 decrypts and computes the hashed value of the password .
( v ) @xmath31 matches the computed hash value with the previously stored value in the memory location when the account was first created ; if the two hash values match , then @xmath31 grants access to the account to @xmath21 .
( vi ) in parallel , @xmath22 generates and sends a nonce to @xmath23 .
( vii ) @xmath23 generates a nonce and responds to @xmath22 .
[ [ security - property ] ] security property + + + + + + + + + + + + + + + + + the prescribed programs in our example aim to achieve the property that only the user who created the account and password ( in this case , @xmath21 ) gains access to the account . [
[ compromised - notaries - attack ] ] compromised notaries attack + + + + + + + + + + + + + + + + + + + + + + + + + + + we describe an attack scenario and use it to illustrate nuances in formalizing program actions as actual causes .
@xmath21executes its prescribed program .
@xmath21 sends an access request to @xmath31 .
an @xmath5 intercepts the message and sends a public key to @xmath21 pretending to be @xmath31 .
@xmath21 checks with @xmath41 , @xmath42 and @xmath43 who falsely verify @xmath5 s public key to be @xmath31 s key .
consequently , @xmath21 sends the password to @xmath5 .
@xmath5 then initiates a protocol with @xmath31 and gains access to @xmath21 s account . in parallel , @xmath22sends a request to @xmath31 and receives a response from @xmath31 .
following this interaction , @xmath22 forwards the message to @xmath23 .
we assume that the actions of the parties are recorded on a _ log _ , say @xmath6 .
note that this log contains a violation of the security property described above since @xmath5 gains access to an account owned by @xmath21 .
first , our definition finds _ program actions as causes _ of violations . at a high - level ,
as mentioned in the introduction , our definition applies in two phases . the first phase ( section [ sec : definitions ] , definition [ definition : cause1 ] ) identifies a minimal prefix ( phase 1 , _ minimality _ ) of the log that can account for the violation i.e. we consider all scenarios where the sequence of actions execute in the same order as on the log , and test whether it suffices to recreate the violation in the absence of all other actions ( phase 1 , _ sufficiency _ ) . in our example
, this first phase will output a minimal prefix of log @xmath6 above . in this case , the minimal prefix will not contain interactions between @xmath22 and @xmath23 after @xmath31 has granted access to the @xmath5 ( the remaining prefix will still contain a violation ) .
second , a nuance in defining the notion of _ sufficiency _ ( phase 1 , definition [ definition : cause1 ] ) is to constrain the interactions which are a part of the actual cause set in a manner that is consistent with the interaction recorded on the log .
this constraint on interactions is quite subtle to define and depends on how strong a coupling we find appropriate between the log and possible counterfactual traces in sufficiency : if the constraint is too weak then the violation does not reappear in all sequences , thus missing certain causes ; if it is too strong it leads to counter - intuitive cause determinations .
for example , a weak notion of consistency is to require that each program locally execute the same prefix in sufficiency as it does on the log i.e. consistency w.r.t .
program actions for individual programs .
this notion does not work because for some violations to occur the _ order of interactions _ on the log among programs is important .
a notion that is too strong is to require matching of the total order of execution of all actions across all programs .
we present a formal notion of _ consistency _ by comparing log projections ( section [ sec : support - defs ] ) that balance these competing concerns .
third , note that while phase 1 captures a minimal prefix of the log sufficient for the violation , it might be possible to remove actions from this prefix which are merely required for a program execution to progress .
for instance note that while all three notaries actions are required for @xmath21 to progress ( otherwise it would be stuck waiting to receive a message ) and the violation to occur , the actual message sent by one of the notaries is irrelevant since it does not affect the majority decision in this example .
thus , separating out actions which are _ progress enablers _ from those which provide information that causes the violation is useful for fine - grained causal determination .
this observation motivates the final piece ( phase 2 ) of our formal definition ( definition [ definition : cause2a ] ) .
finally , notice that in this example @xmath5 , @xmath41 , @xmath42 , @xmath43 , @xmath31 and @xmath22 deviate from the protocol described above .
however , the deviant programs are not sufficient for the violation to occur without the involvement of @xmath21 , which is also a part of the causal set .
we thus seek a notion of sufficiency in defining a set of programs as a joint actual cause for the violation .
joint causation is also significant in legal contexts @xcite .
for instance , it is useful for holding liable a group of agents working together when none of them satisfy the cause criteria individually but together their actions are found to be a cause .
the ability to distinguish between joint and independent ( i.e. , different sets of programs that independently caused the violation ) causes is an important criterion that we want our definition to satisfy .
in particular , phase 2 of our definition helps identify independent causes .
for instance , in our example , we get three different independent causes depending on which notary s action is treated as a progress enabler .
our ultimate goal is to use the notion of actual cause as a building block for accountability the independent vs. joint cause distinction is significant when making deliberations about accountability and punishment for liable parties .
we can use the result of our causal determinations to further remove deviants whose actions are required for the violation to occur but might not be blameworthy ( section [ sec : domains ] ) .
we present our language model in section [ sec : model ] , auxiliary notions in section [ sec : support - defs ] , properties of interest to our analysis in section [ sec : properties ] , and the formal definition of program actions as actual causes in section [ sec : cause1 ] . we model programs in a simple concurrent language , which we call @xmath7 .
the language contains sequential expressions , @xmath1 , that execute concurrently in threads and communicate with each other through ` send ` and ` recv ` commands .
terms , @xmath8 , denote messages that may be passed through expressions or across threads .
variables @xmath9 range over terms .
an expression is a sequence of actions , @xmath10 .
an action may do one of the following : execute a primitive function @xmath11 on a term @xmath8 ( written @xmath12 ) , or send or receive a message to another thread ( written @xmath13 and @xmath14 , respectively ) .
we also include very primitive condition checking in the form of @xmath15 .
@xmath16 each action @xmath10 is labeled with a unique line number , written @xmath17 .
line numbers help define traces later .
we omit line numbers when they are irrelevant .
every action and expression in the language evaluates to a term and potentially has side - effects .
the term returned by action @xmath10 is bound to @xmath9 in evaluating @xmath18 in the expression @xmath19 .
following standard models of protocols , @xmath20 and @xmath21 are untargeted in the operational semantics : a message sent by a thread may be received by any thread .
targeted communication may be layered on this basic semantics using cryptography .
for readability in examples , we provide an additional first argument to @xmath20 and @xmath21 that specifies the _ intended _
target ( the operational semantics ignore this intended target ) .
action @xmath13 always returns @xmath22 to its continuation .
primitive functions @xmath11 model thread - local computation like arithmetic and cryptographic operations .
primitive functions can also read and update a _ thread - local state _
, which may model local databases , permission matrices , session information , etc .
if the term @xmath8 in @xmath15 evaluates to a non - true value , then its containing thread gets stuck forever , else @xmath15 has no effect .
we abbreviate @xmath23 to @xmath24 and @xmath25 to @xmath26 when @xmath9 is not free in @xmath1 . as an example
, the following expression receives a message , generates a nonce ( through a primitive function @xmath27 ) and sends the concatenation of the received message and the nonce on the network to the intended recipient @xmath28 ( line numbers are omitted here ) .
@xmath29 for the purpose of this paper , we limit attention to this simple expression language , without recursion or branching .
our definition of actual cause is general and applies to any formalism of ( non - deterministic ) interacting agents , but the auxiliary definitions of log projection and the function @xmath30 introduced later must be modified . [
[ app : operational : semantics ] ] operational semantics + + + + + + + + + + + + + + + + + + + + + the language @xmath7 s operational semantics define how a collection of _ threads _ execute concurrently .
each thread @xmath31 contains a unique thread identifier @xmath32 ( drawn from a universal set of such identifiers ) , the executing expression @xmath1 , and a local store .
a _ configuration _ @xmath33 models the threads @xmath34 executing concurrently .
our reduction relation is written @xmath35 and defined in the standard way by interleaving small steps of individual threads ( the reduction relation is parametrized by a semantics of primitive functions @xmath11 ) .
importantly , each reduction can either be internal to a single thread or a _ synchronization _ of a ` send ` in one thread with a ` recv ` in another thread .
we make the locus of a reduction explicit by annotating the reduction arrow with a _ label _
this is written @xmath37 .
a label is either the identifier of a thread @xmath32 paired with a line number @xmath17 , written @xmath38 and representing an internal reduction of some @xmath12 in thread @xmath32 at line number @xmath17 , or a tuple @xmath39 , representing a synchronization between a ` send ` at line number @xmath40 in thread @xmath41 with a ` recv ` at line number @xmath42 in thread @xmath43 , or @xmath44 indicating an unobservable reduction ( of @xmath8 or @xmath15 ) in some thread .
labels @xmath45 are called _ local labels _ , labels @xmath46 are called _ synchronization labels _ and labels @xmath44 are called _
silent labels_. an _ initial configuration _ can be described by a triple @xmath47 , where @xmath48 is a finite set of thread identifiers , @xmath49 and @xmath50 .
this defines an initial configuration of @xmath51 threads with identifiers in @xmath48 , where thread @xmath32 contains the expression @xmath52 and the store @xmath53 . in the sequel
, we identify the triple @xmath54 with the configuration defined by it .
we also use a configuration s identifiers to refer to its threads . given an initial configuration @xmath55 , a run is a finite sequence of labeled reductions @xmath56 .
a pre - trace is obtained by projecting only the stores from each configuration in a run .
let @xmath56 be a run and let @xmath57 be the store in configuration @xmath58 .
then , the pre - trace of the run is the sequence @xmath59 .
if @xmath60 , then the @xmath32th step is an unobservable reduction in some thread and , additionally , @xmath61 .
a trace is a pre - trace from which such @xmath44 steps have been dropped .
the trace of the pre - trace @xmath59 is the subsequence obtained by dropping all tuples of the form @xmath62 .
traces are denoted with the letter @xmath8 . to define actual causation
, we find it convenient to introduce the notion of a log and the log of a trace , which is just the sequence of non - silent labels on the trace .
a log is a sequence of labels other than @xmath44 .
the letter @xmath6 denotes logs .
given a trace @xmath63 , the log of the trace , @xmath64 , is the sequence of @xmath65 .
( the trace @xmath8 does not contain a label @xmath66 that equals @xmath44 , so neither does @xmath64 . )
we need a few more straightforward definitions on logs in order to define actual causation . given a log @xmath6 and a thread identifier @xmath32 , the projection of @xmath6 to @xmath32 , written @xmath67 is the subsequence of all labels in @xmath6 that mention @xmath32 .
formally , @xmath68 we call a log @xmath69 a _ projected prefix _ of the log @xmath6 , written @xmath70 , if for every thread identifier @xmath32 , the sequence @xmath71 is a prefix of the sequence @xmath67 .
the definition of projected prefix allows the relative order of events in two different non - communicating threads to differ in @xmath6 and @xmath69 but lamport s happens - before order of actions @xcite in @xmath69 must be preserved in @xmath6 .
similar to projected prefix , we define projected sublog .
we call a log @xmath69 a _ projected sublog _ of the log @xmath6 , written @xmath72 , if for every thread identifier @xmath32 , the sequence @xmath71 is a subsequence of the sequence @xmath67 ( i.e. , dropping some labels from @xmath67 results in @xmath71 ) .
a _ property _ is a set of ( good ) traces and violations are traces in the complement of the set .
our goal is to define the cause of a violation of a property .
we are specifically interested in ascribing causes to violations of safety properties @xcite because safety properties encompass many relevant security requirements .
we recapitulate the definition of a safety property below . briefly
, a property is safety if it is fully characterized by a set of finite violating prefixes of traces .
let @xmath73 denote the universe of all possible traces .
a property @xmath74 ( a set of traces ) is a safety property , written @xmath75 , if @xmath76 . as we explain soon
, our causal analysis ascribes thread actions ( or threads ) as causes .
one important requirement for such analysis is that the property be closed under reordering of actions in different threads if those actions are not related by lamport s happens - before relation @xcite . for properties that are not closed in this sense ,
the _ global order _ between actions in a race condition may be a cause of a violation . whereas causal analysis of race conditions may be practically relevant in some situation , we limit attention only to properties that are closed in the sense described here .
we call such properties reordering - closed or @xmath77 .
two traces @xmath78 starting from the same initial configuration are called reordering - equivalent , written @xmath79 if for each thread identifier @xmath32 , @xmath80 .
note that @xmath81 is an equivalence relation on traces from a given initial configuration .
let @xmath82_\sim$ ] denote the equivalence class of @xmath8 .
a property @xmath74 is called reordering - closed , written @xmath83 , if @xmath84 implies @xmath82_\sim \subseteq p$ ] .
note that @xmath83 iff @xmath85 .
in the sequel , let @xmath86 denote the _ complement _ of a reordering - closed safety property of interest .
( the subscript @xmath87 stands for `` violations '' . ) consider a trace @xmath8 starting from the initial configuration @xmath55 .
if @xmath88 , then @xmath8 violates the property @xmath89 .
a violation of the property @xmath89 is a trace @xmath88 .
our definition of actual causation identifies a subset of actions in @xmath90 as the cause of a violation @xmath88 .
the definition applies in two phases .
the first phase identifies what we call _ lamport causes_. a lamport cause is a minimal projected prefix of the log of a violating trace that can account for the violation . in the second phase ,
we refine the log by removing actions that are merely _ progress enablers _ ; the remaining actions on the log are the _ actual action causes_. the former contribute only indirectly to the cause by enabling the actual action causes to make progress ; the exact values returned by progress enabling actions are irrelevant .
the following definition , called phase 1 , determines lamport causes .
it works as follows .
we first identify a projected prefix @xmath6 of the log of a violating trace @xmath8 as a potential candidate for a lamport cause .
we then check two conditions on @xmath6 .
the _ sufficiency _ condition tests that the threads of the configuration , when executed at least up to the identified prefix , preserving all synchronizations in the prefix , suffice to recreate the violation .
the _ minimality _ condition tests that the identified lamport cause contains no redundant actions .
[ definition : cause1 ] let @xmath88 be a trace starting from @xmath91 and @xmath6 be a projected prefix of @xmath64 , i.e. , @xmath92 .
we say that @xmath6 is the lamport cause of the violation @xmath8 of @xmath86 if the following hold : 1 .
[ sufficiency1 ] * ( sufficiency ) * let @xmath31 be the set of traces starting from @xmath93 whose logs contain @xmath6 as a projected prefix , i.e. , @xmath94 .
then , every trace in @xmath31 has the violation @xmath86 , i.e. , @xmath95 .
( because @xmath96 , @xmath31 is non - empty . ) 2 . [ minimality1 ] * ( minimality ) * no proper prefix of @xmath6 satisfies condition [ sufficiency1 ] . at the end of phase 1 ,
we obtain one or more minimal prefixes @xmath6 which contain program actions that are sufficient for the violation .
these prefixes represent independent lamport causes of the violation . in the phase 2 definition
below , we further identify a sublog @xmath97 of each @xmath6 , such that the program actions in @xmath97 are actual causes and the actions in @xmath98 are progress enabling actions which only contribute towards the _ progress _ of actions in @xmath97 that cause the violation .
in other words , the actions not considered in @xmath97 contain all labels whose actual returned values are irrelevant . briefly , here s how our phase 2 definition works .
we first pick a candidate projected sublog @xmath97 of @xmath6 , where log @xmath6 is a lamport cause identified in phase 1 .
we consider counterfactual traces obtained from initial configurations in which program actions omitted from @xmath97 are replaced by actions that do not have any effect other than enabling the program to progress ( referred to as no - op ) .
if a violation appears in all such counterfactual traces , then this sublog @xmath97 is a good candidate . of all such good candidates , we choose those that are minimal .
the key technical difficulty in writing this definition is replacing program actions omitted from @xmath97 with no - ops .
we can not simply erase any such action because the action is expected to return a term which is bound to a variable used in the action s continuation .
hence , our approach is to substitute the variables binding the returns of no - oped actions with arbitrary ( side - effect free ) terms @xmath8 .
formally , we assume a function @xmath99 that for line number @xmath17 in thread @xmath32 suggests a suitable term @xmath100 that must be returned if the action from line @xmath17 in thread @xmath32 is replaced with a no - op . in our cause definition we universally quantify over @xmath101 , thus obtaining the effect of a no - op . for technical convenience , we define a syntactic transform called @xmath102 that takes an initial configuration , the chosen sublog @xmath97 and the function @xmath101 , and produces a new initial configuration obtained by erasing actions not in @xmath97 by terms obtained through @xmath101 .
[ def : dummify ] let @xmath54 be a configuration and let @xmath97 be a log .
let @xmath99 .
the dummifying transform @xmath103 is the initial configuration @xmath104 , where for all @xmath105 , @xmath106 is @xmath52 modified as follows : * if @xmath107 appears in @xmath52 but @xmath38 does not appear in @xmath97 , then replace @xmath108 with @xmath109 $ ] in @xmath52 . *
if @xmath110 appears in @xmath52 but @xmath111 does not appear in @xmath97 and @xmath112 , then replace @xmath113 with @xmath114 $ ] in @xmath52 .
we now present our main definition of actual causes . [
definition : cause2a ] let @xmath88 be a trace from the initial configuration @xmath54 and let the log @xmath92 be a lamport cause of the violation determined by definition [ definition : cause1 ] .
let @xmath97 be a projected sublog of @xmath6 , i.e. , let @xmath115 .
we say that @xmath97 is the actual cause of violation @xmath8 of @xmath86 if the following hold : 1 .
[ sufficiency2 ] ( * sufficiency * ) pick any @xmath101 .
let @xmath116 and let @xmath31 be the set of traces starting from @xmath117 whose logs contain @xmath97 as a projected sublog , i.e. , @xmath118 .
then , @xmath31 is non - empty and every trace in @xmath31 has the violation @xmath86 , i.e , @xmath95 .
[ minimality2 ] ( * minimality * ) no proper sublog of @xmath97 satisfies condition [ sufficiency2 ] . at the end of phase 2 ,
we obtain one or more sets of actions @xmath97 .
these sets are deemed the independent actual causes of the violation @xmath8 .
the following theorem states that for all safety properties that are re - ordering closed , the phase 1 and phase 2 definitions always identify at least one lamport and at least one actual cause .
suppose @xmath86 is reordering - closed and the complement of a safety property , i.e. , @xmath119 and @xmath120 .
then , for every @xmath88 : ( 1 ) our phase 1 definition ( definition [ definition : cause1 ] ) finds a lamport cause @xmath6 , and ( 2 ) for every such lamport cause @xmath6 , the phase 2 definition ( definition [ definition : cause2a ] ) finds an actual cause @xmath97 .
( 1 ) pick any @xmath88 .
we follow the phase 1 definition .
it suffices to prove that there is a log @xmath92 that satisfies the sufficiency condition .
since @xmath121 , there is a prefix @xmath122 of @xmath8 s.t . for all @xmath123 , @xmath124 .
choose @xmath125 . since @xmath122 is a prefix of @xmath8 , @xmath126 . to prove sufficiency ,
pick any trace @xmath127 s.t .
@xmath128 . it suffices to prove @xmath129 . since @xmath128 , for each @xmath32 , @xmath130 for some @xmath131 .
let @xmath132 be the ( unique ) subsequence of @xmath127 containing all labels from the logs @xmath133 .
consider the trace @xmath134 .
first , @xmath135 extends @xmath122 , so @xmath136 .
second , @xmath137 because @xmath138 . since @xmath139 , @xmath129 .
( 2 ) pick any @xmath88 and let @xmath6 be a lamport cause of @xmath8 as determined by the phase 1 definition . following the phase 2 definition
, we only need to prove that there is at least one @xmath115 that satisfies the sufficiency condition .
we choose @xmath140 .
to show sufficiency , pick any @xmath101 . because @xmath141 , @xmath97 specifies an initial prefix of every @xmath52 and the transform @xmath142 has no effect on this prefix
first , we need to show that at least one trace @xmath127 starting from @xmath143 satisfies @xmath144 . for this
, we can pick @xmath145 .
second , we need to prove that any trace @xmath127 starting from @xmath146 s.t .
@xmath147 satisfies @xmath129 .
pick such a @xmath127 .
let @xmath122 be the prefix of @xmath8 corresponding to @xmath6 .
then , @xmath148 for each @xmath32 .
it follows immediately that for each @xmath32 , @xmath149 for some @xmath150 .
let @xmath132 be the unique subsequence of @xmath127 containing all labels from traces @xmath151 .
let @xmath134 .
first , because for each @xmath32 , @xmath152 , @xmath153 trivially .
because @xmath6 is a lamport cause , it satisfies the sufficiency condition of phase 1 , so @xmath154 . since @xmath120 , and @xmath135 extends @xmath122 , @xmath136 .
second , @xmath155 because @xmath156 and both @xmath135 and @xmath127 are traces starting from the initial configuration @xmath146 . hence , by @xmath119 , @xmath129 .
our phase 2 definition identifies a set of program actions as causes of a violation .
however , in some applications it may be necessary to ascribe thread identifiers ( or programs ) as causes .
this can be straightforwardly handled by lifting the phase 2 definition : a thread @xmath32 ( or @xmath52 ) is a cause if one of its actions appears in @xmath97 .
[ definition : cause2b ] let @xmath97 be an actual cause of violation @xmath86 on trace @xmath8 starting from @xmath157 .
we say that the set @xmath158 of thread identifiers is a cause of the violation if @xmath159 .
[ [ remarks ] ] remarks + + + + + + + we make a few technical observations about our definitions of cause .
first , because lamport causes ( definition [ definition : cause1 ] ) are projected _ prefixes _ , they contain all actions that occur before any action that actually contributes to the violation .
many of actions in the lamport cause may not contribute to the violation intuitively . our actual cause definition filters out such `` spurious '' actions .
as an example , suppose that a safety property requires that the value @xmath160 never be sent on the network .
the ( only ) trace of the program @xmath161 violates this property . the lamport cause of this violation contains all four actions of the program , but it is intuitively clear that the two actions @xmath162 and @xmath163 do not contribute to the violation .
indeed , the actual cause of the violation determined by definition [ definition : cause2a ] does not contain these two actions ; it contains only @xmath164 and @xmath165 , both of which obviously contribute to the violation .
second , our definition of dummification is based on a program transformation that needs line numbers .
one possibly unwanted consequence is that our traces have line numbers and , hence , we could , in principle , specify safety properties that are sensitive to line numbers .
however , our definitions of cause are closed under bijective renaming of line numbers , so if a safety property is insensitive to line numbers , the actual causes can be quotiented under bijective renamings of line numbers .
third , our definition of actual cause ( definition [ definition : cause2a ] ) separates actions whose return values are relevant to the violation from those whose return values are irrelevant for the violation .
this is closely related to noninterference - like security definitions for information flow control , in particular , those that separate input presence from input content @xcite .
lamport causes ( definition [ definition : cause1 ] ) have a trivial connection to information flow : if an action does not occur in any lamport cause of a violation , then there can not be an information flow from that action to the occurrence of the violation .
in this section , we model an instance of our running example based on passwords ( section [ sec : example ] ) in order to demonstrate our actual cause definition . as explained in section [ sec : example ] , we consider a protocol session where @xmath31 , @xmath21 , @xmath22 , @xmath23 and multiple notaries interact over an adversarial network to establish access over a password - protected account .
we describe a formal model of the protocol in our language , examine the attack scenario from section [ sec : example ] and provide a cause analysis using the definitions from section [ sec : definitions ] .
we consider our example protocol with eight threads named \{@xmath31 , @xmath21 , @xmath5 , @xmath41 , @xmath42 , @xmath43 , @xmath22 , @xmath23}. in this section , we briefly describe the protocol and the programs specified by the protocol for each of these threads .
for this purpose , we assume that we are provided a function @xmath166 such that @xmath167 is the program that _ ideally should have been _ executing in the thread @xmath32 . for each @xmath32 , we call @xmath167 the _ norm _ for thread @xmath32 . the violation is caused because some of the executing programs are different from the norms .
these actual programs , called @xmath168 as in section [ sec : definitions ] , are shown later .
the norms are shown here to help the reader understand what the ideal protocol is and also to facilitate some of the development in section [ sec : domains ] .
the appendix describes an expansion of this example with more than the eight threads considered here to illustrate our definitions better .
the proof included in the appendix deals with timestamps and signatures .
the norms in figure [ fig : norms1 ] and the actuals in figure [ fig : actuals1 ] assume that @xmath21 s account ( called @xmath169 in @xmath31 s program ) has already been created and that @xmath21 s password , @xmath170 is associated with @xmath21 s user i d , @xmath171 .
this association ( in hashed form ) is stored in @xmath31 s local state at pointer @xmath172 .
the norm for @xmath31 is to wait for a request from an entity , respond with its ( @xmath31 s ) public key , wait for a username - password pair encrypted with that public key and grant access to the requester if the password matches the previously stored value in @xmath31 s memory at @xmath172 . to grant access
, @xmath31 adds an entry into a private access matrix , called @xmath74 .
( a separate server thread , not shown here , allows @xmath21 to access its account if this entry exists in @xmath74 . )
the norm for @xmath21 is to send an access request to @xmath31 , wait for the server s public key , verify that key with three notaries and then send its password @xmath170 to @xmath31 , encrypted under @xmath31 s public key . on receiving @xmath31 s public key
, @xmath21 initiates a protocol with the three notaries and accepts or rejects the key based on the response of a majority of the notaries . for simplicity , we omit a detailed description of this protocol between @xmath21 and the notaries that authenticates the notaries and ensures freshness of their responses . these details are included in our appendix . in parallel , the norm for @xmath22 is to generate and send a nonce to @xmath23 . the norm for @xmath23 is to receive a message from @xmath22 , generate a nonce and send it to @xmath22 .
each notary has a private database of _
( public_key , principal ) _ tuples .
the notaries norms assume that this database has already been created correctly .
when @xmath21 sends a request with a public key , the notary responds with the principal s identifier after retrieving the tuple corresponding to the key from its database . the programs in this example use several primitive functions @xmath11 . @xmath173 and @xmath174 denote encryption and decryption of message @xmath175 with key @xmath176 and @xmath177 respectively .
@xmath178 generates the hash of term @xmath175 .
@xmath179 denotes message @xmath175 signed with the key @xmath176 , paired with @xmath175 in the clear . @xmath180 and @xmath181 denote the public and private keys of thread @xmath32 , respectively . for readability
, we include the intended recipient @xmath32 and expected sender @xmath28 of a message as the first argument of @xmath182 and @xmath183 expressions .
as explained earlier , @xmath32 and @xmath28 are ignored during execution and a network adversary , if present , may capture or inject any messages .
the security property of interest to us is that if at time @xmath184 , a thread @xmath176 is given access to account @xmath185 , then @xmath176 owns @xmath185 .
specifically , in this example , we are interested in case @xmath186 and @xmath187 .
this can be formalized by the following logical formula , @xmath89 : @xmath188 here , @xmath189 is the state of the access control matrix @xmath74 for @xmath31 at time @xmath184 .
note that , if any two of the three notaries had attested the @xmath5 s key to belong to @xmath31 , the violation would have still happened .
consequently , we may expect three independent program causes in this example : \{@xmath5 , @xmath21 , @xmath31 , @xmath41 , @xmath42 } with the action causes @xmath97 as shown in figure [ fig : log1](c ) , \{@xmath5 , @xmath21 , @xmath31 , @xmath41 , @xmath43 } with the actions @xmath190 , and \{@xmath5 , @xmath21 , @xmath31 , @xmath42 , @xmath43 } with the actions @xmath191 where @xmath190 and @xmath191 can be obtained from @xmath97 ( figure [ fig : log1](c ) ) by considering actions for \{@xmath41 , @xmath43 } and \{@xmath42 , @xmath43 } respectively , instead of actions for \{@xmath41 , @xmath42}. the following theorem states that our definitions determine exactly these three independent causes . [ theorem - passwords2 ] let @xmath192 , and @xmath193 and @xmath168 be as described above .
let @xmath8 be a trace from @xmath157 such that @xmath194 for each @xmath105 matches the corresponding log projection from figures [ fig : log1](a ) and [ fig : log2 ] .
then , definition [ definition : cause2b ] determines three possible values for the program cause @xmath195 of violation @xmath88 : \{@xmath5 , @xmath21 , @xmath31 , @xmath41 , @xmath42 } , \{@xmath5 , @xmath21 , @xmath31 , @xmath41 , @xmath43 } , and \{@xmath5 , @xmath21 , @xmath31 , @xmath42 , @xmath43 } where the corresponding actual causes are @xmath196 and @xmath191 respectively .
it is instructive to understand the proof of this theorem , as it illustrates our definitions of causation .
we verify that our phase 1 and phase 2 definitions ( definitions [ definition : cause1 ] , [ definition : cause2a ] , [ definition : cause2b ] ) yield exactly the three values for @xmath195 mentioned in the theorem .
we show that any @xmath6 whose projections match those shown in figure [ fig : log1](b ) satisfies sufficiency and minimality . from figure
[ fig : log1](b ) , such an @xmath6 has no actions for @xmath23 , @xmath24 , @xmath44 , @xmath32 and only those actions of @xmath22 that are involved in synchronization with @xmath31 . for all other threads ,
the log contains every action from @xmath8 .
the intuitive explanation for this @xmath6 is straightforward : since @xmath6 must be a ( projected ) _ prefix _ of the trace , and the violation only happens because of @xmath197 in the last statement of @xmath31 s program , every action of every program before that statement in lamport s happens - before relation must be in @xmath6 .
this is exactly the @xmath6 described in figure [ fig : log1](b ) .
formally , following the statement of sufficiency , let @xmath31 be the set of traces starting from @xmath198 ( figure [ fig : actuals1 ] ) whose logs contain @xmath6 as a projected prefix . pick any @xmath199 .
we need to show @xmath129 .
however , note that any @xmath127 containing all actions in @xmath6 must also add @xmath200 to @xmath201 , but @xmath202 .
hence , @xmath129 .
further , @xmath6 is minimal as described in the previous paragraph .
phase 2 ( definitions [ definition : cause2a ] , [ definition : cause2b ] ) determines three independent program causes for @xmath195 : \{@xmath5 , @xmath21 , @xmath31 , @xmath41 , @xmath42 } , \{@xmath5 , @xmath21 , @xmath31 , @xmath41 , @xmath43 } , and \{@xmath5 , @xmath21 , @xmath31 , @xmath42 , @xmath43 } with the actual action causes given by @xmath196 and @xmath191 , respectively in figure [ fig : log1](c ) .
these are symmetric , so we only explain why @xmath97 satisfies definition [ definition : cause2a ] .
( for this @xmath97 , definition [ definition : cause2b ] immediately forces @xmath203 . )
we show that ( a ) @xmath97 satisfies sufficiency , and ( b ) no proper _ sublog _ of @xmath97 satisfies sufficiency ( minimality ) .
note that @xmath97 is obtained from @xmath6 by dropping @xmath43 , @xmath22 and @xmath23 , and all their interactions with other threads .
we start with ( a ) .
let @xmath97 be such that @xmath204 matches figure [ fig : log1](c ) for every @xmath32 .
fix any dummifying function @xmath101 .
we must show that any trace originating from @xmath205 , whose log contains @xmath97 as a projected sublog , is in @xmath86 .
additionally we must show that there is such a trace .
there are two potential issues in mimicking the execution in @xmath97 starting from @xmath206 first , with the interaction between @xmath21 and @xmath43 and , second , with the interaction between @xmath31 and @xmath22 .
for the first interaction , on line 7 , @xmath207 ( figure [ fig : actuals1 ] ) synchronizes with @xmath43 according to @xmath6 , but the synchronization label does not exist in @xmath97 .
however , in @xmath208 , the @xmath14 on line 10 in @xmath207 is replaced with a dummy value , so the execution from @xmath208 progresses .
subsequently , the majority check ( assertion [ b ] ) succeeds as in @xmath6 , because two of the three notaries ( @xmath41 and @xmath42 ) still attest the @xmath5 s key .
a similar observation can be made about the interaction between @xmath31 and @xmath22 .
line 4 , @xmath209 ( from figure [ fig : log1](b ) ) synchronizes with @xmath22 according to @xmath6 , but this synchronization label does not exist in @xmath97 .
however , in @xmath208 , the @xmath14 on line 4 in @xmath209 is replaced with a dummy value , so the execution from @xmath208 progresses .
subsequently , @xmath31 still adds permission for the @xmath5 .
next we prove that every trace starting from @xmath208 , whose log contains @xmath97 ( figure [ fig : log1](c ) ) as a projected _ sublog _ , is in @xmath86 .
fix a trace @xmath127 with log @xmath69 .
assume @xmath69 coincides with @xmath97 .
we show @xmath129 as follows : 1 .
since the synchronization labels in @xmath69 are a superset of those in @xmath97 , @xmath31 must execute line 10 of its program @xmath210 in @xmath127 .
after this line , the access control matrix @xmath201 contains @xmath211 for some @xmath212 .
when @xmath209 writes @xmath213 to @xmath201 at line 10 , then @xmath212 is the third component of a tuple obtained by decrypting a message received on line 7 .
since the synchronization projections on @xmath69 are a superset of @xmath97 , and on @xmath97 @xmath214 synchronizes with @xmath215 , @xmath212 must be the third component of an encrypted message sent on line 10 of @xmath216 .
the third component of the message sent on line 10 by @xmath5 is exactly the term `` @xmath5 '' .
( this is easy to see , as the term `` @xmath5 '' is hardcoded on line 9 . )
hence , @xmath217 .
this immediately implies that @xmath129 since @xmath218 , but @xmath219 .
last , we prove ( b ) that no proper subsequence of @xmath97 satisfies sufficiency. note that @xmath97 ( figure [ fig : log1](c ) ) contains exactly those actions from @xmath6 ( figure [ fig : log1 ] ) on whose returned values the last statement of @xmath31 s program ( figure [ fig : actuals1 ] ) is data or control dependent .
consequently , all of @xmath97 as shown is necessary to obtain the violation . in particular ,
observe that if labels for @xmath31 ( @xmath220 ) are not a part of @xmath190 , then @xmath31 s labels are not in @xmath208 and , hence , on any counterfactual trace @xmath31 can not write to @xmath201 , thus precluding a violation .
therefore , the sequence of labels in @xmath220 are required in the actual cause . by sufficiency , for any @xmath101
, the log of trace @xmath127 of @xmath208 must contain @xmath97 as a projected _ sublog_.
this means that in @xmath127 , the assertion [ a ] of @xmath209 must succeed and , hence , on line 7 , the correct password @xmath221 must be received by @xmath31 , independent of @xmath101 .
this immediately implies that @xmath5 s action of sending that password must be in @xmath97 , else some dummified executions will have the wrong password sent to @xmath31 and the assertion [ a ] will fail . extending this logic further
, we now observe that because @xmath5 forwards a password received from @xmath21 ( line 4 of @xmath222 ) to @xmath31 , the send action of @xmath223 will be in @xmath97 ( otherwise , some dummifications of line 4 of @xmath216 will result in the wrong password being sent to @xmath31 , a contradiction ) .
since @xmath223 s action is in @xmath97 and @xmath69 must contain @xmath97 as a _ sublog _ , the majority check of @xmath207 must also succeed .
this means that at least two of @xmath224 must send the confirmation to @xmath21 , else the dummification of lines 8 10 of @xmath225 will cause the assertion [ b ] to fail for some @xmath101 .
since we are looking for a minimal _ sublog _ therefore we only consider the send actions from two threads i.e. @xmath226 . at this point
we have established that each of the labels as shown in figure [ fig : log1](c ) are required in @xmath97 .
hence , @xmath227 . | protocols for tasks such as authentication , electronic voting , and secure multiparty computation ensure desirable security properties if agents follow their prescribed programs .
however , if some agents deviate from their prescribed programs and a security property is violated , it is important to hold agents _ accountable _ by determining which deviations actually caused the violation .
motivated by these applications , we initiate a formal study of _ program actions as actual causes_. specifically , we define in an interacting program model what it means for a set of program actions to be an actual cause of a violation .
we present a sound technique for establishing program actions as actual causes .
we demonstrate the value of this formalism in two ways .
first , we prove that violations of a specific class of safety properties always have an actual cause .
thus , our definition applies to relevant security properties .
second , we provide a cause analysis of a representative protocol designed to address weaknesses in the current public key certification infrastructure .
security protocols , accountability , audit , causation |
the coexistence of ferromagnetism and superconductivity has received substantial attention over the years . around 1980 ,
such states were predicted@xcite and observed,@xcite and the topic later received renewed interest in the context of experimental observations in rare earth borocarbides.@xcite more recently , interest in this subject has been revived by the observation of coexisting superconductivity and ferromagnetism in uge@xmath1@xcite and urhge,@xcite where both types of order are believed to be due to electrons in the same band .
recent theoretical attention has centered on the structure of the phase diagram,@xcite on the existence of spontaneous flux lattices,@xcite and on the question of spin - triplet versus spin - singlet superconductivity.@xcite in contrast , less is known about the properties of superconductors on the paramagnetic side of , but close to , a ferromagnetic instability .
we will refer to `` paramagnetic superconductors '' to describe systems in this regime , although the superconductivity of course leads to the usual strong diamagnetic effects .
such paramagnetic superconductors include systems below the superconducting transition temperature , but above the temperature below which coexistence of superconductivity and ferromagnetism occurs , as well as systems that never develop ferromagnetism , but are close to a ferromagnetic instability in some direction in parameter space other than temperature .
an example of the latter is believed to be the non - oxide perovskite mgcni@xmath2 , which superconducts below a critical temperature @xmath3.@xcite there is no evidence for a ferromagnetic phase in this material , but it has been suggested that a ferromagnetic ground state can be reached upon a relatively small amount of hole doping.@xcite this system may thus be close to a ferromagnetic instability everywhere in its superconducting phase
. a recent study of mgcni@xmath2 microfibers , with @xmath4 , has revealed an anomalous temperature dependence of the critical current density @xmath5.@xcite the critical current density vanishes at @xmath6 according to a power law @xmath7 , with @xmath8 between about 1% and 10% away from the critical point , and no crossover to the usual ginzburg - landau behavior , which predicts @xmath9 .
the authors of ref . have ruled out morphological effects as an explanation , which raises the question whether proximity to a ferromagnetic state may be responsible . indeed , since ferromagnetic fluctuations are expected to weaken ( singlet ) superconductivity , this is a plausible suggestion for the origin of the weaker - than - expected temperature dependence of @xmath5 .
the probable proximity to ferromagnetism has led to a debate about the nature and symmetry of the pairing in mgcni@xmath2.@xcite this point has not been settled ; some experimental evidence points to conventional s - wave pairing ; other , to a superconducting order parameter with nodes .
the nature of the pairing in the other materials mentioned above has not been unambiguously determined either . in this paper
we will focus on the behavior close to @xmath6 , which is qualitatively independent of the symmetry of the order parameter and thus expected to be the same for all nearly ferromagnetic superconductors .
we use a generalized ginzburg - landau theory to theoretically investigate the electrodynamic properties of a superconductor as a ferromagnetic instability is approached .
we treat the superconductivity in the usual mean - field approximation , but the magnetic critical behavior exactly in a scaling sense .
somewhat counter - intuitively , strong magnetic fluctuations make , in a well - defined sense , the superconductivity more robust in certain respects . in particular , the penetration depth becomes anomalously short .
the thermodynamic critical field , on the other hand , becomes weaker , as one might intuitively expect .
the temperature dependencies of the critical field @xmath10 and the penetration depth @xmath11 depend on the magnetic critical exponents @xmath12 and @xmath13 , respectively . for the critical current @xmath14 , this results in an exponent @xmath15 between 1.5 ( the ginzburg - landau result ) and 2.16 in various temperature regimes .
we will discuss both the existing experimental observations , and predictions for new experiments , in the light of these results .
this paper is organized as follows . in sec .
[ sec : ii ] we give elementary phenomenological arguments for the dependence of the thermodynamic critical field , the penetration depth , and the critical current density , on a constant normal - state magnetic permeability @xmath16 .
we then generalize these results to the magnetically critical case , where one needs to distinguish between @xmath16 and the spin susceptibility @xmath17 in a superconduting state , and both @xmath16 and @xmath17 become nonanalytic functions of various control parameters .
in sec.[sec : iii ] we derive these results from a generalized ginzburg - landau theory , and in sec .
[ sec : iv ] we give a discussion of our results .
we are interested in the electromagnetic properties of superconductors with ferromagnetic fluctuations .
we denote the normal - state spin susceptibility , which describes the response of the spin degrees of freedom to an external magnetic field in the absence of superconductivity , by @xmath18 , and the corresponding spin permeability by @xmath19 .
this is in contrast to the spin permeability @xmath20 , which includes the effects of the superconductivity on the spin response , and the magnetic permeability @xmath21 , which describes the response of the total magnetization , including the diamagnetic part .
it is instructive to first recall the dependence of superconducting properties on a constant @xmath22 , neglecting the distinction between @xmath16 and @xmath17.@xcite this can be done by means of elementary arguments .
consider the free energy density @xmath23 of a system in a magnetic field .
it obeys df = df(h=0 ) + hdb , [ eq:2.1 ] where @xmath24 is the thermodynamic magnetic field , and @xmath25 is the magnetic induction . for the sake of simplicity ,
we ignore the vector nature of various quantities in our free energy considerations . for fixed @xmath25 , @xmath23 is the appropriate thermodynamic potential whose minimum determines the equilibrium state .
however , in an experiment @xmath24 is fixed , since @xmath26 is the external current density , and only the latter is experimentally controlled .
one therefore must perform a legendre transform to a thermodynamic potential @xmath27,@xcite which obeys dg = df(h=0 ) -bdh .
[ eq:2.2 ] in a paramagnetic phase , including paramagnetic superconductors , the relation between @xmath25 and @xmath24 is b = h + 4 m = ( 1 + 4)h = h , [ eq:2.3 ] with @xmath28 the magnetization , @xmath29 the magnetic susceptibility , and @xmath30 the magnetic permeability .
integration of eq .
( [ eq:2.2 ] ) yields g(t , h ) = f(t , h=0 ) - _ 0^h dhh .
[ eq:2.4 ] this is generally valid . in a superconducting meissner state ,
@xmath31 , and hence @xmath32 ( ideal diamagnetism ) , and @xmath33 , with @xmath34 the free energy density of the normal state , @xmath35 the superconducting order parameter , @xmath36 the dimensionless distance from the superconducting critical point , and @xmath37 a parameter . in a normal metal
far from a magnetic instability , and ignoring normal - state diamagnetic effects , @xmath38 , or @xmath39 , and @xmath40 . in a normal metal close to a ferromagnetic critical point , @xmath41 is a complicated function of @xmath42 and @xmath24 .
now consider a superconductor with @xmath43 . according to eq.([eq:2.4 ] ) , the magnetic energy density gained by the system allowing magnetic flux to penetrate , i.e. , the free energy density difference between the meissner state with @xmath31 and the normal state with @xmath44 , is @xmath45 .
by contrast , the condensation energy density gained by the system becoming a superconductor is @xmath46 .
the thermodynamic critical field , which is defined by these two energies being equal , is thus = t/ = ^0/ , [ eq:2.5 ] with @xmath47 the critical field for a system with @xmath48 .
an increase in @xmath16 thus decreases the critical field , as one might expect since the externally applied field is amplified inside the material .
the dependence of the london penetration depth @xmath11 on @xmath49 is intuitively less obvious .
consider a large superconducting sample , with linear dimension @xmath50 , surrounded by vacuum and subject to a homogeneous external magnetic field @xmath51 in @xmath52-direction .
along the left edge of the sample , the magnetic induction will be of the form @xmath53 with @xmath54 ( @xmath55 ) . to determine @xmath56 ,
imagine a thin ( thickness @xmath57 ) layer of normal conducting material around the superconductor .
except for the superconductivity , the normal layer should have the same properties as the superconductor , in particular , a magnetic permeability @xmath49 .
then we have @xmath58 in vacuum , and @xmath59 inside the normal layer , see fig .
[ fig:1](a ) .
now let -0 mm @xmath60 .
then we have ( fig . [ fig:1](b ) ) b(x ) = h x<0 _
he^-x/ x0 .
[ eq:2.6 ] now consider the current density associated with @xmath61 . from ampere s law
we have [ eqs:2.7 ] ( x ) = ( x ) = ( 0,j(x),0 ) , [ eq:2.7a ] with @xmath62 the speed of light and j(x ) = b(x ) . [ eq:2.7b ]
this is the _ total _ current density .
it has three contributions , namely , the supercurrent density @xmath63 , the spin or magnetization current @xmath64 , with @xmath65 the spin contribution to the total magnetization , and the external current density @xmath66 .
the latter vanishes in the case we are considering . in a normal metal , the spin current is the only contribution if we ignore normal - state diamagnetic effects . the spin or normal - state susceptibility @xmath67 is defined as the response of @xmath65 to the total magnetic induction @xmath68 minus the contribution to @xmath68 of @xmath65 itself , = _ ( b - 4 ) = ( /)b .
[ eq:2.8 ] for the supercurrent density @xmath69 this implies [ eqs:2.9 ] _
= ( x ) = ( 0,j_(x),0 ) , [ eq:2.9a ] with j_(x ) = b(x ) .
[ eq:2.9b ] now consider one surface ( area @xmath70 ) of the sample .
neglecting corner effects , and for @xmath71 , the total magnetic flux @xmath72 through that surface is = _ 0^l dy _ 0^l
dxb(x ) l_0^dxb(x ) = l_h .
[ eq:2.10 ] on the other hand , the total supercurrent flowing near that surface is , from eq .
( [ eq:2.9b ] ) , i _ = dxj _
l^2_0^dxj_(x ) = .
[ eq:2.11 ] we thus can write the flux = i _
= nqv , [ eq:2.12 ] where @xmath73 is the number of supercurrent carrying particles , @xmath74 is their charge , and @xmath75 is their velocity . if @xmath76 is their mass , then @xmath77 is the total kinetic energy of the supercurrent .
the flux can thus be written = q. [ eq:2.13 ] now we make two observations .
first , @xmath78 , with @xmath79 the particle number density .
second , at the critical field strength the kinetic energy of the supercurrent must equal the condensation energy in the region where the current is flowing , which is ( see sec.[subsec : ii.a ] ) @xmath80 . with @xmath81 the london penetration depth for @xmath48 , _ 0 = 1/t^1/2 , [ eq:2.14 ] this allows to write the flux at the critical field _
= l h_^0_/_0 = l _ , [ eq:2.15 ] where the first equality follows from eq .
( [ eq:2.13 ] ) , and the second one from eq .
( [ eq:2.10 ] ) . we thus obtain [ eqs:2.16 ] h_/= h_^0/_0 , [ eq:2.16a ] or@xcite = _ 0/ .
[ eq:2.16b ] the penetration depth thus _ decreases _ with increasing @xmath49 , as does the critical field .
this is somewhat counterintuitive , as it implies that the superconductivity becomes in some sense more robust .
it also implies that a large normal - state magnetic permeability will make the superconductor necessarily of type i.@xcite we will come back to this observation .
notice that the above derivation relies only on very general energetic considerations and on ampere s law . also notice that it uses an identity at the critical field strength , where the superconductivity vanishes .
this is fine for @xmath10 , but the penetration depth is a property of the superconducting state , and hence the use of @xmath16 is not quite appropriate for this quantity , except in the limit @xmath82 .
more generally , @xmath11 depends on @xmath17 , which in turn depends on the superconducting properties .
this makes no difference deep inside the paramagnetic superconducting phase , and eq.([eq:2.16b ] ) is valid there .
however , as we will see it makes a crucial difference close to a ferromagnetic instability . in order to discuss the critical current ,
we assume a thin - wire geometry with wire radius @xmath83.@xcite the supercurrent density , which is the total current density minus the spin current density , can be written as a generalization of eq .
( [ eq:2.9a ] ) , j_(x ) & = & + & = & ( x ) , [ eq:2.17 ] where we have used eq .
( [ eq:2.3 ] ) and @xmath84 is the local magnetic susceptibility . now integrate over the cross section of the wire . assuming a homogeneous current density within a distance @xmath11 from the surface , and using gauss s theorem on the right - hand side , we have @xmath85 where we have used eq .
( [ eq:2.6 ] ) .
the critical current density @xmath5 is the one that produces the thermodynamic critical field @xmath10 , which yields the familiar london theory result@xcite = c h_/4 .
[ eq:2.18 ] this result is plausible : dimensionally , @xmath5 must be a magnetic field divided by a length .
the relevant length scale is the thickness of the area that supports diamagnetic currents , which is @xmath11
. the relevant field scale should be the field that corresponds to the condensation energy , which is @xmath86 .
to the extent that @xmath87 , as we have assumed in sec .
[ subsubsec : ii.a.2 ] , eq .
( [ eq:2.16a ] ) implies that @xmath5 is independent of @xmath16 , = ^0 .
[ eq:2.19 ] as we will see below , this result changes drastically in the vicinity of a ferromagnetic instability . in the vicinity of a ferromagnetic instability of the normal metal , the normal state magnetic susceptibility @xmath18 , and hence the permeability @xmath16 ,
become large and diverge as the phase transition is approached . at a ferromagnetic critical point ,
the region of linear response shrinks to zero , and @xmath18 and @xmath16 become strongly field dependent .
this field dependence is characterized by the critical exponent @xmath12,@xcite h^1/-1 .
[ eq:2.20 ] the value of @xmath12 depends on the universality class the particular magnetic system belongs to . for all realistic universality classes , @xmath88 ,
whereas in landau or mean - field theory , @xmath89.@xcite substituting eq .
( [ eq:2.20 ] ) into eq .
( [ eq:2.5 ] ) , we find for the thermodynamic critical field t^2/(+1 ) [ eq:2.21 ] this result holds for a system where the distance @xmath90 from the superconducting critical point can be changed while the system remains tuned to magnetic criticality ( more precisely , to the parameter values where magnetic criticality would occur in the absence of superconductivity ) .
generically , the dimensionless distance @xmath91 from magnetic criticality will change as well if @xmath90 is changed , and we will discuss such more realistic situations in sec.[sec : iv ] . for the penetration depth , the situation is more complicated .
in contrast to @xmath10 , which compares the normal - state magnetic energy with the superconducting condensation energy that has nothing to do with spin magnetism , @xmath11 is entirely a property of the superconducting state , and the feedback of the superconductivity on the spin susceptibility , or the difference between @xmath16 and @xmath17 , can not be neglected . as a result of this feedback ,
the magnetic transition in the presence of superconductivity does not occur at @xmath92 , but rather at a value @xmath93.@xcite here @xmath94 is the magnetic correlation length at zero temperature .
this suggests that the spin susceptibility at @xmath92 will be effectively @xmath95 in a mean - field approximation .
more generally , one has @xmath96 , with @xmath13 another critical exponent . using this in eq .
( [ eq:2.16b ] ) with @xmath16 replaced by @xmath17 , we obtain _ 0
^ 1-/2(_^0)^/2 t^-1/2 + /4 .
[ eq:2.22 ] since @xmath97 for ferromagnetic systems,@xcite this implies that the penetration depth at magnetic criticality is anomalously short .
close to the superconducting transition , the superconductor will therefore also be of type i , in agreement with a conclusion drawn from studying the ferromagnetic phase.@xcite for the critical current density , eqs .
( [ eq:2.18 ] ) , ( [ eq:2.21 ] ) , and ( [ eq:2.22 ] ) predict [ eqs:2.23 ] t^ [ eq:2.23a ] with = 2/(+ 1 ) + 1/2 - /4 , ( r=0 ) .
[ eq:2.23b ] with @xmath88 and @xmath98 this yields @xmath99 , in contrast to the ginzburg - landau result @xmath100 .
these results hold at @xmath92 , and again we have assumed that @xmath90 can be varied independently of @xmath91 .
let us relax the former condition .
from the above argument for the effective value of @xmath101 at @xmath92 it also follows that eq.([eq:2.22 ] ) is valid only for @xmath102 .
since @xmath94 is typically on the order of a few @xmath103 , while @xmath81 is typically several hundred @xmath103 or even larger even at zero temperature , and diverges as @xmath104 for @xmath105 , this is a very small range . by contrast , eq .
( [ eq:2.20 ] ) can be valid for @xmath91 as large as several percent , provided @xmath24 is not too small .
not too close to @xmath6 , where @xmath10 goes to zero , eq .
( [ eq:2.21 ] ) can thus be valid in a substantial @xmath91-range , while @xmath106 except in an extremely small interval around @xmath92 . in that case , = 2/(+ 1 ) + 1/2,(1r _ ^0/_0 ) , [ eq:2.23c ] which yields @xmath107 if @xmath88 . finally , at larger values of @xmath91 , or sufficiently close to @xmath6 that @xmath10 is small enough to invalidate eq .
( [ eq:2.20 ] ) , we are back to the paramagnetic case , eq .
( [ eq:2.19 ] ) holds , and thus @xmath100 .
one thus faces a rather complicated situation , where the exponent @xmath15 can take on values between the ginzburg - landau value 3/2 and a value larger than @xmath108 , eq .
( [ eq:2.23c ] ) , depending on various parameters that are not easy to control or even determine experimentally .
we will discuss this in more detail in sec .
[ sec : iv ] .
before we do so , in the following section we will give a more technical and more detailed derivation of all of our results .
we now consider a coupled field theory that describes both superconducting and spin degrees of freedom in order to derive the above results from a more microscopic level and gain a deeper understanding of their origin .
specifically , we consider a generalization of the usual ginzburg - landau equations that includes the spin degrees of freedom .
far from magnetic criticality , the latter can be integrated out to yield ordinary ginzburg - landau theory with @xmath16 entering the magnetic energy density . at magnetic criticality , @xmath16 becomes field dependent , which changes the thermodynamic critical field .
in addition , the leading term in the london equation vanishes , which leads to a generalized london equation that describes exponential decay on a length scale shorter than @xmath81 , in agreement with the qualitative arguments in sec . [ sec : ii ] , and with implications for the critical current as discussed there . unlike in the previous general discussion , in most of this section we will treat the magnetic critical behavior in a mean - field approximation . our starting point is an action for a complex scalar field @xmath35 describing the superconducting degrees of freedom coupled to a vector potential @xmath109 , and a real vector field @xmath65 describing the spin degrees of freedom.@xcite we reiterate that the qualitative behavior near the superconducting @xmath6 does not depend on the symmetry of the order parameter , so our restriction to a scalar order parameter does not imply a loss of generality .
the action reads s & = & dx.[eq:3.1 ] here and in the remainder of this section we use units such that planck s constant and the speed of light are unity , @xmath110 .
the first line is the standard landau - ginzburg - wilson ( lgw ) functional for singlet superconductors .
the first three terms in the second line are a standard vector-@xmath111 theory , with @xmath112 the fluctuating magnetization .
@xmath65 couples to the vector potential via the @xmath113 term,@xcite with @xmath114 , and the last term is necessary to relate @xmath115 to the appropriate gibbs free energy , see eq .
( [ eq:2.2 ] ) .
notice that @xmath35 and @xmath65 are coupled only indirectly via the vector potential @xmath109 .
spin - flip scattering of electrons by the magnetic moments does give rise to a direct coupling of the form @xmath116,@xcite but these terms are not important for our purposes . minimizing this action with respect to @xmath117 , @xmath109 , and
@xmath65 yields the following saddle - point equations , [ eqs:3.2 ] t_1(x ) + u_1(x)^2(x ) - ( - iqa(x))^2(x ) & = & 0 , [ eq:3.2a ] + -i- ( x)^2 a(x ) & = & , [ eq:3.2b ] + t_2m(x ) - a^2m(x ) + u_2m^2(x ) m(x ) & = & ( x ) .
[ eq:3.2c ] if we drop eq .
( [ eq:3.2c ] ) and put @xmath118 in eq .
( [ eq:3.2b ] ) ( this corresponds to dropping @xmath65 from the action ) we recover the usual ginzburg - landau equations.@xcite a non - superconducting solution of the full equations is @xmath119 , @xmath120 , and @xmath121 determined by the magnetic equation of state ( r - a^2)m(x ) + u_2m^2(x)m(x ) = h(x ) , [ eq:3.3 ] where @xmath122 . for a small constant external field @xmath123 a solution of eq .
( [ eq:3.3 ] ) is @xmath124 , with = 1/r [ eq:3.4 ] the normal - state magnetic susceptibility ( see ref . ) . at this point
it is the bare susceptibility , but it is clear that by renormalizing the spin part of the action before constructing the saddle - point solution one can make it the physical susceptibility . now consider the full eqs .
( [ eqs:3.2 ] ) . for a small and
slowly varying @xmath112 we have from eq .
( [ eq:3.2c ] ) ( x ) = ( ^-1 + 4)^-1b(x ) .
[ eq:3.5 ] substituting this into eq .
( [ eq:3.2b ] ) we obtain [ eqs:3.6 ] _ ( x ) = ( x ) - ( x ) , [ eq:3.6a ] where _ ( x ) & = & -i + & & - ( x)^2 a(x ) .
[ eq:3.6b ] together with eq .
( [ eq:3.2a ] ) , these are the equations of motion for an effective action@xcite s _ & = & dx , + [ eq:3.7 ] where we have dropped the now - superfluous subscript on the landau parameters @xmath90 and @xmath37 .
the same result is of course obtained by starting with eq.([eq:3.1 ] ) and integrating out @xmath65 in a gaussian approximation .
the quantity @xmath63 in eqs .
( [ eqs:3.6 ] ) is indeed the supercurrent , as can be seen by comparing eq .
( [ eq:3.6a ] ) with eq.([eq:2.9a ] ) . it does not explicitly depend on @xmath16 , see eq.([eq:3.6b ] ) , and this is important for the flux quantum to be independent of @xmath125 .
the magnetic energy @xmath126 , which does explicitly depend on @xmath16 , does not appreciably contribute to the free energy of a thin film or wire sample , and the standard determination of the critical current , ref .
, thus leads to the usual ginzburg - landau result with no correction due to @xmath22 .
this corroborates the educated guess in sec .
[ subsubsec : ii.a.3 ] . for all other quantities , the usual analysis of ginzburg - landau theory now applies.@xcite one characteristic length scale is given by the square root of the ratio of the coefficients of the gradient - squared term and the @xmath127 term in eq .
( [ eq:3.7 ] ) .
this is the superconducting coherence length @xmath128 .
another one is given by the square root of the ratio of the coefficients of the terms quadratic in @xmath109 .
for a constant @xmath35 , this is the london penetration depth = _ 0/. [ eq:3.8 ] this is identical with eq .
( [ eq:2.16b ] ) , which had been deduced on elementary phenomenological grounds . for the ginzburg - landau parameter @xmath129 we now have @xmath130 , with @xmath131 the value of the parameter for @xmath132 .
this implies that the superconductor is of type i or type ii , respectively , for @xmath133 or @xmath134 .
while one can show this by an explicit analysis of the effective action , a fast way to relate the theory for arbitrary values of @xmath16 to the one for @xmath132 is to rewrite the action in terms of dimensionless quantities.@xcite in conventional ginzburg - landau theory , this is done by introducing = _ 0 , ( x ) = _ 0 ( ) , a(x ) = h_^0_0 ( ) , .
[ eq:3.9 ] here @xmath135 is the superconducting order parameter scale . in terms of these quantities ,
the effective action reads@xcite s _ & = & d [ ( - i())()^2 + & & - ( ) ^2 + ( ) ^4 + ( ( ) ) ^2 + & & - 2()((x ) ) , + [ eq:3.10 ] a simple further rescaling procedure shows that @xmath136 depends only on a single dimensionless parameter , rather than the two parameters @xmath131 and @xmath16 .
define = /_0 , ( /_0 )
= ( ) , ( /_0 ) = ( ) . [ eq:3.11 ] then s _ & = & d [ ( - i())()^2 + & & - ( ) ^2 + ( ) ^4 + ( ( ) ) ^2 + & & - 2 ( ( ) ) ( ( ) ) . +
[ eq:3.12 ] this shows that the theory with an arbitrary @xmath16 maps onto ordinary ginzburg - landau theory with the replacements _ 0 _ 0/ , h. [ eq:3.13 ] @xmath137 thus indeed marks the demarcation between type i and type ii superconductors , and the critical fields can be immediately obtained from the usual results at @xmath132.@xcite for the thermodynamic critical field @xmath86 , the upper critical field @xmath138 , the lower critical field @xmath139 , and the surface critical field @xmath140 we obtain [ eqs:3.14 ] h _ & = & h_^0/ , [ eq:3.14a ] + h_2 & = & h_2 ^ 0/= _ 0 h_^0/ , [ eq:3.14b ] + h_1 & = & h_1 ^ 0 = g(_0/).[eq:3.14c ] + h_3 & = & 1.695h_2 .
[ eq:3.14d ] where the universal function @xmath141 has the limiting behavior g(x ) = x + 0.08 + o(1/x ) & x1/ , 1 & x = 1/. [ eq:3.14e ] if one neglects the weak dependence of @xmath141 on its argument , @xmath139 is approximately independent of @xmath16 .
as one approaches a ferromagnetic instability , @xmath16 keeps increasing and can no longer be treated as a constant .
there are two effects that become important for our purposes .
first , in a normal metal @xmath16 becomes strongly field or induction dependent . at @xmath92
this dependence is nonanalytic and described by the critical exponent @xmath12 .
second , as @xmath91 becomes on the order of @xmath142 ( see sec .
[ subsec : ii.b ] ) in a superconducting phase , the difference between @xmath16 and @xmath17 can no longer be neglected .
related to this , the gradient squared term in eq .
( [ eq:3.2c ] ) must be taken into account .
we now consider these effects , starting with the nonanalytic field dependence in the normal state . at magnetic criticality in the normal state ,
@xmath92 , one has@xcite ( r=0,h ) = _ 0(h / h_0)^1/- 1 .
[ eq:3.15 ] here @xmath143 is a microscopic susceptibility , and @xmath144 is a microscopic field scale .
@xmath28 and , for small values of @xmath24 , @xmath25 are therefore proportional to @xmath145 , or @xmath146 . for small @xmath25 , the number @xmath16 should thus be replaced by a function of @xmath25 with the following leading @xmath25-dependence , ( h_0/b)^- 1 , [ eq:3.16 ] with @xmath147 .
the magnetic energy cost of the flux expulsion that results from the formation of a meissner phase ( which equals minus the normal - state magnetic energy ) is now obtained by using eq .
( [ eq:3.16 ] ) in eq .
( [ eq:3.7 ] ) .
it is @xmath148 .
the condensation energy is still given by @xmath46 , which yields = ( ) ^/(+1 ) t^2/(+1 ) .
[ eq:3.17 ] the thermodynamical critical field is thus weaker than in the paramagnetic case , and the @xmath90-dependence is consistent with eq .
( [ eq:2.21 ] ) . by comparing with eq .
( [ eq:2.5 ] ) , we see that with respect to the thermodynamical critical field , @xmath16 effectively scales like @xmath149 at magnetic criticality
. equations ( [ eq:3.15 ] ) through ( [ eq:3.17 ] ) hold also for small but nonzero values of @xmath91 as long as one is in the field scaling regime , i.e , as long as @xmath24 in appropriate units is large compared to @xmath91 to an appropriate power .
we will discuss this in more detail in sec .
[ sec : iv ] . at this point
we only mention that , since @xmath10 vanishes as @xmath150 , sufficiently close to @xmath151 one will lose the field scaling for any nonzero value of @xmath91 , and @xmath10 will be given by eq .
( [ eq:3.14a ] ) .
the ordinary london equation is obtained from eq .
( [ eq:3.2b ] ) by dropping @xmath112 and treating @xmath152 as a constant ( london approximation ) . with @xmath153
this leads to -_0 ^ -2b(x ) = ( x ) .
[ eq:3.18 ] now take @xmath65 into account . using eq .
( [ eq:3.2c ] ) in eq.([eq:3.2b ] ) , we can eliminate @xmath68 and derive an equation for @xmath121 .
once @xmath65 is known , @xmath68 follows from eq .
( [ eq:3.2c ] ) . within the london approximation one finds \(x ) & = & -(_0 ^ 2/)(x ) + ( _ ^0)^2 ^ 2m(x ) + ( _ ^0)^2_0 ^ 2 ^ 2m(x ) - um^2(x)m(x ) + & & - u_0 ^ 2 ^ 2(x ) m(x ) .
[ eq:3.19 ] here @xmath154 as in sec .
[ subsec : iii.b ] , @xmath155 , and @xmath156 .
as long as @xmath157 , the first term on the right - hand side of eq.([eq:3.19 ] ) leads to a variation of @xmath65 on a length scale @xmath158 .
the second term is a small correction to the first one since @xmath159 .
so is the third term , which is of order @xmath160 relative to the first one . the linearized version of eq .
( [ eq:3.19 ] ) thus reduces to the ordinary london equation , eq .
( [ eq:3.18 ] ) , with @xmath161 . however , for @xmath92 the first term vanishes .
this makes the second term the leading one , and the third term , which is of order @xmath162 compared to the second one , can not be neglected either .
the linearized equation thus reads ( x ) = ( _ ^0)^2 ^ 2 ( x ) .
[ eq:3.20 ] with the same interface geometry as in sec .
[ subsubsec : ii.a.2 ] this takes the form m(x ) = ( _ ^0)^2 m(x ) - ( _ ^0)^2 _ 0 ^ 2 m^(iv)(x ) .
[ eq:3.21 ] this linear quartic ode is solved by an exponential ansatz , @xmath163 .
the real solution that falls off for @xmath164 shows damped oscillatory behavior . from eq .
( [ eq:3.2c ] ) we see that @xmath61 shows the same behavior as @xmath165 , up to corrections of @xmath166 . with the boundary condition
@xmath167 we finally obtain b(x ) = h e^-x/ ( x/ ) .
[ eq:3.22 ] this is the solution of the linearized version of eq .
( [ eq:3.19 ] ) at @xmath92 .
in addition to leaving out the terms of @xmath168 , we have also ignored the fact that the permeability , whether @xmath16 or @xmath17 , does depend on @xmath25 or @xmath28 at magnetic criticality . in a mean - field approximation , @xmath169 at @xmath92 , see eq .
( [ eq:3.16 ] ) , which also leads to terms of @xmath168 in the nonlinear equation .
depending on the ratio of the external field to @xmath170 , these terms may or may not be important for the initial decay of @xmath28 or @xmath25 near the normal metal - to - superconductor boundary . however , once @xmath28 or @xmath25 has decayed sufficiently , these terms become subleading compared to the linear ones in eq .
( [ eq:3.21 ] ) , and the asymptotic behavior as @xmath171 is always given by eq .
( [ eq:3.22 ] ) . in order to make contact with the discussion in sec .
[ subsec : ii.b ] for small but nonzero values of @xmath91 , let us consider the linearized eq.([eq:3.19 ] ) while keeping the first term . instead of eq .
( [ eq:3.21 ] ) we then have m(x ) = ( _ 0 ^ 2/+ ( _ ^0)^2 ) m(x ) - ( _ ^0)^2 _ 0 ^ 2 m^(iv)(x ) . [ eq:3.23 ]
this is solved by [ eqs:3.24 ] m(x ) = m_0e^x , [ eq:3.24a ] with ^2 & = & .
[ eq:3.24b ] here we have chosen the solution for @xmath172 that yields @xmath173 for @xmath174 . equation ( [ eq:3.24b ] ) still provides two solutions for @xmath175 , and the physical solution for @xmath28 is determined by the requirement that @xmath28 be real . a discussion of eq .
( [ eq:3.24b ] ) shows that @xmath172 becomes purely real and negative at @xmath176 .
this is in agreement with the results of blount and varma,@xcite who showed that spiral magnetic order coexisting with the superconductivity occurs at this point . for @xmath177 one
has @xmath178 , which leads to eq.([eq:3.22 ] ) . for @xmath179 one
finds @xmath180 , which leads to b(x ) = h e^-x/_0 , [ eq:3.25 ] in agreement with eq .
( [ eq:3.8 ] ) . equation ( [ eq:3.22 ] ) shows that the effective penetration depth at magnetic criticality is = , ( r_^0/_0 ) , [ eq:3.26 ] in agreement with the conclusions of ref . drawn from studying the ferromagnetic phase , and with eq .
( [ eq:2.22 ] ) with @xmath181 .
the latter approximation results from the fact that our saddle - point equations of motion describe the magnetic equation of state in a mean - field approximation .
the discussion of eq .
( [ eq:3.24b ] ) shows that this result is valid for @xmath182 . by comparing with eq.([eq:2.16b ] ) or
( [ eq:3.8 ] ) , we see that with respect to the penetration depth , @xmath16 at magnetic criticality scales like @xmath183 in mean - field approximation , or @xmath184 in general .
the fact that @xmath185 in eqs .
( [ eq:2.5 ] ) and ( [ eq:2.16b ] ) , respectively , must be interpreted differently for @xmath186 is a consequence of the influence of the superconductivity on the spin response .
for @xmath179 we have , from eq.([eq:3.25 ] ) = _ 0/ , [ eq:3.27 ] in agreement with eq .
( [ eq:3.8 ] ) . the expression for the critical current given by eq .
( [ eq:2.18 ] ) is general within the london approximation .
we have now given a derivation of the behavior of the thermodynamical critical field and the penetration depth given on phenomenological grounds in eqs .
( [ eq:2.21 ] ) and ( [ eq:2.22 ] ) , respectively .
the behavior of the critical current at or near magnetic criticality is thus given by eqs .
( [ eqs:2.23 ] ) .
the critical exponent @xmath13 is positive ( @xmath98 for typical ferromagnetic universality classes in three dimensions@xcite ) .
the result for @xmath11 , eq .
( [ eq:2.22 ] ) or ( [ eq:3.23 ] ) in mean - field approximation , of the previous subsection therefore means that @xmath11 diverges more slowly for @xmath150 than the superconducting coherence length @xmath188 .
consequently , superconductors at magnetic criticality ( @xmath189 ) are necessarily of type i.@xcite this observation notwithstanding , the critical field @xmath187 , which in a type - ii superconductor signalizes the boundary of the vortex phase , still has a physical meaning : it is the minimum field to which the normal metal can be ` supercooled ' before it discontinuously develops a nonzero superconducting order parameter.@xcite it is thus still of interest to determine @xmath187 .
furthermore , the behavior will be necessarily of type i only for @xmath190 in an extremely narrow region . outside of this region , eq.([eq:3.27 ] ) holds , and for a sufficiently large value of @xmath191 the superconductor will still be of type ii .
the determination of @xmath187 is done by linearizing the ginzburg - landau equation , eq.([eq:3.2a ] ) , in @xmath35 .
it then turns into a schrdinger equation for a particle in a vector potential @xmath109 , with @xmath192 playing the role of the energy eigenvalue . by means of standard
arguments@xcite this leads to a critical value of the magnetic induction @xmath193 given by @xmath194 . in a paramagnetic superconductor
, this leads to h _
= h_^0/(= ) , [ eq:3.28 ] which is the same as eq . ( [ eq:3.14b ] ) . at magnetic criticality
, we have , cf .
( [ eq:3.16 ] ) , h _
= b_^/h_0^(- 1 ) t^. [ eq:3.29 ] notice that , in this context , @xmath16 scales as @xmath195 , whereas it scales as @xmath196 if the relevant field scale is @xmath86 . since @xmath187 vanishes much faster than @xmath10 , eq .
( [ eq:3.17 ] ) , the field scaling region will be restricted to larger values of @xmath197 , and @xmath187 will be given by eq .
( [ eq:3.28 ] ) in a substantial range of @xmath90-values . will come back to this in sec .
[ sec : iv ] .
to summarize , we have determined the electrodynamic properties of superconductors close to a ferromagnetic instability , i.e. , materials that , in the absence of superconductivity , would be paramagnetic with large ferromagnetic fluctuations .
this work complements previous studies of the coexistence of superconductivity with ferromagnetic order.@xcite we have treated the superconductivity in mean - field ( ginzburg - landau ) approximation . in addition , we have employed the london approximation , treating the superconducting order parameter as a constant .
the ferromagnetic critical point we have treated explicitly in a mean - field approximation , and we have used scaling arguments to consider the consequences of the exact magnetic critical behavior for the superconductivity . we have found that the thermodynamical critical field @xmath10 decreases due to the ferromagnetic fluctuations , as one would expect , and depends on the magnetic critical exponent @xmath12 , see eqs . ( [ eq:3.17 ] ) and ( [ eq:2.21 ] ) . however , the london penetration depth also decreases , which is intuitively less obvious . at magnetic criticality
the behavior of the magnetic induction at a vacuum - to - superconductor ( or normal metal - to - superconductor ) interface is still characterized by exponential decay , but the characteristic length scale @xmath11 is different from the usual london penetration depth @xmath81 . within a mean - field description of the magnetic criticality
it is the geometric mean of the zero - temperature magnetic correlation length and @xmath81 , see eqs .
( [ eq:3.22 ] ) and ( [ eq:3.26 ] ) ; more generally , it depends on the magnetic critical exponent @xmath13 , see eq .
( [ eq:2.22 ] ) . however , this behavior of the penetration depth is valid only within an extremely small region of width @xmath198 around magnetic criticality . outside of this region , but
still within the ferromagnetic critical region , the temperature dependence of the penetration depth is the same as in ginzburg - landau theory , see eq .
( [ eq:3.27 ] ) . for the critical current @xmath14
this implies a dependence on the reduced temperature given by @xmath199 , where the exponent @xmath15 depends on both @xmath12 and @xmath13 , or on @xmath12 only , depeding on the value of @xmath91 , see eqs .
( [ eqs:2.23 ] ) . with exponent values
appropriate for the usual ferromagnetic universality classes , @xmath200 extremely close to magnetic criticality , and @xmath201 somewhat farther away .
for the temperature dependencies of various observables at magnetic criticality we have assumed that the system stays tuned to magnetic criticality while the temperature is varied .
let us discuss to what extent this assumption is realistic .
consider a phase diagram in a plane spanned by the temperature and some non - thermal control parameter @xmath202 , e.g. , the hole doping concentration in the case of mgcni@xmath2,@xcite and consider the following two qualitatively different possibilities .
figure [ fig:2 ] shows a situation where the magnetic phase separation line does not cross the line @xmath203 .
the stoichiometric compound thus does not enter a magnetic phase upon cooling , although the system is close to a magnetic transition for all temperatures below the superconducting @xmath151 .
this scenario is believed to apply to mgcni@xmath2 .
-0 mm ) - control parameter ( @xmath202 plane .
the solid line denotes the superconducting transition , the dashed line , the magnetic one . along @xmath203
there is only one phase transition at the superconducting @xmath151 .
see the text for additional explanation.,title="fig:",width=264 ] figure [ fig:3 ] shows a situation where the magnetic phase separation line does cross the line @xmath203 , so that the stoichiometric compound enters a phase where superconductivity and magnetism coexist at some temperature below @xmath151 .
this is the situation that was discussed in refs . and and observed in errh@xmath204b@xmath204 and homo@xmath205s@xmath206.@xcite -0 mm , but with a magnetic transition for @xmath203 at a temperature @xmath207 . on the @xmath203 axis
it is shown that @xmath208 splits into the bare magnetic transition temperature @xmath209 and the physical transition temperature @xmath210 to a state with spiral magnetic order , ref .. see the text for additional explanation.,title="fig:",width=264 ] the magnetic transition is to a phase with spiral magnetic order at a temperature @xmath210 slightly below the temperature @xmath209 where ferromagnetism would occur in the absence of superconductivity.@xcite we now can see what is required to keep @xmath91 constant while varying @xmath90 , namely , a situation as shown in fig .
[ fig:2 ] with the dashed line essentially parallel to the @xmath42-axis .
@xmath91 is then given by the dimensionless distance between the two lines . in order for the penetration depth to display the non - ginzburg - landau behavior described by eq .
( [ eq:3.26 ] ) or , more generally , eq . ( [ eq:2.22 ] ) , the two lines would have to be extremely close , in order to keep @xmath91 smaller than @xmath198 , see eq .
( [ eq:3.26 ] ) .
this would result in a temperature dependence of the critical current given by eqs .
( [ eq:2.23a ] , [ eq:2.23b ] ) . while this is possible , it is a very non - generic situation , and it would result in a very large magnetic susceptibility of the normal metal just above the superconducting transition temperature . a situation that is still very non - generic , but requires somewhat less fine - tuning , is one where the dashed line is still essentially parallel to the @xmath42-axis , but in a somewhat larger @xmath91-range , say , with @xmath91 on the order of a few percent . in this case
the penetration depth will show the usual @xmath211 temperature dependence , see eq .
( [ eq:3.27 ] ) .
the temperature dependence of the thermodynamic critical field will be more complicated in this case .
the generalization of eq .
( [ eq:2.20 ] ) to nonzero values of @xmath91 is = r^-f_(h / r^ ) , [ eq:4.1 ] with @xmath212 , @xmath213 , and @xmath12 the usual critical exponents for the magnetic transition . in order for eq .
( [ eq:2.20 ] ) to hold , the @xmath24 must be large compared to @xmath214 in suitable units .
the latter are not determined by any universal arguments , but an analysis of the critical equation of state for both the high - temperature ferromagnet ni ( @xmath215)@xcite and the low - temperature ferromagnet crbr@xmath2 ( @xmath216)@xcite shows that in either case the relevant energy or field scale ( we use units such that @xmath217 ) is given by @xmath208 , which is plausible .
the crossover between the field scaling that leads to eq .
( [ eq:2.21 ] ) and the static scaling that leads to eq .
( [ eq:2.5 ] ) thus occurs at a crossover field h _ t_^0r^. [ eq:4.2 ] @xmath218 for ferromagnetic phase transitions , and with @xmath219 and @xmath220 , one finds @xmath221 .
for mgcni@xmath2 in the vicinity of @xmath151 , this leads to @xmath222 . with @xmath223 at _ zero temperature _ on the order of @xmath224 and @xmath225,@xcite one expects @xmath226 . since @xmath10 vanishes at @xmath151 , this means that @xmath10 will be given by eq .
( [ eq:2.21 ] ) sufficiently far away from @xmath151 , but cross over to @xmath227 near @xmath151 .
consequently , the critical current exponent @xmath15 will be given by eq .
( [ eq:2.23c ] ) at some distance from @xmath151 , and cross over to the ginzburg - landau result @xmath228 as @xmath150 . in the experiment of ref .
, no such crossover was observed down to @xmath229 . at least within the london approximation
, our results confirm the conclusion of ref . that superconductors near a ferromagnetic instability are necessarily of type i. however
, we have also shown that this conclusion is inevitable only within an extremely small region around the ( bare ) magnetic critical point .
the fact that mgcni@xmath2 is observed to be of type ii@xcite is therefore not necessarily in contradiction to the notion that this material is almost ferromagnetic .
however , eq . ( [ eq:3.29 ] ) predicts a strong deviation from ginzburg - landau behavior for the upper critical field @xmath187 .
since @xmath187 goes to zero rapidly as @xmath150 , this behavior will show only at substantial values of @xmath197 even if @xmath91 is very small .
no anomalous behavior was observed for @xmath197 up to @xmath230.@xcite this is reconcilable with close proximity to a magnetic instability only if @xmath91 is very small close to @xmath151 , and grows with decreasing temperature , in which case @xmath187 might never show the magnetic critical behavior .
a signature of this situation would be a large magnetic susceptibility in the normal state just above @xmath151 .
the conclusion from this discussion with respect to the experimental observations in ref .
is as follows . while it is possible that proximity to a ferromagnetic instability is the cause of the observed anomalous behavior of the critical current , such an explanation requires fine tuning of the phase diagram , and would have to be accompanied by a very large enhancement of the spin susceptibility in the normal phase just above @xmath151 . explaining the lack of an anomaly in the temperature dependence of @xmath187
probably requires that the material is closer to the magnetic instability near @xmath151 than at @xmath231 ( i.e. , the dashed line in fig.[fig:2 ] comes closer to the @xmath42-axis with increasing @xmath42 ) .
a direct measurement of the spin susceptibility in the normal phase would be of great interest in this context .
finally , we discuss our predictions for the case of a superconductor that does undergo a transition to a magnetic state below @xmath151 , i.e. , the situation represented by fig .
[ fig:3 ] . in the ( very small ) temperature interval of width @xmath232 around @xmath209 , both the thermodynamic critical field @xmath10 and the penetration depth
@xmath11 will show an anomalous temperature dependence , and the critical current exponent will be given by eq .
( [ eq:2.23b ] ) .
outside of this region , but not too close to @xmath151 , @xmath10 will be anomalous , but @xmath11 will be conventional , and the critical current exponent will be given by eq .
( [ eq:2.23c ] ) . upon approaching @xmath151
, @xmath10 will fall below the crossover field given by eq.([eq:4.2 ] ) , and its temperature dependence will cross over to the usual linear ginzburg - landau behavior .
the critical current exponent close to @xmath151 will thus be the conventional @xmath100 .
the location of this crossover depends on the critical field scale , and will thus be material dependent .
critical current measurements in the materials like errh@xmath204b@xmath204 , or homo@xmath205s@xmath206 , which are believed to fall into this class , would be very interesting .
we thank phil adams for alerting us to his results on mgcni@xmath2 , for sharing his data prior to publication , and for helpful correspondence .
this work was supported by the nsf under grant nos .
dmr-05 - 29966 , and dmr-05 - 30314 . | the electromagnetic properties of superconductors near a ferromagnetic instability are investigated by means of a generalized ginzburg - landau theory .
it is found that the magnetic flux expulsion capability of the superconductor gets stronger , in a well - defined sense , as the normal - state magnetic susceptibility increases .
the temperature dependencies of the london penetration depth , the critical fields , and the critical current are all strongly affected by ferromagnetic fluctuations .
in particular , for the critical current we find a temperature exponent @xmath0 over an appreciable temperature range .
the extent to which proximity to magnetic criticality may be a viable explanation for recent observations in mgcni microfibers , which find @xmath0 , is discussed . |
cooling flows in galaxy clusters , once thought to be on the order of 100 1000 m@xmath2 yr@xmath3 , are now understood to be depositing cool gas on the order of 1 10 m@xmath2 yr@xmath3 onto the central cluster galaxy ( hereafter abbreviated to ccg ; voigt @xmath4 fabian 2004 ) .
it is now generally accepted that some form of feedback ( agn , conduction etc . )
prevents the intracluster medium from cooling , allowing only small amounts of cool gas to accrete onto the ccg .
thus , ccgs lie at the interface where it is crucial to understand the role of feedback and accretion in star formation . within these cooling - flow ccgs , cool
molecular clouds , warm ionized hydrogen , and the cooling intracluster medium are related . a complete view of the star formation process incorporates the stars with the gas and an understanding of the processes by which these phases interact , and therefore , requires information from several wavelength regimes . various previous studies have reported several examples of ongoing star formation in ccgs , in particular those hosted by cooling - flow clusters ( cardiel , gorgas @xmath4 aragn - salamanca 1998 ; crawford et al . 1999 ; mcnamara et al .
2006 ; edwards et al . 2007 ; odea et al .
2008 ; bildfell et al .
2008 ; pipino et al .
2009 ; loubser et al . 2009 ; odea et al .
2010 ; liu , mao @xmath4 meng 2012 ) .
although active star formation in these central galaxies is compelling , the young populations only contribute a very small mass fraction ( pipino et al .
2009 ) , and it has been shown that star formation in these ccgs is correlated with the cooling time of the gas ( rafferty , mcnamara @xmath4 nulsen 2008 ; liu et al . 2012 ) .
recent studies have even claimed that star forming central galaxies are exclusively hosted by cooling flow clusters ( hoffer et al .
2012 , and references therein ) , although not all cooling - flow clusters contain star forming galaxies .
the empirical boundary between clusters that host active central galaxies and clusters that never host them is @xmath5 kev @xmath6 , an entropy corresponding to an intracluster medium cooling time of @xmath7 gyr ( voit et al .
the origin of the gas fuelling this star formation is not yet fully understood .
probable explanations include processes involving the cooling flows , but cold gas deposited during a merging event can not be conclusively eliminated ( bildfell et al .
2008 ) . with the advent of more accurate methods to fit complex star formation histories ( sfh ) from high - quality optical spectroscopy , we are now in a position to directly compare the light / mass fraction of any possible present young stellar population component to the cooling rate of the host cluster .
we proposed and obtained long - slit observations of ccgs with confirmed h@xmath0-filaments .
we selected galaxies with near - ir ( 2mass ) , ultraviolet ( galex ) , x - ray data ( chandra ) , and vla 1.4 ghz fluxes , already available ( mcdonald et al .
detailed properties of the host clusters , which are reported to influence the activity in the central galaxy ( as described above ) , such as central cooling times and the offset between the cluster x - ray peak and the central galaxy , have been derived in previous studies and are available in the literature .
this data is complimentary to the emission - line , long - slit spectroscopy ( on keck and magellan ) along the h@xmath0 filaments of nine ccgs ( not including the four studied here ) by mcdonald et al .
to get maximum possible information , we observed line ratios over a very long wavelength range ( around h@xmath0 and h@xmath8 ) with the slit aligned on , or as close as possible , to the major axis of the galaxy .
we can now place the derived information from the stellar population analysis of the optical spectra in context with multiwavelength data over the full spectrum , to explain the diverse nature of these galaxies .
we introduce the sample and detail of the data reductions in sections [ sample ] and [ reduction ] .
we then derive the kinematics as well as the stellar populations in section [ extinction ] .
we proceed to discuss the four individual cases in section [ figures_individual ] .
we summarise the findings of this paper in section [ summary ] .
we have used the following set of cosmological parameters : @xmath9 = 0.3 , @xmath10 = 0.7 , h@xmath11 = 70 km s@xmath3 mpc@xmath3 .
we have chosen our sample of ccgs from the h@xmath0 imaging presented in mcdonald et al .
( 2010 ) , who in turn , selected their sample from white , jones @xmath4 forman ( 1997 ) .
mcdonald et al .
( 2010 ) enforced the cuts : @xmath12 and @xmath13 , after which they selected 23 clusters to cover the full range of properties , from very rich clusters with high cooling rates to low - density clusters with very small cooling flows . from their 23 clusters , we selected all the clusters with clearly detected h@xmath0 in their centres ( albeit filamentary , extended or nuclear emission ) .
in addition , all of these central galaxies have optical imaging , near - ir ( 2mass ) and uv ( galex data ) available .
thereafter , we selected all the central galaxies with detailed x - ray ( chandra ) data , as well as vla 1.4 ghz fluxes , available .
this resulted in a sub - sample of 10 galaxies .
we observed four of these galaxies with the gmos ifu ( as presented in loubser @xmath4 soechting 2013 ) , and we observed four of these galaxies with the salt rss ( as shown in figure [ fig : thumbnails ] ) .
this does not constitute a complete sample as we merely chose the objects with the most auxiliary information available . [
cols="<,^,^,^,^,^,^,^ " , ]
we have obtained , analysed and interpreted detailed , high spatial and spectral resolution , long - slit observations of four ccgs ( abell 0085 , 0133 , 0644 and ophiuchus ) recently obtained on salt , to probe the spatially - resolved kinematics and stellar populations of the stars .
we use the pegase.hr model with the elodie v3.1 stellar library to determine the sfhs of the galaxies using the full spectrum fitting .
we perform ssp as well as csp fits to acount for more complex sfhs .
monte - carlo simulations and @xmath1-maps are used to check the reliability of the solutions .
we emphasize that we only interpret the presence of young stellar components which is much more reliable than interpreting the absolute ages , metallicities and @xmath0-enhancements .
hudson et al . (
2010 ) use the largest complete sample of 64 galaxy clusters ( highest x - ray flux galaxy cluster sample ) with available high - quality x - ray data from chandra , and apply 16 cool - core diagnostics to them . to segregate cool core and non - cool - core clusters , they find that central cooling time is the best parameter for low redshift clusters with high quality data .
they further show that the discrepancy in classical and spectroscopic mass deposition rates can not be explained with a recent formation of the cool cores , demonstrating the need for a heating mechanism to explain the cooling flow problem .
only a few ccgs have been reported in the literature with ongoing or recent bursts of star formation , and the majority if not all of them are hosted in cooling flow clusters .
rafferty et al .
( 2008 ) , cavagnolo et al .
( 2008 , 2009 ) and hoffer et al .
( 2012 ) show that central galaxies in clusters with low entropy ( @xmath14 ) thus with large cooling rates are the only ccgs to exhibit signs of vigorous star formation . here
, we detect very small , but observable young stellar components in two of our four clusters which have no mass deposition rates ( or very little so that it is below the x - ray detection limit ) .
this suggests another possible mechanism for the recent star formation such as galactic cannibalism that appears in cluster environments due to dynamical friction .
we note that the spatially resolved recession velocity of these two galaxies ( 2masxj17122774 - 2322108 and eso541 - 013 ) show measurable rotation ( see figure [ fig : velocity ] ) , although the sample is too small to infer a possible difference between the kinematics of these galaxies and the galaxy with a younger component hosted by a cooling flow cluster .
abell 0133 ( host of eso541 - 013 ) showed extended h - alpha filaments in mcdonald et al .
( 2010 ) , whereas the other three clusters examined here showed nuclear h - alpha emission . therefore no obvious correlation between the h - alpha gas and stellar kinematics exists .
mergers that are not dissipation - less can supply fresh cold gas that can trigger an episode of star formation ( mihos @xmath4 hernquist 1996 ) . a large library of smooth particle hydrodynamic simulations of galaxy mergers ( di matteo et al .
2007 , 2008 ) show that the average enhancement of the star formation rate in a random galaxy collision is only a factor of a few ( 3 - 4 being the median factor ) and only lasts 200 - 400 myr .
this process is more efficient in dense clusters but the star formation rate increase remains in general below a factor of 10 or less ( martig @xmath4 bournaud 2008 ) . since we have carefully measured the possible presence of emission lines ( with particular attention to the balmer lines ) , and could nt find any detections above three sigma of the noise of the spectra in any of the four objects - we conclude that our slit placement did not intercept the h@xmath0 filaments .
thus in our data , the relatively young stars do not necessarily co - exist with the h@xmath0 emitting gas .
this study , focussing on the detailed analysis of four objects , is complementary to our previous studies investigating the evolution and environmental influence in a larger sample of central cluster galaxies ( loubser et al .
2008 , loubser @xmath4 sanchez - blazquez 2012 ) . throughout the investigation and analysis of the larger sample ,
a picture emerged indicating that the kinematics and stellar population properties of the central galaxies are not very dependent on the galaxy mass ( with the exception of the metallicity gradients which are dependent on the galaxy velocity dispersions ) .
similarly , various cluster properties also seem to have very limited influence on the central galaxy stellar population properties , such as cluster temperature or density . on the other hand , the offset between the x - ray peak of the cluster and the galaxy and the cooling flows in the clusters do influence the probability that the central galaxy will form stars . in loubser
et al . (
2008 ) , we found that one of the six younger galaxies were hosted by a non - cooling flow cluster whereas the other five younger galaxies were hosted by a cooling - flow cluster .
this previous study analysed only the ssp - equivalent ages , but reinforces the findings of the current study .
all of these observations point to very diverse star formation histories of these central galaxies , and excludes the naive picture of very passive , homogeneous evolution .
sil is financially supported by the south african national research foundation .
sil gratefully acknowledges the constructive comments from the anonymous reviewer as well as the help of alet de witt ( hartebeesthoek radio observatory ) with the radio data interpretation .
all of the observations reported in this paper were obtained with the southern african large telescope ( salt ) under program 2011 - 2-rsa@xmath15oth-003 and 2012 - 1-rsa@xmath15oth-003 ( pi : ilani loubser ) .
99 bertelli g. , bressan a. , chiosi c. , fagotto f. , nasi e. , 1994 , a@xmath4as , 106 , 275 bildfell c. , hoekstra h. , babul a. , mahdavi a. , 2008 , mnras , 389 , 1637 brzan l. , rafferty d.a . , mcnamara b.r . ,
wise m.w . ,
nulsen p.e.j . , 2004 ,
apj , 607 , 800 burgh e.b . ,
nordsieck k.h . ,
kobulnicky h.a .
, williams t.b .
, odonoghue d. , smith m.p . , percival j.w .
, 2003 , spie , 4841 , 1463 bouchard a. , prugniel p. , koleva m. , sharina m. , 2010 , a@xmath4a , 513a , 45 cappellari m. , emsellem e. , 2004 , pasp , 116 , 138 cardiel n. , gorgas j. , aragn - salamanca a. , 1998 , mnras , 298 , 977 cavagnolo k.w . , donahue m. , voit g.m . , sun m. , 2008 , apj , 683 , 107 cavagnolo k.w . , donahue m. , voit g.m . , sun m. , 2009 , apjs , 182 , 12 condon j.j .
, cotton w.d . ,
greissen e.w . ,
, perley r.a .
, taylor g.b . ,
broderick j.j . , 1998 , 115 ,
1693 conroy c. , wechsler r.h .
, kravtsov a.v . , 2007 , apj , 668 , 826 crawford c.s . ,
allen s.w .
, ebeling h. , edge a.c .
, fabian a.c
. , 1999 , mnras , 306 , 857 crawford s.m . ,
2010 , spie , 7737e , 54 de lucia g. , blaizot j. , 2007 , mnras , 375 , 2 di matteo p. , combes f. , melchior a .-
, semelin b. , 2007 , a@xmath4a , 468 , 61 di matteo p. , bournaud f. , martig m. , combes f. , melchior a .-
, semelin b. , 2008 , a@xmath4a , 492 , 31 du w. , luo a.l . , prugniel ph .
, liang y.c . ,
zhao y.h . , 2010 , mnras , 409 , 567 dubinski j. , 1998 , apj , 502 , 141 edwards l.o.v . , hudson m.j . , balogh m.l .
, smith r.j . , 2007 , mnras , 379 , 100 fisher d. , illingworth g. , franx m. , 1995 , apj , 438 , 539 hicks a.k . , mushotzky r. , donahue m. , 2010 , apj , 719 , 1844 hoffer a.s . , donahue m. , hicks a. , barthelemy r.s . , 2012 , apjs , 199 , 23 hudson d.s . , mittal r. , reiprich t.h . , nulsen p.e.j .
, andernach h. , sarazin c.l .
, 2010 , a@xmath4a , 513 , 37 jarrett t.h .
, chester t. , cutri r. , schneider s.e .
, huchra j.p . , 2003 , aj , 125 , 525 kawata d. , mulchaey j.s . , gibson b.k . , snchez - blzquez p. , 2006 ,
apj , 648 , 969 kennicutt r.c . , 1998 , apj , 498 , 541 kobulnicky h.a . ,
nordsieck k.h . ,
burgh e.b . ,
smith m.p .
, percival j.w . , williams t.b .
, odonoghue d. , 2003 , spie , 4841 , 1634 koleva m. , prugniel ph . ,
ocvirk p. , le borgne d. , soubiran c. , 2008 , mnras , 385 , 1998 koleva m. , prugniel ph .
, bouchard a. , wu y. , 2009 , a@xmath4a , 501 , 1269 liu f.s . , mao s. , meng x.m . , 2012 , mnras , 423 , 422 loubser s.i . ,
sansom a.e .
, snchez - blzquez p. , soechting i.k .
, bromage g. , 2008 , mnras , 391 , 1009 loubser s.i . , snchez - blzquez p. , sansom a.e . , soechting i.k . , 2009 , mnras , 398 , 133 loubser s.i .
, snchez - blzquez p. , 2012 , 425 , 841 loubser s.i .
, soechting i.k . , 2013 , mnras , 431 , 2933 martig m. , bournaud f. , 2008 , mnras , 385 , l38 mccarthy i.g . , balogh m.l .
, babul a. , poole g.b .
, horner d.j . , 2004 , apj , 613 , 811 mcdonald m. , veilleux s. , rupke d.s.n .
, mushotzky r. , 2010 , apj , 721 , 1262 mcdonald m. , veilleux s. , rupke d.s.n . , 2012 , apj , 746 , 153 mcnamara b.r .
, 2006 , apj , 648 , 164 mihos j. , hernquist l. , 1996 , apj , 464 , 641 murgia m. , eckert d. , govoni f. , ferrari c. , pandey - pommier m. , nevalainen j. , paltani s. , 2010 , a@xmath4a , 514 , 76 odea c.p .
et al . , 2008 ,
apj , 681 , 1035 odea k.p .
et al . , 2010 , apj , 719 , 1619 perez - torres m.a . ,
zandanel f. , guerrero m.a . , pal s. , profumo s. , prada f. , panessa f. , 2009 , mnras , 396 , 2237 pipino a. , kaviraj s. , bildfell c. , hoekstra h. , babul a. , silk j. , 2009 , mnras , 395 , 462 prugniel ph . , soubiran c. , koleva m. , le borgne d. , 2007 , arxiv : astro - ph/0703658 rafferty d.a . , mcnamara b.r .
, nulsen p.e.j . , 2008 ,
apj , 687 , 899 salpeter e.e .
, 1955 , apj , 121 , 161 snchez - blzquez p. et al . , 2006 , mnras , 371 , 703 snchez - blzquez p. et al .
, 2009 , a@xmath4a , 499 , 47 sarzi m. et al . , 2006 , mnras , 366 , 1151 van dokkum p.g . , 2001 ,
pasp , 113 , 1420 viljoen d. , loubser s.i . , 2013 , mnras , submitted voigt l.m .
, fabian a.c . , 2004 , mnras , 347 , 1130 voit g.m . , cavagnolo k.w . ,
donahue m.,rafferty d.a . ,
mcnamara b.r . ,
nulsen p.e.j . , 2008 ,
apj , 681 , 5 white d.a . ,
jones c. , forman w. , 1997 , mnras , 292 , 419
the spatially resolved kinematic data for all 4 ccgs are available as online - data . | we present detailed , high spatial and spectral resolution , long - slit observations of four central cluster galaxies ( abell 0085 , 0133 , 0644 and ophiuchus ) recently obtained on the southern african large telescope ( salt ) .
our sample consists of central cluster galaxies ( ccgs ) with previously - observed h@xmath0 filaments , and have existing data from the x - ray to radio wavelength regimes available . here , we present the detailed optical data over a broad wavelength range to probe the spatially - resolved kinematics and stellar populations of the stars .
we use the pegase.hr model with the elodie v3.1 stellar library to determine the star formation histories of the galaxies using full spectrum fitting .
we perform single stellar population ( ssp ) as well as composite stellar population ( csp ) fits to account for more complex star formation histories .
monte - carlo simulations and @xmath1-maps are used to check the reliability of the solutions .
this , combined with the other multiwavelength data , will form a complete view of the different phases ( hot and cold gas and stars ) and how they interact in the processes of star formation and feedback detected in central galaxies in cooling flow clusters , as well as the influence of the host cluster .
we find small , young stellar components in at least three of the four galaxies , even though two of the three host clusters have zero spectrally - derived mass deposition rates from x - ray observations .
[ firstpage ] galaxies : formation galaxies : elliptical and lenticular , cd galaxies : clusters : individual : abell 0085 galaxies : clusters : individual : abell 0133 galaxies : clusters : individual : abell 0644 galaxies : clusters : individual : ophiuchus |
suppression of radio frequency interference ( rfi ) with adaptive beamforming is widely used in radio astronomy , radar and telecommunications . the main idea behind many algorithms proposed for use in radio astronomy consists of weighting the outputs of array elements in such a way as to create zero values in the synthesized array pattern in the direction of rfi and to keep the signal of interest ( soi ) , the radio source to be observed , in the maximum of the main lobe without significant loss of gain @xcite . during recent years
there has been a growing interest in radio astronomy for applying these methods of rfi mitigation both to existing radio telescopes and to future generation projects .
there are several specific features of the large connected radio interferometers ( ri ) used in radio astronomy such as westerbork synthesis radio telescope ( wsrt ) , very large array ( vla ) and giant metrewave radio telescope ( gmrt ) which make the straightforward application of this adaptive beam - forming different and difficult when compared to classic phased arrays : \1 .
connected ri are highly sparse arrays .
their main mode of operation is correlation processing .
direction of arrival ( doa ) of a signal of interest is a known and time - dependent vector .
there is no computer - controlled amplitude weighting in the existing ri backend hardware .
there is an auxiliary _ tied - array _ facility which is used during vlbi and pulsar observations .
the mode of observation is similar to that of standard phase arrays : the signals from the antennas are added but without amplitude weighting , because the antennas of ri are identical .
there is a phase - only control allowing coherent adding .
the ri works as a single dish " .
noise - like radio source signals are usually much weaker than system noise ( antenna + receiver ) and rfi .
phase - only adaptive nulling is proposed for rfi mitigation during tied - array observations .
small phase perturbations are introduced to the signals of every antenna .
the values of these perturbations are optimized in such a way that the signal from soi is preserved and the rfi signals suppressed .
this techniques has been widely discussed @xcite and is well suited to tied - array observations .
there are two approaches to adaptive beam - forming : narrow - band ( complex weighting of amplitudes and phases ) and wide - band ( digital filtering , delay - tap weighting ) . the narrow - band approach will be used in the following text .
let us consider an equidistant m - element linear array .
the @xmath0-dimensional array output vector @xmath1 , as a function of an angle , i. e. , the complex amplitude of the temporal signal @xmath2 consists of the following components : @xmath3 where @xmath4 is the signal vector corresponding to the plane wave coming from the direction @xmath5 is the @xmath6th rfi vector , coming from any direction @xmath7 is the system noise vector .
these three components are uncorrelated .
vector @xmath4 depends on the incidence angle @xmath8 of the plane wave , measured with respect to the normal to linear array @xmath9^{t}\ ] ] where phase shift @xmath10 is the spacing between array elements , @xmath11 is the wavelength . the phase of the first antenna is chosen to be equal to 0 .
the beamformer , in general , consists of the complex weights @xmath12 which form the beamformer vector @xmath13 @xmath14^{t}.\ ] ] the output of the phased array is @xmath15 the beamformer should satisfy both following requirements : \a ) _ steering capability _ : the soi is protected ( @xmath16 , for a prescribed direction @xmath17 the response of the array is constant regardless of what values are assigned to the weights @xmath13 ; \b ) the effects of rfi should be minimized .
the _ minimum - variance distortionless response _ ( mvdr ) beamforming algorithm , subject to this constraint when @xmath18 is proposed in order to minimize the variance of the beamformer output@xcite .
the solution for @xmath13 in this case is @xmath19^{-1},\ ] ] where @xmath20 is the correlation matrix of @xmath21 as was mentioned in the introduction , wsrt and other large radio astronomy arrays can not use this algorithm in real time with the existing equipment because there are no amplitude control facilities .
that is to say that `` rfi nulling '' is limited to _ phase - only _ control .
phase - only weights can be found to be the solution to the following system of nonlinear equations : @xmath22 @xmath23 are the weights in our phase - only case .
the vector @xmath24 is known and is determined by the soi coordinates .
the construction of the @xmath25 vector requires the knowledge of rfi s doa , which may be known beforehand or could be obtained from the observed correlation matrix @xmath26 , because soi is always much weaker than rfi .
but it is necessary to have a special correlator for this purpose in order to follow all rapid scintillations of rfi which are usually averaged by the main radio interferometer correlator .
so , in principle , the system of equations ( 6 - 9 ) can be solved and the phase corrections @xmath27 introduced into the phase control system . the optimal solution ( 5 ) can be used as a zero approximation for @xmath28 the difficulties in implementing the solution of the system ( 6 - 9 ) are not mentioned here .
a more practical method of calculating the phase corrections @xmath27 in the tied array case is proposed here .
the tied array total power detector output ( @xmath29 ) is @xmath30^{2}dt>=tpd_{sig}+tpd_{rfi}+tpd_{n},\ ] ] where the total power components @xmath31 correspond to the soi , rfi and system noise , respectively , @xmath32 means the statistical expectation .
the mean value of the signal component is @xmath33dt\right\}^{2 } > , \ ] ] the mean value of the n - th rfi component is @xmath34dt\right\}^{2 } > , \ ] ] the mean value of the system noise component is constant .
we assume also that @xmath35 and the different @xmath36 are uncorrelated . considering the tpd output as a function of @xmath0 variables
@xmath37 the following criterium for a `` good '' vector @xmath38^{t}$ ] can be proposed : @xmath39 the denominator @xmath40 is the total tpd output under the assumption @xmath41 the numerator @xmath42 can be calculated for each given @xmath43 and @xmath8 ( the doa of soi ) . therefore ,
maximizing @xmath44 with a proper choice of @xmath43 , a higher signal - to - rfi - plus - noise ratio at the tied array output can be achieved .
this is a classic @xmath0-variable optimization problem which is difficult to solve by the common gradient methods because of the multimodality of @xmath45 there are many local ( secondary ) maximums and a searching algorithm will `` get stuck '' at one of them without finding the global maximum .
genetic algorithms ( ga ) search the solution for the set of variables through the use of simulated evolution , i.e. , _ survival of the fittest _ strategy .
in contrast to the gradient algorithms which are , in general , calculus - based algorithms , ga , first introduced by @xcite , exploits a _ guided random techniques _ during optimization procedure @xcite .
the multimodality problem is successfully overcome by this algorithm .
a simplified block diagram of ga implementation in a radio interferometer is depicted in figure 1 .
a phase control subsystem introduces a certain initial phase distribution @xmath46 corresponding to radio source coordinates and preliminary phase calibration corrections .
the output of the tpd is then continiously measured and used to supply the ga program with the data ( cost function samples ) which monitor the performance of the tied array with respect to rfi .
the ga uses these data to calculate new phases @xmath47 with the aim to maximize @xmath48 these new phases are introduced into the phase control subsystem after each iteration and a new value of the tpd output signal is used for the next step .
thus the feedback loop , _ phase control subsystem - tpd - ga _ , maintains the low value of @xmath49 and therefore the high value of the fitness function @xmath44 , i. e. , the high signal - to - rfi ratio .
computer simulation was performed to illustrate the effectiveness of the phase - only nulling in rfi mitigation .
first , a 14-element half - wavelength linear array was modelled .
the soi direction is equal to @xmath50 and there are two rfi signals : one at the angle @xmath51 , and the other at the angle @xmath52 figure 2 shows in logarithmic scale the quiescent ( dash line ) and adapted ( solid line ) array patterns .
the significant suppression of rfi with the adapted pattern is clearly visible , while the quiescent pattern has the secondary lobe maximums at the rfi positions .
figure 3 shows the corresponding array phase distribution .
sparse 14-element array patterns are represented in figures 4 , 5 , 6 and 7 .
the distance between the elements is @xmath53 and the central frequency is @xmath54 .
the lobes are much narrower than for the half - wavelength array , so the different figures are given to illustrate the result of phase - only nulling .
figure 4 ( a linear scale presentation of the pattern ) is given here to illustrate the loss and distortions of the main lobe .
this is more visible in the linear scale , whereas rfi suppression is better seen in the logarithmic scale .
5 - the quiescent and adapted patterns around angle @xmath50 logarithmic scale , figures 6 , 7 - the same patterns around the directions of rfi1 ( doa@xmath55 and rfi2 ( doa@xmath56 the corresponding array phase distribution is shown in fig .
the subsequent figures illustrate this phase - only nulling for a two - dimensional planar array .
adaptive nulling was simulated for the half - wavelength array with 10x10 elements , the central frequency is equal to @xmath54 .
rectangular coordinates @xmath57 and @xmath58 are angles measured from the @xmath59 and @xmath60 axes , respectively , to the line from the array to the radio source ; thus the soi is at the zenith , with coordinates @xmath61 coordinates of rfi were chosen so as to put them on the maximums of the secondary lobes , the values of rfi suppression are shown in the captions .
the following sequence of figures is given : figure 9 : normalized ( a(@xmath62=1 ) quiescent array with indicated rfi positions ; figure 10 : array s pattern after adaptation ; figure 11 : this array s phase distribution after adaptation .
\1 . existing large radio interferometers ( wsrt , vla , gmrt ) have only phase control facilities and the _ real - time _ adaptive nulling in the rfi direction should take this constraint into account .
capon , j. ( 1969 ) , high - resolution frequency - wavebumber spectrum analysis , _ proc . of the ieee _
, 57 , 14081418 .
davis , r. m. ( 1998 ) , phase - only lms and perturbation adaptive algorithms , _ ieee trans on aerosp . and electr .
_ , 34 , 169178 .
giusto , r. , and de vincenti , p. ( 1983 ) , phase - only optimization for generation of wide deterministic nulls in the radiation patern of phased arrays , _ ieee trans . on aerosp . and
electr . syst .
_ , 31 , 814817 .
haupt , r. l. ( 1995 ) , an introduction to genetic algorithms for electromagnetics , _ ieee ant . and propag .
mag , 37 , _ 715 .
haupt r. l. ( 1997 ) , phase - only adaptive nulling with a genetic algorithm , _ ieee trans .
ant . and propag . , ap-45 , _ 10091015 .
gabriel w. f. ( 1992 ) , adaptive processing array systems , _ proc .
ieee _ , 80 , 152162 .
goldberg , d. e. ( 1989 ) , _ genetic algorithms in search , optimization and machine learning _ , addison - wesley publishing company , inc . , ny .
holland , j.h .
( 1975 ) , _ adaptation in natural and artificial systems _
university of michigan press , ann arbor ; 2nd ed . : 1992 , mit press , cambridge .
krim , h. and viberg , m. ( 1996 ) two decade of array signal processing research , _ ieee sig .
proc . mag .
_ , 6794 , july .
leavitt , m. k. ( 1976 ) , a phase adaptation algorithm , _ ieee trans . on ant . and
_ , ap-24 , 754756 .
michalewicz , z. ( 1992 ) , _ genetic algorithms + data structure = evolution programs _ , springer .
smith , s. t. ( 1999 ) , optimum phase - only adaptive nulling , _ ieee trans .
_ 47 , 1835 -1843 .
steyskal , h. ( 1983 ) , simple method for pattern nulling by phase perturbation , _ ieee trans . on ant . and propag . ,
31 , 163166 .
thompson , p.a .
( 1976 ) , adaptation by direct phase - shift adjustment in narrow - band adaptive antenna systems , _ ieee trans . on ant . and
_ , 24 , 756760 .
widrow , b and stearns , s. ( 1985 ) adaptive signal processing , prentice - hall , inc .
, englewood cliffs , n. j. | connected radio interferometers are sometimes used in the tied - array mode : signals from antenna elements are coherently added and the sum signal applied to a vlbi backend or pulsar processing machine . usually there is no computer - controlled amplitude weighting in the existing radio interferometer facilities . radio frequency interference ( rfi ) mitigation with phase - only adaptive beamforming
is proposed for this mode of observation .
small phase perturbations are introduced in each of the antenna s signal .
the values of these perturbations are optimized in such a way that the signal from a radio source of interest is preserved and rfi signals suppressed .
an evolutionary programming algorithm is used for this task .
computer simulations , made for both one - dimensional and two - dimensional array set - ups , show considerable suppression of rfi and acceptable changes to the main array beam in the radio source direction .
* rfi mitigation with phase - only adaptive beamforming * + p. a. fridman + astron , dwingeloo , p.o .
box 2 , 7990 aa the netherlands + e - mail : [email protected] |
planetary nebulae ( pne ) form around stars of low and intermediate mass ( @xmath1 m@xmath2 ) towards the end of their evolution .
the nebula originates from the interaction between a fairly dense ( @xmath310@xmath4 m@xmath2 yr@xmath5 ) ` slow wind ' ( @xmath310 km s@xmath5 ) lost by the star as a red giant on the asymptotic giant branch ( agb ) with a tenuous ( @xmath310@xmath6 m@xmath2 yr@xmath5 ) ` fast wind ' ( @xmath31000 km s@xmath5 ) that follows after the agb phase .
the nebula material becomes ionized once the effective temperature of the central star , which evolves into a white dwarf , exceeds 25000 k ( for a review see @xcite ) .
ngc 6302 ( pn [email protected] ) is a bipolar pn .
its complex and clumpy morphology shown in h@xmath0 and [ nii ] _ hubble space telescope _
( _ hst _ ) images @xcite can be roughly approximated by a bipolar shape with the two main lobes extending in the east - west direction .
it presents a highly pinched waist which is characteristic for butterfly - shaped bipolar pne @xcite .
the central star of ngc 6302 was directly detected for the first time by @xcite .
it was partially obscured by a dense equatorial lane @xcite and molecular material observed in co emission @xcite that is tracing an expanding torus oriented in the north - south direction , approximately perpendicular to the optical nebula axis . the distance to this nebula
was not known accurately .
@xcite gave a firm lower limit of [email protected] kpc based on a radio expansion proper - motion ( pm ) measurement of the nebular core , and estimate a distance of [email protected] kpc from measurements of pressure - broadening of radio recombination lines .
lately , @xcite determined an unambiguous distance of [email protected] kpc , derived from expansion pms of 15 knots located at the northwestern lobe of the nebula .
these pms were measured comparing two ground - based images with a time separation of @xmath351 yr , indicating a hubble - type expansion .
hubble - type outflows are characterised by an expansion velocity that increases linearly with the distance to the central star .
this type of outflow has been observed also in other objects , i.e. , the nebula around the symbiotic system hen 2 - 104 @xcite , ngc 6537 @xcite , mz 3 @xcite , see @xcite for other examples .
recently , szyszka et al . (
2011 ) measured the expansion pms in ngc 6302 , comparing two _ hst _ images in [ nii ] separated by 9.4 yr .
the velocity field follows a hubble law in agreement with the previous results of @xcite .
the pm vectors present a pattern mostly radial pointing back to the central source with a position close to the central star detected by @xcite .
their results show that the lobes of ngc 6302 were ejected during a brief event 2250@xmath835 yr ago ( in agreement with the result of @xcite ) , and they find evidence for a subsequent acceleration that increased the velocity of the inner regions by 9.2 km s@xmath5 , possibly related to the onset of the ionization . in this work
, we investigate whether or not the morphological and kinematical characteristics of ngc 6302 can result from the interaction between an isotropic fast wind with a toroidally - shaped slow wind .
we have also considered a clumpy structure for the slow wind , based on observations of agb shells , which look like clumpy and filamentary shell structures @xcite .
the paper is organised as follows . in section 2 ,
we describe the basic assumptions and the initial conditions used in the numerical simulations . in section 3
, we present the results including synthetic h@xmath0 emission maps , simulated pms of nebular knots , and their comparison with the observations .
finally , in section 4 , we discuss the implications of our results and give our conclusions .
our simulations are based on the generalisation of the interacting stellar winds model ( gisw ) in which , an isotropic fast wind launched by a star expands into a previously ejected toroidally - shaped slow wind @xcite .
we assumed that the slow wind has a density distribution with a high contrast between the equator and the pole , which is described by the following equation ( given by @xcite ) @xmath9 with @xmath10,\ ] ] where @xmath11 is the distance from the central star and @xmath12 is the polar angle ( @xmath13 at the pole , and @xmath14 at the equator ) .
the parameter @xmath0 determines the ratio between the value of the density at the equator and that at the pole , while @xmath15 determines the shape of the variation , and therefore that of the slow wind ( see @xcite ) .
the value of @xmath16 can be calculated from the mass - loss rate @xmath17 as : @xmath18 where @xmath19 is the constant terminal velocity of the agb wind and @xmath20 is the radius of the region where the stellar wind is imposed . the 3d numerical simulations were performed with the hydrodynamical code @xcite , which integrates the gas dynamical equations with a second - order accurate scheme ( in time and space ) using the ` flux - vector splitting ' method of @xcite on a binary adaptive grid . a rate equation for neutral hydrogen is integrated together with the gas dynamics equations to include the radiative losses through a parametrized cooling function that depends on the density , temperature and hydrogen ionization fraction @xcite . to model the ngc 6302 nebula ,
we have used a computational domain of @xmath21 cm along the @xmath22- , @xmath23- , and @xmath24-axes , respectively .
five refinement levels were allowed in the adaptive cartesian grid , achieving a resolution of @xmath25 cm at the finest level . for the initial condition we have filled the computational domain with the density distribution obtained from equation ( 1)-(3 ) with @xmath17 , @xmath19 , and @xmath20 equal to @xmath26 m@xmath2 yr@xmath5 , @xmath27 , and @xmath28 cm , respectively .
we have used @xmath29 and @xmath30 in order to reproduce the observed morphology . to simulate a ` clumpy ' circumstellar medium ( csm ) , we have modulated the density by a fractal structure with a spectral index of @xmath31 , which is consistent with a turbulent interstellar medium ( see @xcite ) .
this clumpy density is imposed on the initial condition and has fluctuations on the order of @xmath32 of the mean density value . in this csm
, we start an isotropic and fast stellar wind with mass injection @xmath33 of @xmath34 m@xmath2 yr@xmath5 , and a velocity @xmath35 of @xmath36 .
the 3d hydrodynamical simulations of two interacting winds were carried out with the setup given in section 2 .
the fast wind generates a shock wave which propagates into the surrounding csm .
a global bipolar morphology is generated and a similar size to the observed one is achieved after a time of 2000 yr , which is in agreement with the dynamical age given by szyszka et al .
( 2011 ) . from the density and temperature distribution given by our 3d hydrodynamical simulations
, we can perform synthetic h@xmath0 maps . in fig .
[ fha ] , we show the synthetic h@xmath0 emission maps obtained for the @xmath37 and @xmath38 projections ( left and right panels of fig .
[ fha ] , respectively ) , using an angle @xmath39 .
this angle @xmath40 is the angle between the nebula axis ( the @xmath24 axis of the computational domain ) and the plane of the sky , in agreement with previous observational results @xcite .
[ fha ] shows a bipolar morphology for both projections , with a waist of @xmath41 cm , which has a similar size to the observed one ( @xmath42 if a distance of 1.17 kpc is considered ) .
furthermore , both panels display a filamentary and clumpy structure , which is a consequence of the interaction between the fast wind and the clumpy slow wind .
emission maps obtained at an integration time of 2000 yr and for the @xmath37 and @xmath38 projections ( left and right panel , respectively ) .
the angle @xmath40 is the angle between the nebula axis ( the @xmath43 axis of the computational domain ) and the plane of the sky .
the vertical and horizontal axes are in units of 10@xmath44 cm .
the horizontal colour bar gives the h@xmath0 flux in units of erg s@xmath5 @xmath45 sr@xmath5.,width=309 ] to calculate the pms of the nebular knots we have re - started our simulation at an integration time of 2000 yr , and leave it to evolve by 10 additional yr .
then , we obtain the pms of the knots found in the simulation using a cross - correlation method .
we have used the `` pm mapping '' technique described by @xcite , in which , the pms from pair of images are derived by defining boxes including emitting knots and carrying out a cross - correlation function of the emission within the boxes .
the pm is obtained from a fit to the peak of the cross - correlation function .
this method has proven to be better than carrying out direct fits ( e.g. , a gaussian or a two - dimensional paraboloid ) to the observed emission features , because the cross - correlation functions , which are integrals of the emission within the chosen boxes , have higher signal - to - noise ratios than the images .
the cross - correlation boxes have a size of 30@xmath4630 pixels equivalent to 10@xmath4610 arcsec@xmath47 . for a given box of size @xmath48
, we first check whether or not the condition @xmath49 is satisfied in at least one pixel within the central inner box of size @xmath50 ( see appendix of @xcite ) . here
@xmath51 is the h@xmath0 flux of the image and @xmath52 is set equal to @xmath34 erg s@xmath5 @xmath45 sr@xmath5 .
if this condition is met at least for a single pixel in each of the two epochs that are being analysed , the cross - correlation function ( within the @xmath48-size box in the two images ) and the pm are computed .
[ fhapm ] shows the @xmath38 projection of the synthetic h@xmath0 map overlaid by white arrows which represent pm vectors calculated with the `` pm mapping '' technique .
interestingly , the pms show deviations from the radial direction mostly for filamentary features in the outer regions of the outflow lobes . for a spatially extended filamentary structure ,
the pm perpendicular to the locus of the emission is well determined . while the component of the pm of the flow along the filament is highly uncertain , resulting in a large dispersion of the pm vectors determined for filamentary features .
pm map of ngc 6302 .
the nebula axis is oriented at @xmath53 with respect to the plane of the sky .
the vertical and horizontal axes have the same dimensions as in fig . [ fha ] .
the vertical colour bar gives the h@xmath0 flux in units of erg s@xmath5 @xmath45 sr@xmath5 .
the length of each arrow indicates the relative pm of the h@xmath0 emission knots computed in 10@xmath4610 arcsec@xmath47 boxes .
the arrow at bottom left corresponds to a value of 200 km s@xmath5.,width=309 ] fig .
[ figpms ] shows pms vs. distance to the central star plots for the top lobe ( left panel ) and bottom lobe ( right panel ) . from this figure
, we note that the more distant filaments and knots move faster than those located near the source .
however , the pms obtained from our model do not resemble the behaviour given by a hubble - type expansion , such as found in the ngc 6302 observations ( szyszka et al . 2011 ) .
the hubble - type expansion is represented in both panels of fig .
[ figpms ] by dashed lines .
we can see from the fig .
[ figpms ] that many of the knots far from the source have velocities comparable to the observations , however we see a majority with velocities below the hubble - type expansion . at the same time
, the knots that are closer to the source ( @xmath54 ) , have a remarkably constant pm , well below the observations .
this result suggest that an additional acceleration mechanism is needed , particularly near the central star , which is consistent with the findings of szyszka et al .
a possible candidate for this acceleration mechanism is the ionizing flux from the central star , which would photoevaporate the clumps and push them through a rocket effect ( see * ? ? ?
this , and any other acceleration mechanism is not included in our models , and will be pursued in a following work .
we have carried out 3d hydrodynamical simulations of a two interacting - winds model , in order to explain both the global morphology and the tangential velocity of the pn ngc 6302 .
we reproduce the morphology and size of the nebula after an integration time of 2000 yr , which is similar to the dynamical time estimate given by szyszka et al .
we note the that it was necessary to use a high equator to pole density contrast of @xmath31000 to obtain the observed morphology .
a pm study was done on the synthetic h@xmath0 images during a time span of 10 yr similar to the time span between observed images .
the magnitude of the pms as well as the separation from the central star are quantitatively similar , although the obtained pms do not follow a hubble - type expansion as the observations indicate .
this suggests that an additional acceleration mechanism is acting , in particularly near the central star where the pm discrepancy is larger . a possible candidate for this mechanism
is the rocket effect from the photoevaporation of the clumps , due to the radiation of the central star .
we would like to dedicate this paper to the memory of our dear colleague yolanda gmez , who sadly passed away in 2012 .
she was also involved in this project .
lu and pb are grateful to john meaburn for his comments related to a very early draft version of this paper .
we are thankful to our referee , myfanwy lloyd , for her valuable comments on the manuscript .
lu acknowledges support from grant pe9 - 1160 of the greek general secretariat for research and technology in the framework of the program support of postdoctoral researchers .
pfv , acr , jc , and ae acknowledge support from the conacyt grants 167611 , 167625 , and unam dgapa grant ig100214 . 99 balick b. , frank a. , 2002 , ara&a , 40 , 439 corradi r. l. m. , 2004 , aspc , 313 , 148 corradi r. l. m. , livio m. , balick b. , munari u. , schwarz h. e. , 2001 , apj , 553 , 211 corradi r. l. m. , schwarz h. e. , 1993 , a&a , 269 , 462 cox n. l. j. , et al . , 2012 ,
a&a , 537 , a35 dinh - v - trung , bujarrabal v. , castro - carrizo a. , lim j. , kwok s. , 2008 , apj , 673 , 934 esquivel a. , lazarian a. , 2005 , apj , 631 , 320 esquivel a. , lazarian a. , pogosyan d. , cho j. , 2003 , mnras , 342 , 325 esquivel a. , raga a. c. , 2007 , mnras , 377 , 383 gmez y. , rodrguez l. f. , moran j. m. , garay g. , 1989 , apj , 345 , 862 icke v. , preston h. l. , balick b. , 1989 , aj , 97 , 462 matsuura m. , zijlstra a. a. , molster f. j. , waters l. b. f. m. , nomura h. , sahai r. , hoare m. g. , 2005 , mnras , 359 , 383 meaburn j. , lloyd m. , vaytet n. m. h. , lpez j. a. , 2008 , mnras , 385 , 269 mellema g. , 1995 , mnras , 277 , 173 mellema g. , frank a. , 1995 , mnras , 273 , 401 ossenkopf v. , esquivel a. , lazarian a. , stutzki j. , 2006 , a&a , 452 , 223 peretto n. , fuller g. , zijlstra a. , patel n. , 2007 , a&a , 473 , 207 raga a. c. , navarro - gonzlez r. , villagrn - muniz m. , 2000 , rmxaa , 36 , 67 raga a. c. , noriega - crespo a. , carey s. j. , arce h. g. , 2013 , aj , 145 , 28 raga a. c. , reipurth b. , 2004 , rmxaa , 40 , 15 redman m. p. , oconnor j. a. , holloway a. j. , bryce m. , meaburn j. , 2000 , mnras , 312 , l23 steffen w. , koning n. , esquivel a. , garca - segura g. , garca - daz m. t. , lpez j. a. , magnor m. , 2013 , mnras , 436 , 470 szyszka c. , walsh j. r. , zijlstra a. a. , tsamis y. g. , 2009 , apj , 707 , l32 szyszka c. , zijlstra a. a. , walsh j. r. , 2011 , mnras , 416 , 715 van leer b. , 1982 , lecture notes in physics , 170 , 507 | we present 3d hydrodynamical simulations of an isotropic fast wind interacting with a previously ejected toroidally - shaped slow wind in order to model both the observed morphology and the kinematics of the planetary nebula ( pn ) .
this source , also known as the butterfly nebula , presents one of the most complex morphologies ever observed in pne . from our numerical simulations , we have obtained an intensity map for the h@xmath0 emission to make a comparison with the _ hubble space telescope _ ( _ hst _ ) observations of this object .
we have also carried out a proper motion ( pm ) study from our numerical results , in order to compare with previous observational studies .
we have found that the two interacting stellar wind model reproduces well the morphology of , and while the pm in the models are similar to the observations , our results suggest that an acceleration mechanism is needed to explain the hubble - type expansion found in _ hst _ observations .
[ firstpage ] methods : numerical planetary nebulae : general planetary nebulae : individual : ngc 6302 |
although quantum chromodynamics ( qcd ) is widely accepted as the underlying theory of strong interactions , _ ab initio _ calculations of nuclear properties in lattice qcd remain a large theoretical challenge @xcite . in nuclear physics ,
the relevant degrees of freedom are pions and nucleons , and much of the computational effort in such a calculation would be required for generating the correct degrees of freedom from quarks and gluons rather than their interactions .
traditionally , nucleon nucleon interactions are described via phenomenological nuclear forces fitted to scattering data .
effective field theory ( eft ) provides a powerful method to construct nuclear forces with a direct connection to qcd in a systematic , model - independent way @xcite . for very low energies and momenta @xmath1 , the non - analyticities from pion exchange
can not be resolved and one can hence use an eft including only short - range contact interactions between nucleons @xcite . this theory is constructed to reproduce the effective range expansion @xcite in the two - body system and recovers efimov s universal approach to the three - nucleon problem @xcite .
an advantage of the eft formulation is that it can be extended to higher - body systems and external currents in a straightforward way .
the extension of this eft to include the long - range coulomb interaction was first discussed by kong and ravndal for the proton
proton channel @xcite . in ref .
@xcite , this analysis was extended to next - to - next - to - leading order . a renormalisation group analysis of proton proton scattering in a distorted wave basis
was carried out in refs .
moreover , this theory was applied to proton proton fusion in refs .
an extension of this formalism to three charged particles would be important for the possible interpretation of the hoyle state in @xmath2c as an efimov state of @xmath3 particles @xcite and the cluster eft for halo nuclei @xcite . in this work , we are interested in the simpler problem of a three - body system with two charged particles .
close to threshold , the coulomb interaction is strong .
its long - range nature requires special care in the non - perturbative treatment using momentum space integral equations . at higher energies ,
the coulomb interaction becomes perturbative .
rupak and kong have formulated a power counting for the coulomb contributions in the quartet channel of proton deuteron ( @xmath4@xmath5 ) scattering .
they calculated the phase shifts to next - to - next - to - leading order ( ) in the pionless eft and included coulomb effects to next - to - leading order ( ) @xcite .
however , they were not able to extend their calculation to the threshold region below center - of - mass momenta of 20 @xmath6
. they did not consider the doublet channel and the @xmath0he bound state .
a leading order calculation of the @xmath0he nucleus including non - perturbative coulomb interactions was recently presented by ando and birse @xcite .
including isospin breaking effects in the nucleon
nucleon scattering lengths , they obtain a surprisingly accurate description of the @xmath0he@xmath0h binding energy difference but they did not consider scattering observables .
a similar study to in the pionless eft was carried out using the resonating group method @xcite .
their results do not include isospin breaking and are consistent with other determinations of the @xmath0he@xmath0h binding energy difference . in this paper , we focus on @xmath4@xmath5 scattering observables in the quartet and doublet channels .
we extend the power counting by rupak and kong for the coulomb contribution to the doublet channel . by using a special integration mesh ,
we are able to calculate the phase shifts in both channels down to momenta of order 3 @xmath6 .
we also provide a perturbative evaluation of the coulomb contribution to the @xmath0he@xmath0h binding energy difference .
in this section , we briefly summarise the formalism required for calculating @xmath4@xmath5 scattering in the pionless theory .
more technical details can , _ e.g. _ , be found in refs .
@xcite .
we use the effective lagrangian @xmath7d^i -t^{a\dagger}\left[{\sigma_t}+\left({\mathrm{i}}d_0+\frac{{\boldsymbol{d}}^2}{4{m_n}}\right)\right]t^a \\[0.2 cm ] + { y_d}\left[d^{i\dagger}\left(n^t p^i_d n\right)+{\mathrm{h.c.}}\right ] + { y_t}\left[t^{a\dagger}\left(n^t p^a_t n\right)+{\mathrm{h.c.}}\right ] + \mathcal{l}_\mathrm{photon}+\mathcal{l}_3 \ , , \label{eq : l - nd}\end{gathered}\ ] ] with the nucleon field @xmath8 and two dibaryon fields @xmath9 ( with spin 1 and isospin 0 ) and @xmath10 ( with spin 0 and isospin 1 ) , corresponding to the deuteron and the spin - singlet virtual bound state in s - wave nucleon nucleon scattering .
both dibaryon fields are formally ghosts since their kinetic terms have a negative sign .
this is required to avoid the wigner bound and reproduce the positive value of the effective range with short - range interactions @xcite .
spin and isospin degrees of freedom are included by treating the field @xmath8 as a doublet in both spaces , but for notational convenience we have suppressed the spin and isospin indices of @xmath8 .
the projection operators , @xmath11 with @xmath12 ( @xmath13 ) operating in spin ( isospin ) space , project out the @xmath14 and @xmath15 nucleon nucleon partial waves , respectively .
the covariant derivative @xmath16 where @xmath17 is the charge operator , includes the coupling to the electromagnetic field .
furthermore , we have the kinetic and gauge fixing terms for the photons , @xmath18 of which we only keep contributions from coulomb photons .
these correspond to a static coulomb potential between charged particles , but for convenience we introduce feynman rules for a coulomb photon propagator , @xmath19 which we draw as a wavy line , and factors @xmath20 for the vertices . following @xcite , we have introduced a small photon mass @xmath21 to regulate the singularity of the propagator at zero momentum transfer .
as we will discuss later on , this regulator will be removed by numerical extrapolation back to vanishing photon mass . in the doublet - channel of the three - nucleon system ,
a three - body contact interaction is required for renormalisation at leading order @xcite .
it can be written as @xmath22 \bigg ) \
, , \label{eq : l-3}\end{gathered}\ ] ] where @xmath23 is a momentum cutoff applied in the three - body equations discussed below and @xmath24 a known log - periodic function of the cutoff that depends on a three - body parameter @xmath25 . in the strong sector
, we adopt the standard power counting for large s - wave scattering length @xcite . a nucleon bubble together with a bare dibaryon propagator scales as @xmath26 .
the bare dibaryon propagators therefore are dressed by nucleon bubbles to all orders .
the resulting geometric series for the full propagators are shown in fig .
[ fig : dibaryonprop ] .
state ( _ i.e. _ the deuteron ) and ( b ) the @xmath15 state . ] for convenience , we also resum the effective range corrections .
if desired , the perturbative expressions can always be obtained by re - expanding the propagators .
we do not go into the details of the calculations here and simply quote the results for the renormalised propagators , which we obtain by demanding that the effective range expansions @xmath27 around the deuteron pole , and @xmath28 for the singlet channel are reproduced . in writing eq .
( [ eq : er - t ] ) we have used that @xmath29 to the order we are working .
this yields the expressions @xmath30 and analogously @xmath31 for the spin - singlet dibaryon .
these expressions are valid to . at leading order , effective range corrections
are not included and the dibaryon kinetic terms do not contribute .
the corresponding propagators are obtained by setting @xmath32 and @xmath33 in eqs .
( [ eq : prop - d - high ] ) and ( [ eq : prop - t - high ] ) .
the deuteron wave function renormalisation constant is given as the residue at the bound state pole : @xmath34 the coulomb interaction breaks the isospin symmetry that is implicit in the dibaryon propagators from the previous subsection . for the @xmath35-part of the singlet dibaryon we can also have coulomb photon exchanges inside the nucleon bubble .
these can be resummed to all orders , yielding a dressed nucleon bubble @xcite , which is then used to calculate the full singlet dibaryon propagator in the @xmath35-channel , as shown in fig .
[ fig : dressedbubble ] . -channel . ]
the result for the leading order propagator is @xcite @xmath36 with @xmath37 and @xmath38 where @xmath39 denotes the logarithmic derivative of the @xmath40-function .
effective range corrections can be included in the same way as described above .
the power counting of pionless effective field theory has been extensively discussed in the literature ( see the reviews @xcite and references therein ) .
we will thus be rather brief on this subject here .
we will , however , elaborate a bit on the power counting for the coulomb sector of the theory , as it was introduced in @xcite .
the low - energy scale @xmath41 of the theory is set by the deuteron binding momentum @xmath42 .
we can formally count the external momenta @xmath43 to be of the same order .
since we are working in a theory without explicit pions , the natural ultraviolet cutoff of our theory is of the order of the pion mass , @xmath44 .
which cutoff is best to use in practice depends on whether one discusses the quartet - channel system ( where short - range effects are suppressed by the pauli principle ) , or the doublet - channel system . in the first case one
finds that already @xmath45 is sufficient for an accurate description , whereas in the latter case we have to set the cutoff to a few hundred @xmath6 to reach convergence .
the combination of the two scales yields the expansion parameter @xmath46 .
a further relevant scale in our system is the nucleon mass @xmath47 .
it appears explicitly in kinetic energies , which scale as @xmath48 . as a consequence ,
the nucleon propagator scales as @xmath49 and the loop integration measure @xmath50 scales as @xmath51 .
we assume @xmath52 for the nucleon
dibaryon coupling constants and @xmath53 for the bare dibaryon propagator constants . from the form of the coulomb potential in momentum space ,
@xmath54 it is clear that coulomb contributions are dominant for small momentum transfers .
as noted in @xcite , they enter @xmath55 .
this behaviour is not captured by the power counting for the strong sector .
hence , when one wants to perform calculations including coulomb effects for small external momenta , one can no longer assume that all momenta scale with @xmath56 .
instead , one has to keep track of the new scale introduced by the external momenta separately .
we generically denote this scale by @xmath4 and assume @xmath57 for the power counting . as noted in @xcite
, this means that we make a simultaneous expansion in _ two _ small parameters @xmath58 and @xmath59 . for @xmath60 ,
the coulomb contributions are small and the results in both schemes agree . with this modified counting , it is not straightforward to deduce the scaling of loops anymore .
kinetic energies always scale like @xmath41 , so the scaling of @xmath61 and the nucleon propagator is not modified in the presence of coulomb effects .
however , where we could simply assume that all loop momenta scale like @xmath41 before , we now have to check first which contribution is picked up ( or rather enhanced ) after carrying out the @xmath61-integral . in general
, we have that 1 .
the loop integration measure @xmath62 scales as @xmath63 , and 2 .
the photon propagator scales as @xmath64 , where either @xmath65 or @xmath66 .
these rules will become more transparent when we apply them to deduce the scaling of the diagrams shown below .
@xmath5 scattering involving the exchange of coulomb photons . ] in this subsection , we discuss several diagrams contributing to @xmath4@xmath5 scattering that include coulomb photons ( see fig .
[ fig : coulombdiagrams ] ) . for the discussion
we always assume @xmath67 .
diagram ( a ) simply scales as @xmath68 .
the diagram ( b ) is enhanced relative to ( a ) by a factor @xmath69 from the nucleon bubble and hence gives the leading - order coulomb contribution . diagram ( a ) enters only at since the dibaryon kinetic energy operators , which generate the direct coupling of the photons to the dibaryons ,
enter only as effective range corrections .
diagram ( c ) contains two loops , which have to be analysed separately .
the upper nucleon bubble does not contain any photon propagators , so there all momenta scale as @xmath41 . in the lower loop , momenta certainly scale @xmath70 due to the two photon propagators that involve the external momentum . including the remaining nucleon propagator , which cancels the contribution from the integration measure , we are left with a total scaling @xmath71 for diagram ( c ) .
this means that compared to diagram ( b ) it is suppressed by a factor @xmath72 . for diagrams of the form ( c ) with more than two photons attached to the bubble we simply quote the results from @xcite .
the diagram with three photons could contribute with a factor @xmath73 , whereas the diagrams with @xmath74 photons attached to the bubble are even infrared finite and suppressed by factors @xmath75 . following @xcite , we neglect them all and also the logarithmically - scaling diagram with three photons ( which is already small for @xmath76 ) . diagram ( d ) is a little ambiguous since _ a priori _ it is not clear whether the loop momentum should scale @xmath77 or @xmath70 . in @xcite , the first alternative is chosen , yielding that compared to the same diagram without the photon it is suppressed by a factor @xmath78 .
a direct numerical calculation shows that it is a seven - percent effect at threshold .
the diagram ( e ) obviously is irrelevant for the quartet - channel system ( there are never two protons in the dibaryon ) , but , at least in principle , it can play a role in the doublet - channel system .
the power counting , however , yields the same suppression factor as for diagram ( c ) , only in this case the scaling of the loop momentum is not ambiguous .
a direct numerical evaluation yields that at threshold it is a 15% contribution ( again compared to the simple nucleon - exchange diagram without the photon ) .
we take this value as the _ a priori _ theoretical uncertainty of our doublet - channel calculation .
the bottom line of the discussion above is that , as done in @xcite , we iterate the diagrams ( a ) and ( b ) to all orders and do not include any of the other diagrams shown in fig . [
fig : coulombdiagrams ] .
the claim is that this procedure is adequate for both the quartet - channel and the doublet - channel system .
the coulomb effects are thus included at accuracy in our calculation .
we start with a review of @xmath8@xmath5 quartet - channel scattering , where the spin 1 of the deuteron and the spin @xmath79 of the nucleon couple to a total spin of @xmath80 . since this coupling is only possible when the spins of all three nucleons taking part in the reaction are aligned , the pauli principle applies .
hence , the system is rather insensitive to short - range physics .
furthermore , only the dibaryon field representing the deuteron can appear in the intermediate state . in the quartet channel . ] in fig .
[ fig : nd - inteq - q ] we show a diagrammatic representation of the strong ( neutron deuteron ) scattering amplitude @xmath81 , which does not include any coulomb effects .
it is projected onto the spin quartet channel by setting @xmath82 and @xmath83 for the in- and outgoing deuteron spin indices , respectively , and @xmath84 to select the neutron .
after furthermore projecting onto s - waves , we get @xmath85 where @xmath86 and @xmath4 are the incoming and outgoing momenta of the particles in the center - of - mass frame , and @xmath87 more details on the derivation of this equation and the required projections can be found in appendix [ sec : scatteq - details ] .
we have introduced a cutoff @xmath23 to regularise the loop integral , which is particularly convenient for a numerical treatment of the equation .
strictly , however , this regulator is only required for the full amplitude including coulomb interactions and is applied here for convenience only . after setting the energy @xmath88 to the total center - of - mass energy , @xmath89 the equation
is solved numerically with standard linear algebra routines after discretising the integrals . from the result
we then obtain the s - wave scattering phase shift @xmath90 which can be compared to experimental data . in order to simplify the expressions in the following sections , we introduce a short - hand notation for the scattering equations . defining @xmath91 and @xmath92 along with the operation @xmath93
we find that we can write as @xmath94 \,,\ ] ] where we have omitted all arguments . in the quartet channel .
] . ] in order to include coulomb effects and hence discuss proton
deuteron scattering we follow @xcite and define a full scattering amplitude @xmath95 ( see fig . [ fig : pd - inteq - q ] ) that includes both strong and coulomb interactions and a pure coulomb scattering amplitude @xmath96 ( see fig . [
fig : pd - inteq - coulomb ] ) .
after spin- , isospin- and s - wave projection we find the integral equations @xmath97 \label{eq : pd - inteq - q - full}\ ] ] and @xmath98 \label{eq : pd - inteq - q - c}\ ] ] with @xmath99 after solving the individual equations , we calculate the phase shifts @xmath100 and @xmath101 according to .
the final result , which we will compare to experimental data , is the coulomb - subtracted phase shift @xcite , @xmath102 note that in the integral equations above an artificial dependence on the bare coupling constant @xmath103 is kept for notational convenience . in all observables ,
this dependence drops out .
we now go on to the doublet channel , where the spins of the nucleon and the deuteron couple to a total spin of @xmath79 .
the spin - singlet dibaryon can now appear in the intermediate state , which leads to two coupled amplitudes that differ in the type of the outgoing dibaryon .
in contrast to the quartet channel , the three nucleon spins no longer need to be aligned in the same direction , which means that a non - derivative three - nucleon interaction is no longer prohibited by the pauli principle .
this channel is expected to be more sensitive to short - range physics in general , and in fact the three - body interaction ( [ eq : l-3 ] ) is _ needed _ at leading order to ensure correct renormalisation @xcite . in the doublet channel .
the diagrams involving the three - body force have been omitted . ] as we did in the quartet channel , we look at the neutron
deuteron system first .
[ fig : nd - inteq - d ] shows a diagrammatic representation of the coupled - channel integral equation for the scattering amplitude @xmath81 , of which we only needed to consider the upper left part for the quartet - channel system .
the contribution of the three - body interaction ( [ eq : l-3 ] ) is omitted here and will be included below .
after projecting onto the @xmath104@xmath5 doublet channel with @xmath105 we find @xmath106 + { { { \mathcal{t}}_{\mathrm{s}}}^\mathrm{d , b}}\otimes \left[\frac{3{m_n}{y_d}{y_t}}2 d_t { k_{\mathrm{s}}}\right ] \
\label{eq : nd - inteq - d - b - no - h } { { { \mathcal{t}}_{\mathrm{s}}}^\mathrm{d , b}}&= -\frac{3{m_n}{y_d}{y_t}}2{k_{\mathrm{s}}}+ { { { \mathcal{t}}_{\mathrm{s}}}^\mathrm{d , a}}\otimes \left[\frac{3{m_n}{y_d}{y_t}}2 d_d { k_{\mathrm{s}}}\right ] - { { { \mathcal{t}}_{\mathrm{s}}}^\mathrm{d , b}}\otimes \left[\frac{{m_n}{y_t}^2}2 d_t { k_{\mathrm{s}}}\right ] \,.\end{aligned}\ ] ] [ eq : nd - inteq - d - no - h ] more details on the derivation and projection are again given in appendix [ sec : scatteq - details ] . we still need to include the contribution of the three - nucleon interaction ( [ eq : l-3 ] ) in eqs . ( [ eq : nd - inteq - d - a - no - h ] ) and ( [ eq : nd - inteq - d - b - no - h ] ) .
a straightforward calculation shows that this can be achieved with the replacement @xcite @xmath107 it is , however , important to note that for the terms with an additional factor of @xmath108 in front of @xmath109 , it is cancelled in the @xmath24-part by the additional factor @xmath110 in . with this
, we arrive at the final version of our integral equations , @xmath111 { { { \mathcal{t}}_{\mathrm{s}}}^\mathrm{d , b}}\end{pmatrix } = \begin{pmatrix } g_{dd}\left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right)\\[{0.5em } ] -g_{dt}\left(3{k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right ) \end{pmatrix } \\
+ \begin{pmatrix } -g_{dd}d_d\left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right ) & g_{dt}d_t\left(3{k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right ) \\[{0.5em } ] g_{dt}d_d\left(3{k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right ) & g_{tt}d_t\left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right ) \end{pmatrix } \otimes\begin{pmatrix}{{{\mathcal{t}}_{\mathrm{s}}}^\mathrm{d , a}}\\[{0.5em}]{{{\mathcal{t}}_{\mathrm{s}}}^\mathrm{d , b}}\end{pmatrix } \,,\end{gathered}\ ] ] written in a compact matrix
vector notation .
we have furthermore introduced the abbreviations @xmath112 again , the dependence on the bare coupling constants @xmath103 and @xmath113 is only kept for notational convenience and drops out in all observables . finally , we have all the ingredients to discuss the proton deuteron system in the doublet channel . in the doublet channel .
the diagrams representing the three - nucleon force have been omitted . ]
due to the fact that the electromagnetic interaction does not couple to isospin eigenstates we now need two different projections for the amplitude @xmath114 with the outgoing spin - singlet dibaryon : @xmath115 [ eq : proj - d - b1b2 ] the latter corresponds to the amplitude with the outgoing spin - singlet dibaryon in a pure @xmath35-state . for the diagrams that have this component in the intermediate state
, we have to insert the propagator into the unprojected equations in appendix [ sec : scatteq - details ] and find @xmath116{{{\mathcal{t}}_{\mathrm{full}}}^\mathrm{d , b1}}\\[{0.5em}]{{{\mathcal{t}}_{\mathrm{full}}}^\mathrm{d , b2}}\end{pmatrix } = \begin{pmatrix } g_{dd}\left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right)\\[{0.5em } ] -g_{dt}\left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{3\lambda^2}\right)\\[{0.5em } ] -g_{dt}\left(2{k_{\mathrm{s}}}+\frac{4h(\lambda)}{3\lambda^2}\right ) \end{pmatrix } + \begin{pmatrix}g_{dd}{{k_{\mathrm{c}}}^{(d)}}\\[{0.5em}]0\\[{0.5em}]0\end{pmatrix } \\ + \begin{pmatrix }
-g_{dd}d_d\left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right ) & g_{dt}d_t\left(3{k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right ) & 0\\[{0.5em } ] g_{dt}d_d\left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{3\lambda^2}\right ) & g_{tt}d_t\left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right ) & 0\\[{0.5em } ] g_{dt}d_d\left(2{k_{\mathrm{s}}}+\frac{4h(\lambda)}{3\lambda^2}\right ) & -g_{tt}d_t\left(2{k_{\mathrm{s}}}+\frac{4h(\lambda)}{\lambda^2}\right ) & 0 \end{pmatrix } \otimes \begin{pmatrix } { { { \mathcal{t}}_{\mathrm{full}}}^\mathrm{d , a}}\\[{0.5em}]{{{\mathcal{t}}_{\mathrm{full}}}^\mathrm{d , b1}}\\[{0.5em}]{{{\mathcal{t}}_{\mathrm{full}}}^\mathrm{d , b2}}\end{pmatrix } \\ + \begin{pmatrix } -g_{dd}d_d{{k_{\mathrm{c}}}^{(d ) } } & 0 & g_{dt}d_t^{pp } \left(3{k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right)\\[{0.5em } ] 0 & -g_{tt}d_t{{k_{\mathrm{c}}}^{(t ) } } & -g_{tt}d_t^{pp } \left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right)\\[{0.5em } ] 0 & 0 & 0 \end{pmatrix } \otimes\begin{pmatrix } { { { \mathcal{t}}_{\mathrm{full}}}^\mathrm{d , a}}\\[{0.5em}]{{{\mathcal{t}}_{\mathrm{full}}}^\mathrm{d , b1}}\\[{0.5em}]{{{\mathcal{t}}_{\mathrm{full}}}^\mathrm{d , b2}}\end{pmatrix } \label{eq : pd - inteq - d - full}\end{gathered}\ ] ] with @xmath117 the terms in have been separated in such a way that the sub - channels with coulomb contributions can be easily identified .
the equation for the coulomb scattering amplitude @xmath96 is exactly the same as in the quartet channel .
the dibaryon propagators with the resummed kinetic energy insertions have an unphysical deep bound state pole at the radius of convergence of the geometric series . in the quartet channel
the cutoff can be chosen low enough to avoid that pole . due to the larger cutoff needed in the doublet channel , however
, we can not use the resummed propagators here .
instead , we use linear and quadratic insertions of the kinetic energy operator in in the kernel of the integral equations in order to include effective range corrections and obtain the next - to - leading order and next - to - next - to - leading order propagators @xmath118 and @xmath119 @xcite .
alternatively , we can think of this as re - expanding the renormalised propagators and up to linear and quadratic order in @xmath120 , respectively .
this still resums some higher order effective range contributions , but removes the unphysical pole .
the question of when higher - order three - body forces enter in the doublet channel is still under discussion . in ref .
@xcite , a subleading three - body force was included as required by nave dimensional analysis .
a lepage - plot analysis showed that its inclusion reduces the errors in the calculation .
this was supported by ref .
@xcite , where a corresponding logarithmic divergence at n@xmath121lo , requiring a subleading three - body force , was identified .
more recently , platter and phillips , using the subtractive renormalisation scheme , showed that the leading three - body force is sufficient to achieve cutoff independence up to @xcite .
a perturbative analysis recently showed that there is a new three - body parameter already at if the scattering length is not fixed @xcite . in this work , the scattering length _
is _ fixed and we will not include a subleading three - body force .
assuming the counting of @xcite , our calculation will correspond to in the quartet channel and to in the doublet channel .
we will also perform a calculation including only the two - body interactions to in the doublet channel .
the integral equations presented in the previous sections have to be solved numerically .
we do so by discretising the integrals , using gaussian quadrature , principal value integration to deal with the singularity of the deuteron propagator , and appropriate transformations of the integration domain .
the latter are especially important to deal with the numerical difficulties caused by the coulomb photon propagators .
even though we have regulated the singularity with the artificial photon mass @xmath21 , the latter has to be kept small , which then yields strongly peaked functions .
it turns out that the coulomb peaks in the inhomogeneous parts of eqs . , and are the major numerical problem .
we solve it by concentrating the quadrature points around this peak .
together with always putting half of the quadrature points into the low - momentum region ( the interval from zero to the peak position ) , we are able to ( linearly ) extrapolate our results for the scattering phase shifts back to the physical value @xmath122 ( screening limit ) . as a typical example
, we show the quartet - channel phase shift at @xmath123 as a function of @xmath21 in fig .
[ fig : phase - l - q ] .
the linear dependence is clearly visible , with deviations only for very small @xmath21 . in the doublet channel
, the qualitative behaviour is the same , only the curvature becomes visible already for somewhat larger photon masses .
this can be understood by noting that the absolute value of the phase shift is smaller in the doublet channel , especially for low center - of - mass momenta @xmath86 ( _ cf .
_ sec .
[ sec : scatt - results ] ) .
we settled to use the intervals @xmath124 and @xmath125 for the extrapolations in the quartet and doublet channel , respectively .
we note that the error introduced by the extrapolation to @xmath122 can generally be neglected compared to the theoretical error from the eft expansion discussed below .
@xmath5 quartet channel s - wave scattering phase shift at for center - of - mass momentum @xmath123 and cutoff @xmath126 as a function of the regulating photon mass @xmath21.,scaledwidth=80.0% ] we have used the experimental input parameters shown in tab . [ tab : params ] in the numerical calculation .
.parameters used in the numerical calculation , @xmath127 . [ cols="^,^,^,^",options="header " , ]
@xmath5 quartet channel s - wave scattering phase shifts as functions of the center - of - mass momentum @xmath86 .
error bands generated by cutoff variation from 120 to 160 @xmath6 .
experimental @xmath4@xmath5 phase shift data taken from @xcite ( diamonds ) and @xcite ( circles).,scaledwidth=80.0% ] in fig . [ fig : phase - nd - q ] we show the phase shift results for both neutron deuteron and proton deuteron scattering as functions of the center - of - mass momentum @xmath86 .
the error bands are generated by varying the cutoff within a range of @xmath128 to @xmath129 . since the cutoff variation is small ,
we conclude that the calculation is well converged at these cutoffs . for @xmath130
some numerical artifacts show up from integrating over the unphysical second pole in the full deuteron propagator . for the @xmath104@xmath5 curve in fig .
[ fig : phase - nd - q ] we have used @xmath126 .
the fact that the bands do not overlap is no point of concern since they only give a lower bound on the error of the calculation . from the expansion parameter @xmath131 of the eft , the error can be estimated as 30% , 10% , and 3% at , , and , respectively .
thus , at , the 30% error from the expansion parameter clearly dominates . at and , however , the band from the cutoff variation gives a reasonable estimate of the total error in the calculation .
the result to the right of the dotted line at @xmath132 agrees nicely with the results presented in @xcite and also with the experimental data included in the plot . at this point
we remark , however , that for @xmath133 the coulomb parameter @xmath134 is of order @xmath110 , which means that in this regime the non - perturbative treatment of coulomb effects might not even be necessary . more important
are hence the @xmath4@xmath5 results for small momenta ( @xmath135 mev ) to the left of the dotted line , which we could obtain thanks to our optimised numerical procedure .
it would of course be good to have some data points in this region to test our prediction .
@xmath5 doublet channel s - wave scattering phase shifts as functions of the center - of - mass momentum @xmath86 .
error bands generated by cutoff variation from 200 to 600 @xmath6 .
experimental @xmath4@xmath5 phase shift data taken from @xcite.,scaledwidth=80.0% ] the doublet - channel results for the @xmath4@xmath5 scattering phase shifts as functions of the center - of - mass momentum @xmath86 are shown in fig .
[ fig : phase - nd - d ] . as in the quartet channel ,
the @xmath4@xmath5 curve lies above the @xmath104@xmath5 curve and agrees quite well with the experimental data .
a more quantitative comparison is , unfortunately , not possible since there are no errors given for the data points .
the error bands are generated by varying the cutoff within a natural range of @xmath136 to @xmath137 ( _ i.e. _ a few times the pion mass ) .
assuming the power counting of @xcite , our calculation is incomplete since the subleading three - body force is not included . a full calculation
, however , is beyond the scope of the paper since gauging the subleading three - body force creates new three - body contributions to the photon coupling .
our partial result is stable to within about ten percent under the cutoff variation , which is consistent with a 715% error estimate based on the neglected coulomb diagrams ( see sec . [
sec : power - counting ] and @xcite ) . figure [ fig : phase - nd - d ] furthermore shows how the results improve from order to order .
the stability of the partial result with respect to variation of the cutoff suggests that the scattering is relatively insensitive to the subleading three - body interaction . at higher energies , however , there is some room for such a contribution as a our partial result consistently lies two to four degrees above the data .
we observe that the shift from to is of the same order of magnitude as the shift from to .
this behaviour is typical for effective range corrections in the doublet channel @xcite .
the smallness of the corrections can be understood as a cancellation between two different contributions to this correction .
the two contributions are proportional to @xmath138 and @xmath139 , respectively , where @xmath140 is the typical momentum scale of the process .
furthermore , it is known that at observables are often described better than expected from the power counting once the exact pole position of the two - body propagator is reproduced @xcite . as a consequence ,
the shifts in observables from to can be small and of a size comparable to the corresponding shifts from to .
since the power counting we used for the coulomb contributions is not valid for the regime of typical bound state energies , we can not simply use the above equations to calculate @xmath0he .
there are various strategies to proceed .
we could extend the power counting and include additional coulomb diagrams in our equations .
alternatively , we could use an analytic expression for the full off - shell coulomb amplitude as it is done in ref .
@xcite for an calculation . here
, we choose the much simpler approach of calculating the coulomb energy shift for @xmath0he in first order perturbation theory as the expectation value of the coulomb interaction between proton and deuteron using trinucleon wave functions in the isospin limit . in order to obtain the trinucleon wave functions we need to solve the homogeneous coupled - channel equation @xmath141 with @xmath142 , @xmath143 , and @xmath144 g_{dt}\left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{3\lambda^2}\right ) & g_{tt}\left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right ) & -g_{tt}\left({k_{\mathrm{s}}}+\frac{2h(\lambda)}{\lambda^2}\right)\\[{0.5em } ] g_{dt}\left(2{k_{\mathrm{s}}}+\frac{4h(\lambda)}{3\lambda^2}\right ) &
-g_{tt}\left(2{k_{\mathrm{s}}}+\frac{4h(\lambda)}{\lambda^2}\right ) & 0
\end{pmatrix } \,.\ ] ] it is obtained by applying the projections to the raw equation without coulomb contributions . the reason to separate the wave function in this way is that the part with the dibaryon leg in the @xmath35-channel does not contribute to a perturbative calculation of the energy shift since in that case the third nucleon in the system necessarily is a neutron .
the energy in the equation is set to the experimental triton binding energy , @xmath145 and the existence of solution is ensured by adjusting the three - nucleon force @xmath24 appropriately , as it was already done to renormalise the scattering equations .
having obtained the wave functions as solutions of , we still need to normalise them properly .
since the eft generates an energy - dependent interaction in the three - body system , this is done by demanding that @xmath146 where @xmath147 with @xmath148 a short derivation of this normalisation condition can be found in appendix [ sec : triton - norm ] . with the normalised trinucleon wave functions we can obtain the coulomb - induced energy shift in first order perturbation theory and
find @xmath149 with the s - wave projected coulomb potential @xmath150 in momentum space .
our prediction for the @xmath0he binding energy is then given by @xmath151 the results are shown in fig .
[ fig : en - wf ] .
it is remarkable that the result , which should be accurate to about 10% , agrees very well with the experimental value @xmath152 over a large cutoff range . in our partial calculation ,
the results are still quite stable against cutoff variations within @xmath136 to about @xmath153 , but our value lies about @xmath154 above the experimental value .
this shift again leaves room for a natural - sized contribution of the subleading three - body we have not included .
brandenburger , coon and sauer have determined the @xmath0he@xmath0h binding energy difference in a largely model - independent way using experimental charge form factors @xcite .
they found that the coulomb contribution is about 10% below the experimental value for the total energy difference . within the expected error of 10% ,
our result is consistent with this .
ando and birse have carried out a non - perturbative calculation to leading order in the pionless eft including the full off - shell @xmath155-matrix for the coulomb interaction and found @xmath156 @xcite .
their calculation included isospin breaking effects in the nucleon
nucleon scattering lengths .
et al_. found the smaller value @xmath157 in an calculation using the resonating group model to solve the pionless eft with a charge - independent value of the spin - singlet scattering length and non - perturbative coulomb interactions @xcite .
binding energy difference in dependence of the cutoff @xmath23 .
bottom curve : result .
middle curve : result .
top curve : result.,scaledwidth=80.0% ] the increase in our result for @xmath158 that is seen in fig .
[ fig : en - wf ] also occurs at and , but for larger cutoffs . in fig .
[ fig : h - en - nlo ] we show the prediction for @xmath159 together with the three - nucleon force obtained from fitting the triton binding energy .
it is obvious that a drop in the binding energy prediction occurs whenever the three - nucleon force has gone through a pole .
we interpret this as an artifact of the theory which is related to the efimov effect . for cutoffs beyond the position of the first pole in @xmath24 ,
the triton is not the true ground state of the system anymore ; after each pole transition a new ( unphysical ) deep bound state emerges .
he binding energy together with the three - nucleon force ; results .
solid curve : @xmath0he binding energy prediction .
dashed curve : three - nucleon force . dotted line : experimental value for the @xmath0he binding energy.,scaledwidth=80.0% ] these unphysical deep states lead to additional nodes in the triton wave function at short distances which affect our perturbative results .
this suggests that an additional short - distance counterterm is required to cancel these contributions if one wants to go to cutoffs much larger than the pion mass .
in this paper , we have investigated s - wave proton
deuteron scattering in pionless effective field theory . in the quartet channel , we have calculated the elastic scattering phase shift up to using the power counting for coulomb contributions suggested by rupak and kong @xcite . the coulomb effects are included at accuracy in our calculation . using an optimised integration mesh we were able to extend their calculation into the threshold region were the coulomb interaction becomes highly non - perturbative .
we found good agreement both with available phase shift analyses and with the results of rupak and kong at momenta @xmath160 .
moreover , we extended the power counting to the doublet channel and performed a complete calculation of the phase shifts to in agreement with the available phase shift data .
we also carried out a partial calculation that neglected the contribution of the subleading three - body force entering at this order .
the results of this calculation are stable under variations of the cutoff .
furthermore , there is good agreement with the phase shift data at low momenta and room for a small contribution of the neglected three - body force at larger momenta .
overall , however , the doublet channel phase shifts are only weakly sensitive to the subleading three - body force entering at .
although we were mainly interested in @xmath4@xmath5 scattering , we have also calculated the coulomb contribution to the @xmath0he@xmath0h binding energy difference @xmath161 .
this observable has previously been calculated in the pionless theory by treating the coulomb interaction non - perturbatively @xcite . here
, we treat the coulomb potential between proton and deuteron in first order perturbation theory using trinucleon wave functions .
higher order corrections to this quantity are expected to be small .
our result is in reasonable agreement with the experimental value and other evaluations .
we find @xmath161 to be more sensitive to the subleading three - body force .
the partial result is about 10% too large , thus leaving room for a contribution from the omitted three - body force .
we also observe steps in the calculated value of @xmath161 as the cutoff is increased beyond its natural range . whenever the leading three - nucleon force has gone through a pole , a drop in the calculated binding
energy occurs .
we interpret this as an artifact of the theory related to the efimov effect . at higher cutoffs , spurious deep three - body
bound states appear and the triton is not the true ground state anymore .
it appears that an additional short - distance counterterm is required to cancel these contributions if one wants to go to cutoffs much larger than the pion mass .
a further study of this issue would be interesting . in the future ,
a full calculation including the subleading three - body force and the electromagnetic interaction terms generated from gauging its momentum dependence should be carried out .
such an accuracy will , _
e.g. _ , be required for high - precision calculations of low - energy astrophysical processes in pionless effective field theory and the effective field theory for halo nuclei .
we thank a. rusetsky , b. metsch , m. hoferichter and p. hagen for discussions and s .-
ando , s. coon , d. r. phillips and u. van kolck for comments on the manuscript .
this research was supported in part by the dfg through sfb / tr 16 `` subnuclear structure of matter '' and the bmbf under contract no . 06bn9006 .
s.k . was supported by the `` studienstiftung des deutschen volkes '' and by the bonn - cologne graduate school of physics and astronomy .
using the feynman rules that follow from the lagrangian and inserting appropriate symmetry factors , we find @xmath162 for the @xmath104@xmath5 scattering equation depicted in fig . [
fig : nd - inteq - q ] . in the same way we get @xmath163 \\
+ \int{{\!\!\frac{\mathrm{d}^3q}{(2\pi)^3}}}\,\delta_d\left(e-\frac{{\mathbf{q}}^2}{2{m_n}},{\mathbf{q}}\right ) \cdot({\mathrm{i}}{{\mathcal{t}}_{\mathrm{full}}}^{ik})^{\gamma c}_{\alpha a}(e;{\mathbf{k}},{\mathbf{q } } ) \cdot\left\{\frac{{m_n}{y_d}^2}{2}\frac{(\sigma^j\sigma^k)^\beta_\gamma\delta^b_c } { { \mathbf{q}}^2+{\mathbf{q}}\cdot{\mathbf{p}}+{\mathbf{p}}^2-{m_n}e-{\mathrm{i}}{\varepsilon}}\right . \\ + \left.\,\alpha{m_n}^2{y_d}^2\cdot\delta^{kj}\delta^\beta_\gamma \left(\frac{{\mathbf{1}}+\tau_3}{2}\right)^b_c \left[\frac{\mathcal{i}_\mathrm{bubble}(e;{\mathbf{q}},{\mathbf{p}})}{({\mathbf{q}}-{\mathbf{p}})^{2}+\lambda^2 } -\frac{{\rho_d}}{2}\frac{1}{({\mathbf{q}}-{\mathbf{p}})^2+\lambda^2}\right]\right\}\end{gathered}\ ] ] for the full @xmath4@xmath5 scattering equation shown in fig .
[ fig : pd - inteq - q ] , and @xmath164 \\
+ \int{{\!\!\frac{\mathrm{d}^3q}{(2\pi)^3}}}\,\delta_d\left(e-\frac{{\mathbf{q}}^2}{2{m_n}},{\mathbf{q}}\right ) \cdot({\mathrm{i}}{{\mathcal{t}}_{\mathrm{c}}}^{ik})^{\gamma c}_{\alpha a}(e;{\mathbf{k}},{\mathbf{q } } ) \\ \times\alpha{m_n}^2{y_d}^2\cdot\delta^{jk}\delta^\beta_\gamma \left(\frac{{\mathbf{1}}+\tau_3}{2}\right)^b_c \left[\frac{\mathcal{i}_\mathrm{bubble}(e;{\mathbf{q}},{\mathbf{p } } ) } { ( { \mathbf{q}}-{\mathbf{p}})^{2}+\lambda^2}-\frac{{\rho_d}}{2}\frac{1}{({\mathbf{q}}-{\mathbf{p}})^2+\lambda^2}\right]\end{gathered}\ ] ] for the pure coulomb scattering equation ( fig .
[ fig : pd - inteq - coulomb ] ) , where @xmath165 corresponds to the loop integral in the diagram with the photon attached to a nucleon bubble .
the expression looks quite complicated , but it can be simplified .
the dominant terms of @xmath166 are those with @xmath167 and @xmath168 , respectively , due to the prefactors of @xmath169 and @xmath170 . in the latter case
we can furthermore assume that @xmath171 because of the pole at this position in the propagator . furthermore inserting the total center - of - mass energy @xmath172
, we get @xmath173 and @xmath174 [ eq : bubble - approx ] where we have used and the expansion @xmath175 .
the same simplifications , which effectively amount to keeping only loop contributions with @xmath66 , are used in @xcite and appear to be well supported by comparing the results with experimental data ( see sec . [
sec : scatt - results ] ) . in the doublet channel
we find @xmath176 @xmath177 [ eq : nd - inteq - d - raw ] for the coupled - channel @xmath104@xmath5 equation shown in fig .
[ fig : nd - inteq - d ] , and analogously we have @xmath178 \\
+ \int{{\!\!\frac{\mathrm{d}^3q}{(2\pi)^3}}}\,\delta_d\left(e-\frac{{\mathbf{q}}^2}{2{m_n}},{\mathbf{q}}\right ) \cdot({\mathrm{i}}{{\mathcal{t}}_{\mathrm{full}}}^{\mathrm{a},ik})^{\gamma c}_{\alpha a}(e;{\mathbf{k}},{\mathbf{q } } ) \cdot\biggl\{\frac{{m_n}{y_d}^2}{2}\frac{(\sigma^j\sigma^k)^\beta_\gamma\delta^b_c } { { \mathbf{q}}^2+{\mathbf{q}}\cdot{\mathbf{p}}+{\mathbf{p}}^2-{m_n}e-{\mathrm{i}}{\varepsilon } } \\ + \alpha{m_n}^2{y_d}^2\cdot\delta^{kj}\delta^\beta_\gamma \left(\frac{{\mathbf{1}}+\tau_3}{2}\right)^b_c \left[\frac{\mathcal{i}_\mathrm{bubble}(e;{\mathbf{q}},{\mathbf{p}})}{({\mathbf{q}}-{\mathbf{p}})^{2}+\lambda^2 } -\frac{{\rho_d}}{2}\frac{1}{({\mathbf{q}}-{\mathbf{p}})^2+\lambda^2}\right]\biggr\ } \\
+ \int{{\!\!\frac{\mathrm{d}^3q}{(2\pi)^3}}}\,\delta_t\left(e-\frac{{\mathbf{q}}^2}{2{m_n}},{\mathbf{q}}\right ) \cdot({\mathrm{i}}{{\mathcal{t}}_{\mathrm{full}}}^{\mathrm{b},ic})^{\gamma c}_{\alpha a}(e;{\mathbf{k}},{\mathbf{q } } ) \cdot\frac{{m_n}{y_d}{y_t}}{2}\cdot\frac{(\sigma^j)^\beta_\gamma(\tau^c)^b_c } { { \mathbf{q}}^2+{\mathbf{q}}\cdot{\mathbf{p}}+{\mathbf{p}}^2-{m_n}e-{\mathrm{i}}{\varepsilon } } \label{eq : pd - inteq - d - a - raw}\end{gathered}\ ] ] @xmath179\biggr\ } \label{eq : pd - inteq - d - b - raw}\end{gathered}\ ] ] [ eq : pd - inteq - d - raw ] for the @xmath4@xmath5 equation depicted in fig .
[ fig : pd - inteq - d ] .
in this section we will give a brief derivation of the normalisation condition for the triton wave functions used in sec .
[ sec : he-3 ] . in order to do that we need to introduce a little more formalism . for simplicity
, we work with a simplified nucleon deuteron system , where the virtual spin - singlet state is neglected . the discussion could easily be carried out for the full coupled - channel system , but that would only complicate the notation .
we start by considering the full two - body nucleon deuteron propagator ( green s function ) @xmath180 , which fulfils the ( inhomogeneous ) bethe salpeter equation in momentum space : @xmath181 @xmath182 is essentially a product of a nucleon propagator @xmath183 and a deuteron propagator @xmath184 .
more precisely , we have @xmath185 where @xmath186 is the total four - momentum of the system and @xmath187 .
@xmath188 represents the ( doublet - projected ) one - nucleon exchange diagram , @xmath189 as shown , for example , in fig .
[ fig : nd - inteq - q ] .
we now consider a bound state at rest , _
@xmath192 , and define the _ amputated _ wave function @xmath193^{-1 } \cdot\left[\delta_n\left(-{\eta_ne_b}-p_0,{\mathbf{p}}\right)\right]^{-1 } \ , , \label{eq : bs - amputated}\ ] ] which fulfils the homogeneous equation @xmath194 carrying out the @xmath61-integration picks up the residue from the nucleon propagator pole at @xmath195 . from the resulting right hand side of we then find that @xmath196 fulfils the equation @xmath197 this is essentially the single - channel equivalent of ( before s - wave projection ) , so we have established the connection of our current formalism to the triton wave functions @xmath198 in sec .
[ sec : he-3 ] .
note furthermore that @xmath199 is a schrdinger wave function .
we now write in an operator notation as @xmath200 defining @xmath201 we find @xmath202 where @xmath203 corresponds to the wave function given in , and @xmath204 from this we readily derive the normalisation condition @xmath205 where @xmath206 a straightforward calculation shows that @xmath207^{-1 } \ ,
, \label{eq : g0m1 - 3dbs}\ ] ] and we also see that @xmath208 . the expression for @xmath209 _
a priori _ looks more complicated , but one finds that in the formal expansion @xmath210{{\widetilde{g}}_0}^{-1 } \label{eq : v-3dbs - exp}\ ] ] everything but the term @xmath211 drops out , and we have @xmath212 the essential ingredient to see this is @xmath213 which , in turn , follows from the fact that the nucleon propagator residues are always picked up in such a way that one deuteron propagator is cancelled by the inverse propagator in @xmath214 , _ cf . _
altogether , we have shown that is just the single - channel version of the normalisation condition stated in sec .
[ sec : he-3 ] ( modulo s - wave projection ) , where the functions @xmath215 correspond to @xmath214 .
s. r. beane , w. detmold , k. orginos and m. j. savage , prog .
* 66 * ( 2011 ) 1 [ arxiv:1004.2935 [ hep - lat ] ] . s. r. beane , p. f. bedaque , w. c. haxton , d. r. phillips and m. j. savage , arxiv : nucl - th/0008064 .
p. f. bedaque and u. van kolck , ann .
nucl . part .
* 52 * ( 2002 ) 339 [ arxiv : nucl - th/0203055 ] .
e. epelbaum , h .- w .
hammer and u .- g .
meiner , rev .
* 81 * ( 2009 ) 1773 [ arxiv:0811.1338 [ nucl - th ] ] .
d. b. kaplan , m. j. savage and m. b. wise , nucl .
b * 534 * , ( 1998 ) 329 [ arxiv : nucl - th/9802075 ] .
u. van kolck , nucl .
a * 645 * , ( 1999 ) 273 [ arxiv : nucl - th/9808007 ] .
h. a. bethe , phys . rev .
* 76 * ( 1949 ) 38 .
v. efimov , nucl .
a * 362 * ( 1981 ) 45 .
hammer and l. platter , ann .
* 60 * ( 2010 ) 207 [ arxiv:1001.1981 [ nucl - th ] ] .
x. kong and f. ravndal , phys .
b * 450 * ( 1999 ) 320 [ arxiv : nucl - th/9811076 ] .
x. kong and f. ravndal , nucl .
phys . a * 665 * ( 2000 ) 137
[ arxiv : hep - ph/9903523 ] .
ando , j. w. shin , c. h. hyun and s. w. hong , phys .
c * 76 * ( 2007 ) 064001 [ arxiv:0704.2312 [ nucl - th ] ] .
t. barford and m. c. birse , phys .
c * 67 * ( 2003 ) 064006 [ arxiv : hep - ph/0206146 ] .
ando and m. c. birse , phys .
c * 78 * ( 2008 ) 024004 [ arxiv:0805.3655 [ nucl - th ] ] .
x. kong and f. ravndal , phys .
c * 64 * ( 2001 ) 044002 [ arxiv : nucl - th/0004038 ] .
ando , j. w. shin , c. h. hyun , s. w. hong and k. kubodera , phys .
b * 668 * ( 2008 ) 187 [ arxiv:0801.4330 [ nucl - th ] ] .
v. efimov , phys .
b * 33 * ( 1970 ) 563 . c. a. bertulani , h .- w .
hammer and u. van kolck , nucl .
phys . a * 712 * ( 2002 ) 37 [ arxiv : nucl - th/0205063 ] .
g. rupak and x. kong , nucl .
a * 717 * ( 2003 )
73 [ arxiv : nucl - th/0108059 ] . s. ando and m. c. birse , j. phys .
g : nucl . part .
* 37 * ( 2010 ) 105108 [ arxiv:1003.4383 [ nucl - th ] ] . j. kirscher , h. w. griehammer , d. shukla and h. m. hofmann , eur .
phys . j. a * 44 * ( 2010 ) 239 [ arxiv:0903.5538 [ nucl - th ] ] .
f. bedaque , h .- w .
hammer and u. van kolck , nucl .
phys . a * 676 * ( 2000 ) 357 [ arxiv : nucl - th/9906032 ] .
f. bedaque , g. rupak , h. w. griehammer and h .- w .
hammer , nucl .
a * 714 * ( 2003 ) 589 [ arxiv : nucl - th/0207034 ] .
e. braaten and h .- w .
hammer , phys .
rept . * 428 * ( 2006 ) 259 [ arxiv : cond - mat/0410417 ] .
f. gabbiani , p. f. bedaque and h. w. griehammer , nucl .
phys . a * 675 * ( 2000 ) 601 [ arxiv : nucl - th/9911034 ] .
d. r. phillips and t. d. cohen , phys .
b * 390 * ( 1997 ) 7 [ arxiv : nucl - th/9607048 ] .
j. d. jackson and j. m. blatt , rev .
* 22 * ( 1950 ) 77 .
h. w. griehammer , nucl .
a * 760 * ( 2005 ) 110 [ arxiv : nucl - th/0502039 ] .
l. platter , d. r. phillips , few body syst .
* 40 * ( 2006 ) 35 [ arxiv : cond - mat/0604255 ] . c. ji , d. r. phillips , l. platter , europhys .
* 92 * ( 2010 )
13003 [ arxiv:1005.1990 [ cond-mat.quant-gas ] ] . c. van der leun and c. anderliesten , nucl .
a * 380 * ( 1982 ) , 261 .
j. j. de swart , c. p. f. terheggen and v. g. j. stoks , arxiv : nucl - th/9509032 .
j. r. bergervoet , p. c. van campen , w. a. van der sanden and j. j. de swart , phys .
c * 38 * ( 1988 ) 15 .
j. arvieux , nucl .
phys . a * 221 * ( 1973 ) 253 . | we calculate low - energy proton
deuteron scattering in the framework of pionless effective field theory . in the quartet channel , we calculate the elastic scattering phase shift up to next - to - next - to - leading order in the power counting . in the doublet channel , we perform a next - to - leading order calculation .
we obtain good agreement with the available phase shift analyses down to the scattering threshold .
the phase shifts in the region of non - perturbative coulomb interactions are calculated by using an optimised integration mesh . moreover ,
the coulomb contribution to the @xmath0he@xmath0h binding energy difference is evaluated in first order perturbation theory .
we comment on the implications of our results for the power counting of subleading three - body forces . |
exponentially increasing demand for data rates as well as the spectrum congestion in the lower parts of the electromagnetic spectrum forces the wireless communications industry to explore systems adapted to frequencies within the so - called mmwave band @xcite .
mobile operators are investigating the possibility of using mmwave frequencies for future generation of mobile communications , while migrating the backhaul networks to beyond 30 ghz @xcite .
satellite industry has already established its presence in higher frequency bands by using the @xmath0 ghz band for feeder links as well as fixed - satellite - service ( fss ) user terminals in the downlink , and @xmath1 ghz band for the user terminal uplinks @xcite .
such a migration to tens of ghz bands provides the wireless industry with a much higher bandwidths , and thus enables them to deploy multiple gbps services . however , the development of communications systems in the mmwave bands is a challenging task .
mmwave signals suffer from severe propagation loss , penetration loss and rain fading compared to signals in lower frequencies @xcite .
fortunately , the short wavelength of mmwave frequencies enables the denser packing of more antennas in the transceiver ends which now may employ large array structures for providing high beamforming gains .
furthermore , the large - scale transmit and receive arrays enable pre / post - coding techniques for multiple stream transmission resulting in significant spectral efficiency improvements . even though conventional digital pre / post - coding techniques that were developed in the past years for lower frequency mimo systems are independent of the carrier frequency , they can not be applied in mmwave ones due to high demands in hardware complexity and power consumption .
this is due to the one rf chain per antenna requirement for the implementation of a fully digital technique .
each rf chain includes a number of different electronic elements among which are digital - to - analog / analog - to - digital converters that have high requirements in hardware and power consumption .
thus , the implementation of a digital only mmwave transceiver is impractical and alternative approaches must be sought . a first solution to the problem was given via analog only beamforming approaches @xcite .
the core of these techniques is a network of analog phase shifters @xcite that imposes constant modulus constraints on the beamformer and requires only one rf chain , which is highly desirable from a hardware complexity / power consumption point of view . nevertheless , the major drawback of the analog only approaches is that they can not support multi - stream communication .
therefore , in several cases they perform poorly compared to the fully digital approaches since they can not exploit the inherent spatial multiplexing gain of the channel .
an alternative approach replaces the phase shifters with even simpler components , analog switches @xcite .
however , this results in loss in array gain which again leads to performance degradation , especially for highly correlated channels , like the ones experienced in the mmwave band .
a more efficient solution is based on a two stage hybrid beamformer that consists of a low dimensional digital pre - coder applied in the baseband ( bb ) and an analog beamformer applied in the rf domain @xcite .
the analog processing part is usually implemented via a network of phase shifters .
this architecture can be applied to systems with limited the number of rf chains and achieve performance close to the one of a fully digital approach under some conditions .
the concept of hybrid pre - coding was firstly introduced as the so - called `` antenna soft selection '' for point - to - point mimo systems @xcite .
an interesting result in @xcite showed that for single stream transmission , a hybrid beamformer can realize the corresponding fully digital one provided that both the transmitter and receiver ends are occupied with at least two rf chains .
recent works @xcite generalize this result to multi - stream communications by showing that the required number of chains at both transceiver ends must be two times the number of transmitted streams in order for a hybrid architecture to achieve the optimal digital only solution . in mmwave bands ,
a similar approach to the one of antenna soft selection was introduced by the seminal papers @xcite . in @xcite
, it is shown that the spectral efficiency maximization problem can be closely approximated by minimizing the frobenious norm of the difference of the fully digital solution to the product of the analog - digital counterparts of the hybrid pre - coder .
the analog counterpart is selected from a pre - determined codebook and thus , the resulting problem is simplified by the recent advances in sparse reconstruction signal processing literature @xcite .
the performance of this approach is satisfactory for large array systems that have in general strictly more rf chains than the transmitting streams and for low rank channels . in @xcite , the frobenious norm approximation problem of @xcite
was solved without using a codebook for the analog counterpart solution .
their solution was based on an alternating minimization procedure which it is shown to achieve significantly better performance than the one of @xcite when the number of the available rf chains at both transceiver ends is equal to the number of the transmitting streams . in @xcite , a codebook free approach
is again proposed , though now the analog and digital parts of the beamformer are derived by solving two independent optimization problems after decoupling the original one .
the decoupling of the analog - digital design is based on the assumption that the digital part is approximately orthonormal which in general holds for very large array systems .
the performance of the latter approach is also superior compared to the one of @xcite for equal number of rf chains and transmitting streams . in the literature of mmwave communications cited above
, the wireless system is assumed as a stand - alone mimo communications system .
considering the fact that mmwave bands are also highly allocated to several services ( e.g. point - to - point ( p2p ) or point - to - multipoint ( p2mp ) backhaul microwave links , satellite links , high resolution radar , radio astronomy , amateur radio e.t.c . )
@xcite , such an assumption seems to be very optimistic .
given also the current 5 g projections for traffic increase that imply an increase on backhaul demand @xcite , new generation backhaul networks will have to coexist with older generations since traditional solutions like network planning can not meet the demands .
therefore , it is necessary to take the constraints on the imposed interference on the incumbent services into account while designing the mmwave transceivers .
while several digital only cr transceiver designs have been proposed in literature so far @xcite , a mmwave system can not afford to accommodate the number of rf chains that these techniques require for the reasons discussed above in this section .
this in turn can be alleviated by designing hybrid analog - digital transceivers subject to interference temperature constraints as in the case of traditional underlay cognitive radios .
this is the subject of research in this paper . in detail , the contributions of the present paper are as follows .
two cognitive hybrid analog / digital transceiver designs for point - to - point mimo systems are presented with the view to maximize the spectral efficiency subject to interference to the primary user , power and hardware related ( limited number of rf chains ) constraints .
both of the proposed approaches are based on the underlay cognitive radio paradigm @xcite where the secondary user ( su ) may access a spectrum area licensed to a primary user ( pu ) , simultaneously with the latter and provided that the interference power on the pu transmissions is below a predefined threshold .
the proposed approaches derive the pre - coding and post - coding ( combining ) matrices by decoupling the transmitter receiver optimization problem .
thus , at first the pre - coder is derived such that the system s mutual information is maximized subject to the constraints mentioned previously in this paragraph and then for that pre - coder , the optimal linear post - coding matrix is derived in a minimum mean square error ( mmse ) sense .
the proposed transceiver designs are not based on a codebook and involve the solution of difficult non - convex optimization problems . the developed algorithms for their solution
are based on the so - called alternating direction of multipliers method ( admm ) @xcite which is known for its good behaviour in several non - convex optimization problems @xcite , similar to the ones that are dealt in the present paper .
the first approach derives the analog and digital pre - coding matrices by solving directly the mutual information maximization problem .
as it was discussed before , existing solutions are based on approximations of the original problem or on a decoupled version with respect the analog and digital counterparts . as a consequence ,
the solution derived in the present paper achieves satisfactory performance under any channel model and not only for low rank / highly correlated channels , similar to the ones that are based on the parametric mmwave models .
thus , it may also be successfully applied in large array systems that function in lower frequencies and under typical fading channel models in order to reduce the hardware complexity / power consumption .
moreover , it can be further applied to systems that have relative small number of antennas since it does not require orthonormal properties to be satisfied by the digital pre - coder part , as it is the case for the designs that are based on the decoupled analog - digital problem .
while the first hybrid solution presents excellent performance under any conditions , its complexity is rather high due to the @xmath2 function that appears in the cost function of the mutual information maximization problem . to overcome this
, we follow the approach of @xcite and approximate the solution to the mutual information maximization problem by the one that minimizes the frobenious norm of the difference of the hybrid beamformer to the optimal digital only one . a novel algorithm that requires significantly reduced complexity
is derived to provide the solution to the last problem while it achieves close performance to the one of the first hybrid solution ( mutual information maximization problem ) for typical mmwave environments .
furthermore , a novel solution for the cognitive receiver side post - coder is also developed .
an efficient code - book free hybrid analog / digital post - coder is derived based on the mmse criterion and subject to the hardware constraints related to the phase shifters of the analog processing network .
on the contrary , existing approaches are either codebook based @xcite or are considering the codebook free design of only the hybrid pre - coding matrix at the transmitter side @xcite .
the convergence properties of the proposed algorithms are theoretically studied and their performance is evaluated via simulations for typical channel models . the remainder of the paper is organized as follows . in section ii , the system description is given . in section iii , the description of the optimal digital only solutions is given .
section iv describes the hybrid transceiver design based on the mutual information maximization criterion . in section
v , the alternative hybrid pre - coder design which is based on the frobenious norm approximation criterion , is presented .
section vi presents some numerical results and section vii concludes the work .
let us assume that a secondary user ( su ) @xmath3 mimo system access a spectrum area allocated to a @xmath4 primary user ( pu ) by employing a typical cognitive underlay approach ( fig .
1 ) @xcite . and @xmath5 rf chains at the transmitter s and receiver s side , respectively.,scaledwidth=50.0% ] the su system ( fig .
2 ) is assumed to be equipped with @xmath6 and @xmath7 rf chains at the transmitter and the receiver , respectively and transmits a number of @xmath8 streams .
moreover , the su system applies a @xmath9 hybrid pre - coding matrix at the transmitter given by @xmath10 where @xmath11 is the @xmath12 analog rf precoder implemented via a phase shift network and @xmath13 is the @xmath14 digital bb one . in a similar manner ,
the @xmath15 post - coding matrix is given by @xmath16 where @xmath17 is the @xmath18 analog post - coding matrix and @xmath19 is the @xmath20 digital bb one , respectively .
the received signal @xmath21 at the su receiver before and after post - coding , assuming a narrow - band block fading propagation channel is given by @xmath22 where @xmath23 is the @xmath3 channel matrix between the two su ends , @xmath24 is the @xmath25 vector of the su transmitted symbols such that @xmath26 , @xmath27 is the expectation operator , @xmath28 is the variance of the symbols transmitted by the su transmitter , @xmath29 is the @xmath30 identity matrix , @xmath31 , @xmath32 is the @xmath33 channel matrix between the pu transmitter and the su receiver and @xmath34 is the @xmath35 pre - coding matrix applied at the pu transmitter , @xmath36 is the number of the transmitted pu streams , @xmath37 is the @xmath38 vector of the pu transmitted symbols such that @xmath39 , @xmath40 is the variance of the symbols transmitted by the pu transmitter and @xmath41 is i.i.d .
complex gaussian noise modeled as @xmath42 . note that the su does not require explicit knowledge of matrix @xmath34 since it requires knowledge of only the overall matrix @xmath43 .
moreover , the pu system may also apply a hybrid pre - coding technique in order to optimize its transmission or not ( i.e conventional mimo system ) , though as the proposed technique is independent of the latter , we chose an abstracted description to simplify the analysis . in a similar manner ,
the received signal at the pu receiver is given by @xmath44 where @xmath45 is the @xmath4 channel matrix between the two pu ends , @xmath46 is the @xmath47 channel matrix between the su transmitter and the pu receiver and @xmath48 is i.i.d .
complex gaussian noise modeled as @xmath49 .
the mmwave channel presents limited spatial diversity due to the high free space pathloss and high antenna correlation due to the dense packing of large number of antennas .
a geometrical channel model captures better these characteristics of the mmwave band .
thus , by assuming that each of the involved systems employ an uniform linear array ( ula ) of antennas and @xmath50 propagation paths are between a transmitter @xmath51 and a receiver @xmath52 , the discrete narrow - band channel is given by @xcite@xcite @xmath53 where @xmath54 is the complex gain of the @xmath55th path , @xmath56 and @xmath57 are the azimuth angles of arrival and departure , respectively , that are uniformly distributed in @xmath58 .
vectors @xmath59 and @xmath60 are the normalized receive and transmit response vectors at the azimuth angle of arrival and departure @xmath56 and @xmath61 , respectively .
furthermore , the response of an ula of @xmath62 antenna elements is given by @xmath63,\ ] ] where @xmath64 , @xmath65 is the carrier wavelength and @xmath66 is the inter - element spacing .
let us now assume that the su has perfect knowledge of @xmath23 , @xmath43 and @xmath46 channels .
note that this is a typical requirement in cr underlay approaches @xcite .
in general , since both the su and pu systems are operating in the same frequency and under the channel reciprocity assumption , it is straightforward to estimate the required information related to the cross su - pu channels via the training symbols of the pu system . since this a matter well - studied within the cr literature @xcite we avoid any further discussion .
the aim is to derive the pre / post - coding matrices in order to maximize the spectral efficiency of the su system subject to constraints on the total transmission power ( @xmath67 ) and on the interference to the pu transmissions ( @xmath68 ) . to that end
, the su system must solve the following optimization problem @xmath69 where the spectral efficiency @xmath70 is defined as @xmath71 with @xmath72 denoting the interference plus noise covariance matrix at the su receiver and @xmath73 is the determinant of a matrix .
it is instructive to note that by imposing the interference power constraint on , the capacity loss of the pu due to the su transmissions is well regulated as it is proved in @xcite .
thus , by setting @xmath68 to a sufficiently small value with respect to the noise power @xmath74 , the aforementioned capacity loss can be made arbitrary small . for the sake of simplicity ,
we consider in this paper the case where interference constraints are posed from only one pu . extensions to the multiple pu case are straightforward by adding the corresponding interference related constraints on @xmath75 . prior to presenting the proposed hybrid pre
/ post - coding solutions , the equivalent digital only solution is discussed which will serve as a benchmark to the performance of the aforementioned hybrid approaches .
in the digital only design , the su pre / post - coding matrices can be found by solving directly @xmath75 under the assumption that both the transmitter and the receiver ends have equal number of rf chains and antennas .
this problem is in general intractable @xcite and the common approach is to temporally decouple the designs at the transmitter and the receiver . to that end
, the optimal pre - coding matrix @xmath76 is designed such that the mutual information achieved by gaussian signaling over the wireless channel is maximized .
therefore , the following optimization problem should be solved @xmath77 where @xmath78 denotes a positive semi - definite matrix , @xmath79 is the covariance matrix of the interference plus noise signal and @xmath80 .
in fact , @xmath81 is a whitening post - processing filter that maximizes the mutual information between the input and the output of the su system @xcite .
note that @xmath82 is convex and thus , it can be solved efficiently by standard convex optimization techniques ( i.e. interior point methods @xcite ) .
then , by applying the eigenvalue value decomposition on the optimal matrix @xmath83 , we may find @xmath84 .
problem @xmath82 aims at maximizing the su s channel capacity assuming that the su receiver has full csi and can perform optimal nearest - neighbor decoding based on the @xmath85 dimensional received signal @xcite .
this decoding approach requires high computational complexity even for typical mimo systems that have very few antennas compared to a mmwave one . a common approach that avoids the high complexity of the previous solution is to use a linear receiver design . to that end , in this paper a typical minimum mean square error ( mmse ) receiver is considered for the receive side of the su system .
thus , matrix @xmath86 is derived as the solution to the optimization problem @xmath87 where @xmath88 denotes the frobenius norm .
the solution to the previous problem is given by @xcite @xmath89 where we used - and the assumption that @xmath90 , where @xmath91 is a @xmath92 matrix of zero entries .
in this section first , the hybrid analog - digital transmitter is developed by solving the mutual information maximization problem @xmath75 after casting it to the hybrid case .
then , the hybrid mmse based solution for the post - coding matrix at the receiver is presented . in the case of the hybrid approaches ,
the solution to the su s spectral efficiency maximization problem @xmath75 is even harder to be derived since the constraints related to the phase shifters network force the analog pre-/post - coding matrices to lie in the non - convex space of constant modulus complex matrices . in order to achieve a simple solution
, the transceiver design should be again temporally decoupled .
let us first present the pre - coder design by considering the mutual information maximization problem @xmath82 which is now cast to the hybrid pre - coding case , that is @xmath93 where @xmath94 and @xmath95 for an arbitrary matrix @xmath96 with complex entries @xmath97 .
problem @xmath98 is non - convex due to the bi - convex cost function and the non - convex constraint set @xmath99 . in the following ,
an efficient solution is derived by employing the so - called admm @xcite .
this method is an extension of the augmented lagrangian multiplier method @xcite which first performs an alternating optimization with respect to each one of the involved variables individually and then , it updates the corresponding lagrange multipliers via a gradient based rule .
let us first write @xmath98 in the following equivalent form @xmath100 where @xmath101 is an auxiliary @xmath102 matrix variable , the set @xmath103 for a matrix @xmath96 is defined as @xmath104 and the indicator function @xmath105 of sets @xmath103 , @xmath99 is defined as @xmath106 the augmented lagrangian function of @xmath107 is given by @xmath108 where @xmath109 , for two matrices @xmath96 and @xmath110 , @xmath111 is the @xmath112 lagrange multiplier matrix and @xmath113 is a scalar penalty parameter . following the admm approach ,
the solution to @xmath107 is given by the following alternating minimization steps @xmath114 where @xmath115 is the iteration index .
let us first derive the solution of @xmath116 .
unfortunately , the solution to this problem does not admit a closed form due to its cost function and the constraints on the feasible solution set .
however , it is a convex problem ( with respect to the optimizing matrix parameter @xmath101 ) . therefore , a projected gradient based approach will be developed in the following for its solution .
the gradient of the cost function with respect to @xmath101 is given by @xmath117 then , the projected gradient descent based solution @xmath118 is calculated by updating at each step the solution @xmath119 based on the previous one @xmath120 as , @xmath121 where @xmath122 is a step size parameter and @xmath123 is the projection onto the set @xmath103 operator that can be found by solving the following optimization problem @xmath124 where @xmath96 is an arbitrary matrix and @xmath125 is its projection onto the set @xmath103 . the karush - kuhn - tucker ( kkt ) conditions for @xmath126
are given by , @xmath127 where @xmath128 and @xmath129 are the corresponding multipliers and @xmath130 is the optimal solution . from - it can be shown after some manipulations that operator @xmath131 is given by @xmath132^{-1}\mathbf{a } , \ & \mathbf{s } \notin \mathcal{s } \end{cases } \label{eq : s_proj}\end{aligned}\ ] ] where @xmath128 and @xmath129 are set such that the power ( @xmath67 ) and interference constraints ( @xmath68 ) are met , i.e. via a bisection method @xcite .
the iterations of are running until the following termination criterion is met @xmath133 where @xmath134 is a pre - defined tolerance .
let us move now to the derivation of the solution to @xmath135 which can be written in the following simplified form , @xmath136 the solution to this problem can be given by solving the unconstrained problem and then projecting onto the set @xmath99 .
that is , @xmath137 where @xmath138 is the projection onto the set @xmath99 operator that can be found by solving the following optimization problem @xmath139 where @xmath96 is an arbitrary matrix and @xmath140 is its projection onto the set @xmath99 .
it is straightforward to see that the solution to @xmath141 is given by the phase of the complex elements of @xmath96 , that is for @xmath142 we have @xmath143 where @xmath144 and @xmath97 are the elements at the @xmath145th row - @xmath115th column of matrices @xmath146 and @xmath96 respectively and @xmath147 is the modulus of a complex number .
we move now to the solution of @xmath148 which can be written as @xmath149 by equating the gradient of to zero , we can show that the solution to the corresponding optimization problem admits the following closed form @xmath150 finally , for the lagrange multiplier ( @xmath151 ) we use a gradient update rule , that is @xmath152 according to the termination criteria and the convergence properties of the admm sequence discussed later in this sub - section , the primal feasibility condition of @xmath107 is satisfied in the sense that @xmath153 .
thus , it is possible for the solution given by the admm sequence to violate the power and interference constraints of the original problem @xmath98 . to tackle this , upon convergence of the admm sequence , we project the digital pre - coder solution to the set @xmath154 defined as , @xmath155 where @xmath156 is the solution for the analog pre - coder provided by the admm sequence and
@xmath96 is again , an arbitrary matrix . following a similar procedure to the one that used to derive the projection onto the set @xmath103 -
, it can be shown that for @xmath157 the projection onto @xmath154 is given by @xmath158^{-1}\mathbf{a } , \label{eq : projj}\end{aligned}\ ] ] where the lagrange multipliers @xmath159 and @xmath160 are again set such that the power ( @xmath67 ) and interference constraints ( @xmath68 ) are met .
we may now present some results regarding the convergence of the proposed algorithm under mild conditions . note that in literature so far , strong convergence results for the admm have been derived for convex problems that involve only two blocks of variables .
moreover , strong convergence results for non - convex problems are in general unknown and also an open research problem .
problem @xmath107 involves three blocks of variables and on top of that , it is non - convex .
thus , obtaining strong convergence results is an intractable task and beyond the scopes of the present paper . here
, the following theorem regarding the convergence of the admm sequence - to an optimal point is given .
+ _ theorem 1 _ : let @xmath161 is a sequence generated by the alternating minimization steps - .
let us further assume that the multiplier sequence @xmath162 is bounded and satisfies @xmath163 then , the sequence @xmath164 converges always to an optimal point of @xmath107 .
the proof is given in appendix a. in the final part of this section , the implementation aspects of the proposed approach are discussed . the core of the developed technique is the four admm alternating minimization problems @xmath116-@xmath165 .
, @xmath166 and @xmath167 are initialized with random values .
the lagrange multiplier matrix @xmath168 is initialized with zeros . for the termination criteria of this alternating minimization procedure
we propose the ones given by @xmath169 where @xmath170 and @xmath171 are the corresponding tolerances .
the first termination criterion guarantees the convergence of variable @xmath118 and further the convergence of @xmath116 to its optimal value .
the second one guarantees that the primal feasibility condition of @xmath172 is satisfied .
as it can be seen by the optimality conditions of @xmath172 , the convergence of the previous two quantities imply the convergence of the admm sequence to an optimal point of @xmath172 ( check also the proof of theorem 1 at the appendix ) .
initialize @xmath173 , @xmath174 , @xmath175 with random values and @xmath176 with zeros @xmath177 update @xmath119 from @xmath178 update @xmath166 , @xmath167 and @xmath179 from - , and @xmath166 , @xmath180 note that due to the non - convex nature of the latter problem , it is also useful to add a termination criterion related to the maximum permitted number of iterations of the admm sequence which here is defined as @xmath181 .
we close this subsection by discussing the implementation aspects of @xmath116 which is solved by the iterative projected gradient technique of .
the required complexity can be reduced by smooth starting the gradient descent part of the admm s @xmath115th iteration by setting @xmath182 .
furthermore , the tolerance parameter @xmath134 can be set to a larger value at the beginning of the admm sequence and decrease its value as the latter converges for better accuracy @xcite@xcite .
a good rule of thumb is to decrease tolerance @xmath134 by a power of 10 every time the quantities of reach its value .
the complete procedure is given for reference in algorithm 1 .
let us now move to the derivation of the post - coding matrix . for complexity issues , here we assume a typical linear receiver based on the mmse criterion , as it was also discussed in section iii .
based on the constraints related to the structure of the phase only analog counterpart , we may express the corresponding optimization problem as @xmath183 in the absence of hardware limitations that pose the phase only constraints on the receiver , one may find the corresponding optimal digital only post - coder by solving again the unconstrained problem of given that the hybrid pre - coding solutions of algorithm 1 are applied on the transmitter side .
thus , by denoting with @xmath184 the optimal digital only mmse post - coder , it is straightforward to see from that , @xmath185 where we used the following equations @xmath186 returning now to the non - convex problem @xmath187 , we apply the methodology of @xcite , @xcite related to the design of linear mmse estimators with complex structural constraints in order to derive the following equivalent form , @xmath188 that is , the optimal hybrid mmse can be found as the weighted projection of the unconstrained optimal digital mmse onto the set of post - coders that admit the form @xmath189 with @xmath190 .
the proposed solution is based again on the admm by expressing the optimization problem of in the following form @xmath191 where @xmath192 is an auxiliary @xmath193 matrix variable and the indicator function of set @xmath99 is defined in .
the augmented lagrangian function of @xmath194 is given by @xmath195 where @xmath196 is the @xmath197 lagrange multiplier matrix and @xmath198 is a scalar penalty parameter .
following the admm approach , the solution to the optimization problems @xmath194 is given by the following alternating minimization steps initialize @xmath199 , @xmath200 , @xmath201 with random values and @xmath202 with zeros update @xmath203 , @xmath204 , @xmath205 and @xmath206 from - , respectively @xmath204 , @xmath205 @xmath207 where @xmath115 is the iteration index .
let us now derive the solution to @xmath208-@xmath209 .
problem @xmath208 can be directly solved by equating the gradient of the augmented lagrangian to zero .
thus , we have @xmath210 problem @xmath211 is equivalent to @xmath135 through which the analog pre - coder @xmath166 is updated . therefore , it is straightforward to show that the solution to @xmath211 is given by @xmath212 where the projection operator @xmath138 is defined in . in a similar manner , the solution of @xmath213 is equivalent to the one of @xmath148 for the update of matrix @xmath167 .
thus , it can be shown that @xmath214 finally , for problem @xmath165 , a gradient update rule is again used , that is @xmath215 termination criteria similar to the ones of problems @xmath116-@xmath165 may be derived under the same arguments .
therefore , the conditions that must be met are @xmath216 where @xmath217 and @xmath218 are the corresponding pre - defined tolerances .
convergence analysis results can also be derived by following the proof given in appendix for theorem 1 provided that the multiplier sequence @xmath219 is bounded and satisfies @xmath220 .
we omit any further reference to avoid repetition .
the procedure is summarized in algorithm 2 .
as it is evident from subsection iv.a , problem @xmath98 requires quite high complexity to be solved due to the involved @xmath2 function , since problem @xmath116 of the admm sequence via which the solution is derived , does not admit a closed form .
thus , its solution is iteratively calculated by the projected gradient technique in every time step of the admm sequence , as described in subsection iv.a .
therefore , in order to improve the complexity requirements , we need to seek for more efficient solutions to tackle this problem . to that end
, the approach of @xcite is followed where the hybrid pre - coder is designed such that the frobenious norm of its difference to the optimal digital only solution is minimized . for typical point - to point mimo systems in mmwave band , this approach is shown to perform satisfactory @xcite@xcite , while it requires significantly reduced computational complexity . therefore , the hybrid pre - coder is derived as the solution to the optimization problem @xmath221 where the optimal digital only solution @xmath76 is computed by solving @xmath82 in section iii .
following once more the admm approach , problem @xmath222 can be cast in the following form @xmath223 where @xmath224 is an auxiliary @xmath102 matrix variable .
the augmented lagrangian function of @xmath225 is given by @xmath226 where @xmath227 is the corresponding lagrange multiplier and @xmath228 is a scalar penalty parameter .
the admm sequence for the solution of @xmath225 is given by @xmath229 problem @xmath230 can be solved by setting the gradient of the augmented lagrangian with respect the variable @xmath224 to zero and then projecting the result onto the set @xmath103 .
that is , @xmath231 moving on , problems @xmath232 are equivalent to @xmath233 ones .
thus , in a similar manner it can be shown that @xmath234 and @xmath235 finally , the gradient update rule @xmath236 is used to update the value of the lagrange multiplier . from - , it is clear that this approach involves closed form solutions for problems @xmath230-@xmath237 and thus requires significantly reduced complexity compared to the one of @xmath107 . for the receiver side ,
algorithm 2 can be again used to provide the hardware constrained mmse based solution .
termination criteria similar to the ones of problems @xmath116-@xmath165 and @xmath208-@xmath209 may be derived under the same arguments .
therefore , the conditions that must be met are @xmath238 compute @xmath76 by solving @xmath82 as described in sec .
iii initialize @xmath239 , @xmath174 , @xmath175 with random values and @xmath240 with zeros update @xmath241 , @xmath166 , @xmath167 and @xmath242 from - @xmath166 , @xmath180 where @xmath243 and @xmath244 are the corresponding pre - defined tolerances .
convergence analysis results can be derived by assuming that the lagrange multiplier sequence @xmath245 is bounded and satisfies @xmath246 , similar to problems @xmath107 and @xmath194 .
any further details are omitted in order to avoid repetitions .
the procedure is summarized in algorithm 3 .
in this section , numerical results are presented for evaluating the performance of the proposed hybrid approaches .
an environment of @xmath247 propagation paths is assumed for all the involved links ( su - to su , the su - to - pu and pu - to - su ) and the element spacing on the ula of each node is set to @xmath248 .
the performance of the proposed approaches is examined in terms of the achieved mean spectral efficiency versus the signal - to - noise - ratio ( snr ) over 100 channel realizations .
let us now refer to the parameter tuning of the proposed algorithms .
the following numbers are the same for all the experiments presented in this paper . for algorithm 1 ,
the parameters are set as @xmath249 , @xmath250 , @xmath251 , @xmath252 and @xmath253 ( initial value which is updated based on the convergence of the admm sequence as described in section iv.a ) . for algorithm 2 ,
the parameters are set as @xmath254 , @xmath255 and @xmath256 .
finally , for algorithm 3 , the parameters are set as @xmath257 , @xmath258 and @xmath259 .
the maximum number of iterations per admm sequence is set to @xmath260 for all the previous algorithms . for comparison purposes ,
we also plot in the following figures a ) the performance of the optimal digital only transceiver which is derived as described in section iii , and b ) the codebook based approach of @xcite which is adapted to the cognitive scenario proposed here .
this technique aims at minimizing the distance of the hybrid beamformer to the optimal digital one by solving a sparse reconstruction problem .
mimo system and @xmath261 , scaledwidth=70.0% ] the analog counterpart is determined by a codebook and the digital one is computed by applying the well - known orthogonal matching pursuit ( omp ) technique . in the cognitive radio setup examined here ,
the pre - coding part is derived by first applying this approach to approximate the solution of @xmath82 and then , the digital counterpart is projected onto @xmath154 , so as to force the hybrid solution to satisfy the interference and power constraints .
the post - coding matrix at the receiver is derived by applying directly the approach of @xcite to minimize the distance to the optimal mmse receiver of by using - in place of the corresponding values in the original approach . in figure 3 , the performance of the different hybrid techniques is examined for a su @xmath262 mimo system with @xmath263 rf chains . as it was discussed above , the performance of the optimal digital only solution is plotted for @xmath264 for a fair comparison , since the hybrid approaches can support transmission of maximum @xmath265 streams . the maximum transmission power and interference constraints are set to @xmath266 and @xmath267 , respectively .
the pu is assumed to transmit a signal of rank @xmath268 . from the results depicted in figure 3 ,
it is evident that both of the proposed hybrid approaches achieve close performance to the one of the digital only solution for all the examined cases .
the one based on the mutual information maximization problem ( `` algorithms 1 & 2 '' ) achieves better performance compared to the one that minimizes the distance of the hybrid pre - coder to digital only solution ( `` algorithms 3 & 2 '' ) .
moreover note that as the snr increases , the gap in the performance of these two approaches increases too .
this can be explained by the fact that at low snr values , the performance is mainly dominated by noise and thus the approximation of the original problem by the solution of algorithms 3 & 2 has lesser impact on the system s performance .
mimo system and @xmath261 , scaledwidth=70.0% ] on the contrary , for high snr values , the noise has negligible impact on the performance and thus the degradation due to the aforementioned approximation is more profound . a final point in figure 3
is related to the performance of the technique in @xcite which is plotted for all the considered cases .
as it is evident , the performance of that technique is severely inferior to the proposed ones , especially for high snr values .
this is due to the codebook use that restricts the solution set of the analog counterpart , as it was also observed in @xcite and @xcite .
similar results are also observed in figure 4 where the experimental set - up is the same with figure 3 , though now a @xmath269 su system is assumed .
all the techniques benefit from the larger number of antennas and achieve improved performance .
the proposed hybrid techniques achieve again performance very close to the one of the digital only solution for systems with larger number of antenna elements .
it is also noteworthy that the performance gap between the proposed techniques and the one in @xcite is smaller for larger array systems though still quite large , especially for high snr values . in figure 5 , we examine the impact of the pu signal s rank on the performance of the proposed techniques .
the number of rf chains is fixed to @xmath270 and @xmath267 .
we consider pu signals of rank @xmath271 respectively .
the rest of the parameters are the same with the ones of the experiments of figures 3 and 4 .
as it was expected , the performance of all the techniques is degraded as the rank of pu signal s increases . mimo system and @xmath272 , scaledwidth=70.0% ] this can be attributed to the fact that more of the available degrees of freedom are occupied by the pu , as its signal s rank increases and thus , the su has to reduce further its transmitted power in order to satisfy the interference temperature constraint .
both of the proposed hybrid approaches achieve once again close performance to the one of the digital only optimal solution , independently of the pu signal s rank . on the contrary
, the performance of the approach in @xcite is degraded with an increase on the pu signal s of rank . finally , in figure 6
, the impact of the interference constraint on the performance of the proposed techniques is examined .
the experimental set - up is the same with the one of figure 5 , though now the values @xmath273 are considered and the pu signal s rank is fixed to @xmath274 . under both cases ,
the proposed hybrid techniques achieve once more close performance to the one of the digital only case . on the other hand ,
the approach in @xcite exhibits a severe degradation in the performance for low @xmath68 values , as it shown in the same figure .
in this paper , two novel hybrid analog / digital transceiver designs were proposed for cr large - scale antenna systems . the proposed approaches
aim at maximizing the spectral efficiency of the su while keeping the interference to the pu bellow a pre - defined threshold .
both of the approaches decouple the transmitter - receiver optimization problem and are based on the so - called admm .
the first technique derives the optimal pre - coder by maximizing the mutual information between the transceiver s ends .
the post - coding matrix at the receiver side is derived by a novel hardware constrained mmse approach .
the second technique aims at reducing the required computational complexity by deriving the optimal pre - coding matrix at the transmitter by minimizing the distance ( with respect to the frobenious norm ) of the hybrid solution to the optimal digital one . at the receiver side ,
the post - coding matrix is based on the same solution to the one of the first technique .
the second approach requires reduced complexity as it replaces the highly complex mutual information maximization problem with a simpler least squares one .
simulations show that both the proposed approaches achieve close performance to the one of the corresponding digital only transceiver while having less requirements in hardware and power consumption than the latter .
future works that are currently under development include the extension of the proposed approach to partially connected hybrid solutions and frequency selective channels .
the proof is divided into two parts . at first
, we show that given the convergence of the lagrange multiplier sequence , as it is defined in , the sequence @xmath164 has always a limit point . to that end , we begin by showing that the augmented lagrangian function given in is bounded from below .
the equivalent scaled form @xcite of the latter function is given by @xmath276 the claim then is justified by considering the boundness of @xmath277 in .
moreover , the optimization problems @xmath116 , @xmath148 and @xmath165 are convex with respect the optimizing variables and hence , @xmath118 , @xmath167 and @xmath168 are the corresponding unique minimizers .
problem @xmath135 is non - convex , though set @xmath99 is closed and bounded and thus the sequence @xmath166 converges always to a limit point .
let us further assume that variable @xmath113 is set to a large value such that the augmented lagrangian function is strongly convex with respect the optimizing variable in all the @xmath278 .
due to the aforementioned strong convexity of the lagrange function the following equations hold @xcite @xmath279 where @xmath280 denotes the sub - gradient with respect @xmath96 .
+ moreover , since @xmath281 , @xmath282 and @xmath283 are the minimizers of @xmath284 , @xmath285 and @xmath286 respectively , we have @xmath287 by combining - with - we may show that @xmath288 where @xmath289 . taking the summation of the last equality and considering that @xmath290 is bounded from below we have , @xmath291 since the second term is bounded by assumption
, we can deduce that @xmath292 which implies the existence of the limit point of the admm sequence in the sense that @xmath293 we now move to the second part of the proof where the aim is to show that any limit point of the admm sequence - is a local optimal point of @xmath107 . by the kkt conditions ,
an optimal point @xmath294 of @xmath107 should satisfy the set of equations given by @xmath295 now , from the optimality conditions of - we equivalently have @xmath296 since convergence of the admm sequence is guaranteed , we have that @xmath297 , as @xmath298 , where @xmath299 is the corresponding limit point . furthermore , as @xmath298 , @xmath300 and @xmath301 from and , respectively . the previous with their turn
imply that @xmath302 by taking the limit of - and applying in order to simplify the results we get @xmath303 by comparing - to - , it is straightforward to see that the limit point @xmath304 satisfies the kkt conditions and thus it is optimal .
@xmath305 t. s. rappaport , s. sun , r. mayzus , h. zhao , y. azar , k. wang , g. n. wong , j. k. schulz , m. samimi , and f. gutierrez , `` millimeter wave mobile communications for 5 g cellular : it will work ! '' _
ieee access _ ,
vol . 1 , pp . 335349 , 2013 .
s. k. yong and c .- c .
chong , `` an overview of multigigabit wireless through millimeter wave technology : potentials and technical challenges , '' _ eurasip journal on wireless communications and networking _ , vol .
2007 , no . 1 ,
110 , 2006 .
s. maleki , s. chatzinotas , b. evans , k. liolis , j. grotz , a. vanelli - coralli , and n. chuberre , `` cognitive spectrum utilization in ka band multibeam satellite communications , '' _ ieee communications magazine _ , vol .
53 , no . 3 , pp .
2429 , march 2015 .
t. s. rappaport , g. r. maccartney , m. k. samimi , and s. sun , `` wideband millimeter - wave propagation measurements and channel models for future wireless communication system design , '' _ ieee transactions on communications _
63 , no . 9 , pp . 30293056 , sept 2015 .
t. baykas , c. s. sum , z. lan , j. wang , m. a. rahman , h. harada , and s. kato , `` ieee 802.15.3c : the first ieee wireless standard for data rates over 1 gb / s , '' _ ieee communications magazine _ , vol .
49 , no . 7 , pp .
114121 , july 2011 .
j. wang _ et al .
_ , `` beam codebook based beamforming protocol for multi - gbps millimeter - wave wpan systems , '' _ ieee journal on selected areas in communications _ , vol
27 , no . 8 ,
13901399 , october 2009 .
s. hur , t. kim , d. j. love , j. v. krogmeier , t. a. thomas , and a. ghosh , `` millimeter wave beamforming for wireless backhaul and access in small cell networks , '' _ ieee transactions on communications _ , vol .
61 , no .
10 , pp . 43914403 , october 2013 .
o. e. ayach , r. w. heath , s. abu - surra , s. rajagopal , and z. pi , `` the capacity optimality of beam steering in large millimeter wave mimo systems , '' in _ 2012 ieee 13th international workshop on signal processing advances in wireless communications ( spawc ) _ , june 2012 , pp .
100104 .
a. gorokhov , d. a. gore , and a. j. paulraj , `` receive antenna selection for mimo spatial multiplexing : theory and algorithms , '' _ ieee transactions on signal processing _ ,
51 , no . 11 , pp . 27962807 , nov 2003 .
p. sudarshan , n. b. mehta , a. f. molisch , and j. zhang , `` channel statistics - based rf pre - processing with antenna selection , '' _ ieee transactions on wireless communications _
, vol . 5 , no . 12 , pp . 35013511 ,
december 2006 .
o. e. ayach , s. rajagopal , s. abu - surra , z. pi , and r. w. heath , `` spatially sparse precoding in millimeter wave mimo systems , '' _ ieee transactions on wireless communications _ ,
vol . 13 , no . 3 , pp .
14991513 , march 2014 .
x. zhang , a. f. molisch , and s .- y .
kung , `` variable - phase - shift - based rf - baseband codesign for mimo antenna selection , '' _ ieee transactions on signal processing _ , vol .
53 , no . 11 , pp .
40914103 , nov 2005 . j. ahmadi - shokouh , s. h. jamali , and s. safavi - naeini , `` optimal receive soft antenna selection for mimo interference channels , '' _ ieee transactions on wireless communications _
, vol . 8 , no . 12 , pp .
58935903 , december 2009 .
r. mendez - rial , c. rusu , n. gonzalez - prelcic , and r. w. heath , `` dictionary - free hybrid precoders and combiners for mmwave mimo systems , '' in _ signal processing advances in wireless communications ( spawc ) , 2015 ieee 16th international workshop on _ , june 2015 , pp .
151155 .
a. alkhateeb , o. e. ayach , g. leus , and r. w. heath , `` channel estimation and hybrid precoding for millimeter wave cellular systems , '' _ ieee journal of selected topics in signal processing _ ,
vol . 8 , no . 5 , pp .
831846 , oct 2014 .
x. yu , j. c. shen , j. zhang , and k. b. letaief , `` alternating minimization algorithms for hybrid precoding in millimeter wave mimo systems , '' _ ieee journal of selected topics in signal processing _ ,
10 , no . 3 , pp . 485500 , april 2016 .
m. d. mueck , i. karls , r. arefi , t. haustein , and w. keusgen , `` licensed shared access for wave cellular broadband communications , '' in _ cognitive cellular systems ( ccs ) , 2014 1st international workshop on _ , sept 2014 , pp .
r. zhang and y. c. liang , `` exploiting multi - antennas for opportunistic spectrum sharing in cognitive radio networks , '' _ ieee journal of selected topics in signal processing _ ,
vol . 2 , no . 1 , pp .
88102 , feb 2008 .
s. m. perlaza , n. fawaz , s. lasaulce , and m. debbah , `` from spectrum pooling to space pooling : opportunistic interference alignment in mimo cognitive networks , '' _ ieee transactions on signal processing _ , vol .
58 , no . 7 , pp .
37283741 , july 2010 . c. g. tsinos and k. berberidis ,
`` blind opportunistic interference alignment in mimo cognitive radio systems , '' _ ieee journal on emerging and selected topics in circuits and systems _ , vol . 3 , no . 4 , pp .
626639 , dec 2013 .
s. k. sharma , s. chatzinotas , and b. ottersten , `` interference alignment for spectral coexistence of heterogeneous networks , '' _ eurasip journal on wireless communications and networking _ , vol .
2013 , no . 1 , pp . 114 , 2013 .
a. goldsmith , s. a. jafar , i. maric , and s. srinivasa , `` breaking spectrum gridlock with cognitive radios : an information theoretic perspective , '' _ proceedings of the ieee _ , vol .
97 , no . 5 , pp .
894914 , may 2009 .
s. boyd , n. parikh , e. chu , b. peleato , and j. eckstein , `` distributed optimization and statistical learning via the alternating direction method of multipliers , '' _ foundations and trends in machine learning _ ,
vol . 3 , no . 1 , pp . 1122 , 2011 .
m. webb , m. beach , and a. nix , `` capacity limits of mimo channels with co - channel interference , '' in _ vehicular technology conference , 2004 .
vtc 2004-spring .
2004 ieee 59th _ , vol . 2 , may 2004 , pp .
703707 vol.2 .
j. eckstein and d. p. bertsekas , `` on the douglas
rachford splitting method and the proximal point algorithm for maximal monotone operators , '' _ mathematical programming _ ,
55 , no . 1 - 3 , pp .
293318 , 1992 . | milimeter wave ( mmwave ) band mobile communications can be a solution to the continuously increasing traffic demand in modern wireless systems .
even though mmwave bands are scarcely occupied , the design of a prospect transceiver should guarantee the efficient coexistence with the incumbent services in these bands . to that end , in this paper , two underlay cognitive transceiver designs are proposed that enable the mmwave spectrum access while controlling the interference to the incumbent users .
mmwave systems usually require large antenna arrays to achieve satisfactory performance and thus , they can not support fully digital transceiver designs due to high demands in hardware complexity and power consumption .
thus , in order to develop efficient solutions , the proposed approaches are based on a hybrid analog - digital pre - coding architecture . in such hybrid designs
, the overall beamformer can be factorized in a low dimensional digital counterpart applied in the baseband and in an analog one applied in the rf domain .
the first cognitive solution developed in this paper designs the cognitive hybrid pre - coder by maximizing the mutual information between its two ends subject to interference , power and hardware constraints related to the analog counterpart .
the second solution aims at reduced complexity requirements and thus derives the hybrid pre - coder by minimizing the frobenious norm of its difference to the optimal digital only one . a novel solution for the post - coder at the cognitive receiver part
is further proposed here based on a hardware constrained minimum mean square error criterion .
simulations show that the performance of both the proposed hybrid approaches is very close to the one of the fully digital solution for typical wireless environments .
submitted on the ieee transactions on signal processing millimeter wave ( mmwave ) , cognitive radio , multiple - input multiple - output ( mimo ) , large - scale antenna arrays , hybrid pre - coding , alternating direction method of multipliers ( admm ) |
estimating covariance matrices is of fundamental importance for an abundance of statistical methodologies . nowadays
, the advance of new technologies has brought massive high - dimensional data into various research fields , such as fmri imaging , web mining , bioinformatics , climate studies and risk management , and so on .
the usual sample covariance matrix is optimal in the classical setting with large samples and fixed low dimensions @xcite , but it performs very poorly in the high - dimensional setting @xcite . in the recent literature ,
regularization techniques have been used to improve the sample covariance matrix estimator , including banding @xcite , tapering @xcite and thresholding @xcite . banding or tapering is very useful when the variables have a natural ordering and off - diagonal entries of the target covariance matrix decays to zero as they move away from the diagonal . on the other hand ,
thresholding is proposed for estimating permutation - invariant covariance matrices .
thresholding can be used to produce consistent covariance matrix estimators when the true covariance matrix is bandable @xcite . in this sense
, thresholding is more robust than banding / tapering for real applications .
let @xmath1 be the sample covariance matrix .
@xcite defined the general thresholding covariance matrix estimator as @xmath2 where @xmath3 is the generalized thresholding function .
the generalized thresholding function covers a number of commonly used shrinkage procedures , e.g. the hard thresholding @xmath4 , the soft thresholding @xmath5 , the smoothly clipped absolute deviation thresholding @xcite and the adaptive lasso thresholding @xcite .
consistency results and explicit rates of convergence have been obtained for these regularized estimators in the literature , e.g. @xcite , @xcite , @xcite , @xcite . the recent work by @xcite has established the minimax rate of convergence under the @xmath0 matrix norm over a fairly wide range of classes of large covariance matrices , where the thresholding estimator is shown to be minimax rate optimal .
the existing theoretical and empirical results show no clear favoritism to a particular thresholding rule . in this paper
we focus on the soft - thresholding because it can be formulated as the solution of a convex optimization problem .
let @xmath6 be the frobenius norm and @xmath7 be the element - wise @xmath0-norm of all non - diagonal elements .
then the soft - thresholding covariance estimator is equal to [ soft - thresholding ] = _ 2-_n_f^2+||_1 . however , there is no guarantee that the thresholding estimator is always positive definite .
although the positive definite property is guaranteed in the asymptotic setting with high probability , the actual estimator can be an indefinite matrix , especially in real data analysis . to illustrate this issue
, we consider the michigan lung cancer gene - expression data @xcite which have @xmath8 tumor samples from patients with lung adenocarcinomas and @xmath9 gene expression values for each sample .
more details about this dataset are referred to @xcite and @xcite .
we randomly choose @xmath10 genes ( @xmath11 ) , and obtain the soft - thresholding sample correlation matrix for these genes .
we repeat the process ten times for @xmath12 and @xmath13 respectively , and each time the thresholding parameter @xmath14 is selected via the 5-fold cross validation .
we found that none of the soft - thresholding estimators would become positive definite for both @xmath12 and @xmath13 . on average , there exist @xmath15 and @xmath16 negative eigenvalues for the soft - thresholding estimator for @xmath12 and @xmath17 , respectively .
figure [ plot : michigan_data ] displays the @xmath18 smallest eigenvalues for @xmath12 and the @xmath19 smallest eigenvalues for @xmath17 .
[ plot : michigan_data ] to deal with the indefiniteness , one possible solution is to utilize the eigen - decomposition of @xmath20 , and project @xmath20 into the convex cone @xmath21 .
assume that @xmath20 has the eigen - decomposition @xmath22 , and then a positive semidefinite estimator @xmath23 can be obtained by setting @xmath24 .
however , this strategy does not work well for sparse covariance matrix estimation , because the projection destroys the sparsity pattern of @xmath20 .
consider the michigan data again .
after semidefinite projection , the soft - thresholding estimator has no zero entry . in order to simultaneously achieve sparsity and positive semidefiniteness ,
a natural solution is to add the positive semidefinite constraint to ( [ soft - thresholding ] ) .
consider the following constrained @xmath0 penalization problem [ soft-thresholding.plus0 ] ^+=_0 -_n_f^2/2+||_1 .
note that the solution to ( [ soft-thresholding.plus0 ] ) could be positive semidefinite . to obtain a positive definite covariance estimator
, we can consider the positive definite constraint @xmath25 for some arbitrarily small @xmath26 .
then the modified @xmath27 is always positive definite . in this work ,
we focus on solving the positive definite @xmath27 as follows [ soft-thresholding.plus ] ^+= _ -_n_f^2/2+||_1 . despite its natural motivation , ( [ soft-thresholding.plus ] )
is actually a very challenging optimization problem due to the positive semidefinite constraint . to our best knowledge ,
the first attempt for solving ( [ soft-thresholding.plus ] ) was recently proposed by @xcite who added the log - determinant barrier function to ( [ soft-thresholding.plus ] ) : [ rothman ] ^+=_0 -_n_f^2/2-()+||_1 , where the barrier parameter @xmath28 is a small positive constant , say @xmath29 . from the optimization viewpoint , ( [ rothman ] ) is similar to the graphical lasso criterion @xcite which also has a log - determinant part and the element - wise @xmath0-penalty .
@xcite derived an iterative procedure to solve ( [ rothman ] ) .
@xcite s proposal is based on heuristic arguments and its convergence property is unknown .
in this paper we present an alternating direction algorithm for solving ( [ soft-thresholding.plus ] ) directly .
numerical examples show that our algorithm is much faster than the log - barrier method .
we further prove the convergence properties of our algorithm and discuss the statistical properties of the positive - definite constrained @xmath0 penalized covariance estimator .
we use an alternating direction method to solve ( [ soft-thresholding.plus ] ) directly .
the alternating direction method is closely related to the operator - splitting method that has a long history back to 1950s for solving numerical partial differential equations , see e.g. , @xcite .
recently , the alternating direction method has been revisited and successfully applied to solving large scale problems arising from different applications .
for example , @xcite introduced the alternating linearization methods to efficiently solve the graphical lasso optimization problem .
we refer to @xcite for more details on operator - splitting and alternating direction methods . in the sequel , we propose an alternating direction method to solve the @xmath0 penalized covariance matrix estimation problem ( [ soft-thresholding.plus ] ) under the positive - semidefinite constraint .
we first introduce a new variable @xmath30 and an equality constraint as follows [ soft-thresholding.alm ] ( ^+,^+)= _ , \{-_n_f^2/2+||_1 : = , } . the solution to ( [ soft-thresholding.alm ] )
gives the solution to ( [ soft-thresholding.plus ] ) . to deal with the equality constraint in ( [ soft-thresholding.alm ] )
, we shall minimize its augmented lagrangian function for some given penalty parameter @xmath31 , i.e. [ auglagfunction ] l ( , ; ) = -_n_f^2/2+||_1 - , -+-_f^2/(2 ) , where @xmath32 is the lagrange multiplier .
we iteratively solve [ subproblem1 ] ( ^i+1,^i+1)=l(,;^i)and then update the lagrangian multiplier @xmath33 by @xmath34 for ( [ subproblem1 ] ) we do it by alternatingly minimizing @xmath35 with respect to @xmath30 and @xmath36 . to sum up , the entire algorithm proceeds as follows : : : for @xmath37 , solve the following three sub - problems sequentially till convergence ^i+1 = _
l(,^i;^i ) [ alg : adm-1 ] ^i+1 = _
l(^i+1,;^i ) [ alg : adm-2 ] ^i+1=^i-(^i+1-^i+1)/. [ alg : adm-3 ] to further simplify the alternating direction algorithm , we derive the closed - form solutions for ( [ alg : adm-1])([alg : adm-2 ] ) . consider the @xmath30 step .
define @xmath38 as the projection of a matrix @xmath39 onto the convex cone @xmath40 .
assume that @xmath39 has the eigen - decomposition @xmath41 , and then @xmath38 can be obtained as @xmath42 .
then the @xmath30 step can be analytically solved as follows ^i+1 & = & _ l(,^i;^i ) + & = & _ -^i,+-^i_f^2/(2 ) + & = & ( ^i+^i)_+ .
next , define an entry - wise soft - thresholding rule for all the non - diagonal elements of a matrix @xmath39 as @xmath43 with @xmath44 then the @xmath36 step has a closed - form solution given below ^i+1 & = & _ l(^i+1,;^i ) + & = & _ -_n_f^2/2+||_1 + ^i,+-^i+1_f^2/(2 ) + & = & \{((_n-^i)+^i+1 , ) } /(1 + ) .
algorithm 1 shows the complete details of our alternating direction method for ( [ soft-thresholding.plus ] ) . in section 4
we provide the convergence analysis of algorithm 1 and prove that algorithm 1 always converges to the optimal solution of from any starting point .
[ alg : adm-1-formal ] 1 .
input : @xmath31 , @xmath45 and @xmath46 .
2 . iterative alternating direction augmented lagrangian step : for the @xmath47-th iteration 1 . solve @xmath48 ; 2 .
solve @xmath49 ; 3 .
update @xmath50 .
3 . repeat the above cycle till convergence .
in our implementation we use the soft - thresholding estimator as the initial value for both @xmath51 and @xmath45 , and we set @xmath46 as a zero matrix .
the value for @xmath31 is 2 . before invoking algorithm 1
, we always check whether the soft - thresholding estimator is positive definite .
if yes , then the soft - threhsolding estimator is the final solution to ( [ soft-thresholding.plus ] ) .
before delving into theoretical analysis of the algorithm and the resulting estimator , we first use simulation to show the competitive performance of our proposal . in all examples
we standardize the variables to have zero mean and unit variance . in each simulation model
, we generated @xmath52 independent datasets , each with @xmath53 independent @xmath10-variate random vectors from the multivariate normal distribution with mean @xmath54 and covariance matrix @xmath55 for @xmath56 .
we considered two covariance models with different sparsity patterns : model 1 : : : @xmath57 .
model 2 : : : partition the indices @xmath58 into @xmath59 non - overlapping subsets of equal size , and let @xmath60 denote the maximum index in @xmath61 .
@xmath62 model 1 has been used in @xcite and @xcite , and model 2 is similar to the overlapping block diagonal design used in @xcite .
first , we compare the run times of our estimator @xmath63 with the log - barrier estimator @xmath64 by @xcite . as shown in table [
sim : timing ] , our method is much faster than the log - barrier method . in what follows ,
we compare the performance of @xmath63 , @xmath64 and the soft - thresholding estimator @xmath20 .
for all three regularized estimators , the thresholding parameter was chosen by @xmath65-fold cross - validation @xcite .
the estimation performance is measured by the average losses under both the frobenius norm and the spectral norm .
the selection performance is examined by the false positive rate @xmath66 and the true positive rate @xmath67 moreover , we compare the average number of negative eigenvalues and the percentage of positive - definiteness to check the positive - definiteness .
table [ stcor : model1 ] and table [ stcor : model2 ] show the average metrics over 100 replications .
the soft - thresholding estimator @xmath20 is positive definite in 19 or fewer out of 100 simulation runs , while @xmath63 and @xmath64 can always guarantee a positive - definite estimator . the larger the dimension , the less likely for the soft - thresholding estimator to be positive definite . in terms of estimation ,
both @xmath63 and @xmath64 are more accurate than @xmath20 . as for the selection performance , @xmath63 and @xmath64 achieve a slightly better true positive rate than @xmath20 .
overall , @xmath63 is the best among all three regularized estimators . to demonstrate our proposal
we further consider two gene expression datasets : one from a small round blue - cell tumors microarray experiment @xcite and the other one from a cardiovascular microarray study @xcite .
the first dataset has 64 training tissue samples with four types of tumors ( 23 ews , 8 bl - nhl , 12 nb , and 21 rms ) , and 6567 gene expression values for each sample .
we applied the pre - filtering step used in khan et al .
( 2001 ) and then picked the top 40 and bottom 160 genes based on the f - statistic as done in @xcite .
the second dataset has 63 subjects with 44 healthy controls and 19 cardiovascular patients , and 20426 genes measured for each subject .
we used the f - statistic to pick the top 50 and bottom 150 genes . by doing so
, it is expected that there is weak dependence between the top and the bottom genes .
we considered the soft - thresholding estimator @xcite , the log - barrier estimator @xcite and our estimator .
for all three estimators , the thresholding parameter was chosen by 5-fold cross validation .
eigenvalues of all three regularized estimators for the small round blue - cell data ( a ) and the cardiovascular data ( b ) : @xmath20 ( solid ) , @xmath63 ( dashed ) and @xmath64 ( dotted).,title="fig : " ] ( a ) eigenvalues of all three regularized estimators for the small round blue - cell data ( a ) and the cardiovascular data ( b ) : @xmath20 ( solid ) , @xmath63 ( dashed ) and @xmath64 ( dotted).,title="fig : " ] ( b ) as evidenced in plot [ real : eigen ] , the soft - thresholding estimator yields an indefinite matrix for both real examples whereas the other two regularized estimators guarantee the positive - definiteness .
the soft - thresholding estimator contains @xmath68 negative eigenvalues in the small round blue - cell data , and @xmath69 negative eigenvalues in the cardiovascular data . regularized correlation matrix estimation has a natural application in clustering when the dissimilarity measure is constructed using the correlation among features .
for both datasets we did hierarchical clustering using the three regularized estimators .
the heat maps are shown in figure [ real : heatmap ] in which the estimated sparsity pattern well matches the expected sparsity pattern .
( a1 , b1 ) , @xmath63 ( a2 , b2 ) and @xmath64 ( a3 , b3 ) .
the genes are ordered by hierarchical clustering using the estimated correlations.,title="fig : " ] ( a1 ) ( a1 , b1 ) , @xmath63 ( a2 , b2 ) and @xmath64 ( a3 , b3 ) .
the genes are ordered by hierarchical clustering using the estimated correlations.,title="fig : " ] ( a2 ) ( a1 , b1 ) , @xmath63 ( a2 , b2 ) and @xmath64 ( a3 , b3 ) .
the genes are ordered by hierarchical clustering using the estimated correlations.,title="fig : " ] ( a3 ) ( a1 , b1 ) , @xmath63 ( a2 , b2 ) and @xmath64 ( a3 , b3 ) .
the genes are ordered by hierarchical clustering using the estimated correlations.,title="fig : " ] ( b1 ) ( a1 , b1 ) , @xmath63 ( a2 , b2 ) and @xmath64 ( a3 , b3 ) .
the genes are ordered by hierarchical clustering using the estimated correlations.,title="fig : " ] ( b2 ) ( a1 , b1 ) , @xmath63 ( a2 , b2 ) and @xmath64 ( a3 , b3 ) .
the genes are ordered by hierarchical clustering using the estimated correlations.,title="fig : " ] ( b3 ) finally , we compared the average run times over @xmath65 cross validations for both @xmath63 and @xmath64 , as shown in table [ real : timing ] .
it is obvious that our proposal is much more efficient .
in this section , we prove that the sequence @xmath70 produced by the alternating direction method ( algorithm 1 ) converges to @xmath71 , where @xmath72 is an optimal solution of and @xmath73 is the optimal dual variable .
this automatically implies that algorithm 1 gives an optimal solution of .
we define some necessary notation for ease of presentation .
let @xmath74 be a @xmath75 by @xmath75 matrix defined as @xmath76 define the norm @xmath77 as @xmath78 and the corresponding inner product @xmath79 as @xmath80 . before we give the main theorem about the global convergence of algorithm 1 , we need the following lemma . [
lem : fejer - monotone ] assume that @xmath72 is an optimal solution of and @xmath73 is the corresponding optimal dual variable associated with the equality constraint @xmath81
. then the sequence @xmath82 produced by algorithm 1 satisfies [ lem : conclusion - eq ] u^i - u^*_g^2 - u^i+1-u^*_g^2 u^i - u^i+1_g^2 , where @xmath83 and @xmath84 .
now we are ready to give the main convergence result of algorithm 1 .
[ the : main - convergence ] the sequence @xmath82 produced by algorithm 1 from any starting point converges to an optimal solution of .
define @xmath45 as the true covariance matrix for the observations @xmath85 , and define the active set of @xmath86 as @xmath87 with the cardinality @xmath88 .
denote by @xmath89 the hadamard product @xmath90 .
define @xmath91 as the maximal true variance in @xmath45 .
[ theorem : stat - convergence ] assume that the true covariance matrix @xmath45 is positive definite .
* under the exponential - tail condition that for all @xmath92 and @xmath93 @xmath94 we also assume that @xmath95 . for any @xmath96
, we pick the thresholding parameter as @xmath97 where @xmath98 and @xmath99 with probability at least @xmath100 , we have @xmath101 * under the polynomial - tail condition that for all @xmath102 , @xmath103and @xmath93 @xmath104 we also assume that @xmath105 for some @xmath106 . for any @xmath96 ,
we pick the thresholding parameter as @xmath107 with probability at least @xmath108 , we have @xmath101 define @xmath109 and assume that @xmath110 is bounded by a fixed constant , then we can pick @xmath111 to achieve the minimax optimal rate of convergence under the frobenius norm as in theorem 4 of @xcite that p^+-^0_f^2=o_p((1+p)n)= o_p(dn ) .
however , to attain the same rate in the presence of the log - determinant barrier term , @xcite instead would require that @xmath112 , the minimal eigenvalue of the true covariance matrix , should be bounded away from zero by some positive constant , and also that the barrier parameter should be bounded by some positive quantity .
we would like to point out that if @xmath112 is bounded away from zero , then the soft - thresholding estimator @xmath113 will be positive - definite with an overwhelming probability tending to @xmath114 , @xcite .
therefore the theory requiring a lower bound on @xmath112 is not very appealing .
the soft - thresholding estimator has been shown to enjoy good asymptotic properties for estimating large sparse covariance matrices . but its positive definiteness property can be easily violated , which means the soft - thresholding estimator could be in principle an inadmissible estimator for covariance matrices . in this paper
we have put the soft - thresholding estimator in a convex optimization framework and considered a natural modification by imposing the positive definiteness constraint .
we have developed a fast alternating direction method to solve the constrained optimization problem and the resulting estimator retains the sparsity and positive definiteness properties simultaneously .
the algorithm and the new estimator are supported by numerical and theoretical results .
we thank adam rothman for sharing his code .
shiqian ma s research is supported by the national science foundation postdoctoral fellowship through institute for mathematics and its applications at university of minnesota .
hui zou s research is supported in part by grants from the national science foundation and the office of naval research .
since @xmath71 is optimal to , it follows from the kkt conditions that the followings hold .
[ kkt-1 ] ( -^+-^++_n)_j /|_j^+|,j=1, ,p , = 1, ,p j , [ kkt-1.5 ] ( ^+-_n)_jj + ^+_jj=0 , j=1, ,p , [ kkt-2]^+=^+,[kkt-2.5 ] ^+ , and [ kkt-3]^+,-^+ 0 , .
note that the optimality conditions for the first subproblem in algorithm 1 , i.e. the subproblem with respect to @xmath30 in , are given by [ opt - cond - theta]^i-(^i+1-^i)/,-^i+10,.using the updating formula for @xmath115 in algorithm 1 , i.e. , [ update - lambda]^i+1=^i-(^i+1-^i+1)/ , can be rewritten as [ opt - cond - theta-1 ] ^i+1-(^i+1-^i)/,-^i+10,.now by letting @xmath116 in and @xmath117 in , we can get that [ proof - lemma - theta - inequa-1]^+,^i+1-^+0 , and [ proof - lemma - theta - inequa-2]^i+1-(^i+1-^i)/,^+-^i+10 . summing and yields [ proof - lemma - theta - inequa-3]^i+1-^+ , ( ^i+1-^+)+(^i-^i+1)/0 .
the optimality conditions for the second subproblem in algorithm 1 , i.e. , the subproblem with respect to @xmath36 in are given by [ opt - cond - sigma - tmp1]0(^i+1-_n)_j+|^i+1_j| + ^i_j + ( ^i+1-^i+1)_j/,j=1, ,p,=1, ,p , j , and [ opt - cond - sigma - tmp2 ] ( ^i+1-_n)_jj+^i_jj+(^i+1-^i+1)_jj/=0,j=1, ,p.note that by using , and can be respectively rewritten as : [ opt - cond - sigma ] ( -^i+1-^i+1+_n)_j/|^i+1_j| , j=1, ,p,=1, ,p , j , and [ opt - cond - sigma-1.5 ] ( ^i+1-_n)_jj+^i+1_jj=0,j=1, ,p.using the fact that @xmath118 is a monotone function , , , and imply [ proof - lemma - sigma - inequa-1 ] ^i+1-^+ , ( ^+-^i+1)+(^+-^i+1)0 . the summation of and gives [ proof - lemma - combine - inequa-1 ] using the notation of @xmath123 and @xmath124 , can be rewritten as [ proof - lemma - final - inequa-3 ] u^i - u^*,u^i - u^i+1 _ g u^i - u^i+1_g^2+^i+1-^+ _ f^2-^i-^i+1,^i-^i+1 . combining with the following identity @xmath125 we get [ proof - lemma - final - inequa-4 ] now , using and for @xmath47 instead of @xmath126 , we get , [ opt - cond - sigma - k ] ( -^i-^i+_n)_j/|^i_j| , j=1, ,p,=1, ,p , j , and [ opt - cond - sigma - k-1.5 ] ( ^i-_n)_jj+^i_jj=0,j=1, ,p . combining , , , and using the fact that @xmath118 is a monotone function , we obtain , @xmath127 which immediately implies , [ proof - lemma - final - inequa-5 ] ^i-^i+1,^i+1-^i^i+1-^i_f^2 0 . by substituting into , we get the desired result .
it follows from ( i ) that @xmath131 and @xmath132 .
then implies that @xmath133 and @xmath134 . from ( ii ) we obtain that , @xmath123 has a subsequence @xmath135 that converges to @xmath136 , i.e. , @xmath137 and @xmath138 . from @xmath134
we also get that @xmath139 .
therefore , @xmath140 is a limit point of @xmath82 . without loss of generality
, we may always assume that @xmath141 for all @xmath142 . by the condition that @xmath45 is positive definite
, we can always choose some very small @xmath26 such that @xmath143 is smaller than the minimal eigenvalue of @xmath45 .
we introduce @xmath144 , and then we can write in terms of @xmath145 as follows , = _ : = ^,+^0 2+^0-_n_f^2+|+^0|_1(f ( ) ) .
note that it is easy to see that @xmath146 .
now we consider @xmath147 . under the probability event @xmath148
, we have f()-f ( ) & = & 2+^0-_n_f^2 - 2 ^ 0-_n_f^2 + |+^0|_1-|^0|_1 + & = & 2_f^2+<,^0-_n > + |_a_0^c|_1 + ( |_a_0+^0_a_0|_1-|^0_a_0|_1 ) + & & 2_f^2 - + |_a_0^c|_1 -|_a_0|_1 + & & 2_f^2 - 2 + & & 2_f^2 - 2 ^ 1/2_f + & & 2 ^ 2 ( s+p ) + & > & 0 note that @xmath149 is also the optimal solution to the following convex optimization problem = _ : = ^,+^0 f()-f()(g ( ) ) . under the same probability event ,
@xmath150 would always hold . otherwise , the fact that @xmath151 for @xmath152 should contradict with the convexity of @xmath153 and @xmath154 .
therefore , we can obtain the following probability bound @xmath155 now we shall prove the probability bound under the exponential - tail condition .
first it is easy to verify two simple inequalities that @xmath156 and @xmath157 .
the first inequality can be proved by using the taylor expansion , and the second one can be easily derived using the obvious facts that @xmath158 and @xmath159 .
let @xmath160 , @xmath161 and @xmath162 .
for any @xmath96 , we can apply the markov inequality to obtain that ( _ ix_ij > n_0 ) & & ( -t_0n_0)_i=1^ne[(t_0x_ij ) ] + & & ( -t_0n_0)_i=1^n \{1 + 2e[x_ij^2(t_0|x_ij| ) ] } + & & p^-c_0 ^ 1/2(2_i=1^ne[x_ij^2(t_0|x_ij| ) ] ) + & & p^-c_0 ^ 1/2(2_i=1^ne[(t_0 ^ 2x_ij^2 + 1 ) ] ) + & & p^-c_0 ^ 1/2(2ek_1p)(=p^-m-1 ) , where we apply @xmath163 in the second inequality and @xmath164 in the third inequality , and then use @xmath157 in the fourth inequality .
moreover , the simple facts that @xmath165=0 $ ] ( @xmath166 ) and @xmath167 are also used .
let @xmath168 and @xmath169 .
define @xmath170 .
for any @xmath96 , we first apply the cauchy inequality to obtain that e[x_ij^2x_ik^2(2|x_ijx_ik| ) ] & & e[x_ij^2x_ik^2(4(x^2_ij+x^2_ik ) ) ] + & & ( e[x_ij^4(x^2_ij/2)])^1/2(e[x_ik^4(x^2_ik/2)])^1/2 + & & 4 ^ -2(e[(x^2_ij)])^1/2 ( e[(x^2_ik)])^1/2 + & & 4k_1 ^ -2 , where we use the simple inequality @xmath159 in the third inequality .
then , combining this result with the cauchy inequality again yields that & & e[(x_ijx_ik-^0_jk)^2(t_1|x_ijx_ik-^0_jk| ) ] + & & 2e[x_ij^2x_ik^2(2|x_ijx_ik-^0_jk|)]+ 2(^0_jk)^2e[(2|x_ijx_ik-^0_jk| ) ] + & & 8k_1 ^ -2(2 ^ 0_jk ) + 2(^0_jk)^2(2 ^ 0_jk)e[(4(x_ij^2+x_ik^2 ) ) ] + & & 8k_1 ^ -2(2_)+2_^2(2 _ ) ( e[(2x_ij^2)])^1/2(e[(2x_ik^2)])^1/2 + & & 2k_1(4 ^ -2+_^2)(2 _ ) where we use the fact that @xmath171 in the first inequality , and then use @xmath172 in the third inequality .
now , we can apply the markov inequality to obtain the following probability bound & & ( _ i\{x_ijx_ik-^0_jk } > n_1 ) + & & ( -t_1n_1)_i=1^ne + & & p^-2c_1_i=1^n \{1 + 2t_1 ^ 2e } + & & p^-2c_1 ( 2t_1 ^ 2_i=1^ne ) + & & p^-2c_1(k_1(1 + 4 ^ 2_^2)(2_)p)(=p^-m-2 ) , where we apply @xmath163 and @xmath173=\sigma^0_{jk}$ ] for @xmath174 in the second inequality , and we use @xmath164 in the third inequality . recall that @xmath175 and @xmath176 therefore , we can complete the probability bound under the exponential - tail condition as follows ( _ j , k|^n_jk-^0_jk| > ) & & p^2(_ix_ijx_ik >
n(^0_jk+_1))+2p(_ix_ij > n_0 ) + & & 3p^-m . in the sequel
we shall prove the probability bound under the polymonial - tail condition .
first , we define @xmath177 and @xmath178 .
define @xmath179 , @xmath180 and @xmath181 .
then we have @xmath182 and @xmath165=e[y_{ij}]+e[z_{ij}]$ ] . by construction , @xmath183
are bounded random variables , and @xmath184 $ ] are bounded by @xmath185 due to the fact that @xmath186\le k_2\delta_n^{-3}=o(\varepsilon_2 ) .
$ ] now we can apply the bernstein s inequality @xcite to obtain that ( _ i\{y_ij - e[y_ij ] ) } > 2n_2 ) & & ( ) + & & ( ) + & = & o(p^-m-1 ) , where the fact that @xmath187\le e[x^2_{ij}i_{\{|x_{ij}|\ge 1\}}]+e[x^2_{ij}i_{\{|x_{ij}|\le 1\}}]\le k_2 + 1 $ ] is used in the second inequality .
besides , we can apply the markov inequality to obtain that @xmath188 \le k_2(\log n)^{2(1+\gamma+\varepsilon)}n^{-1-\gamma-\varepsilon}.\ ] ] then , we can derive the following probability bound ( _
n_2 ) & = & ( _ i\{y_ij+z_ij - e[y_ij+z_ij ] } > n_2 ) + & & ( _ i\{y_ij - e[y_ij ] } > 2n_2)+ ( _ i\{z_ij - e[z_ij ] } > 2n_2 ) + & & o(p^-m-1)+(_i\{z_ij - o(_2 ) } > 2n_2 ) + & & o(p^-m-1)+_i(|x_ij|>_n ) + & & o(p^-m-1)+k_2(n)^2(1++)n^-- . let @xmath189 and @xmath190 . recall that @xmath191 , and define @xmath192 .
then we have @xmath193 and @xmath194=e[y_{ij}y_{ik}]+e[r_{ijk}]$ ] . by construction , @xmath195
are bounded random variables , and @xmath196 $ ] is bounded by @xmath197 due to the fact that |e[r_ijk]| & & & & _ n^-2 - 4e[x^4(1+)_iji_\{|x_ij|>_n}]e[x^2_ik]+_n^-2 - 4e[x^4(1+)_iki_\{|x_ik|>_n}]e[x^2_ij ] + & & 2 k_2_n^-2 - 4 ( = o(_3 ) ) .
again , we can apply the bernstein s inequality to obtain that ( _ i\{y_ijy_ik - e[y_ijy_ik ] } > 2n_3 ) & & ( ) + & & ( ) + & = & o(p^-m-2 ) , where the fact that @xmath198\le ( e[x^4_{ij}]e[x^4_{ik}])^{1/2}\le k_2 + 1 $ ] is used .
& & ( _ j , k|_i(x_ijx_ik-^0_jk)| > n_3 ) + & & ( _ j , k|_i\{y_ijy_ik - e[y_ijy_ik]}| > 2n_3)+(_j , k|_i\{r_ijk - e[r_ijk]}| > 2n_3 ) + & & 2_j , k(_i\{y_ijy_ik - e[y_ijy_ik ] } >
2n_3 ) + ( _ j , k|_i\{r_ijk - o(_3)}| >
2n_3 ) + & & o(p^-m)+_i , j(|x_ij|>_n ) + & & o(p^-m)+k_2p(n)^2(1++)n^-- recall that @xmath199 . therefore , we can prove the desired probability bound under the polynomial - tail condition as follows & & ( _ j , k|^n_jk-^0_jk| > ) + & & ( _ j , k|_i\{x_ijx_ik-^0_jk}| > n_3 ) + ( _ j|_ix_ij| > n_2 ) + & & o(p^-m)+3k_2p(n)^2(1++)n^-- . beer , d. , kardia , s. , huang , c. , giordano , t. , levin , a. , misek , d. , lin , l. , chen , g. , gharib , t. , thomas , d. , et al . ( 2002 ) , `` gene - expression profiles predict survival of patients with lung adenocarcinoma , '' _ nature med .
_ , 8 , 816824 .
khan , j. , wei , j. , ringnr , m. , saal , l. , ladanyi , m. , westermann , f. , berthold , f. , schwab , m. , antonescu , c. , peterson , c. , et al .
( 2001 ) , `` classification and diagnostic prediction of cancers using gene expression profiling and artificial neural networks , '' _ nature med .
_ , 7 , 673679 .
subramaniana , a. , tamayoa , p. , moothaa , v. , mukherjeed , s. , eberta , b. , gillettea , m. , paulovichg , a. , pomeroyh , s. , goluba , t. , landera , e. , et al .
( 2005 ) , `` gene set enrichment analysis : a knowledge - based approach for interpreting genome - wide expression profiles , '' _ proc .
_ , 102 , 1554515550 . | the thresholding covariance estimator has nice asymptotic properties for estimating sparse large covariance matrices , but it often has negative eigenvalues when used in real data analysis . to simultaneously achieve sparsity and positive definiteness , we develop a positive definite @xmath0-penalized covariance estimator for estimating sparse large covariance matrices . an efficient alternating direction method is derived to solve the challenging optimization problem and its convergence properties are established . under weak regularity conditions ,
non - asymptotic statistical theory is also established for the proposed estimator .
the competitive finite - sample performance of our proposal is demonstrated by both simulation and real applications .
= 1 * keywords * : alternating direction methods ; large covariance matrices ; matrix norm ; positive - definite estimation ; sparsity ; soft - thresholding . |
one of the most convenient ways to trace magnetic field is related to emission and absorption of polarized radiation by aligned dust .
therefore aligned dust is widely used for this purpose both in diffuse interstellar medium and molecular gas .
moreover , there is evidence of aligned dust around young stellar objects and evolved stars as well as astrophysical environments . at the same time , the processes that aligned dust and their relation to magnetic field are still the subject of debates and require further studies ( see lazarian 2007 for a review ) . while radiative torques ( dolginov & mytrophanov 1976 , draine & weingartner 1996 , 1997 , weingartner & draine 2003 , cho & lazarian 2005 , lazarian & hoang 2007 , hoang & lazarian 2007 ) are currently seen as the most promising candidate mechanism , the variety of astrophysical conditions may enables other mechanisms to dominate in particular environments .
mechanical alignment was pioneered by gold ( 1951 , 1952 ) and then quantified and elaborated by other researchers ( e.g. purcell 1969 , purcell & spitzer 1971 , dolginov & mytrophanov 1976 , lazarian 1994 , 1997 , roberge et al .
while the original mechanism could deal with thermally rotating grains only , two modifications of the mechanism introduced in lazarian ( 1995 ) and elaborated later in lazarian & efroimsky ( 1996 ) , lazarian , efroimsky & ozik ( 1997 ) , efroimsky ( 1999 ) enabled the alignment of grains rotating at much higher rates .
the latter were introduced to the field by purcell ( 1979 ) ( see also lazarian & draine 1999ab , where the limitations on the size for suprathermally rotating grains are discussed ) .
the main shortcoming of the mechanical alignment processes was that they required supersonic gas - dust drift to get any appreciable degree of alignment ( see purcell 1969 ) .
although later studies indicated that such drifts can be produced by ambipolar diffusion ( roberge & hanany 1990 , roberge et al .
1995 ) or interactions of charged grains with mhd turbulence ( lazarian 1994 , lazarian & yan 2002 , yan & lazarian 2002 , yan , lazarian & draine 2004 ) , the degree of alignment that is achievable for the mach number drifts of the order of unity is insufficient to explain observations ( see estimates in lazarian 1997 ) . the possibility of subsonic flows to mechanically align grains was mentioned in passing in lazarian ( 2007 ) and lazarian & hoang ( 2007 , henceforth lh07 ) , but no relevant calculations were provided there .
this possibility is related to irregular grains demonstrating _ helicity _ , i.e. ability of spin up in a _ regular way _ while interacting with a flow of particles .
in lh07 radiative torques that arise from the interaction of photons with a grain were considered .
similar torques , however , should emerge when atoms bombard the surface of an irregular grain . in figure
1 we provide a simple model of a helical grain .
the grain consists of an oblate spheroid or an ellipsoid with a mirror attached to it at an angle @xmath0 .
the difference with lh07 is that we assume that the mirror is reflective to atoms rather than photons . for our purposes in the paper , the actual properties of a spheroid or an ellipsoid do not matter .
those geometric shapes are too symmetric and do not produce torques that can spin up a grain in a regular way , i.e. with angular velocity growing in proportion to time .
the stochastic , i.e. with angular velocity growing as a square root of the time , spin up is the essence of the gold - type alignment processes and it is ignored here , where we deal with subsonic flows . for the sake of simplicity , we assume that the larmor precession of the grain in the external magnetic field is much faster than precession arising from the interaction of the ellipsoidal body with the gaseous flow . as a result ,
following arguments similar to those in lh07 , one can disregard the effects of the gaseous flow on the ellipsoidal body altogether .
thus , the only torques to consider are those arising from the mirror . as in lh07 ,
the the major role of the ellipsoidal body is to provide a steady rotation about its axis corresponding to the maximal moment of inertia .
this ensures that the model grain is subject to regular torques while it interacts with the gaseous flow .
the difference between the calculations in lh07 and those in the current paper stems from the fact that , while in lh07 the photons are coming as a beam from a single direction , for subsonic flow that we deal with , atoms hit the mirror from all directions characterized by the flux of atoms given by equation ( [ ap2 ] ) .
this induces averaging of the torques that we implement below .
consider a helical grain ( see fig .
[ f0 ] ) drifting across a gas chamber with a velocity @xmath1 along the @xmath2 axis .
the velocity of an atom with respect to the grain is @xmath3 where @xmath4 is the thermal velocity of the atom . the torque from the perfect reflecting of atoms on a surface area @xmath5 of the mirror is given by @xmath6 a f({\bf s}-{\bf s}_{d } ) d^{3 } s,\label{ap1}\ ] ] where @xmath7 , vectors @xmath8 is the radius vector directed from the spheroid to the mirror
, @xmath9 and @xmath5 is the surface area of the mirror , and @xmath10 is the flux of incoming atoms that can collide with the grain .
angular momentum element @xmath11 is antiparallel to the normal vector and given by @xmath12 where @xmath13
. here @xmath14 are radius and normal vector which are functions of angles @xmath15 describing the orientation of the grain in the lab system defined by @xmath16 and @xmath17 ( see fig .
[ f0 ] ) , as given in lh07 . and
@xmath18 are components of @xmath19 in the grain coordinate system . ] substituting equations ( [ ap2 ] ) and ( [ ap3 ] ) into equation ( [ ap1 ] ) we get @xmath20 ( ( { \bf s}-{\bf s}_{d}).{\bf n})^{2 } a e^{-s^{2 } } d^{3}s,\label{ap4}\ ] ] taking into account the contribution from the reflection on the other surface , total torque becomes then @xmath21\int_{({\bf s}-{\bf s}_{d}).{\bf n}<0 } ( ( { \bf s}-{\bf s}_{d}).{\bf n})^{2 } a e^{-s^{2 } } d^{3}s\nonumber\\ -2
n m_{h } v_{th}^{2 } [ { \bf r}\times { \bf n}]\int_{({\bf s}-{\bf s}_{d}).{\bf n}>0 } ( ( { \bf s}-{\bf s}_{d}).{\bf n})^{2 } a e^{-s^{2 } } d^{3}s , \label{ap5}\ ] ] integral ( [ ap5 ] ) can be analytically evaluated and the resulting torques are given by @xmath22\lbrace |n_{1}| n_{1}(2 s_{d } e^{-s_{d}^{2}}+\sqrt{\pi } erf(s_{d})\nonumber\\ + 2 s_{d}^{2 } \sqrt{\pi } erf(s_{d}))+\frac{n_{1}}{|n_{1}|}(n_{2}^{2}+n_{3}^{2})erf(s_{d})\rbrace,\label{eq8}\ ] ] where @xmath23 and @xmath24 are components of @xmath19 in the lab coordinate system .
equation ( [ eq8 ] ) can be written as @xmath25 where @xmath26 is the vector torque efficiency consisting of components @xmath27 and @xmath28 on axes @xmath29 and @xmath17 , respectively .
two first components for subsonic and supersonic cases are shown in figure [ f1 ] . the third component produced by the mirror
is found to be zero .
it is clear that the component @xmath28 is responsible for the precession ( lh07 ) .
this component is non - zero for a spheroidal body of the grain .
in fact , as we discussed above , we disregard this component , as it does not induce either alignment or spin up . in figure [ f1 ]
we show torque components for subsonic case @xmath30 and a supersonic case @xmath31 .
it can be seen that the component @xmath32 has zero points at @xmath33 and @xmath34 in both subsonic and supersonic case .
the component @xmath35 is very symmetric in the latter .
however , the torques are distorted due to the thermal collisions of gaseous atoms in the former case .
the essential properties ( i.e. symmetry of @xmath35 and zero points of @xmath32 ) of mechanical torques which are similar to those of radiative torques indicate that mechanical torques can align grains in the same ways as radiative torques do , i.e. , grains tend to aligned with long axes perpendicular to magnetic fields .
we study this problem in the section below .
when the larmor precession of a grain is larger than the precession induced by mechanical torques ( see 11.6 in lh07 ) , the alignment occurs with respect to magnetic field ( either @xmath36 or @xmath37 ) which defines the axis of alignment . to study the grain alignment induced by mechanical torques , similar to lh07 , we solve the equations of motion for angular momentum @xmath38 in time t : @xmath39 where @xmath40 is given by equation ( [ eq8 ] ) and @xmath41 is the gaseous damping time . for simplicity , we assume a perfect internal alignment ; i.e. the axis of major inertia @xmath42 is always parallel to @xmath38 ( dw97 ; lh07 ) .
we consider a model grain drifting across magnetic field by an angle @xmath43 acted upon by a subsonic flow with @xmath44 .
representing equation ( [ eq10 ] ) in the spherical coordinate system defined by @xmath45 where @xmath46 is the magnitude of @xmath38 , @xmath47 is the angle between @xmath38 and @xmath48 , and @xmath49 is the larmor precession angle , we get three equations for @xmath50 and @xmath49 ( see sec . 7 in lh07 ) . averaging over @xmath49 from @xmath51 to @xmath52
, we obtain only equations for @xmath53 and @xmath47 .
we construct trajectory map based on the obtained solution @xmath54 .
figure [ f2 ] shows the obtained trajectory map with initial condition @xmath55 and the angle @xmath47 generated from a uniform distribution .
we see that grain phase trajectories that start at different angles @xmath47 converge on either high-@xmath46 attractor points a and b corresponding to perfect alignment , or a low-@xmath46 attractor point c. the latter point formally corresponds to @xmath56 .
however , as it was explained in lh07 and shown by numerical simulations in hoang & lazarian ( 2007 ) , this is an artifact of our ignoring thermal fluctuations within the grain .
if those are taken into account the low-@xmath46 attractor points correspond to the angular momentum of the order of the thermal angular momentum corresponding to the temperature of the grain body ( see also weingartner & draine 2003 , hoang & lazarian 2007 ) .
what should be noted is that the attractor points for @xmath44 correspond to the alignment of @xmath42 parallel to @xmath48 ; i.e. the grain gets aligned with long axes perpendicular to magnetic field .
this is similar to what the davis - greenstein ( 1951 ) mechanism of paramagnetic relaxation predicts , but the process we discuss here does not invoke paramagnetic relaxation .
gaseous bombardment randomizes grain phase trajectories .
however , rather than tracing phase trajectories , it is practically convenient to have a criterion for the efficiency of grain alignment .
the maximal rotational velocity of grains induced by torques can be used for this purpose .
a study in hoang & lazarian ( 2007 ) which solves the lagevin equation in order to account for gaseous bombardment shows that this velocity which us a function of @xmath57 should be approximately 3 times larger than the thermal rotational velocity of the grain . for practical calculations of the rotational velocity
we assume that 10% of the atoms impinging on the grain mirror are reflected .
figure [ f3 ] shows the angular velocity of a grain as a function of the mach number of the flow .
the horizontal line corresponds to @xmath58 for @xmath59 .
it is clear from looking at figure [ f3 ] that for our model , helical grains get aligned for relatively low velocities of gas - grain drift .
all the earlier mechanisms of mechanical alignment are inefficient for such low subsonic velocities .
indeed , drifting velocities of the order of @xmath60 of sound speed should be quite common in the ism ( see yan , lazarian & draine 2004 ) , so potentially the mechanism is widely applicable .
however , it requires further studies .
one may wonder whether the condition of perfect reflection is absolutely necessary .
it can be shown that the torques are changed by a factor of unity if impinging atoms are absorbed on the grain surface , thermalized there , and emitted from the point that they hit the surface . moreover , for regular torques to act on a helical grain , it is sufficient to have a correlation of the place that the atom impinges on the grain surface and is evaporated from it .
it is only when there is no correlation at all , that our model grain does not experience _ regular _ torques .
therefore we believe that we do not overestimate the torques with our assumption of 10% reflection . in general , to characterize the interaction of grain surface with the gaseous flow , one can introduce a `` reflection efficiency factor '' @xmath61 .
a different issue is the degree of helicity of an irregular grain .
naturally , most of the grains do not have facets , but have numerous irregularities .
an idealized model grain ( see fig .
1 ) , for which the damping arising from the gas interactions with the ellipsoidal body tends to zero , has the maximal possible value of the torques @xmath62 for a given grain size @xmath63 and the mach number @xmath64 and its `` helicity reduction factor '' @xmath65 .
this factor , in a general case , can be characterized by @xmath66 .
further numerical studies should clarify the value of @xmath67 for actual irregular grains . in figure [ f3 ]
we show results for different @xmath68 factors .
in our letter we use the same toy model of a helical grain as in lh07 . in lh07
we proved the validity of the model ( which uses a geometric optics approximation ) comparing the functional dependencies of model s torques and those numerically calculated for the actual irregular grains . in comparison ,
the justification of the model for the gaseous bombardment is self - evident .
however , the combination @xmath68 presents a combined uncertainty factor , which characterizes both the uncertainties in grain - gas interactions and grain shape .
testing of the former requires laboratory studies , while the latter can be established via the numerical research . what is the relative role of traditional mechanisms of mechanical alignment ?
those mechanisms require supersonic motions to be efficient .
it is only mechanical alignment of helical grains that allows alignment for the subsonic motions .
thus we may claim that mechanical alignment of helical grains constitutes a new class of mechanical alignment processes .
in fact , the difference between the mechanical alignment of helical grains and the gold mechanism of grain alignment is similar to the difference between the radiative torque alignment and the harwit ( 1970 ) process .
the latter appeals to stochastic torques arising from absorption of photons from a light beam and was shown by purcell & spitzer ( 1971 ) to be inefficient for most interstellar situations .
similarly , as the radiative torque alignment is more efficient than the harwit ( 1970 ) alignment , we believe that the mechanical of helical grains is more efficient than the gold processes . therefore , at supersonic velocities we expect the mechanical alignment of the actual irregular grains to be governed by their helicity , rather than the degree of their oblateness of prolateness , which is the case for the gold mechanisms .
the consequence of this is that the mechanical alignment of grains will happen with long grain axes perpendicular to magnetic field irrespective on the gas - grain velocities .
this is in contrast to the gold alignment , for which the change of alignment happens at the van vleck angle @xmath69 .
in hoang & lazarian ( 2007 ) , the effects of gaseous bombardment and uncompensated purcell s torques ( purcell 1979 ) , e.g. h@xmath70 torques , was studied in connection with the alignment of helical grains subjected to radiative torques .
that study shows that , in the case when both high-@xmath46 and low-@xmath46 attractor points coexist , gaseous bombardment transfers grains from low-@xmath46 to high-@xmath46 thus increasing attractor points compared to the low-@xmath46 ones . ] , counter - intuitively , the degree of alignment .
this is surely true only if @xmath71 is sufficiently high , e.g. higher than @xmath72 .
the purcell torques , do not change @xmath46 in the low-@xmath46 attractor points because grains flip fast at those points causing thermal trapping , as discussed in lazarian & draine ( 1999a ) .
they , however , can increase the values of @xmath46 at high-@xmath46 attractor points . for most situations , this does not affect the degree of alignment .
our considerations above deal with ordinary paramagnetic grains . for those
, the influence of paramagnetic torques is mostly negligible for the typical ism conditions .
the situation changes if grains are superparamagnetic ( see jones & spitzer 1967 ) .
in this situation one can expect always to have alignment with high-@xmath46 attractor points , thus having degree of alignment close to 100% .
the assumption of the presence of superparamagnetic inclusions is an extra assumption , however . as we have stressed above , the mechanical alignment we discuss here and the radiative torque alignment are different incarnations of the alignment of helical grains .
while the radiative torques have attracted a lot of attention recently , the mechanical alignment of helical grains has only be mentioned in a couple of publications ( lazarian 1995 , 2007 , lazarian et al .
1997 , lh07 ) . the relative role of the mechanisms depends on the yet uncertain factor @xmath68 for the mechanical processes .
the observations tends to be in agreement with the radiative torque predictions ( see lazarian 2007 and ref .
therein ) .
it is encouraging that both mechanisms predict the alignment with long grain axes perpendicular to magnetic field .
therefore one may hope that the subsonic alignment of helical grains can reveal via polarimetry the magnetic fields in the situations when the radiative torques fail .
the variety of astrophysical circumstances ensures that there are situations when grains are aligned by mechanical subsonic flows .
note that aligned grains were reported not only for the ism , but also comets and circumstellar regions .
they are also likely to be present in the disks around evolved and young stars .
the goal of this letter above is to attract the attention of the community to the possibility of mechanical alignment of helical dust grains . while further studies of the process are necessary , the following points can be made at this moment : 1952 , mnras , 112 , 215 harwit , m. 1970 , nature , 226 , 61 - 63 hoang , t. , & lazarian , a. 2007 , mnras , in press ( astro - ph/0707.3645 ) jones , j.v . ,
spitzer , l. 1967 , apj , 146,943 lazarian , a. 2007 , journal of quantitative spectroscopy and + radiative transfer , 106,225 lazarian , a. & hoang , t. 2007 , mnras , 378 , 910 ( lh07 ) lazarian , a. , goodman , a.a .
& myers , p.c .
1997 , apj , 490 , 273 - 280 ( lgm97 ) lazarian , a. 1994 , mnras , 268 , 713 - 723 , ( l94 ) 2000 , apj , 536 , l15-l18 lazarian , a. , & draine , b.t .
, 1997 , apj , 487 , 248 - 258 lazarian , a. , & yan , h. 2002 , apj , 566 , l105-l108 lazarian , a. , & efroimsky , m. , ozik , j , 1996 , apj , 472 , 240 - 244 lazarian , a. , goodman , a.a . & myers , p.c .
1997 , apj , 490 , 273 - 280 ( lgm97 ) purcell , e.m .
1969 , on the alignment of interstellar dust , physica , 41 , 100 roberge , w.g .
, & hanany , s. 1990 , b.a.a.s .
, 22 , 862 roberge , w.g . , hanany , s. , & mesinger , d.e .
1995 , apj , 453 , 238 yan , h. , & lazarian , a. 2003 , apj , 592 , l33-l36 yan , h. , lazarian , a. , draine , b. 2004 , apj , 616 , 895 weingartner , j. , & draine , b. 2003 , apj , 589 , 289 ( wd03 ) and @xmath74 averaged over the rotation are shown for different mach numbers @xmath64 of the relative grain - gas flow . for large @xmath31 the torque @xmath75 is symmetric and
@xmath76 has three zeros at @xmath77 and @xmath34 . for subsonic flows
@xmath44 , the torques are modified due to the effect of thermal velocity , but zeros points of @xmath76 are unchanged.,scaledwidth=40.0% ] and @xmath78 , in diffuse gas with @xmath79 , temperature @xmath80 corresponding to the march number @xmath44 for an angle @xmath57 between @xmath1 and magnetic field @xmath48 equal @xmath81 has attractor points a and b of high angular momentum , c with @xmath56 . in the presence of thermal grain wobbling the point c becomes a low-@xmath46 attractor point with @xmath82.,scaledwidth=40.0% ] .
the horizontal line corresponds to @xmath83 which presents the threshold for grain alignment obtained in hoang & lazarian ( 2007 ) .
the results are obtained for a gaseous flow moving parallel to the magnetic field . for grains with velocity mostly perpendicular to the magnetic field ( see yan & lazarian 2003 ) , the angular momentum @xmath46 can be further decreased by a factor 10 - 100,scaledwidth=40.0% ] | we show that grains can be efficiently aligned by interacting with a subsonic gaseous flow .
the alignment arises from grains having irregularities that scatter atoms with different efficiency in the right and left directions .
the grains tend to align with long axes perpendicular to magnetic field , which corresponds to davis - greenstein predictions .
choosing conservative estimates , scattering efficiency of impinging atoms and conservative `` degree of helicity '' , the alignment of helical grains is much more efficient than the gold - type alignment processes . |
thanks to the outstanding work by n. walborn and collaborators over the last 30 years , a detailed classification scheme for ob - stars of solar metallicity has been developed , based on the morphology of the line spectrum .
for the specific case of o - type stars , an alternative scheme relying on quantitative criteria has been provided by conti & alschuler @xcite and mathys @xcite as well .
consequently , the classification of galactic ob stars is rigorously defined .
the classification of extragalactic stars , on the other hand , is still somewhat problematic . in particular
, the calibration between spectral types and physical parameters differs as a function of metallicity when criteria for milky way stars are applied . with respect to b - stars ,
this point has been extensively discussed by various authors ( e.g. , monteverde et al .
@xcite , lennon @xcite , urbaneja et al .
@xcite ) who pointed out that the use of the mk classification criteria for stars with metallicities different than solar can significantly influence the derived spectral types and luminosity classes . unlike b - type stars , where the relative strengths of metal to he i optical lines are used for classification purposes ,
o - type stars are generally classified by comparing the strengths of he i and he ii lines ( see , e.g. , conti & alschuler @xcite ; mathys @xcite , walborn @xcite ; walborn & fitzpatrick,@xcite ) .
thus , one might conclude that there should be no problem to employ the standard classification schemes of either walborn or conti to o - type stars in various metallicity environments . however , theoretical considerations predict that helium line strengths should also depend , though indirectly , on metallicity , due to the processes of mass - loss and euv line - blocking / blanketing ( e.g. , abbott & hummer @xcite , sellmaier et al .
@xcite , herrero et al .
@xcite ) . whilst in stars with weaker winds ( e.g. , dwarfs ) the effect is controlled , to a major extent , by the blanketing alone , in stars with stronger winds ( e.g. , supergiants ) , the processes of wind emission and additional wind - blanketing seem to be of similar or even dominating impact ( e.g. , repolust et al .
as noted in the previous section , the classification of o - type stars in the optical band relies either on the walborn or on the conti classification schemes , both developed to be used at solar metallicity . in the walborn scheme visual estimates of individual line strengths or line strength ratios of strategic helium transitions are used as classification criteria , while in the conti scheme the logarithm of the equivalent width ( ew ) ratio of he i 4471 to he ii 4541 is used instead .
given that an eye - estimated ratio is roughly equivalent to the logarithmic ratio of the corresponding ews , these two approaches should lead to similar results ( but see heap et al .
@xcite ) . to address the metallicity issue regarding the spectral classification of o - type stars , we developed the following strategy .
as a first step and using the fastwind code ( puls et al .
@xcite ) , we constructed 6 model grids , corresponding to two values of metallicity , solar , @xmath1 , and 0.3 @xmath1 , of about -0.5 was found , corresponding to an average metallicity of about 0.3 ( for more information , see efremova et al . @xcite and references therein ) . ] , and three luminosity classes : dwarfs ( dws ) , giants ( gs ) and supergiants ( sgs ) . by means of these models we investigated the predicted variations in helium line strengths , caused by the adopted change in metallicity , as a function of and luminosity class .
subsequently , we evaluated the shifts in spectral type ( and ) and luminosity class , that would eventually appear if galactic classification criteria were used to o - type stars of metallicity 0.3 .
our model grids comprise 64 nlte , line - blanketed models with stellar winds . for all models and independent of metallicity , stellar parameters ( , , and )
were taken from the empirical calibrations of martins et al .
@xcite for galactic stars of luminosity classes i , iii and v. we used a solar chemical composition as derived by asplund et al.@xcite , subsequently scaled to z=0.3 @xmath2 .
a standard @xmath3-velocity law with @xmath3=0.9 was adopted for all models .
terminal velocities , , for solar metallicity models were determined by means of the spectral type - calibration from kudritzki & puls @xcite .
the effects of metallicity on were taken into account by employing the scaling relation from leitherer et al .
@xcite , namely @xmath4 .
mass - loss rates were calculated using the mass - loss recipe from vink et al .
a microturbulent velocity of 10 was used to calculate the atmospheric structures , whilst values of 5 , 10 and 15 were assumed for calculating the emergent spectrum .
based on these model grids , the ew variations of several strategic he i ( e.g. , he i 4026 , he i 4471 and he i 4387 ) and he ii ( e.g. , he ii 4200 , he ii 4541 and he ii 4686 ) lines were investigated , as a function of metallicity ( and 0.3 ) , and luminosity class . our results , partly illustrated in figures [ he_lines ] and 3 , indicate that in the o - star temperature domain he i and he ii are both influenced by metallicity .
the effect is opposite to that for metal lines , i.e. , while metal lines become weaker due to the lower metal content , the helium line strengths are predicted to increase in absorption .
moreover , individual lines react differently .
members of the he i singlet series are more sensitive to metallicity than those of the triplet series , which in turn are more sensitive than he ii transitions ( except for he ii 4686 , see next section ) .
in addition , the effect depends on and , being more pronounced in hotter sgs ( where mass loss effects are strongest ) , and almost negligible ( within @xmath510% ) at intermediate and cooler temperature dws .
our findings imply that , due to the lower metal content , the optical helium line spectrum of low metallicity o - stars would appear somewhat different from that of their galactic conterparts . in particular , at z=0.3 and
if the strength of the primary classification lines , he i 4471 and he i 4387 , were to be considered , these lines would resemble those of a galactic star with _ lower _
( assuming the same and ): about 2 kk lower for sgs and up to 1 kk lower for dws .
vice versa , if compared to galactic counterparts with the _ same observed _ ews in he i , stars with a lower metal content should be hotter . for the strategic he ii 4200 and he ii 4541 transitions
we find the following situation : for intermediate and low temperature dws ( 39 @xmath6 @xmath6 32 kk ) and cooler sgs ( below 34kk ) , the corresponding lines strengths are predicted to remain rather unaffected by metallicity , whilst at higher they are amplified by up to 20 percent , mimicking the lines of a galactic star of _ higher _ .
the effect of metallicity on the spectral classification of o - stars can be evaluated by investigating the behaviour of he i / he ii line ratios instead of individual line strengths .
thus , we have calculated and compared the logarithmic he i(+ii ) 4026/he ii 4200 , he i 4471/he ii 4541 and he
i 4387/he ii 4541 line ratios , as a function of , for the cases of dws , gs and sgs and for the two values of metallicity , and 0.3 .
the results of these calculations are partly illustrated in figure [ he_ratio ] , where the lower and upper limits of the line strength ratios from the mathys @xmath7 - spectral type calibration are overplotted as vertical lines .
the x - position of each of these lines has been placed at the appropriate -value following from the martins et al .
empirical calibration for the corresponding spectral type .
apparently , this criterion to classify o - stars with z=0.3 would lead to temperature differences that increase with and luminosity class .
in particular , for sgs these differences are predicted to range from about 1 to about 2 kk ( i.e. about one sub - type ) , while for dws the expected values are a factor of 2 smaller .
another important result to be noted is that whilst for galactic dws a good correspondence between the calculated and the calibrated values of the he i 4471/he ii 4541 line ratios is found , for cooler sgs ( subtypes later than o7 ) the calculated values are systematically smaller .
this discrepancy results from the ( still not understood ) so - called `` generalized dilution effect '' , which makes the synthetic he i 4471 lines in low temperature sg models weaker than observed ( see repolust et al . @xcite and references therein ) .
since this effect seems to be independent on metallicity , it is not expected to influence our general findings though . in the walborn classification scheme ,
the main luminosity criterion is he ii 4686 . in sgs
this line appears only in emission , while in dws it is observed as a pure absorption feature . since the amount of emission depends on the wind properties , which in turn depend on metallicity , one expects the strength of he ii 4686 to depend on metallicity as well ( see , e.g. , massey et al .
@xcite ) . in figure
3 the behaviour of the he ii 4686 ews ( positive for emission and negative for absorption ) are shown as a function of , for two luminosity classes ( sgs and dws ) and for z= and 0.3 .
expectedly , our calculations for solar metallicity models produce he ii 4686 in emission for sgs and in absorption for dws .
interestingly , however , our calculations predict that the metallicity effects on this line can be so strong that at z=0.3 its sensitivity to luminosity vanishes almost completely , and he ii 4686 appears in absorption for all calculated models .
[ heii4686 ] finally let us note that although an increase in microturbulent velocity ( from 5 to 15 ) is predicted to shift the assigned spectral types by half ( for dws ) to one ( for sgs ) subtype towards later spectral types , this effect does not seem to depend on metallicity . due to this reason model predictions corresponding to = 10 only are shown throughout this paper .
based on a grid of nlte , line blanketed model atmospheres with stellar winds as calculated by means of fastwind , we found that @xmath8 in o - stars the optical helium spectrum can be significantly influenced by metallicity effects , where a decrease in metal content results in stronger he line absorptions _ from both ionization stages _ , where lines from he i are more severely affected than those from he ii .
individual lines react differently in dependence of and luminosity class , but do not depend on microturbulence .
this result is particularly important when _
individual line strengths _ are used to classify stars , as it is the case for the walborn scheme where the detection of weak he i 4471 and he i 4387 is used to fix the subtypes o4 and o6 , respectively . according to our calculations at @xmath9=0.3
, the use of these lines as classification indicators would shift the assigned spectral types by up to one subtype to the later side , provided the detection threshold is the same .
@xmath8 furthermore , our calculations showed that not only individual helium lines , but helium line ratios are also varying with metallicity .
in particular , the use of the he i 4471/he ii 4541 ratio to classify o - stars of @xmath9=0.3 would shift the derived temperature scale by up to 2 kk ( i.e. , about one sub - type ) for sgs , and up to 1 kk ( i.e. , about half a sub - type ) for dws to the hotter side .
the latter result is consistent with the predictions by mokiem et al .
@xcite , who found , using cmfgen ( hillier & miller @xcite ) , that variations in the metal content from z=2 to z=0.1 can change the spectral type of a hot dw by up to one and a half sub - type . extended extragalactic surveys of hot massive stars showed that o - stars in low metallicity environment are systematically hotter than their galactic counterparts with the same spectral type ( e.g. , mokiem et al .
@xcite , massey et al .
@xcite , efremova et al .
2009 ) , thus confirming our principal predictions . however , the observed shifts in are significantly larger than those predicted to appear from a shift in the assigned spectral types , as calculated here and in other work . in particular , for o - type dws temperature differences of 3 to 4 kk , i.e. , a factor of 3 to 4 larger than those predicted by mokiem et al .
@xcite were established between objects in our galaxy and in the smc ( z=0.2 ) would result in a shift of only 1 kk in of a hot dw , i.e. , about half a subtype . ] .
the corresponding results for dws in ngc 6822 are roughly the same ( efremova et al .
@xcite ) .
thus , only part ( @xmath1025 to 30 percent ) of the observed differences in of o - type dws in the smc and in ngc 6822 could be explained by metallicity effects on the spectral type indicators , while the rest must be attributed to other , more `` physical''reasons , such as , e.g. , different stellar structures as a function of metallicity or differences between observed and theoretical wind parameters etc . additional calculations
are required to separate and evaluate these effects .
@xmath8 our calculations predict that the amount of wind emission in he ii 4686 diminishes towards lower metallicities , so that at @xmath00.3 this line would appear only in absorption and could no longer serve as a reliable luminosity indicator .
in particular , using this line to classify stars in low metallicity environments might artificially increase the number of low luminosity ( dws and gs ) o - stars identified in galaxies with lower metallicity , on the expense of the number of o - sgs .
some hints about such effects seem to have been established already .
for instance , massey et al .
@xcite recently noted that in several of their mc targets the he ii 4686 emission is weaker than it might be expected , given their absolute magnitudes , .
and although one might argue that a discrepantly high could be due to spectroscopic binarity , an alternative explanation in terms of metallicity effects is also possible , accounting for our present results .
moreover , the number of sgs identified in the smc appears to be less than expected ( d. lennon , priv .
comm . ) , in agreement with what might be expected from our calculations .
_ acknowledgments : _ part of this work was performed during a visit of nm at the dept . of physics and astronomy of the johns hopkins university .
the work was supported by nasa grant nag5 - 9219 ( nra-99 - 01-ltsa-029 , pi bianchi ) .
j.p . acknowledges a travel grant by the dfg ( under grant pu 117/6 - 1 ) and by the bg nsf ( under grant f-1407/2004 ) . | based on an exteded grid of nlte , line blanketed model atmospheres with stellar winds as calculated by means of fastwind , we have investigated the change in the strengths of strategic helium transitions in the optical as caused by a 0.3 decrease in metallicity with respect to solar abundances .
our calculations predict that only part of the observed increase in of o - type dwarfs could be explained by metallicity effects on the spectral type indicators , while the rest must be attributed to other reasons ( e.g. , different stellar structures as a function of metallicity or differences between observed and theoretical wind parameters etc . ) .
in addition , we found that using the he ii 4686 line to classify stars in low metallicity environments ( @xmath0 0.3 ) might artificially increase the number of low luminosity ( dwarfs and giants ) o - stars , on the expense of the number of o - supergiants . |
several studies , both observational and computational , of groups and clusters of galaxies , and of individual early - type galaxies , assume that the hot gas present in these systems is in hydrostatic equilibrium ( he ) in the overall gravitational potential .
even though the assumption of he has been shown to not always be robust ( e.g. bertin et al . 1993 ; ciotti & pellegrini 2004 ) , many authors assume an isothermal @xmath2-model ( cavaliere & fusco - femiano 1976 , 1978 ) to describe the gas profile , but several solve the he equation , assuming that the gravitational potential is due to dark matter ( dm ) only , usually described by an nfw ( navarro , frenk & white 1996 ) profile ( e.g. makino , sasaki & suto 1998 ; suto , sasaki & makino 1998 ) , or by more general profiles ( e.g. ciotti & pellegrini 2008 ) .
while this approximation seems to hold for groups and clusters of galaxies , it is well known that stars contribute a significant fraction of the total and baryonic matter of elliptical galaxies , and even become the dominant part at the very centre of these systems ( e.g. ferreras , saha & williams 2005 ) .
the purpose of this paper is to extend previous calculations that only considered the potential due to dm , and explicitly include the contribution of the stellar component as well .
providing an analytic fitting formula for calculating the response of the gas to the presence of the gravitational potential induced by both dm and stars is useful while interpreting x - ray data of early - type galaxies and for setting up initial conditions for simulations of elliptical galaxies .
the derivation of an equilibrium gas profile neglecting the effects of stars leads to an incorrect estimate of the enclosed gas mass , particularly in the central regions of the galaxy , and therefore has implications for the evaluation of the baryon fraction as a function of radius , and for the calculation of the rate of accretion onto a central supermassive black hole .
moreover , the known relations ( e.g. makino et al . 1998 ) between observed properties of the hot gas , described via a @xmath2-model , and the parameters of the underlying dm gravitational potential , are significantly affected by the inclusion of the stellar component . in addition , an analytic formula that includes the effects of stars can be important to provide a more realistic formulation of the initial equilibrium conditions for simulations of galaxy formation and evolution ( e.g. di matteo , springel & hernquist 2005 ; springel , di matteo & hernquist 2005 ; hopkins et al . 2006 ) .
it has long been observed , both at low redshift ( e.g. graham et al .
1996 ; kormendy et al . 2009 ) and high redshift ( e.g. mancini et al . 2010 ; van dokkum et al .
2010 ) , that the surface brightness of many elliptical galaxies can be well described by the empirically derived de vaucouleurs ( 1948 ) profile or , more generally , by the srsic ( 1963 , 1968 ) profile .
unfortunately , the de - projected stellar mass density profile can not be calculated analytically ( but see young 1976 for a numerical study ) .
there is , however , a family of stellar mass density profiles ( dehnen 1993 ; see also tremaine et al .
1994 ) that also describes observed properties well .
not only is the dehnen family analytically simple , but also , for a defined range of parameters , it quite successfully resembles the de vaucouleurs profile in both surface density and distribution function .
furthermore , it includes , as special cases , the widely used jaffe ( 1983 ) and hernquist ( 1990 ) profiles .
it is this family of density profiles that we investigate further in this paper , and use to model the stellar component . throughout the paper
, we use the following values for cosmological parameters when needed : @xmath4 , @xmath5 , @xmath6 , @xmath7 km s@xmath8 mpc@xmath9 km s@xmath8 mpc@xmath8 , @xmath10 ( i.e. @xmath11 ) .
the outline of the paper is as follows . in section [ sec : formulation ] , we formulate the problem and derive the new formalism for he gas profiles in the presence of dm and stars . in section [ sec : analysis ] , we analyse the he solutions for a realistic model of an isolated elliptical galaxy , compare the results for a wide range of parameters , and present new relations between @xmath2-model and dm parameters .
we conclude in section [ sec : conclusions ] with a discussion of our results and their implications for interpreting x - ray observations , and for simulations of elliptical galaxies .
we model the composite system of gas in equilibrium with stars and a dm halo . to start with
, we consider the case of an ideal gas in a generic total gravitational potential @xmath12 with a finite central value @xmath13 , and solve the he equation , @xmath14 , where @xmath15 and @xmath16 refer to the pressure and mass density of the gas .
note that the general equation simplifies in the case of spherical symmetry to the familiar form @xmath17 where we have made use of poisson s equation , @xmath18 , and @xmath19 and @xmath20 are the total enclosed mass and total mass density , respectively .
if the gas is barotropic , i.e. @xmath21 , we can further write @xmath22 where the free parameter @xmath23 was chosen as the boundary condition . for simplicity , we only consider the two common barotropic cases : the isothermal and the polytropic case . in the isothermal ideal gas case ,
pressure and density are related by @xmath24 \rho(r ) = ( p_0/\rho_0 ) \rho(r)$ ] , where @xmath25 is the boltzmann constant , @xmath26 is the gas mean molecular weight , @xmath27 is the proton mass , and @xmath28 and @xmath29 are the central gas temperature and density , respectively .
the solution to the he equation in this case can be reduced to the familiar integral : @xmath30 = \rho_0\exp\left[-\frac{g\rho_0}{p_0 } \int_0^r{\frac{m_t(r')}{r'^2}\,dr'}\right ] , \label{eq : iso_sol}\ ] ] where @xmath31 . for the polytropic ideal gas case ,
pressure and density are related by @xmath32 , where @xmath33 and @xmath34 is the polytropic index .
the solution to the he equation in this case is : @xmath35^{\frac{1}{\gamma-1 } } = \rho_0 \left[1-\frac{\gamma-1}{\gamma}\frac{g \rho_0}{p_0}\int_0^r{\frac{m_t(r')}{r'^2}\,dr'}\right]^{\frac{1}{\gamma-1}}. \label{eq : poly_sol}\ ] ] it is worth pointing out that the polytropic solution is only physically meaningful if the term in square brackets of the r.h.s of equation is non - negative .
moreover , the isothermal quantities ( and , in most cases , the polytropic quantities ) are non - zero at infinity : with our choice of the boundary condition , regardless of the slope of the gravitational potential at large radii , only a potential which is infinite at the origin would be able to cause the gas density at infinity to be zero .
for this reason , results at large radii should be treated with caution .
see bulbul et al .
( 2009 ) for an alternate approach in the polytropic case , where they impose @xmath36 as the boundary condition .
this is equivalent to setting @xmath37 in equation .
we further assume that gas does not contribute to the total gravitational potential ( i.e. we neglect gas self - gravity ) , and therefore solve the simplified he equation instead , where now @xmath12 is the gravitational potential due to all mass _ excluding _ gas , and @xmath38 is the total mass _ excluding _ gas .
this assumption is supported , especially for the inner regions , by previous studies ( e.g. suto et al .
1998 ) , and is justified by the fact that gas is _ globally _ negligible , i.e. the enclosed gas mass is one or more orders of magnitude smaller than the total enclosed mass for all radii up to the virial radius .
stars , on the other hand , while having a virial enclosed mass comparable to that of the gas , are not negligible in the inner parts of the galaxy : in fact , they represent the dominant matter fraction at least up to the effective radius ( e.g. ferreras et al . 2005 ) .
we initially consider a system with gas and dm only , and assume that the dm density can be described by an nfw profile , @xmath39 where @xmath3 is a scale radius , @xmath40 is a characteristic ( dimensionless ) density , and @xmath41 is the critical density of the universe . the corresponding gravitational potential and enclosed mass are @xmath42 , \label{eq : nfw_pot_and_mass}\ ] ] respectively , where @xmath43 , and we have assumed @xmath44 . substituting these into equations and , we have , as in makino et al .
( 1998 ) for the isothermal case , @xmath45 , \label{eq : makino_sol}\ ] ] and , as in suto et al .
( 1998 ) for the polytropic case , @xmath46^{\frac{1}{\gamma-1 } } , \label{eq : suto_sol}\ ] ] where @xmath47 .
now we move on to the case of an nfw dm halo with the stellar component modelled using a dehnen profile , neglecting the effects of adiabatic contraction .
the dehnen ( 1993 ; see also tremaine et al .
1994 ) family of stellar mass density profiles is given by
@xmath48 where @xmath49 , @xmath50 is the total stellar mass ( i.e. integrating from the origin to infinity ) , and @xmath51 is a scale radius . the dehnen profile is useful as its projected distribution accurately matches the de vaucouleurs ( 1948 ) surface density profile which is an excellent fit to observations .
this is especially true for values of @xmath52 ranging between 1 ( which corresponds to the hernquist profile ) and 2 ( the jaffe profile ) , with @xmath53 being the value for which the match is closest .
the dehnen density profile corresponds to a gravitational potential @xmath54 } & \mbox{if $ \psi \neq 2 $ , } \cr
\displaystyle{\ln\left(\frac{r}{r+r_*}\right ) } & \mbox{if $ \psi = 2 $ , } \cr \end{cases}\ ] ] where we have assumed @xmath55 . since equations and require a finite central gravitational potential , we only consider dehnen profiles with @xmath56 , for which the gravitational potential and enclosed mass are @xmath57 , \hspace{0.5 cm } m_d(r ) = m _ * \left(\frac{r}{r+r_*}\right)^{3-\psi } , \label{eq : dehnen_pot_and_mass}\ ] ] respectively , where @xmath58 $ ] .
we consider a system that consists of an nfw dm halo and stars described by a dehnen profile , for which the total gravitational potential and total enclosed mass are given by explicitly adding the contributions of the stars and dm given in equations and , respectively .
using equations and , we obtain @xmath45 \exp\left[-\delta_d\left(\frac{r / r_*}{1+r / r_*}\right)^{2-\psi}\right ] \label{eq : capelo_iso}\ ] ] for the isothermal case , and @xmath59^{\frac{1}{\gamma-1 } } \label{eq : capelo_poly}\ ] ] for the polytropic case , where @xmath60 $ ] .
notice that equations ( [ eq : capelo_iso]-[eq : capelo_poly ] ) reduce to equations ( [ eq : makino_sol]-[eq : suto_sol ] ) when @xmath61 .
in this section , for a system with an nfw dm halo and stars described by a dehnen profile , we compare equilibrium gas density profiles as a function of @xmath62 , the stellar mass fraction at the virial radius , @xmath52 , the dehnen parameter , and @xmath28 , the central gas temperature .
we also consider the case when stars are included in the total mass budget , but not in the total gravitational budget .
we consider a system at redshift @xmath63 , which consists of an isolated elliptical galaxy at the centre of an isolated , collapsed dm halo , wherein the density is given by @xmath64 where @xmath65 is the dm mass enclosed within the virial radius @xmath66 , and @xmath67 is a function of the concentration parameter @xmath68 , which depends on redshift and ( total ) virial mass @xmath69 as @xmath70 where @xmath71 and @xmath72 are constants that can be inferred from simulations ( e.g. bullock et al .
2001 ; hennawi et al . 2007 ) and observations ( e.g. comerford & natarajan 2007 ; mandelbaum , seljak & hirata 2008 ) , and @xmath69 is in units of m@xmath73 .
the virial radius is , by definition , the radius of a spherical region within which the mean ( total ) mass density is @xmath74 times the mean critical density ; other authors , e.g. nfw , use @xmath75 . ] : @xmath76 where the virial overdensity @xmath77 ( bryan & norman 1998 ) , @xmath78 , and @xmath79 .
note that other authors ( e.g. bullock et al .
2001 ) use equivalent but different definitions , @xmath80 and @xmath81/\omega_m(z)$ ] , so that @xmath82 . ] , the matter fraction @xmath83 $ ] , and the mean critical density @xmath84/(8 \pi g)$ ] .
the nfw density profile function can be also re - written as in equation , with @xmath85 , and @xmath86 , where @xmath87 and @xmath88 are the dm and baryon fraction , respectively , within the virial radius , @xmath89 is the baryon fraction relative to the universal value , and we have assumed that the universal baryon fraction , @xmath90 , is independent of redshift .
the elliptical galaxy stellar density profile is given by equation .
the total stellar mass @xmath50 is calculated by setting @xmath91 in equation and by imposing ) and by varying the dehnen parameter , the total stellar mass @xmath50 will necessarily vary , albeit only very slightly . ]
the virial stellar mass fraction can be written , following e.g. mamon & okas ( 2005b ) , as @xmath93 where @xmath94 is the stellar mass - to - light ratio , @xmath95 is the universal mass - to - light ratio , @xmath96 is the galactic mass - to - light ratio ( all in m@xmath97l@xmath73 ) , and @xmath98 is the ( dimensionless ) mass - to - light ratio bias , all in the b - band . the scale radius @xmath51 is related to the effective radius @xmath99 through a numerical fit ( dehnen 1993 ) , @xmath100 whereas the effective radius itself can be derived from empirical fits to observations does not depend on the stellar fraction . ]
( mamon & okas 2005a ) as @xmath101 , where @xmath102 , @xmath103 , and @xmath104 and @xmath69 are in units of l@xmath73 and m@xmath73 , respectively .
finally , we can calculate the central gas density @xmath29 by imposing @xmath105 where we have assumed that all matter is comprised of dm , gas , and stars only ( i.e. we have neglected other forms of matter , such as dust , and the presence of a supermassive central black hole ) , and that @xmath106 is given by equation or when stars are included in the gravitational potential , and by equation or when stars are excluded .
note , however , that the gas virial mass is the same with and without the stellar component in the gravitational potential , since stars are included in the total mass budget in both cases .
we note that a very small number of parameters are needed to compute the equilibrium profiles for the gas .
these are , for the dm halo system , the redshift @xmath63 , the ( total ) virial mass @xmath69 of the system , the fitting parameters @xmath71 and @xmath72 of the empirical relation between concentration and total virial mass , and the baryon fraction @xmath89 relative to the universal value .
for the stars , the parameters are the galactic b - band mass - to - light ratio @xmath96 ( or , alternatively , the universal b - band mass - to - light ratio @xmath95 and the b - band mass - to - light ratio bias @xmath98 ) , the stellar b - band mass - to - light ratio @xmath94 , and the dehnen parameter @xmath52 . finally , for the gas , the parameters are the mean molecular weight @xmath26 , the polytropic index @xmath107 , and the central temperature @xmath28 ( or , alternatively , the ratio between central density and pressure @xmath108 ) .
in our fiducial model , we use the mamon & okas ( 2005b ) mass - to - light ratio values and choose @xmath109 , @xmath110 m@xmath97l@xmath73 ( so that @xmath111 m@xmath97l@xmath73 precisely ) , and @xmath112 m@xmath97l@xmath73 . as for the halo parameters ,
we choose a system with universal baryon fraction ( @xmath113 ) at redshift @xmath114 , a ( total ) virial mass @xmath115 m@xmath73 , and the @xmath116-@xmath69 relation parameters @xmath117 and @xmath118 ( bullock et al . 2001 ) .
finally , we fix the central temperature of the gas to be @xmath0 kev ( diehl & statler 2008 ) , the mean molecular weight of the gas to be @xmath119 , the polytropic index to be @xmath120 , and the dehnen parameter to be @xmath121 ( hernquist 1990 ) .
all these quantities are listed in table [ tab : parameters ] , in addition to the parameters calculated throughout this section .
the choice of the central temperature @xmath28 is particularly important , since it greatly influences the values of @xmath122 and @xmath123 .
if we use the ( cluster and group ) mass - temperature ( m - t ) relation ( e.g. sanderson et al . 2003 ) or , rather , extrapolate the m - t relation down to the mass ranges typical of early - type galaxies , we obtain e.g. @xmath124 kev for a system with @xmath115 m@xmath73 .
however , the extrapolation of the m - t relation is not without risk , as there is some evidence for a gradual steepening of the relation , with decreasing mass ( sanderson et al .
using individual observations of isolated elliptical galaxies , the central temperatures measured are usually slightly higher , ranging from @xmath125 kev ( e.g. memola et al . 2009 ) to @xmath126 kev ( e.g. osullivan & ponman 2004 ; osullivan , sanderson & ponman 2007 ) . finally , diehl & statler ( 2008 ) give a convincing argument for a central temperature @xmath127 kev , using observations of 36 normal early - type galaxies .
for this reason , here we study cases with three different central temperatures : @xmath128 kev . , and for two models with lower ( left panels ) and higher ( right panels ) gas temperature , we plot in the top ( middle ) panels , the mass density ( enclosed mass ) profiles of dm ( red , dotted curve ) , stars ( red , dashed curve ) , and isothermal gas ( blue , dot - dashed curve ) in he with dm and stars .
the black , solid curve represents the isothermal gas in he with dm only . in the bottom panels ,
we plot ( solid curve ) the fractional change of enclosed gas mass between the case when stars are included in the total gravitational potential , and the case when they are not ; we also plot ( dashed curve ) the ratio between the change in gravitational energy of a gas particle , due to the addition of the stellar component , and its thermal energy .
the four vertical lines denote the stellar scale radius @xmath51 , the effective radius @xmath99 , the nfw scale radius @xmath3 ( solid line ) , and the virial radius @xmath66 , from left to right , respectively .
the inclusion of the stellar component dramatically alters the gas profile at the very centre of the early - type galaxy , with the gas being more centrally concentrated , due to a deeper and steeper gravitational potential well .
the effect of gas concentration is more pronounced for lower values of the gas temperature.,width=434 ] m@xmath97l@xmath73 , @xmath129 ; left panels ) , whose parameters are listed in table [ tab : parameters ] , and for two models with lower stellar baryon fraction ( @xmath130 m@xmath97l@xmath73 , @xmath131 , central panels ; and @xmath132 m@xmath97l@xmath73 , @xmath133 , right panels ) , we plot in the top ( middle ) panels , the mass density ( enclosed mass ) profiles of dm ( red , dotted curve ) , stars , and isothermal gas in he with dm and stars .
the stellar ( gas ) profiles are the three upper , red ( lower , blue ) bundled curves , for three values of the dehnen parameter : @xmath134 ( dashed curve ) , @xmath121 ( solid curve ) , and @xmath135 ( dot - dashed curve ) .
the black , long - dashed curve represents the gas profile when the stellar component of the total gravitational potential is neglected . in the bottom panels , we plot the fractional change of enclosed gas mass between the case when stars are included in the total gravitational potential , and the case when they are not , for the same three values of the dehnen parameter .
the three vertical lines in each panel denote the effective radius @xmath99 , the nfw scale radius @xmath3 ( solid line ) , and the virial radius @xmath66 , from left to right , respectively .
the effect of gas concentration is more pronounced for higher values of the stellar mass fraction and for higher values of the dehnen parameter.,width=434 ] m@xmath97l@xmath73 ( @xmath133 ) , for two different gas temperatures : @xmath0 kev ( black , solid lines ) and @xmath136 kev ( black , dashed lines ) . the dotted , red curves are the best - fitting @xmath2-models in the @xmath137 range , denoted by the two dotted vertical lines .
the inclusion of stars significantly affects the relation between the @xmath2-model scale radius and the nfw scale radius . for higher stellar mass fractions ( not shown here ) , a @xmath2-model is unable to accurately reproduce our solution .
when the observed surface brightness profile of an isolated elliptical galaxy is found to be well fitted by a @xmath2-model , the stellar mass fraction can not be much greater than @xmath110%.,width=411 ] in fig .
[ fig : dependence_on_t ] we plot the density ( top panels ) and enclosed mass ( middle panels ) profiles of dm , stars , and isothermal gas , for the fiducial model ( with @xmath0 kev ; central panels ) and for the lower and higher temperature models ( with @xmath138 kev , left panels ; and @xmath136 kev , right panels ) , with and without the stellar component in the gravitational potential .
we focused on the isothermal solution because the radial range for which the polytropic solution is non - negative is limited kev . ] .
the inclusion of the stellar component dramatically alters the equilibrium profiles for the gas at the very centre of an early - type galaxy , with the gas being more centrally concentrated , due to a deeper and steeper gravitational potential well .
the presence of the stellar component is thus explicitly taken into account and is manifested in the modification to the gas profile in the central region , in contrast to the case when stars are not included in the total gravitational potential . in the bottom panels of fig .
[ fig : dependence_on_t ] , we plot the fractional change of enclosed gas mass @xmath139 as a function of radius : for the fiducial model , in the central panel , the gas concentration ratio between the _ new _ ( i.e. including stars in the gravitational potential ) and the _ old _ ( i.e. excluding stars ) enclosed gas mass is @xmath140 \equiv f_{conc } \sim 4 $ ] at the stellar scale radius @xmath51 , @xmath141 at the effective radius @xmath99 , and @xmath142 at the nfw scale radius by 2 , fixing all other parameters , and found that the effect of gas concentration increases as the effective radius decreases , as expected . ] , @xmath3 . we note that the effect of gas concentration increases with decreasing temperature , because the ratio between the change in gravitational energy of a gas particle , due to the addition of the stellar component , and its thermal energy , decreases with increasing temperature , as is shown in the bottom panels of fig .
[ fig : dependence_on_t ] .
we also point out that our gas ( density and enclosed mass ) profile is very different from that found in mamon & okas ( 2005b ) .
they assume a fixed @xmath2-model to describe the isothermal gas in the galaxy , and do not solve the he equation directly .
note that , by construction , the enclosed gas mass at the virial radius is the same .
it is worth pointing out that the value of @xmath62 of the fiducial model corresponds to a value of stellar baryon fraction @xmath129 , which is higher than that inferred from recent gravitational lensing studies , for which @xmath143 ( e.g. hoekstra et al .
2005 ; mandelbaum et al .
2006 ; heymans et al . 2006 ; lagattuta et al .
2009 ; see also fukugita , hogan & peebles 1998 ) .
the discrepancy is likely due to different values of mass - to - light ratios and of mass - to - light ratio bias .
lagattuta et al .
( 2009 ) , for example , use a v - band virial mass - to - light ratio @xmath144 m@xmath97l@xmath73 and a v - band stellar mass - to - light ratio @xmath145 m@xmath97l@xmath73 . using these values , we would obtain @xmath143 .
for this reason , we compare the gas radial profiles for three different values of @xmath146 m@xmath97l@xmath73 , which correspond to stellar mass fractions @xmath147 , and to stellar baryon fractions @xmath148 , respectively . for these values of @xmath94
, we also vary the dehnen parameter to explore dependencies on stellar profiles , choosing values of @xmath134 , 1 ( hernquist 1990 ) , and 1.5 ( the `` best '' value in dehnen 1993 ) . fig . [ fig : dependence_on_stars ] shows the density ( top panels ) and enclosed mass ( middle panels ) profiles for the three chosen values of @xmath94 and of @xmath52 , in the same way we plotted fig .
[ fig : dependence_on_t ] , for a system with @xmath149 kev .
the effect of gas concentration is obviously more pronounced for higher values of the stellar mass fraction and for higher values of the dehnen parameter , as it is shown in the bottom panels .
however , the variation of the dehnen parameter does not cause significant changes in the enclosed gas mass , except in the very inner - most regions of the galaxy ( @xmath150 ) : for example , for the @xmath112 m@xmath97l@xmath73 case , the enclosed gas mass at the effective radius varies by less than 1% when changing the dehnen parameter from @xmath121 to @xmath134 , or to @xmath135 . on the other hand ,
the change in stellar fraction is much more evident : for example , for the hernquist case ( @xmath121 ) , the ratio between the enclosed gas masses at the effective radius @xmath99 varies from @xmath13 in the @xmath112 m@xmath97l@xmath73 case , to @xmath12 in the @xmath130 m@xmath97l@xmath73 case , to @xmath11.3 in the @xmath132 m@xmath97l@xmath73 case . by comparing the theoretically derived equilibrium solution for an isothermal gas to the observationally inspired @xmath2-model profile ( cavaliere & fusco - femiano 1976 , 1978 ) , it is possible to derive relations between observed properties of the hot gas and the parameters of the underlying dm gravitational potential of a given system .
makino et al .
pursue this ( 1998 ; see also suto et al .
1998 and wu & xue 2000 for other dm profiles ) for the case of clusters of galaxies with a dm halo described by an nfw profile .
they fit their he analytic solution for an isothermal gas ( our equation [ eq : makino_sol ] ) to a generic @xmath2-model profile given by @xmath151^{3\beta/2 } } = \frac{\rho_0(a_0+a_1\delta_{nfw})}{[1+[r/(a_2r_s)]^2]^{3a_3\delta_{nfw}/2}},\ ] ] where @xmath152 , @xmath153 , and @xmath2 are the central density , the scale radius , and the characteristic slope parameter of the @xmath2-model , respectively , and @xmath154 ( for @xmath155 0 - 3 ) are fitting parameters , in the @xmath137 range , obtaining @xmath156 and @xmath157 , for a wide range of values of @xmath122 .
we performed a non - linear least squares fitting with the idl routine mpfit ( markwardt 2009 ) , using our own values of @xmath122 , and obtained similar results in the case when stars are not included in the total gravitational potential , regardless of the value of the stellar mass fraction and of the gas temperature : @xmath158 and @xmath159 in the @xmath137 range .
note that , even though the @xmath2-@xmath122 relation is independent of temperature , the value of the characteristic slope parameter @xmath2 does depend on temperature through @xmath122 .
now we go a step further and compare the @xmath2-model profile to the isothermal he gas profile in the case when stars are included in the gravitational potential , given by equation . since we found that there is no strong dependence on the dehnen parameter , we focus on the hernquist case ( @xmath160 ) for simplicity . also , we choose the stellar b - band mass - to - light ratio to be @xmath132 m@xmath97l@xmath73 , as the inferred stellar baryon fraction @xmath133 is closer to what found from gravitational lensing observations ( e.g. hoekstra et al .
2005 ) , and because higher stellar mass fraction systems do not fit the @xmath2-model profiles as well . the fact that it is indeed possible to fit x - ray observations of early - type galaxies with @xmath2-model profiles ( e.g. osullivan , ponman & collins 2003 ) , is an additional indication that the stellar baryon fraction of such galaxies can not be much higher than @xmath110% . in fig .
[ fig : beta_model_comparison ] we show the equilibrium gas density profiles for an isothermal gas in a system with nfw dm and hernquist stars , for two different temperatures , including and excluding the effects of stars on the total gravitational potential .
all he gas profiles are fitted to @xmath2-model profiles in the @xmath137 range .
we chose this particular range to be able to make consistent comparisons with the results of makino et al .
( 1998 ) , and also because it is typically difficult to obtain precise x - ray measurements at larger radii ( due to the low surface brightness of the gas ) and at smaller radii ( due to resolution limitations ) .
we note that , whereas the relation between @xmath2 and @xmath122 does not change significantly ( @xmath159 excluding stars , and @xmath161 including stars ) , the @xmath153-@xmath3 relation between the two scale radii varies substantially , from @xmath158 , when stars are not included in the potential , to @xmath162 , when stars are included .
we note that the particular constant of proportionality between @xmath153 and @xmath3 depends on the assumed stellar mass fraction .
using the @xmath153-@xmath3 relation can therefore lead to an underestimate of the nfw scale radius , when stars are not taken into account . defining @xmath163 as the ratio between the @xmath2-model inferred value of @xmath3 when stars are included in the gravitational potential ( the _ new _ value ) , and when they are not ( the _ old _ value ) , we have @xmath164 . assuming the virial mass @xmath69 and radius @xmath66 of the system
are known , we then have the ratio of concentration parameters @xmath165 .
therefore , the enclosed dm mass at @xmath166 would be overestimated by a factor of @xmath167/[\ln(1+c_{vir - new})-c_{vir - new}/(1+c_{vir - new } ) ] \sim 2 $ ] .
we solve the he equation analytically for a system with dm described by an nfw profile , and stars described by a dehnen profile , for the two common barotropic cases ( the isothermal case and the polytropic case ) .
we thus extend previous calculations which only considered the potential due to dm . in doing so , we do assume spherical symmetry , and we neglect the effects of gas self - gravity and adiabatic contraction .
we compare the newly obtained equilibrium solutions as a function of stellar mass fraction , central gas temperature , and the dehnen parameter , to the case when the stellar component is included in the total mass budget but not in the total gravitational potential .
we determine that neglecting the effects of stars leads to an underestimate of the enclosed gas mass in the inner regions of the galaxy .
we then present new relations between @xmath2-model and dm parameters , and find that the inclusion of the stellar component is necessary to avoid an underestimate of the nfw scale radius .
in particular : * for a system with nfw dm and hernquist stars at redshift @xmath114 , with virial mass @xmath115 m@xmath73 , and central temperature @xmath0 kev , we find in the isothermal case that the gas concentration ratio @xmath168 between the _ new _
( i.e. including stars in the gravitational potential ) and the _ old _
( i.e. excluding stars ) enclosed gas mass , can be as high as @xmath13 at the effective radius , and as high as @xmath11.2 at the nfw scale radius , for a stellar baryon fraction @xmath129 . *
the effect of gas concentration decreases with decreasing stellar baryon fraction as expected , with @xmath169 varying from @xmath13 in the @xmath129 case , to @xmath12 in the @xmath131 case , to @xmath11.3 in the @xmath133 case .
* on the other hand , the effect of gas concentration is virtually independent of the dehnen parameter , except in the very inner - most regions of the galaxy ( @xmath150 ) , with the value of @xmath169 changing by @xmath11% when varying the dehnen parameter in the @xmath170 range .
* we further determine that the effect of gas concentration decreases with increasing temperature , as expected .
* we find similar results for the polytropic case at higher temperatures , where the solution is physically meaningful .
in fact , for a fixed central gas temperature , the effect of gas concentration is more pronounced than in the isothermal case , because the polytropic gas has a lower mean temperature .
* we derive new relations between observed properties of the hot gas , described via a @xmath2-model , and the parameters of the underlying dm gravitational potential of early - type galaxies , which avoid an underestimate of the nfw scale radius by @xmath180% , and an overestimate of the enclosed dm mass at @xmath3 by a factor of @xmath12 , assuming a 10% stellar baryon fraction .
* models with higher stellar mass fraction do not fit @xmath2-models well , further indicating that the stellar baryon fraction of isolated elliptical galaxies whose observed surface brightness profile is well fitted by a @xmath2-model can not be much higher than @xmath110%
. these new analytic solutions will be useful for the interpretation of x - ray data of elliptical galaxies and in particular , will help derive unbiased estimates of the baryon mass fraction as a function of radius .
moreover , the formalism in equations ( [ eq : capelo_iso]-[eq : capelo_poly ] ) provides a new prescription for setting up more realistic initial conditions for simulations of galaxy formation ( e.g. di matteo et al . 2005 ; springel et al .
2005 ; hopkins et al . 2006 ) .
prc is grateful for many helpful discussions with andrew szymkowiak . | we present an analytic formulation for the equilibrium gas density profile of early - type galaxies that explicitly includes the contribution of stars in the gravitational potential .
we build a realistic model for an isolated elliptical galaxy and explore the equilibrium gas configurations as a function of multiple parameters . for an assumed central gas temperature @xmath0 kev
, we find that neglecting the gravitational effects of stars , which can contribute substantially in the innermost regions , leads to an underestimate of the enclosed baryonic gas mass by up to @xmath165% at the effective radius , and by up to @xmath115% at the nfw scale radius , depending on the stellar baryon fraction .
this formula is therefore important for estimating the baryon fraction in an unbiased fashion .
these new hydrostatic equilibrium solutions , derived for the isothermal and polytropic cases , can also be used to generate more realistic initial conditions for simulations of elliptical galaxies . moreover ,
the new formulation is relevant when interpreting x - ray data .
we compare our composite isothermal model to the standard @xmath2-model used to fit x - ray observations of early - type galaxies , to determine the value of the nfw scale radius @xmath3 . assuming a 10% stellar baryon fraction , we find that the exclusion of stars from the gravitational potential leads to ( i ) an underestimate of @xmath3 by @xmath180% , and to ( ii ) an overestimate of the enclosed dark matter at @xmath3 by a factor of @xmath12 , compared to the equivalent @xmath2-model fit results when stars are not taken into account . for higher stellar mass fractions ,
a @xmath2-model is unable to accurately reproduce our solution , indicating that when the observed surface brightness profile of an isolated elliptical galaxy is found to be well fitted by a @xmath2-model , the stellar mass fraction can not be much greater than @xmath110% .
dark matter galaxies : elliptical and lenticular , cd galaxies : ism galaxies : structure galaxies : haloes galaxies : stellar content |
in @xcite and @xcite we presented new evidence for a smaller fine structure constant , @xmath5 , in high redshift quasar absorption clouds : @xmath6 .
this preliminary result is the weighted mean from a many - multiplet ( mm ) analysis of 128 absorption systems and represents a significant improvement on our previously published result from 49 systems @xcite : @xmath7 .
such a potentially fundamental result requires extreme scrutiny , including a thorough consideration of possible systematic errors . in @xcite
we explored a wide range of possibilities : laboratory wavelength errors , wavelength calibration errors , atmospheric dispersion effects , unidentified interloping transitions , isotopic ratio and/or hyperfine structure effects , intrinsic instrumental profile variations , spectrograph temperature variations , heliocentric velocity corrections , kinematic effects and large scale magnetic fields .
all but two of these atmospheric dispersion effects and isotopic abundance variation were shown to be negligible . here
we present an updated analysis of these two effects .
we also present new results from a line removal test which constrains systematic errors associated with individual transitions or species .
[ fig : syserr]a illustrates how atmospheric dispersion results in an angular separation between different wavelengths as they enter the spectrograph slit
. this separation increases with increasing zenith angle , @xmath8 . for
the particular optical design of the keck / hires , the component of angular separation across the slit ( i.e. in the wavelength direction as projected from the ccd ) will cause a _ compression _ of the quasar spectrum .
such a compression will mimic @xmath9 for the mg / feii systems observed at @xmath10 , the effect being complicated at high-@xmath11 by the diversity of @xmath12 coefficients coefficient measures the sensitivity of a transition wavelength to changes in @xmath1 .
see @xcite for details and updated values . ]
( see fig .
[ fig : rmline]a ) .
a wavelength dependent asymmetry will also be introduced into the instrumental profile ( ip ) due to truncation of the seeing discs by the slit edges .
these effects will not be applied to the thorium - argon ( thar ) calibration spectra since the thar lamp illuminates the slit uniformly in all cases .
an image rotator is usually employed to hold the slit axis perpendicular to the horizon as the quasar is tracked across the sky .
this eliminates the compression and ip asymmetry effects .
only 51 of our 128 absorption systems were observed in this way ; 77 systems were observed before august 1996 when keck / hires was fitted with an image rotator .
however , the values of @xmath4 for the 77 affected and 51 unaffected systems are consistent , with weighted mean values @xmath13 and @xmath14 respectively . we also expect to observe a correlation between @xmath4 and @xmath8 , especially for the low-@xmath11 mg / feii systems ( see fig .
[ fig : rmline]a ) .
this expectation is confirmed in the simulations below .
however , no such correlation is observed in our quasar data .
therefore , the data themselves suggest a negligible effect on @xmath4 due to atmospheric dispersion effects .
nevertheless , we used the atmospheric and telescope parameters for each quasar observation to model the potential effect of atmospheric dispersion on @xmath4 .
the angular separation of two wavelengths entering the slit is determined by the refractive index of air at the telescope ( which depends on temperature , pressure and relative humidity ) .
the slit angle with respect to the vertical , @xmath15 , is equal to @xmath8 for the pre - rotator keck / hires .
we applied the resulting wavelength shifts to a high s / n simulation of our data and determined @xmath4 , giving a correction which we apply to the values measured from the quasar spectra .
we compare the corrected values of @xmath4 with the uncorrected values in fig .
[ fig : syserr]c .
overall , our model of atmospheric dispersion has an opposite effect on the low-@xmath11 systems compared to the high-@xmath11 systems : it can not explain the observed deviation from @xmath16 .
once we apply the correction , the apparent evolution of @xmath1 with look - back time is enhanced .
mg , si , cr , fe , ni and zn have several stable isotopes and so the transitions used in the mm method will have unresolved isotopic structure .
however , only the isotopic structures of the mg lines have been determined experimentally @xcite .
the mg and si transitions involve similar energy levels and so , in @xcite , we estimated their isotopic structure by scaling the mgii @xmath172796 structure by the mass isotopic shift .
we fitted the mg and si absorption lines with these isotopic structures using the terrestrial values of the isotopic abundances @xcite
. however , galactic observations of mg @xcite and theoretical models of si @xcite in stars strongly suggest that only the @xmath18 mg and @xmath19si isotopes will exist in quasar absorption clouds with significant abundances . fig .
[ fig : syserr]b illustrates the effect this would have on a measured mgi @xmath172852 absorption line position ( the isotopic separations are exaggerated by a factor of 25 ) . to obtain an upper limit on the effect of isotopic ratio evolution on @xmath4 , we refit the mg and si transitions using only the @xmath18 mg and @xmath19si isotopes , i.e. zero abundances of @xmath20 mg and @xmath21si in the absorption clouds .
we compare these new , corrected values of @xmath4 with the uncorrected values in fig .
[ fig : syserr]c ( middle panel ) .
as expected , the potential effect is most prominent in the low-@xmath11 mg / feii systems .
note that @xmath4 is more negative after the correction : isotopic ratio evolution can not explain our results .
although the isotopic structures of the other transitions in our analysis are not known , they should be more compact since the mass isotopic shift is inversely proportional to the square of atomic mass .
the line removal test described below largely rules out such strong effects .
a non - zero @xmath4 manifests itself as a distinct pattern of line shifts ( fig .
[ fig : rmline]a ) .
thus , with a large set of fitted transitions , one or more transitions can be removed from a given system and @xmath4 will remain well - constrained . in the absence of systematic errors associated with the transition removed or with the wavelength scale of the quasar spectra , @xmath4 should not change systematically for all systems .
for example , removing a single transition ( e.g. mgii @xmath172796 ) from our fits to the quasar data directly probes possible systematic errors due to line blending , isotopic / hyperfine structure effects and errors in the laboratory wavelengths .
removing entire species ( e.g. all transitions of siii ) also tests for isotopic / hyperfine structure effects .
given a set of fitted transitions , we only remove transitions that _ can _ be removed . to clarify this , consider removing mgii @xmath172796 ( fig . [
fig : rmline]b ) . for the low-@xmath11 mg / feii systems , a well - constrained value of @xmath4 can only be obtained when at least one mg line is fitted since the difference between the @xmath12 coefficients for the various feii transitions is small ( see fig .
[ fig : rmline]a ) . if the only available anchor line is mgii @xmath172796 then we can not remove it from the system .
similarly , we only remove the entire mgii species if the mgi @xmath172852 line is fitted . in fig .
[ fig : rmline]c we present line removal results for all transitions and species .
the left panel shows the overall effect line removal has on the @xmath22 systems where line removal is possible ( cf .
upper panel of fig .
[ fig : rmline]b ) .
the right panel shows the effect on the entire sample ( cf .
lower panel of fig .
[ fig : rmline]b ) .
we see no evidence for systematic errors associated with particular transitions or species .
we have extended our previous analysis of systematic errors @xcite to include the new , significantly augmented sample of 128 absorption systems outlined in @xcite and @xcite .
we can not explain the observed value of @xmath23 with known systematic errors .
here we have outlined our analysis of the two most important of these : atmospheric dispersion effects and possible isotopic ratio evolution .
the former is an instrumental effect that should have affected 77 of the 128 absorption systems .
however , the data themselves suggest that atmospheric dispersion had a negligible effect on @xmath4 . furthermore
, modelling indicates that the resulting low - order distortion of the wavelength scale should have shifted the low- and high-@xmath11 values of @xmath4 in opposite senses , clearly at odds with our observations .
an upper limit on the effect on @xmath4 of strong isotopic ratio evolution was obtained by assuming zero isotopic abundances for @xmath20 mg and @xmath21si in the quasar absorption clouds .
although the isotopic structures of the other transitions used in the mm method are not known , strong isotopic ratio evolution for these elements is strongly constrained by the line removal test ( fig .
[ fig : rmline]c ) .
all 128 absorption systems were observed with the keck / hires .
although we have ruled out many instrumental systematic errors , independent measurements of @xmath4 must be carried out with a different telescope and spectrograph .
using the mm method to analyse vlt and subaru spectra will be a crucial check on possible systematic errors . | we have previously presented 5.7@xmath0 evidence for a smaller @xmath1 at redshifts @xmath2 from a sample of 128 keck / hires quasar absorption systems : @xmath3 .
a non - zero @xmath4 manifests itself as a distinct pattern of shifts in the measured absorption line wavelengths .
the statistical error is now small : we do detect small line shifts in the hires data .
whether these shifts are due to systematic errors or due to real variation in @xmath1 is now the central question .
here we summarize the two potentially most important systematic effects : atmospheric dispersion and isotopic abundance evolution .
previously , these have been difficult to quantify / model but here we find that neither of them can explain our results .
furthermore , the hires spectra themselves contain no evidence for these effects .
independent measurements of @xmath4 with a different telescope and spectrograph are now crucial if we are to rule out or confirm the present evidence for a variable @xmath1 . |
jones ( 1976 ) studied the propagation of light in a moving dielectric and showed by experiment that a rotating medium induces a rotation of the polarisation of the transmitted light . player ( 1976 ) confirmed that this observation could be accounted for through an application of maxwells equations in a moving medium . more recently
padgett _ et al . _
( 2006 ) reasoned that the rotation of the medium turns a transmitted image by the same angle as the polarisation .
this is in contrast to the faraday effect ( faraday 1846 ) , where a static magnetic field in a dielectric medium , parallel to the propagation of light , causes a rotation of the polarisation but not a rotation of a transmitted image .
rotation of the plane of polarisation and image rotation in a rotating medium may be attributed respectively to the spin and orbital angular momentum of light ( allen _ et al .
_ 1999 , 2003 ) .
the first theoretical treatment of this problem was published by fermi ( 1923 ) , who considered plane waves and a non - dispersive medium .
the theoretical analysis of player ( 1976 ) was also restricted to the propagation of plane waves , but took the dispersion of the medium into account .
player assumed that the dielectric response does not depend on the motion of the medium . in our treatment
we follow his assumption although a more careful analysis by nienhuis _ et al . _
( 1992 ) showed that there will be an effect of the motion on the refractive index for a dispersive medium near to an absorption resonance ( see also baranova & zeldovich ( 1979 ) for a discussion on the effect of the coriolis force on the refractive index ) .
in contrast to player we allow for more general electromagnetic fields that can carry orbital angular momentum ( oam ) .
this leads to an additional term in our wave equation , which corresponds to a fresnel drag term familiar from analysis of uniform motion . for a rotating medium , however
, this drag leads to a rotational shift of the image .
the propagation of light in a rotating medium thus involves both spin angular momentum ( sam ) and oam .
we solve the wave equation for circularly polarised bessel beams and consider two different superpositions of such bessel beams to quantify the effects of both polarisation and image rotation .
for rotation of the polarisation we examine a superposition of left- and right - circularly polarised bessel beams carrying the same amount of oam . for image rotation
we consider a superposition of bessel beams with the same circular polarisation but opposite oam values .
such a superposition creates an intensity pattern with lobes or ` petals ' . in both cases the constituent bessel beams
propagate differently in the medium , which leads to a change in their relative phase .
this is the origin of the rotation of both the polarisation and the transmitted image . for both phenomena
we derive an expression for the angle per unit length of dielectric through which the image or the polarisation is rotated .
the significance of the total angular momentum can be most easily seen in the wave equation for the propagation of light in a rotating medium .
we derive this wave equation in section [ sec : waveeq ] . in the remaining sections we calculate the rotation of polarisation ( section [ sec : polrot ] ) and the image rotation ( section [ sec : imgrot ] ) and reveal their common form .
the wave equation for a general electric displacment @xmath0 in a rigid dielectric medium rotating with angular velocity @xmath1 is given by : @xmath2 \left [ \mathbf{\omega } \times \dot{\mathbf{d } } - ( \mathbf{v } \cdot \nabla ) \dot{\mathbf{d } } \right].\ ] ] an analoguous wave equation can be derived for the magnetic induction @xmath3 .
compared to the form derived by player ( 1976 ) , who considered the special case of a plane wave propagating along the direction of @xmath1 , these wave equations contain an additional term @xmath4 ( \mathbf{v } \cdot \nabla ) \dot{\mathbf{d } } $ ] .
this term is responsible for the fresnel drag effect which modifies the speed of light in a moving medium ( mccrea 1954 ; barton 1999 ; rindler 2001 ) . in the following
we will derive this wave equation for the electric displacement .
our analysis starts with the same considerations as player ( 1976 ) , by introducing a rest frame and a moving frame . in the rest frame
the dielectric medium rotates with an angular velocity @xmath5 and in the moving frame the medium is at rest .
we restrict our analysis to small velocities with @xmath6 and use maxwell s equations in both reference frames ( landau & lifshitz 1975 ) . for the medium at rest
we assume the following constitutive relations : [ eq : constitutive ] @xmath7 where we have used primes to denote the fields and their frequency @xmath8 in the moving frame .
the fields in the moving frame can be expressed in the rest frame by a lorentz transformation ( stratton 1941 ; jackson 1998 ) , which gives to first order in @xmath9 : [ eq : transforms ] @xmath10 where we have set @xmath11 and work with units in which @xmath12 .
the two constitutive relation in ( [ eq : constitutive ] ) in the rest frame are thus given by @xmath13 the dielectric constant is still given as a function of the frequency in the moving frame .
we also assume that the dielectric constant depends only on the frequency and is otherwise independent of the state of motion of the medium .
on combining these two equations we can express @xmath0 and @xmath3 with the two other fields @xmath14 and @xmath15 to the first order in @xmath16 : [ eq : dbfirstorder ] @xmath17 \mathbf{v } \times \mathbf{h } , \label{eq : dfirstorder}\\ \mathbf{b } & = \mathbf{h } - [ \epsilon(\omega ' ) - 1 ] \mathbf{v } \times \mathbf{e}. \label{eq : bfirstorder}\end{aligned}\ ] ] after taking the curl of ( [ eq : dfirstorder ] ) we can use the maxwell equation @xmath18 and express @xmath19 , with the help of ( [ eq : bfirstorder ] ) , in terms of @xmath20 and @xmath21 .
if we assume @xmath22 to be constant ( see [ app : acceleration ] ) , as in player s paper ( player , 1976 ) this yields @xmath23 \mathbf{v } \times \dot{\mathbf{e } } + [ \epsilon(\omega ' ) -1 ] \nabla \times ( \mathbf{v } \times \mathbf{h}).\ ] ] it follows from ( [ eq : dfirstorder ] ) that @xmath24 , to the first order in @xmath22 , and so we can rewrite ( [ eq : vsteady ] ) as : @xmath25 \mathbf{v } \times \dot{\mathbf{d } } + [ \epsilon(\omega ' ) -1 ] \nabla \times ( \mathbf{v } \times \mathbf{h}).\ ] ] we can now take the curl of ( [ eq : curld ] ) to obtain a wave equation for @xmath0 , as @xmath26 for @xmath27 , and the curl of @xmath20 is given by @xmath28 . in order to express the curl of the vector products we use the identity @xmath29 , where the doubly occurring index denotes a summation over the cartesian components .
the operator @xmath30 represents differentiation with respect to the @xmath31th component and acts on the whole product which gives rise to terms containing the divergences of @xmath32 and @xmath15 .
these terms are either zero , because @xmath33 and @xmath27 or they lead to terms which are of second order in @xmath22 and therefore negligible . the wave equation for @xmath0
is thus given by @xmath34 \left [ ( \dot{\mathbf{d } } \cdot \nabla ) \mathbf{v } - ( \mathbf{v } \cdot \nabla ) \dot{\mathbf{d } } \right ] \\ & + [ \epsilon(\omega ' ) - 1 ] \nabla \times \left [ ( \mathbf{h } \cdot \nabla ) \mathbf{v } - ( \mathbf{v } \cdot \nabla ) \mathbf{h } \right ] .
\end{split}\ ] ] for a rotation @xmath5 we can specify terms of the form @xmath35 by expressing the components of the velocity @xmath22 using the levi - civitta symbol @xmath36 as @xmath37 .
the components of @xmath38 are thus given by @xmath39_i = a_l \partial_l \varepsilon_{ijk } \omega_j r_k = a_l \varepsilon_{ijk } \omega_j \delta_{lk } = \left [ \mathbf{\omega } \times \mathbf{a } \right]_i.\]]if we use the results from ( [ eq : directderiv ] ) in ( [ eq : waveeq ] ) we find for @xmath40 : @xmath41 \left [ \mathbf{\omega } \times \dot{\mathbf{d } } - ( \mathbf{v } \cdot \nabla ) \dot{\mathbf{d } } \right ] \\ & + [ \epsilon(\omega ' ) - 1 ] \nabla \times \left [ \mathbf{\omega } \times \mathbf{h } - ( \mathbf{v } \cdot \nabla ) \mathbf{h } \right ] .
\end{split}\ ] ] the curl of the last bracket requires some some additional calculations . the first term is given by : @xmath42 and the second term can be written as : @xmath43 where the last term originates from @xmath44 .
the terms containing the divergence of @xmath15 cancel and the term @xmath45 can be added to the second term in ( [ eq : waveeqrot ] ) .
the two remaining terms @xmath46 and @xmath47 together give @xmath48 : @xmath49 this concludes the derivation of the wave equation ( [ eq : finalwaveeq ] ) .
it is possible to derive the same wave equation for @xmath3 using similar methods . for a rotation around the @xmath50 axis with constant angular velocity @xmath51 ,
the directional derivative @xmath52 is proportional to an azimuthal derivative , as @xmath53 .
this allows us to identify the two terms @xmath48 and @xmath54 in the wave equation @xmath55 \left [ \mathbf{\omega } \times \dot{\mathbf{d } } - \omega \partial_\phi \dot{\mathbf{d } } \right]\ ] ] as the polarisation rotation and rotary fresnel drag terms , respectively .
player s derivation does not contain the term proportional to @xmath56 because he treated only the case of a plane wave propagating in the @xmath50-direction and for such fields @xmath0 is independent of @xmath57 .
on substituting a monochromatic ansatz of the form @xmath58 into ( [ eq : phiwaveeq ] ) , where @xmath59 is the optical angular frequency in the rest frame , we obtain : @xmath60 \omega \omega \left [ { \mathrm{i}}\mathbf{e}_z \times \mathbf{d}_0 - { \mathrm{i}}\partial_\phi \mathbf{d}_0 \right].\ ] ] if we make an ansatz for @xmath61 with a general polarisation given by the complex numbers @xmath62 and @xmath63 ( with @xmath64 ) in the form of @xmath65 , we find that the @xmath66 and @xmath67 components of the wave equation ( [ eq : monowaveeq ] ) decouple if @xmath68 corresponding to left- and right - circularly polarised light respectively . if we restrict the solutions to these two cases we can write the wave equation as : @xmath69 \omega \omega \left ( \pm 1 - { \mathrm{i}}\partial_\phi \right ) \mathcal{d},\ ] ] where the plus sign refers to left - circular polarisation and the minus sign to right - circular polarisation .
we can then identify @xmath70 as the extreme values of the variable @xmath71 which corresponds to the circular polarisation or sam of the light beam .
similarly we can identify @xmath72 as the oam operator , so that the wave equation contains a term which depends on the total angular momentum @xmath73 : @xmath74 \omega \omega \left ( \sigma + l_z \right ) \mathcal{d}.\ ] ] we shall see that it is the dependence on the optical angular momentum that is responsible for the rotation of both the polarisation and of a transmitted image .
the rotation of the polarisation arises from the difference in the refractive indices for left- and right - circularly polarised light .
the angle per unit length by which the polarisation is rotated is called the specific rotary power . for an optically active medium at rest the specific rotary power is characteristic for a given material , but from ( [ eq : transversewaveeq ] ) it can be seen that light propagates differently in a rotating medium , depending on whether the circular polarisation turns in the same rotation sense as the dielectric or in the opposite sense .
this phenomenon is described by the effective specific rotary power ( jones 1976 ; player 1976 ) . the specific rotary power , defined as ( fowles 1975 ) : @xmath75 is the angle of rotation of the plane of polarisation in an optical active medium . here , the indices @xmath76 and @xmath77 refer to right- and left - circularly polarised light .
it was convenient to set @xmath11 for our derivation in section [ sec : waveeq ] but we reintroduce it here to facilitate the calculation of measurable quantities . in order to illustrate the effect of the oam of light we choose a bessel beam as an ansatz for the electrical displacement in the @xmath78 plane : @xmath79 where @xmath80 and @xmath81 are the transverse and longitudinal components of the wavevector .
bessel beams of this form carry oam of @xmath82 per photon ( allen _ et al .
_ 1992 , 1999 , 2003 ) . substituting the bessel beam ansatz in the wave equation ( [ eq : transversewaveeq ] )
yields the following result for the overall wavenumber @xmath83 : @xmath84 \frac{\omega \omega}{c^2 } ( \sigma + m).\ ] ] the indices @xmath77 and @xmath76 denoting the circular polarisation correspond respectively to @xmath85 and @xmath86 . with the help of the relations @xmath87 and
@xmath88 we can turn the equation for the wavenumbers into an equation for the effective refractive indices for left- and right - circularly polarised light : @xmath89 \frac{\omega}{\omega } ( \sigma + m).\ ] ] following player ( 1976 ) we assume that @xmath90 and we can therefore approximate the square root for the refractive indices @xmath91 by a small parameter expansion to the first order in @xmath92 : @xmath93 \frac{\omega}{\omega } \left ( \sigma + m \right).\ ] ] the frequency in the moving frame @xmath8 is different for left- and right - circularly polarised light ( garetz 1981 ) and , more generally , the azimuthal or rotational doppler shift is proportional to the total angular momentum @xmath94 ( allen _ et al .
_ 1994 ; bialynicki - birula & bialynicka - birula 1997 ; courtial _ et al .
_ 1998 ; allen _ et al .
_ 2003 ) . for left - circularly polarised light with @xmath95
the frequency is thus @xmath96 , and for right - circularly polarised light with @xmath97 the frequency changes to @xmath98 .
following player ( 1976 ) we expand the refractive index of the dielectric in a taylor series to calculate the difference @xmath99 : [ eq : refindices ] @xmath100 \frac{\omega}{\omega } \left(1 + m \right ) , \\
n_r(\omega ) & \simeq
n(\omega ) - \frac{d n}{d \omega } \omega ( -1+m ) - \left [ n(\omega ) - \frac{1}{n(\omega ) } \right ] \frac{\omega}{\omega } \left(-1 + m \right).\end{aligned}\ ] ] higher order derivatives of @xmath101 become comparable in magnitude if @xmath102 .
this will only be case for a strongly dispersive medium , such as atomic or molecular gases , near a resonance .
for such gaseous media the dielectric response in a rotating medium has to examined more closely ( nienhuis _ et al .
_ 1992 ) . for solid materials , such as a rotating glass rod , and for optical frequencies
this condition is not fulfilled and we can neglect higher order derivatives in the expansion ( [ eq : refindices ] ) . within player s assumption that the refractive index is independent of the motion of the medium we find for the effective specific rotary power : @xmath103 on introducing the group refractive index @xmath104 and the phase refractive index @xmath105 , we can rewrite the rotary power as @xmath106 which is identical to player s ( 1976 ) expression . in this form the specific rotary power ( [ eq : playerrotpow ] ) can be used directly with experimental data in the si unit system . in the next section
we look at image rotation caused by a difference in the effective refractive indices for different values of @xmath107 .
the specific rotary power describes the rotation of the propagation , but we can define , analogously , a rotary power of image rotation .
the image can simply be created by the superposition of two light beams carrying different values of oam which leads to an azimuthal variation of the intensity pattern .
in particular we consider an incident superposition of two similarly circularly polarised bessel beams with opposite oam values of the form @xmath108 outside the medium the superposition can be written as one bessel beam with a trigonometric modulation @xmath109 but inside the medium the effective refractive index is different for the two components of the superposition ( allen & padgett 2007 ) .
on propagation this leads to phase difference which causes a rotation of the image ( see figure [ fig : petalimage ] ) .
we define @xmath110 which is the angle per unit length by which the image is rotated .
the factor @xmath107 in the expression for @xmath111 appears because of the @xmath112 and @xmath113 phase structure of the interfering beams and the resulting @xmath114-fold symmetry of the created image ( pagdett _ et al .
_ 2006 ) .
image rotation , title="fig:",scaledwidth=49.0% ] image rotation , title="fig:",scaledwidth=49.0% ] the different effective refractive indices for the components of the superposition ( [ eq : superposition ] ) are given by : @xmath115 \frac{\omega}{\omega } ( \sigma \pm m).\ ] ] here , @xmath71 is fixed in contrast to ( [ eq : refindex ] ) .
the roles of @xmath71 and @xmath107 are reversed for the image rotation and the refractive indices for positive and negative oam are given by : [ eq : rotdiffrefind ] @xmath116 \frac{\omega}{\omega } \left(\sigma + m \right ) , \\
n_-(\omega ) & \simeq
n(\omega ) - \frac{d n}{d \omega } \omega ( \sigma - m ) - \left [ n(\omega ) - \frac{1}{n(\omega ) } \right ] \frac{\omega}{\omega } \left(\sigma - m \right).\end{aligned}\ ] ] on substituting ( [ eq : rotdiffrefind ] ) into ( [ eq : imgrot ] ) we find : @xmath117 which can be written in terms of the group and phase refractive indices as : @xmath118 this verifies the reasoning of padgett _ et al . _
( 2006 ) that the polarisation and the image are turned by the same amount when passing through a rotating medium .
it is the total angular momentum that determines the phase shifts and a linearly polarised image will undergo rotations of both the plane of polarisation and the intensity pattern or image .
we have extended a theoretical study by player ( 1976 ) on the propagation of light through a rotating medium to include general electromagnetic fields . in the original analysis player ( 1976 )
showed that the rotation of the polarisation inside a rotating medium can be understood in terms of a difference in the propagation for left- and right - circularly polarised light .
player s ( 1976 ) analysis was thus concerned solely with the spin angular momentum ( sam ) of light .
our treatment has shown that the general wave equation has an additional term , which is of the same form as the fresnel drag term for a uniform motion . in the context of rotating motion , however , this term is connected to the orbital angular momentum ( oam ) of the light . by extending the theoretical analysis to include oam we have been able to attribute polarisation rotation and image rotation to sam and oam respectively
we have shown that a superposition of bessel beams with the same oam but opposite sam states leads to the rotation of the polarisation , whereas a superposition of bessel beams with the same sam and opposite oam values gives rise to a rotation of the transmitted image .
we have obtained quantitative expressions for the rotation of the polarisation and of the transmitted image and have verified that both are turned through the same angle , as recently suggested by padgett _
( 2006 ) . player ( 1976 ) remarked that the derivation by fermi ( 1923 ) appears to be in error .
the mistake in fermi s treatment seems to be in missing the transformation of the magnetic fields .
whereas the change in the electric fields induced by the motion of the medium is explicitly given in terms of the electric polarisation @xmath119 , a similar transformation for the magnetic field is missing . in terms of our derivation this would mean that ( [ eq : bfirstorder ] ) changes to @xmath120 in the rest frame .
this in turn causes that the term @xmath121 would be missing in ( [ eq : vsteady ] ) .
this term and the term @xmath122 contribute equally to the wave equation ( [ eq : finalwaveeq ] ) , which explains why fermi s result for the specific rotary power is smaller than player s and ours by a factor of two . as pointed out by player ( 1976 ) this missing factor
is cancelled by an additional factor of two in fermi s definition of the specific rotary power .
the assumption that @xmath5 is steady is problematic for a rotating motion ; if we assume @xmath123 to be constant over time , then @xmath124 . in principle
this would invalidate our initial considerations for the transformation of the electromagnetic fields ( [ eq : transforms ] ) which strictly hold only for uniform motion . including the time - derivative of @xmath22 would lead to additional terms in ( [ eq : vsteady ] ) of the form @xmath125 \dot{\mathbf{v } } \times \mathbf{e}$ ] .
if we proceed in taking the curl of this vector product we produce four terms which either can be neglected because they are second order in @xmath9 , or they do not contain the time derivative of an optical field .
the latter are smaller than terms that do contain a time derivative by @xmath126 . for our assumption
@xmath90 all such terms are negligible .
fermi , e. 1923 sul trascinamento del piano di polarizzazione da parte di un mezzo rotante .
lincei _ * 32 * , 115118 . reprinted in : fermi , e. 1962 _ collected papers _ , vol .
chicago : university of chicago press .
padgett m. , whyte g. , girkin j. , wright a. , allen l. , hberg p. & barnett s. m. 2006 polarization and image rotation induced by a rotating dielectric rod : an optical angular momentum interpretation .
_ optics lett .
_ * 31 * ( 14 ) , 22052207 . | image rotation , polarisation , rotating dielectric , specific rotary power when light is passing through a rotating medium the optical polarisation is rotated .
recently it has been reasoned that this rotation applies also to the transmitted image ( padgett _ et al .
_ 2006 ) .
we examine these two phenomena by extending an analysis of player ( 1976 ) to general electromagnetic fields .
we find that in this more general case the wave equation inside the rotating medium has to be amended by a term which is connected to the orbital angular momentum of the light .
we show that optical spin and orbital angular momentum account respectively for the rotation of the polarisation and the rotation of the transmitted image .
[ firstpage ] |
tev @xmath0-ray binaries are a subclass of binaries with a compact object , established in the 2000 s ( e.g. , * ? ? ?
* for a recent review ) .
these systems have a spectral energy distribution with a peak beyond 1 mev and are variable at multi - wavelengths up to tev energies .
there are 5 known binaries of this kind , all of which have either an o - type main - sequence star or a be star with a circumstellar disk as the optical counterpart .
the nature of the compact object is unknown in all systems but one ( ) . for such systems two competing scenarios have been proposed , based on the different nature of the compact object , and hence the different mechanism of high energy emissions .
one scenario assumes that the collision between a relativistic pulsar wind and a stellar wind and/or circumstellar disk produces strong shocks , where the high energy emission arises ( e.g. , * ? ? ?
this pulsar wind scenario has successfully been applied to . on the other hand ,
the other scenario assumes the presence of an accreting black hole ( or neutron star ) . in this microquasar or accretion / ejection scenario ,
a large amount of mass transferred from a companion star powers relativistic jets , where @xmath0-ray emission originates ( e.g. , * ? ? ? * ) .
( @xmath6 ) is a recently established tev @xmath0-ray binary @xcite comprised of a b0ve star and a compact object of unknown nature .
the orbit is wide and highly eccentric ( @xmath7 ; * ? ? ?
the system shows correlated variability in the x - ray and tev energy bands @xcite , with two peaks in one orbital cycle : the primary outburst prior to apastron ( orbital phase @xmath8 = 0.30.4 ) and the secondary outburst after it ( @xmath9 0.60.9 ) , with an x - ray dip in between .
this is puzzling , given that these outbursts and the dip occur when the compact object is orbiting very far from the be star .
understanding the mechanism for these phenomena is one of the key issues of this system .
the optical counterpart of , , is established to be a classical be star , i.e. , a massive star with a geometrically thin , circumstellar disk , where balmer and other emission lines originate @xcite .
@xcite observed the h@xmath2 emission line after apastron ( @xmath9 0.70.8 ) .
their profiles over continuous 35 days showed an `` s - shaped '' variability with a period of @xmath4 60 days .
@xcite reported the orbital modulation in the h@xmath2 line profile parameters , likely caused by the interaction between the be disk and the compact object .
however , the lack of observations in @xmath8=0.60.8 makes it difficult to constrain the modulation after apastron . in order to constrain the interaction between the
be disk and the compact object , we have been monitoring using various methods such as high - dispersion spectroscopy , photometry , and polarimetry . in this paper
we report the initial result of the line profile variabilities over 160 days from 2013 october to 2014 april , covering the second half of the orbital period . in sect .
2 , we summarize our observations . the results are described in sect .
3 . in sect .
4 we discuss the be disk region affected by the compact object and the nature of the interaction .
optical high - dispersion spectroscopic observations of were carried out on 12 nights from 2013 october 31 ( phase 0.555 ) to 2014 march 11 ( phase 0.969 ) at the okayama astrophysical observatory with a 188 cm telescope and hides with fiber - feed system @xcite . here ,
we calculated the orbital phase , @xmath8 , taking the orbital period of 315 days from @xcite and the origin ( jd 2454857.5 ) from @xcite , and setting the periastron phase to be 0 ( in @xcite , it is set to 0.967 ) .
note that , although @xcite determined the ephemeris using the orbital period of 321 days , the orbital parameters remain approximately unchanged @xcite .
the observed spectra covered 42007400 wavelength range , with s / n of @xmath10 around h@xmath2 .
the typical wavelength resolution @xmath11 is @xmath12 .
the data were reduced in the standard way , using the iraf echelle package .
high - dispersion spectra were also taken on two nights ( 2014 february 05 , phase 0.861 , and april 10 , phase 0.064 ) using the canada france hawaii telescope / espadons @xcite in spectropolarimetric mode .
the spectra covered a wavelength range of 370010500 with a resolving power of @xmath13 .
we obtained reduced data using the libre - esprit / upena pipeline , provided by the instrument team .
we rectified the normalized intensity by re - determining the continuum level around each line , in order to compare with hides data . in this paper , we focus on the spectroscopic variability . the polarimetric data will be discussed in a forthcoming paper . in addition to spectroscopic data , @xmath14- , @xmath15- , and @xmath16-band photometric data were obtained from 2013 september to 2014 may using howpol @xcite and honir @xcite attached to kanata 1.5 m telescope at higashi - hiroshima observatory in hiroshima , japan .
we observed on 73 , 71 and 10 nights in @xmath14 , @xmath15 , and @xmath16 bands , respectively , using howpol .
we used honir to observe the source on 8 nights , in these three filters .
the logs of the photometric and spectroscopic observations are given in table [ tbl : log ] .
figure [ fig : profave ] shows the averaged h@xmath2 , h@xmath3 , and h@xmath0 profiles . in these profiles ,
the vertical scale of the h@xmath2 line profile is half that of the other line profiles , because the flux is much stronger in the h@xmath2 .
the h@xmath2 line is very strong ( @xmath17 ) and significantly asymmetric .
the profile exhibits a double peak at @xmath18 and @xmath19 , with the latter being brighter than the former , and a hump at @xmath20 on the blue wing . on the other hand ,
the averaged h@xmath3 line profile ( @xmath21 ) is rather symmetric , with a slightly stronger blue peak .
the h@xmath0 line , which is on a broad absorption component , also exhibits a double - peaked profile with stronger blue peak .
balmer lines exhibited complicated line profile variabilities during the monitoring period .
figure [ fig : prof_ind ] displays all observed profiles of the h@xmath2 , h@xmath3 , and h@xmath0 lines , while fig .
[ fig : prof_diff ] presents the time sequence of the residual spectra of these lines from the average .
the variation in the equivalent width , ew , the full width at half maximum , fwhm , and the centroid velocity , @xmath22 , of the h@xmath2 line profile are shown in fig . [
fig : prof_param_ha ] . note that fwhm and @xmath22 are measured by fitting the whole profile with a single gaussian and the line wings with a voigt profile , respectively . in the long term ( @xmath23 days , from @xmath24 ) ,
the h@xmath2 line changed from a red - enhanced profile to a rather symmetric profile , while the stronger peak of the h@xmath3 and h@xmath0 lines changed from the red side to the blue side . in the residual spectra of these lines ,
s - shaped variability is visible ; the bright part at @xmath25 migrates from the blue side to the red side at @xmath26 , and returns to the blue side at @xmath27 .
simultaneously , the faint part at @xmath28 migrates in the opposite direction . at the same time , the maximum and minimum , respectively at @xmath29 and @xmath30 in the residual spectra , migrate in a similar fashion .
the period of these variations is estimated at @xmath31 days using fourier analysis , which is about twice the @xmath32-day variation period reported by @xcite .
we will discuss this in more detail in the next section . on the other hand , in the short term ( @xmath33 days )
, a bright hump appears at @xmath29 in balmer line profiles at @xmath34 .
this hump disappears by @xmath35 ( see the third , forth , and fifth profiles from the bottom in figs . [
fig : prof_ind ] and [ fig : prof_diff ] ) .
when the hump appears , ew(h@xmath2 ) increases by @xmath36 and fwhm(h@xmath2 ) decreases by @xmath37 .
these long- and short - term variabilities respectively seem to occur in phase in different lines , although time lags between lines can not be ruled out because of the low cadence of the observations .
recently , @xcite reported the orbital modulation of the profile parameters of the h@xmath2 line .
their data cover the phase intervals of 00.6 and 0.81 , using the orbital period of 321 days @xcite .
if we take the same orbital period , our data spans the orbital phase 0.41 , which overlaps with @xcite in the phase range 0.40.6 and 0.81 and provides for the first time the modulation data in the phase interval of 0.60.8 . in fig .
[ fig : prof_param_ha ] , ew(h@xmath2 ) seems to increase ( phase 0.40.5 ) and then decrease back to the previous value in about 30 days ( phase 0.50.6 ) , as reported by @xcite .
afterwards , it seems to fluctuate for about 60 days ( phase 0.60.8 ) , which could be associated with the variabilities mentioned above .
@xmath22(h@xmath2 ) , whose variation is thought to reflect the orbital motion of the be star , shows a similar pattern to @xcite except just after the apastron and just before the periastron .
this difference in @xmath22 is possibly caused by the profile variabilities .
fwhm(h@xmath2 ) , on the other hand , shows a different variation pattern from @xcite .
it stays constant except around apastron ( phase 0.450.55 for the period of 321 days ) in this work , whereas in @xcite it showed a sinusoidal pattern .
although many lines are contaminated by neighboring lines , there are six less - contaminated lines ( @xmath385018 , 5316 , 5363 , 5535 , 6433 , and 6456
) that enable us to analyze the line profile variabilities . in our analysis , we have used these six lines . the averaged @xmath395363 line profile in fig .
[ fig : profave ] is shown as a representative profile .
in contrast to the balmer lines , the emission lines show symmetric double - peaked profiles , as predicted for a rotating , axi - symmetric disk .
figures [ fig : prof_ind ] and [ fig : prof_diff ] also display the time sequence of the observed and residual @xmath395363 profiles , respectively . as seen in these figures , lines exhibited no variation .
@xcite analyzed the spectra of several be stars and confirmed that emission lines arise from a disk region of radius of @xmath40 .
they also found empirical relationships between the projected rotational velocity of the central star ( @xmath41 ) and the profile parameters .
applying their relationships , we have derived @xmath41 to be @xmath42 . finally , we estimate the radii of emitting regions of balmer lines . for the h@xmath2 line , we use the mean peak separation of the last six profiles ( @xmath43 ) , because the averaged profile in fig .
[ fig : profave ] , in which all our observed epochs were used , has too complicated features to determine the peak velocities . for the h@xmath3 and h@xmath0 lines , we use the averaged profiles . as a result
, we have the peak separations of the h@xmath2 , h@xmath3 , and h@xmath0 lines to be @xmath44 , @xmath45 and @xmath46 , respectively . adopting @xmath41
derived above , we obtain the radii of the emitting region of these lines as @xmath47 @xmath5 ( h@xmath2 ) , 11 @xmath5 ( h@xmath3 ) , and 7 @xmath5 ( h@xmath0 ) , where @xmath48 is the radius of the be star . throughout our monitoring period ( from apastron to periastron ) , optical brightness of stayed constant in the range of 9.059.32 ( @xmath14 ) mag , 8.538.71 ( @xmath49 ) mag , and 8.588.67 ( @xmath50 ) mag , within the typical 1-@xmath51 error of 0.05 mag .
the average values are @xmath52 mag , @xmath53 mag and @xmath54 mag .
although the monitoring was performed mostly after apastron , the line profiles exhibited remarkable variabilities .
wide wavelength coverage revealed that the variations are seen not only in the h@xmath2 line but also in the h@xmath3 and h@xmath0 lines .
no variation in lines indicates the inner part of the be disk was kept undisturbed during the observing period .
this fact is in agreement with little variability in the optical brightness , which is thought to originate from the disk region within 23 @xmath5 @xcite .
the slight , but significant , variation in the h@xmath0 line profile implies that the interaction with the compact object affects the be disk down to , at least , a radius of 7 @xmath5 .
the dynamical residual spectra of the balmer lines showed an s - shaped variation ( fig .
[ fig : prof_diff ] ) , as @xcite reported .
high - dispersion residual spectra clearly show that there are two pairs of peaks at @xmath55 and @xmath56 .
the lower velocity peaks are the same as those of @xcite ( see their fig .
the period of the variation is @xmath4150 days , which is about twice the @xmath4 60-day variation period @xcite .
the s - shaped variation seen in many be stars is thought to be caused by global disk oscillations .
they are mostly low - frequency @xmath57 oscillations , where @xmath58 is the azimuthal wave number @xcite , but in eccentric binaries , it is also possible that @xmath59 oscillations are excited by the corresponding fourier components of the tidal potential ( e.g. , * ? ? ?
if the period of observed variation is @xmath4 150 days , about half the orbital period , it might be due to an @xmath60 oscillation mode in the be disk , which could be excited by the @xmath60 fourier component of the tidal potential of the compact object .
the difference of the periods between this work and @xcite might be caused by the difference of the length and frequency of the observations . because we observed for about half an orbital cycle ( @xmath61 days ) ,
variabilities of this timescale can be detected . with low cadence of observation , typically once every 1020 days , however , profile variabilities of shorter timescales are invisible . on the contrary , the short monitoring epoch of @xcite ( 35 days ) would have hidden the variability with longer timescales ( @xmath62 days ) .
moreover , observations of this work and @xcite have an interval of @xmath41800 days . therefore , there are possibilities that the oscillation mode has changed or that both oscillation modes exist .
in addition to the s - shaped variation , a remarkable variability was seen in the balmer lines after apastron ( @xmath8=0.600.65 ) .
the short lifetime of this variability ( @xmath63 days ) indicates that it is temporarily caused by an external force such as the tidal force and the ram pressure of the pulsar wind , because internal waves / oscillations have much longer lifetimes ( @xmath64 days ; * ? ? ? * and references therein ) .
it is surprising , however , that such a remarkable variability appeared after apastron , when the interaction is expected to be weak because of the large distance between the be disk and the compact object . in order to understand this phenomenon , it is essential to constrain the starting phase and repeatability of the variability .
comparison of the ew(h@xmath2 ) between this work and @xcite implies the presence of a regular orbital modulation , but different absolute values due to different spectral resolutions make it difficult to quantitatively analyze it .
given the orbital period close to one year , it is necessary to monitor over two or three more successive cycles , within the timescale of be disk variability ( @xmath65 days ) , in order to cover the full orbital phase with the same spectral resolution .
because the compact object is located far away from the be star around apastron ( @xmath66 ) , the gas in the be disk at the radius of 7 @xmath5 can not be significantly affected by the tidal force of the compact object , whose strength is less than @xmath67 of the gravity of the be central star at this radius .
this leaves the interaction with the pulsar wind as the most likely mechanism for the balmer line variability . at the periastron passage , on the other hand ,
the compact object is thought to pass through the be disk at @xmath68 .
nonetheless , the h@xmath3 line , emitted from the disk region of the similar radius , showed no remarkable variation at @xmath690.06 ( see the top profile in the upper right panel of fig .
[ fig : prof_ind ] ) .
this fact suggests that either the tidal interaction is very weak in this system or the orbital period is longer than 315 days , so that the compact object had not passed the disk yet at the time of observation .
the former possibility could be realized if the be disk rotates in the retrograde direction or is tilted by a large angle with respect to the orbital plane , where the disk gas and the compact object interact on a very short timescale .
future observations at significantly later phases will distinguish between these two possibilities .
as described in sects .
[ sec : short ] and [ sec : imp ] , our observations leave the pulsar wind model as the sole candidate for . if the compact object is a pulsar , little line profile variation around periastron ( the top two profiles in each panel of fig .
[ fig : prof_diff ] ) suggests that the pulsar wind is too weak to significantly affect the be disk during these phases .
@xcite proposed a flip - flop scenario for another gamma - ray binary , , where the pulsar is in a rotationally powered regime in the apastron , while it is in a propeller regime in the periastron ( see also * ? ? ? * ) . in a flip - flop system ,
if the gas pressure of the be disk overcomes the pulsar - wind ram pressure , the pulsar wind is quenched . because the be disk of
is estimated to be about three times larger than the binary separation at periastron , the compact object crosses a dense region of the disk near the periastron .
in such a situation , the strong gas pressure is likely to quench the pulsar wind and hence suppress high - energy emissions . in the framework of the pulsar wind model , there are a few mechanisms that might explain the short - term episodic variability discussed above .
the be disk is likely as large as the binary orbit because of no tidal truncation in highly eccentric , large orbit @xcite . because the disk density rapidly decreases with radius ( e.g. , * ? ? ?
* ) , the wind from the pulsar close to the outer part of the disk effectively changes its structure , giving rise to remarkable variability in emission lines arising from the disk outer part .
if the be disk is misaligned with the orbital plane and a node happens to be in the direction corresponding to the phase of the variation ( @xmath70 ) , the pulsar - wind effect on the be disk will be the strongest in this phase interval .
after @xmath71 the pulsar comes close to the denser disk region , which is not easily affected by the pulsar wind .
this might be a cause of the short - term episodic variation after aspastron .
it is not clear , however , how the pulsar wind can affect the inner disk at the radius of 7 @xmath5 .
alternatively , the short - term post - apastron variations might be explained as the emission from the gas captured by the pulsar , if the pulsar wind is not strong enough to expel the surrounding gas .
this picture seems to fit well with the flip - flop model . at any rate ,
observational investigation in the first half of the orbital cycle is needed in order to further test the pulsar - wind scenario .
the next periastron passage of will take place in 2015 december , according to the ephemeris of @xcite ( 2016 january if the ephemeris is taken from @xcite ) .
observations covering this period will provide more clues to the complex interaction and the nature of the compact object in this puzzling @xmath0-ray binary .
this paper is based on the observations taken at the okayama astrophysical observatory .
we are grateful for dr .
eiji kambe to kindly observe .
this work was supported by research fellowships for the promotion of science for young scientists ( ym , kt ) .
ato acknowledges support by the jsps grant - in - aid for scientific research ( 24540235 ) and a research grant from hokkai - gakuen educational foundation .
acc acknowledges support from cnpq ( grant 307076/2012 - 1 ) .
aharonian , f. a. et al .
2007 , , 469 , l1 aliu , e. , archambault , s. , aune , t. et al .
2014 , , 780 , 168 aragona , c. , mcswain , m. v. & de becker , m. 2010 , , 724 , 306 arias , m. l. , zorec , j. , cidale , l. , ringuelet , a. e. , morrell , n. i. & ballereau , d. 2006 , , 460 , 821 artymowicz , p. , & lubow , s. h. 1994 , , 421 , 651 bongiorno , s. d. , falcone , a. d. , stroh , m. , holder , j. , skilton , j. l. , hinton , j. a. , gehrels , n. & grube , j. 2011 , , 737 , 11 carciofi , a. c. & bjorkman , j. e. 2006 , , 639 , 1081 casares , j. , rib , m. , ribas , i. , paredes , j. m. , vilardell , f. & negueruela , i. 2012 , , 421 , 1103 dubus , g. , 2013 , , 21 , 64 kambe , e. , yoshida , m. , izumiura , h. at al . 2013 , 65 , 15 kawabata , k. s. , nagae , o. , chiyonobu , s. , et al .
2008 , proc .
spie , 7014 , 70144l levinson , a. & blandford , r. 1996 , , 456 , l29 manset , n. & donati , j .- f .
2003 , proc .
spie , 4843 , 425 maraschi , l. & treves , a. 1981 , , 194 , 1 okazaki , a. t. 1991 , , 43 , 75 okazaki , a. t. & negueruela , i. 2001 , , 377 , 161 papaloizou , j. c. , savonije , g. j. & henrichs , h. f. 1992 , , 265 , l45 papitto , a. , torres , d. f. & rea , n. 2012 , , 756 , 188 rivinius , th . , carciofi , a. c. , & martayan , c. 2013 , , 21 , 69 sakimoto , k. , akitaya , h. , yamashita , t. et al .
2012 , proc .
spie , 8446 , 844673 torres , d. f. , rea , n. , esposito , p. , li , j. , chen , y. , zhang , s. 2012 , , 744 , 106 lccc + & date & jd & phase + oao / hides + & 2013.10.31 & 2456597.278 & 0.555 + & 2013.11.08 & 2456605.275 & 0.580 + & 2013.11.11 & 2456608.274 & 0.589 + & 2013.11.19 & 2456616.199 & 0.615 + & 2013.12.01 & 2456628.219 & 0.653 + & 2013.12.19 & 2456646.201 & 0.710 + & 2013.12.29 & 2456656.007 & 0.741 + & 2014.01.02 & 2456660.047 & 0.754 + & 2014.01.15 & 2456673.095 & 0.795 + & 2014.01.23 & 2456681.094 & 0.821 + & 2014.02.08 & 2456696.968 & 0.871 + & 2014.03.11 & 2456728.008 & 0.969 + cfht / espadons + & 2014.02.05 & 2456693.843 & 0.861 + & 2014.04.10 & 2456757.772 &
0.064 + + & filter & jd & nights + kanata / howpol + & @xmath14 & 24565512456795 & 73 + & @xmath15 & 24565512456795 & 71 + & @xmath16 & 24567222456795 & 10 + kanata / honir + & @xmath14 , @xmath15 , @xmath16 & 24567432456769 & 8 + | we report on monitoring observations of the tev @xmath0-ray binary , which were carried out to constrain the interaction between the be circumstellar disk and the compact object of unknown nature , and provide for the first time high - dispersion ( r@xmath1 ) optical spectra in the second half of the orbital cycle , from apastron through periastron .
the h@xmath2 , h@xmath3 , and h@xmath0 line profiles are found to exhibit remarkable short - term variability for @xmath41 month after the apastron ( phase 0.60.7 ) , whereas they show little variation near the periastron .
these emission lines show `` s - shaped '' variations with timescale of @xmath4150 days , which is about twice that reported previously .
in contrast to the balmer lines , no profile variability is seen in any emission line .
we estimate the radii of emitting regions of the h@xmath2 , h@xmath3 , h@xmath0 , and emission lines to be @xmath430 , 11 , 7 , and 2 stellar radii ( @xmath5 ) , respectively .
the amplitudes of the line profile variations in different lines indicate that the interaction with the compact object affects the be disk down to , at least , the radius of 7 @xmath5 after the apastron .
this fact , together with little profile variability near the periastron , rules out the tidal force as the major cause of disk variability . although this leaves the pulsar wind as the most likely candidate mechanism for disk variations , understanding the details of the interaction , particularly the mechanism for causing a large - scale disk disturbance after the apastron , remains an open question . |
angular power spectra ( henceforth , aps ) of the galactic synchrotron polarized radiation are raising up an increasing attention in these years .
the first study , based on the parkes 2.4 ghz survey ( duncan et al . 1997 , hereafter d97 ) , was due to tucci et al .
( 2000 ) , and other papers appeared shortly after ( baccigalupi et al .
, 2001a ; giardino et al .
the main motivation for these works was the need for an angular - scale dependent separation of the cosmic microwave background ( cmb ) signal from the polarized radio foreground ; in fact the first work on the aps of a polarized galactic foreground ( prunet et al .
1998 ) modelled thermal dust emission in view of the scenario to be met by planck hfi at 143 - 217 ghz .
cmb polarization is essential in order to remove degeneracies between important cosmological parameters ( zaldarriaga et al .
1997 ) , but the cosmological window ( the region in the frequency angular - scale plane where the cosmological signal is stronger than foregrounds ) is narrower for polarization than for anisotropy
. a careful study of galactic contamination versus angular scale is thereby necessary .
the study of synchrotron aps may also be important because of its bearing on the knowledge of galactic structure , and in particular of the transverse magnetic field in emission regions and the longitudinal field in compact foreground screens ( tucci et al .
this point is open to future studies .
the purpose of this work is to check the generality of the aps behaviour found by tucci et al .
( 2000 ) : electric and magnetic parity aps are governed ( with reasonable approximation for @xmath9 ) by power laws @xmath10 with @xmath11 in the portion of the southern galactic plane probed by the d97 parkes survey .
although such slopes are close to the values @xmath12 found for thermal dust emission ( prunet et al .
1998 ) , later work cast serious doubts on the generality of this behaviour , as far as synchrotron is concerned . for the aps of the scalar @xmath13 somewhat steeper spectra , with @xmath14 ranging from @xmath15 to @xmath16 ,
are given for the galactic plane by baccigalupi et al .
( 2001a ) and giardino et al .
the former authors also find a much higher slope , @xmath17 for @xmath18 out of the galactic plane from the 1.4 ghz , low - resolution survey of brouw and spoelstra ( 1976 : bs76 ) .
this discrepancy would be important when evaluating the synchrotron contamination at large scales ; it would thereby be rather unconfortable in view of the role of cmb polarization harmonics with @xmath19 in the separation of effects of the primordial gravitational background from those of the secondary ionization of the cosmic medium .
[ this is one of the objectives of the sport project , see e.g. cortiglioni et al .
( 2001 ) . ]
note that the situation is not quite clear for the total intensity @xmath20 either : large values @xmath21 are supported by bouchet and gispert ( 1999 ) and giardino et al .
( 2001 ) , but @xmath22 follows from analysis of the jodrell bank 5-ghz interferometric survey by giardino et al .
( 2000 ) .
here we extend our analysis of aps in the galactic plane considering the 2.7 ghz effelsberg survey ( duncan et al .
1999 : d99 ) ; further , we analyse several patches at intermediate galactic latitudes , @xmath23 , from the 1.4 ghz effelsberg survey ( uyaniker et al .
1999 : u99 ) , and finally three regions from bs76 , at 5 frequencies between 408 and 1411 mhz , covering latitudes up to near the north galactic pole . following tucci et al .
( 2001 ) and thereby extending the analysis of baccigalupi et al .
( 2001a ) , we carefully distinguish between the aps providing a fuller statistical description of the spin-2 polarization field ( i.e. , @xmath24 and @xmath25 , from which @xmath26 is usually computed ) , and the @xmath27 spectrum which takes into account only the magnitude of the polarization pseudovector .
this is necessary because tucci et al .
( 2001 ) find quite significant differences between @xmath28 and @xmath4 both for a gaussian polarization field , as predicted in the standard scenario for cmb , and for synchrotron radiation .
our results confirm that slopes around @xmath29 in the range @xmath9 are preferred on the average , in both the northern and the southern galactic plane for @xmath24 , @xmath25 and @xmath28 , although significant fluctuations around theses values are found in @xmath30 patches .
the fluctuations however become stronger in small patches out of the galactic plane , so that a preferred slope should not regarded as meaningful for the u99 survey . in fact , the partial regularities that we find for polarization aps do not support the usefulness of global ( i.e. , full sky ) @xmath6 for a satisfactory description of the spatial distribution of synchrotron : local @xmath6 based on fourier analysis are much more suitable to this purpose .
a consistent picture , however , emerges at all of the scales we investigated . as a matter of facts , from the low - resolution bs76 survey we find moderate slopes for @xmath31 , at all frequencies and even near the galactic pole . due to the limited range of @xmath32 available for the fits
, we find strong correlations between fitted parameters and , as a consequence , large error bars ; however there is no evidence for steep spectra , since the best values lie in the range @xmath33 almost everywhere and are particularly small at frequencies where faraday rotation is more important . for the @xmath4 spectrum ,
the slope fluctuations are smaller than for @xmath6 .
our best value for @xmath34 is @xmath35 at all resolutions and galactic latitudes , and all frequencies @xmath36 ghz .
the difference between @xmath37 and @xmath4 is significant , although not large in the average .
the latter quantity should thereby be confronted with @xmath38 , not with the @xmath39 spectra which are popular among cosmologists , and this is properly done in the last section of this paper .
all of the results that we provide for polarization aps should not be affected by contamination from point sources ; regions with flat @xmath40 on the other hand , are possibly dominated by point sources in total intensity , and are used in this paper to derive upper limits on their contribution to polarization aps , as discussed in section 3 .
our first high - resolution analysis of galactic - plane synchrotron polarization was performed on the 2.4 ghz parkes survey ( d97 ) ; this study , using twelve @xmath41 square patches ( tucci et al .
2000 ) , is extended here to the 2.7 ghz effelsberg survey ( 1999 : d99 ) covering the additional galactic longitude range @xmath42 ( see table [ surveyspatches ] ) and providing six more @xmath43 patches . a belt extended about 54% of the galactic plane
is thereby covered , although with two slightly different frequencies .
the fwhm resolutions are @xmath44 and @xmath45 respectively , sufficient to achieve angular scales up to @xmath46 .
the nominal rms noise in d97 is 8 mk for total power and 5.3 mk for polarization ( 5.3 and 2.9 mk in some more sensitive areas ) , while in d99 the rms noise is 9 mk .
.sky regions used for computation of aps [ cols="<,<,<,<,<",options="header " , ] the results of the fits are given in table [ bsgppole ] .
the declared uncertainties , which are 1-@xmath47 errors on individual parameters computed from @xmath48 fields , are large because of correlations in 3-parameter fits .
in particular , we find strong , positive correlations between @xmath49 and @xmath50 , as illustrated by an example in fig .
[ chicontour1408 ] , which refers to the 408-mhz low - latitude patch .
this figure shows @xmath51 iso - contours that we found in the ( @xmath49 , @xmath52 ) plane after minimization with respect to @xmath53 . in spite of the above uncertainties , we find moderate slopes for @xmath54 and @xmath55 , generally in the range @xmath56 . at low frequencies , @xmath57 mhz where faraday rotation is larger
, we occasionally find some slopes @xmath58 , and in one case ( near the galactic pole in the 610 mhz map ) a best value @xmath59 we should remark that due to the finite resolution of our sampling in ( @xmath49 , @xmath50 , @xmath53 ) space , and the existence of narrow , elongated valleys with @xmath51 near minimum , the best values reported in table [ bsgppole ] are less significant than the full extension of the confidence regions , described by the error bars in the tables .
however , none of the above considerations would change if the centres of such confidence regions were considered instead of the quoted minimum points .
thus there is sufficient evidence that polarization aps are somewhat flatter at low frequencies . the angular spectral behaviour does not exhibit any clear dependence on galactic latitude at these moderate resolutions .
it makes sense therefore to average over our three patches from the bs76 survey .
rows 1 to 5 of table [ aveslopes ] gives the average slopes @xmath60 and @xmath34 for each bs76 frequency . at low frequencies , where faraday rotation becomes more and more important
, we observe a gradual flattening of both spectra ; the deficit of steepness is more apparent for @xmath60 , in agreement with the considerations of tucci et al .
( 2001 ) . on the other hand , at 1411 ghz where faraday rotation is less important ( but not negligible at all )
, there is no evidence for any difference between @xmath60 and @xmath34 due to the large error bars . also , the result quoted for @xmath34 at 1411 mhz is quite consistent to those found with higher resolution in the galactic plane at 2.4 and 2.7 ghz ( cfr .
the last two rows in the table ) .
the regularities found for polarization aps in this work are not sufficient to establish the usefulness of global ( i.e. , full sky ) @xmath61 for a satisfactory description of the spatial distribution of synchrotron itself , and _ a fortiori _ for the separation of galactic synchrotron from cmb at higher frequencies . on the contrary ,
the local @xmath61 based on fourier analysis are much more suitable for a fuller description of angular structure .
this makes all - sky surveys of polarized synchrotron quite necessary .
quite significant fluctuations are found for parameters fitted in @xmath41 patches .
when we average over larger regions , the most stable behaviour is found for the polarized intensity spectrum @xmath4 which , for all of the surveys analysed by us at @xmath62 ghz , everywhere exhibits a slope @xmath63 this holds in the full range @xmath64 , although we have large error bars for the range @xmath31 investigated on bs76 maps . in spite of the more intricate situation for the other aps @xmath61 ( @xmath65 , @xmath66 and @xmath67 ) , we can state the following : * in the galactic plane , the slopes of electric and magnetic parity aps are quite moderate at 2.4 and 2.7 ghz , with averages @xmath68 in the range @xmath69 . * local fluctuations
do not allow us to establish equally significant average slopes out of the galactic plane in the same angular range .
* at lower resolution , @xmath31 , large correlation between fit parameters cause large error bars ; however , the best - fit slopes @xmath54 and @xmath55 stay in the range @xmath70 for almost all frequencies ( in the range @xmath71 ghz ) and galactic latitudes , and are quite inconsistent with values around 3 .
the above behaviours of polarization aps should be attributed to galactic synchrotron , with no appreciable contamination from point sources . on the other hand , the total intensity aps may be locally dominated by sources , when they exhibit small amplitude and slope @xmath72 close to zero .
our results resolve the seeming discrepancy with other authors for the angular range @xmath69 , showing that investigators simply have to carefully consider which polarization aps is actually being computed .
our galactic plane result , @xmath73 is very close to results found for the polarization aps of thermal dust ( prunet et al .
1998 ) ; this number is maybe deeply connected to galactic structure . on the other hand , we do not confirm the high spectral slope found by baccigalupi et al .
( 2001a ) for @xmath27 at smaller @xmath32 on three bs76 patches at 1.4 ghz .
our bs76 patches however are different from theirs , being expressely chosen to have a larger signal - to - noise ratio .
our best value @xmath74 , which arises from averaging over 1.4-ghz bs76 patches , is consistent with our results at all resolutions and frequencies @xmath36 ghz .
the latter also agree with results obtained by tucci et al .
( 2001 ) in the arcminute range .
the slope of starligth aps , @xmath75 ( fosalba et al . 2001 ) , is closer to our @xmath76 . generally speaking ,
the @xmath77 and @xmath78 & @xmath66 fields contain different physical information .
since @xmath77 does not carry any information on the polarization angle , its spectrum can not properly describe some related effects like , for example , beam depolarization ; it should be used carefully , keeping in mind that it does not provide a complete description of the polarization field . in particular , if @xmath4 is extrapolated to the cosmological window , it is important to make a proper comparison with the corresponding aps of cmb .
figure [ cmbsyn ] compares both @xmath24 and @xmath4 spectra of synchrotron and cmb .
the reported cmb e - mode spectrum is the output of cmbfast for a scdm model with a reionization optical depth @xmath79 .
the corresponding @xmath80 spectrum is obtained through simulations in a @xmath81 box , with the mean value being subtracted off .
[ simulations of @xmath82 in the same box reproduced the cmbfast output to a satisfactory extent , see tucci et al .
( 2000 ) for details . ]
the synchrotron aps in the figure are extrapolated to 60 and 90 ghz from the galactic plane average spectra ( 6th row in table [ avegp ] ) for @xmath83 , and from bs76 patch centered at @xmath84 for @xmath85 . in both cases , the computed normalization applies to high - emission regions and is not expected to be representative of the whole sky . for the extrapolations to high frequencies we assume a spectral index @xmath86 ( platania et al . , 1998 ) .
the results taken altogether offer a quite consistent picture .
they imply that in high emission regions , whathever polarization aps is chosen , the synchroton polarized foreground should be comparable to cmb at 60 ghz even at large @xmath32 . at 90 ghz
the expected scenario looks more favourable for cmb experiments , both at small and large angular scales .
reionization effects on cmb should be investigated at @xmath87 by means of @xmath82 rather than @xmath80 . from the inspection of fig .
[ cmbsyn ] , and recalling that the height of the low-@xmath32 peak of @xmath82 is approximately proportional @xmath88 , we infer that the 90-ghz cmb signal from reionization should prevail on synchrotron at least for @xmath8 .
we finally remark that in the @xmath77 spectrum the `` dc '' signal @xmath89 might be interesting , but fourier analysis is not appropriate to this purpose .
we thank t.a .
spoelstra for kindly providing the bs76 data .
this work was performed within the sport collaboration , and was financially supported by the italian space agency ( asi ) .
giardino , g. , banday , a.j . ,
bennet , k. , fosalba , p. , gorski , k.m . ,
omullane , w. , tauber , j. & vuerli , c. , 2001 , in : mining the sky , proc .
mpa / eso / mpe conference , springer - verlag series `` eso astrophysics symposia '' , to be published .
tucci , m. , carretti , e. , cecchini , s. , nicastro , l. , fabbri , r. , gaensler , b.m . ,
dickey , j.m .
, & mcclure - griffiths , n.m . , 2001 , in : astrophysical polarized backgrounds , aip conf .
proc . , eds . s. cortiglioni , s. cecchini , c. sbarra & r. sault , in press . | we study the angular power spectra of polarized galactic synchrotron in the range @xmath0 , at several frequencies between 0.4 and 2.7 ghz and at several galactic latitudes up to near the north galactic pole . electric- and
magnetic - parity polarization spectra are found to have slopes around @xmath1 in the parkes and effelsberg galactic - plane surveys , but strong local fluctuations of @xmath2 are found at @xmath3 from the 1.4 ghz effelsberg survey .
the @xmath4 spectrum , which is insensitive to the polarization direction , is somewhat steeper , being @xmath5 for the same surveys .
the low - resolution multifrequency survey of brouw and spoelstra ( 1976 ) shows some flattening of the spectra below 1 ghz , more intense for @xmath6 than for @xmath4 . in no case
we find evidence for really steep spectra .
the extrapolation to the cosmological window shows that at 90 ghz the detection of e - mode harmonics in the cosmic background radiation should not be disturbed by synchrotron , even around @xmath7 for a reionization optical depth @xmath8 . , , , , , , background radiations radio continuum : general methods : statistical pacs : 98.70-f , 98.70.vc |
the filament eruption is one of the most spectacular phenomena on the sun .
previous studies have shown that filament eruptions are closely associated with magnetic flux cancellation and emergence on the photosphere @xcite , flares , and coronal mass ejections ( cmes ) , which are large - scale expulsions of mass and magnetic field from the sun into the heliosphere ( for a review , see * ? ? ?
it has been widely accepted that the filament eruption , flare , and cme are three different manifestations of a single physical process originating from the coronal magnetic field , but the relationships among them are not well understood @xcite .
generally speaking , the eruption of a filament always starts from a closed magnetic system in quasi - static equilibrium , in which the upward magnetic pressure force of the low - lying sheared field is balanced by the downward tension force of the overlying field . when the eruption begins , the equilibrium is destroyed catastrophically , and part of the nonpotential magnetic flux and the plasma contained within it are expelled from the sun violently . however ,
when and how the eruption takes place is still not well understood , even though extensive observational and theoretical studies have been made in the past several decades . from the space weather point of view
, it is important to enrich our knowledge about the physical mechanisms of various kinds of eruptive solar phenomena to improve our ability to forecast space weather . for the present ,
investigating the coronal magnetic field configuration and its spatial and temporal evolution with high temporal spatial resolution data are important to provide diagnostics on various stages of solar eruptions .
previous studies have shown that filament eruptions exhibit various eruptive behaviors , including failed eruption ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , partial eruption ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and full eruption ( e.g. , * ? ? ?
@xcite have made a specific definition about the three types of the filament eruptions .
according to their definition , a failed filament eruption is a case in which neither the lifted filament mass nor the supporting magnetic structure escapes the solar gravitational field .
recent observational studies indicated that the failed filament eruptions could possibly be sub - divided into two subtypes : one is that the filament , often strongly kinked , erupts symmetrically below a group of overlying loops .
when the erupting filament reaches a maximum height , it will fall back to the solar surface ( e.g. , * ? ? ?
* ; * ? ? ?
the other is that the filament erupts asymmetrically with respect to the overlying confining loops near either footpoint of the loops .
when the erupting filament reaches up to the apex of the overlying loops , it stops there for a relatively long time and then falls down along one or both legs of the confining loops ( e.g. , * ? ? ?
* ; * ? ? ?
up to the present , several important impact factors are proposed to be the reasons for failed filament eruptions .
for instance , a slowly decreasing gradient of the overlying magnetic field @xcite , strong magnetic field intensity at low altitude @xcite , an asymmetric confinement of the background field above the filament @xcite , and lower energy released than that of a successful eruption in the low corona @xcite .
a partial filament eruption occurs when a reconnection site is located within or above the filament .
the reconnection separates the flux rope structure into two parts : an escaping and a surviving part .
the former often can be identified as a cme with or without a bright core in white - light coronagraph observations , resting on where the reconnection site is . on the other hand ,
the latter often leads to the reformation of a new filament in the same channel shortly after the eruption . in a statistical study presented by @xcite
, they found that the separation in partial filament eruptions often occurs in the height range from 1.20 to @xmath1 .
@xcite proposed a two - dimensional scenario to interpret partial filament eruptions .
they suggested that the reconnection site at different positions of the flux rope would lead to different types of filament eruptions . in this scenario , if the reconnection site is located in ( above ) the filament , it will lead to a cme with ( without ) a three - part structure ; if the reconnection site lies below the filament , it will lead to a successful one .
a corresponding three - dimensional simulation of the partial expulsion of a flux rope was performed by @xcite .
a full filament eruption is a case in which the bulk of the filament material and the associated magnetic structure fully escape from the sun .
such eruption often causes a three - part structure cme observed in white - light , which comprises a bright leading shell of material surrounding a dark cavity within which bright prominence material is found @xcite .
a number of theories have been developed to account for filament / cme eruptions ( for reviews , see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . among many filament / cme models proposed so far ,
a majority of studies only considered the local bipolar region and did not take into account any possible multipolar component of the field on larger scales .
however , many filament eruptions often occur in multipolar topologies , such as quadrupolar and @xmath2 sunspot regions ( e.g. , * ? ? ?
* ; * ? ? ?
a numerical model , called magnetic breakout model , involving multipolar field geometry was proposed to interpret filament eruptions and cmes @xcite .
this model assumes a large - scale quadrupolar field configuration , in which the core field is assumed to be increasingly sheared by photospheric motions .
the core field is also surrounded by an overlying antiparallel loop system that restrains the core from erupting .
within such a magnetic configuration , there is a coronal null point that located between the core and the overlying fields .
the expansion of the core will lead to a current sheet formed around the null point region , and the subsequent reconnection in this current sheet progressively removes the field lines that confine the core field by transferring them to the lateral loop systems and thereby allows the sheared core field to erupt explosively outward . in recent years
, the magnetic breakout model has been enriched to a large extent @xcite , and many observational studies have also been put forward in support of the breakout model @xcite . when multiple solar eruptions occur consecutively within a relatively short time period in one complex @xcite , or different active regions over a large distance @xcite , we call them `` sympathetic eruptions ''
. a critical issue for sympathetic eruptions is whether the close temporal correlation between the consecutive eruptions is purely coincidental , or causally linked .
the key to resolve this issue is to find the physical relationship between them .
previous statistical and detailed case studies indicated that physical connections between them do exist ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
at present , it seems most likely that the basic mechanism that connects sympathetic eruptions is of a magnetic nature .
very recently , @xcite investigated a series of flares , filament eruptions , cmes , and related events that occurred on 2010 august 12 .
they found that all events during the two days were connected by a system of separatrices , separators , and quasi - separatrix layers , which indicates the importance of the magnetic topological structure of the global coronal field in the production of sympathetic eruptions .
@xcite also investigated three consecutive filament eruptions that occurred on 2010 august 1 . in this chain of events ,
one filament was located adjacent to a pseudo - streamer that contained two other filaments .
the filament outside the streamer erupted first , and then the other two filaments also erupted consecutively . by performing a three - dimensional simulation of a configuration similar to the observed event
, they successfully reproduced the consecutive filament eruptions , and found that the filament eruptions within the streamer resulted from the reduction of the stabilizing flux above them due to the reconnections within the streamer , triggered by the disturbance of the nearby filament eruption . here
, we present a partial and a full filament eruption that occurred simultaneously in two neighboring source regions that produced a large - scale quadrupolar configuration . this event occurred on 2011 may 12 and was accompanied by a _ goes _
c1.8 flare and a cme with a three - part structure . to our knowledge , this kind of event has not yet been reported before . combining the high temporal and spatial observations taken by
_ solar dynamic observatory _ ( _ sdo _ ) and _ solar terrestrial relations observatory _ ( _ stereo _ ; * ? ? ? * ) , we can investigate the sympathetic filament eruptions in great detail . based on our analysis results ,
we propose a possible mechanism to interpret the sympathetic filament eruptions studied in this paper , and other possible mechanisms are also discussed briefly . in the rest of the paper , instruments and data sets used in the present study are introduced in section 2 , and results are presented in section 3 . in section 4 , we propose a scenario to interpret our observational results , and discussions are also presented .
summary and conclusions are presented in the last section .
the atmospheric imaging assembly ( aia ; * ? ? ? * ) onboard _ sdo _ has high time cadences up to 12 s and short exposures of 0.12 s. it captures images of the sun s atmosphere out to @xmath3 with a pixel width of @xmath4 in seven euv and three uv visible wavelength bands .
the helioseismic and magnetic imager ( hmi ; * ? ? ?
* ) onboard _ sdo _ provides line - of - sight magnetograms at 45 s cadence with a precision of 10 g. the extreme ultraviolet imager ( euvi ; * ? ? ?
* ) of the sun earth connection coronal and heliospheric investigation ( secchi ; * ? ? ?
* ) onboard _ stereo _ provides full - disk 171 , 195 , 284 , and 304 images continuously with a pixel width of @xmath5 , and it observes the sun out to @xmath6 .
the 171 and 284 images are taken with a time cadence of 2 hours , and the time cadence for 195 ( 304 ) is 3 ( 10 ) minutes .
the cor1 instrument of secchi onboard _
stereo _ is an internally occulted coronagraph .
it observes the inner solar corona in white - light from 1.4 to @xmath7 @xcite .
the cor1 images are taken with a time cadence of 5 minutes , and have a pixel size of @xmath8 . on 2011
may 12 , the event was observed by _ stereo _ behind ( _ stereo_-b ) , and the separation angle of _ stereo_-b with _
sdo _ was about @xmath9 .
the source region of the event presented here was near the northwest limb on the _
stereo_-b images , while it was near the northeast limb from the _ sdo _ viewpoint . for convenience
, we will refer to the instruments _
stereo_-b / euvi and cor1 as euvi - b and cor1-b , respectively .
the h@xmath10 data are taken by kanzelh@xmath11he solar observatory ( kso ) in austria , which provides h@xmath10 full - disk images with a 1 minute cadence and a pixel width of roughly @xmath12 .
the _ reuven ramaty high energy solar spectroscopic imager _
( _ rhessi _ ; * ? ? ? * ) hard x - ray ( hxr ) sources ( 612 kev ) are reconstructed using the pixon algorithm @xcite , which provides a significantly better photometry measurement , accurate position estimation , and allows for the detection of fainter sources .
the detectors 38 are used in our reconstruction and the integration time is 1 minute around the _ rhessi _ flare peak . in addition , the _ geostationary operational environmental satellite _
( _ _ ) soft x - ray ( sxr ) flux is also used .
the event of sympathetic eruptions on 2011 may 12 included a partial and a full filament eruption , a _ goes _ c1.8 flare , and a cme observed by cor1-b . the primary magnetic configuration is a large - scale quadrupolar magnetic region , which evolved from the active regions 11193 and 11191 that were on the disk center on 2011 april 18 . an overview of the configuration before the eruptions is displayed in figure [ fig1 ] .
panel ( a ) shows the magnetic configuration of an hmi line - of - sight magnetogram .
the positive ( negative ) polarities are labeled `` p1 '' and `` p2 '' ( `` n1 '' and `` n2 '' ) , respectively . by overlaying the filament spines on the magnetogram
, one can see that f1 ( f2 and f3 ) was located on the neutral line between p1 and n1 ( n1 and p2 ) .
two loop systems , one connecting p1 and n1 and the other connecting p2 and n2 , are obvious in aia euv wavelengths ( see panels ( c ) , ( d ) , and ( g ) ) .
in addition , there are two cavities that can be identified off the disk limb ( panel ( d ) ) . with the procedure `` scc_measure.pro '' developed by w. thompson , we reconstruct the three - dimensional shapes of the spines of the three filaments by using a pair of 304 images from euvi - b at 11:26:15 ut and aia at 11:26:20 ut . obviously , f2 , exhibiting a snaky structure , is located higher than f1 . from the three - dimensional reconstruction
, we obtain that the height of the highest section of f1 ( f2 ) is about 16 ( 31 ) mm above the photosphere .
according to a statistical study presented by @xcite , the height of f2 is close to the lower threshold of the critical unstable height ( 41 mm ) of disrupted filaments , and f1 should be more stable than f2 .
since f3 and cavity 2 remained stable during the eruptions of f1 and f2 , and therefore they may not belong to the magnetic system involved in the sympathetic eruptions studied here , we only focus on the evolution of f1 , f2 , and cavity 1 .
in addition , a time line of the main sub - events and the corresponding features involved in the present event can be found in table table [ tb ] . before the sympathetic filament eruptions ,
a small plasma ejection was observed near the southern end of f2 .
the start time of the ejection was at about 11:30 ut , and then it split into two parts at around 11:32 ut ( see figure [ fig2](a ) and ( b ) ) .
the northern part moved northward into the active region , while the southern part traveled southward and then interacted with the southernmost section of f2 at about 11:40 ut .
subsequently , f2 rose slowly from the interaction region , most likely due to the perturbation induced by the plasm ejection ( see figure [ fig2](c)(d ) and figure [ fig5](b ) , and animation 1 available in the online version of the journal ) .
the slow rising of f2 lasted for about 23 minutes ( 11:45 ut12:08 ut ) at a speed of about 8 km s@xmath0 in the plane of the sky . during this period , three other interesting phenomena
are observed : ( 1 ) the brightening in the 1600 images , which were located at the both sides of f1 ( see figure [ fig2](e)(f ) and animation 2 ) , ( 2 ) the appearance of bright loops connecting the two brightening in the aia 94 wavelength ( see figure [ fig2](g)(h ) and animation 3 ) , ( 3 ) a _ rhessi _ hxr source detected at 11:56 ut over the bright loop top , located also between the two ribbon - like brightening ( see figure [ fig2](h ) ) .
the peculiar temporal and spatial relationship among these pre - eruption features indicates that they were produced by the same reconnection event .
figure [ fig3 ] shows the time profiles of the _ goes
_ 18 sxr flux , the _ rhessi _ hxr count rates in the lower energy bands , and light - curves of the two brightening regions in 1600 and the region of the hot loops in the 94 images ( see the boxes in figure [ fig2](f ) and ( h ) ) .
the time profiles of _ goes _ 18 flux show a small hump ( peaking at about 12:00 ut ) just before the onset of the main c1.8 flare , which peaked at 12:28 ut and 12:37 ut respectively ( indicated by the two arrows in figure [ fig3](a ) ) .
the _ rhessi _
hxr count rates and the light - curves of 1600 and 94 also exhibited such a hump around the same time ( see figure [ fig3](b ) and ( c ) ) . it should be noted that these phenomena are detected during the slow rise stage of f2 just before the filament eruptions .
the temporal and spatial relationship indicates that they are probably the manifestations of the external reconnection as predicted in the magnetic breakout model ( see , * ? ? ? * for detail ) .
the simultaneous eruptions of f1 , f2 , and of a blob - like structure above the apex of the erupting f1 are all displayed in figure [ fig4 ] , and their detailed kinematics are studied with time - distance diagrams as shown in figure [ fig5 ] . the top row of figure [ fig4 ] shows the full view of the eruptions of f1 and f2 using aia 304 images , while the middle row is a close - up view on the erupting f1 .
the eruption of f2 showed clear untwisting motions ( see animation4 ) and experienced three eruption phases : the slow rising phase ( 11:4512:08 ut ,
@xmath138 km s@xmath0 ) , the acceleration phase ( 12:0812:35 ut , @xmath14 m s@xmath15 ) , and the fast eruption phase ( after 12:35 ut , @xmath13102 km s@xmath0 ; see figure [ fig5](b ) ) . just after the slow rising period of f2
, f1 began to rise with a constant speed of about 63 km s@xmath0 .
the eruption of f1 showed a nice kinked - shape and exhibited obvious writhing motions during this period ( see figure [ fig4](e ) and animation 5 ) .
after f1 reached the maximum height ( @xmath16 mm above the solar surface ) at around 12:33 ut , the erupting material started to drain back to the solar surface .
around 12:20 ut , a strong emission occurred around the intersection of the f1 s two legs , and a _ rhessi _ hxr source is also found at this site ( the _ rhessi _ data available at 12:42 ut , see the cyan contours in figure [ fig4](f ) ) .
this result may indicate the occurrence of reconnection in a current sheet underneath the apex of the writhing f1 , which may formed via a kink instability resulting from the interaction of the two adjacent legs underneath the apex of the writhing filament ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , or by the instability - induced collapse of an x - line or hyperbolic flux tube that runs below the rising f1 ( e.g. , * ? ? ?
* ) . in the falling stage of f1 ,
a bright wedge - like post - eruptive loop structure can be observed ( see figure [ fig4](g ) ) .
after the eruption , the filament reappeared at the same place where f1 had resided in ( see figure [ fig4](h ) ) .
the evolutionary process of f1 is reminiscent of the failed filament eruption on 2002 may 27 that has been studied in detail by @xcite , @xcite , and @xcite . in their case , the filament experienced a kinking motion that was effectively stopped at a height of @xmath17 mm above the photosphere , and an hxr source was also identified under the apex of the strongly kinked filament during the erupting phase @xcite .
in addition , the reformation of the filament suggests that the eruption of f1 was possibly a partial filament eruption @xcite .
moreover , as can be seen in the aia 171 images , an intriguing blob - like feature above the apex of the writhing f1 erupted following the eruption of f1 .
it first appeared at about 12:15 ut and moved along the northeast direction .
meanwhile , an obvious outward motion of cavity 1 was also observed . at about 12:35 ut , the blob - like feature interacted with cavity 1 and
then they were moving outward together , leaving behind a clear dark cavity ( see the bottom row of figure [ fig4 ] and animation 6 ) .
the eruption of the blob - like feature and the reformation of f1 together indicate that the eruption of f1 is a partial filament eruption , in which the reconnection possibly occurs above the apex of the erupting f1 ( also see , * ? ? ?
* ) . to make clear the temporal relationship among the eruptions of f1 , f2 , and the blob - like feature , we plot the time - distance diagrams of them in figure [ fig5 ] , in which the main features are indicated by arrows , the important times are highlighted by the four vertical dotted lines , and the speeds of the moving features in the plane of the sky are also plotted ( see the green dashed lines that show linear fit to the stripes ) .
except for the features described in above paragraph , it is interesting that an obvious moving flare ribbon with a speed of 12 km s@xmath0 was identified at the beginning of the fast eruption phase of f2 ( see figure [ fig5](b ) ) , which may indicate the rising of the reconnection site that is associated with the erupting f2 .
the conjugate ribbon can also be found near the active region , and the moving speed is obviously smaller than the western ribbon ( see animation 5 ) .
in addition , obvious post - eruptive flare loops could be identified on the aia and euvi - b images during this phase .
these signatures suggest the reconnection process associated with the erupting f2 , which is similar to the physical picture described in the catastrophic flare / cme model by @xcite .
it is worth noting that the start time of the activation of f1 was at around 12:03 ut , which was close to the phase transition time of f2 from the slow rising phase to the acceleration phase .
this close temporal relationship suggests a close relation between f1 and f2 .
since the rising of f2 occurred before the activation of f1 , it seems that the eruption of f2 reconfigured its surrounding magnetic condition and thereby triggered its neighboring f1 .
in addition , as mentioned above , the _ goes _
c1.8 flare showed double peaks at 12:28 and 12:37 ut . considering the time of the flare and the eruptions of f1 and f2
, we infer that the flare possibly resulted from two components : one was related to the reconnection underneath the erupting f1 , and the other to the moving ribbon , related to the reconnection below the erupting f2 .
however , the moving ribbon associated with f2 was very weak compared to the emission caused by f1 .
therefore , there is another possibility that both the double peaks of the flare were caused by the strongly heated filament material of f1 during its eruption , in which the emission will come from both the heated filament material and from the related chromospheric ribbons , and their contributions may not be fully simultaneous and therefore produce the double flare peaks observed by _
goes_. along with the rising of f1 , a blob - like loop structure intruded into cavity 1 at about 12:15 ut with a speed of about 32 km s@xmath0 .
it was also accompanied by the expansion of cavity 1 at a speed of about 28 km s@xmath0 , slightly smaller than that of the blob . at about 12:33 ut ,
when f1 reached its maximum height , the majority of the risen loops that confined f1 started to fall back , but the blob kept on moving outward and subsequently interacted with cavity 1 , and then they erupted together with a speed of about 169 km s@xmath0 ( see the bright convex structure in the lower part of figure [ fig5](c ) ) .
for the blob eruption , its transition from the slow rising phase to the fast eruption phase occurred when f1 reached its maximum height , which suggests that the blob was separated from the flux rope structure that contained f1 due to reconnection that occurred within the flux rope system .
since no filament material expulsion could be detected in the eruption of f1 , we conjecture that the reconnection site was located about the apex of f1 . from another angle , _
stereo_-b also catched this event , and the results are shown in figure [ fig6 ] . in the euvi - b and cor1-b observations , except for the cavities , the main features , such as the plasma ejection , the consecutive eruptions of f1 and f2 , and the post - eruptive flare loop related to f2 , can also be observed .
( see animations 79 ) .
an intriguing phenomenon is that each filament was encircled by a preceding bright arch during their eruptions . for convenience
, we call the arches preceding f1 and f2 a1 , and a2 respectively ( see the white curves in figure [ fig6](d ) ) . a2 is possibly the progenitor of the bright front of the cme recorded by cor1-b ( figure [ fig6](i ) ) .
it is clear that the cme was caused by the erupting f2 . by overlying the spines of the erupting f2 at different time on the cor1-b image at 13:35 ut
, a close temporal and spatial relationship between the cme and the erupting f2 can be found ( see figure [ fig6](i ) ) .
obviously , the bright core of the cme corresponds to the erupting f2 , while the cme s bright front is the development of the erupting a2 on the white - light cor1-b observations .
in addition , the erupting f2 showed a northward drift relative to the initial eruption direction , which probably resulted from the orientation of the cavity overlying it .
however , we do not find a cme that could be considered as the developing of the erupting a1 , which is possibly because a1 is too faint to be detected by cor1 . by comparing the observations of _ stereo_-b and _ sdo _
, we infer that the erupting a1 observed at euvi - b 195 should be the counterpart of the erupting blob - like feature on the aia 171 images .
the event presented in this paper occurred near the eastern limb on the _ sdo _ images , and thus there was no reliable direct magnetic field measurement available . so we resort to the potential field source surface ( pfss ) model @xcite to obtain information of the three - dimensional magnetic environment in the vicinity of the quadrupolar source region . in this model ,
the fields in the corona are assumed to be a potential field ( current - free ) and become radial at the source surface at @xmath18 .
the extrapolation is based on a synoptic magnetic map that was composed of a series of consecutive hmi line - of - sight magnetograms , within a limited area around the central meridian . since the eruptions occurred on the eastern limb , the information of the source region on the synoptic map is mainly from that of three weeks ago and several days after , when the source region passed through the central meridian . in spite of this , we believe that the basic magnetic topology of the coronal field is still reliable , since the basic configuration of the source region on the hmi line - of - sight magnetograms did not change too much during the time interval of consideration . the results of the extrapolated coronal field are shown in figure [ fig7 ] .
it should be noted that only a few representative field lines are plotted in the figure . in this figure , one can see that three low - lying lobes ( blue , orange , and green ) connecting the four polarities are enclosed by a group of high , large field lines ( magenta ) connecting p1 and n2 . f1 and f2(f3 )
are located under the eastern and the middle low - lying lobes respectively .
a coronal null point is found between the low - lying middle lobe and the overlying antiparallel flux system .
the position of the coronal null point is indicated by the red `` x '' symbol in figure [ fig7](b ) .
such magnetic configuration is consistent with the magnetic topology in magnetic breakout models ( e.g. , * ? ? ?
* ) . by comparing the coronal field extrapolated from the pfss model and the _ sdo _ and _ stereo_-b observational results
, it seems likely that the overlying large loop system ( magenta ) can be considered as the counterpart of cavity 1 observed on the aia 171 images , while the left and the middle low - lying lobes are possibly the initial states of the erupting a1 and a2 observed on the euvi - b images .
based on the calculated three - dimensional coronal fields , we examine the gradient with respect to the height of the background magnetic fields , which is believed to be an important impact factor to diagnose the ultimate fate of a filament eruption ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this parameter is expressed as the so - called decay index , defined as @xmath19 . here , @xmath20 is the strength of the external field confining the erupting core and @xmath21 the height above the photosphere .
the theoretical instability condition is @xmath22 @xcite .
when the decay index satisfies this condition , a successful eruption can be expected due to the torus instability @xcite or partial torus instability @xcite .
it should be noted that the external field is produced by sources other than the currents in the sheared core field , and only the poloidal component of the external field was used to calculate the decay index in the theoretical work of @xcite . in our calculation
, we use the field extrapolated by the pfss model to approximate the real external field since it is difficult to separate the current - induced field from the pfss model field , and the transverse component of the three - dimensional coronal field is used to calculate the decay index .
we believe that this approximation has little influence on the results of this study , since we just make a comparison between the relevant values for the two filaments . at each height , the transverse component field is averaged over a small pixel region above the filament spine .
figure [ fig7](c ) shows the plots of the decay index varying with height for f1 ( blue ) and f2 ( orange ) . in general
, the decay index of the coronal field above f1 is lower than that of f2 at each height , which suggests that the transverse magnetic field in the lower corona above f1 decreases slower than that for f2 .
this result is in agreement with our observational results that f1 erupted partially while f2 erupted successfully , and it is also consistent with the theoretical prediction that the decay index for a confined eruption is typically smaller than that for a successful one .
it is interesting that the height distribution of the decay index for f1 increases monotonically , whereas for f2 , the decay index first shows a rapid increase and then decreases to a lower constant level .
such changing pattern of the decay indexes in the low corona suggests a difference distribution of the coronal field above f1 and f2 , and that the coronal field above f2 is more complex than that of f1 .
based on our analysis results , we propose a possible interpretation for the sympathetic filament eruptions presented in this paper , within the context of the magnetic breakout model ( figure [ fig8 ] ) .
panel ( a ) shows the initial magnetic topology of the coronal field . in this configuration ,
four polarities ( p1 , n1 , p2 and n2 ) are connected by four groups of flux systems , and two filaments ( f1 and f2 ) are confined by the left and middle low - lying flux systems .
the initiation of the eruptions starts from the disturbance induced by a plasma ejection .
subsequently , f2 rises slowly while f1 remains stable for a certain time .
the rising of f2 will lead to expansion of its overlying fluxes and formation of a current sheet ( cs1 ) between these fluxes and their overlying large antiparallel flux system ( see the red dashed curve in figure [ fig8](b ) ) . the continuously rising of f2 will speed up the magnetic reconnection in cs1 .
this reconnection will progressively remove the field lines confining f2 by transferring them to the lateral low - lying flux systems and thereby reduce the magnetic confinement ability of the fields above f2 . on the other hand , the accelerated particles spiral along the reconnected field lines down to the dense chromosphere , where hxr bremsstrahlung will be created because of their interaction with the thermal ions . during this period ,
brightening at the footpoints of the reconnected field lines can be expected at various lines , including sxr ( hxr ) detected by _ goes _ ( _ rhessi _ ) instruments ( figure [ fig8](c ) ) .
as proposed by @xcite , a magnetic implosion must occur simultaneously during the energy conversion process in coronal transients such as flares and cmes . under the condition of energy conservation ,
the energy release between the states before and after coronal transients implies the reduction of the upward magnetic pressure and thereby results in the contraction of the coronal fields overlying the energy releasing site ( see * ? ? ?
* ; * ? ? ?
* for details ) .
it should be noted that this magnetic implosion just occurs during the onset of the coronal transients in a short timescale ( several minutes ) and with a slow speed of a few km s@xmath0 .
recently , an unambiguous case which shows large - scale contraction of the coronal loops has been reported by @xcite . in their case ,
the contraction sustained for about 10 minutes at a mean speed of 5 km s@xmath0 , and then it was followed by the fast expansion of the coronal loops due to the compensation of energy from other regions .
other indirect evidence for the magnetic implosion phenomenon mainly comes from the observations on both converging motions of the conjugate footpoints and the descending motion of looptop sources during the early phase of some flares ( e.g. , * ? ? ?
* ; * ? ? ?
? * ; * ? ? ?
* ) . in the present case
, we consider such a magnetic implosion mechanism as the linkage between the sympathetic filament eruptions , even though no obvious contraction of the overlying coronal loops could be detected from the observations .
based on the magnetic implosion conjecture , however , we believe that the reduction of the magnetic pressure around the reconnection region must occur due to the releasing of the magnetic free energy during the early stage of the external reconnection in cs1
. a reduction of the magnetic pressure there will lead to the contraction of the overlying loops connecting p1 and n2 , and the expansion of the three low - lying lobes .
the expansion of the left lobe that connects p1 and n1 will successively reduce the stabilizing tension on f1 .
on the other hand , the strong writhing of f1 indicates that its eruption was driven by the kink instability within it .
the reduction of the magnetic tension force will lead to the eruption of f1 , by triggering the kink instability within the filament .
hence , the implosion could be a possible physical linkage of the sympathetic filament eruptions .
the continuously rising of f2 will result in the formation of a new current sheet ( cs2 ) between the legs of the field lines astride f2 .
the reconnection occurring in cs2 cuts the field lines that confine f2 and thus produces the observed cme , the moving flare ribbons , and the post - eruptive flare loops .
the eruption of the blob and the reformation of f1 could be explained using the partial flux rope eruption scenario @xcite , in which reconnection occurs in a current sheet ( cs3 ) that locates above the filament .
it should be noted that the reconnection in cs1 will increase the confinement ability of the fields above f1 by transferring the field lines from the middle lobe to the lateral lobes , and the reconnection in cs3 also increases the confinement of the fields above f1 .
hence , the eruption of f1 is just a partial flux rope eruption and no filament material expulsion .
the decay index analysis indicates that the drop of the field above f1 stayed for a considerable height range below the typical torus instability threshold , which may be an additional reason for the magnetic confinement of f1 . in our interpretation presented in figure [ fig8 ] ,
brightening should be appeared at p1 , n1 , p2 , and n2 during the external reconnection in cs1 , as the case reported by @xcite . whereas , in the case presented here , obvious brightening only observed at p1 and n1 .
the absence of brightening at p2 and n2 are possibly due to the weak and dispersed distribution of the magnetic fields in these regions , as shown by the hmi magnetograms .
in addition , all the brightened loops at 94 during the slow rising phase of f2 , the peculiar hxr source at 11:56 ut , and the small hump detected before the initiation of the flare on the time profiles of _ goes _ 18 sxr flux and _ rhessi _ hxr count rates can be considered as the evidence for the external reconnection , in which the hxr source represents looptop source . moreover , if we consider cavity 1 as the counterpart of the overlying large flux system shown in the cartoon ( figure [ fig8 ] ) , an expansion motion of cavity 1 should be observed .
however , the observational fact is that the obvious expansion of cavity 1 started along with the eruption of the blob , several minutes after the end of the slow rising phase of f2 ( see figure [ fig5 ] ) .
one possibility for this discrepancy is that the expansion of cavity 1 was too weak to be detected during the early phase of the external reconnection .
the other possibility is that cavity 1 did not belong to the quadrupolar magnetic system at all , since we do not observe it on the _
stereo_-b observations . both the post - eruptive flare loops and the moving flare ribbon associated with f2
could be considered to be the evidence for the reconnection in cs2 . for
the reconnection occurred in cs3 , we do not find significant evidence for it .
however , the observational results do indicate the eruption of f1 was a partial flux rope eruption , such as the reformation of f1 .
further detailed studies on such partial flux rope eruptions are needed to clarify this mechanism .
the sympathetic filament eruptions presented in this paper show some similarity to the study presented by @xcite . in their case ,
two filaments , which are confined by a pseudo - streamer , erupt consecutively due to the removal of a sufficient amount of confining fields above them . in our case
, the rising of f2 could lead to the external reconnection in cs1 .
this reconnection successively removes the stabilizing flux above f2 , which further accelerates the eruption of the filament .
it should be noted that such a mechanism is also important for producing coronal blowout jet @xcite , in which external reconnection removes the confining fluxes above a small filament close to the jet base , and thus leads to the eruption of the filament .
both our results and the study by @xcite indicate that the structural properties of the large - scale coronal field are important for producing sympathetic eruptions . in the study performed by @xcite
, they found that a series of explosive and eruptive events were connected by a system of separatrices , separators , and quasi - separatrix layers .
this study highly stresses the importance of the topological features in the coronal field , and proposed three different interpretations to explain the series of consecutive eruptions ( see * ? ? ? * for detail ) .
one of their interpretations is that sympathetic eruptions might be the result of the destabilization of local field configurations by an overall change of the large - scale coronal magnetic field . in this scenario ,
the events that occur simultaneously or in a short timescale over long distances are not a chain in which one triggers another .
this mechanism , to a certain extent , could be apply to explain the sympathetic filament eruptions presented in this paper .
based on our analysis results , we can assume that f1 was close to the kink instability threshold , and f2 was close to the critical height .
these magnetic systems in quasi - static equilibrium were prone to be destabilized catastrophically by the large - scale change in the global magnetic configuration .
this explanation implies that the eruptions of f1 and f2 were independent , and that no direct physical linkage existed between them .
however , we find various evidence that support the magnetic breakout scenario as shown in figure [ fig8 ] , such as the pre - eruptive signatures and the coronal null point found in the three - dimensional coronal field topology by using the pfss extrapolation .
in addition , for the small plasma ejection observed at around 11:30 ut , it first split into two parts and then moved bidirectionally in the southern and the northern direction .
while the southern part interacted with f2 , the northern part traveled into the source region where f1 resided in .
this raises the question whether the consecutive eruptions of f1 and f2 were triggered by the split plasma ejection simultaneously , and whether the brightening around p1 and n1 were caused by , for instance , tether - cutting - like reconnection below f1 initiated by the northward plasma ejection , and the reconnection further led to the destabilization of f1 . to clear these questions , we check the temporal relationship between the plasma ejection and the pre - eruptive signatures .
we find that the northward plasma ejection disappeared at about 11:38 ut , the brightening first appeared at about 11:54 ut , and the start time of the rising of f1 was about 12:08 ut .
these temporal relationships indicate that the pre - eruptive signatures were possibly not triggered by the plasma ejection . on the contrary
, they were possibly resulted by the external reconnection at the null point , as discussed above .
we present the first observational evidence for that both partial and full flux rope eruptions can simultaneously occur in one solar breakout event .
the high temporal and high spatial observations of aia allow us to distinguish the relationship between the both filament eruptions . in the event , the initiation of the large - scale magnetic system started from the perturbation induced by the plasma ejection , which directly acted on f2 and thereby led to the rising of the filament .
the eruption of f2 was successful and associated with a cme observed in white - light observations , which underwent three eruption phases : the slow rising phase , the acceleration phase , and the fast eruption phase . during the slow rising phase , evidence for the external reconnection around the coronal null point are identified , such as the brightening on both sides of f1 , the bright loops at aia 94 wavelength , the hxr source at 11:56 ut , as well as the small hump detected by _ goes _ and _ rhessi _ before the main c1.8 flare . on the other hand , f1 starts to erupt at around the end of the slow rising phase of f2 , and the eruption of f1 was interpreted as a partial flux rope eruption , in which the flux rope was separated by the reconnection occurred above the apex of f1 , and no filament material expulsion could be detected .
the three - dimensional coronal field is obtained based on the pfss extrapolation , which reveals the basic magnetic topology of coronal field above the quadrupolar source region .
the magnetic configuration is composed of three low - lying lobes and a large - scale overlying flux system , in which a coronal null point is found between the middle lobe and the overlying antiparallel field lines . by calculating the decay index of the coronal fields above the two filament
, we find that the decay index for f1 is smaller than that for f2 at the same height , in agreement with the theoretical prediction that the decay index for a confined eruption is typically smaller than that for a successful one .
in addition , the height distribution of the decay index for f2 showed a decrease at the lower height , whereas , the decay index for f1 increases monotonically with height .
the distribution of the decay indexes above the filaments in the low corona indicate the coronal field distribution above f2 is more complex than that of f1 .
a possible mechanism is proposed to interpret the sympathetic filament eruptions . in our interpretation ,
the eruption of f2 was caused by the removal of a sufficient amount of the stabilizing field above the flux rope via the external reconnection around the coronal null point . in the meantime
, the external reconnection also increases the confinement ability of the fields above f1 . to link the two consecutive filament eruptions , we adopt the magnetic implosion mechanism during the early phase of external reconnection to connect them , even though we do not find obvious evidence for the implosion mechanism in our available observations
however , this interpretation is supported by a number of observational evidence , such as the various pre - eruptive signatures .
we stress that the magnetic breakout scenario is just one possibility for the case presented in this paper .
we do not intend to exclude other possibilities , such as the change of the large - scale magnetic configuration . in any case , the topology properties of the large - scale coronal field are important for producing sympathetic eruptions . in summary ,
our analysis results support the breakout scenario proposed by s. antiochos and coworkers , and we believe that the existence of the physical linkage between the sympathetic filament eruptions studied in this paper . more investigations involving similar magnetic structure with high temporal and spatial resolution observations would be helpful to fully understand the physical mechanisms of sympathetic solar eruptions .
sdo is a mission for nasa s living with a star ( lws ) program .
the _ stereo_/secchi data used here were produced by an international consortium of the naval research laboratory ( usa ) , lockheed martin solar and astrophysics lab ( usa ) , nasa goddard space flight center ( usa ) , rutherford appleton laboratory ( uk ) , university of birmingham ( uk ) , max - planck - institut for solar system research ( germany ) , centre spatiale de li@xmath23ge ( belgium ) , institut doptique th@xmath24orique et appliqu@xmath24e ( france ) , and institut dastrophysique spatiale ( france ) .
we thank the kso , _ rhessi _ , and _ goes _ teams for data support .
we would like to thank the anonymous referee for constructive comments and suggestions that improved the content of the manuscript .
this work is supported by the natural science foundation of china under grants 10933003 , 11078004 , and 11073050 , and the national key research science foundation ( 2011cb811400 ) , and open research program of key laboratory of solar activity of national astronomical observatories of chinese academy of sciences ( klsa2011_14 ) .
alexander , d. , liu , r. , & gilbert , h. r. 2006 , , 653 , 719 antiochos , s. k. 1998 , , 502 , l181 antiochos , s. k. , devore , c. r. , & klimchuk , j. a. 1999 , , 510 , 485 aulanier , g. , deluca , e. e. , antiochos , k. s. , mcmullen , r. a. & golub , l. 2000 , , 540 , 1126 aulanier , g. , t@xmath11r@xmath11k , t. , d@xmath24moulin , p. , & deluca , e. e. 2010 , , 708 , 314 bi , y. , jiang , y. , yang , l. h. , & zheng , r. s. 2011 , , 16 , 276 bong , s .- c . , moon , y .-
cho , k .- s .
2006 , , 636 , l169 cheng , x. , zhang , j. , ding , m. d. , guo , y. , & su , j. t. 2011 , , 732 , 87 crifo , f. picat , j. p. , & cailloux , m. 1983 , , 83 , 143 devore , c. r. , & antiochos , s. k. 2005 , , 628 , 1031 devore , c. r. , & antiochos , s. k. 2008 , , 680 , 740 fan , y. , & gibson , s. e. 2004 , , 609 , 1123 fan , y. , & gibson , s. e. 2007 , , 668 , 1232 forbes , t. g. 2000 , , 105 , 23153 forbes , t. g. 2010 , in heliophysics : space storms and radiation : causes and effects , ed . c. j. schrijver & g. l. siscoe ( cambridge : cambridge univ .
press ) , 159 forbes , t. g. , linker , j. a. , chen , j. , cid , c. , k@xmath25ta , j. , et al .
2006 , , 123 , 251 gary , g. a. , & moore , r. l. 2004 , , 611 , 545 gibson , s. e. , & fan , y. 2006 , , 637 , l65 gilbert , h. r. , alexander , d. , & liu , r. 2007 , , 245 , 287 gilbert , h. r. , holzer , t. e. , burkepile , j. t , & hundhausen , a. j. 2000 , , 537 , 503 gilbert , h. r. , holzer , t. e. , low , b. c. , & burkepile , j. t. 2001 , , 549 , 1221 howard , r. a. , moses , j. d. , vourlidas , a. , et al .
2008 , , 136 , 67 hudson , h. s. 2000 , , 531 , l75 hudson , h. s. , & cliver , e. w. 2001 , , 106 , 25199 isenberg , p. a. , & forbes , t. g. 2007 , , 670 , 1453 ji , h. , wang , h. , goode , p. r. , jiang , y. , & yurchyshyn , v. 2004 , , 607 , l55 ji , h. , wang , h. , schmahl , e. j. , moon , y .-
j , & jiang , y. 2003 , , 595 , l135 jiang , y. , shen , y. , yi , b. , yang , j. , & wang , j. 2008 , , 677 , 699 jiang , y. , yang , j. , hong , j. , bi , y. , & zheng , r. , 2011 , , 738 , 179 kaiser , m. l , kucera , t. a. , davila , j. m. , et al .
2008 , , 136 , 5 kliem , b. , linton , m. g. , t@xmath11r@xmath11k , t. , & karlick@xmath26 , m. 2010 , , 266 , 91 kliem , b. , & t@xmath11r@xmath11k , t. 2006 , , 96 , 255002 kumar , p. , srivastava , a. k. , filippov , b. , erd@xmath24lyi , r. , & uddin , w. 2011 , , 272 , 301 lemen , j. r. , title , a. m. , akin , d. j. , et al . 2012 , , 275 , 17 lin , j. , & forbes , t. g. 2000 , , 105 , 2375 lin , j. , soon , w. , & baliunas , s. l. 2003 , , 105 , 47 , 53 lin , r. p. , dennis , b. r. , hurford , g. j. , et al . 2002 , , 210 , 3 liu , c. , lee , j. , karlick@xmath26 , m. , prasad choudhary , d. , deng , n. , et al .
2009a , , 703 , 757 liu , k. , wang , y. , shen , c. , & wang , s. 2012 , , 744 , 168 liu , r. , alexander , d. , & gilbert , h. r. 2007 , , 661 , 1260 liu , r. , wang , h. , & alexander , d. 2009b , , 696 , 121 liu , y. 2008 , , 679 , l151 liu , y. , jiang , y. , ji , h. , zhang , h. , & wang , h. 2003 , , 593 , l137 liu , y. , su , j. , xu , z. , lin , h. , shibata , k. et al . 2009c , , 696 , l70 lynch , b. j. , antiochos , s. k. , devore , c. r. , luhmann , j. g. , & zurbuchen , t. h. 2008 , , 683 , 1192 lynch , b. j. , antiochos , s. k. , macneice , p. j. , zurbuchen , t. h. , & fisk , l. a. 2004 , , 617 , 589 macneice , p. , antiochos , s. k. , phillips , a. , et al .
2004 , 614 , 1028 maia , d. , aulanier , g. , wang , s. j. , et al .
2003 , , 405 , 313 manoharan , p. k. , & kundu , m. r. 2003 , , 592 , 597 metcalf , t. r. , hudson , h. s. , kosugi , t. , puetter , r. c. , & pina , r. k. 1996 , , 466 , 585 miki@xmath27 , z. , & lee , m. a. 2006 , , 123 , 57 moon , y .-
j . , choe , g. s. , park , y. d. , et al .
, & goode , p. r. 2002 , , 574 , 434 moon , y .-
choe , g. s. , wang , h. , & park , y. d. 2003 , , 588 , 1176 mrozek , t. 2011 , , 270 , 191 olmedo , o. , & zhang , j. 2010 , , 718 , 433 phillips , a. d. , & macneice , p. j. 2005 , , 624 , l129 plunkett , s. p. , vourlidas , a. , @xmath28imberov@xmath29 , s. et al .
2000 , , 194 , 371 schatten , k. h. , wilcox , j. m. , & ness , n. f. 1969 , , 6 , 442 schou , j. , borrero , j. m. , norton , a. a. , et al .
2012 , , 275 , 327 schrijver , c. j. , & derosa , m. l. 2003 , , 212 , 165 schrijver , c. j. , & title , a. m. 2011 , , 116 , a04108 shen , y. , liu , y. , su , j. , & deng , y. 2012 , , 745 , 164 shen , y. , liu , y. , & liu , r. 2011 , res .
astrophys . , 11 , 594 sterling , a. c. , & moore , r. l. 2001a , , 560 , 1045 sterling , a. c. , moore , r. l. , qiu , j. , & wang , h. 2001b , , 561 , 1116 sterling , a. c. , & moore , r. l. 2004a , , 602 , 1024 sterling , a. c. , & moore , r. l. 2004b , , 613 , 1221 sui , l. , & holman , g. d. 2003 , , 596 , l251 sui , l. , holman , g. d. , & dennis , b. r. 2004 , , 612 , 546 thompson , w. t. , willian , t. , davila , j. m. , et al . , 2003 , , 4853 , 1 t@xmath11r@xmath11k , t. , berger , m. a. , & kliem , b. 2010 , , 516 , a49 t@xmath30k , t. , kliem , b. , & titov , v. s. 2004 , , 413 , l27 t@xmath30k , t , & kliem , b. 2005 , , 630 , l97 t@xmath30k , t. , panasenco , o. , titov , v. s. , et al . 2011 , , 739 , l63 tripathi , d. , gibson , s. e. , qiu , j. , et al . 2009 , , 498 , 295 van der holst , b. , manchester iv , w. , sokolov , i. v. , et al .
2009 , , 693 , 1178 veronig , a. m. , karlick@xmath26 , m. , vr@xmath31nak , b. , et al .
2006 , , 446 , 675 wang , h. , chae , j. , yurchyshyn , v. , et al . 2001 , , 559 , 1171 wang , h. , liu , c. , jing , j. , & yurchyshyn , v. 2007 , , 671 , 973 wang , t. j. , yan , y. h. , wang , j. l. , kurokawa , h. , & shibata , k. 2002 , , 572 , 580 williams , d. r. , t@xmath30k , t. , d@xmath24moulin , p. , van driel - gesztelyi , l. , & kliem , b. 2005 , , 628 , l163 wuelser , j. , lemen , j. r. , tarbell , t. d. , et al . 2004 , , 5171 , 111 yan , x. l. , qu , z. q. , & kong , d .-
2011 , , 414 , 2803 yang , j. , jiang , y. , zheng , r. , bi , y. , hong , j. , et al .
2012 , , 745 , 9 zhang , y. z. , hu , y. q. , & wang , j. x. 2005 , , 626 , 1096 zhou , g. p. , wang , j. x. , zhang , j. , chen , p. f. , ji , h. s. , et al .
2006 , , 644 , 575 zhukov , a. n. , & veselovsky , i. s. 2007 , , 664 , l131 zuccarello , f. p. , soenen , a. , poedts , s. , zuccarello , f. , & jacobs , c. 2008 , , 689 , l157 cl approx .
time & corresponding feature + 11:30 ut & the first appearance of the plasma ejection + 11:32 ut & the plasma ejection was split into two parts + 11:40 ut & the southern part of the ejection acted on f2 + 11:45 ut & f2 started to rise slowly + 11:56 ut & start of the small pre - flare hump , peaked at 12:00 ut + & brightening appeared at around p1 and n1 + & brighten up of the loops that connect p1 and n1 + & an hxr source detected around the apex of the loops + 12:03 ut & f1 was activated and started to rise + 12:08 ut & the slow rising of f2 turned into the acceleration phase + & the start of the main _ goes _
c1.8 flare peaking at 12:28 and 12:37 ut + 12:15 ut & the first appearance of the blob - like feature + 12:33 ut & f1 reached the maximum erupting height and started to fall back + & the start of the fast eruption of the blob - like feature + & the acceleration of f2 transited into the fast eruption phase + | we report two sympathetic solar eruptions , including a partial and a full flux rope eruption in a quadrupolar magnetic region , where a large and a small filament resided above the middle and the east neutral lines respectively .
the large filament first rose slowly at a speed of 8 km s@xmath0 for 23 minutes and then it was accelerated to 102 km s@xmath0 . finally , this filament erupted successfully and caused a coronal mass ejection . during the slow rising phase , various evidence for
breakout - like external reconnection has been identified at high and low temperature lines .
the eruption of the small filament started around the end of the large filament s slow rising .
this filament erupted partially and no associating coronal mass ejection could be detected .
based on a potential field extrapolation , we find that the topology of the three - dimensional coronal field above the source region is composed of three low - lying lobes and a large overlying flux system , and a null point located between the middle lobe and the overlying antiparallel flux system .
we propose a possible mechanism within the framework of the magnetic breakout model to interpret the sympathetic filament eruptions , in which the magnetic implosion mechanism is thought to be a possible linkage between the sympathetic eruptions , and the external reconnection at the null point transfers field lines from the middle lobe to the lateral lobes and thereby leads to the full ( partial ) eruption of the observed large ( small ) filament .
other possible mechanisms are also discussed briefly .
we conclude that the structural properties of coronal fields are important for producing sympathetic eruptions . |
experimental realization of atomic bose - einstein condensation and degenerated fermi gases , has stimulated a large interest in the physics of ultracold gases @xcite . among others ,
the issue of fluctuations in the number of condensed atoms has been a subject of intensive theoretical studies @xcite-@xcite .
it is well - known that the standard textbook approach based on the grand - canonical ensemble predicts unphysically large fluctuations .
this pathological behavior is avoided , however , when one employs the canonical or the microcanonical ensemble , where the total number of particles is fixed . from the experimental point of view ,
trapped atoms are under conditions of almost complete isolation .
this favors the use of microcanonical ensemble , at least in the situations when predictions of different statistical ensembles are not equivalent .
while in ideal gases the microcanonical fluctuations have been thoroughly investigated @xcite-@xcite , the corresponding problem for interacting gases has been only studied in the first - order perturbation theory @xcite , or for relatively small systems by means of numerical methods @xcite . on the other hand ,
the bogoliubov theory @xcite , which has proved to be immensely successful to describe weakly interacting gases , has been applied to study the fluctuations in the canonical ensemble @xcite , but to our best knowledge , it was not used to investigate the microcanonical fluctuations .
the purpose of this paper is to address the problem of microcanonical fluctuations in the framework of bogoliubov theory . for a homogeneous system
we perform our analysis within the particle - number - conserving formulation of the bogoliubov theory , developed by girardeau and arnowitt @xcite . in this way
we avoid the possible errors resulting from the nonconservation of the number of particles in the bogoliubov method .
we obtain that the microcanonical fluctuations , similarly to the canonical ones @xcite , exhibit anomalous scaling with the total number of particles .
moreover , fluctuations in the microcanonical and the canonical ensemble differ by an extensive quantity , and predictions of both ensembles become equivalent in the thermodynamic limit . for a homogeneous system
this feature is already observed for an ideal gas .
nevertheless , in a harmonic trap , the microcanonical fluctuations of noninteracting condensate remain different from the canonical ones even in the thermodynamic limit .
we show here , that inclusion of weak interactions described in terms of bogoliubov theory , restores the equivalence of microcanonical and canonical description , and gives rise to the anomalous scaling of fluctuations in both considered ensembles .
noteworthy , for a homogeneous system we observe that application of the particle - number - conserving method leads to the similar results as obtained from the standard , nonconserving approach .
therefore , for a trapped gas we carry out our derivations only within the nonconserving theory , which we find more convenient in this case .
we discuss this issue in the last section explaining why at low temperatures the fluctuations are insensitive to the actual number of atoms .
the anomalous scaling of the fluctuations predicted by the bogoliubov theory is closely related with the existence of phonon excitations at low energies . on the contrary ,
application of the first - order perturbation theory results in the normal scaling of the fluctuations @xcite .
while the perturbative treatment is more appropriate for finite size systems with sufficiently small interactions , in the thermodynamic limit the system should be rather described within the bogoliubov theory , since the presence of arbitrary small interactions leads to the phonon excitations at low energies in the thermodynamic limit .
we note that there is some controversy on using the bogoliubov method to calculate the quantities higher than the second order in the field operators @xcite .
an excellent agreement of the bogoliubov theory with the exact statistics calculated for the one dimensional trapped gas @xcite provides , however , strong arguments in favor of the bogoliubov approach . the ultimate test of the theoretical predictions should be provided by experiments , either through the measurement of the second - order correlation functions , or by detecting the statistics of scattered photons @xcite . the paper is organized as follows . in section [ sec : idealgas ] , for the sake of completeness , we analyze the fluctuations of ideal bose gases confined in a three dimensional box and a harmonic trap . the behavior of fluctuations in an interacting gas confined in a three dimensional box is studied in section [ sec : intgas ] . in particular , in section
[ sec : gatheory ] we briefly describe the girardeau - arnowitt particle - number - conserving formalism .
section [ sec : fluctcn ] is devoted to the analysis of fluctuations in the canonical ensemble , while the microcanonical fluctuations are calculated in section [ sec : fluctmc ] .
section [ sec : intgastrap ] presents the results for a weakly interacting trapped gas .
the canonical fluctuations are considered in section [ sec : fluctcntrap ] , whereas section [ sec : fluctmctrap ] describes the results for the microcanonical ensemble .
we end in section [ sec : conclusions ] presenting some conclusions .
finally , four appendices give some technical details .
in this section we analyze fluctuations of the number of condensed atoms in an ideal gas .
first , we consider homogeneous system with @xmath0 noninteracting bosons confined in a box of the size @xmath1 with periodic boundary conditions . below the critical temperature of the bose - einstein condensation , the fluctuations of the number of condensed atoms can be calculated with the help of the _ maxwell s demon ensemble _ ( mde ) @xcite . in this approach
the condensate serves as an infinite reservoir of particles for the subsystem of excited states .
such treatment is justified as long as the probability of states with all the atoms excited is negligible , hence it is not valid in the vicinity of the critical temperature .
the mean number of excited atoms @xmath2 and its fluctuations @xmath3 in the canonical ensemble , can be calculated from the grand canonical partition function @xmath4 of the excited subsystem @xcite @xmath5 in this formula @xmath6 denote the single particle energy of the level @xmath7 , @xmath8 is the inverse temperature and @xmath9 is the fugacity . in the canonical ensemble ,
the total number of particles @xmath0 is conserved , and obviously @xmath10 , and @xmath11 , where @xmath12 and @xmath13 denote the mean number of condensed atoms and its fluctuations , respectively . in the thermodynamic limit ,
summation on the r.h.s . of eqs .
( [ ne ] ) and ( [ dne ] ) , can be replaced by integration . however ,
this is not always possible , in particular , in the case of three dimensional box , where resulting integral for the condensate fluctuations is infrared divergent . to avoid these difficulties , in our study the thermodynamic quantities
are evaluated from the integral representation involving spectral zeta function @xcite . employing integral representation of the exponential function @xmath14
, one can express @xmath2 and @xmath3 as contour integrals of mellin - barnes type @xcite @xmath15 where @xmath16 denotes riemann zeta function , and @xmath17 is the spectral zeta function : @xmath18 .
the contour of integration lies to the right of all the poles of the integrand .
now , the problem of calculating fluctuations has been reduced to determination of the poles of @xmath17 . for the considered case of a three dimensional box with periodic boundary conditions ,
the spectral zeta function reads @xmath19 where @xmath20 denotes the zeta function of epstein @xcite @xmath21 and @xmath22 is the energy of the first excited state .
function @xmath20 has merely single pole at @xmath23 equal to @xmath24
. the straightforward calculation of the contour integrals ( [ neint ] ) and ( [ dneint ] ) with zeta function ( [ zbox ] ) , leads to @xmath25 where @xmath26 . in derivation of eqs .
( [ neideal ] ) and ( [ dneideal ] ) , we have included leading and next to the leading rightmost pole of the integrand .
thus , the result accounts also for correction due to the finite size of the system .
as was first pointed out in @xcite , the canonical fluctuations in a box are anomalous .
this can be observed by rewriting eq .
( [ dneideal ] ) in the form containing explicit dependence on the number of particles @xmath27 where @xmath28 , and @xmath29 is the critical temperature .
now we turn to the microcanonical ensemble . in order to calculate fluctuations of the number of condensed atoms , we employ the following relation @xcite @xmath30 ^ 2 } { \left\langle \delta^2 e \right\rangle_{\mathrm}{cn}},\ ] ] which links microcanonical fluctuations with the quantities calculated in the canonical ensemble : particle - energy correlation @xmath31 , and fluctuations of the system s energy @xmath32 .
the former can be determined from the mean number of excited particles @xmath2 @xmath33 in a similar manner , one can calculate the fluctuations of energy from its expectation value @xmath34 while the mean energy @xmath35 can be expressed as a contour integral containing spectral zeta function @xmath17 @xmath36 after straightforward calculations , we arrive at the following result for the microcanonical fluctuations in a box @xmath37 by comparison with the canonical fluctuations ( [ dneideal ] ) , we see that the leading - order term , which gives rise to the anomalous scaling , is the same .
the difference between fluctuations in the considered ensembles appears in the finite size correction term , thus the microcanonical and canonical fluctuations become equal in the thermodynamic limit .
fluctuations of the number of excited particles in the microcanonical and canonical ensemble for the system of @xmath38 noninteracting bosons confined in a box with periodic boundary conditions ( upper panel ) and in a harmonic trap ( lower panel ) .
analytical results ( solid line ) derived in mde approximation are compared to the exact numerical data ( dashed line ) .
the temperature is scaled in the units of the critical temperature @xmath39 .
, width=321 ] let us turn now to the case of a trapped gas . for simplicity
we assume that the atoms are confined in a spherically symmetric harmonic potential , however generalization of the results for anisotropic traps is straightforward .
the spectral zeta function for the harmonic trap of frequency @xmath40 is given by @xcite @xmath41 with @xmath42 .
substituting this function into eq .
( [ dneint ] ) and taking into account leading and next to the leading rightmost poles , we obtain the following result for the fluctuations in the canonical ensemble @xmath43 where @xmath44 denotes euler s constant .
the logarithmic correction in ( [ dneidealtrap ] ) originates from a double pole of the integrand at @xmath45 . in the thermodynamic limit : @xmath46 , @xmath47 and @xmath48_const _ , the leading order term in eq .
( [ dneidealtrap ] ) is proportional to @xmath0 and the fluctuations exhibit normal behavior . in the microcanonical ensemble we employ the identity ( [ relfl ] ) , which leads to the following result for the fluctuations @xmath49 with @xmath50 we observe that the leading order term , which gives rise to the normal scaling of fluctuation in the thermodynamic limit , is different than in the canonical ensemble . hence the microcanonical and the canonical fluctuations in the harmonic trap remain different even in the thermodynamic limit , in contrast to their behavior in a homogeneous system .
[ fig : fluktideal ] presents fluctuations of the number of condensed atoms in the canonical and the microcanonical ensembles , calculated for the system of @xmath38 atoms confined in a three - dimensional box ( upper panel ) and a three dimensional harmonic trap ( lower panel ) .
it compares the analytical predictions of eqs .
( [ dneideal ] ) , ( [ dnemc ] ) ( box ) and eqs .
( [ dneidealtrap ] ) , ( [ dnemcidealtrap ] ) ( harmonic trap ) with the exact numerical results calculated with the help of the recurrence algorithms for the canonical and microcanonical partition functions .
we see that analytic results agree very well with the numerical data , apart from the region close to the critical temperature , and above it , where the mde approximation ceases to be valid .
in this section we present a brief description of the particle - number - conserving version of the bogoliubov theory , formulated by girardeau and arnowitt ( ga ) @xcite . in the following ,
we restrict our analysis to the regime of temperatures much lower than the critical temperature .
this allows us to make the following assumptions : ( i ) the mean number of condensed atoms is much larger that its fluctuations @xmath51 , thus we can we can exclude the possibility of exciting the state where the condensate is totally depleted , ( ii ) we neglect the influence of a finite temperature on the excitation spectrum , through the thermal depletion of the condensate , which is predicted in the popov theory @xcite .
in addition we assume dilute - gas regime @xmath52 , where @xmath53 is the density of atoms and @xmath54 is the _
s_-wave scattering length , which allows to neglect the influence of the quantum depletion of the condensate on the bogoliubov excitation spectrum .
let us introduce unitary operators @xmath55 , @xmath56 , @xmath57 , where @xmath58 , @xmath59 are , respectively , the annihilation and creation operators of the particles with momentum @xmath60 , and @xmath61 .
the particle - number - conserving ground state of ga theory is @xmath62 with the unitary operator @xmath63 @xmath64 \right].\ ] ] coefficients @xmath65 are real , even function of @xmath66 . in the ga theory
they are determined by minimizing the expectation value of the full hamiltonian in variational ground state ( [ psi0 ] ) .
the inclusion of the total hamiltonian in the variational calculus leads , however , to the excitation spectrum which possess a gap @xcite , which is not appropriate for studying the system in the thermodynamic limit .
gapless bogoliubov spectrum is recovered , when one minimizes the expectation value of the particle - number - conserving bogoliubov hamiltonian @xmath67 where @xmath68 , @xmath69 , and @xmath70 is the operator of the total number of particles .
this approach assumes omitting the hartree - fock and pair - pair interaction terms in the part of the total hamiltonian , which gives nonzero contribution in the variational ground state .
these terms , however , are negligible in the considered range of temperatures .
the minimization procedure yields @xcite @xmath71 where @xmath72 , @xmath73 is the atomic density , and @xmath74 are the energy levels of the bogoliubov spectrum .
the excited states in ga theory are constructed in basically similar manner to the variational ground state @xmath75 where @xmath76 and @xmath77 . here
, @xmath78 is the set of integer numbers , where @xmath79 represents the number of elementary excitations with energy @xmath80 . in an interacting gas ,
canonical fluctuations of the number of condensed atoms can be calculated with the help of the following generating function . @xmath81 where @xmath82 is the operator of the number of noncondensed particles , and the trace has to be taken over all eigenstates of the system with the total number of particles equal to @xmath0 . for @xmath83 ,
generating function @xmath84 is equal , by definition , to the canonical partition function .
the fluctuations of the number of noncondensed atoms @xmath3 are given by @xmath85 obviously , the total number of particles is conserved , and the fluctuations in the number of condensed and noncondensed atoms are equal : @xmath11 .
we calculate @xmath84 by performing trace in the basis of eigenstates @xmath86 of ga theory @xmath87 where summation runs over sets @xmath78 of the occupation numbers of elementary excitations . in principle , according to the definition of @xmath86 , the number of elementary excitation @xmath88 can not exceed the total number of particles @xmath0 . however , well below the critical temperature , the probability of exciting such a state is negligible , and we can consider the summation in eq .
( [ xu ] ) as unconstrained .
similar approximation is assumed in mde , where we treat the condensate as a reservoir of particles for excited subsystem , and as a consequence we consider the partitions of excited atoms in the canonical ensemble without the particle number constrain .
the transformed operators @xmath89 and @xmath90 can be found with the help of the following identity @xcite @xmath91 where @xmath92 and @xmath93 are the usual amplitudes of the bogoliubov transformation .
after some straightforward algebra we obtain @xmath94 where @xmath95 is the ground state energy of the system .
the transformation @xmath96 does not diagonalize the hamiltonian @xmath97 exactly , and in the derivation of ( [ uhbu ] ) we have neglected terms describing interactions between excited atoms .
this approximation is consistent with neglection of the hartree - fock and pair - pair scattering terms in the total hamiltonian , and is well justified in the considered range of temperatures . employing eqs .
( [ dnecnint ] ) , ( [ xu ] ) , ( [ uneu ] ) and ( [ uhbu ] ) we obtain the following result for the fluctuations in the canonical ensemble @xmath98 in comparison to the ideal gas result ( [ dne ] ) , thermal fluctuations of the number of elementary excitations , are multiplied by the factor @xmath99 .
second term on r.h.s . of eq .
( [ dnecnint1 ] ) , describes the quantum fluctuations at zero temperature @xmath100 .
we note that a similar expression can be derived in the framework of the standard , particle - number - nonconserving theory @xcite .
we comment on this issue in the last section of the paper .
now , we rewrite @xmath101 in terms of contour integral involving spectral zeta function .
calculations similar to the derivation of ( [ neint ] ) and ( [ dneint ] ) , yield the following integral representations @xmath102 with spectral zeta function @xmath103 where @xmath104 , and @xmath105 is a vector composed of integer numbers .
spectral zeta function @xmath106 can be expressed as a sum of two simpler functions @xmath107 with @xmath108 in the thermodynamic limit : @xmath46 , @xmath109 , @xmath110 , and @xmath111 .
thus , for sufficiently large system we can consider the regime @xmath112 , and in the subsequent derivation we will approximate function @xmath113 , by its asymptotic expansion for large parameter @xmath114 .
the analytic properties of the function @xmath113 are studied in appendix [ appresid ] , while its asymptotic expansion is developed in appendix [ appasympt ] . in the former
it is shown , that the poles of @xmath113 are located at @xmath115 , @xmath116 , with residues equal to @xmath117 .
hence the residues of @xmath118 are proportional to @xmath119 , and inclusion in the calculation only the leading rightmost poles , would yield the result which is valid only in the regime of temperatures @xmath120 in order to obtain result , which holds also for low temperatures : @xmath121 , in the calculation of contour integral ( [ dnecnint2 ] ) , we have to sum over all the poles which lie to the left of the contour of integration .
this summation can be done analytically , the details of the derivation are presented in appendix [ appcnfluct ] .
the result for the canonical fluctuations reads @xmath122 t^2 + \frac{\pi^2}{6 } ( 2\alpha)^{\frac{3}{2 } } \nonumber \\ & { } + \left [ \zeta_h\left(\frac{1}{2},1+\frac{\alpha}{2 \pi
t}\right)+ \frac{\alpha^2}{2 \pi^2 t^2 } s_1\left(\frac{\alpha}{2 \pi t}\right ) \right .
\nonumber \\ & \left .
\quad \ , { } + \frac{\alpha}{2 \pi t } \zeta_h\left(\frac{3}{2},1 + \frac{\alpha}{2 \pi t}\right ) \right ] ( \pi t)^{\frac{3}{2}},\end{aligned}\ ] ] where @xmath123 and @xmath124 is zeta function of hurwitz : @xmath125 , defined by the series for re @xmath126 . the leading order term : @xmath127 is proportional to @xmath128 in the thermodynamic limit , and the canonical fluctuations in the interacting gas are anomalous , similarly to the ideal gas .
the prefactor of the anomalous term is exactly twice smaller than for the ideal gas , which can be attributed to the coupling of @xmath66 and @xmath129 modes in the bogoliubov hamiltonian @xcite .
we note that our result agree in the leading order with the result of ref .
@xcite , calculated in the framework of the standard , particle - number - nonconserving theory .
it is interesting to investigate the behavior of fluctuations in the two limits : @xmath120 and @xmath130 . for the assumed range of temperatures , the former
can be realized in the system with sufficiently small interactions .
the latter corresponds to the regime where the thermal excitations occur mainly in the phonon part of the energy spectrum . in the limit @xmath120 , we perform taylor expansion of functions @xmath131 and @xmath132 for small @xmath133 , and obtain @xmath134 t^2 + \frac{\pi^2}{6 } ( 2\alpha)^{\frac{3}{2 } } \nonumber \\ \label{dnecnintht } & { } + \zeta\left({{\textstyle}\frac{1}{2}}\right ) ( \pi t)^{\frac{3}{2 } } + \frac{\alpha}{4}\zeta\left({{\textstyle}\frac{3}{2}}\right ) ( \pi t)^{\frac{1}{2}}.\end{aligned}\ ] ] in the regime @xmath130 , we perform asymptotic expansion of the functions @xmath131 and @xmath132 for large @xmath133 @xcite , and obtain the following result @xmath135 this result can be also derived , by assuming phonon dispersion relation for the energy spectrum : @xmath136 , with the sound velocity @xmath137 . in this case
spectral zeta function takes form @xmath138 the last equation can be obtained directly from eq .
( [ defz1 ] ) , by approximating each term of the series by its asymptotic behavior for @xmath139 .
[ fig : dneint ] presents the fluctuations of the number of noncondensed atoms in the canonical ensemble calculated for @xmath140 .
the inset shows the same quantity for the system with stronger interactions : @xmath141 .
the highest temperature presented in the plot , corresponds to @xmath142 for @xmath38 .
[ fig : dneint ] compares the analytical result of eq .
( [ dnecnint3 ] ) , with its low and high temperature expansions : ( [ dnecnintlt ] ) and ( [ dnecnintht ] ) , respectively .
we observe that the low temperature expansion correctly describes the fluctuations in the whole regime of considered temperatures .
this reflects the fact that the main contribution to the fluctuations , comes from the phonon part of the spectrum . on the other hand ,
the high temperature expansion is valid only for temperatures @xmath143 .
fluctuations of the number of noncondensed atoms in the canonical ensemble for interacting bose gas confined in a box with periodic boundary conditions .
plots are shown for @xmath144 and @xmath145 ( inset ) , where @xmath146 .
the analytical result of eq .
( [ dnecnint3 ] ) is compared with its low and high temperature expansions : eqs .
( [ dnecnintlt ] ) and ( [ dnecnintht ] ) , respectively . the temperature is scaled in the units of @xmath22.,width=321 ]
we now turn to the the microcanonical ensemble .
fluctuations of the number of noncondensed atoms can be calculated from the following generating function @xmath147 where delta operator @xmath148 selects the states with the total energy equal to @xmath149 , and the trace is taken over all the eigenstates with a fixed total number of particles . taking into account the definition of @xmath150 , the mean number and fluctuations in the microcanonical ensemble can be evaluated from @xmath151 according to eqs .
( [ nemcint ] ) and ( [ dnemcint ] ) , the fluctuations of the number of noncondensed particles can be expressed as @xmath152 where derivative of the mean number of noncondensed particles should be taken before setting @xmath83 in eq .
( [ nemcint ] ) for @xmath153 . in a similar manner
, one can represent the fluctuations of the number of noncondensed particles in the canonical ensemble @xmath154 now we are ready to establish the relation between fluctuations in the microcanonical and canonical ensemble . in the following
we show that identity ( [ relfl ] ) , derived previously for an ideal gas , holds also in the case of an interacting gas .
we basically repeat the steps of derivation for the ideal gas .
first , we assume that the mean number of noncondensed particles in both considered ensembles coincide .
this must be true in the thermodynamic limit , where we expect that both ensembles are equivalent with respect to the basic thermodynamic quantities . according to this assumption
, we denote the mean number of noncondensed particles by @xmath155 , not distinguishing between microcanonical and canonical ensembles .
we choose energy @xmath149 and fugacity @xmath9 as a pair of state variables and take the total differential of @xmath155 @xmath156 from the total differential @xmath157 it is straightforward to derive the following relation @xmath158 according to eq .
( [ dnecnrel ] ) , l.h.s . of eq .
( [ rel1 ] ) at @xmath83 represents the canonical fluctuations of the number of noncondensed atoms .
the r.h.s . of ( [ rel1 ] ) can be rewritten in the following way @xmath159 now , if we observe that @xmath160 and @xmath161 we arrive at the desired relation stated by eq .
( [ relfl ] ) .
thus , we have expressed the microcanonical fluctuations in terms of the quantities calculated in the canonical ensemble . moreover , the particle - energy correlation @xmath162 and fluctuations of the total energy @xmath163 that enter this relation , according to eqs .
( [ dnede ] ) and ( [ d2e1 ] ) can be determined directly from the mean number of noncondensed particles @xmath155 and the mean energy @xmath149 .
we calculate the latter quantities in a similar manner to the canonical fluctuations , representing them in terms of contour integrals containing spectral zeta function , and summing the contributions from all the poles of the integrand .
this yields @xmath164 and @xmath165 8 \pi^{\frac{7}{2 } } t^{\frac{5}{2 } } + \frac{e_0}{\delta } - { { \textstyle}\frac{4 \pi}{15 } } ( 2\alpha)^{\frac{3}{2 } } - t,\end{aligned}\ ] ] where @xmath166 , and @xmath95 is the ground - state energy .
taking into account the leading order terms in eqs .
( [ neintu ] ) and ( [ ecnu ] ) , after some algebra we obtain the following result for the difference between the canonical and the microcanonical fluctuations @xmath167 since @xmath168 , and @xmath169 remains constant in the thermodynamic limit , the difference between fluctuations scales normally with the number of particles .
thus , the microcanonical fluctuations , similarly to the canonical ones , exhibit anomalous behavior . furthermore
, the relative difference between @xmath170 and @xmath171 tends to zero in the thermodynamic limit , and fluctuations in both ensembles become equal for large @xmath0 .
let us investigate the behavior of microcanonical fluctuations in the two limits : @xmath120 and @xmath130 . to this end
, we can expand eq .
( [ dndeoverd2e ] ) for large and small values of @xmath169 . in the regime @xmath172 , the leading behavior of fluctuations in the canonical and microcanonical ensemble
is governed by @xmath173 t^2 + \zeta({{\textstyle}\frac{1}{2 } } ) ( \pi t)^{\frac{3}{2 } } , \\
\label{dnemclt } \left\langle \delta^2
n_e \right\rangle_{\mathrm}{mc } \simeq & \ , \frac{1}{2 } \left[\zeta_e(2)+\frac{\pi^2}{\sqrt{2 \alpha}}\right ] t^2 \\ & + \pi^{3/2 } \frac{\zeta(\frac{1}{2})^2 - \frac{3}{5}\zeta(\frac{3}{2})}{\zeta(\frac{1}{2 } ) } t^{3/2},\end{aligned}\ ] ] we notice that in this limit , the difference between ensembles is given in the leading order by the ideal gas result ( cf eqs .
( [ dneideal ] ) and ( [ dnemc ] ) ) . in the opposite limit : @xmath174 , we obtain @xmath175 in this regime , the microcanonical fluctuations differ from the canonical ones by the prefactor of the finite size correction term . in order to verify predictions of the analytic results , we performed numerical calculations in the microcanonical ensemble .
the microcanonical partition functions were computed with the help of recurrence algorithms @xcite . from the partition functions ,
we calculate the statistics of elementary excitations , and finally the fluctuations of the number of noncondensed particles from @xmath176 \nonumber \\
\label{fluctmcsum } & { } + 4 \sum_{{\bf k}\neq 0 } u_{\bf k}^2 v_{\bf k}^2 \left ( \langle \hat{n}_{\bf k }
\hat{n}_{\bf -k } \rangle_{\mathrm}{mc } + \langle \hat{n}_{\bf k } \rangle_{\mathrm}{mc } + { { \textstyle}\frac{1}{2 } } \right ) , \end{aligned}\ ] ] where @xmath177 .
this relation can be derived starting from the definition of @xmath150 , and performing differentiation according to eq .
( [ dnemcint ] ) .
the mean number @xmath178 of elementary excitations with energy @xmath80 in the microcanonical ensemble is given by @xmath179 analogous expression holds for the product of the occupations numbers @xmath180 in order to express the final result in terms of the temperature , we calculate the microcanonical temperature @xmath181 from the entropy of the system : @xmath182 . in our case , the entropy is given by @xmath183 , where the r.h.s .
depends on the total number of particles only through the energy spectrum .
[ fig : dneint1 ] presents fluctuations of the number of noncondensed particles in the canonical and microcanonical ensemble , calculated for @xmath140 and @xmath141 ( inset ) .
the highest temperature presented in the plot corresponds to @xmath142 for @xmath38 atoms .
the figure shows the analytic results of eqs .
( [ dnecnint3 ] ) and ( [ dndeoverd2e ] ) for the canonical and microcanonical fluctuations , respectively . they are compared with the canonical fluctuations calculated numerically from eq .
( [ dnecnint1 ] ) and with the microcanonical fluctuations evaluated from the recurrence relations for the microcanonical partition functions .
the analytical predictions for both ensembles agree very well with the numerical calculations .
we note that , the missing low temperature part of the numerical curve for the microcanonical fluctuations is due to large uncertainty in determination of the microcanonical temperature at low energies of the system . on the other hand , we see that in the system of a moderate size ( @xmath184 ) the canonical and microcanonical fluctuations can be clearly distinguished .
fluctuations of the number of noncondensed atoms in the canonical and microcanonical ensemble for a three dimensional box with periodic boundary conditions .
plots are shown for @xmath144 and @xmath145 ( inset ) , where @xmath146 . depicted
are : the analytical result of eqs .
( [ dnecnint3 ] ) and ( [ dndeoverd2e ] ) for the canonical and microcanonical fluctuations , respectively , the canonical fluctuations calculated numerically from eq . ( [ dnecnint1 ] ) , and the microcanonical fluctuations evaluated numerically with the help of recurrence algorithms . the temperature is scaled in the units of @xmath22.,width=321 ]
in this section we calculate the canonical fluctuations of a weakly interacting gas confined in a spherically symmetric harmonic trap of frequency @xmath40 . for a homogeneous system we have demonstrated that the particle - number - conserving approach leads to the same value of fluctuations as the traditional , nonconserving bogoliubov method .
since the trapped gas is more complicated to treat analytically , in the following we apply the standard bogoliubov theory generalized for an inhomogeneous system , limiting our analysis to the leading order behavior in the thermodynamic limit . similarly to the homogeneous case we restrict our considerations to the temperatures much smaller than the critical temperature , neglecting the influence of the thermal depletion on the excitation spectrum .
in addition , we assume the dilute - gas regime , which allows to neglect the effects of the quantum depletion .
furthermore , we assume that the number of atoms is sufficiently large : @xmath185 , where @xmath186 is the harmonic oscillator length , and the condensate can be described in the _ thomas - fermi _ ( tf ) regime .
this is particularly convenient since in this regime the analytic solutions of the bogoliubov - de gennes equations are known @xcite . in the bogoliubov theory the field operator is decomposed into the condensate wave function @xmath187 and the noncondensed part @xmath188 : @xmath189 .
the approximate hamiltonian with neglected third- and fourth - order terms in @xmath188 is diagonalized by the bogoliubov transformation @xmath190 where @xmath191 , @xmath192 are the creation and annihilation operators of the elementary excitations , while the amplitudes @xmath193 , @xmath194 are solutions of the bogoliubov - de gennes equations . in the tf regime , the condensate density has typical `` inverted parabola '' shape with the condensate wave function @xmath195 , which vanishes outside the condensate radius @xmath196 .
the chemical potential is determined by the normalization of @xmath197 and in the tf approximation is given by @xmath198 .
the low - energy collective excitations of the condensate in the tf regime have energies @xcite @xmath199 which is valid for @xmath200 . on the other hand ,
the amplitudes @xmath201 and @xmath202 of those modes are described by @xcite @xmath203^{\pm 1/2 } \nonumber \\ & \times y^l p_{n}^{(l+1/2,0)}(1 - 2y^2 ) y_{lm}(\theta,\phi),\end{aligned}\ ] ] where @xmath204 , @xmath205 , @xmath206 , @xmath207 denote the jacobi polynomial and @xmath208 is the spherical harmonic . in the above formula
@xmath209 is the radial quantum number , @xmath210 and @xmath211 are the quantum numbers of the angular part of the wave function , and @xmath212 when @xmath213 . since we consider only the lowest collective modes with energies @xmath200 , in further calculations we can neglect the amplitude @xmath214 , and use the following approximation : @xmath215 .
we calculate the fluctuations substituting the operator of the number of noncondensed particles @xmath216 into @xmath217 .
this yields @xmath218 where we retain only the leading order term , and @xmath219 . in derivation of eq .
( [ dnetrapint ] ) we assume that elementary excitations are populated according to the grand - canonical statistics : @xmath220 for @xmath221 , @xmath222 , with @xmath223 . this is equivalent with using the mde for an ideal gas , and results from assumption that at sufficiently low temperatures the number of elementary excitations can be considered as unconstrained .
furthermore , in analogy to the homogeneous gas we expect that the main contribution to the fluctuations comes from the low - energy phononlike modes , and in the thermodynamic limit we can approximate @xmath224 by @xmath225 .
the overlap integral @xmath226 is calculated in appendix [ appoverlap ] . substituting the result of integration into eq .
( [ dnetrapint ] ) , we arrive at the final result for the canonical fluctuations @xcite @xmath227 with @xmath228\left(2n+l+\frac 52\right)^{2}}\nonumber \\ & + \frac{\left(\varepsilon_{nl}^2+\frac 12 \right)^2 } { \varepsilon_{nl}^2
\left(2n+l+\frac 12\right)^2 \left(2n+l+\frac 52\right)^2 } \bigg].\end{aligned}\ ] ] in the above equation @xmath229 and @xmath230 for @xmath213 .
the numerical calculation of eq .
( [ a ] ) yields @xmath231 .
we observe that in the thermodynamic limit ( @xmath46 , @xmath47 , and @xmath232 _ const _ ) the fluctuations exhibit anomalous scaling , similarly to the behavior in a homogeneous system .
moreover , the @xmath233 temperature dependence of fluctuations becomes universal , both for the homogeneous and trapped gases . to calculate the microcanonical fluctuations in a trapped gas we employ the thermodynamic relation ( [ relfl ] ) . in section [ sec : fluctmc ] we have shown that this identity applies also to an interacting gas .
the proof is quite general , and in the case of a trapped gas the only difference with respect to the homogeneous system is the form of the operators @xmath234 and @xmath97 , that enter the definition of @xmath84 and @xmath150 . according to eqs .
( [ dnde ] ) and ( [ d2e ] ) , the particle - energy correlation and the fluctuations of the system s energy , can be directly calculated from the mean number of noncondensed particles and the mean energy .
the former quantity can be evaluated as the expectation value of the operator @xmath216 , where in place of @xmath235 we substitute decomposition ( [ bogtransf ] ) .
for @xmath236 we use the approximation @xmath215 , obtaining the following result @xmath237 where @xmath238 stands for the quantum depletion .
since @xmath238 does not depend on the temperature , its exact value is not important for the subsequent derivation .
now , we utilize the results of appendix [ appoverlap ] for the overlap integral between the amplitudes @xmath201 , and replace the summation by integration , which is valid in the thermodynamic limit .
this yields @xcite @xmath239 a similar method can not be applied , however , to calculate the mean energy of a weakly interacting trapped gas . in this case
main contribution comes from the boundary of the condensate , where the excitations have single - particle character .
derivation based on the semiclassical formulation of the hartree - fock - bogoliubov theory , which includes the effects of collective modes inside , and the single - particle excitations outside the condensate , yields the following law @xcite @xmath240 now , with the help of results ( [ netrap ] ) and ( [ etrap ] ) we calculate the difference between the fluctuations in the canonical and the microcanonical ensemble : @xmath241 it is easy to observe that the difference is proportional to @xmath0 in the thermodynamic limit .
therefore the microcanonical fluctuations , similarly to the canonical ones , are anomalous , and in the thermodynamic limit predictions of both ensembles become equivalent .
in this paper we have analyzed the fluctuations of the number of condensed particles in a weakly interacting homogeneous and trapped bose gases . for a homogeneous system
we apply the particle - number - conserving formulation of the bogoliubov theory , and obtain the analytical results for the fluctuations in the canonical and microcanonical ensemble , including the corrections due to the finite size of the system .
our derivation is based on contour integral representation in terms of spectral zeta function .
we determine the poles of zeta functions resulting from the bogoliubov spectrum , and develop their asymptotic expansions .
this allows us to find analytical formulas for the mean number of noncondensed particles , its fluctuations and the mean energy , which are valid for arbitrary ratio of the temperature to the chemical potential .
we determine the microcanonical fluctuations from the thermodynamic identity which relates them to the quantities calculated in the canonical ensemble .
our analysis shows that both the microcanonical and canonical fluctuations exhibit anomalous scaling , and in the thermodynamic limit fluctuations in these two ensembles become equal .
a similar behavior is observed for a trapped gas . in this case
we carry out our calculations within the standard bogoliubov theory , and evaluate only the leading order behavior in the thermodynamic limit .
our results have been derived in the regime of temperatures much lower than the critical temperature , where the effects of the quantum and thermal depletion on the population of the condensate can be neglected .
we stress , however , that generalization of the obtained results for the case of higher temperatures , can be done in straightforward way in the spirit of popov theory @xcite . to this end , for a homogeneous system parameter @xmath114 should be replaced by @xmath242 , with the density of condensed atoms @xmath243 determined in a self - consistent way from the equation for the mean number of noncondensed particles . in a similar way one can generalize the results for the trapped gas . in this case
the total number of particles @xmath0 in equation for @xmath244 should be replaced by the number of condensed atoms @xmath245 , with @xmath245 calculated in a self - consistent a manner .
finally we would like to comment on the application of the particle - number - conserving method to calculate the fluctuations . for a homogeneous system
we have verified that both conserving and nonconserving theories lead to the identical results .
such insensibility of the condensate statistics , can be attributed to the fact , that below the critical temperature , the condensate acts as a reservoir of particles for the noncondensed states .
thus , as long as the temperature is not too close to the critical temperature , the probability distribution of the number of noncondensed atoms remains insensitive to the actual number of atoms .
the total number of atoms influences only the energy spectrum in the interacting gas .
therefore , predictions of conserving and nonconserving theories are the same with respect to the statistics of noncondensed part of the system .
this work was stimulated by interactions with p. zi .
the author is grateful to l. pitaevskii , s. zerbini and s. giorgini for valuable discussions .
financial support from the esf program bec2000 + is acknowledged .
in this appendix we evaluate residues of the spectral zeta function , which appears in the studies of fluctuations in a weakly interacting bose gas : @xmath246 the series in ( [ appdeftheta ] ) is convergent provided that @xmath247 . in the remaining part of the complex plane , @xmath113 is defined through analytic continuation .
we rewrite eq .
( [ appdeftheta ] ) in the following way @xmath248 for the moment we assume that @xmath249 .
later , we show how the following derivation can be generalized for larger values of @xmath114 .
since @xmath250 , in the regime @xmath249 we can apply the binomial expansion @xmath251 substituting this expansion into ( [ apptheta1 ] ) , and interchanging the order of summation , we arrive at the following series representation for @xmath113 @xmath252 where @xmath253 denote epstein zeta function .
since the singularities of @xmath254 in the numerator occur exactly at the same points as the singularities of @xmath255 in the denominator , the only poles in the series ( [ apptheta2 ] ) , are the ones of @xmath253 .
epstein zeta function @xmath256 has only single pole at @xmath257 , with residue equal to @xmath24 .
hence the poles of @xmath113 are located at @xmath258 for @xmath259 , with residues @xmath260 in the case @xmath261 , we can split the series ( [ appdeftheta ] ) into finite sum @xmath262 , containing terms for which @xmath263 , and the series @xmath264 containing rest of the terms @xmath265 the finite sum @xmath262 is analytic on the whole complex @xmath9-plane . for the second term
@xmath264 we apply the binomial expansion , and obtain @xmath266 where @xmath267 since the function @xmath268 is analytic on the whole complex @xmath9-plane , the residues of @xmath113 are given by eq .
( [ restheta ] ) , also in the case @xmath261 .
in this appendix we develop asymptotic expansion of spectral zeta function @xmath113 . to this end
we apply the method described in @xcite .
first , we rewrite ( [ appdeftheta ] ) in the following way @xmath269 making use of the integral representation of gamma function , eq .
( [ app2theta ] ) can be transformed into @xmath270 for the term @xmath271 we use integral representation of the mellin - barnes type @xmath272 with @xmath273 and @xmath274 .
interchanging the order of integration and performing the integral over @xmath275 we arrive at @xmath276 by choosing @xmath277 , the sum over @xmath278 is absolutely convergent and may be expressed in terms of epstein zeta function @xmath279 for large @xmath114 the function under integral rapidly decay for @xmath280 as long as @xmath281 is not too large .
when @xmath282 becomes comparable to @xmath114 the integrand starts to grow , and finally becomes infinite for @xmath283 .
therefore , we close the contour of integration in ( [ app2theta3 ] ) on the left side , but include only a few leading order poles in the calculation of the integral . if we choose @xmath284 , then the integrand has poles at @xmath285 ( pole of the epstein zeta function ) and at @xmath286 for @xmath259 ( poles of @xmath287 ) . for @xmath288 with @xmath209
all the poles are of the order of one and the asymptotic expansion reads @xmath289
in this appendix we calculate the canonical fluctuations from the integral representation @xmath290 where spectral zeta function @xmath118 is given by @xmath291 the poles of the integrand give the following contributions : _ pole of @xmath292 at @xmath45 . _
its residue is equal to 1 , and this pole gives the contribution @xmath293 .
using asymptotic expansion of @xmath113 , we obtain @xmath294 _ pole of @xmath295 at @xmath296 . _
its residue is equal to 1 , and this pole gives the contribution @xmath297 . applying asymptotic expansion of @xmath113 for large @xmath114 we extract the leading order behavior @xmath298 _ poles of @xmath295 at @xmath299 , for @xmath209 .
_ their residues are equal to @xmath300 .
since @xmath301 , these poles do not give any contribution .
_ pole of @xmath295 at @xmath302 . _
its residue is equal to @xmath303 and this pole gives the contribution @xmath304 . with the help of the asymptotic expansion we find that this term is of the order of @xmath305 , and
therefore may be neglected in the final result .
_ poles of @xmath295 at @xmath306 for @xmath209 .
_ their residues are equal to @xmath307 , and the contribution from these poles is @xmath308 .
however , @xmath309 , as can be deduced either from the binomial expansion ( [ apptheta2 ] ) or the asymptotic expansion ( [ app2asympt ] ) . _ poles of @xmath113 at @xmath310 for @xmath209 . _
their residues are given by ( [ restheta ] ) .
the contribution , which we denote by @xmath311 , from these poles is @xmath312 after some algebra we rewrite ( [ contrc1 ] ) in the following way @xmath313 assuming @xmath314 , the summation can be done analytically with the help of summation formula @xcite @xmath315 valid for @xmath316 .
hence , the final result is @xmath317 .\end{aligned}\ ] ] since the function @xmath311 remains analytic for @xmath318 , validity of the result ( [ contrc1b ] ) can be extended for @xmath318 by analytic continuation .
_ poles of @xmath319 at @xmath320 for @xmath209 . _
their residues are given by ( [ restheta ] ) .
the contribution @xmath321 , from these poles is @xmath322 the sum can be rewritten in the following way @xmath323 the summation can be simplified with the help of the following identity , which can be derived from ( [ sumzeta ] ) @xmath324 where @xmath325 assuming @xmath314 , the final result for @xmath321 reads @xmath326 since the function @xmath327 remains analytic for @xmath328 , validity of the result ( [ contrc2b ] ) can be extended for @xmath318 by analytic continuation .
finally we calculate the quantum fluctuations @xmath329 . expressing @xmath330 in terms of @xmath113 and applying asymptotic expansion ( [ app2asympt ] )
, we obtain @xmath331 summing up the contributions of eqs .
( [ contr1 ] ) , ( [ contr2 ] ) , ( [ contrc1b ] ) , ( [ contrc2b ] ) , and ( [ d2nq ] ) we obtain the canonical fluctuations ( [ dnecnint3 ] ) .
in this appendix we calculate the integral @xmath333 . for @xmath334 ,
we apply the approximation @xmath335 where @xmath336 is defined by eq .
( [ fdef ] ) . performing the integration over the angular part of the function @xmath337
, we arrive at @xmath338 where @xmath339 to perform the latter integral , first we express @xmath207 in terms of @xmath340 and @xmath341 utilizing recurrence relations for the jacobi polynomials , and then perform the resulting integrals with the help of @xcite .
the final result reads @xmath342 for asymptotic expansions of the hurwitz zeta function see for example : e. elizalde , s.d .
odintsov , a. romeo , a.a .
bytsenko , and s. zerbini , _ zeta regularization techniques with applications _ ( world scientific , singapore 1994 ) .
asymptotic expansion of @xmath132 can be found by means of the technique described in appendix [ appasympt ] . | we study fluctuations of the number of bose condensed atoms in weakly interacting homogeneous and trapped gases . for a homogeneous system
we apply the particle - number - conserving formulation of the bogoliubov theory and calculate the condensate fluctuations within the canonical and the microcanonical ensembles .
we demonstrate that , at least in the low - temperature regime , predictions of the particle - number - conserving and traditional , nonconserving theory are identical , and lead to the anomalous scaling of fluctuations .
furthermore , the microcanonical fluctuations differ from the canonical ones by a quantity which scales normally in the number of particles , thus predictions of both ensembles are equivalent in the thermodynamic limit .
we observe a similar behavior for a weakly interacting gas in a harmonic trap .
this is in contrast to the trapped , ideal gas , where microcanonical and canonical fluctuations are different in the thermodynamic limit . |
one of the observational arguments supporting inflation is the evidence of nearly scale - invariant primordial adiabatic fluctuations with a mostly gaussian spectrum , as shown by cobe @xcite and wmap @xcite . in its simplest realizations , adiabatic cosmological fluctuations
are simply related to linear fluctuations of the inflaton generated during the accelerated stage : such fluctuations have a gaussian spectrum . non - linear fluctuations inevitably carry a @xmath0-squared distribution , which may contribute to non - gaussianities in the 3-d power spectrum and in cmb anisotropies depending on the amplitude and spectrum of these non - linearities ( see @xcite for a review ) .
a study of fluctuations beyond the linear level is therefore interesting not only for the issue of the stability of inflationary dynamics , but also because of its connection with the predictions of non - gaussianities in the matter power spectrum and in cmb anisotropies . even in absence of non - linearity in the potential for the scalar field ,
the generality of intrinsic non - gaussianities is guaranteed by the non - linearity of the einstein equations , a fact which has motivated a great interest both from the theoretical and observational point of view @xcite .
our choice of calculating the intrinsic non - gaussianities for a free massive inflationary model is therefore a case of primary interest , since in such a case the whole non - linear sector is given by gravity .
in this paper we shall focus on the perturbative evualuation of _ second order gauge - invariant _ ( gi henceforth ) curvature fluctuations . for this purpose
we discuss different second order gi measures of curvature perturbations ( see also @xcite ) , which to first order coincide with the curvature on flat slicing hypersurfaces .
we then compute the correlation function for differents second order gi variables , finding , for all scales of interest , a growth in the spectrum amplitude as the total number of e - folds which inflation lasted increases . a dependence on the total number of e - folds
was also found for the production of test scalar fields with a mass smaller than that of the inflaton @xcite .
the paper is organized as follows .
we present the formalism and different gauge - invariant measures of curvature perturbations in section ii . in sections iii and
iv we evaluate the quantum correlators for different second order gi variables with both analytical and numerical methods .
we conclude in section v and in appendices a e b we add useful technical formulae for gauge transformations and for the analytical results given in section iii .
let us consider inflation in a flat universe driven by a classical minimally coupled scalar field with a potential @xmath1 and follow the notation in @xcite .
the action is given by : s d^4x l = d^4x [ action ] where @xmath2 is the lagrangian density and we shall study the chaotic model @xmath3 @xcite ( see @xcite for a second order calculation in a different inflationary model ) .
we shall consider , up to second order in the fluctuations , both the inflaton : @xmath4 and the metric of a flat universe ( @xmath5 ) : @xmath6 \ , , \label{metric_second}\end{aligned}\ ] ] where @xmath7 . as is clear from the above we limit ourselves to the scalar sector of metric perturbations , neglecting vector and tensor perturbations . in the above formulation
the gauge is not fixed , and one can eliminate two scalar functions among the four metric coefficients and the inflaton fluctuation .
we wish to extend to second order the gi potential for the curvature perturbation . for this purpose ,
let us note that to first order the so called curvature ( potential ) perturbation is given by = + ^2 e [ cp_order1 ] and this quantity is not gi ( for the general gauge trasformation see appendix a ) .
we now consider the comoving slicing @xcite , which is defined to be the slicing orthogonal to the worldlines of comoving observers : they are free - falling and the expansion defined by them is isotropic .
that is the observers do not measure any flux of energy ( @xmath8 ) and for the universe under consideration this corresponds to having @xmath9 .
the transformation from a general slicing to a comoving slicing with @xmath9 is : _
com= - ^0_(1 ) = 0 ^0_(1 ) = [ trasf_inf_ord_uno_comoving ] and one obtains r^(1)=_com=++^2 e , [ comovingrord1 ] which is the first order gi comoving curvature ( potential ) perturbation .
to second order the situation is more involved due to the presence of terms quadratic in the first order .
we are interested in the definition of gi quantities including second order .
their construction is not unique and we shall therefore construct three different gi second order curvature ( potential ) perturbations . this variable is obtained by repeating consistently up to second order the procedure used for the first order comoving curvature ( potential ) perturbation . starting from ^(2)=
^(2)+ ^2 e^(2 ) , [ cp_order2_linear ] to second order the comoving slicing is caracterized by @xmath11 and @xmath12 , that is @xmath9 and @xmath13 . in this case , in order to fix all the degrees of freedom , we restrict ourselves to considering only an infinitesimal time trasformation up to second order , which implies taking @xmath14 and @xmath15 equal to zero . from the first order we obtain the condition in eq.([trasf_inf_ord_uno_comoving ] ) , while from the second order we have ^(2)_com= ^(2 ) - ^0_(1 ) + = 0 ^0_(2 ) = . [ trasf_inf_ord_due_comoving ]
thus one obtains @xmath16- \frac{1}{6 a}\frac{1}{\dot{\phi}}\left[\partial^i\beta \partial_i\varphi-\frac{1}{2}\nabla^2\left(\beta \varphi \right)\right]+ \frac{1}{4 } \frac{\partial^i \partial^j}{\nabla^2}\left[-2 h \frac{\varphi}{\dot{\phi } } d_{ij } e \right . \nonumber \\ & & \left . - \frac{\varphi}{\dot{\phi } } d_{ij } \dot{e } + \frac{1}{a^2 } \frac{\varphi}{\dot{\phi}^2 } d_{ij } \varphi - \frac{1}{2a } \frac{1}{\dot{\phi } } \left ( \varphi d_{ij } \beta + \beta d_{ij } \varphi
\right ) \right]\ , .
\label{comovingrord2}\end{aligned}\ ] ] let us note that this quantity is analogous to the second order curvature perturbation on a uniform density hypersurface defined in @xcite .
these two quantities agree ( apart from a sign ) in the long wavelength limit ( see also @xcite ) .
we give the second order gi curvature perturbation potential obtained from a form analogous to eq .
( [ comovingrord1 ] ) ( the first three terms of the following equation ) , and add a minimal set of terms necessary to obtain a second order gi variable . using the general transformations of appendix
a , we obtain @xmath18 \right\}\ , . \label{r_second_gen}\end{aligned}\ ] ] this quantity is the true gi generalization of @xmath19 computed in @xcite , which also takes into account infinitesimal spatial ( scalar ) transformation in addition to the infinitesimal time translations considered in @xcite . in order to introduce the third curvature ( potential ) perturbation ,
let us first consider the intrinsic 3-curvature @xmath21 associated with a foliation of space - time with constant proper time @xmath22 for a flat universe . using eq.(49 ) and eq.(58 ) of @xcite we obtain : @xmath23 \label{def3rconc}\end{aligned}\ ] ] with @xmath24 in our notations the intrinsic curvature is : @xmath25 instead of using eq .
( [ cp_order2_linear ] ) , we now consider the potential @xmath26 of the intrinsic curvature to second order @xmath27 , given by @xmath28 .
we then obtain the comoving gi expression by going from a general slicing to a comoving slicing : @xmath29- \frac{1}{6 a}\frac{1}{\dot{\phi}}\left[\partial^i\beta \partial_i\varphi-\frac{1}{2}\nabla^2\left(\beta \varphi \right)\right]+ \frac{1}{4 } \frac{\partial^i \partial^j}{\nabla^2}\left[-2 h \frac{\varphi}{\dot{\phi } } d_{ij } e \right .
\nonumber \\ & & \left .
- \frac{\varphi}{\dot{\phi } } d_{ij } \dot{e } + \frac{1}{a^2 } \frac{\varphi}{\dot{\phi}^2 } d_{ij } \varphi - \frac{1}{2a } \frac{1}{\dot{\phi } } \left ( \varphi d_{ij } \beta + \beta d_{ij } \varphi
\right ) \right ] + \frac{1}{\nabla^2 } \left [ 4 \left ( \psi+\frac{1}{6 } \nabla^2 e+\frac{h}{\dot{\phi}}\varphi \right ) \right . \nonumber \\ & & \left .
\nabla^2 \left ( \psi + \frac{1}{6 } \nabla^2 e + \frac{h}{\dot{\phi } } \varphi \right)+ \frac{3}{2 } \partial^i \left ( \psi+\frac{1}{6 } \nabla^2 e + \frac{h}{\dot{\phi}}\varphi\right)\partial_i \left ( \psi+\frac{1}{6 } \nabla^2 e+\frac{h}{\dot{\phi}}\varphi \right ) \right .
\nonumber \\ & & \left .
-\frac{1}{2 } \nabla^2 e \nabla^2 \left ( \psi+\frac{1}{6 } \nabla^2 e + \frac{h}{\dot{\phi}}\varphi\right ) -\frac{1}{2}\partial^i\partial^j e \partial_i\partial_j \left ( \psi+\frac{1}{6 } \nabla^2 e+\frac{h}{\dot{\phi}}\varphi \right ) \right . \nonumber \\ & & \left . -\frac{1}{2 } \partial^i \left ( \psi+\frac{1}{6 } \nabla^2 e+\frac{h}{\dot{\phi}}\varphi \right)\partial_i \nabla^2 e -\frac{1}{16 } \partial^i\nabla^2 e
\partial_i \nabla^2 e \right . \nonumber \\ & & \left .
+ \frac{1}{16}\partial^i\partial^j\partial^k e \partial_i\partial_j\partial_k e\right ] \label{comovingrord2full}\end{aligned}\ ] ] the preceding three expressions undergo a considerable simplification in the long - wavelength limit . for large scales
we obtain : r^(2)_a=^(2 ) + ^(2)- ( + 2 h ) - + ( -h^2 -+ ) [ comovingrord2_lw_limit ] r^(2)_b = ^(2 ) + ^(2 ) + ( 2 h^2 + -h ) ^-1 ( + 2 h + ) ^2 [ r_second_gen_lw_limit ] @xmath30 \label{comovingrord2full_lw_limit}\end{aligned}\ ] ] in this limit the difference between @xmath31 and @xmath32 is given by @xcite : r^(2)_b - r^(2)_a= .
[ differencer ] let us note that , as for the first order @xmath33 , @xmath31 also is constant in time in this limit @xcite .
further on using different parametrizations of the metric many other gauge invariant quantities can be constructed .
for example from @xcite one can define another variable @xmath34 , which is related to @xmath31 on large scales by r^(2)_a = r^(2)_lr- ( r^(1))^2 .
[ r_lr ] we also consider @xmath35 since it is directly related to a geometrical quantity : the intrinsic 3-curvature .
as in one of our previous papers on back - reaction @xcite , we choose to work in the uniform curvature gauge ( ucg ) , which can be generalized straightforwardly to second order through the conditions : @xmath36 and @xmath37 . in this gauge ,
the evolution equation for inflaton fluctuations is regular also during the coherent oscillation of the scalar field @xcite . in the ucg and in the long wavelenght limit
we have , for the given second order curvature perturbation potentials , the following results : r^(2)_a=^(2)- + ^2 ( -1 -+ ) [ comovingrord2_lw_limit_ucg ] r^(2)_b = ^(2 ) + ( 2 h^2 + -h ) ^-1 ^2 [ r_second_gen_lw_limit_ucg ] r^(2)_c = ^(2)- + ^2 ( -1 -+ ) + . [ comovingrord2full_lw_limit_ucg ] on subtracting the average of the quadratic piece ( in order to also have a zero average for the non - linear piece ) , it is useful to introduce the parameter of nonlinearity @xmath38 @xcite : r = r_l+ f_nl[formnlr ] where @xmath39 is the gaussian part of @xmath40 ( that is the first order part @xmath33 ) . using the first order comoving curvature perturbation together with the first two second order definitions given in eqs .
( [ comovingrord2_lw_limit_ucg ] ) , ( [ r_second_gen_lw_limit_ucg ] ) , we obtain for a @xmath41 chaotic inflation and using the approximation @xmath42 and eq.([relation_ord1_ord2 ] ) , to leading order in the slow - roll parameter @xmath43 , the following results for @xmath38 : @xmath44 @xmath45 let us note that we can not evaluate @xmath46 in analogy with @xcite because the non - local terms are not subleading in the long - wavelength approximation . in the next two sections we proceed with the evaluation of the spectrum of the non - linear corrections to the curvature @xmath10 and of the second order gi field fluctuation @xmath47 defined in @xcite . the evaluation of the quantum correlator of second - order gi variables ( see also @xcite ) involves a sum over momenta , which is plagued by ultraviolet divergencies .
the adiabatic subtraction used in the calculation of the finite part of energy and pressure of fluctuations @xcite is roughly equivalent to only considering super - hubble fluctuations .
motivated by this analogy we shall also use this criterium in the present case : all the integrals which follow have the physical hubble radius as an upper limit of integration .
the same integrals also need a lower limit of integration which defines the region of nearly scale invariant ( red tilted ) spectra generated by inflation : we choose this lower limit of integration , a comoving momentum @xmath48 , of order of the hubble rate when inflation begins , @xmath49 , as previously found in analytical @xcite and numerical investigation @xcite ( see also @xcite ) . at the present state of our knowledge ( i.e. without any theoretical input for the choice of @xmath48 ) ,
we believe that this is a natural choice as compared to considering @xmath48 of the order of the present comoving hubble radius as is done in @xcite . from now on we shall restrict ourselves to the slow - roll stage of @xmath41 chaotic inflation , during which @xmath50 holds . within the above approximation from eq.([comovingrord2_lw_limit_ucg ] ) , to leading order in the slow - roll parameter and using eq.(28 ) of our @xcite , one obtains ^(2)~r^(1 ) 2 [ andamento_varphi2 ] and r^(2)_a = r^(1 ) 2 . [ r2a_leading ] from the above equation and eq .
( [ r_lr ] ) it is evident that @xmath51 is of order a slow - roll parameter times the square of the first - order perturbation .
let us consider the quantum correlation function for this second order gauge invariant variable @xmath52 the quantized scalar field variable @xmath53 is given by ( t , * x * ) = d*k * [ quantumfourier_varphi ] in the slow - roll approximation and on large scales ( but @xmath54 ) we consider @xcite _ k ( t)=- , [ varphi_infrared ] with h(t_k ) h_0 ( 1 - 2 _ 0 ) ^1/2 [ hcrossing ] which is the hubble parameter when the fluctuations crosses the horizon and @xmath55 .
the power spectrum of first order curvature perturbation , at large scales and in the ucg , is therefore : p_r^(1 ) ( k ) |r^(1)_k|^2 = |_k ( t)|^2 = .
[ 1storderspectrum ] it is important to note that by using the expression ( [ hcrossing ] ) for @xmath56 we obtain a power spectrum of curvature perturbations in ( [ 1storderspectrum ] ) having the correct value to leading order of the spectral index and exhibiting running during slow - roll .
the physical content of the correlation function for @xmath31 is conveniently described by the power spectrum associated with its fourier transform 0 |r^(2)_a(*x * ) r^(2)_a(*y*)| 0 = d^3 k ^i * k * ( * x*-*y * ) |r^(2)_ak(t)|^2 [ corr_ra ] which becomes , using eq.([cf_r2a ] ) @xmath57 where we have considered only modes which are super - hubble , but greater then comoving infrared cut - off @xmath58 , using the heaviside functions @xmath59 .
we can perform a fourier transform with respect to @xmath60 of eq.([corr_r2a_varphi ] ) and obtain @xmath61 where _ 0= d^3 k_1 d^3 k_2 ( a h - k_1 ) ( a h - k_2 ) ( k_1-l ) ( k_2-l ) .
[ q_0_r2a ] for @xmath62 and on evaluating the integral in eq.([trasf_fourier_r2a ] ) to leading order and in the momentum window given by the condition @xmath63 one obtains the expression : |r^(2)_ap(t)|^2 \ { _ i=0 ^ 3 a_i ( ) ^i + g()+h ( ) } [ r2a_value ] which is time independent in our approximation .
the coefficients @xmath64 are given in appendix b , @xmath65 and @xmath66 .
substituting for the coefficients @xmath64 we obtain the following result : @xmath67 on taking only the leading term for the coefficents of the different powers in @xmath68 , we obtain : @xmath69 where in the last equality we have used eq.([1storderspectrum ] ) .
this last expression shows compactly one of the main results of our investigation : the spectrum of this second order gi curvature perturbation is proportional to the first order spectrum through logarithmic corrections which encode the scale @xmath58 at which inflation started . without these logarithmic corrections
the second order curvature perturbation would be completely local and much smaller than first order terms since for the scales of interest @xmath70 .
let us consider the second order gi scalar field fluctuation @xmath47 ( which is a possible second order generalization of the first order mukhanov variable @xcite ) defined in @xcite as the second order field fluctuations on uniform curvature hypersurfaces in the longwavelength limit .
the uniform curvature hypersurfaces is defined as the slicing with @xmath71 , and one obtains the following result @xcite : q^(2)=^(2)+ ^(2)+ + ( 1 + - ) ^2 + .
[ q_malik ] in the ucg , as for the first order , this gi variable is simply the second order field fluctuations @xmath72 , studied in our previous paper @xcite .
thus the equation of motion for this second order gi field fluctuation in the ucg , to leading order in the slow - roll parameter and in the long - wavelength approximation is given by @xcite : ^(2 ) + 3 h ^(2)+ 3 h q^(2 ) = m^2 .
[ eq_q_scalata_second_leading_order_approximate ] it is interesting to comment on the self - interaction mediated by gravity corrected for the inflaton . just as the feedback of metric perturbation vanishes in first order perturbation theory for @xmath73 ( see eq . ( 7 ) of @xcite ) , it also vanishes for the self - interaction to second order in eq .
( [ eq_q_scalata_second_leading_order_approximate ] ) . integrating the above in two steps
we obtain : ^(2)+ 3 h q^(2 ) = m^2 dt , [ eq_q_scalata_second_leading_order_approximate_int ] and q^(2 ) ( t , * x * ) = ^t dt a(t)^3 ^t dt m^2 .
[ eq_q_scalata_green_function ] let us consider the quantum correlation function of this second order @xmath47 gauge invariant variable @xmath74 .
the quantized scalar variable @xmath47 is given by @xmath75\ , .
\label{quantumfourier_q}\end{aligned}\ ] ] the physical content of the correlation function for @xmath47 is better described by the power spectrum associated with its fourier transform 0 |q^(2)(t , * x * ) q^(2)(t , * y*)| 0 = d^3 k ^i * k * ( * x*-*y * ) |q^(2)_k ( t)|^2 [ corr_q ] and using eq.([eq_q_scalata_green_function ] ) and eq.([varphi_infrared ] ) , one obtains @xmath76 ^ 2 \theta(k_1-l ) \theta(k_2-l ) \exp^{i ( { \bf k_1}+{\bf k_2 } ) \cdot ( { \bf x}-{\bf y } ) } \right . \nonumber \\ & & \,\,\,\,\ , \,\,\,\,\,\left .
+ \int d^3 k_1 d^3 k_2 \frac{h(t_{k_1})^4}{k_1 ^ 3}\frac{h(t_{k_2})^4}{k_2 ^ 3}\frac{1}{4 } \left [ \int^t dt ' a(t')^3 \int^{t ' } dt '' \frac{1}{h(t'')^3 } \theta(a(t '' ) h(t'')-k_1 ) \right ] \right .
\nonumber \\ & & \,\,\,\,\ , \,\,\,\,\,\left . \left
[ \int^t dt ' a(t')^3 \int^{t ' } dt ''
\frac{1}{h(t'')^3 } \theta(a(t '' ) h(t'')-k_2 ) \right ] \theta(k_1-l ) \theta(k_2-l ) \right\}\ , .
\label{corr_q_varphi_green}\end{aligned}\ ] ] on performing the fourier transformation w.r.t .
@xmath60 of eq.([corr_q_varphi_green ] ) one finds @xmath77 ^ 2
\theta(k - l ) \theta(|{\bf
p}-{\bf k}|-l)+ ( 2 \pi)^3 \delta^{(3)}(p ) q_0 \label{trasf_fourier_green}\end{aligned}\ ] ] where @xmath78 \nonumber \\ & & \left [ \int^t ds ' a(s')^3 \int^{s ' } ds '' \frac{1}{h(s'')^3 } \theta(a(s '' ) h(s'')-k_2 ) \right ] \theta(k_1-l ) \theta(k_2-l ) \ , . \label{q_0_green}\end{aligned}\ ] ] on considering @xmath62 and replacing the @xmath59 function by the domain of integration , it is possible to perform the time integrals exactly : @xmath79 \right .
\nonumber \\ & & \left .
-\left(-\frac{3}{2 \dot{h}}\right)^{3/2 } \left [ \gamma\left(-\frac{1}{2 } , -\frac{3}{2}\frac{h(t)^2}{\dot{h}}\right)- \gamma\left(-\frac{1}{2 } , -\frac{3}{2 } \frac{f(k , p , y)^2}{\dot{h}}\right)\right ] \right\}^2 \ , , \label{trasf_fourier_green_numerical}\end{aligned}\ ] ] where @xmath80 is given by @xmath81 \,.\ ] ] it is also possible to obtain an explicit approximate espression for the integral starting from eq.([trasf_fourier_green ] ) subject to the condition @xmath63 and the approximation @xmath82 .
in particular we find @xmath83 -\left(-\frac{3}{2}\frac{h_0 ^ 2}{\dot{h}}\right)^{3/2 } \left[\gamma\left(-\frac{1}{2},-\frac{3}{2}\frac{h^2}{\dot{h}}\right ) -\gamma\left(-\frac{1}{2},-\frac{3}{2}\frac{h_0 ^ 2}{\dot{h}}\right)\right ] \right\}^2 \label{q2_value_green}\end{aligned}\ ] ] where , as in the previous subsection , the coefficients @xmath64 are discussed in the appendix , @xmath65 and @xmath66 .
let us also investigate a much more crude approximation , obtained on neglecting @xmath84 in eq.([eq_q_scalata_second_leading_order_approximate_int ] ) and considering the infrared limit @xmath85 , which leads to the expression q^(2 ) ^2 = r^(1)2 .
[ relation_ord1_ord2 ] the correlation function for @xmath47 turn simplifies to @xmath86 so , on proceeding as the previous subsection , for @xmath62 , to the leading order and in an `` approximate '' fashion with the condition @xmath63 one obtains |q^(2)_p(t)|^2()^2 \ { _ i=0 ^ 3 a_i ( ) ^i + g()+h ( ) } .
[ q2_value ] substituting the coefficients @xmath64 we obtain the following result : @xmath87 on taking only the leading term for the coefficents of the different powers in @xmath68 , we obtain : @xmath88 analogous considerations to those given after eq.([r2a_value_developed_lead ] ) are also valid in this case .
also , as we can see from eq .
( [ q2_value_developed ] ) , one obtains a spectrum which , according to eqs .
( [ r2a_leading ] ) and ( [ relation_ord1_ord2 ] ) , slowly increases in time during slow - roll as @xmath89 , whereas @xmath31 is constant .
again , we note that both the gi second order field fluctuation @xmath47 and the second order curvature @xmath31 have a nearly scale invariant spectrum with a logarithmic corrections .
let us introduce a few parameters which are useful for the numerical analysis .
the first is @xmath90 , the total number of e - folds , thus we need to specify the time associated with the end of inflation .
this definition may be obtained on requiring the hubble parameter @xmath91 to be equal , for example , to the inflaton mass @xmath92 .
this happens , in our chaotic model for the time @xmath93 when the slow - roll parameter @xmath94 is equal to @xmath95 .
we then solve the equation ( ) = [ eq_general ] obtaining = -- and the final time , defined by @xmath96 , is t_f=-- corresponding to a total number of e - folds given by n_tot= - .
it is also useful to define the variable @xmath97 , which is equal to the number of e - folds from the end of inflation , n(t)== - . in our analysis
we are interested in the fluctuations that cross the horizon at @xmath98 e - folds from the end of inflation .
the time @xmath99 associated with @xmath100 is : t _ * = -- and the comoving wave - number which crosses the hubble radius at @xmath101 is : k _ * = a(t _ * ) h(t _ * ) = e^-- -n _ * . if we wish to take a `` photograph '' of a fluctuation with @xmath102 on super - hubble scales during inflation we should take : > > .
[ cond_photo ] we now proceed with the numerical analysis taking @xmath103 . in particular
we are interested in the comparison of second order fluctuations with first order ones , in order to understand for which condition first order perturbation theory breaks down .
we shall also plot the shape , in @xmath104 , of the spectrum of fluctuations . in figure @xmath105
we consider , for two different values of the initial hubble parameter @xmath49 ( @xmath106 and @xmath107 , respectively ) , the logarithm of the second order contribution to the power spectrum of @xmath47 : @xmath108 ) and the numerical solution of eq.([trasf_fourier_green_numerical ] ) , in order to see the improvement of the roughest analytical approximation , with respect to the exact numerical calculation with increasing @xmath49 , and the shape of the spectrum .
we give the figure at an instant during inflation for which @xmath109 .
corresponding to those two values of @xmath49 we have the following data * for @xmath110 we have @xmath111 , @xmath112 , @xmath113 and @xmath114 .
the time at which we `` photograph '' the inflation is @xmath115 . * for @xmath116 we have @xmath117 , @xmath118 , @xmath119 and @xmath120 .
the time at which we `` photograph '' the inflation is @xmath121 .
( 0.2,0.9)@xmath122 ( -6.5,4)@xmath123 ( -8.3,0.3)@xmath122 ( -14.9,4.1)@xmath123 in order to compare the @xmath124 given by the numerical solution of eq.([trasf_fourier_green_numerical ] ) with the first order contribution to the power spectrum @xmath125 we compare the values calculated at @xmath102 as a function of the total number of e - folds @xmath126 , at the end of inflation @xmath127 . from the figure on the right ( which differs from the figure on the left only because of the bigger range of @xmath126 ) we see that second order effects are of the same order as first order effects for @xmath126 near @xmath128 which corresponds to a @xmath49 near @xmath129 .
therefore for such values first order perturbation theory breaks down ( -7.8,3.6)@xmath126 ( -13.5,4.3)@xmath130 ( -13.5,2)@xmath131 ( 0.1,3.6)@xmath126 ( -5.5,4.3)@xmath130 ( -5.5,2)@xmath131 in figure @xmath132 we show the dependence of the fluctuations on the modes , crossing the horizon n e - folds before the end of inflation , as seen at the end of inflation .
we consider three different cases corresponding to values of @xmath133 equal to 7 , 14 and 30 , compare @xmath134 and @xmath135 , and vary @xmath136 from @xmath137 to @xmath105 with : k_n = e^-- -n [ knvalue ] which is the mode of the fluctuation that crosses the horizon at @xmath136 e - folds before the end of inflation . as before the first and second order fluctuations are of the same order of magnitude for @xmath138 .
( -1,4.6)@xmath139 ( -9.5,4.5)@xmath140 ( -9.5,2)@xmath141 let us proceed in a analogous way for the second order curvature perturbation ( potential ) @xmath31 . as before in figure @xmath142 we shall consider two different cases for inflation with @xmath49 equal to 7 m and 14 m and we shall plot @xmath143 ) , to exhibit the shape of the spectrum .
we give the picture an instant during inflation for which @xmath109 .
( 0.2,0.3)@xmath122 ( -6.5,4)@xmath144 ( -8.3,2.30)@xmath122 ( -14.9,4)@xmath144 for the sake of comparing @xmath144 with the first order contribution to the power spectrum @xmath145 ) , we give two different kinds of figures . in figure
@xmath146 we compare those values calculated at @xmath102 , as a function of the total number of e - folds @xmath126 , at the end of inflation @xmath93 . from the figure on the right ( which again differs from the figure on the left only because of the larger range of @xmath126 ) we see that for this case the second order effect is of the same order as the first order effect for @xmath126 near @xmath147 which corresponds to a @xmath49 near @xmath148 .
therefore first order perturbation theory breaks down for a value of @xmath49 smaller than that obtained before and we have a stronger limit .
( -8,3.5)@xmath126 ( -13.5,4.1)@xmath149 ( -13.5,2.2)@xmath150 ( 0,3.5)@xmath126 ( -5.5,4.1)@xmath149 ( -5.2,2)@xmath150 in figure @xmath151 we again consider three different value of @xmath133 , namely @xmath152 , @xmath153 and @xmath154 , compare @xmath155 and @xmath156 , and vary @xmath136 from @xmath137 to @xmath105 at the end of inflation . as before the first and second orders will be of the same order of magnitude if we take @xmath157 .
( -1,4.7)@xmath139 ( -8.7,4.5)@xmath155 ( -8.7,2.5)@xmath156 finally we wish to compare the critical initial values of @xmath49 which we obtain in this paper , with two different limiting values .
the first one is the value @xmath158 , obtained by requiring that back - reaction in the ucg gauge will become important by the end of inflation @xcite , when @xmath159 .
such a value is : h_br ~(16 ^2 ) ^1/6 m^2/3 m_pl^1/3 .
[ hbr ] the second one is the value obtained by the self - reproduction argument @xcite h_sr ( 8 ) ^3/4()^1/2 m^1/2 m_pl^1/2 .
[ hsr ] this comparison is given in table i. .comparison of the values of @xmath49 for which first order perturbation theory break down with @xmath158 and @xmath160 . [
cols="^,<",options="header " , ] we stress that neither of the two values of @xmath49 or of @xmath158 are related to the self - reproduction scale @xcite .
the values of @xmath49 obtained by considering the breakdown of linear perturbation theory in this paper are of the same order of magnitude as @xmath158 ( see also @xcite for an analogous value ) .
this fact is not surprising : both these calculations include second - order cosmological perturbations .
the results for back - reaction were considered ambiguous since it is not simple to demonstrate the slow - down of the inflationary expansion in terms of a gi quantity . on the other hand in the present paper ,
the breakdown of first order perturbation theory is found for gi variables in the long - wavelength limit .
the results summarized in the table i and derived from figures 2 , 3 , 5 and 6 can also be obtained directly from eq.([r2a_value_developed_lead ] ) and eq.([q2_value_developed_lead ] ) . in those cases and for @xmath161 on using eq.([knvalue ] ) the leading logarithm values ( that is the term with @xmath162 ) of @xmath163 and @xmath164 are given by : @xmath165 and the first and second orders are comparable for h_0 m ( ) ^1/3 [ h0r ] and h_0 m ( ) ^1/3 [ h0q ] respectively , where eq.([h0r ] ) nearly give the same value as table i ( namely near @xmath148 ) , while eq.([h0q ] ) only gives the same order of magnitude ( actually @xmath166 ) , which is not surprising considering the approximations made .
we have studied second order cosmological perturbations for a chaotic @xmath41 inflationary model , and considered three second order gi measures of curvature perturbations . for the gi curvature perturbation studied and for the second order gi scalar field flcutuation presented in @xcite , we have found , for all scales of interest ( see , for example , eq.([r2a_value_developed_leadfinale ] ) and eq.([q2_value_developed_leadfinale ] ) ) , that the amplitude of the spectrum of quadratic corrections grows with the total number of e - folds @xmath167 . for @xmath47
we also obtain a mild time dependence associated with the growth of the slow - roll parameter @xmath94 .
since both the quantities studied here are gi , this dependence on @xmath167 can not be a gauge artifact .
the spectrum of those second order gi variables obtained by fourier trasforming the curvature - curvature quantum correlation function is nearly scale - invariant , with additional logarithmic corrections with respect to the first order spectrum .
on comparing the first and the second order contributions for the two different cases considered , one finds two different values for the initial hubble parameter @xmath49 beyond which first order perturbation theory breaks down .
we found limits which are of the same order of magnitude as that one found by back - reaction in the ucg @xcite .
neither the back - reaction limit nor those found here have anything to do with the self - reproduction scale . in the curvature - curvature quantum correlation function , the cross terms involving first and
third order perturbation may also contribute .
the third order curvature perturbation is clearly beyond the scope of the present paper and will be the subject of future work .
unless such a term cancels or dominates the growth we have found , our results imply that in single field models the amplitude of the spectrum of intrinsic non - gaussianities generated during the slow - roll stage will depend on the total number of e - folds which inflation has lasted .
we wish to thank r. abramo , s. matarrese , a. a. starobinsky for discussions and k. a. malik for useful correspondence .
the gauge in our formulation ( see eq.([metric_second ] ) ) is not fixed , since one can eliminate two scalars among the metric and the inflaton degrees of freedom . after a general analysis
we shall be interested in making a gauge choice in order to match previously studied situations .
an infinitesimal coordinate trasformation to second order is @xcite : x^^= x^+ ^_(1 ) + ( ^_(1 ) , ^_(1 ) + ^_(2 ) ) , [ trasf_gauge_coord ] where @xmath168 and @xmath169 are the coordinate changes to first an second order , respectively , and @xmath170 and @xmath171 .
it induces the following change in a geometric object @xmath172 : t^(1 ) ^(1)= t^(1 ) - _ _ ( 1 ) t^(0 ) [ trasf_tens_ord1 ] t^(2 ) ^(2)= t^(2 ) - _ _ ( 1 ) t^(1 ) + ( _ _ ( 1)^2 t_0 - _ _ ( 2 ) t_0 ) [ trasf_tens_ord2 ] which leads to the following general gauge trasformation for our scalar perturbations .
to first order we have : = - ^0_(1 ) [ trasf_inf_ord_uno ] = - ^0_(1 ) = - ^0_(1 ) + 2 a ^s_(1 ) = + h ^0_(1)+ ^2^s_(1 ) = e -2^s_(1 ) . [ e_transformation ]
to second order one finds : ^(2)= ^(2 ) - ^0_(1 ) + -_k ^k^s_(1 ) , [ trasf_inf_ord_due ] ^(2 ) = ^(2)- ^0_(1 ) -2 ^0_(1)-^0_(2)+ ^0_(1)^0_(1 ) + ( ^0_(1))^2 ^(2)= ^(2 ) - _ ( 2)^0 + a ^s_(2 ) + ( _ ( 1)^0)^2 + @xmath173 @xmath174 \ , . \label{e2_transformation}\end{aligned}\ ] ]
the coefficients @xmath64 are given by : @xmath175 where @xmath176 so one obtains a. kogut _ et al .
_ , astrophys .
j. letters * 464 * ( 1996 ) l29 e. komatsu _ et al .
_ , astrophys .
j. suppl .
* 148 * ( 2003 ) 119 d. n. spergel _
et al . _ , astro - ph/0603449 n. bartolo , e. komatsu , s. matarrese and a. riotto , phys .
rept . * 402 * ( 2004 ) 103 .
v. acquaviva , n. bartolo , s. matarrese and a. riotto , nucl .
b * 667 * ( 2003 ) 119 h. noh and j. hwang , _ phys .
rev . _ * d 69 * , 104011 ( 2004 ) .
k. a. malik and d. wands , class .
* 21 * ( 2004 ) l65 .
d. h. lyth , k. a. malik and m. sasaki , jcap * 0505 * ( 2005 ) 004 . f. vernizzi , phys .
d * 71 * ( 2005 ) 061301 d. langlois and f. vernizzi , phys .
* 95 * ( 2005 ) 091303 .
d. h. lyth and y. rodrguez _ phys .
rev . _ * d 71 * , 123508 ( 2005 ) .
f. finelli , g. marozzi , g. p. vacca and g. venturi , _ phys .
* d 65 * , 103521 ( 2002 ) .
f. finelli , g. marozzi , g. p. vacca , and g. venturi , _ phys . rev . _ * d 69 * , 123508 ( 2004 ) .
a. d. linde , phys .
b * 132 * ( 1983 ) , 317 .
b. losic and w.g .
unruh , _ phys .
rev . _ * d 72 * , 123510 ( 2005 ) .
a. riotto , hep - ph/0210162 .
h. kodama and t. hamazaki , prog .
* 96 * ( 1996 ) 949 ; y. nambu and a. taruya , prog .
* 97 * ( 1997 ) 83 ; f. finelli and r. h. brandenberger , phys .
* 82 * ( 1999 ) 1362 k. a. malik , jcap * 0511 * ( 2005 ) 005 .
s. weinberg , phys .
d * 72 * ( 2005 ) 043514 . v. f. mukhanov , jetp lett . *
41 * , 493 ( 1985 ) ; sov .
jetp * 68 * 1297 ( 1988 ) .
f. finelli , g. marozzi , g. p. vacca and g. venturi , _ energy - momentum tensor of quantum fluctuations in chaotic @xmath180 inflation _ , in proceedings of sorrento 2003 , `` thinking , observing and mining the universe '' .
s. goncharov , a. d. linde , and v. f. mukhanov , int .
phys . a 2 , 561 ( 1987 ) ; a. d. linde , _ particle physics and inflationary cosmology _ ( harwood academic , chur , switzerland , 1990 ) . | the evolution of gauge invariant second - order scalar perturbations in a general single field inflationary scenario are presented .
different second order gauge invariant expressions for the curvature are considered .
we evaluate _
perturbatively _ one of these second order curvature fluctuations and a second order gauge invariant scalar field fluctuation during the slow - roll stage of a massive chaotic inflationary scenario , taking into account the deviation from a pure de sitter evolution and considering only the contribution of super - hubble perturbations in mode - mode coupling .
the spectra resulting from their contribution to the second order quantum correlation function are nearly scale - invariant , with additional logarithmic corrections with respect to the first order spectrum .
for all scales of interest the amplitude of these spectra depends on the total number of e - folds .
we find , on comparing first and second order perturbation results , an upper limit to the total number of e - folds beyond which the two orders are comparable . |
the superconducting transition in conventional superconductors is rather well described by mean field theories , essentially because in the coherence volume @xmath9 a large number of pairs is present . on the contrary , in cuprate superconductors the small coherence length @xmath10 , the reduced carrier density , the marked anisotropy and the high transition temperature @xmath6
strongly enhance the superconducting fluctuations ( sf ) . in a wide temperature range , which can extend up to @xmath11 or @xmath12 k above @xmath6
, a variety of phenomena related to sf @xcite can be detected by several experiments such as specific heat @xcite , thermal expansion @xcite , penetration depth @xcite , conductivity @xcite and magnetization @xcite measurements .
also nmr - nqr relaxation has been used to detect sf @xcite .
recently the field dependence of the nmr relaxation rate @xmath13 and of the knight shift have been studied , by varying the magnetic field from @xmath14 up to @xmath15 tesla @xcite .
mitrovic et al .
@xcite attributed the field dependence of @xmath16 to the corrections in the density of states ( dos ) contribution to the spin - lattice relaxation . on the other hand , gorny et al
@xcite have observed @xmath16 field independent , up to 14 tesla , from 150 k to 90 k , in a sample showing a decrease in @xmath17 starting around @xmath18 k. an analysis of the role of sf in nmr experiments and of the field dependence of @xmath13 has been recently carried out by eschrig et al .
@xcite , by extending the analitical approach of randeria and varlamov @xcite to include short wave - length and dynamical fluctuations . of particular interest
is the role of sf in underdoped high temperature superconductors .
in fact , in these compounds one has the peculiar phenomena of the opening , at @xmath19 , of a spin gap ( as evidenced by nmr and neutron scattering ) and of a pseudo - gap ( as indicated by transport and arpes experiments ) @xcite .
these gaps , which are possibly connected @xcite , have been variously related to sf of particular character @xcite . in this paper magnetization and @xmath0cu(2 ) nmr - nqr relaxation measurements in underdoped ybco
are reported and compared with the ones in the optimally doped compound .
the magnetization data yield information on the fluctuating diamagnetism ( fd ) , while the nuclear relaxation rates @xmath13 convey insights on the effects of sf on the @xmath20-integrated spin susceptibility in the zero frequency limit .
the paper is organized as follows . in sect .
ii some basic equations describing sf and fd are recalled . after some experimental details ( sect .
ii ) , in sect . iii the experimental findings and their analysis are presented , with emphasis on the effect of the field on @xmath13 and on the behavior of the diamagnetic magnetization in chain - ordered underdoped ybco .
the main results and conclusions are summarized in sect .
because of sf the number of cooper pairs per unit volume , which is given by the average value of the square of the modulus of the order parameter @xmath21 , is different from zero above @xmath6 . in the time - dependent ginzburg - landau ( gl ) description @xcite the collective amplitudes and the correspondent decay times of sf are given by @xmath22 where @xmath23 and @xmath24 , with @xmath25 .
@xmath26 plays the role of the fourier components of the average number of pairs per unit volume , while @xmath27 is the correspondent relaxation time
. fluctuating pairs can give rise to a diamagnetic magnetization above @xmath6 .
the diamagnetic magnetization @xmath7 in general is not linear in the field @xmath28 .
the curves @xmath7 vs. @xmath28 and @xmath7 vs. @xmath29 can be tentatively analized by generalizing the lawrence - doniach model , using the free energy functional @xcite @xmath30=\sum_n\int d{\bf r}\biggl [ a\vert \psi_n\vert^2 + { b\over 2}\vert \psi_n\vert^4 + \nonumber \\ + { \hbar^2\over 4 m_{\parallel}}\vert[{\bf \nabla_{\parallel}}- { 2ie\over\hbar c}{\bf a_{\parallel } } ] \psi_n\vert^2 + t \vert \psi_{n+1}- \psi_n\vert^2 \biggr]\end{aligned}\ ] ] where the last term takes into account the tunneling coupling between adjacent layers . from eq .
2 , deriving the free energy in the presence of the field and by means of a numerical derivation ( @xmath31 ) , one can obtain the fluctuating magnetization . according to scaling arguments @xcite for moderate anisotropy ( quasi-3d case )
one expects that the magnetization curves , at constant field , cross at @xmath8 when the magnetization is scaled by @xmath32 .
the amplitude @xmath33 at @xmath34 departs from the prange s result @xcite by a factor @xmath35 , corresponding to the anisotropy ratio @xmath36 .
accordingly @xcite , the data at different fields collapse onto an universal curve when @xmath37 is reported as a function of @xmath38 , for a critical exponent for the coherence length corresponding to @xmath39 . for strong anisotropies , namely quasi-2d systems ,
the @xmath7 curves at constant field cross each other at @xmath40 .
@xmath41 is larger than the value obtained from the gaussian approximation by a factor @xcite around @xmath14 .
collapse of the data onto a universal curve occurs for @xmath42 well above @xmath43 , usually in the range @xmath44 .
the contributions to the relaxation rates w due to sf can be derived within a fermi liquid scenario , withouth specifying the nature of the interactions @xcite .
the direct and most singular contribution , equivalent to the aslamazov - larkin paraconductivity , is not effective as nuclear relaxation mechanism .
the positive maki - thompson ( mt ) contribution @xmath45 results from a purely quantum process , involving pairing of the electron with itself at a previous stage of motion , along intersecting trajectories .
a negative sf contribution @xmath46 comes from the density of states reduction when electrons are subtracted to create pairs .
approximate expressions for @xmath46 and @xmath45 can be derived by resorting to simple physical arguments as follows .
the relaxation rates can be approximated in the form @xmath47 where @xmath48 is an average form factor for cu nuclei and @xmath49 an effective spectral density @xcite .
sf modify the static spin susceptibility @xmath50 and the effective correlation time @xmath51 in eq .
3 . from eq.1
the in - plane density of pairs is @xmath52 and therefore the sf imply @xmath53\end{aligned}\ ] ] for the dynamical part in eq .
3 , the dos contribution can be obtained by averaging over the bz the collective correlation time in eq . 1 : @xmath54 in the dirty limit , differing from the exact calculation @xcite only by a small numerical factor .
@xmath55 is the electron collision time .
the mt contribution has to be evaluated in the framework of diagrammatic theories @xcite .
the final result corresponds to the 2d - average over the bz of a decay rate of diffusive character @xmath56 ( with @xmath57 carrier diffusion constant ) , phenomenologically accounting for phase breaking processes by adding in the decay rate a frequency @xmath58 : @xmath59 where @xmath60 is a dimensionless pair breaking parameter . by indicating with @xmath61 the relaxation rate in the absence of sf and by neglecting the correction to @xmath62 ( eq .
4 ) , from eqs . 3 , 5 and 6
the relaxation rate in 2d systems turns out @xmath63 \label{ram}\end{aligned}\ ] ] to extend this equation to a layered system , when dimensionality crossover ( 2d@xmath64 3d ) occurs , one has to substitute @xcite @xmath65 by @xmath66 $ ] and @xmath67 by @xmath68 $ ] , with @xmath69 anisotropy parameter ( @xmath70 is the interlayer distance ) .
it is noted that the mt contribution ( first term in eq .
7 ) is present only for s - wave orbital pairing , while it is averaged to almost zero for d - symmetry .
the measurements have been carried out in oriented powders of optimally doped ybco in one chain - disordered underdoped and in two chain - ordered underdoped ybco samples @xcite the oxygen content in the underdoped compounds was close to 6.66 , with sligth differences in @xmath6 .
electron diffraction microscopy evidenced the expected tripling of the a - axis , while resistivity measurements show a sharp transition with zero resistivity at t= 62 k and occurrence of paraconductivity below about 75 k @xcite .
table i collects the main properties of the samples , as obtained from a combination of measurements . the @xmath0cu relaxation rates @xmath71 have been measured by standard pulse techniques . in nqr the recovery towards the equilibrium after the saturation of the @xmath72 line is well described by an exponential law , directly yielding @xmath73 . for the nmr relaxation
a good alignment of the @xmath74 axis of the grains along the magnetic field is crucial to extract reliable values of w ( h ) .
the nmr satellite line , corresponding to the @xmath75 transition can be used to adjust the alignment and to monitor the spread in the orientation of the c - axis , the resonance frequency being shifted at the first order by the term @xmath76 , due to the quadrupole interaction ( @xmath77 angle of the @xmath78-axis with the field ) . from the width of the spectrum ( fig .
1a ) the spread in the orientation of the c - axis appears within about two degrees .
the recovery law for the nmr satellite transition is @xmath79 this law has been checked to fit well the experimental data ( fig .
1b ) , proving the lack of background contamination due to other resonance lines .
the magnetization has been measured by means of a metronique ingegnerie squid magnetometer , on decreasing the temperature at constant field and , at selected temperatures , by varying @xmath28 .
the paramagnetic contribution to m was obtained from m vs @xmath28 at t @xmath80 110 k , where practically no fluctuating magnetization is present .
then @xmath7 was derived by subtraction , singling out a small contribution from paramagnetic impurities .
the paramagnetic susceptibility turns out little temperature dependent around @xmath6 and this dependence will be neglected in discussing the much stronger diamagnetic term . in optimally doped ybco the measurements
have been carried out at the purpose to confirm the results recently obtained in single crystals by other authors@xcite .
the isothermal magnetization curves are satisfactorily described on the basis of the anisotropic gl functional ( eq .
2 ) , for @xmath81 . only close to @xmath6 and for @xmath82 tesla
an observable departure is detected , indicating crossover to a region of non - gaussian sf , in agreement with recent thermal expansion measurements @xcite .
the @xmath7 vs. @xmath29 curves cross at @xmath40 when @xmath7 is scaled by @xmath83 , as expected for moderate anisotropy @xcite .
let us first comment the data in optimally doped ybco ( fig .
the comparison between the nmr and nqr @xmath13 s has been already discussed in previous papers @xcite .
here we only add a few comments motivated by more recents works @xcite involving the remarkable aspect of the field dependence .
the nqr @xmath13 can be compared to eq .
7 , by using the values @xcite @xmath84 , @xmath85 and a dephasing time parameter @xmath86 .
a quantitative fitting is inhibited by the fact that the background contribution to w does not follow the korringa law , in view of correlation effects among carriers @xcite . a firm deduction , however , is that an anisotropy parameter @xmath87 , and namely a crossover to a 3d regime , is required to avoid unrealistic values ( 2d line in figure 2a ) , as indicated also by magnetization measurements ( see later on ) .
mitrovic et al .
@xcite have discussed the field dependence of @xmath88 in terms of quenching of the dos term only , by resorting to the theory of eschrig et al . @xcite .
the field dependence of the sf contribution to the relaxation rate is a delicate issue , because of the nontrivial interplay of many parameters which include reduced temperature @xmath89 , reduced field @xmath90 , anisotropy parameter @xmath91 , elastic and phase breaking times @xmath55 and @xmath92 .
eschrig et al .
@xcite have extended the analitical approach @xcite by taking into account arbitrary values of @xmath93 , short - wave fluctuations and dynamical fluctuations .
the price of this generalization is the restriction to a 2d spectrum of sf fluctuations . in view of our experimental findings
the analysis of the field effect in a purely 2d framework is questionable . on the other hand ,
a re - examination of the field effect for a layered system is now possible by applying the method of the dos term regularization devised by budzin and varlamov @xcite , which has indicated how the divergences can be treated .
since the dynamical fluctuations could be relevant only for fields of the order of @xmath94 , up to @xmath95 @xmath96 the fluctuations can be safely treated in the nearly static limit .
the field dependence for weak fields ( @xmath97 ) turns out @xcite @xmath98^{3/2}}\beta^2\end{aligned}\ ] ] where @xmath99 + \psi ^{\prime } ( 1/2)}= \bigg\ { \begin{tabular}{cc } $ ] 14(3)/^3@xmath100t1@xmath1014t@xmath1021t1/@xmath103 ( here @xmath104 ) .
so both effects of increase and of decrease of @xmath13 on increasing the magnetic field are possible , depending on the mean free path in the specific sample .
if @xmath105 the main correction to @xmath13 is due to the mt term and one should observe @xmath13 decreasing with increasing field , while if @xmath106 the dos correction becomes dominant and @xmath13 is expected to increase . as it appears from fig .
2 the effect of the field for @xmath107 is small , if any . it is noted that if eq . 7 is applied to the underdoped compounds the reduction in the fermi energy @xmath108 and the increase in the anisotropy ( i.e. decrease of @xmath91 ) would imply a sizeable increase in @xmath109 .
this enhancement in the underdoped compounds does not occur ( see fig . 2b and 2c for the chain - disordered and chain - ordered ybco s , respectively ) , unless the decrease of @xmath13 over a wide temperature range , usually related to the spin - gap opening @xcite , should be attributed to a field - independent dos term . on the other hand ,
the conventional sf of gl character should occur only close to @xmath6 , where , on the contrary , @xmath50 is little temperature dependent .
eq . 9 , in principle ,
does predict a field dependence .
however , if typical values @xmath110 k , @xmath111 , @xmath112 and @xmath113 are used , for @xmath114 one has @xmath115 , hardly to evidence for @xmath116 .
a comprehensive discussion of the field dependence of the sf contribution to the relaxation rate is given elsewhere @xcite .
here we only remark that the results in strong fields ( @xmath117 ) from mitrovic et al .
@xcite can hardly be justified on the basis of eq .
as regards the magnetization measurements in optimally doped ybco , as already mentioned , our data indicate a 3d regime crossing from gaussian to critical fluctuations close to @xmath6 .
the value of @xmath118 @xcite at the crossing point of the curves @xmath119 vs. @xmath29 turns out around @xmath120 , corresponding to an anisotropy factor @xmath121 .
a collapse onto a common function is obtained when @xmath122 is plotted as a function of @xmath123 .
the collapse fails in magnetic fields less than about 0.3 tesla .
a small - field departure from the universal function has been already observed in underdoped compounds @xcite , but no mention about it is found in the literature for optimally doped ybco . in the underdoped compounds a marked enhancement of the fluctuating magnetization is observed ( figs . 3 and 4 ) . in fig .
4 the susceptibilty , defined as @xmath124 for @xmath125 tesla , is reported .
the inset is a blow up of the results around the temperature where reversing of the sign of the magnetization occurs .
the susceptibilities for chain - ordered and chain - disordered underdoped samples are compared with the one in optimally doped ybco in fig .
5 . in underdoped ybco
the magnetization curves dramatically depart from the ones expected on the basis of eq . 2 and numerical derivative .
no crossing point is observed in the curves @xmath7 vs @xmath29 and no collapse is found on anisotropic 3d or 2d - like curves for @xmath126 or @xmath7 vs @xmath127 ( fig .
these anomalies are more manifested in the chain ordered compound , as it can be realized from the peculiar shape of the isothermal magnetization vs. @xmath28 ( fig .
one could remark that a fraction of a few percent of non - percolating local superconducting regions can account for the absolute value of the diamagnetic susceptibility and for the shape of the magnetization curves .
a trivial chemical inhomogeneity could be suspected .
the fact that the anomalous diamagnetism has been observed in three samples differently prepared and since it is strongly enhanced in the chain - ordered ones , supports the intrinsic origin of the phenomenon .
recent theories @xcite have discussed how a distribution of local superconducting temperatures , related to charge inhomogeneities , can cause anomalous diamagnetic effects .
mesoscopic charge inhomogeneities have been predicted on the basis of various theoretical approaches @xcite .
in underdoped compounds one could expect the occurrence of preformed cooper pairs causing `` local '' superconductivity lacking of long - range phase coherence .
an inhomogeneous distribution of carriers , on a mesoscopic scale , is supported by @xmath128 plane optical conductivity @xcite and it is the basic ingredient of the stripes model @xcite .
since the anomalous diamagnetism is enhanced in chain - ordered underdoped ybco one is tempted to relate it to the presence of stripes .
the insurgence of phase coherence among adjacent charge - rich regions can qualitatively be expected to yield strong screening above the bulk superconducting temperature @xcite .
therefore , one could phenomenologically describe the role of sf in underdoped compounds , as follows . at @xmath129 , just below @xmath130 , fluctuating pairs are formed , without long - range phase coherence .
below @xmath131 , about @xmath132 k above @xmath6 , phase coherence among adjacent charge - rich drops @xcite start to develop , yielding the formation of superconducting loops with strong screening and causing the onset of the anomalous diamagnetism .
close to @xmath6 , sf of gl character occur and long - range phase coherence sets in .
by combining magnetization and nmr - nqr relaxation measurements , an attempt has been done to clarify the role of superconducting fluctuations and of fluctuating diamagnetism in underdoped ybco , vis - a - vis the optimally doped compound . in the latter case
the fluctuations above @xmath6 are rather well described by an anisotropic ginzburg - landau ( gl ) functional and by scaling arguments for slightly anisotropic systems .
a breakdown of the gaussian approximation for small magnetic fields has been observed close to @xmath6 .
the @xmath0cu relaxation rates @xmath13 around @xmath6 show little field dependence , if any .
one can not rule out the presence of a mt contribution , and then of a small @xmath133wave component in the spectrum of the fluctuations , which should be sample - dependent in view of the role of impurities in the pair - breaking mechanisms .
in underdoped ybco an anomalous diamagnetism is observed , on a large temperature range .
the diamagnetic susceptibility at @xmath6 is about an order of magnitude larger than the one in the optimally doped sample and the isothermal magnetization curves can not be described by the anisotropic gl functional .
scaling arguments , such as the search of an universal function in terms of @xmath134 , appear inadequate to justify the experimental findings .
the anomalies are more marked in the chain - ordered samples .
also in the underdoped ybco @xmath0cu @xmath13 in nqr almost coincide with the nmr ones at @xmath135 tesla , in agreement with a theoretical estimate of the effect of the magnetic field .
another conclusion drawn from the field and temperature dependences of the @xmath0cu relaxation rate is that the spin gap does not depend on the magnetic field and that the behavior of the spin susceptibility for @xmath129 can not be ascribed to sf of gl character .
the fluctuating diamagnetism observed in underdoped ybco can be phenomenologically accounted for by the presence of charge inhomogeneities at mesoscopic level .
the anomalies in the magnetization curves can be attributed to `` locally superconducting '' non - percolating drops , present above the bulk @xmath6 and favoured by chain ordering .
a. bianconi , f. borsa , j. r. cooper and m. h. julien are thanked for useful discussions .
the collaboration of paola mosconi in processing the squid magnetization data is gratefully acknowledged .
thanks are due to b.j .
suh and p. manca , for having provided well characterized underdoped ybco compounds .
the availability of the metronique ingegnerie squid magnetometer from the departement of chemistry , university of florence ( prof .
d. gatteschi ) is gratefully acknowledged .
the research has been carried out in the framework of a `` progetto di ricerca avanzata ( pra ) '' , sponsored by istituto nazionale di fisica della materia ( pra - spis 1998 - 2000 ) . for a recent collection of papers , see fluctuations phenomena in high temperature superconductors , edited by m. ausloos and a.a .
varlamov , kluwer academic publisher , ( 1997 ) a.a .
varlamov , g.balestrino , e. milani and d.v .
livanov , advances in physics 48 , 655 ( 1999 ) m.thinkham `` introduction to superconductivity '' , mc graw - hill n.y .
( 1996 ) chapters 8 - 9 see a.junod , in `` studies of high temperature superconductors '' , edited by a.v .
narlikar ( nova science publishers , inc .
new york , 1996 ) , vol .
18 v.pasler et al .
phys . rev .
81 , 1094 ( 1998 ) z.h .
lin et al .
letters 32 , 573 ( 1995 ) and references therein m.a .
howson et al .
phys . rev .
b 51 , 11984 ( 1995 ) and references therein see the review paper by l.n.bulaevskii , int . j. of modern physics b4 , 1849 ( 1990 ) and several papers in ref . 1 p.carretta , d. livanov , a.rigamonti and a.a.varlamov , phys.rev.b 54 , r9682 ( 1996 ) p.carretta , a.rigamonti , a.a.varlamov , d.livanov , nuovo cimento 19 , 1131 ( 1997 ) v. mitrovic et al . , phys .
82 , 2784 ( 1999 ) h.n.bachman et al .
b 60 , 7591 ( 1999 ) k. gorny et al . , phys . rev .
82 , 177 ( 1999 ) m. eschrig et al .
b 59 , 12095 ( 1999 ) m.randeira and a.a.varlamov , phys
b 50 , 10401 ( 1994 ) a.rigamonti , f.borsa and p.carretta , rep.prog.phys.61 , 1367 ( 1998 ) a.g .
loeser et al .
science 273 , 325 ( 1996 ) h.ding et al .
nature 382 , 51 ( 1996 ) g. v. m. williams , j. l. tallon , r. michalak and r. dupree , phys .
b 54 , r6909 ( 1996 ) t. dahm , d. manske and l. tewordt , phys .
b 55 , 15274 ( 1997 ) v.j.emery , s.a.kivelson and o.zachar phys.rev .
b 56 , 6120 ( 1997 ) ; see also v. j. emery and s. a. kivelson , j. phys .
59 , 1705 ( 1998 ) c. castellani , c. di castro and m. grilli z.phys .
b 103 , 137 ( 1997 ) c.baraduc , a. budzin , j - y henry , j.p .
brison and l. puech , physica c 248 138 ( 1995 ) a.budzin and v.dorin , ref . 1 , p.335 m. b. salamon et al .
b 47 , 5520 ( 1993 ) t.schneider and h. keller , inter
. jornal of modern physics b 8 , 487 ( 1993 ) a.junod , j.y .
genoud , g. triscone and t. schneider , physica c 294 , 115 ( 1998 ) r.e .
prange , phys .
b 1 , 2349 ( 1970 ) m.a.hubbard , m.b .
salamon and b.w.veal , physica c 259 , 309 ( 1996 ) c.p.slichter `` principles of magnetic resonance '' , springer verlag , berlin - heidelberg ( 1990 ) p.manca , p.sirigu , g.castellani and a.migliori , il nuovo cimento 19 , 1009 ( 1997 ) ; g.castellani , a.migliori , p.manca and p.sirigu , il nuovo cimento 19 , 1075 ( 1997 ) u.welp et al . , phys.rev.lett .
67 , 3180 ( 1991 ) a.i.buzdin and a.a.varlamov , phys.rev.b 58 , 14195 ( 1998 ) p.mosconi , a.rigamonti and a. a. varlamov , appl . mag .
reson . to be published yu .
n. ovchinikov , s. a. wolf and v. z. kresin , phys .
b 60 , 4329 ( 1999 ) e. z. kuchinskii and m. v. sadovskii , cond - mat/9910261 d.mihailovic , t.mertelj and k.a.muller , phys . rev .
b 57 , 6116 ( 1998 ) ; see also kabanov et al .
phys.rev.b 59 , 1497 ( 1999 ) m. r. norman , cond - mat/9904048 a.bianconi et al . phys.rev.lett .
76 , 3412 ( 1996 ) c. bergemann et al .
, phys . rev .
b 57 , 14387 ( 1998 ) | magnetization and @xmath0cu nmr - nqr relaxation measurements are used to study the superconducting fluctuations in yba@xmath1cu@xmath2o@xmath3 ( ybco ) oriented powders . in optimally doped ybco
the fluctuating negative magnetization @xmath4 is rather well described by an anisotropic ginzburg - landau ( gl ) functional and the curves @xmath5 cross at @xmath6 . in underdoped ybco , instead , over a wide temperature range an anomalous diamagnetism is observed , stronger than in the optimally doped compound by about an order of magnitude .
the field and temperature dependences of @xmath7 can not be described either by an anisotropic gl functional or on the basis of scaling arguments .
the anomalous diamagnetism is more pronounced in samples with a defined order in the cu(1)o chains .
the @xmath0cu(2 ) relaxation rate shows little , if any , field dependence in the vicinity of the transition temperature @xmath8 .
it is argued how the results in the underdoped compounds can be accounted for by the presence of charge inhomogeneities , favoured by chains ordering . |
the coulomb blockade effect in circuits with small - capacitance tunnel junctions provides the means of manipulating single charge quanta ( see , for example , the review paper by averin and likharev @xcite ) .
if periodic signals of frequency @xmath2 are applied to the gate electrodes of the circuit , a train of single charges @xmath3 can move across an array of junctions so that charge pumping , giving rise to the current @xmath4 , is achieved .
it was experimentally proven that the normal - state metallic circuits enable single electrons ( @xmath5 ) to be effectively pumped at frequencies @xmath2 of about several mhz @xcite . moreover
, the accuracy of single - electron pumping can nowadays meet the requirements of fundamental metrology , @xmath6 @xmath7 @xcite .
unlike pumping of electrons , the pumping of cooper pairs ( @xmath8 ) in superconducting circuits has not been that successful so far .
the only experiment had been carried out in 1991 by geerligs _
@xcite with a three - junction al sample .
although this experiment did evidence a pumping of the pairs , the pumping was strongly disturbed by several factors : the landau - zener transitions , cooper pair co - tunneling , quasiparticle tunneling , etc . as a result ,
the shape of the current plateau at @xmath9 in the i - v curve was far from being perfect .
recently , in their theoretical paper pekola _
_ @xcite concluded that pumping of cooper pairs with reasonable accuracy in a three - junction array was impossible . for practical values of parameters ( the ratio of the josephson coupling energy to the charging energy @xmath10 ) they evaluated the inaccuracy of pumping @xmath11 to be as much as 9%-63% .
this is because of intensive co - tunneling of pairs in the short arrays , i.e. the process of tunneling simultaneously across several ( @xmath12 ) junctions . to suppress the co - tunneling and , in doing so , to improve the characteristics of the pump they proposed to considerably increase the number of junctions @xmath13 ) and , hence , of the gates ( @xmath14 )
however , such modification would make operation of the circuit more complex . in this paper
we propose an alternative way to improve cooper pair pumping in a three - junction array .
we modify the bare circuit by attaching to the ends of the array the miniature resistors ( see the electric diagram of this device , which we call r - pump , in fig . [ schema ] ) . their total resistance @xmath15 exceeds the resistance quantum @xmath16 k@xmath1 so that the dimensionless parameter @xmath17 the self - capacitance of these resistors should not be much larger than the junction capacitance @xmath18 .
in this paper we will analyze how such resistors affect tunneling and co - tunneling of cooper pairs in the three - junction array and present preliminary experimental data .
in general terms , the principle of operation of the r - pump remains the same as that of the pump without resistors ( for details of the cooper pair pump operation see ref .
two periodic signals @xmath19 and @xmath20 are applied to the gates to form an elliptic trajectory in the parameter plane @xmath21 . at zero voltage @xmath22 across the pump
, this trajectory encircles in the clockwise ( counter- clockwise ) direction the triple point of the boundaries between the stability domains for adjacent charge configurations with an excess cooper pair . a complete cycle results in the charge @xmath8 @xmath23 being sequentially transferred across all three junctions of the array .
the cooper pair transitions in a circuit without dissipation ( quasiparticle tunneling is neglected ) occur due to josephson tunneling ( supercurrent ) .
the strength of the josephson coupling @xmath24 ( here @xmath25 is the critical current ) is assumed to be @xmath26 , the charging energy , i.e. parameter @xmath27 .
when the domain boundary is crossed , the resonance condition is obeyed and josephson coupling mixes up the two states , each corresponding to the position of the pair on either side of the junction .
( note that , at the same point , the resonance condition for co - tunneling in the opposite direction across two other junctions is met . )
slow passage across the boundary results in transition of the pair in the desired direction .
this process is mapped on the adiabatic level - crossing dynamics ( see , for example , ref .
@xcite ) , and the probability of missing the transition is given by the landau - zener expression : @xmath28 where @xmath29 is the velocity of variation of the energy difference @xmath30 between the two states .
in contrast to the elastic ( and , hence , reversible ) tunneling of pairs in the pump without resistors , tunneling in the r - pump is accompanied by dissipation of energy .
the strength of the dissipation effect depends on the active part re@xmath31 of an electromagnetic impedance seen by the tunneling pair , namely by the dimensionless parameter @xmath32 re@xmath33 .
if the condition of weak josephson coupling @xmath34 is satisfied , i.e. @xmath35 , the rate of cooper pair tunneling is expressed as @xcite @xmath36 where @xmath30 is the energy gain associated with this transition and @xmath37 is the function describing the property of the environment to absorb the energy released in a tunneling event .
( the discussion of properties as well as numerous examples of @xmath37 for different types of electromagnetic environment are given in the review paper by ingold and nazarov @xcite . ) the probability of a missed transition is then given by the expression : @xmath38 dt \right\ } = \exp \left(-{\frac{\pi e_j^2}{2\hbar\dot{e } } } \right ) = p_{\rm lz}.\ ] ] here we used the normalization condition @xmath39 , and the assumption of a value @xmath40 constant in time was exploited .
thus , in the case of weak coupling , dissipation does not affect the total probability of tunneling @xmath41 .
this conclusion is in accordance with the general result obtained by ao and rammer @xcite for quantum dynamics of a two - level system in the presence of substantial dissipation . using the expansion of @xmath42 in the region of small @xmath43 @xcite , @xcite one arrives at the expression @xmath44 for @xmath45 the rate @xmath46 is peaked at @xmath47 , while in the case ( @xmath48 ) the rate is zero at @xmath47 and the maximum
is reached at finite energy @xmath49 , or , in other terms , at finite positive voltage @xmath22 across the junction .
this property of @xmath50 in circuits with substantial dissipation plays the crucial role in the operation of the three - junction r - pump . for one junction in the three - junction array with gate capacitances @xmath51 the effective impedance re@xmath52 , i.e. the parameter @xmath53 .
the factor @xmath54 , which considerably attenuates the effect of resistors , stems from the square of the ratio of the total capacitance of the series array ( @xmath55 ) to the junction capacitance @xmath18 ( see a similar network analysis in ref .
@xcite ) .
accordingly , the equivalent damping for two junctions of the three - junction array is larger and determined by @xmath56 and , hence , @xmath57 .
therefore , the rate of cooper pair transition simultaneously across two junctions , i.e. the rate of the co - tunneling , at small @xmath30 is @xmath58 this expression has been obtained by considering the environmental effect in a similar fashion as was done by odintsov @xmath59 @xcite for the co - tunneling of normal electrons in a single - electron transistor .
accordingly , co - tunneling across all three junctions decays even stronger ( @xmath60 ) , because the whole array experiences the full resistance @xmath15 . the comparison of eq .
( 6 ) with eq .
( 5 ) taken for the value @xmath61 shows that the rate of co - tunneling is drastically depressed if @xmath62 is sufficiently large .
for example , for @xmath63 ( or @xmath64 k@xmath1 ) the rate of cooper pair co - tunneling across two junctions @xmath65 ( i.e. it is similarly small as single - electron co - tunneling across a three - junction normal array @xcite ) .
on the other hand , the direct tunneling rate @xmath66 , is similar to that of ordinary electron tunneling .
the effective junction resistance in the latter case is @xmath67 . for the practical value @xmath68 this
formula yields @xmath69 k@xmath1 , i.e. the typical value of a normal single electron junction .
since the junction resistance determines the characteristic time constant @xmath70 , it can be expected that the frequency characteristics of the cooper pair r - pump are at least not worse than those of the normal electron counterpart without resistors .
the al / alo@xmath71/al tunnel junctions and cr resistors were fabricated @xmath72 by the three - angle shadow evaporation technique through a trilayer mask patterned by e - beam lithography and reactive - ion etching .
the tunnel junction parameters of the best suited available sample were found to be : capacitance @xmath73 af and normal - state resistance @xmath74 k@xmath1 .
the former value yields @xmath75ev . on the assumption that the ambegaokar - baratoff relation between the critical current @xmath25 and the junction resistance is valid , the aforementioned value of @xmath76 yields the value @xmath77ev , where @xmath78ev is the superconducting energy gap of aluminum at low temperature .
these parameters of the sample give the ratio of characteristic energies @xmath79 .
the thin - film cr resistors ( thickness @xmath80 nm ) had lateral dimensions of about 80 nm by 10 @xmath81 m .
for the purpose of an independent characterization of the resistors and junctions , we attached four resistors in twos to either end of the junction array , resulting in the equivalent resistance @xmath82 on each side as depicted in fig .
[ schema ] .
( for the measurement of the i - v curves we used , however , the two - wire configuration . ) either of four resistors had @xmath83 k@xmath1 and reasonably low self - capacitance per unit length , @xmath84 af/@xmath81 m .
the resistors exhibited negligibly small ( @xmath85 % ) non - linearity of the i - v curve in the region of small voltages at the millikelvin temperature @xcite .
the samples were measured in the dilution refrigerator at the temperature @xmath86 of about 10 mk , i.e. well below the critical temperature of aluminum , @xmath87 1.15 k. the bias and gate lines were equipped with pieces of the thermocoax cable which was thermally anchored at the sample holder plate and served as a filter for frequencies above 1 ghz @xcite . the typical i - v curves with and without ac drive are presented in fig .
[ ivc ] . the curves exhibit considerably smeared steps whose position ( @xmath88 ) clearly shows the linear current - versus - frequency dependence ( although the quality of the steps is noticeably degraded with increasing frequency )
another feature of the steps is their position on the voltage axes : they appear at finite voltage @xmath22 applied in the current direction , while at @xmath89 neither co - tunneling nor pumping effect were observed .
we attribute this effect to the damping effect of the resistors , which was too strong in this sample .
thus , the behavior of the cooper pair r - pump differs significantly from that of the cooper pair pump without resistors @xcite .
although our first experiment evidenced the desirable effect of cooper pair pumping , a further improvement is inevitably needed .
first , the sample parameters should be optimized : josephson coupling should be increased ( possibly up to @xmath90 ) while the resistance value @xmath15 should be somewhat reduced .
these parameter values lead to an increase of the tunneling rate and , as a consequence , to cooper pair pumping at zero voltage bias @xmath22 .
secondly , the quasiparticle tunneling , resulting in sporadic translations of the cycle trajectory and , hence , to the pumping errors , should be reliably depressed : as the measurements showed , all our al - cr samples ( as well as those with @xmath91 ) so far suffered from quasiparticle tunneling @xcite and did not show the so - called parity effect @xcite , @xcite .
it is particularly remarkable that our cr resistors do not prevent an uncontrolled poisoning of josephson tunneling by non - equilibrium quasiparticles arriving at the islands from the external circuit .
such `` buffer '' effect of the normal electrodes was proposed and demonstrated in al - cu devices by joyez @xmath59 @xcite .
that is why the problem of the poisoning quasiparticle tunneling requires a more radical solution .
probably manufacture of small - capacitance niobium junctions ( @xmath92 mev @xmath93 ) could improve the situation by making the parity effect in these structures stable . in conclusion , we proposed a simple superconductor - normal metal circuit enabling , in principle , the efficient pumping of cooper pairs . due to the energy dissipation in the resistors , the effect of cooper pair co - tunneling
is heavily damped .
a possibility of pair pumping was demonstrated and improvements of the experiment are proposed .
d. v. averin and k. k. likharev , single - electronics : correlated transfer of single electrons and cooper pairs in small tunnel junctions . in b.
l. altshuler , p. a. lee and r. a. webb editors , _ mesoscopic phenomena in solids , _ pages 173 - 271 , amsterdam , north - holland : elsevier , 1991 .
a. b. zorin , s. v. lotkhov , h. zangerle and j. niemeyer .
coulomb blockade and co - tunneling in single electron circuits with on - chip resistors : towards the implementation of the r - pump .
_ j. appl .
phys . , _ 88:2665 - 2670 , 2000 . in the normal state maintained by application of perpendicular magnetic field @xmath94 t
, this sample exhibited excellent steps at @xmath95 due to the single electron pumping , see s. v. lotkhov , s. a. bogoslovsky , a. b. zorin and j. niemeyer .
the single electron r - pump : first experiment .
_ preprint , available at _ http://arxiv.org/abs/cond-mat/0001206 .
m. t. tuominen , j. m. hergenrother , t. s. tighe and m. tinkham .
experimental evidence for parity - based 2@xmath96 periodicity in a superconducting single - electron transistor .
lett . , _ 69:1997 - 2000 , 1992 ; a. amar , d. song , c. j. lobb and f. c. wellstood .
2@xmath96 and @xmath96 periodic pair currents in superconducting coulomb - blockade electrometers .
lett . , _ 72:3234 - 3237 , 1994 .
p. joyez , p. lafarge , a. filipe , d. esteve and m. h. devoret .
observation of parity - induced suppression of josephson tunneling in the superconducting single electron transistor .
lett . , _ 72:2458 - 2461 , 1994 . | we propose a circuit ( we call it r - pump ) comprising a linear array of three small - capacitance superconducting tunnel junctions with miniature resistors ( @xmath0 k@xmath1 ) attached to the ends of this array .
owing to the coulomb blockade effect and the effect of dissipative environment on the supercurrent , this circuit enables the gate - controlled transfer of individual cooper pairs . the first experiment on operating the r - pump
is described . |
the orthorhombic heavy fermion system ucoge , discovered in 2007 @xcite , is one of the few compounds exhibiting long range coexistence between weak itinerant ferromagnetism ( magnetic moment m@xmath1 @xcite ) and superconductivity .
such coexistence is attested by the observation of two bulk phase transitions in specific heat measurements @xcite . @xmath2sr and nqr measurements on different samples
@xcite reveal that the compound is fully ferromagnetic below the curie temperature ( @xmath3@xmath4k ) while about 50% of the sample is superconducting below @xmath5@xmath6k @xcite .
the upper critical field is extremely anisotropic , exceeding 16 tesla for @xmath7 and @xmath0 while it reaches only 0.5 tesla for @xmath8 @xcite , the easy magnetization axis @xcite .
the coexistence of ferromagnetism and superconductivity and the observed extremely high upper critical field suggest the realization of unconventional superconductivity with equal spin pairing ( triplet ) @xcite .
such a superconductor is inherently two - band ( one for the up and one for the down spins ) , however it is not known whether the two bands are superconducting or only one , in analogy to the a1 phase of @xmath9he , and whether the decoupling between the bands is large enough to induce multigap superconductivity . multigap superconductivity is quite common .
first observed in nb - doped srtio@xmath10 @xcite , it is found in mgb@xmath11@xcite , various heavy fermions @xcite , cuprates @xcite and pnictides @xcite systems .
here we report evidences for multigap superconductivity in the ferromagnetic superconductor ucoge . for a two - band ferromagnetic superconductor , due to the crystal structure and the strong spin orbit coupling @xcite , only two types of odd parity superconducting states are possible @xcite .
these states differ by the position of their nodes , either lying on the northern and southern poles of the fermi surface @xmath12 or on the line of equator @xmath13 .
the symmetry of the superconducting gap has a small effect on the shape of the upper critical field , which was used by hardy and huxley @xcite to suggest a superconducting state with a line of nodes in the parent system urhge .
however , in contradiction to the theoretical prediction this line was proposed to occur at @xmath14 .
thermal conductivity is a strong probe of the gap symmetry of a superconductor .
indeed , the density of non - superconducting quasiparticles , which are the main heat carrier channel at low temperatures , is strongly influenced by the presence and type of gap nodes .
( color online ) raw data for thermal conductivity divided by temperature and electrical resistivity in inset , with ( a ) @xmath0-axis and ( b ) @xmath8-axis .
heat current is applied along the @xmath15 crystallographic axis.,title="fig:",scaledwidth=48.0% ] ( color online ) raw data for thermal conductivity divided by temperature and electrical resistivity in inset , with ( a ) @xmath0-axis and ( b ) @xmath8-axis .
heat current is applied along the @xmath15 crystallographic axis.,title="fig:",scaledwidth=48.0% ] we measured thermal conductivity ( @xmath16 ) with a two thermometers one heater setup in the temperature range 30mk 10k and in magnetic field up to 8.5 tesla .
resistive carbon thermometers were used , held by thin kevlar strings and measured through superconducting nbti wires to insure good external thermal insulation .
thermometers and sample were connected through a gold wire spot welded on the sample side , achieving a contact resistance of @xmath17@xmath18 .
four probes electrical resistivity ( @xmath19 ) was measured simultaneously using the same gold wires for voltage measurement , allowing a direct verification of the setup using the wiedemann franz law ( @xmath20 for @xmath21 with @xmath22 w@xmath23k@xmath24 ) .
this setup was mounted on a piezo rotator which allows fine tuning of the relative angle between magnetic field and sample crystallographic axis .
this is required due to the strong angular dependence of magnetic properties @xcite .
the single crystal of pure ucoge composition was grown using the czochralski method in a tetra
arc furnace , characterized by laue diffraction and specific heat .
the sample was cut in a bar shape with the widest direction of @xmath25 mm along the @xmath15 crystallographic axes . with a residual resistivity ratio of @xmath26 , the sample offers a good compromise between crystallographic quality ( single grain crystal ) and low mean free path .
the latter property is required in order to differentiate between electronic and magnetic thermal conductivity contributions as described below . thermal conductivity divided by temperature ( @xmath27 ) and resistivity ( @xmath19 ) for two field orientations , @xmath0-axis and @xmath8-axis ,
are presented in figures [ fig : kappa_rho_hb ] ( a ) and ( b ) respectively .
the ferromagnetic transition ( @xmath3 ) disappears with a magnetic field applied along the @xmath15 crystallographic axis in agreement with such an orientation for the magnetic moments , due to the absence of symmetry breaking at the transition .
the superconducting transition ( @xmath5 ) is clearly observed by a kink in the two sets of curves below @xmath28k .
the observation of both superconducting and ferromagnetic transitions in thermal conductivity curves indicates that both phases are bulk .
the enhancement of @xmath5 under magnetic field ( @xmath0 ) previously observed in resistivity measurements @xcite is confirmed as a bulk property ( red dashed line in figure [ fig : kappa_rho_hb](a ) ) which indicates a high precision in the sample alignment .
indeed , such enhancement was reported to occur only when the magnetic field is applied within 1@xmath29 of the @xmath30 crystallographic axis @xcite .
it is particularly interesting to note that the bulk re - entrance of superconductivity already occurs at @xmath315tesla , while the resistive one is only observed around 10tesla @xcite .
the method used to extract @xmath5 is described in @xcite .
the residual value in the superconducting state is quite large ( @xmath32% of the normal state value ) . for a fully open superconducting gap ( s - wave )
the ratio @xmath27 vanishes with temperature ( @xmath33 ) , as observed in nb@xcite , nbse@xcite and mgb@xmath11@xcite for example . for systems in which the superconducting gap vanishes at points or lines of nodes ,
a residual value is expected due to impurity scattering .
such residual value was observed in the d - wave high-@xmath5 superconductor tl@xmath11ba@xmath11cuo@xmath34 @xcite and in the possibly p - wave superconductor sr@xmath11ruo@xmath35 @xcite . for systems
in which a superconducting gap with lines of nodes is present , the limit can even be universal , with @xmath36 independent of the amount of impurities @xcite . in ucoge ,
the residual value of thermal conductivity is not universal .
a recently probed sample with rrr@xmath37 has a lower residual term of @xmath38% of the extrapolated non superconducting value @xcite .
the residual term can be due to the presence of nodes in the superconducting gap , a special superconducting phase ( like superfluid phase a1 of @xmath9he ) , a band of gapless superconductivity ( due to impurities ) or an inhomogeneous sample ( partly non superconducting ) .
this last option is supported by nqr measurements , reporting a mixture of superconducting and non - superconducting regions @xcite , and by a specific heat measurement reporting a residual term of about 50% of the normal state value @xcite .
both experiments were performed on samples of similar quality to the one used in this study .
note that the sharpness of both the ferromagnetic and the superconducting transitions exclude the possibility of a distribution of transitions in the bulk phases , which would give rise to broad or no features in thermal conductivity .
the higher value for @xmath5 observed in the resistivity measurement must originate from part of the sample with non bulk superconductivity .
this could be due to filamentary superconductivity or superconductivity occurring first in ferromagnetic domain wall , where magnetization is suppressed which is certainly favorable for superconductivity .
( color online ) lorenz ratio @xmath39 obtained from the thermal conductivity and resistivity data presented in figure [ fig : kappa_rho_hb]a .
all the curves extrapolate to 1 at @xmath40k , reflecting the good quality of the measurement . in inset the residual thermal conductivity contribution ( @xmath41 ) is shown ( see text).,scaledwidth=48.0% ] + in figure [ fig : ll0 ] the lorenz ratio ( @xmath42 ) calculated for different values of the magnetic field applied along the @xmath30 crystallographic direction is plotted . for @xmath21 the wiedemann - franz law @xmath43
is obeyed within 5% which reflect the validity of our measurements .
the equality indicates that at low temperatures , electrical and thermal currents are transported by the same carriers : the electrons . at finite temperatures
two effects can produce a deviation from the unity of @xmath44 .
other thermal carrier channels such as phonon or magnon excitations will enhance this ratio .
in contrary the ratio will be lower than one if a strong electron electron inelastic scattering , reducing the efficiency of electronic thermal conductivity , dominates @xcite . in a conventional metal ,
magnons are absent and as few phonons are present at low temperatures , the ratio is reduced below 1 for the lowest temperatures and exceeds 1 in the higer temperatures regime due to the phonons contribution .
the depth of the minimum in @xmath44 depends on the mean free path of the quasiparticles .
as the sample investigated has a relatively low @xmath45 value the mean free path of the quasiparticles is short and only a weak deviation from 1 is expected for @xmath44 at low temperatures . in figure
[ fig : ll0 ] the increase of @xmath44 above 1 from the lowest temperatures indicates the presence of another type of heat carrier than the electrons .
as ucoge is ferromagnetic , magnons or uniaxial fluctuations are good candidates @xcite .
an analysis of the anisotropy of this contribution with heat current direction re - enforce this hypothesis @xcite .
we can obtain the approximate value of this additional thermal conductivity contribution ( @xmath46 ) by assuming the lorenz ratio is 1 for the electronic contribution : @xmath47 with @xmath48 @xmath49 @xmath50 .
we found that the magnetic contribution is independent of a magnetic field applied along the @xmath30 crystallographic direction ( inset figure [ fig : ll0 ] ) while it is strongly reduced by a magnetic field applied in the @xmath15 crystallographic direction , as expected for longitudinal spin fluctuations @xcite and in agreement with a previous study using a different technique to extract the magnetic contribution @xcite . in order to further analyze the temperature dependence of thermal conductivity
, we subtract the additional contribution obtained previously ( @xmath51 ) and calculated the thermal conductivity one would observe if the compound was not superconducting , hereafter called normal contribution to thermal conductivity ( @xmath52 ) .
this is done by extrapolating the normal state resistivity to @xmath21 assuming a fermi liquid dependence ( @xmath53 ) .
then we calculated the electronic contribution to thermal conductivity using the wiedemann
franz law : @xmath54 .
note that this is only possible due to the moderate @xmath45 value of the sample , when practically no deviation from the wiedemann
franz law are expected . indeed , for a more general study @xcite , a phenomenological model had to be introduced for the electronic thermal conductivity which prevent any further discussion of the temperature dependence .
the ratio of superconducting to `` extrapolated normal state '' thermal conductivity in the superconducting state ( @xmath55 ) is reported in figure [ fig : ks_kn_t ] ( a ) and ( b ) for different magnetic fields applied along the @xmath30 crystallographic direction .
such ratio is related to the fraction of superfluid quasiparticles .
we can clearly distinguish two different temperatures dependencies : linear for @xmath56 t and quadratic otherwise .
panel ( c ) of figure [ fig : ks_kn_t ] emphases the linear and quadratic temperature dependencies with a double logarithmic plot .
the temperature dependence of thermal conductivity is related to the type of nodes ( points , lines ) and their opening angle @xcite .
the different temperature dependencies indicate two different gap structures depending on the field range .
there is no report of a phase transition between two different superconducting phases , neither with temperature nor upon applying a magnetic field , as required in order to modify the symmetry of the superconducting gap and our measurements support a crossover .
therefore , we infer a multigap superconducting state and not multiple superconducting states .
the low field temperature dependence would then result from the addition of the thermal conductivity of the two bands , while at high fields only the band with the larger superconducting gap would be superconducting .
the high magnetic field temperature dependence , @xmath57 , suggests the presence of a line of nodes in the gap of the superconducting band , as expected in analogy to urhge @xcite .
note that the observation of two different field ranges is independent of the temperature dependence of the subtracted @xmath46 contribution .
a similar evolution of the temperature dependence of thermal conductivity was observed in well known two - band superconductors such as mgb@xmath11 @xcite , cecoin@xmath58 @xcite and nbse@xmath11 @xcite although with different power laws . the small deviation from linearity of @xmath55 at low temperatures ( @xmath59 ) for @xmath60 t is understood in the multigap scenario as corresponding to the characteristic energy of the smaller gap ( figures [ fig : kappa_rho_hb ] ( a ) and [ fig : ks_kn_t ] ) .
( color online ) field dependence of the ratio @xmath61 : ( a ) for @xmath8 ( red squares ) , ( b ) for @xmath0 ( orange circles ) .
( c ) comparison between @xmath62 ( see text ) of ucoge and the single gap superconductor nb ( black squares @xcite ) and the two - band superconductors pros@xmath35sb@xmath63 ( green triangles @xcite ) , mgb@xmath11 @xmath64 ( blue diamonds @xcite ) and mgb@xmath11 @xmath8 ( violet stars @xcite ) .
@xmath65 of ucoge is taken as 25 tesla in ( c).,scaledwidth=48.0% ] the idea of a multigap superconductivity is re - enforced by the appearance of a plateau above @xmath66 t in the field evolution of the residual term of thermal conductivity , for @xmath0 ( figure [ fig : ks_kn_h ] ( b ) ) .
the experimental resolution might not allow to observe such feature for @xmath8 ( figure [ fig : ks_kn_h ] ( a ) ) .
the decoupling between the two gaps might also be weaker in this configuration .
if we assume that the residual term for @xmath60 t is due to an inhomogeneous part of the sample never superconducting , we can extend the analysis by comparing ucoge to well established two gaps superconductors , as pros@xmath35sb@xmath63 and mgb@xmath11 ( fig .
[ fig : ks_kn_h ] ( c ) ) .
here we have assumed two parallel contributions to thermal conductivity @xmath67 , with @xmath68 and @xmath69 for the never superconducting contribution .
@xmath69 is assumed to be field independent . independently of this assumption , the three systems : ucoge , pros@xmath35sb@xmath63 and mgb@xmath11 presented in figure [ fig : ks_kn_h ]
are characterized by two energy scales corresponding to the values of the two respective gaps .
the field dependence of @xmath70 for a single gap superconductor is drastically different to the one found in ucoge as demonstrated with the case of nb .
a ferromagnetic system has inherently two bands , for majority and minority electron spins .
it is therefore tempting to map the two superconducting gaps to the two ferromagnetic bands . the different strength of superconductivity
could then be explained by the proximity to the lifshitz phase transition , previously reported @xcite .
the strong increase of density of state would enhance the superconducting coupling .
it is however not clear whether such bands could be decoupled enough electronically to induced multigap superconductivity .
a classical scenario of the two gaps occurring on different fermi pockets is another possibility .
in conclusion , with the study of thermal conductivity ( @xmath71 ) in the ferromagnetic heavy fermion system ucoge , we confirm the re - enforcement of superconductivity under magnetic field and establish the bulk character of this effect .
in addition , we observed two distinct energy scales depending on the value of the magnetic field applied along the @xmath30 crystallographic axis . at high magnetic fields
the temperature dependence of thermal conductivity is compatible with the presence of a line of nodes in the superconducting gap as reported in the parent system urhge .
both the field dependence of the residual term and the different temperature dependencies of @xmath71 suggest the realization of multigap superconductivity .
further experiments on samples with different @xmath45 values and heat currents directions as well as theoretical modeling of thermal conductivity in multigap systems are required in order to obtain the exact nature of the superconducting state in the ferromagnetic superconductor ucoge .
we gratefully acknowledged experimental support from j.p . brison and sample preparation from v. taufour .
this work was supported by the commissariat lnergie atomique of grenoble and by the universit de grenoble .
n. t. huy , a. gasparini , d. e. de nijs , y. huang , j. c. p. klaasse , t. gortenmulder , a. de visser , a. hamann , t. grlach , h. v. lhneysen , superconductivity on the border of weak itinerant ferromagnetism in ucoge , phys .
99 ( 6 ) ( 2007 ) 067006 .
http://dx.doi.org/10.1103/physrevlett.99.067006 [ ] .
n. t. huy , d. e. de nijs , y. k. huang , a. de visser , unusual upper critical field of the ferromagnetic superconductor ucoge , phys .
100 ( 7 ) ( 2008 ) 077002
. http://dx.doi.org/10.1103/physrevlett.100.077002 [ ] .
a. de visser , n. t. huy , a. gasparini , d. e. de nijs , d. andreica , c. baines , a. amato , muon spin rotation and relaxation in the superconducting ferromagnet ucoge , phys .
102 ( 16 ) ( 2009 ) 167003 . http://dx.doi.org/10.1103/physrevlett.102.167003 [ ] .
t. ohta , t. hattori , k. ishida , y. nakai , e. osaki , k. deguchi , n. k. sato , i. satoh , http://jpsj.ipap.jp/link?jpsj/79/023707/[microscopic coexistence of ferromagnetism and superconductivity in single - crystal ucoge ] , j. phys .
jpn . 79 ( 2 ) ( 2010 ) 023707 . http://dx.doi.org/10.1143/jpsj.79.023707 [ ] . http://jpsj.ipap.jp/link?jpsj/79/023707/ t. ohta , y. nakai , y. ihara , k. ishida , k. deguchi , n. k. sato , i. satoh , http://jpsj.ipap.jp/link?jpsj/77/023707/[ferromagnetic quantum critical fluctuations and anomalous coexistence of ferromagnetism and superconductivity in ucoge revealed by co - nmr and nqr studies ] , j. phys .
77 ( 2 ) ( 2008 ) 023707 .
http://dx.doi.org/10.1143/jpsj.77.023707 [ ] .
d. aoki , t. d. matsuda , v. taufour , e. hassinger , g. knebel , j. flouquet , http://jpsj.ipap.jp/link?jpsj/78/113709/[extremely large and anisotropic upper critical field and the ferromagnetic instability in ucoge ] , j. phys .
78 ( 11 ) ( 2009 ) 113709 . http://dx.doi.org/10.1143/jpsj.78.113709 [ ] . http://jpsj.ipap.jp/link?jpsj/78/113709/ g. binnig , a. baratoff , h. e. hoenig , j. g. bednorz , http://link.aps.org/doi/10.1103/physrevlett.45.1352[two-band superconductivity in nb - doped srtio@xmath10 ] , phys .
( 1980 ) 13521355 .
http://dx.doi.org/10.1103/physrevlett.45.1352 [ ] .
http://link.aps.org/doi/10.1103/physrevlett.45.1352 f. bouquet , y. wang , i. sheikin , t. plackowski , a. junod , s. lee , s. tajima , http://link.aps.org/doi/10.1103/physrevlett.89.257001[specific heat of single crystal @xmath73 : a two - band superconductor with two different anisotropies ] , phys .
( 2002 ) 257001 .
http://dx.doi.org/10.1103/physrevlett.89.257001 [ ] .
g. seyfarth , j. p. brison , m .- a .
masson , j. flouquet , k. izawa , y. matsuda , h. sugawara , h. sato , multiband superconductivity in the heavy fermion compound pros@xmath35sb@xmath63 , phys .
95 ( 10 ) ( 2005 ) 107004 .
http://dx.doi.org/10.1103/physrevlett.95.107004 [ ] .
g. seyfarth , j. p. brison , g. knebel , d. aoki , g. lapertot , j. flouquet , multigap superconductivity in the heavy - fermion system cecoin@xmath58 , phys . rev .
lett . 101 ( 4 ) ( 2008 ) 046401 .
http://dx.doi.org/10.1103/physrevlett.101.046401 [ ] .
r. khasanov , a. shengelaya , a. maisuradze , f. lamattina , a. bussmann - holder , h. keller , k. a. mller , experimental evidence for two gaps in the high - temperature @xmath74 superconductor , phys .
98 ( 2007 ) 057007 . http://dx.doi.org/10.1103/physrevlett.98.057007 [ ] .
j. yang , g .- q .
zheng , multiple superconducting gaps , anisotropic spin fluctuations and spin - orbit coupling in iron - pnictides , mod .
b 26 ( 2012 ) 1230008 . http://dx.doi.org/10.1142/s0217984912300086 [ ] . m. samsel - czekaa , s. elgazzar , p. m. oppeneer , e. talik , w. walerczyk , r. tro , http://stacks.iop.org/0953-8984/22/i=1/a=015503[the electronic structure of ucoge by _ ab initio _ calculations and xps experiment ] , j. phys .
: condens .
matter 22 ( 1 ) ( 2010 ) 015503 .
v. p. mineev , t. champel , http://link.aps.org/doi/10.1103/physrevb.69.144521[theory of superconductivity in ferromagnetic superconductors with triplet pairing ] , phys .
b 69 ( 2004 ) 144521 . http://dx.doi.org/10.1103/physrevb.69.144521 [ ] . http://link.aps.org/doi/10.1103/physrevb.69.144521 d. aoki , t. d. matsuda , f. hardy , c. meingast , v. taufour , e. hassinger , i. sheikin , c. paulsen , g. knebel , h. kotegawa , j. flouquet , http://jpsj.ipap.jp/link?jpsjs/80sa/sa008/[superconductivity reinforced by magnetic field and the magnetic instability in uranium ferromagnets ] , j. phys . soc . jpn . 80sa ( supplement a ) ( 2011 ) sa008 .
l. howald , http://tel.archives-ouvertes.fr/tel-00584598[interactions between superconductivity and quantum criticality in cecoin@xmath58 , urhge and ucoge ] , ph.d .
thesis , universit de grenoble ( 2011 ) .
j. lowell , j. b. sousa , http://dx.doi.org/10.1007/bf00628399[mixed-state thermal conductivity of type ii superconductors ] , j. low temp .
phys . 3 ( 1970 ) 6587 .
http://dx.doi.org/10.1007/bf00628399 [ ] .
e. boaknin , m. a. tanatar , j. paglione , d. hawthorn , f. ronning , r. w. hill , m. sutherland , l. taillefer , j. sonier , s. m. hayden , j. w. brill , heat conduction in the vortex state of nbse@xmath76 : evidence for multiband superconductivity , phys .
90 ( 11 ) ( 2003 ) 117003 .
a. v. sologubenko , j. jun , s. m. kazakov , j. karpinski , h. r. ott , thermal conductivity of single - crystalline mgb@xmath11 , phys .
b 66 ( 1 ) ( 2002 ) 014504 . http://dx.doi.org/10.1103/physrevb.66.014504 [ ] .
c. proust , e. boaknin , r. w. hill , l. taillefer , a. p. mackenzie , heat transport in a strongly overdoped cuprate : fermi liquid and a pure @xmath77-wave bcs superconductor , phys .
89 ( 14 ) ( 2002 ) 147003 . http://dx.doi.org/10.1103/physrevlett.89.147003 [ ] .
h. suderow , j. brison , j. flouquet , a. tyler , y. maeno , very low temperature thermal conductivity in the layered perovskite superconductor sr@xmath11ruo@xmath35 , j. phys . : condens .
matter 10 ( 34 ) ( 1998 ) l597l602 .
m. suzuki , m. a. tanatar , n. kikugawa , z. q. mao , y. maeno , t. ishiguro , universal heat transport in sr@xmath76ruo@xmath78 , phys .
88 ( 22 ) ( 2002 ) 227004 . http://dx.doi.org/10.1103/physrevlett.88.227004 [ ] .
m. j. graf , s .- k .
yip , j. a. sauls , d. rainer , http://link.aps.org/doi/10.1103/physrevb.53.15147[electronic thermal conductivity and the wiedemann - franz law for unconventional superconductors ] , phys . rev .
b 53 ( 1996 ) 1514715161 . http://dx.doi.org/10.1103/physrevb.53.15147 [ ] .
m. taupin , l. howald , d. aoki , j. flouquet , j. p. brison , http://link.aps.org/doi/10.1103/physrevb.89.041108[existence of anisotropic spin fluctuations at low temperature in the normal phase of the superconducting ferromagnet ucoge ] , phys .
b 89 ( 2014 ) 041108 . http://dx.doi.org/10.1103/physrevb.89.041108 [ ] . http://link.aps.org/doi/10.1103/physrevb.89.041108 d. aoki , m. taupin , c. paulsen , f. hardy , v. taufour , h. kotegawa , e. hassinger , l. malone , t. d. matsuda , a. miyake , i. sheikin , w. knafo , g. knebel , l. howald , j .-
brison , j. flouquet , http://jpsj.ipap.jp/link?jpsjs/81sb/sb002/[field-induced phenomena in ferromagnetic superconductors ucoge and urhge ] , journal of the physical society of japan 81sb ( supplement b ) ( 2012 ) sb002 . http://dx.doi.org/10.1143/jpsjs.81sb.sb002 [ ] . http://jpsj.ipap.jp/link?jpsjs/81sb/sb002/ k. proke , a. de visser , y. k. huang , b. fk , e. ressouche , http://link.aps.org/doi/10.1103/physrevb.81.180407[anomalous spin distribution in the superconducting ferromagnet ucoge studied by polarized neutron diffraction ] , phys .
b 81 ( 2010 ) 180407 . http://dx.doi.org/10.1103/physrevb.81.180407 [ ] .
y. matsuda , k. izawa , i. vekhter , http://stacks.iop.org/0953-8984/18/i=44/a=r01[nodal structure of unconventional superconductors probed by angle resolved thermal transport measurements ] , j. phys . :
matter 18 ( 44 ) ( 2006 ) r705 .
l. malone , l. howald , a. pourret , d. aoki , v. taufour , g. knebel , j. flouquet , http://link.aps.org/doi/10.1103/physrevb.85.024526[thermoelectricity of the ferromagnetic superconductor ucoge ] , phys .
b 85 ( 2012 ) 024526 . http://dx.doi.org/10.1103/physrevb.85.024526 [ ] . http://link.aps.org/doi/10.1103/physrevb.85.024526 | we performed thermal conductivity measurements on a single crystal of the ferromagnetic superconductor ucoge under magnetic field .
two different temperature dependencies of the thermal conductivity are observed , for @xmath0 : linear at low magnetic field and quadratic for magnetic field larger than 1 tesla . at the same field value
, a plateau appears in the field dependency of the residual term of thermal conductivity .
such observations suggest a multigap superconductivity with a line of nodes in the superconducting gap . thermal conductivity , ferromagnetic superconductor , multigap , re - entrance of superconductivity , ucoge . 74.20.rp,74.25.fc,71.27.+a , |
coherent conversion of an atomic to a molecular bose - einstein condensate ( bec ) is a first step towards ` superchemistry ' @xcite , which is the stimulated emission of molecules in a chemical reaction .
a number of studies of this @xcite have shown that direct conversion via raman photoassociation @xcite appears feasible , based on stimulated free - bound and bound - bound transitions in the presence of two laser fields of different frequencies @xcite . here
pairs of atoms from the two - atom continuum of the ground electronic potential are transferred via an excited bound molecular state to a bound molecular state of a lower energy in the ground potential .
raman photoassociation allows coupling to a single molecular state , which can be selected by the raman laser frequencies .
practical estimates using available lasers and transitions indicate that coherent transfer may be limited by spontaneous emission from the intermediate molecular excited electronic state .
another mechanism that can result in coupled atomic - molecular bec systems @xcite is based on feshbach resonances @xcite .
however , realistic analysis and experimental implementations @xcite indicate that the loss processes due to inelastic atom - molecule collisions occur at a significant rate . a possible route towards minimizing losses and decoherence from spontaneous emission in photoassociation is stimulated raman adiabatic passage ( stirap ) @xcite , in which a counter - intuitive pulse sequence is used , where the first input pulse couples the molecular levels - even when there are no molecules present .
in this situation , a dark superposition state is formed , due to interference effects between the atomic and molecular electronic ground states .
this minimizes the probability of a real transition to the molecular excited state , and hence reduces spontaneous emission .
previous analyses of this problem have not taken into account losses , collisions , or the full three - dimensional structure of the two bose condensates in a trap . in this paper
, we provide an analysis which is relevant in a physically appropriate model that does include the known physics of spontaneous emission losses , @xmath0-wave scattering processes and spatial diffusion of the condensates .
the result is that the stirap process appears feasible at high laser pulse intensities , provided the rabi frequency is much greater than the two - photon detuning due to mean - field interactions . we give a detailed numerical calculation based on both homogeneous and inhomogeneous three - dimensional mean - field ( gross - pitaevskii type ) theories , including couplings to non - resonant intermediate levels , and show how the results scale with the two - photon detuning , pulse duration and raman pulse intensities .
an optimal situation is found by considering an off - resonance operation and different effective rabi frequencies in the two raman channels .
we show that these strategies can greatly enhance the conversion efficiency for given laser intensities , thus making the experimental requirements much more feasible .
we start by considering the theory of coherently interacting atomic and molecular condensates needed to describe this process @xcite , and assume a specific coupling mechanism based on stimulated free - bound raman transitions @xcite , in which two atoms of energy @xmath1 collide to from a molecule of energy @xmath2 , with an excited molecular state forming as an intermediate step .
the raman coupling is induced by two laser fields of frequencies @xmath3 and @xmath4 , and becomes resonant when the raman detuning @xmath5 goes to zero .
this allows coupling to a single molecular state , which can be selected by the raman laser frequencies .
we derive the atom - molecule coupling for a simplified model of the two - body interaction @xcite , in which the atoms interact in their electronic ground state through a single born - oppenheimer potential @xmath6 .
molecules are formed in a single bound vibrational state of energy @xmath2 with radial wave function @xmath7 .
two free atoms with zero relative kinetic energy have a total energy @xmath8 , and a relative radial wave function @xmath9 , normalized so that asymptotically @xmath10 .
we assume that the laser field has two frequency components , with @xmath11 $ ] , @xmath12 . each couples the ground electronic state to a single electronically excited state described by a potential @xmath13 , with ` bare ' electronic rabi frequencies @xmath14 , where @xmath15 is the molecular electric dipole matrix element with a nuclear separation of @xmath16 .
the excited state has vibrational levels @xmath17 with energies @xmath18 and radial wave functions @xmath19 .
all bound levels are normalized so that @xmath20 .
we consider the case of near resonant transitions @xmath21 and denote the resonant excited vibrational level @xmath22 via index @xmath23 ( see fig .
[ diagram ] ) .
the usual quantum field theory hamiltonian @xcite for noninteracting atomic ( @xmath24 ) and molecular ( @xmath25 ) species , in well - defined internal states described by annihilation operators @xmath26 , is given by : @xmath27 .\ ] ] here @xmath28 ( @xmath29 ) are the masses , @xmath30 , and @xmath31 is the trapping potential including the internal energy for the @xmath32-th species , where we define @xmath33 .
including @xmath0-wave scattering processes and laser induced particle inter - conversion , together with the assumption of a momentum cut - off , results in the following terms in the effective interaction hamiltonian : @xmath34 , \\ \hat{h}_{int}^{(2 - 3 ) } & = & \int d^{3}{\mathbf{x}}\left [ \frac{-\hbar { \omega } _ { 2}}{2}e^{-i\omega _ { 2}t}\hat{\psi } _ { 2}^{{}}({\mathbf{x}})\hat{\psi } _ { 3}^{\dagger } ( { \mathbf{x}})+h.c.\right ] .\end{aligned}\ ] ] here @xmath35 ( @xmath12 ) are the molecular rabi frequencies .
these can be treated using the franck - condon overlap integrals @xmath36 , if we take @xmath37 as the mean electronic rabi frequency .
we note that @xmath38 , which connects the atomic and molecular condensates , has units of @xmath39 , and must be multiplied by the atomic condensate amplitude to obtain a true rabi frequency .
the couplings @xmath40 in the diagonal case are given by @xmath41 , where @xmath42 is the @xmath32-th species @xmath0-wave scattering length , while the non - diagonal terms are given by @xmath43 , where @xmath44 is the inter - species scattering length and @xmath45 is the reduced mass .
in addition , we account for losses from each state , at a rate @xmath46 .
the resulting heisenberg equations for the field operators are treated within the mean - field approximation , in which the operators are replaced by their mean values , and a factorization is assumed .
this approximation is expected to be valid at sufficiently high density .
corrections due to quantum correlations @xcite have been treated in greater detail elsewhere @xcite .
next , we introduce rotating frame detunings , defined so that : @xmath47 in the case of uniform condenstates , @xmath31 are equal to @xmath48 , and @xmath49 .
this results in the following set of gross - pitaevskii type of equations for the mean - field amplitudes , in rotating frames such that @xmath50 $ ] : @xmath51 here @xmath52 is the @xmath32 -th gross - pitaevskii mean - field detuning in the rotating frame , defined so that : @xmath53 we also introduce the two - photon laser detuning at trap center : @xmath54 in addition to losses due to spontaneous emission from the electronic excited states , rotationally or vibrationally inelastic atom - molecule collisions may also give rise to losses .
the magnitude of these rates is presently unknown , and we neglect them here .
we note that these rates should decrease rapidly with increasing molecular binding energy and go to zero in the molecular ground state , so that it should be possible to obtain a very low rate by selecting the coupling to a deeply bound molecular level .
this approximation means that we will set @xmath55 in the following treatment , where @xmath56 is the kronecker delta - function .
the simplest stirap scheme employs exact tuning of the laser frequencies to the bare state resonances , i.e. , @xmath57 .
this , however , is not necessarily required as the stirap can still occur with detuned intermediate levels . moreover , as we show below , in bec environments , the off - resonance operation turns out to be more efficient if the detunings compensate for the phase shifts due to mean - field energies . in effect
this is equivalent to a renormalized two - photon `` on - resonant '' operation in which @xmath58 . if , instead of considering near - resonance coupling , we consider a large intermediate level detuning so that the excited state can be adiabatically eliminated , we recover the basic terms in the set of equations analyzed in @xcite .
simultaneously , this would give us the previously known result @xcite for the laser induced modification to the scattering length @xmath59 occurring in a simple single - laser photoassociation of pairs of atoms , thus justifying the above form of the interaction hamiltonian .
in the present paper , however , we assume that the detunings @xmath60 are small compared to the characteristic separation between the vibrational levels so that all other vibrational levels can be neglected .
the near - resonant excited level is treated explicitly , rather than eliminated adiabatically as in @xcite .
before carrying out simulations of the full @xmath61 equations in a trap , it is instructive to start with a simplified version of the theory - in which there are no kinetic energy terms .
we expect this approximation to be valid in the thomas - fermi limit of large , relatively dense condensates , which is a regime of much current experimental interest .
this is described by the following set of equations : @xmath62 where we have introduced as an effective thomas - fermi limit frequency shift , @xmath63 this corresponds to the treatment given in @xcite , except that we explicitly include the loss term @xmath64 due to spontaneous emission , the @xmath0-wave scattering processes due to @xmath40 , and the franck - condon integrals into the coupling coefficients for the free - bound and bound - bound transitions .
in order to understand how this is related to the usual stirap technique in a three - level @xmath65 atomic system , we introduce a new wave - function @xmath66 , which corresponds to the coherent amplitude of a ( virtual ) molecular condensate with the same number of atoms as in the atomic bec . we then introduce a bose - stimulated rabi frequency , which includes a local coherent bec amplitude for the first free - bound transition : @xmath67 .
this leads to the equations : @xmath68 these are precisely the usual stirap equations , except with additional detunings coming from the mean - field interactions , and a rabi frequency in the first transition which is proportional to the amplitude of the atomic bec wave - function . in practise
, the rabi frequency may have an additional space - dependence due to the spatial variation of the laser phase and intensity .
we therefore conclude that , provided we can satisfy the normal adiabatic stirap requirements of slow time - variation in the _ effective _ rabi frequencies in the above equations , the technique will also work for a bec .
this is a simpler proof than previously @xcite . in particular
, we can immediately deduce the expected solution for real rabi frequencies in the adiabatic limit . in order to have @xmath69 ,
one requires : @xmath70 here the space dependent mixing angle @xmath71 is obtained from the ratio of effective rabi frequencies : @xmath72 ^{1/2}\ , .\ ] ] we can see that _ initially _ , while @xmath73 is still close to its initial value , the mixing angle is close to its expected value in normal stirap , since @xmath74 .
however , at the final stages of the adiabatic passage , the nonlinear effects due to the atom - molecular coupling become important .
as the atomic bec amplitude only varies on the time - scale of the input fields , the nonlinear atom - molecular coupling term by itself should not introduce new adiabatic restrictions .
however , there will be time - dependent detunings introduced by the mean - field terms .
finally , we can see that similar conclusions can also be reached in a non - uniform bec , by replacing the uniform detunings with appropriate gross - pitaevskii detunings , that include the spatial potentials . in a typical bec cooled in the thomas - fermi regime ,
we expect the kinetic energy terms to have relatively small effects .
stirap can therefore be implemented as usually by using two laser pulses applied in counterintuitive order .
we choose gaussian pulses of the form @xmath75,\ ; \ ; \ ; \ ; ( i=1,2),\ ] ] or @xmath76,\ ] ] where the peak values are related as follows : @xmath77 .
the pulse at frequency @xmath4 is applied first , with the center at @xmath78 , while the second pulse at frequency @xmath3 is delayed by @xmath79 , i. e. , @xmath80 where @xmath81 is the delay coefficient , and @xmath82 is the pulse duration , which we assume is the same for both pulses . in terms of the rabi frequencies ,
the adiabatic condition for stirap now reads as @xcite : @xmath83 where @xmath84 is the rms rabi frequency , @xmath85 is the duration during which the pulses overlap , and @xmath86 simply corresponds to the detuning of the single - photon transition .
however , there is a second condition , which is often not stated explicitly .
this is that stirap requires an effective two - photon resonance , to avoid dephasing between the initial and final states in the dark - state superposition .
the two - photon resonance condition is different from the usual stirap case , since a detuning of @xmath87 causes a phase rotation _ both _ in _ @xmath88 _ and in _ @xmath89 _ as well , since this also includes a phase term from the condensate . as a result ,
the necessary condition for two - photon resonance is therefore : @xmath90 as is usually the case in stirap , these conditions can not be satisfied very early or late in the pulse sequence , when the rabi frequencies are small ; but they should be satisfied over most of the stirap interaction , and over most of the condensate volume . for simplicity , we will apply these conditions to the peak rabi frequency @xmath91 , and to the total pulse duration @xmath82 .
further , since @xmath89 is itself a function of the stirap evolution , we introduce an effective first rabi frequency , defined in terms of the initial density @xmath92 .
this is an upper bound to the stimulated rabi frequency ; thus @xmath93 ( and sometimes we write @xmath94 , for uniformity ) , where @xmath95,\ ; \ ; \ ; \ ; ( i=1,2)\ , , \ ] ] where @xmath96 and @xmath97 . a typical pulse sequence is shown in fig .
[ stirap - pulses ] . from the definition of @xmath98 ,
the franck - condon overlap integrals are an important issue from the point of view of employing a realistic set of parameters .
since the overlap integrals enter into the definition of the effective rabi frequencies , their values will affect the adiabatic condition rewritten in terms of the `` bare '' electronic rabi frequencies @xmath37 .
we will analyze this in more detail in the next section .
we start by considering a uniform condensate , described by a similar equation to the thomas - fermi case , except with a uniform trap potential for simplicity : @xmath99 here the uniform detuning term @xmath100 is defined as the thomas - fermi detuning , evaluated at the trap center .
we start by considering a uniform condensate in which the @xmath0-wave scattering interactions are negligible ( i.e. , @xmath101 ) , and assume exact resonances with respect to bare state transitions , @xmath102 .
we first simulate the above simplified equations ( [ j-3 ] ) with an initial condition of a pure atomic condensate .
this model is more realistic than that of m. mackie et al .
@xcite , as it includes spontaneous emission .
we find that including the loss term @xmath64 imposes restrictions on the effective rabi frequencies @xmath103 and the duration of the pulses @xmath82 .
the results are best understood with reference to table [ stirap_param ] , which gives the values of typical stirap parameters characteristic of a condensate of @xmath104rb atoms @xcite ) , and corresponding to the pulse sequence in fig . [ stirap - pulses ] .
.typical parameter values for efficient stirap . [ cols="<,<",options="header " , ] the most favorable case that we find corresponds to tuning the raman transitions to @xmath105 in the excited potential bound by @xmath106 @xmath107 or @xmath108 s@xmath109 from the dissociation limit , and @xmath110 in the ground potential bound by @xmath111 ghz or @xmath112 s@xmath109 .
the binding energy of the level @xmath110 sets up the frequency difference @xmath113 @xmath114 s@xmath109 , and the values of all relevant detunings .
the nearby levels around @xmath105 in the excited potential are separated by about @xmath115 s@xmath109 , which is much larger than @xmath116 . for this arrangement of the target levels , the resonant franck - condon overlap integrals
are equal to @xmath117 m@xmath118 and @xmath119 , so that the effective peak rabi frequencies are equal to : @xmath120 s@xmath109 , for @xmath121 m@xmath122 , and @xmath123 s@xmath109 .
the resulting values of calculated loss and light shift coefficients are given in table [ losses and light shifts ] .
an important factor in minimizing the most significant loss coefficient @xmath124 and its `` cousin '' term in eq .
( [ condition - gamma-2(1 ) ] ) is the relatively small value of @xmath125 and of the franck - condon overlaps @xmath126 of the closest nearby levels .
while this is favorable for the undesired loss terms , the small value of @xmath125 also affects the strength of the bound - bound coupling of the primary raman transition , @xmath97 , which must be kept large .
however , the small @xmath125-value is compensated here by a strong bare electronic rabi frequency @xmath127 s@xmath109 so that @xmath128 is still large and the adiabaticity condition is maintained . using the above parameter values and simulating the stirap equations with the additional terms given by eqs .
( [ losses1])-([losses3 ] ) , results in about @xmath129 conversion efficiency ( @xmath130 ) , which is rather high and is an encouraging result . in this simulation ,
the estimated optimum pulse duration @xmath82 , the pulse delay coefficient @xmath81 , and the two - photon detuning @xmath131 , were taken as follows : @xmath132 @xmath133 s@xmath109 , @xmath134 , and @xmath135 s@xmath109 .
the effective rabi frequencies and the particle number densities for this calculation are given in figs .
[ multi - level ] ( a ) and ( b ) .
we note that these results are essentially identical whether the exact complex form or the far - off - resonant approximations to the additional coefficients are employed . in other words , any coherent cancellation process via stirap in the extra vibrational levels , is negligible compared to the incoherent losses via non - stirap excitation of these levels .
the final step in our analysis is to include the trap potential and the kinetic energy terms and simulate the full set of coupled inhomogeneous mean - field equations ( [ psi_3 ] ) in three space dimensions .
the initial state in these simulations is a pure atomic bec , with no molecules present , as given by the standard steady - state gross - pitaevskii equation in a trap .
we consider spherically symmetric trap potentials @xmath136 and choose the trap oscillation frequencies @xmath137 equal to each other : @xmath138 hz ( @xmath29 ) . including these terms , we simulate eqs .
( [ psi_3 ] ) assuming that the initial peak density of the atomic bec is @xmath139 m@xmath122 at the trap center .
this corresponds to the total initial number of atoms @xmath140 equal to @xmath141 .
we note that the characteristic time scale associated with the trap potential can be estimated as @xmath142 s. the time scale associated with the kinetic energy term is estimated from the ` healing ' length @xmath143 which corresponds to a healing time scale of @xmath144 . under adiabatic conversion of an equilibrium ( thomas - fermi like ) atomic bec , this healing time scale coincides with the dephasing time
@xmath145 associated with the mean field energy potential and discussed before .
both these time scales ( @xmath146 and @xmath144 ) are longer than the duration of pulses in stirap we employed earlier .
therefore , addition of these terms can not dramatically change the results and conclusions obtained above for uniform condensates . to show this we first consider the idealized three - level model with parameter values given in tables [ stirap_param ] and [ table_u ] , i.e. the case of equal and relatively strong effective rabi frequencies .
the results of simulations of eqs .
( [ psi_3 ] ) are shown in figs .
[ density ] and [ 3d - numbers ] , where we see a rather high conversion efficiency @xmath147 .
this should be compared with the uniform condensates result of fig .
[ conversion93 ] .
next , we consider the case of more realistic parameter values ( lower rabi frequencies ) , and include the effects of incoherent radiative losses and dephasings as described in the previous sub - section . using the calculated values of all relevant parameters for the raman transitions tuned to @xmath105 and @xmath110 , and adding the trap potential and kinetic energy terms to the earlier eqs .
( [ losses1])-([losses3 ] ) , we simulate the process of stirap governed by the resulting full set of mean - field equations in three dimensions : @xmath148 @xmath149 @xmath150 we switch on the sequence of two raman pulses , as shown in fig .
9 , corresponding to : @xmath120 s@xmath109 , @xmath123 s@xmath109 , @xmath132 @xmath133 s @xmath109 , and @xmath134 .
the values of @xmath82 and @xmath81 , together with the choice of the two - photon detuning @xmath135 s@xmath109 , correspond to the optimized set of parameters as in table [ table_main ] and are obtained for @xmath105 and @xmath110 levels using @xmath151 s@xmath109 and @xmath127 s@xmath109 .
the corresponding values of all other relevant coefficients are given in table [ losses and light shifts ] and [ table_u ] , while the spontaneous decay rate is @xmath152 s@xmath109 , as before .
the optimum value of @xmath131 accounts for nonzero values of the light shift coefficients @xmath153 , so that the actual detuning to be optimized using the table [ table_main ] is the overall effective detuning @xmath154 .
the results of simulations with this set of parameter values are given in fig .
[ final - numbers ] where we plot the total occupation numbers in the atomic and molecular becs as a function of time .
the obtained conversion efficiency is @xmath155 .
this is @xmath156 lower than the efficiency in the corresponding homogeneous case of fig .
[ multi - level ] , but still is a rather encouraging result given the fact that @xmath157 conversion of about @xmath158 atoms would give a molecular bec with the total of @xmath159 molecules , with a peak density of about @xmath160 m@xmath122 .
to summarize , stirap is a potential route towards the coherent conversion of an atomic to a molecular bec .
this process involves stimulated emission of molecules , and is very different to normal chemical kinetics .
as such , it is a form of ` superchemistry ' @xcite .
stirap in an atomic bec can be treated in a very similar way to normal stirap , by introducing an effective rabi frequency for the first photo - association transition .
a potential difficulty is the relatively low values of the effective rabi frequency in the first or photo - association transition .
this can be regarded as physically due to the low densities of atoms in a typical weakly interacting bec , compared with atoms in a molecule .
this means that the corresponding franck - condon coefficient , after multiplying by the relevant bec amplitude , has a very small value .
thus , the laser intensities required may be quite high , in order to obtain rabi frequencies comparable to those used in atomic transitions .
this by itself is not critical , since deep in the adiabatic limit it is normally permissible to use low rabi frequencies , as long as the associated time - scales are long enough . from this point of view
, the use of stirap , and consequent reduction of spontaneous emission , is a physically sensible idea .
however , including @xmath0-wave scattering or mean - field processes into the model for stirap sets up a characteristic two - photon dephasing time scale , so that the pulse durations in stirap have to be _
shorter _ than a certain critical value .
short pulse durations necessarily involve high values for the effective rabi frequencies in the usual symmetric case of @xmath161 , thus requiring a very high laser power for the first transition - if the adiabaticity condition is to be maintained . in order to ease this demanding requirement and be able to achieve highest possible conversion efficiency at smaller total laser power
, we propose to use an off - resonance operation ( thus cancelling part of the mean - field detuning effect ) and effective rabi frequencies @xmath162 and @xmath128 of different magnitudes .
this has the effect of dramatically increasing coherent molecule production , in a physically accessible regime of moderate laser intensity .
further improvements may be possible by tailoring the input pulse frequencies to the time - dependent two - photon detuning , caused by inter - atomic and intermolecular scattering .
we have carried out mean - field calculations in three dimenisons to verify that trap inhomogeneity should not have adverse effects on the stirap process .
finally , we stress the importance of radiative losses and dephasing due to incoherent processes that occur during stirap .
these processes are usually assumed to be negligible in ordinary stirap between purely atomic or molecular states . in the present case of coupled atomic / molecular becs ,
this assumption can not be easily justified since the free - bound transition typically involves a relatively low effective rabi frequency , which necessitates long pulse durations .
instead , we find that the incoherent couplings can be rather destructive unless special care is taken to minimize their effect .
this involves detailed knowledge of the structure of the free - bound and bound - bound transitions and subsequent identification of optimum target levels in stirap , so that the overall conversion efficiency remains comparable to the predictions of the simplified three - level model .
d. and k. k. gratefully acknowledge the arc for the support of this work .
d. h. and r. w. acknowledge the support by the u. s. national science foundation , the r. a. welch foundation , and the nasa microgravity research program .
10 d. j. heinzen , r. wynar , p. d. drummond , and k. v. kheruntsyan , phys .
lett . * * 8**4 , 5029 ( 2000 ) . k. v. kheruntsyan and p. d. drummond , phys .
a * * 5**8 , r2676 ( 1998 ) .
p. d. drummond , k. v. kheruntsyan , and h. he , phys .
rev . lett . * * 8**1 , 3055 ( 1998 ) . j. javanainen and m. mackie , phys .
a * * 5**9 , r3186 ( 1999 ) .
r. h. wynar , r. s. freeland , d. j. han , c. ryu , and d. j. heinzen , science * * 2**87 , 1016 ( 2000 ) . y. b. band and p. s. julienne , phys .
a * * 5**1 , r4317 ( 1995 ) ; a. vardi , d. abrashkevich , e. frishman , and m. shapiro , chem . phys . * * 1**07 , 6166 ( 1997 ) ; j. m. vogels _ et al .
rev . a * * 5**6 , r1067 ( 1997 ) ; p. s. julienne , k. burnett , y. b. band , and w. c. stwalley , phys .
a * * 5**8 , r797 ( 1998 ) ; r. ct and a. dalgarno , chem .
phys . lett . * * 2**79 , 50 ( 1997 ) .
p. tommasini , e. timmermans , m. hussein , and a. kerman , cond - mat/9804015 ; e. timmermans _ et al .
lett . * * 8**3 , 2691 ( 1999 ) ; v. a. yurovsky , a. ben - reuven , p. s. julienne , and c. j. williams , phys .
a * * 6**0 , r765 ( 1999 ) ; f. a. van abeelen and b. j. verhaar , phys .
lett . * * 8**3 , 1550 ( 1999 ) ; f. h. mies , e. tiesinga , and p. s. julienne , phys .
a * * 6**1 , 022721 ( 2000 ) .
e. tiesinga , b. j. verhaar , and h. t. c. stoof , phys .
a. * * 4**7 , 4114 ( 1993 ) ; e. tiesinga , a. j. moerdijk , b. j. verhaar , and h. t. c. stoof , phys .
a. * * 4**6 , r1167 ( 1992 ) ; a. j. moerdijk , b. j. verhaar , and a. axelsson , phys . rev .
a * * 5**1 , 4852 ( 1995 ) ; j. m. vogels _ et al_. , phys .
a * * 5**6 , r1067 ( 1997 ) ; s. inouye _ et al_. , nature * * 3**92 , 151 ( 1998 ) ; ph .
* 8**1 , 69 ( 1998 ) ; j. l. roberts _ et al_. , phys .
_ _ * * 8**1 , 5109 ( 1998 ) ; j. stenger _
lett . * * 8**2 , 2422 ( 1999 ) . m. mackie , r. kowalski , and j. javanainen , phys .
lett . * * 8**4 , 3803 ( 2000 ) .
p. o. fedichev _
et al . _ ,
phys . rev .
lett . * * 7**7 , 2913 ( 1996 ) ; k. burnett , p. s. julienne , and k .- a .
suominen , phys .
lett . * * 7**7 , 1416 ( 1996 ) ; j. l. bohn and p. s. julienne , phys .
a * * 5**6 , 1486 ( 1997 ) .
a. l. fetter and j. d. walecka , _ quantum theory of many - particle systems _ ( mcgraw - hill , new york , 1991 ) ; a. a. abrikosov , l. p. gorkov , and i. e. dzyaloshinski , _ methods of quantum field theory in statistical physics _
( dover , new york , 1963 ) ; n. p. proukakis , k. burnett , and h. t. c. stoof , phys .
a * * 5**7 , 1230 ( 1998 ) .
m. holland , j. park , and r. walser , cond - mat/0005062 .
j. j. hope , m. k. olsen , and l. i. plimak , phys . rev .
a * * 6**3 , 043603 ( 2001 ) .
k. bergman , h. theuer , and b. w. shore , rev .
phys . * * 7**0 , 1003 ( 1998 ) ; k. bergmann and b. w. shore , in _ molecular dynamics and spectroscopy by stimulated emission pumping _ , h - l dai and r. w. field , eds .
( world scientific , 1995 ) .
j. m. vogels _ et al_. , phys .
a * * 5**6 , r1067 ( 1997 ) .
r. h. wynar , phd thesis , university of texas at austin , 2000 ; s. j. j. m. f. kokkelmans and b. j. verhaar , private communication . | the process of stimulated raman adiabatic passage ( stirap ) provides a possible route for the generation of a coherent molecular bose - einstein condensate ( bec ) from an atomic bec .
we analyze this process in a three - dimensional mean - field theory , including atom - atom interactions and non - resonant intermediate levels .
we find that the process is feasible , but at larger rabi frequencies than anticipated from a crude single - mode lossless analysis , due to two - photon dephasing caused by the atomic interactions .
we then identify optimal strategies in stirap allowing one to maintain high conversion efficiencies with smaller rabi frequencies and under experimentally less demanding conditions .
pacs numbers : 03.75.fi , 05.30.jp , 03.65.ge . |
in principle qcd is the fundamental theory that describes hadronic systems , but its non - perturbative character makes a direct application to hadrons difficult , with the exception of the high energy and momentum transfer regime .
fortunately , at low @xmath2 effective field theories based on effective hadronic degrees of freedom have been applied with success .
examples are chiral perturbation theory @xcite , small scale expansion @xcite , soliton models and constituent quark models @xcite .
constituent quark models provided consistent descriptions encompassing the low - energy baryon spectrum , elastic and inelastic form factors , charge and magnetic radii , magnetic moments , axial and pseudoscalar form factors @xcite .
although all those different frameworks may include the essential features of qcd , which are confinement and chiral symmetry , they are only partial simulations of the true underlying theory . recently , significant progress has occurred in calculations of qcd in the lattice , which evaluate the important qcd non - perturbative contributions at low @xmath2 directly from the underlying theory . up to now the applications of lattice qcd
are restricted still to large pion masses ( @xmath3 mev ) corresponding to heavy quarks , and lattice spacings that are one order of magnitude smaller than the size of the nucleon ( @xmath4 fm ) . to extract information on the real world the results
must be extrapolated both to the continuum limit ( @xmath5 ) and to the physical pion mass regime @xcite . however , except for the limit to the physical region , there is no simple way of interpreting the lattice qcd data . in this work
we invert this procedure .
we take here the challenge of describing the lattice data using a quark model .
we start with the constituent quark model obtained from the covariant spectator theory .
our model was presented in refs .
@xcite , where it was adjusted to the experimental data for the four nucleon elastic form factors and the dominant form factor of the @xmath1 transition . in this work
we extend this model to the unphysical region of the lattice data .
although the constituent quark models were originally thought for and applied to the physical limit , a constituent quark model that is simultaneously consistent with the experimental data and lattice qcd data is valuable , because it includes indirectly the constraints of qcd , and therefore satisfies properties of the underlying theory .
a similar procedure was considered in refs .
@xcite using a larger number of parameters .
our model does not include explicit pion cloud effects .
however , as the electromagnetic interaction with the quarks is parametrized according to vector meson dominance ( vmd ) , part of the pion cloud effects are indirectly taken into account .
therefore the model can be applied in the regions where the valence quark degrees of freedom are dominant , and it gives a good description of the nucleon and @xmath6 systems @xcite .
for the nucleon , pion cloud effects are fundamental in the time - like region @xcite , although they are not so important in the region of small positive @xmath2 values @xcite , which is involved in the study of the electromagnetic transition . for the @xmath1 transition , as it happens also to other constituent quark models , without explicit pion cloud effects , the predictions for the magnetic moment , which dominates the transition , deviate necessarily from the data for low @xmath2 , showing that an effective pion cloud within the vmd current is not sufficient .
because of the opening of the @xmath7 channel , an explicit pion cloud contribution must be considered @xcite .
therefore our results in the physical low @xmath2 region give only the constituent quark core contributions @xcite , usually labeled as `` bare '' contributions .
our results for these contributions are consistent with independent calculations of dynamical reaction models @xcite based on hadronic degrees of freedom . as for the e2 and c2 multipoles , in the large @xmath8 limit @xcite
they represent second order corrections and are not considered here . as pion cloud effects in lattice qcd are expected to be small for @xmath9 0.40 gev @xcite , the description of the lattice qcd data appears as the ideal testing ground for our constituent quark model , where pion cloud effects are not explicitly included .
this defines the main goal of this work . at present
, lattice calculations are still restricted to high pion masses @xmath10 350 mev , although technical improvements are increasingly allowing to reach nearer and nearer the physical region @xcite . in this work we took the quenched lattice qcd data from refs .
@xcite . in the unquenched calculations the effects of the sea quarks
are explicitly considered .
although unquenched lattice qcd data is already available , it is not expected that the quenched and unquenched data differ substantially in the region @xmath9 0.40 gev @xcite .
in addition , there are some differences between unquenched data for the @xmath1 transition based on two different unquenched methods at similar pion masses @xcite that still have to be clarified in the future . for these reasons we took the conservative option of using quenched data exclusively .
our new model presented here is based on three specific features : i ) the electromagnetic interaction with the constituent quarks is described within the impulse approximation , and considering vmd ; ii ) the wavefunctions of the quark - diquark system are parametrized by simple monopole factors reduced to the hulthen form , with one or two effective range parameters that balance the details of the short range and the long range behavior of the system ; iii ) the constituent quark magnetic anomalous moment scales with the inverse of its mass , which we write , following @xcite as a function of the current quarks mass .
a current based on vmd is suitable to describe the interaction of the photon with the constituent quark , for the quark - antiquark spin-1 vertex .
depending on the isospin , the intermediate meson pole corresponds to the @xmath11 or the @xmath12 meson , at low @xmath2 , or , in the large @xmath2 regime , to some other effective heavy meson pole @xmath13 .
the photon interaction is then described as proceeding through the production of an intermediate meson state which annihilates subsequently into a quark - antiquark pair .
vmd is successful in the description of the electromagnetic interaction with nucleons .
the nucleon s - state wavefunction includes the correct spin structure for the quark - diquark spin 0 and 1 components , associated to isospin 0 and 1 states , respectively . for the @xmath6 ,
since total isospin is @xmath14 , the s - state wavefunction reduces to the diquark spin 1 with isospin 1 structure . as for the scalar wavefunction
, it takes the phenomenological form _ b ( p , k ) = , [ eqpsis ] where @xmath15 , @xmath16 , with @xmath17 the baryon mass ( for the nucleon or @xmath6 ) , @xmath18 is the diquark mass and @xmath19 is a normalization constant . [ in ref
. @xcite we considered @xmath20 for the nucleon and @xmath21 for the @xmath6 ] . the parameters @xmath23 can be interpreted as yukawa mass or range coefficients that distinguish between two different regimes for the momentum range .
the parametrization of the momentum dependence in terms of @xmath24 absorbs the dependence on the baryon masses @xmath17 .
the range parameters @xmath23 for the nucleon and the @xmath6 were fixed by the nucleon and @xmath1 form factor experimental data @xcite . in the spectator quark model the transition amplitude between a initial ( momentum @xmath25 ) nucleon ( @xmath0 ) and a final ( momentum @xmath26 ) baryon @xmath15 can be written , in impulse approximation @xcite as j^= 3 _ _ k complete baryon wavefunction ( including spin , isospin and momentum dependence ) , @xmath27 is the diquark polarization and @xmath28 the quark current operator dependent of the hadron isospin @xmath29 .
the baryon polarizations are suppressed for simplicity .
the symbol @xmath30 represents the invariant integral in the on - shell diquark moment @xmath31 , with @xmath32 as the diquark energy .
the factor 3 takes account for the isospin symmetrization .
the quark current @xcite takes the general form j_i^&= & ( f_1+(q^2 ) + f_1- ( q^2)_3 ) ^+ + & & ( f_2+(q^2 ) + f_2- ( q^2 ) _ 3 ) .
the constituent quark form factors @xmath33 ( @xmath34 ) are parametrized using a vmd structure , and are normalized according to @xmath35 and @xmath36 , where @xmath37 are defined in terms of the quark anomalous magnetic moments @xmath38 and @xmath39 according with @xmath40 and @xmath41 . see refs .
@xcite for details . in this first calculation
the wavefunctions are reduced to s - states for the quark - diquark system . although it is well known that angular momentum components besides the s - states are essential for the description of the nucleon in deep inelastic scattering , their effects are not so evident in the elastic form factors , particularly for low @xmath2 @xcite . as for the @xmath6 , calculations of the valence quark contribution associated with higher angular momentum states ( d - states ) suggest a small effect @xcite . in fact , even with only s - wave components in the wavefunctions for both the nucleon and the @xmath6 , the model generates the dominant contribution for the @xmath1 transition , the dipole magnetic moment form factor .
the non - zero contributions for the other form factors appear only when d states are included , or when pion cloud effects are taken in consideration @xcite . at the end , we will estimate the effect in the @xmath1 quadrupole transition form factors , in the lattice regime , from adding the d - states to the present model . as shown in @xcite
the diquark mass scales out from the formulas obtained for the electromagnetic form factors .
this allowed us to ignore any explicit dependence on the quark mass .
this mass dependence was present in the nucleon , @xmath6 and vector meson masses in an implicit way only .
however , in order to grant a comparison of our results with the lattice data , we extend here the model to include a dependence of the quark anomalous moment @xmath38 and @xmath39 on the quark mass .
this dependence was not explicitly considered in refs .
@xcite because they dealt only with the physical data . inspired by ref .
@xcite , which combines chiral symmetry with conventional quark models , and applies an analytic continuation of the chiral expansion to the simple su(6 ) model , we then use the smooth variation of the hadronic properties with the current quark masses above 60 mev .
therefore , in the spirit of the constituent quark models , we assume here that @xmath42 ( @xmath43 ) scales with @xmath44 , where @xmath45 is the constituent quark mass .
labeling the quark anomalous moment at the physical point by @xmath46 , we can write @xmath42 , for an arbitrary constituent quark mass @xmath45 , as _ q= _ q^0 , [ eqkq ] where @xmath47 is the constituent quark mass at the physical point . to include an explicit dependence on the quark mass we consider the parametrization due to cloet _ et .
@xcite m_q = m_+ cm_q = m_+ c m_q^phy , [ eqmq ] where @xmath48 ( @xmath49 ) is the ( physical ) current quark mass , @xmath50 is a new parameter corresponding to the constituent quark mass in the chiral limit ( @xmath51 ) , and @xmath52 is a coefficient of the order of the unity . in the same equation @xmath53 stands for the pion mass in the model , a parameter to be fixed by the lattice data , and @xmath54 for the physical mass . following ref
. @xcite we considered @xmath55 mev .
the parametrization in ( [ eqmq ] ) is most sensitive to @xmath50 .
reference @xcite fixes @xmath56 gev .
different descriptions using quark models and different lattice sizes lead to different values @xcite . for this reason we use @xmath50 as the only free parameter allowed to vary in the calculation presented here .
it is needed to introduce an explicit dependence of the nucleon magnetic moment on the pion mass , and to enable the connection of the constituent quark model to the lattice qcd calculations @xcite .
we consider three different cases .
first we considered the case @xmath57 , corresponding to the limit where @xmath42 has no dependence on @xmath53 .
we tested also the original parametrization @xmath58 0.42 gev @xcite .
finally , although we did not performed a systematic fit , we tested several other values of @xmath50 .
furthermore , to extend the spectator model to the region of the quenched lattice qcd data we still need to consider the nucleon and @xmath6 masses determined by the lattice simulations @xcite .
this brings to the calculation an implicit dependence on the pion mass . as for the vectorial mesons included in the vmd , quark current picture , we use the parametrization @xcite m_= c_0 + c_1 m_^2 , [ eqmrho ] where @xmath59 gev and @xmath60 gev@xmath61 . the simple parametrization ( [ eqmrho ] ) describes well the quenched lattice qcd data and is consistent with finite volume corrections @xcite . in this applications we consider this parametrization together with model ii of the refs .
@xcite , where the heavy vectorial meson mass is @xmath62 , with @xmath63 the nucleon mass of the lattice calculations .
in fig . [ fignucleonff ] we compare the predictions of our model for the nucleon form factor with the lattice data from ref .
@xcite , corresponding to the three values of @xmath50 .
we consider in particular the isovector nucleon form factor because the contributions of the disconnected diagrams vanishes in lattice calculations , if the flavor su(2 ) symmetry is assumed as in ref .
we considered the lowest pion masses and the smallest lattice spacings .
in particular , we take the data corresponding to @xmath64 504 mev , 649 mev ( @xmath65 fm ) and @xmath66 mev ( @xmath67 fm ) . for larger lattice spacings a dependence on @xmath68 is observed @xcite . from the figure
we conclude that @xmath56 gev and 0.80 gev gives an excellent description of the nucleon data .
still , when the pion mass increases there is a systematic deviation of our model from the lattice data of ref .
this deviation may be a consequence of the fact that the wavefunction parametrization in terms of a low momentum scale ( @xmath69 ) ( long range behavior ) and high momentum scale ( @xmath70 ) was kept unchanged .
as the pion masses vary , at least the long range parameter may vary , and may vary more rapidly for larger pion masses . as for the short range parameter ( @xmath70 ) it sets the scale below which the short range physics becomes important . as observed in applications of the finite - range regularization effective field theory , the scale associated to the momentum cut - off ( and consequently short range effects )
can be expressed by an universal regulator which describes simultaneously the large pion mass regime and the physical regime @xcite . in our calculation @xmath70
is related to that universal regulator , and it is expected not to depend crucially on the pion mass value . for a finer analysis of our results , we compare our predictions for the isovector nucleon magnetic moment in physical nucleon magnetons , with the lattice qcd data and the chiral result from ref .
@xcite . to convert @xmath71 given by the lattice data in units @xmath72
( @xmath63 nucleon mass in lattice ) to `` physical '' units @xmath73 , we need to use the transformation @xmath74 .
the results are presented in the fig .
[ figgmv ] for @xmath56 gev and @xmath75 gev , as function of @xmath53 .
the agreement of our results with the chiral expression from ref .
@xcite shows the consistency of our calculations with chiral calculations .
finally , in fig . [ figndeltaff ] we compare the quenched lattice data for the @xmath1 transition from ref .
@xcite with the spectator quark model corresponding to @xmath76 gev , 0.80 gev and @xmath77 .
all cases shown correspond to a lattice spacing of @xmath78 fm .
the contribution of the quark core extracted from @xcite at the physical point is also included . to be consistent , we considered the parametrization of eq .
( [ eqmrho ] ) for @xmath79 , although the original ref .
@xcite gives slightly different results .
as for @xmath63 and @xmath80 , we use the values derived directly from the lattice data @xcite . for the larger pion masses
we observe an almost perfect agreement between the predictions of @xmath81 and the data , although @xmath82 gev is also close . for @xmath83 gev
we have also a good agreement , except for a slight deviation from the lattice data for @xmath84 gev@xmath85 2 gev@xmath86 .
this deviation may result from the effect of the pion cloud for light pions , predicted to be important for @xmath87 gev @xcite .
as for the physical pion mass case , our model is coherent with the constituent quark core data labeled bare form factor , which is extracted indirectly from experiment @xcite . in this case all lines coincide . as mentioned already
, the d - states in the @xmath6 wavefunction induce contributions for the subleading electric and coulomb quadrupole transition form factors @xmath88 and @xmath89 .
the exact contributions depend on the specific parametrization , in particular on the admixture coefficients for the two d - waves . in ref .
@xcite it was shown how a percentage of d - states of @xmath90 can provide an excellent description of the quadrupole lattice data , without significantly changing the description of the dominant magnetic dipole form factor .
that application corresponds to the @xmath91 limit and estimates the effect of the valence quark ( or bare ) contributions as only @xmath92 of the total quadrupole for both @xmath88 and @xmath89 , at the physical point ( the remaining contribution being the pion cloud ) . on the other hand , in that particular parametrization , and in the region of study @xmath93 gev@xmath86 , the effect of the constituent quark mass dependence described by eq .
( [ eqkq ] ) corresponds to a correction of less than 20% for both form factors , a correction smaller than the typical lattice errorbands ( @xmath94 ) .
chiral perturbation methods for lattice qcd extrapolations are useful for small @xmath2 ( like @xmath95 gev@xmath86 ) , but are not adequate for the high @xmath2 region .
constituent quark models can supply an alternative guidance for lattice qcd extrapolations . in this work we consider a covariant constituent quark model of the nucleon and the @xmath6 based on the spectator formalism for the quark - diquark system ( covariance is an important issue for the description of the high momentum transfer processes ) .
the model contains no explicit pion cloud , besides the effects included in the @xmath11 term of the vdm picture for the electromagnetic current , and is therefore reduced to the bare quark hadron structure . since pion cloud effects are expected to be negligible for large values of the pion mass , the comparison of the model to the quenched data for pion masses larger than 450 mev
is justified a priori . to accomplish that comparison
the initial covariant quark model was extended to the lattice data region , @xmath96 mev . in light of the work in ref .
@xcite , this extension was done by introducing the constituent quark mass in the chiral limit parameter , @xmath50 , and the nucleon @xmath11 , @xmath6 masses used in the lattice calculations .
the main conclusion is that the covariant constituent quark model which was previously calibrated by means of a quantitative description of the nucleon , @xmath1 and @xmath6 form factor data in the physical region , as shown in previous works @xcite , after a simple extension involving one parameter only , describes also quantitatively well the lattice results for the nucleon isovector form factor and the @xmath1 `` bare '' magnetic form factor . with @xmath56 gev , consistently with the range @xmath97 gev suggested by several constituent quark models @xcite , we obtain a very good description of the lattice nucleon form factors data , but underestimate the @xmath1 magnetic moment form factor data by less than 9% at low @xmath2 for @xmath98 mev .
note that the original parametrization @xmath56 gev @xcite was a result of a phenomenological fit and was not derived from first principles .
an optimal description of the lattice data for both nucleon and @xmath1 transition form factors is achieved , once the scale of the constituent quark mass in the chiral limit is fixed as @xmath75 gev . [ the values @xmath56 and 0.80 gev are better for the nucleon data ; @xmath75 gev and @xmath99 are better for the @xmath1 data ] .
the parameter @xmath75 gev is relatively large when compared with alternative constituent quark models , which may indicate that some of the effective quark - antiquark configurations contained in our model through the vmd mechanism in the electromagnetic current , for instance , do not correspond to the lattice calculations in the quenched approximation .
still , the possibility of adjusting other parameters is very promising .
a refinement of the description can be pursued , by increasing the number of adjustable parameters , or simply by introducing an extra dependence on the pion mass .
for instance , in the future we certainly plan to check the assumption that the @xmath100 parameter which controls the long range regime does not depend on the pion mass value .
in addition , we may consider as well an explicit dependence of the heavier meson mass pole @xmath13 , that regulates the short range effect in vmd mechanism , on the pion mass . alternatively
, a pion mass dependence of the @xmath101 parameter can be introduced .
furthermore , we want to study the quality of the description in the high pion masses region , not probed yet here .
the authors are particularly thankful to gckeler for supply information about ref .
g. r. would like to thank specially to ross young for the detailed explanations of the lattice proprieties and the extrapolations for the real world .
g. r. also thank franz gross , ian cloet , michael pardon , anthony thomas and ping wang for helpful discussions .
this work was partially supported by jefferson science associates , llc under u.s .
doe contract no .
de - ac05 - 06or23177 .
g. r. was supported by the portuguese fundao para a cincia e tecnologia ( fct ) under grant no .
sfrh / bpd/26886/2006 .
this work has been supported in part by the european union ( hadronphysics2 project `` study of strongly interacting matter '' ) .
b. julia - diaz , d. o. riska and f. coester , phys .
c * 69 * , 035212 ( 2004 ) [ erratum - ibid .
c * 75 * , 069902 ( 2007 ) ] [ arxiv : hep - ph/0312169 ] .
s. boffi , l. y. glozman , w. klink , w. plessas , m. radici and r. f. wagenbrunn , eur .
phys . j. a * 14 * , 17 ( 2002 ) [ arxiv : hep - ph/0108271 ] ; r. f. wagenbrunn , s. boffi , w. klink , w. plessas and m. radici , phys . lett .
b * 511 * , 33 ( 2001 ) [ arxiv : nucl - th/0010048 ] .
m. m. giannini , e. santopinto and a. vassallo , eur .
j. a * 12 * , 447 ( 2001 ) [ arxiv : nucl - th/0111073 ] .
e. pace , g. salme , f. cardarelli and s. simula , nucl .
a * 666 * , 33 ( 2000 ) [ arxiv : nucl - th/9909025 ] ; f. cardarelli and s. simula , phys .
c * 62 * , 065201 ( 2000 ) [ arxiv : nucl - th/0006023 ] .
t. sato and t. s. h. lee , phys .
c * 63 * , 055201 ( 2001 ) [ arxiv : nucl - th/0010025 ] ; s. s. kamalov , s. n. yang , d. drechsel , o. hanstein and l. tiator , phys .
c * 64 * , 032201(r ) ( 2001 ) [ arxiv : nucl - th/0006068 ] . | a covariant quark model , based both on the spectator formalism and on vector meson dominance , and previously calibrated by the physical data , is here extended to the unphysical region of the lattice data by means of one single extra adjustable parameter the constituent quark mass in the chiral limit .
we calculated the nucleon ( @xmath0 ) and the @xmath1 form factors in the universe of values for that parameter described by quenched lattice qcd .
a qualitative description of the nucleon and @xmath1 form factors lattice data is achieved for light pions . |
a pulsar parallax can be combined with a measurement of the pulsar s dispersive delay ( the dispersion measure or dm ) to provide an accurate measure of the free electron density along the line of sight ( los ) .
psr j0030 + 0451 is one of fewer than a dozen pulsars to have its parallax measured via timing @xcite .
another dozen have been measured via vlbi ( e.g. see brisken et al .
2002 , chatterjee et al .
2004 , and also this urlshami / psrvlb / parallax.html ] ) .
these measurements are important because they give us most of our knowledge about the galactic thermal electron distribution @xcite .
in addition , psr j0030 + 0451 presents a rare evolutionary case as an isolated millisecond pulsar ( msp ) . in the most popular msp evolutionary model ,
msps are formed via accretion of matter from a companion star .
the incoming matter adds to the pulsar s angular momentum , i.e. , the pulsar is `` spun up . ''
however , isolated msps present a conundrum : they were presumably spun up , yet they are without a companion which would have done so .
one possible scenario is that the pulsar has ablated its companion @xcite .
we expect msps to have lower velocities than the regular population , because the kick from the supernova progenitor had to be small enough to leave the binary intact . to produce an isolated msp , the binary must remain intact during and after the supernova , but then after the spin - up phase the companion must leave the system or be evaporated .
several authors have debated whether isolated msp velocities are lower , higher , or indistinguishable from those of the general population of msps .
@xcite suggest we might expect isolated msps to have higher velocities .
they argue that if isolated msps are formed by ablation , we would expect them to form from the tighter binaries which are more susceptible to ablation .
the correlation between tight binaries and higher velocities is suggested by @xcite .
@xcite present the argument for faster velocities for isolated msps as a counterpoint to their timing proper motion and scintillation measurements which suggest the opposite , as do the measurements of @xcite and @xcite .
@xcite , however , find the velocities of the populations to be indistinguishable .
we present a measurement of the transverse velocity of psr j0030 + 0451 which is unusually small , even compared to the isolated msp population .
we reconsider the question of the velocity of isolated msps as compared to the binary msp population . owing to its small timing
residual , @xmath31 @xmath4s , psr j0030 + 0451 is a good candidate for membership in the pulsar timing array ( pta ) , which is a collection of pulsars that will be used for detecting gravitational radiation @xcite .
for this reason , continued refinement of the timing model of psr j0030 + 0451 is important .
in fact , the pta , as it is conceived , is an interferometer , so a variety of baselines will be important to its operation .
psr j0030 + 0451 may be particularly useful in this regard because it has large angular separation from psrs b1855 + 09 , j1713 + 07 and j0437 - 4715 , which are among the most stable and precise pulsars @xcite . in 2 we present a significant refinement to the timing model previously published @xcite .
we discuss our measurements of parallax and proper motion in
[ sec : pm_px ] .
we include the effects of the solar wind in our analysis , which we discuss in 4 . in 5,6 , and 7
we discuss the space velocity of j0030 + 0451 , corrections to its measured period derivative , and the implications of its measured distance for the local interstellar medium ( lism ) .
in 8 we summarize our conclusions .
we have conducted timing observations over a 6.5 year period , from 1997 december to 2004 july , at the arecibo observatory , using the arecibo - berkeley pulsar processor ( abpp ) and the princeton mark iv system .
details of abpp signal processing can be found in @xcite , where the first three years of these data are presented .
details of mark iv signal processing can be found in @xcite . over the course of each observation , incoming signals from two orthogonal polarizations
were coherently dedispersed , squared to obtain power measurements , and folded modulo the pulse period for intervals of three minutes .
opposite polarizations were summed , using absolute flux calibrations of the receivers whenever possible or , when absolute flux calibration was not available , using the system temperature of the pulsed noise source as published by telescope operations .
pulse times of arrival ( toas ) were derived from the calibrated three minute integrations using conventional algorithms .
the toas from a given frequency and instrument on a given day were averaged into a single effective toa for that day , the uncertainty of which was estimated from the spread of the individual three - minute toas .
we performed a weighted fit to the averaged toas using tempo .
we used ecliptic coordinates to minimize covariance between the two components of the position and proper motion measurements .
this is usually necessary when the ecliptic latitude is low ( 1.44 in our case ) .
table 1 shows updated spin , astrometric , and other parameters for psr j0030 + 0451 .
the root mean squared ( rms ) of the timing residuals quoted in table 1 , is calculated using toas from profiles averaged over 30 minutes .
we estimated uncertainties in all the fitted parameters by doubling errors given by tempo , an ad hoc procedure that attempts to account for timing noise and other possible systematic errors .
the quoted uncertainty in parallax also includes a term added in quadrature ( 0.3 mas ) due to the solar wind ( see [ sec : electron ] ) .
residual arrival times after subtracting the best fit are shown in figure [ fig : resid ] .
ll ecliptic longitude ( deg ) & 8.91036695(6 ) + ecliptic latitude ( deg ) & 1.445692(3 ) + period ( s ) & 0.00486545320829334(3 ) + period derivative ( ss@xmath2 ) & 1.0162(1)@xmath5 + proper motion in ecliptic longitude ( mas yr@xmath2 ) & @xmath65.74(9 ) + absolute value of proper motion in ecliptic latitude ( mas yr@xmath2 ) & @xmath7 + dispersion measure ( pc @xmath8 ) & 4.3328(2 ) + epoch ( mjd ) & 52035 + solar _ n_@xmath9 ( e@xmath10 @xmath8 ) & 6.9(2.1 ) + solar _ @xmath11_@xmath9 ( e@xmath10@xmath8yr@xmath2 ) & 0.0(3 ) + number of epochs of data & 56 + timing data span ( mjd ) & 50,790 @xmath6 53,280 + right ascension(j2000 ) & 00 30 27.4308(6 ) + declination(j2000 ) & + 04 51 39.72(1 ) + galactic longitude ( deg ) & 113.141135(1 ) + galactic latitude ( deg ) & @xmath657.611237(3 ) + dm derived distance ( pc ) & 317(25 ) + characteristic age ( yr ) & @xmath12 + magnetic field strength ( g ) & @xmath13 + column electron density along los ( e@xmath10 @xmath8 ) & 0.014(2 ) + parallax ( mas ) & 3.3(9 ) + parallax derived distance ( pc ) & 300(90 ) + magnitude of transverse velocity ( km s@xmath2 ) & 8@xmath617 + rms residual at 430 mhz ( @xmath4s , abpp / markiv ) & 2.3/1.0 + rms residual at 1410 mhz ( @xmath4s , abpp / markiv ) & 2.2/2.2 + we examined timing solutions in which dm was allowed to vary over time , and we found that they did not significantly improve the fit quality or change the timing model parameters .
however , it is interesting to compare our upper limit on @xmath14 , the time derivative of dm , to the correlation between dm and @xmath14 measured by @xcite .
when @xmath14 is included in the timing model , its best - fit value is @xmath15 @xmath8pc yr@xmath2 , which implies an upper limit ( with 95% confidence ) of @xmath16 @xmath8pc yr@xmath2 . this value is smaller than expected according to @xcite by about a factor of 5 .
if more small - dm small-@xmath14 pulsars are found it might suggest that @xmath14 is a stronger function of dm than we would expect for wedgelike thermal plasma perturbations distributed randomly along the los .
evidence for a significant parallax measurement is shown in figure [ fig : parallax ] , which compares residuals of the pulse arrival times with and without parallax incorporated into the timing model .
the figure shows averaged timing residuals versus day number binned in increments of 18 days and folded over a half - year period to provide the best visibility for the parallax signature .
the data have been fit for proper motion but not for parallax in the left part of figure [ fig : parallax ] .
data on the right have been fit for both proper motion and for parallax .
we have superimposed the best - fit parallax curve onto the pre - fit data . on close examination of figure [ fig : parallax ]
, there is a subtle difference in the positions of the data points relative to the fit curve ( left plot ) and the horizontal axis ( right plot ) .
this is not an unexpected result of ( a ) a global fit to all the parameters with and without parallax and ( b ) the post - fit binning .
our best - fit value for parallax , @xmath17mas , includes both measurement uncertainty of 0.88 mas and systematic uncertainty due to the solar wind model , 0.3mas , which will be discussed in the following section .
we measure a proper motion of @xmath18mas yr@xmath2 in the plane of the ecliptic
. proper motion out of the ecliptic plane is naturally difficult to measure given the pulsar s position , a problem which was compounded by its high covariance with variations in dispersion measure .
we quote a conservative upper limit on the magnitude of the out - of - plane proper motion of 10 mas yr@xmath2 .
we can combine our distance measurement with our proper motion measurements to yield the pulsar s velocity .
we obtain a velocity of @xmath19 km s@xmath2 in the plane of the ecliptic and an upper limit of @xmath20 km s@xmath2 out of the plane of the ecliptic .
thus , the pulsar s transverse velocity lies between 8 and 17 km s@xmath2 .
this confirms the value presented by @xcite , @xmath21 km s@xmath2 , which they found using scintillation measurements . the measured proper motion of psr j0030 + 0451 results not just from its motion relative to its local standard of rest ( lsr ) , but also from the difference between its lsr and the lsr of the sun and from the solar motion .
after considering these effects ( see 5 for details ) , we calculate its transverse velocity relative to its lsr to be between 4 and 20 kms@xmath2 .
this is one of the slowest transverse velocities measured for any pulsar .
young pulsars ( those which have not been spun up ) have a mean velocity of 265 km s@xmath2 @xcite , a factor of three times faster than msps ( 5 ; see also hobbs et al .
2005 ; cordes & chernoff 1997 ; nice & taylor 1995 ) . psr j0030 +
0451 is , in fact , only one tenth as fast as the average msp ( 5 ) .
psr j0030 + 0451 is roughly in the ecliptic plane ( ecliptic latitude is 1.44 ) where the solar system provides additional dispersion caused by the charged particles of the solar wind .
@xcite show that the solar electron density , _
n_@xmath22 , can be modeled as @xmath23 , where @xmath24 is the radial distance to the sun in astronomical units , and @xmath25 is roughly 10 @xmath8 @xcite .
this simple model is the default in tempo . maintaining the @xmath26 dependence , we searched @xmath27 space for the best value of @xmath25 , which we found to be @xmath28@xmath8 .
reducing @xmath25 from 9.0 to 6.9 @xmath8 reduces the fitted parallax by about 0.3 mas which increases the distance by about 30 pc .
these two parameters are covariant because the model of the solar wind yields an annual pattern of pulse delays with a strong cusp when the pulsar is behind the sun with respect to the earth .
a fourier decomposition of this delay pattern yields significant terms with annual periodicity ( covariant with pulsar position ) , semi - annual periodicity ( covariant with parallax ) , as well as higher order terms .
the additional 0.3 mas uncertainty in parallax due to the solar wind was added in quadrature to the doubled tempo error .
note that in the case of psr j1713 + 0747 , another ecliptic plane msp , @xcite found it necessary to eliminate data within 30 of the sun . in our case
we did not find that this significantly improved the fit , nor did it substantially change any fitted parameters .
we wondered if it would be possible to measure a change in the solar wind on its 11-year cycle . using tempo
we mapped @xmath27 space in the range of @xmath610 @xmath8 yr@xmath29 @xmath8 yr@xmath2 but found that the minimum in @xmath27 occurred at an essentially null value of @xmath30@xmath8 yr@xmath2 , indicating that the change is currently beyond our measurement capability .
@xcite presents an extensive study of the velocities of various sub - groups of pulsars , including isolated and binary recycled pulsars .
they find that isolated recycled pulsars are not significantly slower ( @xmath31 km s@xmath2 ) than binary recycled pulsars ( @xmath32 km s@xmath2 ) .
when the velocity of psr j0030 + 0451 ( using 9 km s@xmath33 ) is added to the sample the average velocity of isolated recycled pulsars becomes @xmath34 km s@xmath2 , which is still not significantly different from the average of the binary recycled pulsars .
in contrast , both @xcite and @xcite previously claimed to see evidence that isolated msps are slower than binary msps .
however , when one uses a more recent galactic electron density model to estimate the distances to the pulsars @xcite , the discrepancy disappears . this is a potent reminder that a distance model has a significant effect on conclusions drawn from it .
@xcite also find the isolated msp population to have a slightly lower average velocity ( 70 km s@xmath2 ) as compared to the binary population .
they caution against drawing conclusions from velocity data which are based on imprecise distances .
we performed an analysis of the velocity data that is slightly different from the analysis done by @xcite with similar results .
we defined an msp to be any pulsar with period @xmath35s , which provides a narrower sample than that used by @xcite who defined msps to be any pulsar with period @xmath36s and period derivative @xmath37 .
our criterion means that the binary pulsars in our sample nearly all have helium white dwarf companions , whereas hobbs et al .
used binaries with a mix of companion types .
the 29 msps in the galactic disk with measured proper motions are shown in table [ tab : velocities ] .
psr j1730 - 2304 , which has no measured declination proper motion , has been included in the table for completeness but has not been included in any of the following calculations .
we corrected each pulsar s velocity to its lsr as follows .
we used the measured proper motion and distance to calculate a three dimensional vector representing the ( two dimensional ) transverse motion of the pulsar in the reference frame of the sun .
we then removed the solar motion and rotated the resulting vector from the lsr of the sun to the lsr of the pulsar .
finally , we recovered those two components of the vector which are perpendicular to the line of sight .
this computation required selecting a value for the unknown los velocity of the pulsar ; we chose a value appropriate for a star at rest in the pulsar s lsr .
r@lr@llr@llrrl j & 0030 + 0451 & @xmath38 & .84 & @xmath390.09 & @xmath40 & & & 310 & @xmath4120 & this work + j & 0711@xmath66830 & @xmath615&.7 & @xmath390.5 & 15&.3 & @xmath390.6 & 860 & 113 & @xcite + j & 1024@xmath60719 & @xmath641&&@xmath392 & @xmath670 & & @xmath393 & 200 & 70 & @xcite + j & 1730@xmath62304 & 20&.5 & @xmath390.4 & & & 510 & @xmath4250 & @xcite + j & 1744@xmath61134 & 18&.64 & @xmath390.08 & @xmath610&.3 & @xmath390.5 & 360 & 31 & @xcite + b & 1937 + 21 & @xmath60&.130 & @xmath390.008 & @xmath60&.469 & @xmath390.009 & 3600 & 87 & @xcite + j & 1944 + 0907 & 12&.0 & @xmath390.7 & @xmath618 & & @xmath393 & 1800 & 173 & @xcite + j & 2124@xmath63358 & @xmath614&&@xmath391 & @xmath647 & & @xmath391 & 270 & 48 & @xcite + j & 2322 + 2057 & @xmath617&&@xmath392 & @xmath618 & & @xmath393 & 790 & 79 & @xcite + j & 0437@xmath64715 & 121&.438 & @xmath390.006 & @xmath671&.438 & @xmath390.007 & 140 & 84 & @xcite + j & 0613@xmath60200 & 2&.0 & @xmath390.4 & @xmath67 & & @xmath391 & 1700 & 60 & @xcite + j & 0751 + 1807 & @xmath43&.35&@xmath390.03 & @xmath44&&@xmath392 & 1150 & 22 & @xcite + j & 1012 + 5307 & 2&.4 & @xmath390.2 & @xmath625&.2 & @xmath390.2 & 840 & 107 & @xcite + j & 1045@xmath64509 & @xmath65&&@xmath392 & 6 & & @xmath391 & 1940 & 119 & @xcite + j & 1455@xmath63330 & 5&&@xmath396 & 24 & & @xmath3912 & 530 & 71 & @xcite + j & 1640 + 2224 & 1&.66 & @xmath390.12 & @xmath611&.3 & @xmath390.2 & 1160 & 67 & @xcite + j & 1643@xmath61224 & 3&&@xmath391 & @xmath68 & & @xmath395 & 2320 & 96 & @xcite + j & 1709 + 2313 & @xmath63&.2 & @xmath390.7 & @xmath69&.7 & @xmath390.9 & 1390 & 57 & @xcite + j & 1713 + 0747 & 4&.917 & @xmath390.004 & @xmath63&.933 & @xmath390.010 & 1100 & 30 & @xcite + b & 1855 + 09 & @xmath62&.94 & @xmath390.04 & @xmath65&.41 & @xmath390.06 & 910 & 17 & @xcite + j & 1909@xmath63744 & @xmath69&.6 & @xmath390.2 & @xmath635&.6 & @xmath390.7 & 820 & 131 & @xcite + j & 1911@xmath61114 & @xmath66&&@xmath394 & @xmath623 & & @xmath3913 & 1220 & 128 & @xcite + b & 1953 + 29 & @xmath61&.0 & @xmath390.3 & @xmath63&.7 & @xmath390.3 & 4610 & 128 & @xcite + b & 1957 + 20 & @xmath616&.0 & @xmath390.5 & @xmath625&.8 & @xmath390.6 & 2490 & 325 & @xcite + j & 2019 + 2425 & @xmath69&.41 & @xmath390.12 & @xmath620&.60 & @xmath390.15 & 1490 & 142 & @xcite + j & 2051@xmath60827 & 1 & & @xmath392 & @xmath65 & & @xmath393 & 1040 & 42 & @xcite + j & 2129@xmath65721 & 7&&@xmath392 & @xmath64 & & @xmath393 & 1340 & 48 & @xcite + j & 2229 + 2643 & 1 & & @xmath394 & @xmath617 & & @xmath394 & 1440 & 130 & @xcite + j & 2317 + 1439 & @xmath61&.7 & @xmath391.5 & 7&.4 & @xmath393.1 & 820 & 20 & @xcite + this corrected velocity is listed in the second to last column of table 3 .
the average corrected velocity of the isolated msps is @xmath45 km s@xmath2 whereas the average corrected velocity of all the binary msps is @xmath46 km s@xmath2 .
( the uncorrected averages are 79 and 90 km s@xmath2 respectively ) .
if one allows the sample to include only those proper motions which have been measured to better than 2@xmath47 the average corrected velocities are @xmath45 km s@xmath2 and @xmath48 km s@xmath2 respectively .
( uncorrected averages are 79 and 99 km s@xmath2 . ) in each case the isolated msp population is indistinguishable from the binary msp population .
the 2@xmath47 cutoff in velocity introduces a selection of higher velocity pulsars .
thus , the average velocities are higher in that case .
an alternative statistic for evaluating the dynamics of pulsar populations is the distribution of heights above or below the galactic plane , @xmath49 . for the pulsars listed
, one finds that the standard deviation from zero for the binary msp population is twice that of the isolated msp population : @xmath50pc vs @xmath51pc .
figure [ fig : zheight ] shows a histogram of @xmath49 for each population .
the isolated msp population is represented in the upper half of the figure , the binary msp population in the lower half .
figure [ fig : zheight ] shows that the known isolated msps are closer to the plane than are the known binary msps .
this could be either a reflection of differences in the intrinsic spatial distributions of the two types of msps , or a selection effect .
a smaller intrinsic spread in scale heights for isolated msps is only possible if that population also has a smaller intrinsic velocity distribution , so that the objects do not travel as far from the plane as they oscillate in the galaxy s potential .
our determination that the two velocity distributions are in fact indistinguishable makes this scenario unlikely .
however , with identical velocity distributions , a difference in intrinsic _ luminosity _ distributions would cause the less - luminous population to be detected only to smaller distances and hence only to smaller scale heights .
in fact , @xcite find that luminosities of isolated and binary msps are different at the 99.5% confidence level , with the isolated msps being intrinsically dimmer .
we have confirmed their results with an updated catalog ; also , a simple examination of the median distance of the isolated population ( 510 pc ) compared to the median distance of the binary population ( 1155 pc ) suggests that the isolated msps must be less luminous .
using our upper limit on the proper motion , we can calculate the upper limit of the shklovskii correction to the period derivative @xmath52 .
we find that @xmath53 or not quite 5% of the measured @xmath52 may be due to proper motion .
the acceleration toward the disk of the galaxy is about the same size in the other direction , @xmath54 .
the acceleration in the disk makes a much smaller contribution , @xmath55 . combining these corrections changes the measured @xmath52 by less than 1% .
the parallax measurement of @xmath56mas yields a distance of @xmath57 pc .
this falls in between the distance estimates made by the old ( taylor & cordes 1993 , hereafter tc93 ) and new ( cordes & lazio 2002 , hereafter ne2001 ) distance models : 230 and 317 respectively .
both estimates are based on a dm of @xmath58@xmath8pc @xcite . in the second galactic quadrant ,
between 90@xmath59 and 180@xmath59 galactic longitude , psr j0030 + 0451 is the first pulsar with a measured parallax , so it provides an important check of the ne2001 model .
in addition to a parameterized model of the lism , the model adds a number of clumps and voids to the previous tc93 model , but nothing in the direction of psr j0030 + 0451 .
the agreement of our distance with theirs suggests no significant clumps or voids exist along this los .
nearby pulsars with known parallaxes are very useful for studying the lism .
scintillation parameters measure density fluctuations and have been used to map out the lism , but are unable to measure densities .
the density measurements must come from parallax @xcite .
these authors model the lism explicitly as a low density bubble surrounded by a shell of much higher density fluctuations , but more pulsars with known distances are required to confirm the model .
psr j0030 + 0451 will therefore be a marvelous tool for studying the lism .
we have measured the parallax of psr j0030 + 0451 to be @xmath60mas .
we have measured its proper motion to be @xmath18 mas yr@xmath2 in the plane of the ecliptic and have established an upper limit on its motion out of the plane at 10 mas yr@xmath2 .
the is one of the lowest velocities measured for any pulsar and is noteworthy even within the relatively low - velocity millisecond pulsar population . combining proper motion data from this pulsar with the collection of existing msp proper motion measurements , we find the statistical properties of the transverse velocities of isolated and binary msps are indistinguishable from each other .
we do , however , find that the average @xmath49-height of the isolated msps is half that of the binary msps .
we suggest that a luminosity difference between the two classes of objects , such as that suggested by @xcite , @xcite , and @xcite , is the simplest way to account for both the observed difference in @xmath49-height and the similarity of the velocity distributions .
we are grateful to the arecibo telescope operators .
we thank robert ferdman , paul demorest , paulo freire , duncan lorimer , ramachandran , and kiriaki xilouris , for valuable discussions and for assisting with observations .
we thank the referee , simon johnston , for substantial comments that significantly improved the manuscript .
the arecibo observatory is a facility of the national astronomy and ionosphere center , operated by cornell university under a cooperative agreement with the national science foundation ( nsf ) .
anl acknowledges a research corporation award in support of this research .
djn is supported by nsf grant ast-0206205 .
ihs holds an nserc ufa and is supported by a discovery grant .
dcb acknowledges support from nsf ast-9987278 for abpp instrumentation and nsf ast-0206044 for the science program . | we report the parallax and proper motion of millisecond pulsar j0030 + 0451 , one of thirteen known isolated millisecond pulsars in the disk of the galaxy .
we obtained more than 6 years of monthly data from the 305 m arecibo telescope at 430 mhz and 1410 mhz .
we measure the parallax of psr j0030 + 0451 to be @xmath0mas , corresponding to a distance of @xmath1pc .
the cordes and lazio ( 2002 ) model of galactic electron distribution yields a dispersion measure derived distance of 317 pc which agrees with our measurement .
we place the pulsar s transverse space velocity in the range of 8 to 17 km s@xmath2 , making this pulsar one of the slowest known .
we perform a brief census of velocities of isolated versus binary millisecond pulsars .
we find the velocities of the two populations are indistinguishable .
however , the scale height of the binary population is twice that of the isolated population and the luminosity functions of the two populations are different .
we suggest that the scale height difference may be an artifact of the luminosity difference . |
when measuring the intensity of optical radiation , the registered signal is inevitably burdened with photon ( poisson ) noise , and the statistics concerning the received signal differ from those of incident radiation .
this noise is not additive and in the case of fluctuating incident radiation its power changes depending on the radiation s instantaneous intensity .
the problem concerning the contribution of photon noise in the measured statistical characteristics is considered a classical problem that has been solved many years ago .
particular interest in the problem was triggered during the interpretation of the non - classical interference of light , study of raman scattering , and photon correlation spectroscopy ( see for e.g. , @xcite ) .
calculating the contribution of photon noise is important to several experimental works , particularly those that involve the photon counting method in low - light applications .
it has been shown that the registered mean , variance , and correlations are related with the corresponding set of statistics for stochastic radiation in a simple manner .
a simple relationship called the mandel relation @xcite connects the moments of any order of fluctuating intensity with the corresponding factorial moments of photon counts .
the mandel relation is a convenient expression that has been repeatedly used , including in the study of stellar scintillation statistics @xcite .
however , this relation is valid for an ideal detector while practical photoelectric detectors are not ideal .
firstly , the photon - counting method involves nonlinearity caused by the presence of dead time in the registration chain .
for example , in the measurements of stellar scintillation with the multi - aperture scintillation sensor ( mass ) @xcite , accounting for nonlinearity is an essential step in signal processing @xcite , because in order to achieve the necessary statistical accuracy , we are forced to choose bright stars .
the attenuation of the light flux is not a solution , since in this case , a well - known effect that a single - photon detector may have after - pulses of different nature , becomes more noticeable .
this effect also distorts the statistics of registered events .
the after - pulse effect has not thus far been properly studied from the theoretical viewpoint , and ill - grounded empirical arguments are used to describe the effect . in this study , we present analogues of the mandel relation , which connect the moments of a stochastically fluctuating light flux ( hereafter reffered to as scintillation ) with the factorial moments of the registered counts for the case of noticeable nonlinearity of any given photon counter .
the expressions are derived for both non - paralyzable and paralyzable dead time , i.e. , counters of both the first ( i ) and the second ( ii ) types .
the relations consider the effect of after - pulses as well .
experimental results confirming our theoretical conclusions are presented in sec .
[ sec : exper ] .
the applicability of the relations is tested using actual measurements of stellar scintillation .
finally , we present simple algorithms to solve the practical problem of the determination of the statistical moments of scintillation from measured data .
the measurement of statistical characteristics involves a large number of exposures with a exposure time @xmath0 over a time interval @xmath1 .
thus , the number of discrete counts is given as @xmath2 .
each count @xmath3 is a random variable , possibly correlated with adjacent counts .
we assume that the scintillation is a stationary random process , and that it is described by the probability distribution @xmath4 of the scintillation @xmath5 which is the normalized flux incident on the detector .
the value of @xmath6 is essentially non - negative , its expectation @xmath7 = 1 $ ] , and its variance @xmath8 = s^2 $ ] is the scintillation index . in the ideal case ,
the probability to detect @xmath3 output events for an input flux @xmath9 is described by the poisson distribution @xmath10 ( pd ) .
the total probability of the registration of @xmath3 events with a mean flux @xmath11 can be evaluated using the expression @xmath12 this description in the form of a doubly stochastic process ( cox process ) is commonly accepted .
alternatively , the inhomogeneous poisson process is used when the count rate is a smoothly varying function of time ( see for e.g. , @xcite ) .
this approach is more convenient for deterministic light signals , but of little use in scintillation studies .
let us calculate the mean value of @xmath3 , which is distributed in accordance with the function @xmath13 , directly from its definition .
@xmath14 = \sum\limits_{x=0}^{\infty } x\ , \phi(x;\xi , s^2 ) = \int\limits_{0}^{\infty}\sum\limits_{x=0}^{\infty } x\,p(x,\xi s)\,f(s ; s^2)\mathrm{\,d}s = \int\limits_{0}^{\infty}\xi s\,f(s ; s^2)\mathrm{\,d}s = \xi .
\label{eq : mu1}\ ] ] this expression confirms the interpretation of the parameter @xmath15 as the mean value of the input flux of photon events .
to calculate the variance @xmath16 $ ] , we use the fact that for the pd , the factorial moment @xmath17 } = \theta^r = \mu_\mathrm{[1]}^r$ ] , where @xmath18 denotes the order of the moment . the factorial moment @xmath19}$ ] of the registered count @xmath3
can be written as @xmath20 } = \sum\limits_{x=0}^{\infty } ( x)_r \int\limits_{0}^{\infty } p(x;\xi s)\,f(s ; s^2)\mathrm{\,d}s , \label{eq : fmom_calc0}\ ] ] where @xmath21 denotes the pochhammer symbol for the lower factorial .
after summation under the integral and replacing the sum with the expression for @xmath17}$ ] , we obtain @xmath20 } = \int\limits_{0}^{\infty } ( \xi s)^r f(s ; s^2)\mathrm{\,d}s % \\= \xi^r \int\limits_{0}^{\infty } s^r f(s ; s^2)\mathrm{\,d}s = \xi^r\,\nu^{\prime}_r , \label{eq : mandell}\ ] ] where @xmath22 denotes the raw moment of order @xmath18 of the scintillation @xmath6 . this equation is called the mandel relation . applying the trivial formulas @xmath23 } = \mathsf{e}[x^2 ] - \mathsf{e}[x]$ ] , @xmath16 = \mathsf{e}[x^2 ] -
( \mathsf{e}[x])^2 $ ] , and @xmath24 , we obtain the following well - known expression : @xmath25 = \xi + \xi^2 s^2 .
\label{eq : mu2}\ ] ] equations ( [ eq : mu1 ] ) and ( [ eq : mu2 ] ) explain why sometimes instead of the process with the distribution given by eq .
( [ eq : probab1 ] ) , poisson noise is considered incorrectly as additive noise with power @xmath11 superimposed on the signal with mean @xmath11 and variance @xmath26 . in this interpretation
, the scintillation has a very different distribution , although has the same the first two moments .
higher moments , which can be calculated by eq .
( [ eq : mandell ] ) , demonstrate the difference between these interpretations . equation ( [ eq : mandell ] ) can also be obtained from the representation of the probability generating function ( pgf ) @xmath27 through the moment generating function ( mgf ) @xmath28 for the @xmath4 distribution as below : @xmath29 where @xmath30 over all values of @xmath3 is the poisson pgf with the parameter @xmath9 . since @xmath3 is a non - negative integer
, @xmath31 is also the generating function of factorial moments .
the relation between moments can be obtained in the usual way . in the case of nonlinear detectors , the distribution @xmath10 is no longer a poisson distribution . as we have shown in @xcite , an adequate description of registered events with a counter having a dead time @xmath32
is achieved by using the generalized poisson distribution ( gpd ) , introduced in @xcite .
this distribution is the two - parameter distribution and it is usually written as @xmath33 where @xmath34 , and the parameter @xmath35 defines the truncation of @xmath36 from the inequality @xmath37 .
when @xmath38 , the distribution reduces to an ordinary pd with the parameter @xmath39 .
the expectation ( mean ) and variance of gpd are respectively equal to @xmath40 = \frac{\theta}{1-\lambda } , \quad \mathsf{var}[k ] = \frac{\theta}{(1-\lambda)^3}. \label{eq : gpd_mom}\ ] ] in our case , the parameter @xmath39 denotes the flux of input events @xmath9 , and the parameter @xmath41 . for counter type i , @xmath42 , and for the counter with paralyzable dead time , @xmath43 @xcite .
the parameter @xmath44 is the nonlinearity parameter , or dead time , reduced to the interval @xmath0 , i.e. , @xmath45 . in the situations discussed below , @xmath46 and hence @xmath47 . from eq .
[ eq : gpd_mom ] , it follows that modulo @xmath35 is the ratio of the number of missing events to the number of registered ones .
the truncation has a formal character as @xmath36 is always considerably more than @xmath48 $ ] .
the parameter @xmath35 can be simply obtained from the non - poisson parameter @xmath49/\mathsf{e}[x ] = ( 1 - \lambda)^{-2}$ ] , which is usually measured during detector tests @xcite .
the generating functions for the gpd are known , and thus , the pgf can be written @xcite as : @xmath50 where @xmath51 denotes the lambert function ( see for e.g. , @xcite ) .
it is easy to verify that @xmath52 , and in the limit @xmath53 , it coincides with the poisson pgf . the method applied in the case of eq .
( [ eq : fmgf ] ) , can not be repeated directly , since in the discussed case , @xmath35 is dependent on @xmath39 , that does not allow to present eq .
( [ eq : fgf ] ) as the scintillation mgf with compound argument .
a relation similar to eq .
( [ eq : fmgf ] ) can only be obtained by neglecting the variation of @xmath35 with light intensity .
the setting a certain value of @xmath35 leads to a linear approximation , corresponding to the local slope of the nonlinear curve at this chosen value .
however , it is known that in the case of strong scintillation , the maximal intensity exceeds the average flux by an order of magnitude .
therefore , it is required to perform a detailed analysis taking into account the curvature of the nonlinearity function , is desired . in the following sections , the nonlinearity for constant flux
is refered as the _ static nonlinearity_. the effect , that caused by the fluctuations of light intensity and curvature of nonlinearity function , is reffered to as the _ dynamic nonlinearity_. using eq . (
[ eq : fgf ] ) , we can calculate any factorial moment @xmath17}$ ] for the gpd . in the normalized form
, the first four moments can be written as follows : @xmath54 } & = \frac{\theta}{1-\lambda } , \notag \\ \frac{\mu_\mathrm{[2]}}{\mu_\mathrm{[1]}^2 } & = 1 + \frac{\lambda\,(2-\lambda)}{\theta\,(1-\lambda ) } , \label{eq : fmomgpd } \\
\frac{\mu_\mathrm{[3]}}{\mu_\mathrm{[1]}^3 } & = 1 + \frac{3\,\lambda\,(2-\lambda)}{\theta\,(1-\lambda ) } + \frac{\lambda^2(2\,\lambda^2 - 8\,\lambda+9)}{\theta^2(1-\lambda)^2 } , \notag \\ \frac{\mu_\mathrm{[4]}}{\mu_\mathrm{[1]}^4 } & = 1 + \frac{6\,\lambda(2-\lambda)}{\theta\,(1-\lambda ) } + \frac{\lambda^2(11\,\lambda^2 - 44\,\lambda+48)}{\theta^2\,(1-\lambda)^2 } + \dots \notag\end{aligned}\ ] ] since @xmath55 , the second term on the right - hand side is of the order @xmath44 , the third is of the order of @xmath56 , etc . in general , the numerical coefficient of the second term is equal to @xmath57 for a moment of the order @xmath18 . for a typical dead time of @xmath5830 ns and exposure longer than 3 @xmath59s ( @xmath60 ) , we can ignore the third and subsequent terms .
in most situations , only the first term is needed , but for higher moments , the factor @xmath61 may be @xmath62100 and the second term can be significant despite the small value of @xmath44 .
we restrict our analysis to the first two terms , and represent the second term in the form @xmath63 , where the coefficient @xmath64 depends on the dead time type ; in either case , it differs from unity by a factor less than @xmath65 .
subsequently , @xmath17}$ ] can be written as @xmath66 } = \mu_\mathrm{[1]}^r \left(1 - \omega r\,(r-1)\,\tau\right ) = g_r\,\mu_\mathrm{[1]}^r .
\label{eq : fmom_appr}\ ] ] since @xmath64 is multiplied by the small parameter @xmath44 , it may be set to 1 .
for brevity , we denote the factor @xmath67 as @xmath68 .
the relations in the previous section are valid for the case where all registered events are independent . however , in actual detectors , after - pulses of various natures can be observed , when one extraneous pulse appears with a probability @xmath69 , subsequent to the pulse corresponding to a photon event .
it is known that binomial selection applied to a random poisson value @xmath3 leads also to the pd of variable @xmath70 with the parameter @xmath71 .
the unselected part @xmath72 is also poissonian with the parameter @xmath73 , and it is statistically independent from @xmath70 .
let us assume that @xmath70 represents the events that give rise to after - pulses , and hence , instead of @xmath70 , we will register @xmath74 .
obviously , in this case @xmath75 = \theta + q\theta$ ] . the variance @xmath76 = ( 1-q)\theta$ ] , and for the sample @xmath70 with after - pulses , @xmath77 = 4\,q\theta$ ] .
the total variance of @xmath3 amounts to @xmath16 = ( 1-q)\theta + 4\,q\theta = \theta + 3\,q\theta$ ] .
as can be seen , an excess of variance appears , i.e. the process is not quite poissonian , and the non - poissonian factor @xmath49/\mathsf{e}[x ] \approx 1 + 2\,q$ ] . in terms of the pgf
, this poisson process with after - pulses is written very simply as @xmath78 using this expession , we can obtain the factorial moments of the process as @xmath54 } & = \theta(1+q ) , \\ \mu_\mathrm{[r ] } & = \theta^r(1+q)^r + \theta^{r-1}r\,(r-1)\,q\,(1+q)^{r-2 } + o(q^2 ) , \notag \label{eq : after_moms}\end{aligned}\ ] ] or in the normalized form as @xmath79}}{\mu_\mathrm{[1]}^r } = 1 + \frac{r(r-1)q}{\mu_\mathrm{[1]}(1+q ) } + o(q^2/\mu_\mathrm{[1]}^2 ) .
\label{eq : reatr}\ ] ] from this formula , it follows that if the light intensity is decreasing , the normalized factorial moments increase infinitely , as opposed to the purely poissonian case .
when @xmath80 } < q$ ] , the terms of @xmath81}^2)$ ] also begin to influence the moments with @xmath82 .
the distortion of statistics caused by after - pulses leads to the invalidity of the mandel relation ( [ eq : mandell ] ) . in its right - hand side ,
the factor @xmath83 appears before the moment @xmath22 , and the term @xmath84 is another additional term . the combined effect of detector nonlinearity and after - pulses can be estimated by constructing the appropriate pgf : @xmath85 using this function , we calculated the first few factorial moments : @xmath54 } & = \frac{\theta(1+q)}{1-\lambda } , \notag \\ \frac{\mu_\mathrm{[2]}}{\mu_\mathrm{[1]}^2 } & = 1 \!+\ !
\frac{\lambda(2-\lambda)}{\mu_\mathrm{[1]}(1-\lambda)^2}\!+\!\frac{2q}{\mu_\mathrm{[1]}(1-\lambda)^2(1+q)},\label{eq : jointq } \\
\frac{\mu_\mathrm{[3]}}{\mu_\mathrm{[1]}^3 } & = 1 \!+\ !
\frac{3\lambda(2-\lambda)}{\mu_\mathrm{[1]}(1-\lambda)^2 } \!+\!\frac{6q}{\mu_\mathrm{[1]}(1-\lambda)^2(1+q ) } \!+\ !
o\bigl(\frac{\lambda^2}{\theta^2}\bigr ) . \notag\end{aligned}\ ] ] evaluating the following moments , we ensured that the moment of order @xmath18 can be presented in the form @xmath79}}{\mu_\mathrm{[1]}^r } = 1 + \frac{r(r-1)\lambda(2-\lambda)}{2\,\theta(1-\lambda)(1+q ) } + \frac{r(r-1)q}{\theta(1-\lambda)(1+q)^2 } + o\bigl(\frac{\lambda^2}{\theta^2}\bigr)+o\bigl(\frac{q^2}{\theta^2}\bigr ) .
\label{eq : momgpd - after}\ ] ] this expression satisfies two extreme cases : the absence of dead time ( perfect linearity ) and complete absence of after - pulses . as in sec .
[ sec : facmomgpd ] , we replace the first two terms with the coefficient @xmath86 while neglecting the emerged factor @xmath87 .
the third term at high fluxes introduces a small contribution , and therefore we can ignore distinction of @xmath88 from unity after multiplying eq .
( [ eq : momgpd - after ] ) by @xmath80}^r$ ] . denoting @xmath89 ,
we finally obtain @xmath66 } = g_r\,\mu_\mathrm{[1]}^r + h_r\,\mu_\mathrm{[1]}^{r-1}. \label{eq : fmom_appraft}\ ] ] the coefficient @xmath90 for @xmath91 is equal to zero ; the increasing of the order leads to an almost quadratic growth of the coefficient .
we recall that these are the factorial moments of the distribution @xmath92 . to proceed further to the statistical moments of registered counts in the case of scintillation , at first we expand @xmath93}$ ] ( [ eq : momgpd - after ] ) in series powers of @xmath44 . for counter type i , @xmath80 } = ( 1 + q)\xi s/(1 + \tau\xi s)$ ] and the expansion
is given as @xmath66 } = g_r(1+q)^r\sum_{k=0}\frac{r^{(k)}}{k ! } ( \xi s)^{r+k}(-\tau)^k + h_r(1+q)^{r-1}\sum_{k=0}\frac{(r+1)^{(k)}}{k ! } ( \xi s)^{r+k-1}(-\tau)^k , \label{eq : fmom_ser1}\ ] ] where @xmath94 denotes the rising factorial @xmath95 .
for counter type ii , @xmath80 } = ( 1 + q)\xi s e^{-\tau\xi s}$ ] , and its corresponding expression can be obtained from eq .
( [ eq : fmom_ser1 ] ) by replacing @xmath94 by @xmath96 .
that is , the expressions for types i and ii are almost identical , differing only in the coefficients with the third and higher terms .
these expressions form an alternating series what facilitate their convergence for a wide range of input parameters .
let us substitute the expansion into the integral in eq .
( [ eq : mandell ] ) , thereby defining the factorial moments of the registered counts @xmath19}$ ] .
the integration of the variable @xmath6 provides the raw moments @xmath97 of the scintillation , and as a result , we have @xmath20 } = g_r(1+q)^r\xi^r\sum_{k=0 } \frac{r^{(k)}}{k ! } ( -\xi\tau)^k \nu^\prime_{r+k } + h_r(1+q)^{r-1}\xi^{r-1}\sum_{k=0}\frac{(r+1)^{(k)}}{k!}(-\xi\tau)^k\nu^\prime_{r+k-1}. \label{eq : amom1}\ ] ] accordingly , in the expression for counter type ii , instead of using the rising factorial @xmath94 , the power @xmath98 is substituted .
evidently , with @xmath99 and @xmath100 , these expressions reduce to the mandel relation . from eq .
( [ eq : amom1 ] ) , it follows that in the case of nonlinearity , the factorial moment of the registered counts is defined by the corresponding and subsequent raw moments of the scintillation with rapidly decreasing weights if @xmath101 .
after - pulses add a preceding raw moment to the sum .
we clarify eq .
( [ eq : amom1 ] ) for the example of @xmath102 } \equiv \mathsf{e}[x]$ ] ( @xmath103 ) for the paralyzable dead time counter : @xmath104 } = ( 1+q)\,\xi\sum_{k=0 } \frac{(-\xi\tau)^k}{k ! } \nu^\prime_{k+1 } = \\= ( 1+q)\xi ( 1 - \tau\xi - s^2\tau\xi + \frac{1}{2}\nu^\prime_3\tau^2\xi^2 - \frac{1}{6}\nu^\prime_4\,\tau^3\xi^3 + \dots )
. \label{eq : alpha1_dec1}\ ] ] here , the moment @xmath105 is expressed via the scintillation index @xmath106 , and @xmath107 . setting all the values of @xmath108
, we obtain the standard formula for static nonlinearity : @xmath75 = \xi\,e^{-\tau\xi}$ ] .
we note that for strong scintillation ( @xmath109 ) , a significant additional bias appears in the form of dynamic nonlinearity , which amounts to the static nonlinearity .
[ sec : normfmom ] the factorial moment @xmath19}$ ] is proportional to the value @xmath110 , which is unknown in advance , and we have to somehow normalize the moments to avoid dependence on the brightness of the light source . for an ideal detector , normalization can be executed at once by using the value of @xmath110 . in our case ,
the measured mean @xmath102}$ ] is more convenient .
let us denote the normalized moment by @xmath111 } = \alpha_\mathrm{[r]}/\alpha_\mathrm{[1]}^r$ ] . for a constant flux ( @xmath112 , the distribution @xmath4 is degenerate , all the raw moments @xmath113 , and @xmath19 } = \mu_\mathrm{[r]}$ ] ) , all sums in the expression for @xmath111}$ ] are reduced , and naturally , we obtain @xmath114 } = g_r + \frac{h_r}{\alpha_\mathrm{[1]}}. \label{eq : amom_const}\ ] ] this equation indicates that detector nonlinearity has almost no effect on @xmath111}$ ] , and the effect of after - pulses should be very noticeable for low light intensities
. for stochastic radiation , @xmath111}$ ] immediately provides an approximate estimate of the desired moment @xmath115 of the scintillation .
since the detector nonlinearity reduces the values of the numerator and denominator of @xmath111}$ ] , it is expected that the normalized moment is changed smaller . indeed , in the linear approximation @xmath114 }
= \nu^\prime_{r } - r(\nu^\prime_{r+1 } - \nu^\prime_{r}\nu^\prime_{2})\,\tau\xi + \dots,\ ] ] the coefficient of @xmath65 is certainly less than the corresponding coefficient in eq .
( [ eq : amom1 ] ) . from eqs .
( [ eq : probab1][eq : mandell ] ) we can infer that for the calculation of the moment of a doubly stochastic process , it is not necessary to use the inner distribution @xmath10 if the corresponding moment is known .
further calculations are reduced to the weighted averaging of this moment with the outer distribution @xmath4 . in measurements of scintillation ( e.g. , for probing of atmospheric optical turbulence ) ,
the most important values are the relative variance and covariance .
given the importance of these parameters , we calculate the mean , variance and covariance of registered counts without using a specific form of the distribution @xmath10 .
this calculation is easy to perform for counter type ii . to obtain the mean of the registered flux , we use the well - established fact that @xmath116 = \xi s e^{-\tau \xi s}$ ]
therefore , we have @xmath117 = \int\limits_{0}^{\infty}\xi s e^{-\tau \xi s}\,f(s;s^2)\mathrm{\,d}s .
\label{eq : munl1}\ ] ] in addition , we consider the generating function of the central moments @xmath118 for the distribution @xmath4 in the form @xmath119 = \int\limits_{0}^{\infty}e^{u(s-1)}\,f(s ; s^2)\mathrm{\,d}s , \label{eq : munl2}\ ] ] while assuming that it exists in the neighborhood of zero .
it is well - known that this function and its derivatives are expressed through central moments by means of the following relationships : @xmath120 using @xmath121 , it is easy to show from eqs .
( [ eq : munl1 ] ) and ( [ eq : munl2 ] ) that @xmath122 substituting in this formula the series expansion of the mgf and its derivatives , i.e. , eq .
( [ eq : serz ] ) , we obtain @xmath123 in the expanded formula , we replace @xmath124 , @xmath125 , and @xmath126 .
the resulting expression is slightly different in its form from eq .
( [ eq : alpha1_dec1 ] ) because the additional factor of @xmath127 is separated from the sum .
the effect of after - pulses can be accounted by merely multiplying the equation with the factor @xmath87 . to calculate the variance of the registered events , we use the well - known classical expression @xcite for variance in the case of a paralyzable counter : @xmath128 = \theta e^{-\tau\theta}(1 - 2\tau\theta e^{-\tau\theta})$ ] .
let us represent it in the form of @xmath129 = \mathsf{e}[k ] - 2\tau\mathsf{e}[k]^2 $ ] , which implies the equation @xmath130 } = \mu_\mathrm{[1]}^2(1 - 2\,\tau ) \label{eq : mu2cl}\ ] ] the factor of @xmath131 is precisely the coefficient @xmath132 , which has been introduced in section [ sec : facmomgpd ] . using this equation ,
we can write @xmath133 } = g_2\int\limits_{0}^{\infty } \xi^2
s}\,f(s;s^2)\mathrm{\,d}s \label{eq : f2int}\ ] ] this integral can be calculated in a manner similar to eq .
( [ eq : munl1 ] ) using the mgf @xmath28 . omitting the intermediate transforms ,
we obtain @xmath134 } = g_2\,\xi^2 e^{-2\tau\xi}\sum_{k=0}^\infty ( \nu_k+2\nu_{k+1}+\nu_{k+2 } ) \frac{(-2\tau\xi)^k}{k ! } = \\= g_2\,\xi^2 e^{-2\tau\xi}\bigl(1+s^2 - 2\,(2\,s^2+\nu_3)\,\tau\xi + \dots\bigr ) .
\label{eq : amom2cl}\end{gathered}\ ] ] as in the case of eq .
( [ eq : alpha1 ] ) , this formula does not exactly have the form of eq .
( [ eq : amom1 ] ) ; however , in most cases , they coincide . setting @xmath99
, we obtain @xmath23 } = \xi^2 ( 1 + s^2)$ ] , which after transition to variance , redices to known expression @xmath16 = \xi + \xi^2 s^2 $ ] .
based on the discussion in sec [ sec : afterpulse ] , we account for the effect of after - pulses by multiplying the obtained expression by the factor of @xmath135 and adding the term @xmath136 . to calculate the covariance @xmath137 of registered events , instead of using eq .
( [ eq : fmom_calc0 ] ) , we write @xmath138 = \\= \iint\limits_{0}^{\infty } \xi_1\xi_2 s_1 s_2e^{-\tau\xi_1 s_1 -\tau\xi_2 s_2}\,f(s_1,s_2))\mathrm{\,d}s_1 \mathrm{\,d}s_2 , \label{eq : cov_int}\ ] ] where the function @xmath139 is a joint distribution of the correlated random variables @xmath140 and @xmath141 .
owing to uncorrelated photon events , the joint distribution @xmath142 , and the product of @xmath143\mathsf{e}[x_2 ; \xi_2 s_2]$ ] appears under the integral .
the variables @xmath140 and @xmath141 may represent scintillation values for different apertures ( channels ) or at different temporal points . in any case
, the moment @xmath144 $ ] or @xmath145 characterizes their correlation .
similar to the approach assumed to calculate variance , we make use of the generating function of the joint central moments @xmath146 $ ] , and we represent @xmath137 in the form @xmath147 where @xmath148 and @xmath149 . function @xmath150 and its derivatives can be expanded as in eq .
( [ eq : serz ] ) in the series of joint moments @xmath151 : @xmath152 for clarity , we write the first few terms of this series : @xmath153.\ ] ] subsequent to the subtraction of the products of the means , we note that the required moment @xmath154 is a key term in the series .
the presence of after - pulses in channels 1 and 2 with respective probabilities of @xmath155 and @xmath156 is accounted for by the multiplicative factor @xmath157 . for practice ,
the inverse task , i.e. , calculation of the scintillation moments on the base of measured moments @xmath111}$ ] , is more interesting .
it greatly simplifies matters that only the scintillation moment of order @xmath18 enters into the right - hand side of the equations with weight nearly equal to unity .
the following moments are included in the inverse calculation with small weights of the order of @xmath65 , @xmath158 , etc .
in such a situation , we can use a simple iterative scheme starting from eq .
( [ eq : mandell ] ) for perfectly linear counters . however , certain factors can complicate these calculations . for certain distributions of @xmath4 , raw moments quickly and indefinitely grow with their order , and therefore , decrease in @xmath159 does not guarantee convergence of the series . the sufficient conditions for the convergence of the alternating series given by eq .
( [ eq : amom1 ] ) are @xmath160 for counters of types i and ii , respectively , for all @xmath36 starting from some number .
these inequalities indicate the following : 1 ) for counter type i , the restrictions are less stringent than for counter type ii , 2 ) the convergence of the series depends on the nonlinearity @xmath161 and a series converging at a small flux may diverge at a large flux , 3 ) convergence depends on order @xmath18 , and it can only occur for the lowest moments . for example , a log - normal distribution , which is often used as a representative distribution of stellar scintillation , has a moment @xmath162 .
eq . ( [ eq : cond1 ] ) leads to the condition @xmath163
obviously , for any @xmath18 and arbitrarily small @xmath65 , the condition is never satisfied starting from some @xmath36 , since @xmath164 due to the normalization . for the expansions given by eqs .
( [ eq : alpha1 ] ) and ( [ eq : amom2cl ] ) as regards central moments , the conditions are similar , but include a combination of few consecutive moments .
thus , the series for @xmath23}$ ] converges if @xmath165 given that the central moments increase at a slower rate than raw ones , and that the right - hand side includes the sum of three consecutive moments , we expect that condition given by eq .
( [ eq : cond2 ] ) permits a wider range of initial data . for practical purposes , a convergence in mathematical sense is not sufficient . in order to ensure that errors in measurements do not deteriorate a performance of the iterative scheme and its result ,
the actual convergence of the series should be provided to occur faster , for e.g. , by considering only the linear and quadratic terms .
measured light may contain an additional constant component : background radiation @xmath166 .
the background is described by a degenerate distribution in which all the central moments are equal zero except the mean @xmath167 $ ] .
we need to convert the previously obtained moments of the variable @xmath168 , where the relative fraction of the background @xmath169 , to the moments of @xmath170 .
obviously , @xmath171 = \mathsf{e}[s ] = 1 $ ] . using the properties of mgf ,
we obtain @xmath172 that leads to the equality @xmath173 for the central moments .
this implies that the results obtained in the previous step are transformed to the moments of the scintillation itself according to the relation @xmath174 for raw moments , the equation can also be obtained from the mgf using the shift and scaling formula for a random variable : @xmath175 the differentiation of which can provide the required relation . on the left - hand side ,
we obtain @xmath176 , and on the right , we obtain a polynomial of degree @xmath18 for @xmath177 and the moments @xmath178 . considering that @xmath179 , we get the approximate expression : @xmath180 a similar relation is given in @xcite , p. 87 . from this equation , it follows that even the background radiation @xmath181 can significantly affect higher moments when using the relation given by eq .
( [ eq : back ] ) .
in the case of constant intensity , the moments @xmath19 } = \mu_\mathrm { [ r]}$ ] , and thus , the distribution @xmath92 can be verified experimentally .
we carried out these measurements with the mass instrument in the laboratory using an internal light source @xcite .
the experiment was conducted at a device temperature of about @xmath182c with photomultiplier tubes ( pmts ) manufactured in 2003 . as expected ,
these factors have led to a significant increase in the fraction of after - pulses . for each level of intensity over a wide range ,
the factorial moments of order @xmath183 were measured for an accumulation time of 60 s. a short exposure time of @xmath184 was set in order to amplify desired effects . [ cols="^,^ " , ] the behavior of the measured normalized moments @xmath185}$ ] of various order @xmath186 depending on the measured mean @xmath102}$ ] , is shown in the left panel of fig . [
fig : exp2 ] for low light fluxes .
we note that the normalized moments increase with decreasing value of @xmath102}$ ] in agreement with eq .
( [ eq : amom_const ] ) .
the right panel of the figure shows the dependence of @xmath111}$ ] for large fluxes .
all measured points lie below unity , as predicted by the theory . using measurements for @xmath187
, we determined the probability of after - pulses as @xmath188 .
the value of the nonlinearity parameter @xmath189 was measured previously .
the curves , constructed in accordance with the definition of the coefficients @xmath190 and @xmath89 as per eq .
( [ eq : amom_const ] ) are indicated by thin solid lines . from eq .
( [ eq : momgpd - after ] ) , we note that the sum of the second and the third terms is equal to zero when @xmath191 regardless of the order @xmath18 .
this fact is represented as the intersection of all the curves at one point with @xmath102 } = 580 $ ] and @xmath111 } = 1 $ ] .
the observed agreement between the theoretical description and the experimental data confirms that the distribution of registered events is close to the gpd and the accounting of after - pulses via eq .
( [ eq : momgpd - after ] ) is valid .
we note that the modified poisson distribution @xcite does not predict the observed behavior of @xmath111}$ ] for large fluxes .
the experimental results confirm our theoretical conclusions , and in particular , that the use of factorial moments at very low counts per exposure requires a guaranteed absence of the after - pulses .
the measurements shown in fig .
[ fig : exp2 ] can be effectively used to determine the fraction of after - pulses @xmath192 with improved accuracy due to the large coefficient @xmath193 for higher moments .
it is noteworthy that in actual detectors , the value @xmath192 depends on several external factors that vary over time , and requires periodic inspection .
numerous measurements of the statistical moments of stellar scintillation @xcite show that , in fact , the moments are growing slower than the corresponding moments of the log - normal distribution
. we also performed a series of measurements with the mass instrument during monitoring of the optical turbulence in the atmosphere above mount shatdzhatmaz @xcite in the period january
april 2013 . at this point , we recall that the device measures the fluctuations in light intensity with an exposure time of @xmath194s simultaneously in four entrance apertures a , b , c , and d , whose outer diameters are 2.0 , 3.5 , 6.3 and 9.0 cm , respectively .
the nonlinearity parameters for all the channels are nearly identical , and @xmath195 .
raw moments from the third up to the ninth , measured at the smallest aperture a , are shown in fig .
[ fig : raw_mom ] as a function of the second raw moment @xmath196 .
the moments were calculated as @xmath197}$ ] , that is , without taking into account the nonlinearity of the detectors .
each point was calculated as an average of 20 estimates with an accumulation time of 1 s ( 40000 counts ) .
this procedure provides reasonably good statistical accuracy , and it allows the minimization of the impact of trends on the results . although periods with very strong scintillations are rarely observed , our measurements cover a wide range of scintillation power .
the figure illustrates the fact that in the case of strong scintillation , the actual instances of observed higher moments are considerably smaller than the moments of the log - normal distribution .
for example , at @xmath198 , the seventh moment @xmath199 is less than the corresponding log - normal moment by an order of magnitude . using these data we tested the convergence of eq .
( [ eq : amom1 ] ) for moments of different orders . to perform this test
, we attempted to determine for the nonlinearity @xmath65 for which the relative error , introduced by a cubic term , amounts to @xmath200 in the worst case .
the relative contributions of the first three terms , depending on the scintillation power , are shown in fig . [
fig : terms ] .
as might be expected , the allowable nonlinearity decreases rapidly with order and @xmath65 becomes as low as 0.01 for @xmath201 which is usually considered to lie within the linearity range .
the contribution of the quadratic term is negligible ( less than 0.01 ) under weak scintillation ; however , it comes up to 0.050.08 under strong scintillation , and it should be taken into account .
linear term should always be considered , and under strong scintillation , it can be neglected only if the nonlinearity @xmath65 is less than 0.001 even for the second moment . in our measurements , this condition corresponds to the mean @xmath202 counts , what greatly impairs statistical accuracy . with the same parameters ,
similar estimates are made for eqs .
( [ eq : alpha1 ] ) and ( [ eq : amom2cl ] ) .
it is confirmed that these expansions converge more rapidly ; for extremely strong scintillation , the contribution of the quadratic term is reduced by almost half , while the contribution of the cubic one becomes @xmath2030.005 . in this section ,
we first describe the series given by eq .
( [ eq : amom1 ] ) obtained under the assumption of gpd .
the algorithm for the calculation of the required scintillation moment is sufficiently simple ; we allocate the desired moment @xmath115 ( @xmath204 ) and the sum of corrections thereto @xmath205 ( @xmath206 ) in the right - hand side . because the measurements yield the normalized factorial moment @xmath111}$ ] , before substitution this moment
must be multiplied by the factor @xmath207 , where @xmath208 represents the sum in eq .
( [ eq : alpha1_dec1 ] ) , and it equals @xmath102}/\xi$ ] ( see sec . [
sec : registered ] ) .
to calculate @xmath205 , we use the measured @xmath111}$ ] as estimates of the scintillation moments . by transposing the sum @xmath209 to the left - hand side
, we obtain the corrected moment of the scintillation to the right . to assess the contribution of after - pulses , crude estimates of the moments are sufficient . under conditions of our measurements with mass ,
@xmath208 is nearly unity , for e.g. , for scintillation in the a and b apertures , this factor is @xmath2030.995 , and for apertures c and d , it is about @xmath210 , reaching @xmath211 for brightest stars .
therefore , the correction for detector nonlinearity is small ( its relative value is only a few percent ) , and one iteration is sufficient for the calculation .
however , relative corrections for higher moments ( @xmath212 ) can be as large as @xmath213 . despite the fact that the nonlinearity factor for aperture a is small ( @xmath214 ) , the corrections for a are comparable with the corrections for aperture d ( @xmath215 ) because the scintillation in aperture d is approximately five times less intense . in principle , for the comparison with theoretical behaviors , the dependencies shown in fig .
[ fig : raw_mom ] can be used without correction .
the reason is that in the coordinates of this graph , the nonlinearity effect shifts data points along the measured dependencies . as result
, the data points are not divided into subgroups , corresponding to weak and bright program stars , although the brightness of the stars differs more than 10 times . for the processing of the measurements of the mean , variance and correlation of the scintillation , eqs .
( [ eq : alpha1 ] ) , ( [ eq : amom2cl ] ) and ( [ eq : rho_fin ] ) are better suited in practice . in these formulas ,
the effect of static nonlinearity is already separated out , and only the effect of dynamic nonlinearity requires decomposition .
we recall that these formulas are valid for the counter with paralyzable dead time , which is a typical situation when using a photomultiplier as detector .
let us denote the sum in eq .
( [ eq : alpha1 ] ) by @xmath216 , and let us evaluate it using the estimates of the scintillation moments @xmath217}$ ] .
the mean flux at the counter input can be calculated by using the formula from @xcite : @xmath218 where @xmath219 denotes the lambert function , which is already known to us .
we select all the terms containing @xmath220 from the sum in eq .
( [ eq : amom2cl ] ) , and the residual of the sum is denoted as @xmath221 which is also of the order of unity .
dividing eq .
( [ eq : amom2cl ] ) by @xmath222 , we obtain the expression for the normalized factorial moment : @xmath223 } = ( 1 - 2\,\tau)\frac{\nu_2\,(1 - 4\tau\xi+2\tau^2\xi^2 ) + \sigma_2}{\sigma_1 ^ 2 } + \frac{2\,q}{\alpha_1 } \label{eq : normal}\ ] ] when using the factorial moments , the contribution of the poisson noise is already accounted for , and the second term in the equation represents the difference between poisson noise and photon noise in the presence of after - pulses .
this equation is easily solved for @xmath220 .
strictly speaking , @xmath220 is included in the term @xmath216 as well , and we should use an iterative process . to refine the higher moments we can use the general eq .
( [ eq : amom1 ] ) .
we note that @xmath224 is included in eq .
( [ eq : amom2cl ] ) with weight @xmath2032@xmath65 and its contribution to the actual measurements with mass can range upto 3% under strong scintillation conditions .
the double sum in eq .
( [ eq : rho_fin ] ) for the covariance is of the order @xmath225 . as in the case of the variance
, we split this sum into the terms with @xmath154 and the remainder @xmath226 . dividing this expression by the product of the averages @xmath227
, we obtain the following equation for the measured @xmath228 : @xmath229 in the case of auto - covariance , it is possible to further simplify the formula since in this case , @xmath230 , @xmath231 .
in addition , the relation @xmath232 holds for mixed moments .
in this study , we considered the impact of two well - known effects on the registration of photodetector counts : dead time @xmath44 and after - pulses with probability @xmath192 , which are inherent in different degrees in all counters of single - photon events .
these effects distort the registered statistics in the measurements of the statistical characteristics of fluctuating radiation . although our starting point is the measurement of stellar scintillation for remote sensing of the optical turbulence via the mass measurement method , these relations are applicable to any similar problems , for example , even to those associated with fast and strong chaotic variability of physical and astronomical light sources . assuming that dead time leads to the distribution of counts ( the number of detected photons per exposure ) described by the generalized poisson distribution , we derived general expressions for the factorial moments @xmath19}$ ] of the registered counts in the form of a series expansion with powers of the nonlinearity @xmath65 combined with the raw moments @xmath233 ( @xmath234 ) of light intensity . using well - known expressions for the nonlinearity of the counter with paralyzable dead time
, we obtained the dependencies of the mean , variance , and covariance on the central moments of light scintillation .
these expansions are more practical than the general ones , and they are not directly associated with a particular type of distribution of photon events .
a detailed examination of these expansions proves that their convergence depends on the order of the factorial moment , the value of the nonlinearity @xmath65 , and the actual growth of scintillation moments . for processes with the log - normal distribution of radiation ,
these series diverge at an arbitrarily small nonlinearity .
the expansions indicate that under strong scintillation ( when the relative variance of light fluctuations is of the order of unity ) , the dynamic nonlinearity owing to the curvature of the nonlinearity function is comparable to the static nonlinearity corresponding to its slope . in such a scenario ,
the usual treatment of the effect becomes insufficient .
the evaluation of the effect of after - pulses on the registered moments demonstrates that the contribution of after - pulses is usually underestimated .
even a small fraction ( @xmath2031% ) of after - pulses leads to significant errors in the higher moments ( of the order of tens of percent ) . at very weak fluxes
, the usage of the factorial moments ( mandel relation ) can lead to errors by several times .
we performed simple experiments to confirm the theoretically obtained relations .
the results prove that the gpd adequately represents the distribution of photon counts in the case of nonlinearity caused by dead time during the registration process .
e. jakeman and p. n. pusey , `` photon - counting statistics of optical scintillation , '' in _ inverse scattering problems in optics .
series : topics in current physics , _ * 20 * ( springer berlin heidelberg , 1980 ) , 73116 .
v. kornilov , a. a. tokovinin , o. vozyakova , a. zaitsev , n. shatsky , s. f. potanin , and m. s. sarazin , `` mass : a monitor of the vertical turbulence distribution , '' in _ adaptive optical system technologies ii , _
p. l. wizinowich and d. bonaccini , eds .
, _ proc of the spie _ , * 4839 * , 837845 ( 2003 ) . r. m. corless , d. j. jeffrey , and d. e. knuth , `` a sequence of series for the lambert w function , '' in _ proceedings of the 1997 international symposium on symbolic and algebraic computation , _
w. w. kchlin , ed .
( ny : acm press , maui , hawaii , 1997 ) , 197204 . | many physical experiments require analysis of the statistics of fluctuating radiation . in the case of an ideal single - photon detector ,
the contribution of photon noise to the statistics of the registered signal has been thoroughly examined . however , practical photon counters have a dead time , leading to miscounting of certain true events , and sometimes the counters generate false after - pulses .
this study investigate the impact of these two effects , and it presents the theoretical relations between the statistical moments of the radiation and the registered counts while also accounting for dead time and the probability of after - pulses .
expressions for statistical moments of any order are obtained on the basis of the generalized poisson distribution ( gpd ) . for counters with paralyzable dead time
, alternative relations for the mean and variance are derived using generally accepted formulas . as an example
, the measurements of stellar scintillation and the result of simple experiment are considered .
the results of the experimental verification of the theoretical expression confirm the need to account for the non - ideal nature of detectors in almost all similar measurements . |
there has been significant recent work on vlasov - maxwell ( vm ) equilibria that are consistent with nonlinear force - free @xcite and ` nearly force - free ' @xcite magnetic fields in cartesian geometry .
therein , force - free refers to a magnetic field for which the associated current density is exactly parallel , which is the definition we shall also use , @xmath0 these works consider one - dimensional ( 1d ) collisionless current sheets , with refs . specifically calculating vm equilibrium distribution functions ( dfs ) that are self - consistent with a given specific magnetic field configuration .
a natural question to consider is whether it is also possible to find self - consistent force - free ( or nearly force - free ) vm equilibria for other geometries , in particular cylindrical geometry . in this paper
we shall present particular vm equilibria for 1d magnetic fields which are nearly force - free in cylindrical geometry , i.e. flux tubes / ropes two of the archetypal field configurations in cylindrical geometry are the @xmath1-pinch ( with axial current and azimuthal magnetic field ) , a classical example of which is the bennett pinch@xcite ; and the @xmath2-pinch ( azimuthal current and axial magnetic field ) .
consideration of ` vlasov - fluid ' models of @xmath1-pinch equilibria was given in ref .
, with ref .
calculating @xmath1-pinch equilibria and an extension with azimuthal ion - currents .
others have also constructed kinetic models of the @xmath2-pinch , see refs . for examples . in the same year as pfirsch @xcite , cylindrical kinetic equilibria with only azimuthal current
were studied in ref . .
for examples of treatments of the stability of fluid and kinetic linear pinches , see refs respectively .
recently there have been studies on ` tokamak - like ' vm equilibria with flows @xcite , starting from the vm equation in cylindrical geometry and working towards grad - shafranov equations for the vector potential .
we also note two vlasov equilibrium dfs in the literature that are close in style to the one that we shall present .
the first is described in a brief paper @xcite , with an equilibrium presented for a cylindrical pinch .
however , their distribution describes a different magnetic field and the df appears not to be positive over all phase space .
the second df is a very recent paper @xcite that actually describes a magnetic field much like the one that we discuss .
their df is designed to model ` ion - scale ' flux tubes in the earth s magnetosphere .
formally , their quasineutral model approaches a nonlinear force - free configuration in the limit of a vanishing electron to ion mass ratio . in their model ,
current is carried exclusively by electrons and the non - negativity of the df depends on a suitable choice of microscopic parameters .
finally , we mention that in beam physics , much work on constructing cylindrical vm equilibria is done by looking for mono - energetic distributions with conserved angular momentum , see refs . for some examples .
magnetic flux tubes and flux ropes are prevelant in the study of plasmas , with a wide variety of observed forms in nature and experiment , as well as uses and applications in numerical experiments and theory . some examples of the environments and fields of study in which they feature include solar @xcite ; solar wind @xcite ; planetary magnetospheres @xcite and magnetopauses @xcite ; astrophysical plasmas @xcite ; tokamak @xcite , laboratory pinch experiments @xcite and the basic study of energy release in magnetised plasmas @xcite , to give a small selection of references .
one application of flux tubes is in the study of solar active regions @xcite and the onset of solar flares and coronal mass ejections @xcite . a classic magnetohydrodynamic ( mhd ) model for magnetic flux tubes
was first presented by t. gold and f. hoyle ( gh ) @xcite , initially intended for use in the study of solar flares .
the gh model is an infinite , straight , 1d and nonlinear force - free magnetic flux tube with constant ` twist'@xcite .
mathematically , the gh magnetic field could be regarded as the cylindrical analogue@xcite of the force - free harris sheet@xcite ( a planar current sheet model ) , as the bennett pinch @xcite might be to the ` original ' harris sheet @xcite .
it is typical to consider solar , space and astrophysical flux tubes within the framework of mhd , e.g. see ref . .
however , many of these plasmas can be weakly collisional or collisionless , with values of the collisional free path large against any fluid scale @xcite , making a description using collisionless kinetic theory necessary .
it is the intention of this paper to study the gh flux tube model beyond the mhd description , since - apart from the very recent work in ref .
- we see no attempt in the literature of a microscopic description of the gh field .
other than any interesting theoretical advances , a possible application of the results of this study could be to implement the obtained model in kinetic ( particle ) numerical simulations . in cartesian geometry , the work in refs .
used the method proposed by channell@xcite to tackle the vm inverse problem , i.e. to determine self - consistent equilibrium dfs for a given magnetic field configuration .
channell described the extension of his work to cylindrical geometry as ` not possible in a straightforward manner . ' as explained in ref .
( in which cylindrical coordinates are used to model a torus ) , this is due in part to the ` toroidicity ' of the problem , i.e. the @xmath3 factor in the equations . as we shall see in this paper
, another potential complication is the need to allow at least in principle a non - zero charge density .
the work in this paper does not present a generalised method for the vm inverse problem in cylindrical geometry , but instead some particular solutions for a specific given magnetic field .
the paper is structured as follows . in section 2
we first review the theory of the equation of motion consistent with a collisionless df in cylindrical geometry , and discuss the question of the possibility of 1d force - free equilibria .
then we introduce the magnetic field to be used .
we note that whilst the work in this paper is applied to a particular magnetic field from subsection 2.1 onwards , the steps taken to calculate the equilibrium df seem as though they could be adaptable to other cases . in section 3
we present the form of the df that gives the required macroscopic equilibrium , and proceed to ` fix ' the parameters of the df by explicitly solving ampre s law and poisson s equation .
note that whilst we choose to consider a two - species plasma of ions and electrons , we see no obvious reason preventing the work in this paper being used to describe plasmas with a different composition . in section 4 .
we present a preliminary analysis of the physical properties of the equilibrium .
particularly technical calculations are in the appendices .
appendix a contains the zeroth and first order moment calculations , used to find the number densities and bulk flows directly , and in turn the charge and current densities .
appendix b contains the mathematical details of the existence and location of multiple maxima of the df in velocity - space .
a collisionless equilibrium is characterised by the 1-particle distribution function , @xmath4 , a solution of the steady - state vlasov equation ( e.g. see ref . ) .
the vlasov equation in cylindrical coordinates is @xmath5=0,\label{eq : vlasov}\end{aligned}\ ] ] see for example refs . .
here @xmath6 and @xmath7 are used as ` spatial ' indices running over @xmath8 , and @xmath9 is used as the particle species index .
individual particle positions and velocities are given by @xmath10 and @xmath11 respectively , for @xmath12 the horizontal distance from the @xmath1 axis , and @xmath2 the azimuthal angle .
the totally antisymmetric unit tensor of rank 3 ( the levi - civita tensor ) is @xmath13 , and the einstein summation convention is applied ( such that repeated indices are summed over , with subscript and superscript indices used to describe co- and contravariant components respectively .
the mass and charge of particle species @xmath9 are @xmath14 and @xmath15 respectively .
the electric and magnetic fields are defined as @xmath16 and @xmath17 , for @xmath18 the scalar potential and vector potential @xmath19 .
the ` fluid ' equation of motion of a particular species @xmath9 is found by taking first - order velocity moments of the vlasov equation . after a routine but laborious moment - taking calculation
, we see that - in equilibrium ( @xmath20 ) , assuming a one - dimensional configuration with only radial dependence ( @xmath21 , and letting @xmath4 be an even function of the radial velocity @xmath22 - force balance for species @xmath9 is maintained according to @xmath23 the pressure tensor for species @xmath9 is a rank-2 tensor and is defined by @xmath24 where @xmath25 , for @xmath26 the bulk velocity of species @xmath9 and @xmath27 the individual particle velocity .
note that the assumption of @xmath28 to be an even function of @xmath29 automatically implies that @xmath30 .
equation ( [ eq : species ] ) can be summed over species to give @xmath31 where @xmath32 is the force density associated with the rotating bulk flows of the ions and electrons .
equation ( [ eq : cmhde_alt ] ) is a cylindrical analogue of the force balance equation in cartesian geometry ( e.g. see ) .
there are ` extra inertial terms ' as compared to the case of cartesian geometry . from the point of view of a particular magnetic field @xmath33 ( which is the point we take by specifying a particular macroscopic equilibrium )
, we see that equilibrium is maintained by a combination of density / pressure variations as in the case of cartesian geometry , but with additional contributions from centrifugal forces and as an inevitable result of the resultant charge separation , an electric field .
this clearly demonstrates that ` sourcing ' an exactly force - free macroscopic equilibrium with an equilibrium df in a 1d cylindrical geometry is inherently a more difficult task than in the cartesian case .
the presence of ` extra ' positive definite inertial forces and , almost inevitably , forces associated with charge separation raises the question of whether exactly force - free equilibria are possible at all in this paradigm . before proceeding ,
we comment that given certain macroscopic constraints on the electromagnetic fields or fluid quantities - such as the force - free condition , or a specific given magnetic field ( for example ) - it is not _ a priori _ known how to calculate a self - consistent vlasov equilibrium , or if one even exists within the framework of the assumptions made .
hence one has to proceed more or less on a case by case basis , with the intention of achieving consistency with the required macroscopic conditions , upon taking moments of the df . in refs .
for example , a method used to calculate a df , given a prescribed 1d magnetic field was inverse fourier transforms ( ift ) .
a distribution function of the form @xmath34 was used , with @xmath35 , @xmath36 and @xmath37 the conserved particle hamiltonian and canonical momenta in the @xmath38 and @xmath39 directions , and @xmath40 an unknown function , to be determined . since
our problem is one of a 1d equilibrium with variation in the radial direction , the three constants of motion are the hamiltonian , and the canonical momenta in the @xmath2 and @xmath1 directions : @xmath41 a function of a subset of the constants of motion is automatically a solution of the vm equation ( e.g. see ref . ) .
one can try to calculate an equilibrium distribution for the gold - hoyle force - free flux tube without a background field by a similar method , assuming a df of the form @xmath42 by exploiting the convolution in the definition of the current density ,
@xmath43 ampre s law can be solved by ift , with the quantity @xmath44 defined by @xmath45 notice how when written in this integral form , @xmath46 is not only a function of @xmath47 , but - in contrast with the cartesian case - also of the relevant spatial co - ordinate , @xmath12 . in the case of zero scalar potential ,
the result of the calculation is to give a distribution function that is not a solution of the vlasov equation as it is not a function of the constants of motion only .
in essence , an additional @xmath48 factor is required in the df to counter @xmath49 terms that manifest by completing the square in the integration .
the physical cause here would appear to be the inertial forces associated with the rotational bulk flow .
if one assumes a non - zero scalar potential , then it seems impossible to satisfy ampre s law .
the physical cause seems to be that , in the case of force - free fields , one would require a ` different ' electrostatic potential to balance the inertial forces for the ions and electrons , which is of course nonsensical .
thus , our investigation seems to suggest that it is not possible to calculate a df of the form of equation ( [ eq : fansatz ] ) for the exact gh field . to make progress
, we introduce a background field in the negative @xmath1 direction .
the mathematical motivation for this change is to balance the ` @xmath49 problem ' .
physically , it seems that the background field introduces an extra term ( whose sign depends on species ) into the force - balance , to allow for both the ion and electrons to be in force balance simultaneously , given one unique expression for the scalar potential .
the vector potential , magnetic field and current density used in this paper are as follows ( gh+b ) : @xmath50 the magnetic permeability _ in vacuo _ is given by @xmath51 and the characteristic magnetic field strength by @xmath52 .
the constant @xmath53 has units of inverse length , and we use @xmath54 to represent the characteristic length scale of the system ( @xmath55 ) ( see table 1 for a concise list of the dimensionless quantities used in this paper , all denoted with a tilde , @xmath56 )
. the dimensionless constant @xmath57 controls the strength of the background field in the @xmath1 direction , and as a result there are now two different interpretations to be made .
we could either consider the system as a gh flux tube of uniform twist embedded in an untwisted uniform background field , or consider the whole gh+b magnetic field as a non - uniformly twisted flux tube .
we note that flux tubes embedded in an axially directed background field have recently been observed during reconnection events in the earth s magnetotail , by the cluster spacecraft @xcite . in the first interpretation
, @xmath53 is a direct measure of the ` twist ' of the embedded flux tube ( see ref . ) , with the number of turns per unit length ( in @xmath1 ) along a field line given by @xmath58 @xcite . in the second interpretation
, we see that the system is not uniformly twisted , with the @xmath1 distance traversed when following a field line ( e.g. ) given by @xmath59 the fact that this depends on @xmath12 demonstrates that the system as a whole has non - uniform twist .
the number of turns per unit length in @xmath1 of the gh+b field : the ` twist ' is given by @xmath60 and is plotted in figure [ fig : twist ] for three values of @xmath7 .
since @xmath61 corresponds to the field - reversal regime , we see a mixture of positive and negative twists ( figure [ fig : twist1 ] ) .
however , for @xmath62 we see only negative values of the twist ( figures [ fig : twist2 ] and [ fig : twist3 ] ) , i.e. we travel in the negative @xmath1 direction as we wind round the gh+b flux tube in the anti - clockwise direction .
the magnetic field is plotted in figures [ fig:0.3]-[fig:0.5 ] for two values of @xmath7 .
the @xmath63 case contains a reversal of the @xmath64 field direction and as such is akin to a reversed field pinch ( e.g. see ref . for a laboratory interpretation ) : this configuration may be of use in the study of astrophysical jets , see ref .
for example .
the value @xmath65 gives zero @xmath64 at @xmath66 , and as such is the value that distinguishes the two different classes of field configuration , namely unidirectional ( @xmath62 ) or including field reversal ( @xmath61 ) .
the value of @xmath67 for which the @xmath64 field reverses is plotted in figure [ fig : rev ] .
the magnitude of the gh+b magnetic field is plotted in figure [ fig : bmag ] for three values of @xmath7 .
for all values of @xmath7 , @xmath68 for large @xmath67 , i.e. to a potential field .
the primary task of this paper is to calculate self - consistent collisionless equilibrium distribution functions for the gh+b field .
this problem essentially reduces to solving ampre s law such that equation ( [ eq : vlasov ] ) is satisfied .
we assume nothing about the electric field however , and in fact use that degree of freedom to solve ampre s law .
the resultant form of the scalar potential is then substituted into poisson s equation , to establish the final relationships between the microscopic and macroscopic parameters of the equilibrium .
although the ift method did not yield a self - consistent equilibrium df for the gh field without a background field , the outcome of the calculation can still be used as an indication of possible forms for the df for the gh+b field . using trial and error we arrived at the distribution function @xmath69,\label{eq : ansatz}\end{aligned}\ ] ] which is a superposition of two terms that are consistent macroscopically with a ` rigid - rotor ' ,
@xcite for example . a rigid - rotor is microscopically described by a df of the form @xmath70 ( with @xmath71 in the second term of the df in equation ( [ eq : ansatz ] ) ) .
each @xmath72 term corresponds to an average macroscopic motion of rigid rotation with angular frequency @xmath73 , and rectilinear motion with velocity @xmath74 .
the dimensionless constants @xmath75 , @xmath76 , @xmath77 and @xmath78 are yet to be determined , with @xmath79 for positivity of the distribution .
note that the thermal beta is @xmath80 and @xmath81 is the thermal velocity of species @xmath9 .
the ratio of the thermal larmor radius , @xmath82 ( for @xmath83 ) to the macroscopic length scale of the system @xmath84 , is given by @xmath85 typically known as the ` magnetisation parameter ' @xcite . in our system , the magnitude of the magnetic field and hence @xmath86 itself is spatially variable .
for the purposes of the calculations in this paper however , we set @xmath87 as a characteristic value ( see table [ tab : param ] for a concise list of the micro and macroscopic parameters of the equilibrium ) . by insisting on a specific magnetic field configuration ( the gh+b field ) we have made a statement on the macroscopic physics . in searching for the equilibrium df
, we are trying to understand the microscopic physics . in this sense
we are tackling an ` inverse problem ' .
once an assumption on the form of the df is made then should the assumed form be able to reproduce the correct moments this inverse problem reduces to establishing the relationships between the microscopic and macroscopic parameters of the equilibrium . in this section
we ` fix ' the free parameters of the df in equation ( [ eq : ansatz ] ) , such that maxwell s equations are satisfied ; @xmath88 note that the solenoidal constraint and faraday s law are automatically satisfied for the gh+b field in equilibrium , since @xmath17 implies that @xmath89 and @xmath16 implies that @xmath90 . in appendix [ app :
moments ] we have calculated the @xmath91 current density , found by summing first order moments in @xmath92 of the df .
we now substitute in the macroscopic expressions for @xmath93 , @xmath94 and @xmath95 from ( [ eq : current ] ) and ( [ eq : vecfield ] ) into the expression for the @xmath96 current density of equation ( [ eq : jzapp ] ) . after this substitution
, we can calculate a @xmath97 that makes the system consistent .
the substitution of the known expressions for @xmath91 , @xmath98 and @xmath99 gives @xmath100 in order to satisfy the above equality we can construct a solution by introducing a ` separation constant ' @xmath101 .
we multiply the above equation by @xmath102 which makes the left - hand side constant , whilst the right - hand side is a sum of two terms , one depending on ion parameters and the second depending on electron parameters
. then we can define @xmath103 by @xmath104 associating the ` ion term ' with the first term on the right - hand side of ( [ eq : a1a2 ] ) , and the ` electron term ' with the second term on the right - hand side of ( [ eq : a1a2 ] ) .
after some algebra we can rearrange these two associations to give two expressions for the scalar potential , one in terms of the ion parameters , and one in terms of the electron parameters : @xmath105\right\}\nonumber\\
\phi(r)&=&\frac{1}{q_e\beta_e}\ln\left\{\frac{\mu_0n_{0e}q_ev_{th , e}}{2\tau b_0\gamma_1}\left[{\tilde{{u}}_{ze}}e^{({\tilde{{u}}_{ze}}^2+\tilde{r}^2{\tilde{\omega}_{e}}^2)/2+{\tilde{\omega}_{e}}\tilde{r}^2k/\delta_e}\left(1+\tilde{r}^2\right)^{2-({\tilde{\omega}_{e}}-{\tilde{{u}}_{ze}})/(2\delta_e)}+\tilde{{v}}_{ze}{c}_ee^{\tilde{{v}}_{ze}^2/2}\left(1+\tilde{r}^2\right)^{2+\tilde{{v}}_{ze}/(2\delta_e ) } \right]\right\}\nonumber\end{aligned}\ ] ] the two values of the scalar potential above must be made identical by a suitable choice of relationships between the ion and electron parameters .
given enough freedom in parameter space , we could say that the @xmath1 component of ampre s law is _ implicitly _ solved the above equations , in that one just needs to choose a consistent set of parameters .
however , we seek a solution in an _ explicit _ sense . in order to make progress we non - dimensionalise the above equations by multiplying both sides by @xmath106 with @xmath107 once this
is done we can write the scalar potential in the form @xmath108^{\frac{e\beta_r}{q_i\beta_i}}\right\},\label{eq : phi3}\\ e\beta_r\phi(r)&=&\ln\left\{\left[\text{electron terms}\right]^{\frac{e\beta_r}{q_e\beta_e}}\right\}.\label{eq : phi4}\end{aligned}\ ] ] specifically , equations ( [ eq : phi3 ] ) and ( [ eq : phi4 ] ) require the equality of the arguments of the logarithm to hold in order for a meaningful solution to be obtained for the scalar potential .
a first step towards this is made by requiring consistent powers of the @xmath109 ` profile ' in the right - hand side of the above expression to allow factorisation .
hence @xmath110 and hence the rigid - rotation , @xmath75 , is fixed by the difference of the rectilinear motion , @xmath111 . on top of this , we require that the power of the @xmath109 ` profile ' on the right - hand side is the same for both the ions and electrons , thus @xmath112 this condition seems to be a statement on an average potential energy associated with the particles . once more to allow factorisation of the @xmath109 ` profile ' , we insist that net @xmath113 terms cancel , i.e. @xmath114 the physical meaning of this condition seems to be that the frequencies of the rigid rotor for each species are matched according to the relevant magnetisation , and the background field magnitude .
the remaining task is to ensure equality of the ` coefficients ' @xmath115\right\}^{\frac{e\beta_r}{q_i\beta_i}}={\mathcal{d}}=\left\{- \frac{1}{4\delta_e\gamma_1 } \frac{n_{0e}m_ev_{th , e}^2}{b_0 ^ 2/(2\mu_0 ) } \left[{\tilde{{u}}_{ze}}e^{{\tilde{{u}}_{ze}}^2/2}+\tilde{{v}}_{ze}{c}_ee^{\tilde{{v}}_{ze}^2/2 } \right]\right\}^{\frac{e\beta_r}{q_e\beta_e}}\label{eq : d1fix}\end{aligned}\ ] ] these seem to be conditions on the ratios of the energy densities associated with the bulk rectilinear motion and the magnetic field respectively .
thus far we have 8 constraints and 12 unknowns ( @xmath116 ) given fixed characteristic macroscopic parameters of the equilibrium @xmath52 , @xmath53 , and @xmath7 .
we can now write down an expression for @xmath18 that explicitly solves the @xmath1 component of ampre s law ; @xmath117 with @xmath118 clearly , we require that @xmath119 for the expression above to make sense .
it is clear that the sign of @xmath103 could , in principle , affect the sign of @xmath120 .
it is seen from ( [ eq : d1fix ] ) that positivity of @xmath120 implies that @xmath121&>&0,\\ \frac{1}{\gamma_1}\left [ \tilde{u}_{ze}e^{\tilde{u}_{ze}^2/2}+\tilde{v}_{ze}c_ee^{\tilde{v}_{ze}^2/2 } \right]&<&0.\end{aligned}\ ] ] by rearranging the above inequalities to make @xmath122 the subject , it can be seen after some algebra that positivity of @xmath120 and @xmath122 is guaranteed when @xmath123 note that these conditions are sufficient , but not necessary , i.e. it is possible to have @xmath119 and @xmath124 for any value of @xmath125 , and even for @xmath126 in the case of @xmath127 .
thus far we have only considered the @xmath96 component , and it is premature to consider all components of ampre s law satisfied . let us move on to consider the @xmath2 component . in a process similar to that above , we substitute in the macroscopic expressions for @xmath128 , @xmath94 and @xmath95 for the gh+b field into the expression for the @xmath129 current density of equation ( [ eq : jthetageneral ] ) in appendix [ app : moments ] . after this substitution
, we can once more calculate the @xmath18 that makes the system consistent .
the substitution gives @xmath130 using the parameter relations as above , we determine that the scalar potential is again given in the form of ( [ eq : scalarpot ] ) , @xmath131 hence , this form of the scalar potential is consistent provided @xmath132^{\frac{e\beta_r}{q_i\beta_i}}={\mathcal{d}}=\left [ -\frac{1}{\gamma_2 } \frac{1}{4\delta_e}\frac{n_{0e}m_ev_{th , e}\omega_e/\tau}{b_0 ^ 2/(2\mu_0 ) } e^{{\tilde{{u}}_{ze}}^2/2 } \right]^{\frac{e\beta_r}{q_e\beta_e}}\label{eq : d2fix}\ ] ] for @xmath133 another separation constant .
these seem to be conditions on the ratios of the energy densities associated with the bulk rotation and the magnetic field respectively .
this has added two more constraints .
once again we must ensure that @xmath119 .
since @xmath134 , the right - hand side of the above equation implies that @xmath135 to ensure that @xmath119 . whilst the left - hand side implies that @xmath136 for positivity of @xmath120 since @xmath137 .
hence we can say that for positivity @xmath138 we can now consider ampre s law satsified , given a @xmath18 that solves poisson s equation . as a result ,
the problem of consistency is now shifted to solving poisson s equation , where the remaining degrees of freedom lie .
the final step in ` self - consistency ' is to solve poisson s equation .
frequently in such equilibrium studies , this step is replaced by satisfying quasineutrality and in essence solving a first order approximation of poisson s equation , see for example refs . .
here we solve poisson s equation exactly , i.e. to all orders .
poisson s equation in cylindrical coordinates with only radial dependence gives @xmath139 the electric field is calculated as @xmath140 , giving @xmath141 we can now take the divergence of the electric field @xmath142 and so @xmath143 this gives a non - zero net charge per unit length ( in @xmath1 ) of @xmath144 the charge density derived in equation ( [ eq : edivergence ] ) must equal the charge density calculated by taking the zeroth moment of the df .
the expression for the charge density calculated in ( [ eq : sigmaapp ] ) gives @xmath145 the second equality is found by substituting the form of the vector potential from equation ( [ eq : vecfield ] ) , and the final equality is reached by using the conditions derived in equations ( [ eq : omegafix ] ) - ( [ eq : scalarpot ] ) .
we can now match equations ( [ eq : edivergence ] ) and ( [ eq : chargedensity ] ) to get @xmath146 we now have 12 physical parameters ( @xmath147 ) with 11 constraints ( [ eq : omegafix]-[eq : d1fix ] ) , ( [ eq : d2fix ] ) & ( [ eq : poisscon ] ) .
for example , if one picks @xmath52 , @xmath53 , @xmath7 and one microscopic parameter , say @xmath148 , then the remaining parameters of the equilibrium , ( @xmath149 , @xmath150 , @xmath151 ) , are now determined .
one could of course choose the values of a different set of parameters , and determine those that remain by using the constraints derived .
note that whilst the constants @xmath125 and @xmath152 are system parameters , they are not physically meaningful as they only represent a change in the gauge of the scalar potential .
it is seen from equations ( [ eq : edivergence ] ) and ( [ eq : coulomb ] ) that basic electrostatic properties of the equilibrium described by @xmath4 are encoded in @xmath153 .
the equilibrium is electrically neutral only when @xmath154 , and non - neutral otherwise .
specifically , there is net negative charge when @xmath155 , and net positive charge when @xmath156 .
this net charge is finite in the @xmath157 plane and given by @xmath158 in equation ( [ eq : coulomb ] ) .
physically , the sign of @xmath153 seems to be related to the respective magnitudes of the bulk rotation frequencies , @xmath75 . from equations ( [ eq : omegafix ] ) and ( [ eq : gfix ] ) we see that @xmath155 implies that latexmath:[\ ] ] we see again that @xmath230 has unit slope for two values of @xmath248 . once again , after some graphical consideration of the problem , it becomes apparent that @xmath249 should be bounded above and below for multiple maxima .
after elementary consideration of the functional form of ( [ eq : linearz ] ) , for example with graph plotting software we see that multiple maxima in the @xmath186 direction can only occur , for a given @xmath67 , when @xmath243 ( and hence @xmath187 ) satisfies these inequalities for ions @xmath250 and these for electrons @xmath251
[ cols="^,^",options="header " , ] 0.25 ) of the gh+b field for three values of @xmath7 .
[ fig : twist1 ] shows the twist for @xmath61 , and as such there are both negative and positive twists , due to the field reversal .
[ fig : twist2 ] and [ fig : twist3 ] both show negative twist , since there is no magnetic field reversal .
, title="fig : " ] 0.25 ) of the gh+b field for three values of @xmath7 .
[ fig : twist1 ] shows the twist for @xmath61 , and as such there are both negative and positive twists , due to the field reversal .
[ fig : twist2 ] and [ fig : twist3 ] both show negative twist , since there is no magnetic field reversal .
, title="fig : " ] 0.25 ) of the gh+b field for three values of @xmath7 .
[ fig : twist1 ] shows the twist for @xmath61 , and as such there are both negative and positive twists , due to the field reversal .
[ fig : twist2 ] and [ fig : twist3 ] both show negative twist , since there is no magnetic field reversal .
, title="fig : " ] 0.17 and [ fig:0.5 ] show the gh+b magnetic field in the @xmath252 plane , for two values of @xmath7 .
the curved arrows indicate the direction of the @xmath253 components , whilst the blue - black - red shading denotes the magnitude and direction of the @xmath64 component .
the @xmath63 case contains a reversal of the @xmath64 field direction and as such is a reversed field pinch whilst @xmath254 gives zero @xmath64 at @xmath66 .
[ fig : rev ] shows the radius at which @xmath64 changes its direction , for @xmath255 .
@xmath64 does not reverse for @xmath62 .
, title="fig : " ] 0.17 and [ fig:0.5 ] show the gh+b magnetic field in the @xmath252 plane , for two values of @xmath7 .
the curved arrows indicate the direction of the @xmath253 components , whilst the blue - black - red shading denotes the magnitude and direction of the @xmath64 component .
the @xmath63 case contains a reversal of the @xmath64 field direction and as such is a reversed field pinch whilst @xmath254 gives zero @xmath64 at @xmath66 .
[ fig : rev ] shows the radius at which @xmath64 changes its direction , for @xmath255 .
@xmath64 does not reverse for @xmath62 .
, title="fig : " ] 0.25 and [ fig:0.5 ] show the gh+b magnetic field in the @xmath252 plane , for two values of @xmath7 .
the curved arrows indicate the direction of the @xmath253 components , whilst the blue - black - red shading denotes the magnitude and direction of the @xmath64 component .
the @xmath63 case contains a reversal of the @xmath64 field direction and as such is a reversed field pinch whilst @xmath254 gives zero @xmath64 at @xmath66 .
[ fig : rev ] shows the radius at which @xmath64 changes its direction , for @xmath255 .
@xmath64 does not reverse for @xmath62 .
, title="fig : " ] 0.25 -[fig : bmag3 ] show the magnitude of the gh+b magnetic field for @xmath256 and @xmath192 respectively , normalised by @xmath52 . for @xmath257 , @xmath68 from above , whereas for @xmath62 , @xmath68 from below.,title="fig : " ] 0.25 -[fig : bmag3 ] show the magnitude of the gh+b magnetic field for @xmath256 and @xmath192 respectively , normalised by @xmath52 . for @xmath257 , @xmath68 from above , whereas for @xmath62 , @xmath68 from below.,title="fig : " ] 0.25 -[fig : bmag3 ] show the magnitude of the gh+b magnetic field for @xmath256 and @xmath192 respectively , normalised by @xmath52 . for @xmath257 , @xmath68 from above , whereas for @xmath62 , @xmath68 from below.,title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath259 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath259 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath259 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath259 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath259 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath259 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath259 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath259 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath261 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath261 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath261 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath261 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath261 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath261 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath261 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath261 ) and @xmath196 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium with field reversal ( @xmath258 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
the white horizontal / vertical lines indicate the regions in which multiple maxima in either the @xmath189 or @xmath186 directions can occur , if at all .
a single line indicates that the ` region ' is a line .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
note that there are not any multiple maxima in this case .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
note that there are not any multiple maxima in this case .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
note that there are not any multiple maxima in this case .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
note that there are not any multiple maxima in this case .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
note that there are not any multiple maxima in this case .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
note that there are not any multiple maxima in this case .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
note that there are not any multiple maxima in this case .
, title="fig : " ] 0.245 in @xmath197 space for an equilibrium without field reversal ( @xmath260 ) , for a variety of parameters ( @xmath262 ) and @xmath263 .
note that there are not any multiple maxima in this case .
, title="fig : " ] 63ifxundefined [ 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 10.1063/1.3077307 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.135003 [ * * , ( ) ] link:\doibase 10.1063/1.3268771 [ * * , ( ) ] link:\doibase 10.1063/1.3623740 [ * * , ( ) ] link:\doibase 10.1063/1.4826502 [ * * , ( ) ] link:\doibase 10.1063/1.4935488 [ * * , ( ) ] link:\doibase 10.1063/1.4934611 [ * * , ( ) ] , link:\doibase 10.1017/s0022377816000519 [ * * , ( ) ] link:\doibase 10.1063/1.3552141 [ * * , ( ) ] link:\doibase 10.1103/physrev.45.890 [ * * , ( ) ] link:\doibase 10.1017/s0022377801001507 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1017/s002237780002554x [ * * , ( ) ] link:\doibase 10.1515/zna-1962 - 1004 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1016/0003 - 4916(60)90023 - 3 [ * * , ( ) ] link:\doibase 10.1142/p251 [ _ _ ] ( , ) link:\doibase 10.1088/1751 - 8113/40/27/f08 [ * * , ( ) ] link:\doibase 10.1140/epjd / e2014 - 50007 - 9 [ * * , ( ) ] , link:\doibase 10.1063/1.4928111 [ * * , ( ) ] link:\doibase 10.1016/0375 - 9601(76)90315 - 7 [ * * , ( ) ] link:\doibase 10.1063/1.4958319 [ * * , ( ) ] link:\doibase 10.1063/1.861887 [ * * , ( ) ] link:\doibase 10.1103/physreva.31.2556 [ * * , ( ) ] link:\doibase 10.1063/1.1693161 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1086/341539 [ * * , ( ) ] link:\doibase 10.1086/367611 [ * * , ( ) ] link:\doibase 10.1086/168805 [ * * , ( ) ] link:\doibase 10.1029/2007ja012684 [ * * , ( ) ] link:\doibase 10.1029/gl013i008p00801 [ * * , ( ) ] link:\doibase 10.1029/gl017i001p00049 [ * * , ( ) ] link:\doibase 10.1016/0032 - 0633(89)90116 - 5 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1086/501499 [ * * , ( ) ] , link:\doibase 10.1063/1.2428280 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.116.235001 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.84.3326 [ * * , ( ) ] link:\doibase 10.1098/rspa.2014.0913 [ * * , ( ) ] , link:\doibase 10.1007/lrsp-2009 - 4 [ * * ( ) , 10.1007/lrsp-2009 - 4 ] \doibase 10.1051/0004 - 6361:20030692 [ * * , ( ) ] link:\doibase 10.1086/344799 [ * * , ( ) ] , link:\doibase 10.3847/0004 - 637x/817/1/5 [ * * , ( ) ] , @noop * * , ( ) @noop _ _ ( , ) link:\doibase 10.1063/1.2988338 [ * * , ( ) ] @noop * * , ( ) @noop _ _ ( ) link:\doibase 10.1007/lrsp-2006 - 1 [ * * ( ) , 10.1007/lrsp-2006 - 1 ] @noop * * , ( ) @noop _ _ ( , ) @noop `` , '' ( ) @noop * * , ( ) link:\doibase 10.5194/angeo-30 - 761 - 2012 [ * * , ( ) ] @noop _ _ ( , ) link:\doibase 10.1142/9789814644839_0009 [ _ _ ] ( , ) chap . , pp .
@noop _ _ ( , ) https://books.google.co.uk/books?id=veotr-i_teyc[__ ] , cambridge atmospheric and space science series ( , ) link:\doibase 10.1063/1.1692077 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.95.245004 [ * * , ( ) ] link:\doibase 10.1063/1.1978888 [ * * , ( ) ] link:\doibase 10.1063/1.2186187 [ * * , ( ) ] , | we calculate exact one - dimensional collisionless plasma equilibria for a continuum of flux tube models , for which the total magnetic field is made up of the ` force - free ' gold - hoyle magnetic flux tube embedded in a uniform and anti - parallel background magnetic field . for a sufficiently weak background magnetic field
, the axial component of the total magnetic field reverses at some finite radius .
the presence of the background magnetic field means that the total system is not exactly force - free , but by reducing its magnitude the departure from force - free can be made as small as desired .
the distribution function for each species is a function of the three constants of motion ; namely the hamiltonian and the canonical momenta in the axial and azimuthal directions .
poisson s equation and ampre s law are solved exactly , and the solution allows either electrically neutral or non - neutral configurations , depending on the values of the bulk ion and electron flows .
these equilibria have possible applications in various solar , space and astrophysical contexts , as well as in the laboratory .
= 1 |
coronal ( or x - ray ) jets are collimated , transient eruptions of hot plasma observed predominantly in coronal holes and at the periphery of active regions , driven by flux emergence or cancellation ( e.g. , * ? ? ?
`` standard jets '' are launched by reconnection between open magnetic fields and expanding closed fields , while the more energetic `` blowout jets '' additionally involve an eruption of the closed fields @xcite . for detailed descriptions of coronal jets
see the recent reviews by @xcite , @xcite , and @xcite .
coronagraph and in - situ observations revealed that coronal jets can produce signatures in the outer corona , sometimes even at earth or beyond @xcite .
it was therefore suggested that they contribute to powering the corona and solar wind ( e.g. , * ? ? ?
* ; * ? ? ?
quantifying this contribution from observations is difficult , and only few attempts have been made so far .
@xcite obtained a `` crude estimate '' for the energy budget of coronal jets , which was @xmath1 times smaller than required for maintaining the sw .
@xcite suggested a mass - contribution of coronal jets to the sw of @xmath210% , though @xcite remarked that this value should be 2.5 times smaller . using heliospheric images from smei @xcite , @xcite estimated mass and energy contributions of 3.2% and 1.6% , respectively . finally , estimated the energy provided by coronal jets to be less than 1% of what is required to heat a polar coronal hole .
they emphasized , though , that many high - temperature jets may be missed by the observations @xcite .
thus , while it appears that the contribution of coronal jets to the sw may be small compared to the potential contribution by the much more numerous jet - like events on smaller scales @xcite , its exact amount remains uncertain .
numerical models provide a powerful tool to address this question .
however , mhd simulations of coronal jets were so far restricted to a limited treatment of the energy transfer in the corona and to heights far below the sw - acceleration region .
recently , @xcite presented mhd simulations of coronal jets that overcame both limitations . using these simulations ,
we evaluate here the mass and energy contribution of coronal jets to the sw , and compare our results to previous observational estimates .
the simulations we use for our analysis are described in @xcite , and a more detailed account will be provided in a forthcoming publication ( trk _ et al .
_ , in preparation ) .
they were performed using the 3d , viscous , resistive mhd model in spherical coordinates developed at predictive science inc . @xcite .
the model incorporates thermal conduction parallel to the magnetic field , radiation losses , and parametric coronal heating .
the nonuniform grid covers 120@xmath3 , where @xmath3 is the solar radius , using @xmath4 points in @xmath5 .
the smallest radial grid interval at @xmath6 is @xmath7 km ; the angular resolution is @xmath8 in the area containing the jet
. a purely radial coronal magnetic field , produced by a magnetic monopole located at sun center is prescribed .
starting from an initial thermodynamic sw - solution @xcite , we relax the system to a spherically symmetric , steady - state solar wind .
afterwards , using the method of @xcite , we emerge a flux rope whose structure and evolution was previously computed with the model of @xcite .
this is done by prescribing the values of the magnetic field and calculating the velocities using the characteristic equations at @xmath9 . ) .
( b ) magnitude of elssser variable in a plane intersecting the jet ; same time as in ( a ) .
( c ) same as ( a ) 1.66hrs later , showing coronal signatures of the blowout jet .
( d ) same as ( b ) at the time shown in ( c ) .
the times correspond to peak energy fluxes at @xmath10 ( cf .
figures[fig - poynting_wave ] and [ fig - f1_f2 ] ) . ,
scaledwidth=47.7% ] here we analyze three runs . in all cases ,
the system is in a steady - state at @xmath11 , when the flux emergence is started . in our main run , _ jet 1 _ ,
emergence is imposed for 4.5hrs , and the evolution is modeled for a total of 12hrs . the simulation produces a standard jet at @xmath12h , launched by the first strong occurrence of magnetic reconnection in the current layer that forms between the emerging flux and the open coronal field .
several reconnection episodes follow , driven by the continuous expansion of the emerging flux rope . at @xmath13h a blowout jet is produced , associated with the eruption of the rope , which again is followed by several reconnection episodes , now occurring in the current layer that forms below the eruption .
the blowout jet is much stronger than the standard jet , with a peak kinetic energy of @xmath14erg , compared to only @xmath15erg for the standard jet .
plasma is ejected upwards by reconnection outflows during the jets , and additionally by the flux - rope eruption during the blowout jet .
both jets produce alfvn waves , which travel much faster than the ejected plasma , due to the relatively large coronal alfvn speed in our model ( several 1000kms@xmath16 ) . in figure[fig - jet1 ]
we show manifestations of the jets in the outer corona , using synthetic difference - images of the polarization brightness @xcite and the magnitude of the elssser variable , @xmath17 , which represents an outward propagating alfvn perturbation along a radial magnetic field line ( e.g. , * ? ? ?
* ) . at @xmath18h an elongated brightness - enhancement and a clear wave - signature
are visible in panels ( a ) and ( b ) , respectively .
the former is associated with plasma ejected by the standard jet , while the latter is associated with the alfvn wave triggered by the blowout jet ( the significantly weaker alfvn wave triggered by the standard jet has already left the field - of - view at this time ) . at @xmath19h
the plasma ejected by the blowout jet becomes visible , while there are no strong wave - signatures anymore .
the brightness - image shows density - enhancements and rarefactions , reminiscent of narrow coronal mass ejections and `` streamer blobs '' ( e.g. , * ? ? ?
* ) . our second run , _ jet 2 _ , has the same settings as _ jet 1 _ , but here we stop the flux emergence after 2hrs , so that only the standard jet occurs .
this run is calculated for 6hrs . in both runs ohmic heating
is included , so that magnetic energy dissipated through the resistivity ( corresponding to a resistive diffusion time @xmath20 hrs ) contributes to the energy equation . finally , in _ jet 3 _
, we again impose the emergence for4.5 hrs , but turn off ohmic heating .
an inspection of this run reveals that it is essentially a lower - energy version of _ jet 1 _ , so we stopped the calculation shortly after the occurrence of the blowout jet , at @xmath21h . at @xmath22 , @xmath23 , @xmath24 , which contains the jet at all times . for our analysis
we then select an area @xmath25 that corresponds to the jet s maximum transverse expansion .
, scaledwidth=47.7% ]
following @xcite and , we calculate the mass and energy fluxes associated with our jets , and integrate them in time and space to find the total mass and energy injected into the sw .
different from these authors , we perform our analysis high in the corona , at @xmath10 , and use the poynting flux , rather than the wave - energy flux , for a more accurate calculation of the total energy flux ( [ ss : wave_vs_poynting ] ) . for the integration we use a spherical quadrangle @xmath26 at @xmath27 that contains the jet - signatures at all times ( figure[fig - flux ] ) .
the area @xmath25 corresponds to the maximum lateral expansion of the blowout jet as it passes through @xmath26 at @xmath28h ( see figure [ fig - jet1](c ) ) .
we define @xmath25 as the area within which the perturbation in the sum of the kinetic and magnetic energies is larger than @xmath29 at that time . as in the above - mentioned papers , we consider the kinetic , potential , enthalpy , and wave energy fluxes : @xmath30 and the radial component of the poynting flux , @xmath31 fluxes due to radiation and thermal conduction are neglected . for each flux
@xmath32 we determine the constant background @xmath33 produced by the steady - state sw ( which is zero for @xmath34 and @xmath35 ) .
then we calculate the net power through @xmath26 due to the fluxes by integrating @xmath36 and removing the background : @xmath37 dividing each @xmath38 by @xmath25 , we obtain the average , background - subtracted energy fluxes associated with the jets as functions of time : @xmath39 for _ jet 13_. the magenta line shows the respective fluxes of the background solar wind . ]
first we compare in figure [ fig - poynting_wave ] the wave - energy flux with the radial poynting flux . for the former we use the transverse velocities and the magnetic field strength .
we see that the wave - energy flux is always smaller than the poynting flux .
in fact , the wave - energy flux formula assumes equipartition of ( perpendicular ) magnetic and kinetic energy , therefore missing some energy contribution when this condition does not fully apply .
henceforth we use only the poynting flux to measure the wave energy injected into the sw .
prior to the occurrence of the standard jet , no significant wave flux is transferred into the sw . as mentioned above ,
alfvn waves launched low in the corona require only a few minutes to arrive at @xmath26 in our model .
consequently , we see for all runs the first clear signal shortly after the first strong reconnection associated with the onset of the standard jet at @xmath40h .
_ jet 2 _ , for which the flux emergence is terminated at @xmath41h , does not produce any significant signal afterwards . for _
jet1 _ and _ jet3 _ , activity increases strongly shortly after @xmath42h , marking stronger reconnection episodes associated with the increasing expansion of the emerging flux rope prior to its eruption .
the strongest peaks are around @xmath43h , when signals associated with the blowout jet arrive at @xmath26 . for _ jet1 _ , which was calculated for a longer time than _ jet3 _ ,
the following decay phase contains two secondary peaks .
the first one results from small reconnection events in the current sheet that forms below the erupting flux rope , while the second one may be related to a third jet that grows out of the flaring current sheet ( trk _ et al .
_ , in preparation )
. the overall evolution of the poynting flux is very similar for _ jet1 _ and _ jet3 _ , suggesting a negligible influence of ohmic heating on the strength of the alfvn waves generated by reconnection .
a clear deviation can be seen only during the strongest peak , which is associated also with the flux - rope eruption .
figure[fig - f1_f2 ] shows @xmath44 , @xmath45 , and @xmath46 together with the background sw flux .
_ jet 2 _ shows relatively weak activity , associated with plasma accelerated during the standard jet , which needs about 2hrs to arrive at @xmath26 .
very similar activity is visible for _ jet 1 _ but much less for _ jet 3 _ , which lacks ohmic heating and produces weaker plasma upflows .
an additional increase , coinciding with the major peak in the poynting flux ( figure [ fig - poynting_wave](b ) ) , occurs in @xmath44 at @xmath47h for _ jet 1 _ and _ jet 3_. its absence in the other fluxes indicates that it is due to a ( transverse ) velocity increase rather than a density increase , as the alfvn wave associated with the blowout jet passes through @xmath26 . indeed , @xmath48 at this time , comparable to the background sw speed ( @xmath49 ) , while @xmath50 and @xmath51 increase by 10 and 30% at most , respectively . _ jet 1 _ then shows a strong increase of all fluxes for @xmath52h , when the blowout jet ( i.e. , the erupted flux rope ) passes through @xmath26 ( which at @xmath19 increases @xmath50 and @xmath51 by factors of up to 1.2 and 3.1 , respectively ) .
the flux maxima ( @xmath19h ) are very similar and about one order of magnitude smaller than the peak poynting flux associated with the preceding alfvn wave ( cf .
figure[fig - poynting_wave](b ) ) ; their ratios change if the analysis is performed at a different radius . ] . afterwards
all fluxes show a strong oscillation and a final slow decay .
the oscillation has similar amplitude for all fluxes , and is due to the density enhancements and rarefactions visible in figure[fig - jet1](c ) .
all this activity is barely noticeable in the poynting flux ( @xmath53 = 28kms@xmath16 at @xmath19 , much smaller than at @xmath18 ) . for @xmath54h @xmath44 remains smaller than @xmath55 , since @xmath50 falls below the sw speed at @xmath56 once the ejecta has passed . in order to compare our results with observational estimates , we calculate the mass and total energy injected during _
jet 1 _ into the sw : @xmath57 where @xmath58 is the average mass flux calculated from @xmath59 according to eqs .
( [ eq - pt]-[eq - fbar ] ) .
figure[fig - total ] shows the mass flux and the total energy flux as functions of time . to calculate @xmath60 and @xmath61
, we extrapolate after @xmath62h the integrands to infinity and obtain @xmath63 and @xmath64 , the latter corresponding to @xmath21% of the magnetic - energy release during the blowout jet .
@xmath60 equals the order - of - magnitude estimate obtained from uvcs observations at @xmath65 by @xcite .
in contrast , both @xmath60 and @xmath61 are about one order of magnitude smaller than the mass and kinetic energy estimated by @xcite . assuming that _ jet 1 _ represents an average coronal jet , a jet - occurrence frequency @xmath66 @xcite , and a fully open magnetic field at @xmath56
, we obtain a total energy flux @xmath67 and a proton - density flux @xmath68 , where @xmath69 is the proton mass and @xmath70 the spherical surface at @xmath56 . scaling these fluxes to 1 astronomical unit ( au ) , assuming spherical sw - expansion at constant speed
, we get @xmath71 @xmath72 is significantly smaller than the 1au estimate @xmath73 by @xcite and @xcite and the correction by @xcite . ] . in order to compare with the energy - flux estimate by , which was obtained for the low corona
, we calculate @xmath74 for a typical height extension of coronal jets , @xmath75 @xcite , and employ the same assumptions for jet frequency and coronal hole area as those authors .
we get @xmath76 , in very good agreement with their value @xmath77 .
finally , we compare our results with sw fluxes derived from in - situ spacecraft measurements . according to @xcite , the total energy flux of the sw at 1au
is @xmath21.5erg@xmath78s@xmath16 , and very similar for the slow and fast solar wind .
the proton - density flux varies by a factor of about 2 between the slow and fast wind , and falls into the range @xmath79 ( * ? ? ?
* ; * ? ? ?
* and references therein ) .
comparing these values to our jet - produced fluxes scaled to 1au ( equation[eq:1au ] ) , we estimate that coronal jets provide @xmath80% and @xmath0% to the total energy and mass content of the sw , respectively .
we used an mhd simulation with an advanced energy equation and a large spherical grid to estimate the mass and energy contributions of coronal jets to the sw . the simulation ( _ jet 1 _ ) produced both a standard and a blowout jet and included ohmic heating .
we followed the previous , observational analysis by , except that we calculated mass and energy fluxes at @xmath27 rather than in the low corona , and used the poynting flux instead of the wave - energy flux to obtain the total energy flux .
our results can be summarized as follows .
\(1 ) the blowout jet produces a strong perturbation in the sw , the standard jet only to a minor one .
this is not surprising , as the former is much more energetic than the latter in our simulation ( the peak kinetic energies differ by almost two orders of magnitude ) .
such a large difference is in line with the results of @xcite , who estimated the magnetic energy released during a blowout jet to be about 2030 times larger than for a standard jet .
\(2 ) the blowout jet ( and associated flux - rope eruption ) triggers a strong alfvn wave that arrives at @xmath56 more than one hour before the ejecta .
this significant lag is due to the relatively large coronal alfvn speed in our model , and is expected to decrease if more realistic values for the alfvn speed in coronal holes are chosen .
the wave yields the dominant contribution to the poynting flux , while the contributions of alfvn waves triggered by the standard jet and by interim reconnection events remain relatively small .
\(3 ) the bulk kinetic energy , potential energy , and enthalpy fluxes are also dominated by the blowout jet .
in contrast to the poynting flux , the evolution of these fluxes is characterized by a pronounced oscillation , which reflects plasma density enhancements and rarefactions associated with the eruption .
( 4 ) the mass and total energy contributions of the blowout jet ( the contributions by the standard jet are negligible ) at @xmath56 are @xmath81 and @xmath82 , respectively .
these are only @xmath83 of the heliospheric mass and kinetic energy estimates by @xcite , but those authors considered three of the largest jets during a three - week _ hinode _ campaign @xcite . on the other hand , @xmath60 agrees with the order - of - magnitude estimate by @xcite obtained at @xmath84 .
\(5 ) using observed occurrence - rates of coronal jets , we estimated ranges for the total energy and proton - density fluxes produced by these events in the low corona and at 1au .
we found very good agreement with the low - corona energy - flux estimate by .
our estimate for the proton - density flux at 1au is almost an order of magnitude smaller than the values given by @xcite , @xcite , but it must be kept in mind that their estimate was obtained merely from _
hinode_/xrt observations .
( 6 ) comparing our results with in - situ measurements of sw fluxes , we estimate that coronal jets provide @xmath80% and @xmath0% to the total energy and mass content of the sw , respectively .
our results are largely consistent with existing observational estimates of the contribution of coronal jets to the mass and energy content of the sw , and thus support the present conjecture that the these contribution are relatively small .
however , it must be kept in mind that ( i ) obtaining such estimates requires simplifying assumptions , ( ii ) the measurement uncertainties are large , and ( iii ) our simulation is relatively idealized and only one parameter - set of the model ( assumed to be representative ) was considered so far .
more advanced observations ( for instance from the upcoming solar orbiter and solar probe plus missions ) and simulations are needed to finally pin down the contribution of coronal jets to the sw .
we thank the referee for helpful comments and acknowledge support by nasa s h - sr and lws programs .
m.g.l . and j.e.l .
were is part supported by the chief of naval research .
this work benefited from discussions within the issi international team 258 `` understanding solar jets and their role in atmospheric structure and dynamics '' . , j. w. , golub , l. , lundquist , l. , van ballegooijen , a. , savcheva , a. , shimojo , m. , deluca , e. , tsuneta , s. , sakao , t. , reeves , k. , weber , m. , kano , r. , narukage , n. , & shibasaki , k. 2007 , science , 318 , 1580 , b. v. , buffington , a. , hick , p. p. , altrock , r. c. , figueroa , s. , holladay , p. e. , johnston , j. c. , kahler , s. w. , mozer , j. b. , price , s. , radick , r. r. , sagalyn , r. , sinclair , d. , simnett , g. m. , eyles , c. j. , cooke , m. p. , tappin , s. j. , kuchar , t. , mizuno , d. , webb , d. f. , anderson , p. a. , keil , s. l. , gold , r. e. , & waltham , n. r. 2004 , , 225 , 177 , n .- e .
2011 , in astronomical society of the pacific conference series , vol .
437 , solar polarization 6 , ed .
j. r. kuhn , d. m. harrington , h. lin , s. v. berdyugina , j. trujillo - bueno , s. l. keil , & t. rimmele , 99 raouafi , n. e. , patsourakos , s. , pariat , e. , young , p. r. , sterling , a. c. , savcheva , a. , shimojo , m. , moreno - insertis , f. , devore , c. r. , archontis , v. , trk , t. , mason , h. , curdt , w. , meyer , k. , dalmasse , k. , & matsui , y. 2016 , space science reviews , 1 , k. , ishido , y. , acton , l. w. , strong , k. t. , hirayama , t. , uchida , y. , mcallister , a. h. , matsumoto , r. , tsuneta , s. , shimizu , t. , hara , h. , sakurai , t. , ichimoto , k. , nishino , y. , & ogawara , y. 1992 , , 44 , l173 , k. , nakamura , t. , matsumoto , t. , otsuji , k. , okamoto , t. j. , nishizuka , n. , kawate , t. , watanabe , h. , nagata , s. , ueno , s. , kitai , r. , nozawa , s. , tsuneta , s. , suematsu , y. , ichimoto , k. , shimizu , t. , katsukawa , y. , tarbell , t. d. , berger , t. e. , lites , b. w. , shine , r. a. , & title , a. m. 2007 , science , 318 , 1591 , o. c. , howard , r. a. , simnett , g. m. , gurman , j. b. , plunkett , s. p. , sheeley , n. r. , schwenn , r. , koomen , m. j. , brueckner , g. e. , michels , d. j. , andrews , m. , biesecker , d. a. , cook , j. , dere , k. p. , duffin , r. , einfalt , e. , korendyke , c. m. , lamy , p. l. , lewis , d. , llebaria , a. , lyons , m. , moses , j. d. , moulton , n. e. , newmark , j. , paswaters , s. e. , podlipnik , b. , rich , n. , schenk , k. m. , socker , d. g. , stezelberger , s. t. , tappin , s. j. , thompson , b. , & wang , d. 1997 , in esa special publication , vol .
415 , correlated phenomena at the sun , in the heliosphere and in geospace , ed .
a. wilson , 103 , t. , lionello , r. , titov , v. s. , leake , j. e. , mikic , z. , linker , j. a. , & linton , m. g. 2016 , in astronomical society of the pacific conference series , vol .
504 , coimbra solar physics meeting : ground - based solar observations in the space instrumentation era , ed .
i. dorotovic , c. fischer , & m. temmer , 185 | transient collimated plasma eruptions in the solar corona , commonly known as coronal ( or x - ray ) jets , are among the most interesting manifestations of solar activity .
it has been suggested that these events contribute to the mass and energy content of the corona and solar wind , but the extent of these contributions remains uncertain .
we have recently modeled the formation and evolution of coronal jets using a three - dimensional ( 3d ) magnetohydrodynamic ( mhd ) code with thermodynamics in a large spherical domain that includes the solar wind .
our model is coupled to 3d mhd flux - emergence simulations , i.e , we use boundary conditions provided by such simulations to drive a time - dependent coronal evolution .
the model includes parametric coronal heating , radiative losses , and thermal conduction , which enables us to simulate the dynamics and plasma properties of coronal jets in a more realistic manner than done so far .
here we employ these simulations to calculate the amount of mass and energy transported by coronal jets into the outer corona and inner heliosphere . based on observed jet - occurrence rates
, we then estimate the total contribution of coronal jets to the mass and energy content of the solar wind to @xmath0% and ( 0.3 - 1.0)% , respectively .
our results are largely consistent with the few previous rough estimates obtained from observations , supporting the conjecture that coronal jets provide only a small amount of mass and energy to the solar wind .
we emphasize , however , that more advanced observations and simulations are needed to substantiate this conjecture . |
the recent advance in submicron technology enables fabrication of truly one dimensional ( 1d ) quantum wires .
the electron liquids in these systems are usually described in terms of the tomonaga
luttinger ( tl ) model@xcite .
edge states in a two dimensional electron gas , under conditions of the fractional quantum hall effect , were argued to be tl liquids as well@xcite .
it is well known that in the tl model with a repulsive electron
electron interaction the effective strength of a backward scattering by an impurity defect increases with decreasing temperature@xcite . for this reason
, the conductance of tl liquids with a single defect has been intensively discussed recently using various theoretical methods@xcite .
it has been concluded that even a weak impurity eventually causes the conductance to vanish at low temperatures .
the physical interpretation of the conductance vanishing is based on the assumption _ _ made by kane and fisher@xcite , that in the limit of low temperatures the behavior of a tl system with an impurity may be described as tunneling between two disconnected semi - infinite tl wires .
the effective amplitude of tunneling between the half
wires scales to zero with decreasing temperature , because the tunneling density of states at the ending point of a tl liquid vanishes when the electron
electron interaction is repulsive@xcite .
this interpretation corresponds to a scenario in which the effective strength of the impurity increases in the course of the renormalization , so that at the final stage a weak impurity transforms into a strong barrier , and disconnects the tl wire .
however , a direct calculation of the tunneling density of states@xcite , obtained by a mapping of the weak impurity problem onto a coulomb gas theory , apparently contradicts this intuitive picture .
it has been found that at the location of a weak impurity the tunneling density of states is enhanced , rather than vanishing .
the scenario of ref.@xcite is based on the assumption that no other fixed points intervene in the scaling from the repulsive fixed point of a weakly scattering defect to the attractive fixed point corresponding to a tunneling junction of two half - wires .
the contradiction of this scenario to the calculations of single particle properties , such as the tunneling density of states , and the fermi edge singularity@xcite , indicates that maybe this is not the case . in this work the problem of a single impurity in tl liquids with a repulsive electron
electron interaction is reinvestigated .
we concentrate on the limit when the fermi wave length is much larger than the defect size .
this situation is typical for semiconductors , where the filling of the conduction band is far from one half .
then it is possible to describe the problem as a continuous model with an appropriately chosen point like defect .
we come to the conclusion that the low energy physics of a weak impurity is controlled by a fixed point that differs from the one corresponding to a tunneling junction of two half - wires .
this aspect of the discussed problem proves to be similar to an overscreened two channel kondo problem , while the original idea of the theory of ref.@xcite has been orientated to a situation similar to the ordinary kondo problem ( see also ref.@xcite for a discussion ) .
the above statement may seem to be strange .
indeed , in the overscreened two channel kondo problem the infinitely strong exchange interaction corresponds to a repulsive fixed point@xcite , while in the problem under discussion the only apparent candidate for a fixed point in the strong coupling regime is the fixed point of a tunneling junction@xcite , which is attractive .
we find here that there exists another fixed point corresponding to the infinitely strong amplitude of the impurity backward scattering , and it is repulsive .
this implies that in addition there should be an attractive fixed point at a finite value of the backward scattering amplitude .
the renormalization group ( rg ) flow diagram of the problem depends on several parameters characterizing the impurity , and is rather complex .
each of the fixed points describing the backward scattering problem , as well as that of the tunneling junction , corresponds to a manifold of fixed points .
an important fact is , that the tunneling junction manifold , and the two of the new fixed points describing backward scattering , are located in different parts of the rg flow diagram . for a point like defect ,
all fourier components of an impurity potential are practically the same .
therefore , one may characterize the strength of the potential by an initial ( bare ) value of a dimensionless parameter @xmath0 where @xmath1 is the fermi velocity .
when @xmath2 is small , the backward scattering problem is described by the partition function @xmath3 , depending on a backward scattering amplitude @xmath4 , and on the electron electron interaction parameter @xmath5 .
alternatively , when @xmath2 is large , i.e. , the impurity is strong , the appropriate description is a tunneling model with a partition function @xmath6 , depending on a tunneling amplitude @xmath7 .
for free electrons , the two partition functions are deeply connected , because of the relation between the reflection from and the transmission through the defect . in this case , in principle , each one of them can represent an impurity of an arbitrary strength .
the situation becomes quite different when the electron electron interaction is switched on in tl liquids .
then , the partition functions @xmath3 and @xmath8 are not equivalent anymore .
instead , they describe two complementary limiting cases of the phase diagram , and give completely different structures of the rg fixed points for these cases . in this paper , the rg flow diagram of the continuous model is analyzed by examining different
corners of the scattering and the tunneling limits , and tailoring them together .
the flow diagram is constructed in axes representing the strength of the impurity @xmath2 , and the backward scattering amplitude @xmath9 ( and/or @xmath10 ) , respectively .
the presentation of the phase diagram in the space of the impurity potential parameters , rather than the conventional presentation in terms of the backward scattering amplitude together with the parameter @xmath5 , helps clarify the different character of the fixed points controlling the low energy physics in the cases of a weak and a strong bare impurity potential .
in the course of renormalization of a weak impurity the effective barrier evolves in a special manner . only the backward scattering amplitude @xmath9 increases , while @xmath2 , being marginal , remains small .
that is why a weak local impurity in a continuous model does not evolve towards a strong barrier , but proceeds to scale in the other , complementary , part of the rg flow diagram .
the low energy physics in this case is controlled by a line of fixed points @xmath11 , at a finite value of @xmath9 .
( more precisely , @xmath11 is a manifold of fixed points due to the presence of other marginal parameters .
since we use a two dimensional plot for the rg - diagram , we will call such manifolds lines . )
the other part of the flow diagram contains a line of attractive fixed points , @xmath12 , controlling the low energy physics when the bare potential of the impurity is strong .
this line describes the vanishing of the tunneling amplitude @xmath7 , and corresponds to two disconnected half - wires .
we think that the attraction of the weak impurity problem to @xmath11 , but not to @xmath12 , is the reason for the difference between the results obtained by means of the coulomb gas theory for the single particle correlators , and those obtained relying on the scenario of two disconnected wires .
the paper is organized as follows : in sec .
[ se : weakstrong ] we discuss the renormalization of the scattering and the tunneling hamiltonians , i.e. , the rg flow for two different limits of the impurity strength .
the existence of several parameters describing a local defect in the continuous model is emphasized . in sec .
[ se : rgphaseplane ] a rg flow diagram unifying both cases of a weak and of a strong impurity potential is discussed . in appendix [ app1 ] the repulsive character of the limit @xmath13 and , consequently , the existence of the attractive @xmath11-line , is demonstrated by mapping the impurity problem onto a spin-@xmath14 heisenberg chain .
in this section we reproduce the main results concerning the renormalization of the problem in the limits of a weak and of a strong impurity potential .
we assume that the fermi wave length is much larger than the defect size .
this situation occurs in semiconductor wires where the filling of the conduction band is far from one half and the fermi wave length is large .
several authors used the half - filled tight binding model with a link defect , to analyze the problem of backward scattering in quantum wires .
the tight binding model corresponds to a fixed choice of the parameters describing the defect , that makes this model a specific one . in this case
the fermi wave length is commensurate with the impurity size , and the final fixed point depends on the internal structure of the defect@xcite .
the continuous model considered here is not constrained by the specific structure of the local defect . * a weak potential scattering : the repulsive line * @xmath15 the hamiltonian of the tl model is given by @xmath16 here the electron spectrum is linearized near the fermi points @xmath17 ; @xmath18 are the field operators of fermions that propagate to the right with wave vectors @xmath19 , and @xmath20 are the field operators of left propagating fermions with wave vectors @xmath21 ; @xmath22 are the electron density operators ; @xmath23 describes the density
density interaction with a momentum transfer much smaller than @xmath24 .
the hamiltonian ( [ rgeq : ffh ] ) describes the 1d electron liquid when the backward scattering amplitude of the electron
electron interaction may be ignored , and it is a fixed point hamiltonian for a broad class of 1d systems . for low energy physics only processes of electron scattering with a momentum transfer close to zero and to
@xmath25 are essential .
let us denote @xmath26 here @xmath27 and @xmath28 are the fourier transform amplitudes of the impurity potential @xmath29 for a weak impurity the forward and backward scattering amplitudes are @xmath30 respectively ; @xmath1 is the fermi velocity .
( to avoid confusion , we will not use here the notations @xmath31 for the dimensionless amplitudes @xmath32 , unlike in @xcite , because the relations of the amplitudes @xmath32 with the scattering phase shifts are ill defined for interacting electrons in the tl model . ) note , that since the bare impurity potential @xmath33 is assumed to be local , it has fourier components of the same order for practically all momenta .
therefore , the line of the bare parameters corresponds to @xmath34 .
in addition to @xmath2 , @xmath9 , and @xmath35 , another parameter , @xmath36 , describing the asymmetry of the forward scattering of left and right movers will be introduced . in the presence of time reversal symmetry
@xmath37 , but it is not necessarily zero for the quantum hall edge states .
the scattering of the conduction band electrons by a single local defect at @xmath38 is given by @xmath39 , \end{array } \label{rgeq : fih}\ ] ] where @xmath40 .
the bosonization technique ( for review see @xcite ) allows one to reduce the tl hamiltonian to a quadratic form in terms of operators of bosonic fields @xmath41 and @xmath42 : @xmath43 , \label{rgeq : dpd } \\
\tilde{\phi}_{0}(x ) & = & \frac{-i}{\sqrt{4\pi } } \frac{2\pi } { l}\sum_{p}\frac{% \exp \left ( { -\eta |p|/2-ipx}\right ) } { p}\left [ \rho _ { r}(p)-\rho _ { l}(p)\right ] , \end{aligned}\ ] ] where @xmath44 is the system length , and @xmath45 is an ultraviolet cutoff , @xmath46 is equal to the average density of the electron liquid .
the fields @xmath41 and its dual partner @xmath47 are conjugate variables , i.e. , @xmath48 = i\delta \left ( x - y\right ) .
\label{rgeq : cr}\ ] ] after rescaling the operators @xmath49 the bosonized representation of the hamiltonian ( [ rgeq : ffh ] ) becomes @xmath50 here @xmath5 is an effective parameter of the electron
electron interaction @xmath51 @xmath52 when the electron
electron interaction is repulsive , while for the attractive interaction @xmath53 .
the bosonic representations of the operators @xmath54 and @xmath55 are given as @xmath56 , } \label{rgeq : fermion } \\
\psi _ { l}^{b}(x ) & = & i\frac{e^{-ik_{f}x}}{\sqrt{2\pi \eta } } \exp { \left [ -% \frac{i}{2}\left ( \frac{4\pi } { \beta } \tilde{\phi}-\beta \phi \right ) \right ] } .\end{aligned}\ ] ] then , the scattering by a weak impurity may be written in terms of the @xmath57-field as @xmath58 the rg equations for this problem are @xmath59 @xmath60 @xmath61 @xmath62 the last equation here reflects the well known fact that in the tl model the interaction parameter @xmath5 is not renormalized .
since @xmath5 describes a bulk electron liquid , the presence of a single impurity can not influence this parameter .
the decoupling of the rg - equations for amplitudes @xmath2 from the rest of the parameters is a consequence of an important property of the hamiltonian @xmath63 .
namely , it can be split into two commuting parts containing the forward @xmath2 , and the backward @xmath9 , scattering terms separately@xcite , see eq .
( [ eq : lah ] ) . since the forward scattering @xmath2 is given by a linear term in eq .
( [ rgeq : bih ] ) , and the amplitude @xmath9 can not influence the renormalization of @xmath2 because of the separation , the fulfillment of eq .
( [ eq : rgb ] ) becomes evident .
hence , under the condition that the impurity can be described as a local weak potential in a tl liquid , the forward scattering amplitude @xmath2 is a marginal parameter .
the parameters @xmath64 and @xmath65 are marginal as well , see eq .
( [ eq : rgc ] ) .
this is because the @xmath64-term , being linear in @xmath66 , can be removed from the hamiltonian by a canonical transformation , while @xmath35 does not reveal itself in the partition function @xmath67 .
in contrast , the backward scattering amplitude is relevant for the repulsive case .
( [ eq : rga ] ) has been derived @xcite ( see also @xcite ) for small @xmath9 .
the renormalization of @xmath9 in the strong coupling regime is discussed in sec .
[ se : rgphaseplane ] .
( [ eq : rga][eq : rgd ] ) describe a repulsive manifold of fixed points @xmath15 , denoted as the @xmath15-line at the left - bottom corner , @xmath68 , of the rg - plane depicted in fig .
[ fg : rgtl ] . * a strong potential barrier : the line of attractive fixed points * @xmath12 when the bare impurity is strong enough , the description of the problem in terms of the scattering amplitudes @xmath69 , and @xmath35 ceases to be adequate . on the other hand
, the fact that the impurity is strong does not contradict the assumption of locality @xmath70 .
a local and strong impurity can also be considered as a point like problem , but here it should be described by two semi infinite tl liquids , with a weak tunneling junction between their ending points . like in the case of the weak potential scattering , there are four parameters that describe the tunneling and reflection processes at the tunneling junction of the two half wires .
these parameters are the tunneling amplitude @xmath7 , its phase @xmath71 , and the two parameters , @xmath72 and @xmath73 , characterizing the phases that an electron acquires when it is reflected at the ends of the half wires .
the parameter @xmath73 describes the asymmetry of the left and the right parts of the tunneling junction . in the particular case of a strong @xmath74-function potential , where @xmath75 @xmath76 , the amplitudes @xmath77 @xmath78 .
the low energy physics of each semi infinite wire may be described by a single _ _ chiral mode@xcite .
it will be assumed that there is no density
density interaction between the half
wires , but inside each of them the density density interaction is present .
the effects of the electron
electron interaction inside a half wire can be taken into consideration by a canonical transformation ( see , e.g. , sec .
@xmath79 in ref.@xcite )
. then the appropriate bosonic description of tunneling between two oppositely moving chiral modes corresponding to each of the half - wires is given by @xmath80 here @xmath81 is related to the total density of the two chiral modes , and @xmath82 is a field dual to @xmath83 .
the rg equations in this case are analogous to eqs .
( [ eq : rga]-[eq : rgd ] ) : @xmath84 @xmath85 @xmath86 in contrast to @xmath9 , the tunneling amplitude @xmath7 scales to zero for repulsive electron electron interaction ( @xmath52 ) .
therefore , the manifold of fixed points described by the tunneling hamiltonian @xmath87 is attractive . in the two dimensional plot depicted in fig .
[ fg : rgtl ] it is presented in the upper right corner as the @xmath12 -line .
the problem of an impurity in a tl liquid has been discussed in sec .
[ se : weakstrong ] for the two limiting models : an impurity scattering , eq.([rgeq : bih ] ) , and a tunneling junction , eq.([eq : ht ] ) .
the scaling flow diagrams of these complementary models are depicted together in fig .
[ fg : rgtl ] as a combined plot . in this plot ,
the scattering model is represented on the left side , and the tunneling model on the right one .
the center of the plot corresponds to the crossover region when the impurity potential has an intermediate strength . since for a large barrier
the backscattering amplitude is large , and the tunneling amplitude is small , we use the vertical axis to represent @xmath9 and @xmath10 .
the horizontal axis represents @xmath2 together with the other parameters , which in a model with a linearized electron spectrum are not renormalized . due to the presence of these additional parameters , lines on the symbolic two - dimensional plot correspond to hyperplanes . for a point - like impurity
there are certain relations between the bare parameters characterizing the potential : @xmath88 , and @xmath89 .
hence , only the left - bottom and the right - upper corners of the phase diagram in fig .
[ fg : rgtl ] are attainable for bare parameters of a local impurity in the continuous model . however , according to eqs .
( [ eq : rga ] ) , ( [ eq : rgb ] ) the amplitude @xmath9 renormalizes to a strong coupling regime , while @xmath2 being marginal , remains small . therefore , to understand the nature of this regime , we have to analyze the scattering hamiltonian ( [ rgeq : bih ] ) for the region of the parameters corresponding to the upper left corner of the phase diagram .
( since large @xmath4 together with small @xmath2 does not correspond to a physical realization of the bare parameters , we use in the phase diagram `` @xmath9 '' with quotation marks . ) in appendix [ app1 ] , the line @xmath90 with @xmath91 and a not too large @xmath2 is argued to be a repulsive fixed line . for this purpose ,
the hamiltonian @xmath92 of the impurity scattering in the tl model is mapped onto a semi - infinite spin-@xmath14 heisenberg chain . in this mapping
the impurity backscattering amplitude corresponds to a magnetic field @xmath93 , that acts on the spin located at the origin of the chain .
in the course of the rg treatment small @xmath94 scales , as @xmath9 does , to a strong coupling regime . to investigate the nature of this regime
, we analyze the stability of the limit @xmath95 , following the spirit of the treatment of the two channel kondo problem by nozires and blandin@xcite .
when @xmath96 is very large , the spin
spin coupling in the spin chain can be considered as a small perturbation in a real space rg analysis .
this procedure generates a new operator , which is not irrelevant .
if one assumes that @xmath97 is a stable fixed point , the latter fact makes the rg process nonconvergent .
this is in contradiction to the assumption of stability , and therefore we have to conclude that @xmath97 is a repulsive fixed point .
since @xmath93 , we get that the limit of an infinitely strong backward scattering is repulsive .
both limits , @xmath98 and @xmath13 , are repulsive , and therefore there necessarily should be an attractive fixed point at some finite value of the backward scattering amplitude @xmath9 .
( this analysis is valid when a description of the local impurity scattering in terms of the tl model is a good approximation .
then the hamiltonian @xmath99 can be split into two commuting parts describing the backward and the forward scattering separately@xcite , see eq .
( [ eq : lah ] ) .
the discussed rg - regime of the @xmath9-amplitude develops inside the backward scattering part alone . ) _
_ thus , on the left side of the diagram there are two repulsive lines @xmath15 and @xmath100 , and in addition an attractive line @xmath11 , which is sandwiched between the opposite directions of the flow .
the right side of the diagram corresponding to the tunneling hamiltonian ( [ eq : ht ] ) , contains an attractive fixed line @xmath12 which describes scaling to zero of the tunneling amplitude @xmath7 .
the information collected up to now is presented on the combined plot in fig .
[ fg : rgtl ] . to complete the central part of the flow diagram , a region of an intermediate impurity strength should be studied .
none of the two limiting models describes the problem faithfully in this crossover region , and a consideration of a more comprehensive hamiltonian , which covers both limiting cases , is needed .
moreover , to study the rg - flow in the crossover region , one has to give up the approximations of the linearized electron spectrum and/or of the locality of the defect .
then , the decoupling of forward and backward scattering is no longer valid .
( in the bosonized representation terms @xmath101 describe the curvature of the electron spectrum .
one of the possible ways to consider the nonlocality of the impurity is to add a term @xmath102 .
when @xmath103 the coefficient of this term is very small . )
effects arising due to the nonlinearity of the electron spectrum and the nonlocality of the defect should be studied by a loop expansion in higher orders .
these effects may have highly important influence on the renormalization of the parameters @xmath2 and @xmath9 .
for example , it may cause the flow lines near the @xmath11-line to bend to the left or to the right .
the plot in fig .
[ fg : rgtl ] is based on the idealized models , and as a draft it gives a hint how the known limiting cases could be matched together . since the curvature of the electron spectrum is not universal , different scenarios can occur .
the two most apparent versions of the flow diagram are presented in figs .
[ fg : rgnl]@xmath104 and [ fg : rgnl]@xmath105 , but more sophisticated variants can be imagined due to multidimensionality of the problem , which up to now was hidden by the linearized spectrum approximation . in the version presented in fig .
[ fg : rgnl]@xmath104 , the limiting cases of a weak and a strong impurity evolve completely independently .
in contrast , the rg - flow presented in fig
. [ fg : rgnl]@xmath105 corresponds to a scaling from @xmath106to @xmath107 , i.e. , from a weak impurity scattering to a limit of two disconnected half - wires , as it has been assumed in ref.@xcite . however , this version of the rg - flow acquires , in the present discussion , a new essential element .
namely , the flow trajectory after the first stage where it reaches the @xmath108 -line dwells at length in its vicinity , and this leads to an intermediate asymptotic behavior . for a weak enough impurity this intermediate regime can be very long , and then it determines the low energy physics in a certain temperature range .
-5 cm 7.5 cm we emphasize that in considerations based on the approximation of a linearized electron spectrum , rg - trajectories that start at @xmath106end at @xmath11 .
this approximation has been utilized in the mapping of the problem onto a coulomb gas theory @xcite . therefore , the tunneling density of states , and the fermi edge singularity exponent , found in refs .
@xcite correspond to the physics near @xmath109 , and not to a tunneling junction , i.e. , not near @xmath12 .
we have studied a single impurity in tl liquids with a repulsive electron electron interaction .
the problem has been described by a continuous model with a point - like defect .
two complementary limiting models corresponding to the weak and the strong impurity potential limits have been presented on a unified flow diagram depicted in fig .
[ fg : rgtl ] .
apart from the backward scattering amplitude for a weak impurity , and the tunneling amplitude at the opposite extreme , another parameter characterizing the strength of the impurity potential controls the rg flow .
the unified plot of the flow diagram is rather complex . because of many marginal parameters , fixed points of the limiting problems generate manifolds which on the symbolic two - dimensional plot of the flow diagram we present as lines .
when the potential barrier of the impurity is strong , the low temperature behavior is controlled by the @xmath12-line corresponding to two disconnected semi infinite wires .
in addition , the flow diagram contains another attractive fixed line , @xmath11 , controlling the low temperature physics when the bare potential of the impurity is weak .
the existence of this line has been established by considering a scattering hamiltonian ( [ rgeq : bih ] ) with the backward scattering parameter taken to be very strong , @xmath110 , while the forward one , @xmath2 , was small .
this somewhat fictitious case is important due to the fact that in the tl model only the backward scattering scales to strong coupling while @xmath2 remains marginal . by mapping this case onto a spin-@xmath14 heisenberg chain
, it has been shown that the line @xmath90 describing the limit @xmath111 is unstable .
the existence of the attractive line @xmath109 at an intermediate value of the parameter @xmath9 , follows from the fact that both limiting lines @xmath90 and @xmath15 , have proved to be unstable .
the renormalization of the @xmath9-amplitudes resembles the situation of the overscreened two channel kondo problem @xcite .
this is not accidental . at a particular value of the electron
electron interaction parameter @xmath112 , the problem of a weak impurity in a tl liquid can be mapped onto the two channel kondo model with a specific value of the longitudinal exchange coupling ( for details see the end of appendix [ app1 ] ) .
it is well known that in the overscreened two channel kondo problem the limit of infinite exchange interaction is unstable , and there is an anomalous fixed point at a finite coupling .
since the problem under consideration , and the overscreened two channel kondo model are equivalent at one point , it is natural that the line @xmath13 proves to be repulsive , as it has been found here .
note , however , that the present treatment is not restricted to a specific value of the electron electron interaction .
the presentation of the phase diagram in the space of parameters characterizing the impurity potential , helps clarify the difference between the @xmath12- and the @xmath11-lines of fixed points
they are located in different parts of the phase diagram . in the continuous model with a linearized electron spectrum the two limiting cases of a point - like defect s
strength evolve completely independently .
this is constructive for understanding the low temperature physics in the case of a weak impurity . on the other hand , an extension beyond the assumption of the model such as the linearization of the spectrum ,
the locality of the defect , as well as other possible mechanisms of a crossover between different regimes , remain open questions
. a novel line of fixed points can be essential for the scenario of ref.@xcite .
this scenario is based on the assumption of scaling from a weak impurity scattering to a strong barrier .
the existence of the @xmath11 -line indicates that the situation is more complicated ( see the discussion of figs .
[ fg : rgnl]@xmath104 , [ fg : rgnl]@xmath105 at the end of sec .
[ se : rgphaseplane ] ) .
we think that the attraction of the weak impurity problem to @xmath11 , but not to @xmath12 , is the reason for the results obtained by means of the coulomb gas theory@xcite .
the enhancement of the tunneling density of states obtained in this theory corresponds to decrease of the escape rate of an electron from a defect center , as a result of multiple backward scattering in combination with the electron
electron interaction .
we believe that the physics of the @xmath11-line may have a relation to the strengthening of the role of the friedel oscillations in the tl model , see e.g. , ref .
@xcite .
to conclude , we have developed a continuous model of a single local defect in a tl liquid .
this theory is related to the physics of semiconducting @xmath113 quantum wires , and edge states in the quantum hall effect .
we identify a new attractive fixed point controlling the strong coupling regime of the backward scattering in the tl model .
this novel point may also have implications for some other related problems , in particular to the theory of the motion of a quantum particle in a dissipative environment , see ref.@xcite and references therein .
we thank y. gefen and d. e. khmelnitskii for useful discussions .
a. f. is supported by the barecha fund award .
the work is supported by the israel academy of science under the grant no .
801/94 - 1 and by german - israel foundation ( gif ) .
in this appendix the problem of a local impurity in the tl model is mapped onto a semi - infinite spin-@xmath14 heisenberg chain , with a magnetic field @xmath114 acting on the spin located at origin of the chain .
the mapping onto a spin chain is an appropriate way to study the nature of the strong coupling regime of the impurity problem .
this can be done by analyzing the stability of the point @xmath95 .
when @xmath115 , one can consider the spin
spin coupling in the spin chain as a small perturbation in a real space rg analysis .
this procedure generates a new operator , which is not irrelevant , and therefore the fixed point @xmath95 proves to be unstable . both limits , @xmath116 and @xmath95 , are repulsive , and consequently there should be an attractive fixed point at an intermediate finite value of @xmath96 . since @xmath93 , this analysis yields the existence of the repulsive @xmath90-line and the attractive @xmath11 -line on the left part of the rg flow diagram depicted in fig .
[ fg : rgtl ] . [ [ a - mapping - of - impurity - scattering - in - tl - liquids - onto - a - spin-12-semi - infinite - chain . ] ] a mapping of impurity scattering in tl liquids onto a spin-@xmath14 semi - infinite chain . ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ that obey the commutation relations @xmath118 & = & \left [ \theta _ { o}(x),\theta _ { o}(y)\right ] & = & -\frac{i}{4}\mbox { sgn}(x ) , \\ & & \left [ \theta _ { e}(x),\theta _ { o}(y)\right ] & = & -\frac{i}{4}. \end{array } \label{eq : thetacom}\ ] ] in terms of @xmath119 the hamiltonian @xmath92 , see eqs .
( [ rgeq : bfh ] ) and ( [ rgeq : bih ] ) , can be rewritten in the form @xmath120 @xmath121 , \label{eq : lah}\ ] ] @xmath122 .\ ] ] although @xmath123 and @xmath124 do not commute , the hamiltonian @xmath99 is divided into even and odd parts , @xmath125 and @xmath126 , respectively , because the even part contains only derivatives of @xmath123 . for small bare scattering amplitudes @xmath31 , and a not too strong electron
electron interaction , the linearization of the spectrum of the electrons used in the tl model is valid .
the decoupling of forward and backward scattering holds under this approximation . for simplicity
we omit the phase @xmath35 , and the @xmath64-term related to time reversal asymmetry , in @xmath126 .
we now show that the odd part @xmath126 is effectively equivalent to the hamiltonian of a semi - infinite spin-@xmath14 antiferromagnetic chain with anisotropy @xmath127 : @xmath128 a hamiltonian of this type , with @xmath129 , has been introduced by guinea @xcite for the description of a quantum particle interacting with a dissipative environment , at a particular value of the friction coefficient .
it has also been used to discuss the transmission through barriers in tl liquids@xcite , for a given value of the electron electron interaction @xmath112 . here
we introduce the @xmath130-term in order not to be limited to a particular value of the electron electron interaction .
note , that in @xmath131 , unlike the standard tight binding model ( see , e.g. , ref .
@xcite ) , the defect is located on a single site @xmath132 , and therefore it has no internal structure .
this is consistent with our purpose to study a point - like impurity . to show the equivalence of @xmath126 and @xmath131
, one should perform a sequence of transformations . after applying the inverse of the jordan - wigner transformation@xcite @xmath133 @xmath131
transforms into @xmath134 , where @xmath135 in the absence of the local magnetic field @xmath96 , the average of @xmath136 over the ground state @xmath137 of the spin chain is equal to zero . since @xmath138 the sector @xmath139 corresponds to half filling . the local operator @xmath140 can not change the bulk properties of the chain .
therefore , it will be assumed that the fermion system @xmath134 of eq.([eq : spinfermion ] ) is at half filling .
the continuum limit of @xmath134 ( e.g. , see ref .
@xcite ) corresponds to the effective hamiltonian @xmath141 , where @xmath142 , \\ h_{h } & = & v_{f}\frac{1}{\sqrt{\eta } } h\left ( r(x=0)+l(x=0)+r^{\dagger } ( x=0)+l^{\dagger } ( x=0)\right ) . \end{array } \label{eq : rl}\ ] ] here @xmath143 , where @xmath144 is the lattice spacing , and the operators @xmath44 and @xmath145 represent left and right movers on a semi - infinite line .
the remnant of the discrete structure of the chain is the last term in @xmath146 which corresponds to the umklapp processes at half filling@xcite . this umklapp
term scales to zero for @xmath147 , and also renormalizes the electron
electron interaction with a small momentum transfer .
however , at small @xmath148 the latter effect is negligible , and for the low energy description one may substitute @xmath149 by the effective term @xmath150 the last step , which needs to be carried out in order to establish the equivalence between @xmath126 and @xmath131 , is to unfold the semi - infinite line with left and right movers into a full line with a single chiral bosonic field .
this yields : @xmath151 \\ & & -\frac{4v_{f}}{\sqrt{2\pi } \eta } h\cos \left ( \sqrt{4\pi } \theta _ { ch}(x=0)\right ) , \end{array } \label{eq : chirall}\ ] ] where @xmath152 diagonalizing the quadratic part of the hamiltonian , we find : @xmath153 , \label{eq : chirallbar}\ ] ] where @xmath154 @xmath155 , @xmath156arctanh@xmath157 , and @xmath158 .
notice , the important role of the @xmath127-term it modifies @xmath159 inside the cosine term .
thus , as a result of the sequence of transformations @xmath160 we obtain that the hamiltonians @xmath131 and @xmath126 are equivalent when @xmath161 and@xmath162 . due to the @xmath163-term , this equivalence is extended here to a finite interval of the electron
electron interaction .
we will use the equivalence of @xmath126 and @xmath131 to analyze the stability of the fixed line at @xmath13 and small @xmath2 .
a simple variant of the nozires and blandin approach@xcite in their analysis of the two channel kondo problem will be considered . following this approach
, we will assume that @xmath164 and check whether the fixed point @xmath165 is a stable one . in the presence of a strong magnetic field @xmath115
, the spin at site @xmath166 is oriented along the direction opposite to the magnetic field .
its coupling to the nearest neighbor at the lattice site @xmath167 can be treated as a perturbation .
the reduced hamiltonian that includes only sites @xmath166 and @xmath168 is given by @xmath169 after performing the permutation @xmath170 , @xmath171 , and @xmath172 we obtain the hamiltonian @xmath173 + \frac{j}{4}\left ( 1+% \bar{\gamma}\right ) \left [ s_{0}^{-}s_{1}^{+}+s_{0}^{+}s_{1}^{-}\right ] -2hjs_{0}^{z}. \label{eq : jper2}\ ] ] for @xmath115 the spin at site @xmath166 is in the state @xmath174 . up to the first order in @xmath175 we have
@xmath176 this means that the spin at site @xmath168 is under the action of an effective magnetic field @xmath177 . under the assumption that @xmath115 , the higher orders in the perturbation theory give small corrections of the order of @xmath178 . as a result of this renormalization procedure step ,
we arrive at a problem equivalent to the initial one : a semi - infinite spin-@xmath179 heisenberg chain , with a local magnetic field acting on the site at origin of the chain ( now it will be site @xmath168 ) .
if one assumes that the discussed fixed point is such that the local magnetic field at the origin of the spin chain flows to infinity , then the renormalization procedure generates a relevant operator that makes this process nonconvergent .
this is in contradiction with the initial assumption that @xmath95 is a stable fixed point .
we have to conclude that @xmath180 is a repulsive fixed point .
note , that this conclusion holds for a finite interval of the parameter @xmath181 , because @xmath182 does not radically influence the effective magnetic field acting on site @xmath168 .
one can see this from the second order corrections to the energy of the anti - parallel and parallel spin configurations , that are @xmath183 .
it is not accidental that the present discussion resembles the analysis of the overscreened two channel kondo problem@xcite .
indeed , the spin chain model , in the absence of the @xmath127-term , is equivalent to the tl - impurity problem at a particular value of an electron
electron interaction parameter @xmath112 @xcite .
the latter problem can in turn be reduced to a resonant level model@xcite . moreover , the overscreened two channel kondo model at a specific value of the longitudinal exchange coupling is equivalent to the same resonant level model
. thus , the spin chain model and the kondo model are equivalent at one point . on the other hand
, it is well known that in the overscreened two channel kondo problem the limit of infinite exchange interaction is unstable , and there is an anomalous fixed point at a finite coupling@xcite .
this property is preserved in the presence of the spin exchange anisotropy , which is irrelevant@xcite .
since the spin chain model and the two channel kondo model are equivalent at one point , it is natural that we have found that the point @xmath95 is repulsive . since @xmath93
, it follows from this analysis that the line @xmath184 is a repulsive fixed line for the problem of impurity scattering in a tl liquid .
the present treatment is not restricted to the special point of a tl liquid with @xmath112 .
this has been accomplished by the @xmath185 term in the spin chain model . | the problem of a single impurity in one dimensional tomonaga luttinger liquids with a repulsive electron
electron interaction is discussed .
we find that the renormalization group flow diagram for the parameters characterizing the impurity is rather complex .
apart from the fixed points corresponding to two weakly connected semi infinite wires , the flow diagram contains additional fixed points which control the low temperature physics when the bare potential of the impurity is not strong .
specialpagefalse evenheadoddhead |
recent observations of the corona with the lasco coronagraph @xcite on board of the soho spacecraft showed that the corona can be highly structured even during the solar activity minimum .
the observations revealed a triple structure of the streamer belt which was existent for several consecutive days .
the observations further showed that these triple structures occasionally go unstable leading to a seemingly new and extraordinarily huge kind of coronal mass ejection ( global cmes ) .
natural questions arising from these observations are whether the helmet streamer triple structure is directly connected with or responsible for the occurrence of global cmes and what is the physical mechanism of their formation .
the aim of this paper is to provide a first step towards a better theoretical understanding of these phenomena . the structure of helmet streamers and their stability has been studied both observationally and theoretically for a long time ( e.g. @xcite , @xcite ; @xcite , @xcite ; @xcite , @xcite ; @xcite , @xcite ; @xcite , @xcite ; @xcite , @xcite ; @xcite , @xcite ; @xcite , @xcite ; @xcite , @xcite ) .
a natural association seems to exist between helmet streamer stability and coronal mass ejections ( cmes ) and sometimes cmes are accompanied by erupting prominences or filaments .
coronal streamers are also thought to be the source regions of the slow solar wind and the activity processes discussed above may well contribute to this component of the solar wind .
multiple streamer structures have already been observed before soho , however mainly during the maximum phase of the solar activity cycle .
also , observations of the heliospheric plasma sheet seem to indicate a multiple current sheet substructure of the plasma sheet itself @xcite which has initiated several theoretical investigations of the stability of multiple current sheets @xcite . the aim of this work is to present a simple method by which triple helmet streamer structures can be calculated self - consistently .
we want to keep the method simple enough so that we can obtain largely analytical results without rendering our models too unrealistic .
since the observed structures seem to be quite elongated ( the radial magnetic field component is considerably larger than the longitudinal magnetic field component ) , we can employ a method which has already been applied very successfully to the equilibrium structure of the magnetotail of the earth @xcite .
the method is based on an asymptotic expansion .
here we generalize this method and show how it can be applied to find solutions with internal multiple structures .
obvious generalizations of the original magnetospheric versions to solar applications are the use of a more suitable coordinate system ( spherical rather than cartesian ) , the inclusion of magnetic shear , of the solar gravity field for large scale structures and of plasma flow on open field lines . for reasons of mathematical simplicity and
since we only deal with the subcritical part of the flow , we confine the discussion to incompressible flow .
our models provide a flexible quantitative description of structured helmet streamers .
they can also be used as the starting configurations for future studies of the dynamical evolution of such structures .
the outline of the paper is as follows . in sect .
[ sec : method ] we present the mathematical basis of the method .
section [ sec : results ] contains representative results obtained by applying the method under various sets of assumptions . in sect .
[ sec : discussion ] we summarize the paper , give a discussion of our results and an outlook on future work .
we use the equations of stationary ideal magnetohydrodynamics to describe the coronal plasma : @xmath0 here , @xmath1 stands for the plasma pressure , @xmath2 for the magnetic field , @xmath3 for the plasma density , @xmath4 for the solar gravitational potential , @xmath5 for the plasma velocity , @xmath6 for the electric field , @xmath7 is the gas constant , @xmath8 the temperature and @xmath9 the vacuum permeability . due to the high conductivity of the coronal plasma the magnetic field is frozen into it ( eq .
[ ohm ] ) and we can make the assumption that the observed plasma features outline the magnetic field structure . throughout the rest of the paper
we make the following assumptions : * the streamer structure is very extended in azimuth and varies only very slowly along this direction ( this assumption is supported by the observations @xcite which showed that the triple structure existed for several days ) ; we can then describe the structure as approximately two - dimensional in the sense of rotational invariance . *
the streamers are elongated in the radial direction and the radial magnetic field component is considerably stronger than the latitudinal component ; this is the basic assumption allowing to apply an asymptotic expansion procedure . *
if flow is included , it is purely field - aligned and confined to the open field line region ; since we describe only the low lying parts of the streamer structure the flow velocities should be subcritical and we will therefore only consider incompressible flow . * for simplicity , we consider only isothermal equilibria ; in principle , other temperature structures or even the inclusion of an energy equation into the scheme is possible . in the following ,
we normalize the magnetic field by a typical value @xmath10 , the plasma pressure @xmath11 by @xmath12 ( where @xmath13 is the maximum value of @xmath11 ) , the mass density @xmath3 by @xmath14 , the length @xmath15 by a solar radius and the current density by @xmath16 . to demonstrate the method we first use a two - dimensional cartesian geometry with no spatial variation in the @xmath17-direction ( @xmath18 ) and assume static structures ( @xmath19 ) without magnetic shear ( @xmath20 ) . in other geometries the basic properties of the method stay similar though the details may vary due to the geometry .
writing the magnetic field as @xmath21 we find from equation ( [ stat ] ) that the flux function @xmath22 has to obey the equation ( e.g. @xcite @xcite ; @xcite @xcite ; @xcite @xcite ; @xcite @xcite ) @xmath23 with the assumption of a constant temperature and because @xmath24 in cartesian geometry we get : @xmath25 with @xmath26 .
for slow variation of @xmath27 with height @xmath28 one can neglect gravity and finds @xmath29 .
if we assume that the magnetic field perpendicular to the photosphere ( @xmath30 in the cartesian geometry ) is much larger than the parallel component ( @xmath31 in the cartesian geometry ) , or in other words , that streamers are rather elongated in the radial direction , we can use the method of asymptotic expansion and apply it to eq .
( [ gsg1 ] ) .
mathematically stretched configurations are characterized by the ordering @xmath32 if we neglect terms of the order @xmath33 in eq .
( [ gsg1 ] ) we obtain after one integration with respect to @xmath34 : @xmath35 where @xmath36 is an integration constant ( total pressure on the @xmath28-axis ) , which depends parametrically on @xmath28 .
we solve this differential equation by separation @xcite : @xmath37 and fix the constant @xmath38 so that @xmath39 .
unfortunately it is impossible to derive the pressure function @xmath40 directly from the observations presently available .
we will therefore choose @xmath40 by a compromise between physical reasoning and mathematical simplicity .
we do not choose pressure functions which have a too simple dependence on @xmath22 like @xmath41 , @xmath22 or @xmath42 because solutions of this type are a ) known to be linearly stable and b ) have already been discussed in @xcite .
a popular and convenient choice is @xmath43 because it corresponds to the physical assumption that the plasma is in local thermodynamic equilibrium and it has the mathematical advantage that one can find analytical solutions . using @xmath44 where @xmath45 may be a function of @xmath28 and @xmath46 may be positive or negative , we get from equation ( [ gl8 ] ) : @xmath47 \right\ } + \frac{1}{c}\log\left(\frac{p_0}{k}\right ) \label{axz}\ ] ] the components of the magnetic field are then given by @xmath48 @xmath49 we remark that one can not evaluate @xmath50 and @xmath51 in equation ( [ bx ] ) directly because in general @xmath52 depends on @xmath53 and @xmath54 as well
. the plasma pressure can be obtained from the eq .
( [ druck ] ) @xmath55 and the current density @xmath56 from @xmath57 : @xmath58 the method just outlined has been applied very successfully to calculate the structure of the earth s
. however , since we want to apply this method to describe multiple streamer structures a few details of the method have to be changed though the basic approach stays the same . in an arcade structure
the direction of the @xmath30 component of the magnetic field changes its sign as one crosses the centre of the arcade . in a triple structure , one will therefore encounter three such changes as one moves across the structure . in a stretched configuration described by the asymptotic expansion procedure , the only possibility to achieve
this is that the current density @xmath56 changes its sign between the streamers , or , in other words , in the central streamer the current flows in the opposite direction with respect to the two outer streamers . in terms of the sign of @xmath30
this means that in order to paste the three arcades continuously together , the @xmath30 component in the central streamer has to change just in the opposite sense to the outer two streamers . since the current density @xmath56
is directly linked to the plasma pressure by the relation @xmath59 , this implies that the derivative of the pressure with respect to the flux function @xmath22 has to change its sign , too .
we can therefore not use a single exponential function to model the pressure because the exponential function and its derivative have no zero .
the simplest alternative is that we use different pressure functions for the middle ( @xmath60 say ) and for the outer streamers ( @xmath61 ) . in principle
, we could also use a different pressure function for each of the three streamers , but since we do not want to make the model too complicated we assume the structure to be mirror symmetric with respect to the @xmath28-axis for simplicity . due to the different pressure functions our solution will have two separatrix field lines characterized by @xmath62 which separate the regions of different pressure from each other .
these separatrices have to be calculated self - consistently together with the solution but due to our symmetry assumption we only have to calculate one of those separatrix field lines . in the asymptotic expansion procedure
the location of the separatrix field line will be given by @xmath63 where @xmath64 is only varying weakly with @xmath28 .
if we require a smooth transition of magnetic field across the separatrix from the outer to the inner streamer , the pressure functions must be continuous .
therefore we choose @xmath65 for the middle streamer and @xmath66 for the outer streamers , while the coefficients @xmath67 , @xmath68 , @xmath69 and @xmath70 ( @xmath71 ) are related by @xmath72 on the separatrix labeled by @xmath62 .
this choice has the consequence that the current density @xmath56 has a discontinuity at the separatrix .
we remark , however , that the absolute value of @xmath56 is only rather small at the separatrix and only changes from a small negative to a small positive value and therefore the discontinuity of the current density should not influence the solution structure very much .
we choose @xmath73 to put the centre of the middle streamer on the @xmath28-axis . to calculate the separatrix between the middle and one of the outer streamer we use equation ( [ gl8 ] ) with the form of @xmath40 inside the middle streamer to obtain @xmath74 note that both @xmath75 and @xmath67 may depend weakly on @xmath28 .
if we call the centre of the outer streamer @xmath76 , we can calculate this central field line relative to the separatrix field line again with equation ( [ gl8 ] ) , but this time using @xmath61 and setting @xmath77 : @xmath78 if one takes into consideration that @xmath79 and equation ( [ sep1 ] ) one finds : @xmath80 a sketch of the whole structure is shown in figure [ skizze ] .
we can now calculate @xmath22 , @xmath30 , @xmath31 , @xmath11 and @xmath56 as functions of @xmath34 and @xmath28 from the equations ( [ axz ] ) , ( [ bz ] ) , ( [ bx ] ) , ( [ pxz ] ) and ( [ jxz ] ) and by using the symmetry properties of these quantities . as an input we need the total pressure @xmath53 which is the same in all three streamers , the parameters @xmath69 , @xmath70 and @xmath67 and @xmath68 .
if gravitation is included , @xmath67 and @xmath68 may depend weakly on @xmath28 , but could be different if we allow different temperatures inside the streamers .
the method as outlined so far neglects that coronal streamers are the source regions of the slow solar wind .
it is an open question if a stationary slow solar wind component exists or whether the slow solar wind is the result of many small eruptions .
recent observations ( schwenn and inhester , private communication 1997 ) seem to favour the second alternative . if that is true , a steady state model providing the start configuration for a time - dependent computation of the eruption process need not include the flow , at least not near the symmetry plane .
nevertheless we want to show how a stationary plasma flow can be included in our model .
it may be relevant for the open field regions at larger distances from the ecliptic , to which we confine the flow .
the inclusion of plasma flow into the asymptotic expansion procedure has been carried out for both incompressible @xcite and compressible flow @xcite .
there are three critical points in the solar wind : the slow magnetosonic point @xmath81 , the alfvn point @xmath82 and the fast magnetosonic point @xmath83 with @xmath84 for the coronal plasma .
if one assumes that the flow velocity at the basis of the corona is less than @xmath81 the solar wind has to pass through all these critical points .
@xcite showed that in stretched configurations the total pressure function @xmath53 is related to the flow velocity .
for @xmath85 the flow velocity decreases with decreasing @xmath53 and for @xmath86 the flow velocity increases with decreasing @xmath53 .
as it is necessary that @xmath53 decreases with respect to @xmath28 to get closed arcade structures , the solar wind velocity should have a minimum at the helmet streamer cusp .
this result is of course only valid if a stationary slow solar wind exists .
the assumption of incompressible flow in the solar wind , however , can not be justified in a strict sense , because generally gravity is not negligible . since in the paper our main emphasis is placed on closed structures without directed flow , this model assumption does not influence our conclusions significantly .
( in fact , all our explicit examples are structures without flow . )
we note that the above mentioned variation of the solar wind velocity remains qualitative valid , if one neglects gravity and assumes incompressible flow .
we will assume that the flow is purely field - aligned and since we will not extend our solution to the first critical point , we will for simplicity make the assumption of the flow being incompressible and subalfvenic .
we mention that , as a useful nice property , such stationary solutions with flow , can be directly calculated by transformation from static solutions @xcite . here
we follow the approach developed in @xcite .
equation ( [ gsg2 ] ) is then substituted by : @xmath87 where @xmath88 is the alfvn mach number defined by @xmath89 , where @xmath90 is the alfvn velocity .
note that for incompressible flow both the alfvn mach number and the plasma density are constant along open field lines . to extend the method outlined so far to include flow on open field lines
, we have to define a second separatrix field line @xmath91 between the regions of open and closed magnetic flux . since there shall be no flow in the closed field line regions , we choose @xmath92 on the left hand side of the separatrix and @xmath93 ( subalfvnic flow ) on the right hand of the separatrix .
in the special case @xmath94 we find with equation ( [ gsg3 ] ) that one can easily include incompressible subalfvenic plasma flow ( @xmath95 ) , if we replace the term @xmath96 in the static theory by @xmath97 .
the location of the separatrix field line @xmath91 could be calculated by a similar procedure as the inner separatrix , but it is more convenient to choose it to be at the same distance from the centre of the outer streamer as the inner separatrix so that we have a symmetric situation again and one finds @xmath98 .
we remark that this condition is necessary if one investigates configurations with a cusp structure .
one can then calculate the shape of the separatrix defined by @xmath99 simply from @xmath100 ( see figure [ skizze ] for illustration ) .
it is possible to extend our method and include magnetic shear writing @xmath101 as : @xmath102 using the @xmath17 component of equation ( [ stat ] ) one finds @xmath103 and instead of ( [ gsg1 ] ) we get : @xmath104 with @xmath105 using the method of asymptotic expansion we get : @xmath106 we remark that the reason for magnetic shear should be a displacement of magnetic foot points and that only closed field lines will be sheared .
we assume @xmath107 for mathematical simplicity .
in general we can not solve equation ( [ scher2 ] ) analytically .
this is only possible if we choose the special case @xmath108 which we will do to illustrate the effect of magnetic shear .
then we get : @xmath109 and we can easily include magnetic shear by substituting @xmath54 by @xmath110 in equation ( [ axz ] ) .
it is possible to investigate other values ( @xmath111 ) numerically , but one can see the main effect of magnetic shear already in this special case .
we calculate triple structures analogous to the method without shear and substitute @xmath112 by @xmath113 .
we have to consider shear only on closed field lines .
we use @xmath114 in the middle streamer , @xmath115 in the outer streamers and @xmath116 on open field lines .
thus there is a jump in @xmath117 and a corresponding jump in the magnetic field at the boundary between outer streamers and open field lines and consequently there is a thin current sheet . unlike the calculations on closed field lines one can not fix @xmath118 .
instead of ( [ gl8 ] ) one has to use : @xmath119 where @xmath120 is the value of the last closed field line ( boundary field line between open and closed regions ) .
one finds @xmath98 for configurations with cusp structure .
we remark that violating the condition @xmath121 could still make sense .
for example , one could use @xmath122 but @xmath123 .
of course this mismatch in the value of @xmath117 would have to be compensated by a corresponding mismatch in @xmath30 and this would lead to additional current sheets between the middle and the outer streamers . in this subsection
we formulate the method in spherical coordinates @xmath124 , which are more realistic to describe the solar corona on scales comparable to or larger than a solar radius .
we assume static @xmath125 configurations and do not consider spatial variation in the @xmath126-direction @xmath127 .
thus we can present the magnetic field as : @xmath128 here @xmath129 is the flux function in spherical coordinates .
we remark that in spherical coordinates the flux function is not identical with a component of the magnetic vector potential .
one finds @xmath130 , where @xmath131 is the @xmath126-component of the vector potential @xmath132 .
we find from ( [ stat ] ) that @xmath133 has to obey the equation : @xmath134 where @xmath135 is constant on field lines . with the assumption of a constant temperature and because @xmath136 one finds : @xmath137 with @xmath138 .
we assume that the magnetic field perpendicular to the photosphere @xmath139 is much larger than the parallel component @xmath140 .
for simplicity we do not consider magnetic shear here @xmath141 .
stretched configurations are characterized by @xmath142 this approximation is valid for configurations close to the equatorial plane ( @xmath143 ) and @xmath144 has to vary only slowly with respect to @xmath145 .
the second condition limits the outer boundary of our model corona to a few solar radii . as
the observed configurations are indeed concentrated near the equatorial plane and the closed streamer field lines seem to be radially bounded to a few ( about @xmath146 ) solar radii , our asymptotic model should describe these configurations approximately correct . if we neglect terms of the order @xmath33 in eq .
( [ gradkugel ] ) we obtain after one integration with respect to @xmath147 : @xmath148 we use ( analogous to our calculations in cartesian geometry ) @xmath149 and solve this differential equation by separation . instead of ( [ axz ] )
we get : @xmath150 we remark that one can not find a one dimensional solution of equation ( [ gradkugel ] ) with @xmath151 like in cartesian geometry .
thus no analogy to the harris sheet in cartesian geometry exists in spherical coordinates . to calculate triple structures in spherical geometry
one uses a similar procedure as described for cartesian geometry .
we choose @xmath152 to put the centre of the middle streamer and calculate the separatrix field line analogous to ( [ sep1 ] ) and get : @xmath153 if we call the center of the outer streamer @xmath154 we get analogous to our calculations in cartesian geometry : @xmath155 the quantities @xmath129 , @xmath139 , @xmath140 , @xmath11 , @xmath156 can then be calculated as functions of @xmath145 and @xmath147 in the usual way .
.parameter sets used to calculate the solutions presented in figure [ fig1 ] [ cols="<,<,<,<,<,<,<",options="header " , ] [ table2 ] in a next step we use our model to describe helmet streamer configurations with a cusp . up to now
we have not given a radial boundary of the closed magnetic field lines .
now we define such a point at @xmath157 , @xmath158 which defines the last closed field line .
we use as a function for the total pressure : @xmath159 field line plots for three different sets of parameters are shown in figure [ figcusp ] .
the parameter sets used are listed in table [ table2 ] .
it can be seen that the effect of changing the parameters @xmath69 and @xmath70 stays the same as in the case without cusp .
an increase of the exponent @xmath160 leads to a stronger decrease of @xmath53 as the cusp point is approached from below and this leads to a sharper cusp structure .
these examples illustrate the flexibility of our model .
one can describe different types of triple helmet streamers by specifying the free parameters in the general solutions . in this section
we investigate the influence of the solar gravitational field on the structure of the solutions .
though we are still calculating in cartesian geometry we will not take the gravitational force to be constant as it is usually done in cartesian geometry if the length scale under consideration is much smaller than a solar radius ( e.g. @xcite @xcite ; @xcite @xcite ) , but use the more appropriate potential @xmath161 .
together with the assumption of constant temperature we get : @xmath162 where @xmath163 is the constant of gravitation , @xmath164 the solar mass , @xmath165 the solar radius and @xmath166 . under the assumption of a pure hydrogen plasma and a constant temperature of @xmath167 k
one finds @xmath168 corresponding to @xmath169 .
one difficulty of the method in cartesian geometry is that @xmath54 must decrease more slowly with respect to @xmath28 than @xmath53 because otherwise one can not find arcade type solutions .
this is difficult to achieve with a realistic value of @xmath170 , but since we only want to investigate the qualitative effect of gravitation on the solution we use the unrealistic but mathematically more convenient value @xmath171 .
we remark that this problem is alleviated in spherical geometry which is more realistic anyway ( see section [ seckugel ] ) the inclusion of shear only makes sense on closed field lines .
as we have a well defined boundary between open and closed regions only in solutions with cusp structure , we only investigate such configurations here . in figure [ figscher ]
we fix the parameters @xmath172 and use for the total pressure : @xmath173 & z \leq z_{\mbox{\tiny cusp } } \\
b\left(k(z)+\frac{\lambda^2}{2 } \right ) & z > z_{\mbox{\tiny cusp } } \end{array } \right.\ ] ] the factor @xmath174 can be used to fix the normal component of the magnetic field at the solar surface , which we could use as a boundary condition .
the main effect of magnetic shear is that there is a jump in the magnetic field and consequently a thin current sheet between the closed and open field lines .
indications of such a current sheet are also found in recent observations @xcite . in figure [ figbz]a ) we plot @xmath30 as a function of @xmath34 at the height @xmath175 for the same parameters as in figures [ figscher]a ) and c ) . if we do not include shear the structure above the helmet streamer cusp @xmath176 becomes equivalent to a harris sheet and @xmath30 goes smoothly through zero in the center of the configuration .
the inclusion of shear on closed field lines leads to a jump in @xmath30 from a positive to a negative value in the center of the whole configuration ( see figure [ figbz ] b ) ) .
this causes a thin current sheet , which is more pronounced than the current sheet at the outer boundary of the streamer configuration below the cusp .
this current sheet corresponds to the heliospheric current sheet .
the effect can clearly be seen in figure [ figscher ] . in the figures [ figscher]b ) and
c ) we included magnetic shear and in c ) the shear is stronger than in b ) leading to a larger jump in @xmath30 .
this can be seen in the contourplot of @xmath22 as the field line density is much higher in the center above the cusp in figure [ figscher]c ) than in b ) . in the lower panel ( figure [ figscher]d)-f ) )
we investigated the same configurations as in the upper pictures but included gravitation .
one can see that without shear ( figure [ figscher]d ) ) , the configuration gets wider with increasing @xmath28 .
( if one uses the realistic value @xmath177 instead of @xmath171 the configuration will become unrealistically wide without shear . )
the inclusion of shear reduces this effect and if one investigates force free configurations @xmath178 or nearly force free solutions @xmath179 which is accomplished in c ) and
f ) the effect of gravitation vanishes .
now we present some example solutions of helmet streamer configurations in spherical coordinates .
this geometry is of course more realistic to describe the coronal magnetic field on large scales . in this section
we also include the influence of the solar gravitational field on the structure of the solutions from the beginning . with @xmath180
we get as in cartesian geometry ( see section [ shear ] ) @xmath181 here we use the realistic value @xmath177 . in figure [ figkugel]a ) we show a field line plot of a solution in which we prescribed the total pressure as @xmath182 $ ] and used the same parameters as in figure [ fig1]a ) ( see table [ table1 ] , case a ) . thus @xmath183 varies in the same way as @xmath53 in the cartesian case without gravitation .
the effects of the different geometry show up as a bending of the out streamers towards the equatorial plane . in figure [ figkugel]b )
( see table [ table2 ] , case b ) we present a solution with cusp structure in spherical geometry with the total pressure @xmath184 & r \leq r_{\mbox{\tiny cusp } } \\
k(r ) b & r > r_{\mbox{\tiny cusp } } \end{array } \right.\ ] ] and use the same parameters as in figure [ figcusp]c ) . again
the bending of the outer streamers is obvious .
we remark that prescribing the parameters @xmath185 and @xmath186 is equivalent to prescribing the normal component of the magnetic field at the solar surface .
this property of the solutions could be used to match measured magnetic fields and generate the corresponding streamer structure .
we have developed a comparatively simple procedure based on the method of asymptotic expansion to calculate equilibria of triple helmet streamer structures .
the procedure represents a generalization of the asymptotic expansion method used in the theory of magnetotail equilibria @xcite .
we have applied this procedure to the calculation of stationary states of triple helmet streamers under various conditions and investigated the dependence of the solution properties on the model parameters , the influence of the solar gravitational field , the inclusion of field aligned flow and on the basic geometry used . as a next step
we plan to use the calculated configurations as start equilibria for mhd simulations .
these simulations will answer the question of stability which we did not address in this paper and provide important insights into possible mechanisms of eruptive phenomena in the solar atmosphere , especially those connected with a triple structure of the solar corona .
the simulations should have relevance both for huge coronal mass ejections and very small but frequent eruptions which may represent the source mechanism of the slow solar wind .
the authors thank alan hood , bernd inhester , eric priest and rainer schwenn for discussions and useful comments .
we also thank the referee , r.a .
kopp , for his useful remarks .
we acknowledge financial support by the dfg graduiertenkolleg `` hochtemperaturplasmaphysik '' ( tw ) , by pparc ( tn ) and by a british - german academic research collaboration grant . here
we calculate the terms @xmath50 and @xmath51 which where not explicitly given in equation ( [ bx ] ) .
bavassano , b. , woo , r. and bruno , r. : 1997 _ geophys .
* 24 * , 1655 .
birk , g.t . and
otto , a. : 1997 , _ adv . space res .
_ * 19 * , 1879 .
birk , g.t .
, konz , c. and otto , a. : 1997 , _ phys .
plasmas , in press _ birn , j. : 1991 , _ phys .
fluids b _ * 3 * , 479 .
birn , j. : 1992 , _ j. geophys .
res . _ * 97 * , 16817 .
birn , j. and schindler , k. : 1981 , in priest , e.r .
( ed . ) _ solar flare magnetohydrodynamics _ ; gordon and breach , new york , 337 .
birn , j. , goldstein , h. and schindler , k. : 1978 , _ solar phys . _ * 57 * , 81 .
birn , j. , sommer , r. and schindler , k. : 1975 , _ astrophys .
space sci . _ * 35 * , 389 .
crooker , n.u.,siscoe , g.l . ,
shodan , s. , webb , d.f . , gosling , j.t . and smith , e.j .
: 1993 , _ j. geophys .
_ * 98 * , 9371 .
cuperman , s. , bruma , c. , dryer , m. and semel , m. : 1995 , _ astron .
astrophys . _ * 299 * , 389 .
cuperman , s. , detman , t.r . ,
bruma , c. and dryer , m. : 1992 , _ astron .
astrophys . _ * 265 * , 785 .
cuperman , s. , ofman , l. and dryer , m. : 1990 , _ astrophys .
j. _ * 350 * , 846 .
dahlburg , r.b . and karpen , j.t .
: 1995 , _ j. geophys
. res . _ * 100 * , 23489 .
gebhardt , u. : 1989 , diplom thesis , ruhr - universitt bochum .
gebhardt , u. and kiessling , m.k - h . : 1992 , _ phys .
fluids b _ * 4 * , 1689 .
hundhausen , a.j . : 1995 , in k. strong , j. saba and b , haisch ( eds . ) , _ the many faces of the sun _ , in press .
koutchmy , s. and livshits , m. : 1992 , _ space sci .
rev . _ * 61 * , 393 .
low , b.c . : 1975 , _ astrophys .
j. _ * 197 * , 251 .
low , b.c .
: 1980 , _ solar phys . _ * 65 * , 147 .
otto , a. and birk , g.t.:1991 , _ phys .
fluids b _ * 4 * , 3811 .
platt , u. and neukirch , t. : 1994 , _ solar phys . _ * 153 * , 287 .
pneuman , g.w . and kopp , r.a .
: 1971 , _ solar phys .
_ * 18 * , 258 .
schindler , k. : 1972 , in b.m .
mccormac ( ed . ) , _ earth s magnetospheric processes _ , d. reidel publ . comp . , dordrecht 1972 , p.200 schindler , k. and birn , j. : 1982 , _ j. geophys .
res . _ * 87 * , 2263 .
schwenn , r. , inhester , b. , plunkett , s.p . ,
epple , a. , podlipnik , b. et al . : 1997 _ solar phys .
_ , in press .
wang , a .- h .
, wu , s.t . , suess , s.t . and
poletto , g. : 1993 , _ solar phys . _
* 147 * , 55 .
wang , s. , liu , y.f . and zheng , h.n .
: 1997 , _ solar phys . _ * 173 * , 409 .
wiegelmann , t. : 1997 , _ physica scripta _ , in press .
wiegelmann , t. and schindler , k. : 1995 , _ geophys .
* 22 * , 2057 .
woo , r. , armstrong , j.w .
, bird , m.k . and
ptzold , m. : 1995 , _ astrophys . j. _ * 449 * , l91 .
, guo , w.p . and wang , j.f .
: 1995 , _ solar phys . _
* 157 * , 325 .
yan , m. , otto , a. muzzell , d. and lee , l.c . : 1994 , _
j. geophys .
_ * 99 * , 8657 .
young , r. and hameiri , e. : 1992 , _ j. geophys .
res . _ * 97 * , 16789 .
zwingmann , w. : 1987 , _ solar phys . _ * 111 * , 309 . | recent observations of the solar corona with the lasco coronagraph on board of the soho spacecraft have revealed the occurrence of triple helmet streamers even during solar minimum , which occasionally go unstable and give rise to particularly huge coronal mass ejections .
we present a method to calculate ( semi-)analytically self - consistent stationary configurations of triple helmet streamers which can serve as input for stability considerations and dynamical calculations .
the method is based on an asymptotic expansion procedure using the elongated structure of the streamers .
the method is very flexible and can be used in both cartesian and spherical geometry .
we discuss the effects of magnetic shear , gravity and field - aligned flow on open field lines .
example solutions illustrating the influence of each of these features on the solution structure are presented . |
chromospheric filaments ( prominences at the limb ) are one of the major features characterizing the solar activity in low solar atmosphere ( chromosphere and low corona ) .
filament eruptions are at the core of many solar - terrestrial effects or space weather .
the filaments are formed along magnetic polarity inversion lines , but only a minority of these neutral lines ( nls ) has filaments above them .
several studies were made trying to find the necessary conditions for filament formation , but still there is no clear understanding of this @xcite .
it was found by previous studies that filaments form only in filament channels ( for review , see ) , which overlay nls .
some studies have indicated that the convergence of magnetic flux leading to either flux cancellation or reconnection is required for filament formation ( see ; ; , and references therein ) .
recently , suggested that formation of filament and filament channels is part of development of large - scale chiral systems . in our present study , we concentrate only on properties of neutral lines , and do not consider other aspects of filament formation ( such as , for example , role of helicity ) .
have shown that one important condition for the formation of a filament channel is a strong horizontal component of magnetic field aligned with the magnetic neutral line .
thus , one can ask if the properties ( and evolution ) of magnetic nls could be sufficient to identify potential location of filament formation ? indeed , several studies ( _ e.g. _ , ; ) have suggested that the gradient of magnetic field transverse to the neutral line may play a significant role in filament formation . according to one of the necessary conditions for the filament formation is a low gradient of magnetic field transverse to nl .
here we conduct a statistical study of horizontal gradients of magnetic field across neutral lines , and investigate how these properties differ for nls with filaments and without them .
we also investigate whether the properties of neutral lines change with solar cycle , and whatever this change can be the cause of variation in filaments length and total number with the solar cycle .
we employ daily h@xmath0 full - disk images from mauna loa solar observatory ( mlso ) and soho / mdi ( _ michelson doppler imager _ ) synoptic maps for 1997 - 2010 years . for our main h@xmath0 data
set , we included daily full - disk images corresponding to two complete carrington rotations ( crs ) per year with a time separation between selected crs about five months .
however , due to lack of data , only one rotation was taken for the years 1997 , 2003 , and 2010 ; and for year 2000 , we selected three ( non - consecutive ) solar rotations .
in addition to mdi synoptic maps ( for same carrington rotations as our h@xmath0 data ) , we also use the _ synoptic optical long - term investigations of the sun _ ( solis , ) synoptic maps and the full - disk magnetograms taken at ca ii 854.2 nm chromospheric line to study properties of the chromospheric neutral lines . because of the statistical nature of our study , we automated the identification of filaments ( using h@xmath0 full - disk images ) and neutral lines ( using mdi synoptic maps ) .
several studies were made in developing filament automatic detection ( _ e.g. _ , ; ; ; ) .
most of them have an extensive algorithm of filament detection from an image taken in a single wavelength or use multiple images taken in different wavelengths . in our algorithm
, we used both h@xmath0 daily full - disk images and magnetograms .
this allowed us to significantly simplify the process of filament detection by comparing the location of potential filaments ( dark features initially found on h@xmath0 images ) with underlying magnetic field .
this approach also helped to reduce ( computer ) processing time .
the criterion for filament identification is an intensity threshold which is defined as an average intensity over the full - disk image minus two standard deviations of the average .
all pixels satisfying the above criterion ( but excluding pixels near the limb ) are considered as potential filaments .
the final selection takes into account existence of nl inside the filament contour .
magnetic neutral lines were identified on synoptic maps after smoothing them by 50 x 50 pixels running window .
figure 1 shows an example of a h@xmath0 full - disk image with filament contours found by our automated procedure .
it is clear that our procedure successfully identifies all major filaments .
small clumps of filament material could be missed especially in locations near the solar limb .
for example , while the procedure identifies one clump of filament material near the north polar region ( near the top of solar disk in figure 1 ) , it misses several other clumps in the same location . as an additional test , for a few selected solar rotations
, we compared our filament identification with the data from the bass2000 solar survey database ( bass2000.obspm.fr ) .
figure 2 demonstrates a very good agreement between filaments from bass2000 data base ( green ) , and those identified by our method ( red ) .
figure 3 is an mdi synoptic map with neutral lines drawn in different colors .
the colors are used to indicate the type of nl : green is for the lines between magnetic polarities inside one active region , and red is for the lines separating polarities of two different active regions . for the reasons of automation ,
the determination of a type of nl ( inter / intra- active region ) is based on the polarity of fields to the west and east of the neutral line . if the orientation of these two polarities followed the hale polarity rule , we called the neutral line as intra - active region . otherwise , it was classified as inter - region nl .
inevitably , this approach mis - identifies non - hale polarity active regions .
but the fraction of such regions is very minor , and thus , the effect on overall statistics is insignificant .
figure 4 provides example of neutral lines associated with the chromospheric filaments as identified by our procedure .
the portions of neutral lines with filaments are marked by red color .
although we analyze only two solar rotations per year , our data set of filaments is representative of a complete data . for example , a latitudinal distribution of filaments in our data set exhibits well - known latitudinal drift ( _ e.g. _ , ) . during the rising phase of solar activity
, the filaments distribution ( in our set ) shows a clear preference for high latitudes , and as the cycle progresses , more filaments are found at lower latitudes following the equatorward drift similar to sunspot s butterfly diagram .
while the chromospheric filaments are located above the magnetic neutral lines , it is not uncommon to see a displacement between the location of a nl ( as measured in the photosphere ) and the filament body ( as observed in h@xmath0 ) in full - disk images .
one can contribute such a displacement to the fact that filaments are the features located at certain heights above the photosphere and to their projection to the image plane .
however , our analysis suggests that this displacement could not be explained by a simple projection effect alone . to further explore this displacement
, we analyzed a number of filaments observed near the central meridian .
high - latitude filaments ( @xmath1lat@xmath2 60 degrees ) were excluded from this analysis .
we also compared the change in filament - nl displacements for selected filaments near the east and west limbs .
the displacements between the filament body and the neutral line were present in different locations independent of closeness to the limb or the central meridian . in some cases ,
only a part of a filament was displaced relative to closest neutral line , while the rest of the filament was perfectly aligned with nl .
figure 5 ( left panel ) shows one such an example .
since the filaments could be located at a certain height above the photosphere , one could argue that their location should correlate better with the neutral lines of the chromospheric fields ( which are formed higher in the atmosphere ) .
contrary to this expectation , some filaments show a larger displacement relative to closest chromospheric neutral line as compared with the photospheric nl ( compare left and right panels in figure 5 ) .
thus , the displacement of filament location ( as seeing in image plane of the photospheric and chromospheric magnetograms ) is clearly not due to a projection effect alone .
one possible explanation for the displacements described above is that the `` neutral lines '' are not lines but planes separating opposite polarities .
these `` polarity inversion planes '' are not always vertical , but they could be inclined depending on flux systems they separate .
one can visualize this as a ribbon ( perhaps , a current sheet ) `` waving '' around the isolated coronal flux systems .
the intersection of this polarity inversion plane with the photosphere corresponds to the photospheric neutral line , and the filament can form within this plane higher up .
if this interpretation is correct , the displacement between the photospheric nl and filament body would be an interplay between the location ( height ) of filament within the polarity inversion plane and the tilt of the plane relative to solar surface .
the idea that ( some ) chromospheric filaments are sheet - like structures is not new .
we simply use it to demonstrate that a broadly held belief that the filaments are located above magnetic neutral lines should be taken with a clear understanding that in image plane that location of filament bodies and the neutral lines could be displaced from each other , and not only due to elevation of filament above the photosphere , but also because of its height geometry . in some cases of relatively weak ( and fragmented ) magnetic fields ,
the exact location of the photospheric neutral line is very uncertain . in some of these cases ,
a continuous photospheric neutral line may not even exist underneath of a filament .
for example , filament segment in the low - right corner of figure 6 overlies patch of negative polarity ; there appears no clear neutral line ( of photospheric field ) that can be associated with this filament segment .
the inconsistencies between the location of filaments and the photospheric/ chromospheric neutral lines described in this section may indicate that the filament location is governed by the coronal magnetic fields rather than the photospheric or even the chromospheric fields . as a rule , we find a better agreement between the neutral lines and filament positions when we use synoptic maps .
we speculate that since these maps are based on an averaged magnetic field , they may better represent the large - scale magnetic domains . even though for filament identification we employ daily full - disk images , we do average filament positions over several days , and thus , any projection effects are minimized . based on the results of the analysis presented in this section , for our study of magnetic properties of nls we chose to use the synoptic magnetograms .
first , we investigate gradients of magnetic field across the neutral line and their variation with level of solar activity .
figure 7 shows temporal variation of gradient distribution for nls with filaments ( left panel ) and without them ( right panel ) .
the gradients were computed at uniformly spaced intervals along magnetic neutral lines .
then , each segment was identified either as having a filament or not based on proximity of filament material to this part of neutral line .
we did not attempt any averaging of field gradients for individual filaments .
different colors are used to indicate the fraction of nls with a specific property .
for example , contour labeled 0.95 means that 95% of the nls have gradients less or equal of gradient shown as abscissa on figure 7 .
for instance , for the year 1997 , 95% of nls with filaments ( figure 7 , left ) have gradients @xmath3 0.011 g km@xmath4 , while 95 % of nls without filaments ( figure 7 , right ) have gradients @xmath3 0.0088 g km@xmath4 . at the first glance ,
the distributions of gradients across nls with and without filaments are very similar . however , neutral lines with filaments exhibit systematically stronger gradients as compared with nls without filaments ( compare figure 7 left and right panels ) .
filaments are observed above the nls with gradients ranging from 6.6 @xmath5 10@xmath6 g km@xmath4 to 0.11 g km@xmath4 .
only a small fraction of filaments are formed above the nls with gradients outside that range .
a fraction of the neutral lines with gradients smaller than about 0.001 g km@xmath4 is about 20% for nls with or without filaments , and it does not seem to vary with solar activity cycle . a fraction of neutral lines with gradients higher than 0.005 g km@xmath4 shows a strong variation with solar activity cycle . by its amplitude , nls with and without filaments show comparable change between solar minimum and maximum : for nls with filaments , 95% level decreases from about 0.019 g km@xmath4 ( year 2002 ) to about 0.007 g km@xmath4 ( year 2009 ) , while for nls without filaments it changes from 0.017 g km@xmath4 ( year 2002 ) to 0.006 g km@xmath4 ( year 2009 ) . on the other hand ,
the changes for nls with filaments occur over a longer period of time ( _ i.e. _ , peak corresponding to 95% level on figure 7 left panel is much `` broader '' as compared with figure 7 , right panel ) . in 2002 , about 50% ( between 0.95 and 0.45 contours in figure 7 ) nls with filaments fall within 0.019 - 0.0035 g km@xmath4 . in the same year , 35% of all filaments
( between 0.95 and 0.45 contours in figure 7 ) formed above nls within 0.019 - 0.005 g km@xmath4 range of gradients . in comparison , 50% of nls without filaments fell within 0.017 - 0.0025 g km@xmath4 , and 35% of nls ( without filaments ) were in 0.017 - 0.004 g km@xmath4 range . in 2004 , 50% ( 0.95 - 0.45 levels in figure 7 ) of filaments developed above nls with
0.0155 - 0.0022 g km@xmath4 magnetic field gradients , while 50% nls without filaments fall within 0.0095 - 0.0018 g km@xmath4 range of gradients . during periods of low sunspot activity ( _ i.e. _ , 2007 - 2009 ) ,
neutral lines with and without filaments have similar range of gradients ( _ e.g. _ , in 2007 , 85% of neutral lines with and without filaments were below about 0.005 g km@xmath4 ( c.f
. , figure 7 left and right panels ) .
thus , our data show a clear tendency for nls with filaments to have stronger magnetic gradients during the rising and declining phases and the maximum of solar cycle , as compared with nls without filaments .
although filaments are formed above the nls with a wide range of gradients , there appears to be a preferred range of gradients where filaments are observed more often .
this preferred range covers gradients from about 0.0025 to 0.019 g km@xmath4 .
less than 20% of filaments were formed at gradients below about 0.001 g km@xmath4 , and less than about 5% of filaments have developed above nls with 0.0019 g km@xmath4 ( at solar maximum ) . on the other hand ,
having `` preferred '' horizontal gradient does not guarantee the filament formation there ; large fraction of nls which exhibit similar gradients remains empty of filament material .
furthermore , on average , only a small fraction of neutral lines is associated with filaments .
this fraction varies from about 2% at solar minimum to 15% at solar maximum ( figure 8) .
next , we investigate the properties of magnetic field gradients for different types of filaments .
one of the earliest classification schemes ( _ e.g. _ , ) segregates filaments on two categories based on the nature of the nl above which the filament lies .
if filament overlays a neutral line separating two polarities forming an active region , it is classified as a ` bipolar region filament ' , or type a. when filament is formed above a nl that separates two different active regions , it is called a ` between bipolar region filament ' , or type b. to automate the categorization of nls , we assigned their type based on magnetic polarity orientation and the hale polarity rule ( _ i.e. _ , if the ( leading - following ) polarity orientation for a given nl agrees with the hale - nicholson polarity rule , this nl and filament above it were classified as type a ) .
otherwise , the nl ( and a filament ) were classified as type b. example of such classification is shown in figure 2 .
green color is used for type a lines and red color for type b. although this simplified approach allows an automatic identification of the type of nls and filaments , it has a limitation of mis - identifying nls and filaments associated with non - hale polarity regions .
likely , only a small fraction ( about a few percent ) of all active regions disobeys the hale - nicholson polarity rule .
two types of nls ( and filaments ) do not appear to show any significant difference in properties of magnetic gradients .
we did find , however , a slight preference for filaments to form in between active regions ( type b ) at a rising phase of solar cycle , while type a filaments ( inside active region ) were more abundant at the declining phase of cycle .
lastly , we expanded our investigation to the neutral lines derived from the chromospheric magnetic fields observed by _ vector spectromagnetograph _ ( vsm ) on solis .
we applied the same technique as for the photospheric magnetograms .
the results are very similar to those derived from the photospheric fields with respect to the pattern of gradients described above .
we studied the properties of horizontal gradients across neutral lines depending on presence or absence of filaments above them .
we found a preferred range of nl gradients where filaments are observed relatively more frequently .
still only about 18% of neutral lines within that `` preferred range '' of gradients have filaments above them .
thus , a mere fact that a horizontal gradient across magnetic neutral line falls within a `` preferred '' range does not guarantee that a filament will form in that location .
one can not exclude , however , that together with other factors these gradient properties could play a role in creating overall conditions for filament formation .
thus , for example , several studies suggested that convergence of magnetic field may be required for the filament formation ( _ e.g. _ , ; ; ) .
another factor that could play important role in filament formation is the rate of canceling magnetic features that were hypothesized to supply the material to filaments @xcite . due to spatial and temporal limitation of the data set used in this study
, we did not investigate the role of these other factors in filament formation .
the gradients for neutral lines with filaments show similar variations with the solar cycle as do the neutral lines without filaments ( _ i.e. _ , an increase in fraction of stronger gradients during the maximum of solar activity ) . however , this change in properties of gradients for nls with filaments is enhanced during the rising and declining phases of cycle , while the properties of nls without filaments are more narrowly defined to a maximum of solar cycle . during these periods ( of rising and declining phases of solar cycle ) , a larger fraction of nls within preferred range of gradients
is occupied by filaments .
although we did not find a strong preponderance of filament formation on the properties of magnetic gradients of neutral lines , our data do suggest that there is a range of magnetic gradients preferred for filament formation . a cyclic increase in fraction of neutral lines ( with these preferred gradients )
may provide enhancement in favorable conditions for more filament formation during solar maximum .
however , there is a need for an additional mechanism(s ) that takes advantage of the above favorable conditions to create more filaments .
such additional condition could be the amount of material in the corona available for filament formation .
one can speculate that during solar maximum , overall conditions in the photosphere and chromosphere ( _ e.g. _ , enhanced network fields , larger number of small - scale reconnection events _
etc_. ) may result in a significantly larger amount of material supplied to the corona . in combination with the increased fraction of nls with stronger gradients , having more material in the corona
may create more opportunities for condensation of this material to form the chromospheric filaments .
finally , the noted displacement of filaments ( with respect to the photospheric neutral lines ) and the lack of a strong correlation between the gradient properties of nls and filament formation may suggest that filament formation is governed by properties of coronal magnetic fields rather than the photospheric or chromospheric magnetic fields . in itself , this may fit better with the model in which the excess of material in the corona combined with changes in properties of nls at solar cycle maximum is the prime reason for a solar cycle variation in chromospheric filaments . at the present , no such model exists .
the authors acknowledge funding from nasa s nnh09al04i inter agency transfer and nsf grant ags-0837915 .
this work utilizes solis data obtained by the nso integrated synoptic program ( nisp ) , managed by the national solar observatory , which is operated by the association of universities for research in astronomy ( aura ) , inc . under a cooperative agreement with the national science foundation .
soho is a project of international cooperation between esa and nasa .
national solar observatory ( nso ) is operated by the association of universities for research in astronomy , aura inc under cooperative agreement with the national science foundation ( nsf ) .
dmoulin , p. : 1998 , _
magnetic fields in filaments _ , in a.s.p .
series , vol . * 150 * , iau colloq .
167 : new perspectives on solar prominences , eds .
david webb , david rust , brigitte schmieder ( san francisco : asp ) , 78 .
gaizauskas , v. : 1998 , _
`` filament channels : essential ingredients for filament formation '' _ , in a.s.p .
series , vol . * 150 * , iau colloq .
167 : new perspectives on solar prominences , eds .
david webb , david rust , brigitte schmieder ( san francisco : asp ) , 257 . | we investigate gradient of magnetic field across neutral lines ( nls ) , and compare their properties for nls with and without the chromospheric filaments .
our results show that there is a range of preferred magnetic field gradients where the filament formation is enhanced . on the other hand
, a horizontal gradient of magnetic field across a nl alone does not appear to be a single factor that determines if a filament will form ( or not ) in a given location . |
hopf algebras were introduced in the 1950 s from three different perspectives : algebraic groups in positive characteristic , cohomology rings of lie groups , and group objects in the category of von neumann algebras .
the study of non - commutative non - cocommutative hopf algebras started in the 1960 s .
the fundamental breakthrough is drinfeld s report @xcite . among many contributions and ideas ,
a systematic construction of solutions of the quantum yang - baxter equation ( qybe ) was presented .
let @xmath0 be a vector space .
the qybe is equivalent to the braid equation : @xmath1 if @xmath2 satisfies , then @xmath3 is called a braided vector space ; this is a down - to - the - earth version of a braided tensor category @xcite .
drinfeld introduced the notion of quasi - triangular hopf algebra , meaning a pair @xmath4 where @xmath5 is a hopf algebra and @xmath6 is invertible and satisfies the approppriate conditions , so that every @xmath5-module @xmath0 becomes a braided vector space , with @xmath2 given by the action of @xmath7 composed with the usual flip .
furthermore , every finite - dimensional hopf algebra @xmath5 gives rise to a quasi - triangular hopf algebra , namely the drinfeld double @xmath8 as vector space .
if @xmath5 is not finite - dimensional , some precautions have to be taken to construct @xmath9 , or else one considers yetter - drinfeld modules , see [ subsec : repns ] . in conclusion , every hopf algebra is a source of solutions of the braid equation .
essential examples of quasi - triangular hopf algebras are the quantum groups @xmath10
@xcite and the finite - dimensional variations @xmath11 @xcite . in the approach to the classification of hopf algebras exposed in this report , braided vector spaces and braided tensor categories play a decisive role ; and the finite quantum groups are the main actors in one of the classes that splits off . by space limitations
, there is a selection of the topics and references included
. particularly , we deal with finite - dimensional hopf algebras over an algebraically closed field of characteristic zero with special emphasis on description of examples and classifications .
interesting results on hopf algebras either infinite - dimensional , or over other fields , unfortunately can not be reported .
there is no account of the many deep results on tensor categories , see @xcite .
various basic fundamental results are not explicitly cited , we refer to @xcite for them ; classifications of hopf algebras of fixed dimensions are not evoked , see @xcite .
i thank all my coauthors for pleasant and fruitful collaborations .
i am particularly in debt with hans - jrgen schneider , matas graa and ivn angiono for sharing their ideas on pointed hopf algebras with me .
let @xmath12 and @xmath13 .
the base field is @xmath14 . if @xmath15 is a set , then @xmath16 is its cardinal and @xmath17 is the vector space with basis @xmath18 .
let @xmath19 be a group : we denote by @xmath20 the set of isomorphism classes of irreducible representations of @xmath19 and by @xmath21 the subset of those of dimension 1 ; by @xmath22 the centralizer of @xmath23 ; and by @xmath24 its conjugacy class . more generally we denote by @xmath25 the set of isomorphism classes of simple objects in an abelian category @xmath26 .
the group of @xmath27-th roots of 1 in @xmath14 is denoted @xmath28 ; also @xmath29 . the group presented by @xmath30 with relations @xmath31
is denoted @xmath32 .
the notation for hopf algebras is standard : @xmath33 , @xmath34 , @xmath35 , denote respectively the comultiplication , the counit , the antipode ( always assumed bijective , what happens in the finite - dimensional case ) .
we use sweedler s notation : @xmath36 . similarly ,
if @xmath37 is a coalgebra and @xmath0 is a left comodule with structure map @xmath38 , then @xmath39 . if @xmath40 are subspaces of @xmath37 , then @xmath41 ; also @xmath42 and @xmath43 for @xmath44 .
the first examples of finite - dimensional hopf algebras are the group algebra @xmath45 of a finite group @xmath19 and its dual , the algebra of functions @xmath46 .
indeed , the dual of a finite - dimensional hopf algebra is again a hopf algebra by transposing operations . by analogy with groups ,
several authors explored the notion of extension of hopf algebras at various levels of generality ; in the finite - dimensional context , every extension @xmath47 can be described as @xmath37 with underlying vector space @xmath48 , via a heavy machinery of actions , coactions and non - abelian cocycles , but actual examples are rarely found in this way ( extensions from a different perspective are in @xcite ) .
relevant exceptions are the so - called _ abelian extensions _ @xcite ( rediscovered by takeuchi and majid ) : here the input is a matched pair of groups @xmath49 with mutual actions @xmath50 , @xmath51 ( or equivalently , an exact factorization of a finite group ) .
the actions give rise to a hopf algebra @xmath52 .
the multiplication and comultiplication can be further modified by compatible cocycles @xmath53 , producing to the abelian extension @xmath54 . here
@xmath53 turns out to be a 2-cocycle in the total complex associated to a double complex built from the matched pair ; the relevant @xmath55 is computed via the so - called kac exact sequence .
it is natural to approach hopf algebras by considering algebra or coalgebra invariants .
there is no preference in the finite - dimensional setting but coalgebras and comodules are locally finite , so we privilege the coalgebra ones to lay down general methods .
the basic coalgebra invariants of a hopf algebra @xmath5 are : @xmath56 the group @xmath57 of group - like elements of @xmath5 .
@xmath56 the space of skew - primitive elements @xmath58 , @xmath59 ; @xmath60 . @xmath56 the coradical @xmath61 , that is the sum of all simple subcoalgebras .
@xmath56 the coradical filtration @xmath62 , where @xmath63 ; then @xmath64 .
the category @xmath65 of left modules over a hopf algebra @xmath5 is monoidal with tensor product defined by the comultiplication ; ditto for the category @xmath66 of left comodules , with tensor product defined by the multiplication . here
are two ways to deform hopf algebras without altering one of these categories .
@xmath67 let @xmath68 be invertible such that @xmath69 and @xmath70
. then @xmath71 ( the same algebra with comultiplication @xmath72 ) is again a hopf algebra , named the _ twisting _ of @xmath5 by @xmath73 @xcite .
the monoidal categories @xmath65 and @xmath74 are equivalent . if @xmath5 and @xmath75 are finite - dimensional hopf algebras with @xmath65 and @xmath76 equivalent as monoidal categories , then there exists @xmath73 with @xmath77 as hopf algebras ( schauenburg , etingof - gelaki ) .
examples of twistings not mentioned elsewhere in this report are in @xcite .
@xmath67 given a linear map @xmath78 with analogous conditions , there is a hopf algebra @xmath79 ( same coalgebra , multiplication twisted by @xmath80 ) such that the monoidal categories @xmath66 and @xmath81 are equivalent @xcite .
a yetter - drinfeld module @xmath82 over @xmath5 is left @xmath5-module and left @xmath5-comodule with the compatibility @xmath83 , for all @xmath84 and @xmath85 . the category @xmath86 of yetter - drinfeld modules is braided monoidal.[page : yd ] that is , for every @xmath87
, there is a natural isomorphism @xmath88 given by @xmath89 , @xmath84 , @xmath90 .
when @xmath5 is finite - dimensional , the category @xmath86 is equivalent , as a braided monoidal category , to @xmath91 .
the definition of hopf algebra makes sense in any braided monoidal category .
hopf algebras in @xmath86 are interesting because of the following facts discovered by radford and interpreted categorically by majid , see @xcite : @xmath92 if @xmath7 is a hopf algebra in @xmath86 , then @xmath93 with semidirect product and coproduct is a hopf algebra , named the _ bosonization _ of @xmath7 by @xmath5 .
@xmath92 let @xmath94 be hopf algebra maps as in @xmath95 & \ar@/^-1pc/_{\iota}[l ] h}$ ] with @xmath96 .
then @xmath97 is a hopf algebra in @xmath86 and @xmath98 . for instance ,
if @xmath99 , then the tensor algebra @xmath100 is a hopf algebra in @xmath86 , by requiring @xmath101 .
if @xmath102 satisfies @xmath103 , @xmath104 the usual flip , then the exterior algebra @xmath105 is a hopf algebra in @xmath86 .
there is a braided adjoint action of a hopf algebra @xmath7 in @xmath86 on itself , see e.g. ( * ? ? ?
* ( 1.26 ) ) .
if @xmath106 and @xmath107 , then @xmath108 .
a quasitriangular hopf algebra @xmath109 is _ triangular _ if the braiding induced by @xmath7 is a symmetry : @xmath110 for all @xmath111 .
a finite - dimensional triangular hopf algebra is a twisting of a bosonization @xmath112 , where @xmath19 is a finite group and @xmath113 has @xmath114 @xcite .
this lead eventually to the classification of triangular finite - dimensional hopf algebras @xcite ; previous work on the semisimple case culminated in @xcite .
the algebra of functions @xmath46 on a finite group @xmath19 admits a haar measure , i.e. , a linear function @xmath115 invariant under left and right translations , namely @xmath116 sum of all elements in the standard basis of @xmath45 .
this is adapted as follows : a right integral _ on _ a hopf algebra @xmath5 is a linear function @xmath117 which is invariant under the left regular coaction : analogously there is the notion of left integral .
the notion has various applications .
assume that @xmath5 is finite - dimensional .
then an integral _ in _
@xmath5 is an integral on @xmath118 ; the subspace of left integrals in @xmath5 has dimension one , and there is a generalization of maschke s theorem for finite groups : @xmath5 is semisimple if and only if @xmath119 for any integral @xmath120 .
this characterization of semisimple hopf algebras , valid in any characteristic , is one of several , some valid only in characteristic 0 .
see @xcite .
semisimple hopf algebras can be obtained as follows : @xmath92 a finite - dimensional hopf algebra @xmath5 is semisimple if and only if it is cosemisimple ( that is , @xmath118 is semisimple ) .
@xmath92 given an extension @xmath121 , @xmath5 is semisimple iff @xmath75 and @xmath122 are .
notice that there are semisimple extensions that are not abelian @xcite .
@xmath92 if @xmath5 is semisimple , then so are @xmath71 and @xmath79 , for any twist @xmath73 and cocycle @xmath80 . if @xmath19 is a finite simple group , then any twisting of @xmath45 is a simple hopf algebra ( i.e. , not a non - trivial extension ) @xcite , but the converse is not true @xcite .
@xmath92 a bosonization @xmath123 is semisimple iff @xmath7 and @xmath5 are . to my knowledge ,
all examples of semisimple hopf algebras arise from group algebras by the preceding constructions ; this was proved in @xcite in low dimensions and in @xcite for dimensions @xmath124 , @xmath125 , where @xmath126 , @xmath127 and @xmath128 are primes .
see ( * ? ? ?
* question 2.6 ) . an analogous question in terms of fusion categories : is any semisimple hopf algebra weakly group - theoretical ?
see ( * ? ? ?
* question 2 ) . there are only finitely many isomorphism classes of _ semisimple _ hopf algebras in each dimension @xcite , but this fails in general @xcite .
( kaplansky ) .
let @xmath5 be a semisimple hopf algebra .
the dimension of every @xmath129 divides the dimension of @xmath5 .
the answer is affirmative for iterated extensions of group algebras and duals of group algebras @xcite and notably for semisimple quasitriangular hopf algebras @xcite .
nichols algebras are a special kind of hopf algebras in braided tensor categories .
we are mainly interested in nichols algebras in the braided category @xmath86 , where @xmath5 is a hopf algebra , see page .
in fact , there is a functor @xmath130 from @xmath86 to the category of hopf algebras in @xmath86 .
their first appearence is in the precursor @xcite ; they were rediscovered in @xcite as part of a quantum differential calculus " , and in @xcite to present the positive part of @xmath10 .
see also @xcite .
there are several , unrelated at the first glance , alternative definitions .
let @xmath99 .
the first definition uses the representation of the braid group @xmath131 in @xmath27 strands on @xmath132 , given by @xmath133 , @xmath2 in @xmath134 tensorands ; here recall that @xmath135 .
let @xmath136 be the matsumoto section and let @xmath137 be the quantum symmetrizer , @xmath138 . then define @xmath139 hence @xmath140 is a graded hopf algebra in @xmath86 with @xmath141 , @xmath142 ; by the algebra structure depends only on @xmath2 .
to explain the second definition , let us observe that the tensor algebra @xmath100 is a hopf algebra in @xmath86 with comultiplication determined by @xmath143 for @xmath144 .
then @xmath145 coincides with the largest homogeneous ideal of @xmath100 generated by elements of degree @xmath146 that is also a coideal .
let now @xmath147 be a graded hopf algebra in @xmath86 with @xmath148 .
consider the conditions @xmath149 these requirements are dual to each other : if @xmath150 has finite - dimensional homogeneous components and @xmath151 is the graded dual of @xmath150 , i.e. , @xmath152 , then @xmath150 satisfies if and only if @xmath7 satisfies .
these conditions determine @xmath153 up to isomorphisms , as the unique graded connected hopf algebra @xmath150 in @xmath86 that satisfies @xmath154 , and .
there are still other characterizations of @xmath145 , e.g. as the radical of a suitable homogeneous bilinear form on @xmath100 , or as the common kernel of some suitable skew - derivations .
see @xcite for more details . despite all these different definitions ,
nichols algebras are extremely difficult to deal with , e.g. to present by generators and relations , or to determine when a nichols algebra has finite dimension or finite gelfand - kirillov dimension .
it is not even known a priori whether the ideal @xmath145 is finitely generated , except in a few specific cases .
for instance , if @xmath2 is a symmetry , that is @xmath155 , or satisfies a hecke condition with generic parameter , then @xmath153 is quadratic . by the efforts of various authors ,
we have some understanding of finite - dimensional nichols algebras of braided vector spaces either of diagonal or of rack type , see [ subsec : nichols - diagonal ] , [ subsec : nichols - rack ] .
we now explain how nichols algebras enter into our approach to the classification of hopf algebras .
recall that a hopf algebra has the _ dual chevalley property _ if the tensor product of two simple comodules is semisimple , or equivalently if its coradical is a ( cosemisimple ) hopf subalgebra .
for instance , a _ pointed _ hopf algebra , one whose simple comodules have all dimension one , has the dual chevalley property and its coradical is a group algebra . also , a _ copointed _ hopf algebra
( one whose coradical is the algebra of functions on a finite group ) has the dual chevalley property .
the lifting method is formulated in this context @xcite .
let @xmath5 be a hopf algebra with the dual chevalley property and set @xmath156 . under this assumption ,
the graded coalgebra @xmath157 associated to the coradical filtration becomes a hopf algebra and considering the homogeneous projection @xmath158 as in @xmath159 & { \operatorname{gr}}h \ar@<-0.5ex>@{->}_{\quad \pi}[r ] & \ar@<-0.5ex>[l ] k } $ ] we see that @xmath160 .
the subalgebra of coinvariants @xmath7 is a graded hopf algebra in @xmath161 that inherits the grading with @xmath162 ; it satisfies since the grading comes from the coradical filtration .
let @xmath163 be the subalgebra of @xmath7 generated by @xmath164 ; then @xmath165 . the braided vector space @xmath0 is a basic invariant of @xmath5 called its _
infinitesimal braiding_. let us fix then a semisimple hopf algebra @xmath75 .
to classify all finite - dimensional hopf algebras @xmath5 with @xmath166 as hopf algebras , we have to address the following questions . 1 .
[ item : method - nichols ] determine those @xmath167 such that @xmath168 is finite - dimensional , and give an efficient defining set of relations of these .
[ item : method - gen - deg - one ] investigate whether any _ finite - dimensional _ graded hopf algebra @xmath7 in @xmath161 satisfying @xmath162 and @xmath169 , is a nichols algebra .
[ item : method - lifting ] compute all hopf algebras @xmath5 such that @xmath170 , @xmath0 as in [ item : method - nichols ] .
since the nichols algebra @xmath153 depends as an algebra ( and as a coalgebra ) only on the braiding @xmath2 , it is convenient to restate question [ item : method - nichols ] as follows : 1
. determine those braided vector spaces @xmath171 in a suitable class such that @xmath172 , and give an efficient defining set of relations of these .
2 . for those @xmath0 as in ( a@xmath173 ) ,
find in how many ways , if any , they can be realized as yetter - drinfeld modules over @xmath75 .
for instance , if @xmath174 , @xmath175 a finite abelian group , then the suitable class is that of braided vector spaces of diagonal type . in this context
, question ( a@xmath176 ) amounts to solve systems of equations in @xmath175 .
the answer to [ item : method - nichols ] is instrumental to attack [ item : method - gen - deg - one ] and [ item : method - lifting ] .
question [ item : method - gen - deg - one ] can be rephrased in two equivalent statements : 1 .
investigate whether any _ finite - dimensional _ graded hopf algebra @xmath150 in @xmath161 with @xmath148 and generated as algebra by @xmath177 , is a nichols algebra .
2 . investigate whether any _ finite - dimensional _ hopf algebra @xmath5 with @xmath178 is generated as algebra by @xmath179 .
we believe that the answer to [ item : method - gen - deg - one ] is affirmative at least when @xmath75 is a group algebra . in other words , by the reformulation ( b@xmath176 ) : [ conj : gendegone ] _
@xcite _ every _ finite - dimensional _ pointed hopf algebra is generated by group - like and skew - primitive elements .
as we shall see in [ subsec : gen - deg-1 ] , the complete answer to [ item : method - nichols ] is needed in the approach proposed in @xcite to attack conjecture [ conj : gendegone ] .
it is plausible that the answer of ( b@xmath176 ) is affirmative for every semisimple hopf algebra @xmath75 .
question [ item : method - lifting ] , known as _ lifting _ of the relations , also requires the knowledge of the generators of @xmath145 , see [ subsec : deformations ] . before starting with the analysis of the various questions in [ subsec : chevalley ]
, we discuss a possible approach to more general hopf algebras @xcite .
let @xmath5 be a hopf algebra ; we consider the following invariants of @xmath5 : @xmath56 the _ hopf coradical _
@xmath180}$ ] is the subalgebra generated by @xmath61 .
@xmath56 the _ standard filtration _ @xmath180 } \subset h_{[1 ] } \subset \dots$ ] , @xmath181}=\wedge^{n+1 } h_{[0]}$ ] ; then @xmath182}$ ] .
if @xmath5 has the dual chevalley property , then @xmath181 } = h_n$ ] for all @xmath183 . in general , @xmath180}$
] is a hopf subalgebra of @xmath5 with coradical @xmath61 and we may consider the graded hopf algebra @xmath184}/h_{[n-1]}$ ] . as before , if @xmath185}$ ] is the homogeneous projection , then @xmath186 is a hopf algebra in @xmath187}}^{h_{[0]}}\mathcal{yd}}$ ] and @xmath188}$ ] .
furthermore , @xmath189 with grading inherited from @xmath190 .
this discussion raises the following questions . 1 .
[ item : gen - method - coalgebras ] let @xmath37 be a finite - dimensional cosemisimple coalgebra and @xmath191 a bijective anti - coalgebra map .
classify all finite - dimensional hopf algebras @xmath122 generated by @xmath37 , such that @xmath192 .
[ item : gen - method - r ] given @xmath122 as in the previous item , classify all finite - dimensional connected graded hopf algebras @xmath7 in @xmath193 .
[ item : gen - method - lifting ] given @xmath122 and @xmath7 as in previous items , classify all deformations or liftings , that is , classify all hopf algebras @xmath5 such that @xmath194 .
question [ item : gen - method - coalgebras ] is largely open , except for the remarkable ( * ? ? ?
* theorem 1.5 ) : if @xmath5 is a hopf algebra generated by an @xmath35-invariant 4-dimensional simple subcoalgebra @xmath37 , such that @xmath195 , then @xmath5 is a hopf algebra quotient of the quantized algebra of functions on @xmath196 at a root of unity @xmath197 .
nichols algebras enter into the picture in question [ item : gen - method - r ] ; if @xmath198 , then @xmath153 is a subquotient of @xmath7 .
question [ item : gen - method - lifting ] is completely open , as it depends on the previous questions .
here we expose two important notions introduced in @xcite .
let @xmath199 and @xmath200 .
a _ basic datum _ of type @xmath201 is a pair @xmath202 , where @xmath203 is a set and @xmath204 is a map such that @xmath205 for all @xmath206 .
let @xmath207 be the quiver @xmath208 over @xmath209 , with @xmath210 , @xmath211 ( here @xmath212 means target , @xmath213 means source ) .
let @xmath214 be the free groupoid over @xmath207 ; in any quotient of @xmath214 , we denote @xmath215 i.e. , the implicit superscripts are those allowing compositions .
coxeter datum _ is a triple @xmath216 , where @xmath202 is a basic datum of type @xmath201 and @xmath217 is a family of coxeter matrices @xmath218 with @xmath219 the _ coxeter groupoid _ @xmath220 associated to @xmath216 ( * ? ? ?
* definition 1 ) is the groupoid presented by generators @xmath207 with relations @xmath221 a generalized root system ( grs for short ) is a collection @xmath222 , where @xmath223 is a family of generalized cartan matrices @xmath224 , cf .
@xcite , and @xmath225 is a family of subsets @xmath226 .
we need the following notation : let @xmath227 be the canonical basis of @xmath228 and define @xmath229 by @xmath230 , @xmath231 , @xmath232 .
the collection should satisfy the following axioms : @xmath233 for all @xmath232 , @xmath234 .
we call @xmath235 , respectively @xmath236 , the set of _ positive _ , respectively _ negative _ , roots .
let @xmath237 , @xmath238 , @xmath239 , @xmath232 , and @xmath240 the subgroupoid of @xmath241 generated by all the @xmath242 , i.e. , by the image of the morphism of quivers @xmath243 , @xmath244 .
there is a coxeter matrix @xmath218 , where @xmath245 is the smallest natural number such that @xmath246 .
then @xmath217 fits into a coxeter datum @xmath216 , and there is an isomorphism of groupoids @xmath247 & \cw = \cw(\cx , \rho , \cc ) } $ ] @xcite ; this is called the _ weyl groupoid _ of @xmath248 . if @xmath249 , then @xmath250 , by .
the sets of _ real _ roots at @xmath232 are @xmath251 ; correspondingly the _ imaginary _ roots are @xmath252 .
assume that @xmath253 is connected .
then the following conditions are equivalent ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* lemma 2.11 ) : * @xmath254 is finite for some @xmath255 , * @xmath254 is finite for all @xmath255 , * @xmath256 is finite for all @xmath255 , * @xmath253 is finite .
if these hold , then all roots are real @xcite ; we say that @xmath248 is _
finite_. we now discuss two examples of grs , central for the subsequent discussion .
[ exa : grs - lie - super ] @xcite let @xmath257 be a field of characteristic @xmath258 , @xmath259 , @xmath260 and @xmath261 .
we assume @xmath262 for simplicity .
let @xmath263 .
let @xmath264 be the kac - moody lie superalgebra over @xmath257 defined as in @xcite ; it is generated by @xmath265 , @xmath266 and @xmath267 , @xmath206 , and the parity is given by @xmath268 , @xmath269 , @xmath270 , @xmath271 .
let @xmath272 be the root system of @xmath264 .
we make the following technical assumptions : @xmath273 the matrix @xmath274 is _ admissible _ if holds @xcite .
let @xmath275 be given by @xmath276 we need the following elements of @xmath257 : @xmath277 with the help of these scalars , we define a reflection @xmath278 , where @xmath279 , with @xmath280 , and @xmath281 , with @xmath282 [ thm : isomorfismo ti ] there is an isomorphism @xmath283 of lie superalgebras given ( for an approppriate basis @xmath284 of @xmath265 ) by @xmath285 assume that @xmath286 ; then and hold .
let @xmath287 then @xmath288 , where @xmath289 and @xmath290 , is a finite grs , an invariant of @xmath264 .
[ exa : grs - nichols - ss - yd ] let @xmath5 be a hopf algebra , assumed semisimple for easiness .
let @xmath291 be finite - dimensional , with a fixed decomposition @xmath292 , where @xmath293 .
then @xmath294 and @xmath295 are @xmath296-graded , by @xmath297 for all @xmath298 , @xmath299 .
recall that @xmath300 .
[ th : grs - semisimple ] _ @xcite _ if @xmath301 , then @xmath82 has a finite grs .
we discuss the main ideas of the proof .
let @xmath302 .
we define @xmath303 , @xmath304 then @xmath305 and @xmath306 is a generalized cartan matrix @xcite . also , @xmath307 is irreducible @xcite , @xcite .
let @xmath209 be the set of objects in @xmath86 with fixed decomposition ( up to isomorphism ) of the form @xmath308 then @xmath309 , where @xmath310 , is a crystallographic datum .
next we need : @xmath67 ( * ? ? ?
* theorem 4.5 ) ; @xcite there exists a totally ordered index set @xmath311 and families @xmath312 in @xmath313 , @xmath314 such that @xmath315 as @xmath296-graded objects in @xmath86 , where @xmath316 for all @xmath317 , @xmath318 .
let @xmath319 , @xmath320 , @xmath321 .
then @xmath322 is a finite grs .
[ th : classif - grs ] _
@xcite _ the classification of all finite grs is known .
the proof is a combinatorial tour - de - force and requires computer calculations .
it is possible to recover from this result the classification of the finite - dimensional contragredient lie superalgebras in arbitrary characteristic @xcite .
however , the list of @xcite is substantially larger than the classifications of the alluded lie superalgebras or the braidings of diagonal type with finite - dimensional nichols algebra .
let @xmath19 be a finite group .
we denote @xmath323 for @xmath324 .
so @xmath325 is a left @xmath19-module with a @xmath19-grading @xmath326 such that @xmath327 , for all @xmath328 . if @xmath329 , then the braiding @xmath88 is given by @xmath330 , @xmath331 , @xmath90 , @xmath332
now assume that @xmath333 is a finite abelian group .
then every @xmath334 is a @xmath175-graded @xmath175-module , hence of the form @xmath335 , where @xmath336 is the @xmath337-isotypic component of @xmath338 .
so @xmath339 is just the category of @xmath340-graded modules , with the braiding @xmath88 given by @xmath341 , @xmath342 , @xmath343 , @xmath344 , @xmath345 .
let @xmath346 , @xmath200 .
[ def : diagonal ] let @xmath347 be a matrix with entries in @xmath348 .
a braided vector space @xmath3 is of _ diagonal type _ with matrix
@xmath349 if @xmath0 has a basis @xmath350 with @xmath351 thus , every finite - dimensional @xmath352 is a braided vector space of diagonal type .
question [ item : method - nichols ] , more precisely ( a@xmath173 ) , has a complete answer in this setting .
first we can assume that @xmath353 for @xmath206 , as otherwise @xmath354 .
also , let @xmath355 and @xmath356 a braided vector space with matrix @xmath357 . if @xmath358 and @xmath359 for all @xmath360 , then @xmath361 as braided vector spaces .
[ th : classif - diagonal ] _
@xcite _ the classification of all braided vector spaces of diagonal type with finite - dimensional nichols algebra is known .
the proof relies on the weyl groupoid introduced in @xcite , a particular case of theorem [ th : grs - semisimple ] .
another fundamental ingredient is the following result , generalized at various levels in @xcite .
[ th : pbw ] _
@xcite _ let @xmath0 be a braided vector space of diagonal type .
every hopf algebra quotient of @xmath100 has a pbw basis .
the classification in theorem [ th : classif - diagonal ] can be organized as follows : @xmath92 for most of the matrices @xmath362 in the list of @xcite there is a field @xmath257 and a pair @xmath363 as in example [ exa : grs - lie - super ] such that @xmath364 , and @xmath365 has the same grs as the nichols algebra corresponding to @xmath349 @xcite .
@xmath92 besides these , there are 12 ( yet ) unidentified examples .
we believe that theorem [ th : classif - diagonal ] can be proved from theorem [ th : classif - grs ] , via example [ exa : grs - lie - super ] .
[ th : presentation - diagonal ] _
@xcite _ an efficient set of defining relations of each finite - dimensional nichols algebra of a braided vector space of diagonal type is known .
the proof uses most technical tools available in the theory of nichols algebras ; of interest in its own is the introduction of the notion of convex order in weyl groupoids . as for other classifications above
, it is not possible to state precisely the list of relations .
we just mention different types of relations that appear . *
quantum serre relations , i.e. , @xmath366 for suitable @xmath367 .
* powers of root vectors , i.e. , @xmath368 , where the @xmath369 s are part of the pbw basis . * more exotic relations
; they involve 2 , 3 , or at most 4 @xmath370 s in @xmath201 .
we now consider nichols algebras of objects in @xmath371 , where @xmath19 is a finite not necessarily abelian group .
the category @xmath371 is semisimple and the simple objects are parametrized by pairs @xmath372 , where @xmath373 is a conjugacy class in @xmath19 and @xmath374 , for a fixed @xmath375 ; the corresponding simple yetter - drinfeld module @xmath376 is @xmath377 as a module .
the braiding @xmath2 is described in terms of the conjugation in @xmath373 . to describe the related suitable class
, we recall that a rack is a set @xmath378 with a map @xmath379 satisfying * @xmath380 is a bijection for every @xmath381 . *
@xmath382 for all @xmath383 ( self - distributivity ) .
for instance , a conjugacy class @xmath373 in @xmath19 with the operation @xmath384 , @xmath385 is a rack ; actually we only consider racks realizable as conjugacy classes .
let @xmath15 be a rack and @xmath386 a decomposition of @xmath15 , i.e. , a disjoint family of subracks with @xmath387 for all @xmath388 .
[ def : braiding - rack]@xcite a _
2-cocycle of degree @xmath389 _ , associated to @xmath390 , is a family @xmath391 of maps @xmath392 such that @xmath393 given such @xmath349 , let @xmath394 and let @xmath395 be given by @xmath396 then @xmath397 is a braided vector space called of _ rack type _ ; its nichols algebra is denoted @xmath398 .
if @xmath399 , then we say that @xmath349 is _
principal_. every finite - dimensional @xmath113 is a braided vector space of rack type ( * ? ? ?
* theorem 4.14 ) .
question ( a@xmath173 ) in this setting has partial answers in three different lines : computation of some finite - dimensional nichols algebras , nichols algebras of reducible yetter - drinfeld modules and collapsing of racks .
the algorithm to compute a nichols algebra @xmath153 is as follows : compute the space @xmath400 of relations of degree @xmath370 , for @xmath401 ; then compute the @xmath402-th partial nichols algebra @xmath403 , say with a computer program .
if lucky enough to get @xmath404 , then check whether it is a nichols algebra , e.g. via skew - derivations ; otherwise go to @xmath405 .
the description of @xmath406 is not difficult @xcite but for higher degrees it turns out to be very complicated .
we list all known examples of finite - dimensional nichols algebras @xmath398 with @xmath15 indecomposable and @xmath349 principal and abelian ( @xmath407 ) .
[ exa : nichols - sim ] let @xmath408 be the conjugacy class of @xmath409-cycles in @xmath410 , @xmath411 .
we start with the rack of transpositions in @xmath410 and the cocycles @xmath412 , @xmath337 that arise from the @xmath413 with @xmath414 , see ( * ? ? ?
* ( 5.5 ) , ( 5.9 ) ) .
let @xmath0 be a vector space with basis @xmath415 and consider the relations @xmath416 the quadratic algebras @xmath417 and @xmath418 were considered in @xcite , @xcite respectively ; @xmath419 are named the _ fomin - kirillov algebras_. it is known that @xmath56 the nichols algebras @xmath420 and @xmath421 are twist - equivalent , hence have the same hilbert series .
ditto for the algebras @xmath422 and @xmath419 @xcite .
@xmath56 if @xmath423 , then @xmath424 and @xmath425 are finite - dimensional @xcite ( for @xmath426 part of this was done by graa ) . in fact @xmath427 but for @xmath428 , it is not known whether the nichols algebras @xmath420 and @xmath421 have finite dimension or are quadratic .
[ exa : nichols - cube ] @xcite the nichols algebra @xmath429 is quadratic , has the same hilbert series as @xmath430 and is generated by @xmath431 with defining relations @xmath432 [ exa : nichols - affines ] @xcite let @xmath274 be a finite abelian group and @xmath433 .
the _ affine _ rack @xmath434 is the set @xmath274 with product @xmath435 , @xmath436 .
let @xmath437 be a prime , @xmath438 a power of @xmath126 , @xmath439 and @xmath440 the multiplication by @xmath441 ; let @xmath442 .
assume that @xmath443 , 4 , 5 , or 7 , with @xmath444 , @xmath445 , 2 or 3 , respectively .
then @xmath446 , @xmath447 being the euler function , and @xmath448 , where @xmath449 is generated by @xmath450 with @xmath451 ; and @xmath452 is generated by @xmath453 and @xmath454 of course @xmath455 ; also @xmath456 . by duality
, we get @xmath457 [ exa : nichols - tetra]@xcite there is another finite - dimensional nichols algebra associated to @xmath458 with a cocycle @xmath349 with values @xmath459 , where @xmath460 .
concretely , @xmath461 and @xmath462 can be presented by generators @xmath463 with defining relations @xmath464 the ideas of example [ exa : grs - nichols - ss - yd ] in the context of decomposable yetter - drinfeld modules over groups were pushed further in a series of papers culminating with a remarkable classification result @xcite .
consider the groups @xmath465
@xmath56 @xcite let @xmath19 be a quotient of @xmath466 .
then there exist @xmath467 , @xmath468 such that @xmath469 and @xmath470 .
@xmath56 @xcite let @xmath19 be a quotient of @xmath471 .
then there exist @xmath472 , @xmath473 , @xmath474 , @xmath475 such that @xmath476 , @xmath477 , @xmath478 and @xmath479 , @xmath480 .
@xmath56 @xcite let @xmath19 be a quotient of @xmath481 .
then there exist @xmath482 , @xmath483 such that @xmath484 , @xmath485 and @xmath486 .
@xmath56 @xcite let @xmath19 be a quotient of @xmath150 .
then there exist @xmath487 , @xmath488 such that @xmath489 , @xmath490 and @xmath491 . _
let @xmath19 be a non - abelian group and @xmath492 such that @xmath19 is generated by the support of @xmath493 .
assume that @xmath494 and that @xmath495 .
then @xmath493 is one of @xmath496 , @xmath497 , above , and correspondingly @xmath19 is a quotient of either @xmath498 , @xmath499 , or @xmath150 .
implicit in question ( a@xmath173 ) in the setting of racks is the need to compute all non - principal 2-cocycles for a fixed rack @xmath15 .
notably , there exist criteria that dispense of this computation .
to state them and explain their significance , we need some terminology .
all racks below are finite .
@xmath56 a rack @xmath15 is _ abelian _ when @xmath500 , for all @xmath501 .
@xmath56 a rack is _
indecomposable _ when it is not a disjoint union of two proper subracks .
@xmath56 a rack @xmath15 with @xmath502 is _ simple _ when for any projection of racks @xmath503 , either @xmath158 is an isomorphism or @xmath504 has only one element . _
@xcite , @xcite _ every simple rack is isomorphic to one of : 1 .
affine racks @xmath505 , where @xmath126 is a prime , @xmath506 , and @xmath150 is the companion matrix of a monic irreducible polynomial @xmath507 $ ] of degree @xmath212 , @xmath508 .
[ item : twisted - conjugacy ] non - trivial ( twisted ) conjugacy classes in simple groups .
[ item : twisted - homogeneous ] twisted conjugacy classes of type @xmath509 , where @xmath510 , with @xmath122 a simple non - abelian group and @xmath511 ; and @xmath512 acts by @xmath513 , where @xmath514 .
[ def : rack - typed ] @xcite we say that a finite rack @xmath15 is _ of type d _ when there are a decomposable subrack @xmath515 , @xmath516 and @xmath517 such that @xmath518 . also , @xmath15 is _ of type f _
@xcite if there are a disjoint family of subracks @xmath519 and a family @xmath520 with @xmath521 , such that @xmath522 , @xmath523 , for all @xmath524 .
an indecomposable rack @xmath15 _ collapses _ when @xmath525 for every finite _ faithful _
2-cocycle @xmath349 ( see @xcite for the definition of faithful ) .
[ th : type4 ] _
@xcite ; @xcite _ if a rack is of type d or f , then it collapses . the proofs use results on nichols algebras from @xcite .
if a rack projects onto a rack of type d ( or f ) , then it is also of type d ( or f ) , hence it collapses by theorem [ th : type4 ] .
since every indecomposable rack @xmath15 , @xmath502 , projects onto a simple rack , it is natural to ask for the determination of all _ simple _ racks of type d or f. a rack is _ cthulhu _ if it is neither of type d nor f ; it is _ sober _ if every subrack is either abelian or indecomposable @xcite .
sober implies cthulhu .
@xmath56 let @xmath526 .
let @xmath373 be either @xmath527 , if @xmath528 , or else @xmath529 if @xmath530 . the type of @xmath80 is formed by the lengths of the cycles in its decomposition .
@xmath92 @xcite if the type of @xmath80 is @xmath531 , @xmath532 , @xmath533 , @xmath534 , @xmath535 , @xmath536 , @xmath537 , or @xmath538 , then @xmath373 is cthulhu . if the type of @xmath80 is @xmath539 , then @xmath373 is sober .
@xmath92 @xcite let @xmath437 be a prime .
assume the type of @xmath80 is @xmath540 .
if @xmath541 or not of the form @xmath542 , @xmath128 a prime power , then @xmath373 is sober ; otherwise @xmath373 is of type d. assume the type of @xmath80 is @xmath543 .
if @xmath544 or not of the form @xmath542 , @xmath128 a prime power , then @xmath373 is sober ; otherwise @xmath373 is of type d. @xmath92 @xcite for all other types , @xmath373 is of type d , hence it collapses .
@xmath56 @xcite let @xmath545 and @xmath127 be a prime power .
let @xmath546 not semisimple and @xmath547 .
the type of a unipotent element are the sizes of its jordan blocks .
@xmath92 assume @xmath548 is unipotent .
if @xmath548 is either of type @xmath549 and @xmath127 is even or not a square , or of type @xmath550 and @xmath551 , then @xmath373 is sober .
if @xmath548 is either of type @xmath552 and @xmath127 is even , or of type @xmath553 and @xmath554 then @xmath373 is cthulhu .
if @xmath548 is of type @xmath555 and @xmath556 is even , then @xmath373 is not of type d , but it is open if it is of type f. @xmath92 otherwise , @xmath373 is either of type d or of type f , hence it collapses .
@xmath56 @xcite let @xmath373 be a conjugacy class in a sporadic simple group @xmath19 .
if @xmath373 appears in table [ tab : notd ] , then @xmath373 is not of type d. if @xmath557 is the monster and @xmath373 is one of , then it is open whether @xmath373 is of type d. otherwise , @xmath373 is of type d. @xmath150 & & @xmath558 & + @xmath559 & & @xmath560 & + @xmath561 & & @xmath562 & + @xmath563 & & @xmath564 & + @xmath565 & & @xmath566 & + @xmath567 & & @xmath568 & + @xmath569 & & @xmath570 & + @xmath571 & & @xmath572 & + @xmath573 & & @xmath574 & + @xmath575 & & @xmath576 & + @xmath577 & & @xmath578 & + @xmath579 & + here is the scheme of proof proposed in @xcite to attack conjecture [ conj : gendegone ] : let @xmath150 be a finite - dimensional graded hopf algebra in @xmath161 with @xmath148 and generated as algebra by @xmath177 .
we have a commutative diagram of hopf algebra maps @xmath580 & & { { \mathfrak b}}(v ) \\ & t(v)\ar@{->}[ul]^p \ar@{->}[ur ] & } $ ] . to show that @xmath158 is injective , take a generator @xmath128 ( or a family of generators ) of @xmath145 such that @xmath581 and consider the yetter - drinfeld submodule @xmath582 of @xmath100
; if @xmath583 , then @xmath584 . then @xmath126 factorizes through @xmath585 , where @xmath586 is the ideal generated by primitive generators of @xmath145 , and so on .
the conjecture has been verified in all known examples in characteristic 0 ( it is false in positive characteristic or for infinite - dimensional hopf algebras ) . a finite - dimensional pointed hopf algebra @xmath5 is generated by group - like and skew - primitive elements
if either of the following holds : @xmath92 _ @xcite _ the infinitesimal braiding is of diagonal type , e. g. @xmath587 is abelian .
@xmath92 _ @xcite . _
the infinitesimal braiding of @xmath5 is any of @xmath588 , @xmath589 ( @xmath590 ) , @xmath591 , @xmath592 , @xmath593 , @xmath594 , @xmath595 .
we address here question [ item : method - lifting ] in [ subsec : chevalley ] .
let @xmath15 be a finite rack and @xmath596 a 2-cocycle .
a hopf algebra @xmath5 is a _ lifting _ of @xmath597 if @xmath61 is a hopf subalgebra , @xmath5 is generated by @xmath179 and its infinitesimal braiding is a realization of @xmath598 .
see @xcite for liftings in the setting of copointed hopf algebras .
we start discussing realizations of braided vector spaces as yetter - drinfeld modules .
let @xmath346 and @xmath599 .
first , a _
yd - datum of diagonal type _ is a collection @xmath600 where @xmath601 , @xmath602 , @xmath603 ; @xmath19 is a finite group ; @xmath604 ; @xmath605 , @xmath206 ; such that @xmath606 , @xmath603 .
let @xmath3 be the braided vector space of diagonal type with matrix @xmath607 in the basis @xmath608
. then @xmath113 by declaring @xmath609 , @xmath206 .
more generally , a _
yd - datum of rack type _
@xcite is a collection @xmath610 where @xmath15 is a finite rack ; @xmath596 is a @xmath611-cocycle ; @xmath19 is a finite group ; @xmath612 is an action of @xmath19 on @xmath15 ; @xmath613 is equivariant with respect to the conjugation in @xmath19 ; and @xmath614 is a family of 1-cocycles @xmath615 ( that is , @xmath616 , for all @xmath617 , @xmath618 ) such that @xmath619 and @xmath620 for all @xmath621 .
let @xmath622 be the associated braided vector space .
then @xmath0 becomes an object in @xmath371 by @xmath623 and @xmath624 , @xmath625 , @xmath617 .
second , let @xmath626 be a yd - datum of either diagonal or rack type and @xmath627 as above ; let @xmath628 .
the desired liftings are quotients of @xmath629 ; write @xmath630 in these quotients instead of @xmath631 to distinguish them from the elements in @xmath632 . let @xmath241 be a minimal set of generators of @xmath145 , assumed homogeneous both for the @xmath633- and the @xmath19-grading
. roughly speaking , the deformations will be defined by replacing the relations @xmath634 by @xmath635 , @xmath636 , where @xmath637 belongs to a lower term of the coradical filtration , and the ideal @xmath638 generated by @xmath639 , @xmath636 , is a hopf ideal .
the problem is to describe the @xmath639 s and to check that @xmath640 has the right dimension .
if @xmath581 has @xmath19-degree @xmath440 , then @xmath641 for some @xmath642 ; depending on the action of @xmath19 on @xmath128 , it may happen that @xmath643 should be 0 . in some cases ,
all @xmath636 are primitive , so all deformations can be described ; see @xcite for quantum linear spaces ( their liftings can also be presented as ore extensions @xcite ) and the examples [ exa : lifting - transpositions - s3 ] and [ exa : lifting - transpositions - s4 ] .
but in most cases , not all @xmath636 are primitive and some recursive construction of the deformations is needed .
this was achieved in @xcite for diagonal braidings of cartan type @xmath644 , with explicit formulae , and in @xcite for diagonal braidings of finite cartan type , with recursive formulae .
later it was observed that the so obtained liftings are cocycle deformations of @xmath632 , see e.g. @xcite .
this led to the strategy in @xcite : pick an adapted stratification @xmath645 @xcite ; then construct recursively the deformations of @xmath646 by determining the cleft extensions of the deformations in the previous step and applying the theory of hopf bi - galois extensions @xcite . in the examples below , @xmath647 for all @xmath617 by (
* ( d ) ) .
[ exa : lifting - transpositions - s3 ] @xcite let @xmath648 be a yd - datum .
let @xmath649 be such that @xmath650 let @xmath651 be the quotient of @xmath629 by the relations @xmath652 then @xmath653 is a pointed hopf algebra , a cocycle deformation of @xmath654 and @xmath655 ; @xmath656 iff @xmath657 for some @xmath658 .
conversely , any lifting of @xmath659 is isomorphic to @xmath660 for some yd - datum @xmath648 and @xmath649 satisfying , .
[ exa : lifting - transpositions - s4 ] @xcite let @xmath661 be a yd - datum .
let @xmath662 be such that @xmath663 let @xmath651 be the quotient of @xmath629 by the relations @xmath664 then @xmath653 is a pointed hopf algebra , a cocycle deformation of @xmath665 and @xmath666 ; @xmath656 iff @xmath657 for some @xmath658 .
conversely , any lifting of @xmath667 is isomorphic to @xmath660 for some yd - datum @xmath661 and @xmath662 satisfying , .
[ exa : lifting - affine - f4 ] @xcite let @xmath668 be a yd - datum .
let @xmath662 be such that @xmath669 let @xmath651 be the quotient of @xmath629 by the relations @xmath670 @xmath671 then @xmath653 is a pointed hopf algebra , a cocycle deformation of @xmath654 and @xmath672 ; @xmath656 iff @xmath657 for some @xmath658 .
conversely , any lifting of @xmath673 is isomorphic to @xmath660 for some yd - datum @xmath668 and @xmath662 satisfying , .
[ exa : lifting - affine - f5 ] @xcite let @xmath674 be a yd - datum .
let @xmath662 be such that @xmath675 let @xmath651 be the quotient of @xmath629 by the relations @xmath676 @xmath677 .
then @xmath653 is a pointed hopf algebra , a cocycle deformation of @xmath654 and @xmath678 ; @xmath656 iff @xmath657 for some @xmath658 .
conversely , any lifting of @xmath679 is isomorphic to @xmath660 for some yd - datum @xmath674 and @xmath662 satisfying , .
[ exa : lifting - affine - f5 - 3 ] @xcite let @xmath680 be a yd - datum .
let @xmath662 be such that @xmath681 let @xmath651 be the quotient of @xmath629 by the relations @xmath682 where @xmath683 .
then @xmath653 is a pointed hopf algebra , a cocycle deformation of @xmath654 and @xmath678 ; @xmath656 iff @xmath657 for some @xmath658 .
conversely , any lifting of @xmath684 is isomorphic to @xmath660 for some yd - datum @xmath680 and @xmath662 satisfying , .
here is a classification from @xcite . let @xmath685 be a yd - datum of diagonal type as in with @xmath175 a finite abelian group and let @xmath352 be the corresponding realization .
we say that @xmath626 is a _
cartan datum _ if there is a cartan matrix ( of finite type ) @xmath686 such that @xmath687 , @xmath688 .
let @xmath689 be the root system associated to @xmath690 , @xmath691 a choice of simple roots , @xmath692 the set of connected components of the dynkin diagram of @xmath689 and set @xmath693 whenever @xmath694 belong to the same @xmath695 .
we consider two classes of parameters : * @xmath696 is a family in @xmath697 with @xmath698 when @xmath699 or @xmath700 .
* @xmath701 is a family in @xmath14 with @xmath702 when @xmath703 , @xmath704 , and @xmath705 or @xmath706 . here
@xmath707 for an arbitrary @xmath708 .
we attach a family ( @xmath709 in @xmath710 to the parameter @xmath711 , defined recursively on the length of @xmath712 , starting by @xmath713 . from all these data
we define a hopf algebra @xmath714 as the quotient of @xmath715 by the relations @xmath716 the hopf algebra @xmath717 is pointed , @xmath718 and @xmath719 .
let @xmath5 be a pointed finite - dimensional hopf algebra and set @xmath720 .
assume that the prime divisors of @xmath721 are @xmath722 .
then there exists a cartan datum @xmath626 and parameters @xmath643 and @xmath711 such that @xmath723 .
it is known when two hopf algebras @xmath717 and @xmath724 are isomorphic .
the proof offered in @xcite relies on @xcite .
some comments : the hypothesis on @xmath725 forces the infinitesimal braiding @xmath0 of @xmath5 to be of cartan type , and the relations of @xmath153 to be just quantum serre and powers of root vectors .
the quantum serre relations are not deformed in the liftings , except those linking different components of the dynkin diagram ; the powers of the root vectors are deformed to the @xmath726 that belong to the coradical .
all this can fail without the hypothesis , see @xcite for examples in rank 2 .
we present some classification results of pointed hopf algebras with non - abelian group .
we say that a finite group @xmath19 _ collapses _ whenever any finite - dimensional pointed hopf algebra @xmath5 with @xmath727 is isomorphic to @xmath45 .
@xmath67 @xcite let @xmath729 and let @xmath626 be the corresponding yd - datum .
let @xmath5 be a finite - dimensional pointed hopf algebra with @xmath730
. then @xmath5 is isomorphic either to @xmath731 , or to @xmath732 , or to @xmath733 , cf .
example [ exa : lifting - transpositions - s3 ] .
@xmath67 @xcite let @xmath734 be a finite - dimensional pointed hopf algebra with @xmath735 .
let @xmath736 , @xmath737 , @xmath738 , with corresponding data @xmath739 , @xmath740 and @xmath741 .
then @xmath5 is isomorphic to one of @xmath742 here @xmath743 is as in example [ exa : lifting - transpositions - s4 ] ; @xmath744 is the quotient of @xmath745 by the relations @xmath746 , @xmath747 , @xmath748 and @xmath749 is the quotient of @xmath750 by the relations @xmath751 clearly @xmath752 , @xmath753 , @xmath754 .
also @xmath755 iff @xmath756 for some @xmath757 , and @xmath758 iff @xmath759 for some @xmath757 .
@xmath67 @xcite let @xmath5 be a finite - dimensional pointed hopf algebra with @xmath760 , but @xmath761 .
it is not known whether @xmath762 .
let @xmath739 , @xmath740 be the data corresponding to @xmath763 , @xmath764 . if the infinitesimal braiding of @xmath5 is not @xmath765 , then @xmath5 is isomorphic to one of @xmath743 , @xmath766 ( defined as in example [ exa : lifting - transpositions - s4 ] ) , or @xmath767 , or @xmath768 ( defined as above ) .
@xmath67 @xcite let @xmath769 .
let @xmath770 be a finite - dimensional pointed hopf algebra with @xmath771 .
then the infinitesimal braiding of @xmath5 is @xmath772 , where the type of @xmath80 is @xmath773 and @xmath774 , @xmath775 or @xmath776 ; it is an open question whether @xmath777 , see example [ exa : nichols - sim ] . if @xmath778 , there are two more nichols algebras with unknown dimension corresponding to the class of type @xmath537 , but they are conjugated to those of type @xmath779 by the outer automorphism of @xmath780 .
@xmath67 @xcite let @xmath781 , @xmath782 with @xmath783 .
let @xmath784 be the dihedral group of order @xmath785 .
then there are infinitely many finite - dimensional nichols algebras in @xmath371 ; all of them are exterior algebras as braided hopf algebras .
let @xmath5 be a finite - dimensional pointed hopf algebra with @xmath786 , but @xmath787 .
then @xmath5 is a lifting of an exterior algebra , and there infinitely many such liftings .
we say that a semisimple hopf algebra @xmath75 _ collapses _ if any finite - dimensional hopf algebra @xmath5 with @xmath788 is isomorphic to @xmath75 .
thus , if @xmath19 collapses , then @xmath46 and @xmath789 collapse , for any twist @xmath73 .
next we state the classification of the finite - dimensional copointed hopf algebras over @xmath790 @xcite .
let @xmath729 as a yetter - drinfeld module over @xmath791 .
let @xmath792 be such that @xmath793 .
let @xmath794 be the quotient of @xmath795 by the relations @xmath796 , @xmath797 , @xmath798 , for @xmath799 .
then @xmath800 is a hopf algebra of dimension 72 and @xmath801 .
any finite - dimensional copointed hopf algebra @xmath5 with @xmath802 is isomorphic to @xmath803 for some @xmath643 as above ; @xmath804 iff @xmath643 and @xmath805 are conjugated under @xmath806 . | this is a survey on the state - of - the - art of the classification of finite - dimensional complex hopf algebras .
this general question is addressed through the consideration of different classes of such hopf algebras .
pointed hopf algebras constitute the class best understood ; the classification of those with abelian group is expected to be completed soon and there is substantial progress in the non - abelian case .
_ dedicado a biblioco 34 _ 16t05 , 16t20 , 17b37 , 16t25 , 20g42 .
hopf algebras , quantum groups , nichols algebras . |
it has been proven , beginning in the early 1960 s @xcite , that there always exist states with negative energy density in quantum field theory .
some specific examples include the casimir effect @xcite and squeezed states @xcite , both of which have been experimentally realized .
( although the energy density itself is far too small to be directly measured . )
negative energy is also required for black hole evaporation , and hence for the consistency of the laws of black hole physics with those of thermodynamics . on the other hand
, unrestricted amounts of negative energy could produce bizarre effects , for example , violations of the second law of thermodynamics @xcite .
however , the same laws of quantum field theory which allow the existence of negative energy also appear to severely restrict its magnitude and duration in such a way as to prevent gross large - scale effects .
these bounds are known as quantum inequalities , and quite a large body of work now exists on the subject .
for some recent reviews of quantum inequalities , see refs .
quantum inequality bounds have been proven , for example , for the minimally coupled scalar , electromagnetic , and dirac fields .
it should be pointed out that the potential macroscopic problems arise not because of the existence of negative energy per se , but from the arbitrary separation of negative and positive energy .
it is this behavior which the quantum inequalities prohibit .
many possible configurations of separated negative and positive energy can easily be ruled out , and known permitted examples involve the subtle intertwining of the two @xcite .
whether the currently known examples are representative of the general case is unknown .
hence , the study of further examples could prove useful .
since the negative energy densities in these states are too small to be directly measurable , experiments in quantum optics may offer the best possibilities for indirect detection of negative energy .
( however , see also refs . @xcite . ) a first link between quantum optics and the work on quantum inequalities has been forged in a recent paper by marecki @xcite . for squeezed states , he proved quantum inequality - type bounds on the magnitude and duration of the squeezing .
quantum optics has seen enormous experimental and theoretical advances in the last twenty years .
this marriage of optics with quantum field theory has resulted in experiments which were formerly purely `` gedanken '' becoming those which are now routinely performed in the laboratory .
highly non - classical states , such as schrdinger `` cat states '' and squeezed states , have been produced and play a part in everything from quantum computers to noise reduction in laser interferometer gravitational wave detectors .
the `` cat states '' of the electromagnetic field are superpositions of coherent states and have been created experimentally @xcite .
the experiments which have been done so far have produced mesoscopic superpositions , in which the mean photon number is of order @xmath0 .
this is somewhat short of a true schrdinger cat state , which would be a superposition of two or more classical configurations , that is , coherent states with very large occupation numbers .
more recently , there have been proposals for methods of creating superpositions of squeezed vacuum states @xcite .
an interesting question arises : can one start with two quantum states which do not involve negative energy and by superposing them obtain negative energy ?
the answer is yes ; the classic standard example being the vacuum @xmath1 two - particle state ( for a nice discussion see ref .
@xcite ) . is this true for the superposition of other states as well ?
more generally , what effects does superposition have on negative energy
? could one also go the other way , i.e. , start with two states involving negative energy and by superposing them diminish or eradicate the negative energy ? in this paper , we will address such questions for several classes of states . in sect .
[ sec : rho ] , we develop some formalism for parameterizing the maximum magnitude of negative energy that can occur for states of a minimally coupled scalar field in minkowski spacetime with either one or two modes excited . we give several examples of superpositions for a single mode in sect . [ sec : super ] , including superpositions of two coherent states , two squeezed vacuum states , and a coherent state with a squeezed vacuum state . in sect .
[ sec : entangled ] , we move to the two - mode case .
this allows us to consider examples of entangled states involving either squeezed vacua or coherent states for the two modes .
a summary of our conclusions is presented in section [ sec : sum ] .
in this paper , we will consider a massless scalar field in flat spacetime , for which the stress tensor operator is @xmath2 the normal - ordered energy density operator is @xmath3 \,,\ ] ] where @xmath4 with the @xmath5 being the mode functions .
we wish to consider the case where all modes except for two are in the vacuum state . for the first mode ,
let @xmath6 , @xmath7 , and @xmath8 be the mode function , annihilation operator , and creation operator , respectively , and let @xmath9 , @xmath10 , and @xmath11 be the corresponding quantities for the second mode .
the expectation value of the energy density in an arbitrary quantum state can be expressed as @xmath12 + \langle b^\dagger b \rangle\ , ( |{\dot f_2}|^2 + |{\nabla f_2}|^2 ) \nonumber \\ & + & \langle b^2 \rangle\ , [ { \dot f_2}^2 + ( { \nabla f_2})^2 ] + 2 \langle a^\dagger b \rangle\ , ( { \dot f_1}^ * { \dot f_2 } + { \nabla f_1}^ * \cdot { \nabla f_2 } ) + 2 \langle a b \rangle\ , ( { \dot f_1 } { \dot f_2 } + { \nabla f_1 } \cdot { \nabla f_2 } ) \biggr\ } \ , .
\label{eq : rho}\end{aligned}\ ] ] let @xmath13 all of the information needed to give the two - mode energy density , eq .
( [ eq : rho ] ) , at a given quantum state is encoded in the above set of six amplitudes and four phases . in the case of a traveling waves
, we may take the mode functions to be @xmath14 where @xmath15 , for @xmath16 and @xmath17 is a normalization volume . in this case
, the mean energy density may be expressed as @xmath18 + r_2\ , \omega_2\ , \cos[2(\mathbf{k}_2 \cdot \mathbf{x } - \omega_2 t)+\gamma_2 ] \nonumber \\ & + & r_3\ , \sqrt{\omega_1 \omega_2}\ , ( 1 + \mathbf{\hat{k}}_1 \cdot \mathbf{\hat{k}}_2)\ , \cos[(\mathbf{k}_2-\mathbf{k}_1 ) \cdot \mathbf{x } -(\omega_2-\omega_1)t + \gamma_3 ] \nonumber \\ & + & r_4\ , \sqrt{\omega_1 \omega_2}\ , ( 1 + \mathbf{\hat{k}}_1 \cdot \mathbf{\hat{k}}_2)\ , \cos[(\mathbf{k}_2+\mathbf{k}_1 ) \cdot \mathbf{x } -(\omega_2+\omega_1)t + \gamma_4 ] \biggr\ } \ , .
\label{eq:2mode_rho}\end{aligned}\ ] ] we will also consider the case of a standing wave which depends upon only one space coordinate , in which case the mode functions can be taken to be @xmath19 the energy density now becomes @xmath20 \
, \cos[(\omega_2-\omega_1)t -\gamma_3 ] \nonumber \\ & + & 2r_4\ , \sqrt{\omega_1 \omega_2}\ , \cos[(\omega_1+\omega_2)x ] \ , \cos[(\omega_1+\omega_2)t -\gamma_4 ] \biggr\ } \ , .
\label{eq:2mode_rho_stand}\end{aligned}\ ] ] a useful special case is when only one mode is excited . in this case , we may set @xmath21 , @xmath22 , @xmath23 , and @xmath24 . in this case , we need only the three real numbers @xmath25 , @xmath26 , and @xmath27 to determine the energy density in a given state . for the case of a traveling wave
, we have @xmath28 \right\}\ , \label{eq : rho1}\ ] ] we can see from eq .
( [ eq : rho1 ] ) that the minimum value of @xmath29 is @xmath30 and hence we can have negative energy density only if @xmath31 . in the case of a standing wave , eq .
( [ eq:2mode_rho_stand ] ) becomes @xmath32\ , .
\label{eq : rho2}\ ] ] again , the minimum value of @xmath29 is given by eq .
( [ eq : rho_min ] ) .
in this section , we examine some explicit examples of superpositions involving a single mode . in each case , we need only calculate the quantity @xmath33 to determine the maximum magnitude of the negative energy .
first we consider a superposition of coherent states .
coherent states are eigenstates of the annihilation operator , that is @xmath34 let @xmath35 \ , , \label{eq : scs}\ ] ] where @xmath36 and @xmath37 are two different coherent states for the same mode , @xmath38 is a complex number , and @xmath39 is a normalization factor ( see , for example , sec .
7.6 of ref .
we also assume that the states are normalized so that @xmath40 . as a result
we have that @xmath41 \,.\ ] ] the coherent states are not orthonormal ; their overlap integral is given by ( see for example , eq.(3.6.24 ) of ref .
@xcite ) : @xmath42 therefore the square of the normalization factor is @xmath43}^{-1 } \,.\ ] ] the mean number of particles is found to be @xmath44\,,\ ] ] and @xmath45\,.\ ] ] let @xmath46 then the quantities @xmath25 , @xmath26 , and @xmath27 are functions of six real parameters , the magnitudes and phases of @xmath47 , @xmath48 , and @xmath38 .
however , one finds that only @xmath27 depends upon all six . the magnitudes @xmath25 and @xmath26 depend only upon the difference @xmath49 , and are hence functions of five parameters .
we are primarily interested in the quantity @xmath33 , which measures the maximum magnitude of the negative energy density . hence set @xmath50 and write @xmath51 and let @xmath52 be a five - dimensional vector given by @xmath53 one may use eqs .
( [ eq : f ] ) and ( [ eq : g ] ) as the basis of a numerical algorithm to search for points of maximum @xmath54 and hence maximally negative energy density .
start at a random point in the five - dimensional parameter space , and compute @xmath54 and @xmath52 . if @xmath55 , then this choice of parameters is a quantum state with negative energy density .
the components of @xmath52 indicate the direction in which @xmath54 is increasing most rapidly .
one then moves along this direction until a local maximum of @xmath54 is located .
a preliminary , non - exhaustive , search located two such local maxima , at @xmath56 and at @xmath57 .
( one can trivially generate a third maximum by interchange of @xmath47 and @xmath48 in the latter case . )
the first example corresponds to @xmath58 and the second to a superposition of a coherent state and the vacuum .
interestingly , the maximum magnitude of the negative energy density is about the same in both examples , with @xmath59 , and hence @xmath60 .
we do not have an explanation as to why these two choices give the same value of @xmath33 .
the mean particle number is @xmath61 in the first example and @xmath62 in the second .
this example illustrates that a superposition of two coherent states can produce negative energy density , and the maximum negative energy density arises for mean particle number of order one .
we begin with a review of the features of the expectation value of the energy density in a single squeezed vacuum state .
our state is given by : @xmath63 where @xmath64 is the squeeze parameter and @xmath65 is a phase parameter .
the squeeze operator @xmath66 is given by @xmath67 } \,.\ ] ] this operator is unitary since @xmath68 the single - mode squeezed state @xmath69 is produced by the squeeze operator acting on the vacuum state @xmath70 the state @xmath69 can be written , after some work ( see eq .
( 3.7.5 ) of ref .
@xcite ) , in terms of the even fock states as @xmath71}^n \,|2n \rangle \ , .
\label{eq : xi}\ ] ] we also have that @xmath72 in the state @xmath69 we have the expectation values @xmath73 where we have made use of eqs .
( [ eq : sdagas ] ) .
thus @xmath74 and @xmath75 attains its maximum value of @xmath76 as @xmath77 .
we now calculate the energy density in a superposition of two single - mode squeezed vacuum states of the form @xmath78 \ , , \label{eq : esvs}\ ] ] where , for simplicity , we will choose @xmath79 and set @xmath80 in this state we have @xmath81 \nonumber \\ & = & n^2 \biggl[{\rm sinh}^2 r ( 1+{|\eta|}^2 ) - 2 |\eta|\ , \cos \theta \ , \frac{{\rm sech}r \ , { \rm tanh}^2 r}{{(1+{\rm tanh}^2 r)}^{3/2 } } \biggr ] \ , , \label{eq : n_2sq}\end{aligned}\ ] ] where we have made use of eq .
( [ eq : ada ] ) in the appendix . a similar calculation , using eq . ( [ eq : a2 ] ) , yields @xmath82 \,,\ ] ] and @xmath83^\frac{1}{2 } \ , .
\label{eq : r_2sq}\ ] ] the normalization of our state is given by @xmath84 \,.\ ] ] the square of the normalization factor is then @xmath85}^{-1 } \
, , \label{eq : nsquared}\ ] ] where we have used eq .
( [ eq : sech ] ) of the appendix . from eqs .
( [ eq : n_2sq ] ) and ( [ eq : r_2sq ] ) , we can compute the quantity @xmath33 , which gives the maximum magnitude of the negative energy density , as function of @xmath86 and @xmath64 . for fixed @xmath86 ,
one typically finds that @xmath33 attains a maximum value for some value of @xmath64 , usually of order one . a typical case of @xmath87 is illustrated in fig .
[ fig:2sq ] .
the case @xmath88 is just the single squeezed vacuum state discussed in sect .
[ sec : single_sq ] .
this case gives the maximum negative energy density , @xmath89 , for large @xmath64 .
all other values of @xmath38 , corresponding to superposed squeezed vacua , give somewhat smaller amounts of negative energy density , and attain their maximum negative energy density at finite values of @xmath64 . in this subsection
, we consider states of the form @xmath90\ , , \label{eq : csv}\ ] ] where @xmath69 is a squeezed vacuum state , and @xmath91 is a coherent state .
we may use eq .
( [ eq : in_prod ] ) to find @xmath92\right\}^{-1 } \,.\ ] ] similarly , we find @xmath93\biggr\}\end{aligned}\ ] ] and @xmath94\ , { \rm e}^{i \delta}\ , \tanh r \ ; \exp[-\frac{1}{2}\ , { \rm e}^{i \delta}\,(\alpha^*)^2\ , \tanh r ] \biggr\}\,,\end{aligned}\ ] ] using eqs .
( [ eq : ada2 ] ) and ( [ eq : ad2 ] ) .
let us consider the case where @xmath95 , @xmath96 , and @xmath47 is real . in this case , @xmath33
is plotted in fig .
[ fig : scs ] for various values of @xmath47 . here
the maximum negative energy density , @xmath97 is attained for large @xmath64 . note that this state has less negative energy than does the squeezed vacuum by itself .
[ see eq .
( [ eq : rnsqvac ] ) . ] for @xmath47 non - zero , we find that @xmath33 initially decreases , reaches a minimum value , and then increases again . for the case @xmath98 , for example , there is negative energy for @xmath99 and again for @xmath100 , but not for intermediate values of @xmath64 .
note that @xmath101 is a superposition of the vacuum and a coherent state , a special case of the state treated in sect .
[ sec : scs ] .
a limit of special interest is when @xmath102 and we have the superposition of the vacuum with a squeezed vacuum state . in this case ,
@xmath103^{-1}\,,\ ] ] @xmath104 and @xmath105\,.\ ] ] the @xmath106 curve in fig .
[ fig : scs ] is this limit for @xmath107 .
if @xmath38 is real and negative , @xmath108 , we then have @xmath109 and @xmath110 in the case that @xmath111 the right - hand side of eq .
( [ eq : r - n_svv ] ) is plotted as a function of @xmath64 in fig .
[ fig : sv_v ] . here
we find the maximum negative energy , @xmath112 at @xmath113 .
this is slightly less negative energy than can be found in a single squeezed vacuum state .
note that as @xmath114 , this state becomes @xmath115 , a two - particle state with positive energy density everywhere .
this is the reason that the behavior in fig .
[ fig : sv_v ] differs from the @xmath102 curve in fig .
[ fig : scs ] . in the latter case , @xmath116 , and there is negative energy for all values of @xmath64 .
in this section , we will consider several examples of entangled states involving two modes .
our first example of a two - mode entangled squeezed state was described by barnett and radmore @xcite and is defined by @xmath117 where @xmath118 is the vacuum state for both modes , and @xmath119 is a two - mode squeeze operator . if one were to expand the state @xmath120 in terms of number eigenstates , the expansion would contain states with an even total number of particles , with half of these particles in each mode .
one has the following identities @xcite @xmath121 note that @xmath122 one may use these relations to show that @xmath123 the minimum energy density in this state is @xmath124\,.\ ] ] this is never more negative than the minimum energy density that would be obtained if the two modes were individually in squeezed vacuum states .
the latter energy density is @xmath125 , where @xmath33 is given by eq .
( [ eq : rnsqvac ] ) .
thus we can write @xmath126 which is always non - negative and approached zero only when the two modes have nearly the same frequency .
in this subsection , we will consider a second possibility for an entangled two - mode squeezed state , which was discussed by zhang @xcite .
this state is defined by @xmath127 where @xmath128 and @xmath129 are single - mode squeezed vacuum states for mode @xmath7 , and @xmath130 and @xmath131 are such states for mode @xmath10 . in general , @xmath132 , @xmath133 , @xmath38 , and @xmath134 can be four arbitrary complex parameters .
however , we will restrict our attention to the case where they are real and satisfy @xmath135 in this case , @xmath136\right\}^{-\frac{1}{2 } } = \left [ 2\,(1+\cos \theta \,{\rm sech}\ , 2r ) \right]^{-\frac{1}{2 } } \ , , \label{eq : zhangn}\ ] ] where we have used eq .
( [ eq : sech ] ) for each of the two modes .
similarly , we find @xmath137 \ , , \label{eq : zhangn}\ ] ] and @xmath138 here we have use eqs .
( [ eq : ada ] ) and ( [ eq : a2 ] ) , as well as the identity @xmath139 .
in addition , we find @xmath140 and @xmath141 . in this case , the energy density , eq .
( [ eq:2mode_rho ] ) , becomes @xmath142 + \omega_2\ , \cos[2 ( \mathbf{k}_2 \cdot \mathbf{x } - \omega_2 t)+\gamma_1 ] \right\ } \biggr)\ ] ] we can always choose the spatial position @xmath143 and time @xmath144 so as to make both cosine functions equal to @xmath145 , in which case we achieve the minimum allowed energy density in this state of @xmath146 from eqs .
( [ eq : zhangn ] ) , ( [ eq : zhangn ] ) and ( [ eq : zhangr ] ) , we find @xmath147 as a function of @xmath86 and @xmath64
. in general , the behavior of this entangled state is similar to that of the superposed squeezed vacua illustrated in fig .
[ fig:2sq ] .
however , there is one limit of particular interest , which is when @xmath148 and @xmath149 .
( note that if @xmath150 , then @xmath151 , and there is no negative energy . )
if we take the limit @xmath148 , for fixed @xmath152 , then we find the asymptotic forms @xmath153 and @xmath154 in the case that @xmath149 , the coefficient in the expression for @xmath155 can be large , so we can have an unusually large particle number in relation to the value of @xmath64 . the quantities @xmath156 and @xmath155
are plotted in fig .
[ fig : zhang ] for two values of @xmath86 close to @xmath157 . in this case
, we can have a reasonable amount of negative energy at very small values of the squeeze parameter , @xmath64 . in this subsection
, we consider a state of the same form as that in eq .
( [ eq : zhang ] ) , but involving entangled coherent states for two modes , which was also discussed by zhang @xcite .
let @xmath158 where @xmath159 , _ etc _ are single - mode coherent states .
we will restrict our attention to the case where the magnitudes of the four complex coherent state parameters are all equal , and @xmath160 and @xmath161 .
thus @xmath162 and @xmath163 here @xmath164 are the phases of @xmath165 , respectively . in this case
, we find @xmath166^{-\frac{1}{2 } } \,,\ ] ] and @xmath167 as well as @xmath168 , @xmath169 , @xmath170 , and @xmath171 .
let @xmath172 and @xmath173 .
we then set @xmath174 which can always be done by a suitable choice of @xmath143 and @xmath144 .
the energy density for a two - mode traveling wave state , eq .
( [ eq:2mode_rho ] ) now becomes @xmath175\,.\ ] ] if the two modes are close in wavenumber , so that @xmath176 and @xmath177 , and we set @xmath178 then @xmath179 where @xmath180 the function @xmath181 is plotted in fig .
[ fig : en_coherent ] , where we see that it attains a maximum value of about @xmath182 at @xmath183 .
this corresponds to a negative energy density of @xmath184 , which is about three times as negative as the maximally negative energy density found in the superposed coherent states discussed in sect .
[ sec : scs ] .
in this paper we have developed a formalism for parameterizing the energy density in states of a massless scalar field in which either one or two modes are excited .
we found explicit expressions for the energy density for the cases of traveling waves and of standing waves in one spatial direction . in all cases , the maximum negative energy density which can be achieved in a given state
can be expressed in terms of our parameters .
we next applied this approach to find the maximum negative energy density in several states , including some states which are of current interest in quantum optics . for the case of a single mode
, we considered three possible superposition states : ( 1 ) two coherent states , ( 2 ) two squeezed vacuum states , and ( 3 ) a coherent state and a squeezed vacuum state .
the superposition of two coherent states can be described as a schrdinger `` cat state '' in the sense that it would be a superposition of two classical configurations in the limit of large coherent state parameter . here
we find that the maximal negative energy density is achieved with mean photon numbers slightly less than one .
this is an example where a quantum superposition state has negative energy density , even though each component of the superposition would have positive energy density by itself . in the case of the superposition of two squeezed vacuum states ,
one finds the opposite effect .
although here the superposition state does has negative energy density , it is somewhat less negative than in the case of a single squeezed vacuum state .
furthermore , the most negative energy density now occurs for small mean photon number , as opposed to large number in the case of a single squeezed vacuum state . in the case of a superposition of a coherent state and a squeezed vacuum state
, we find that for fixed coherent state parameter , there is negative energy density for small squeeze parameter , and again for larger values , but there is an intermediate range where the energy density is always positive .
we next examined some two - mode states involving entanglement between the two modes , including two examples of entangled squeezed vacuum states .
the first example , the barnett - radmore state @xcite , exhibits somewhat less negative energy density than would be found if each mode were separately in a squeezed vacuum state . in the second example , the zhang state @xcite , we find results similar to those in the superposition of squeezed vacuum states .
there is negative energy in the zhang state , but only for small mean particle numbers . finally , we examined a two - mode entangled coherent state , which also exhibits negative energy for small mean particle number .
it is also similar to the case of a superposition of coherent states of a single mode , but the entangled state has somewhat more negative energy density .
one of the motivations for this investigation is to draw links between theoretical studies of violations of the weak energy condition , and experimental work in quantum optics .
we hope that this line of work will lead to further experimental studies of subvacuum phenomena . in this appendix
, we will calculate some of the matrix elements of operators such as @xmath185 and @xmath186 which are needed to find the energy density in the states treated in this paper .
we begin with matrix elements between squeezed vacuum states .
the diagonal matrix elements @xmath187 and @xmath188 are given by eq .
( [ eq : diag_sq ] ) .
we need off - diagonal matrix elements of the form @xmath189 and @xmath190 , where @xmath132 is real .
if we set @xmath95 , so that @xmath191 , then eq . ( [ eq : xi ] ) becomes @xmath192 this leads to the result @xmath193 use the fact that @xmath194 to find @xmath195 similarly , we may use eq .
( [ eq : xi2 ] ) to show that @xmath196 because these matrix elements are real , we have that @xmath197 in the present case , the squeeze operator is @xmath198}\ , .
\label{eq : squeezer}\ ] ] from this relation , we see that @xmath199 next we derive the matrix elements involving both a coherent state and a squeezed vacuum state that are needed in sect .
[ sec : csv ] . a coherent state may be represented in terms of number eigenstates as @xcite @xmath200 this may be combined with eq .
( [ eq : xi ] ) to show that @xmath201^n \nonumber \\ & = & { \rm e}^{-\frac{1}{2 } |\alpha|^2 } \ , \sqrt{{\rm sech}r}\ , \exp\left(-\frac{1}{2 } { \rm e}^{-i\delta } \alpha^2 \tanh r \right)\ , , \label{eq : in_prod}\end{aligned}\ ] ] and @xmath202 note that @xmath203 finally , we show that @xmath204 t.a .
roman , `` some thoughts on energy conditions and wormholes '' , in _ proceedings of the tenth marcel grossmann meeting on general relativity _ , edited by s.p .
bergliaffa and m. novello , ( world scientific , singapore , 2006 ) , gr - qc/0409090 .
fewster , `` energy inequalities in quantum field theory '' , in _ xivth international congress on mathematical physics _ , edited by j.c .
zambrini ( world scientific , singapore , 2005 ) , see updated version , math - ph/0501073 .
pfenning and l.h .
ford , _ quantum inequality restrictions on negative energy densities in curved spacetimes _ doctoral dissertation , ( dept . of physics and astronomy , tufts university ) , gr - qc/9805037 , sec . | we examine the maximum negative energy density which can be attained in various quantum states of a massless scalar field .
we consider states in which either one or two modes are excited , and show that the energy density can be given in terms of a small number of parameters .
we calculate these parameters for several examples of superposition states for one mode , and entangled states for two modes , and find the maximum magnitude of the negative energy density in these states .
we consider several states which have been , or potentially will be , generated in quantum optics experiments . |
in the absence of external forcing the fluid can not maintain its steady state because of the continous dissipation of energy .
the decay of kinetic energy in incompressible fluid turbulence has been explored in several studies ( see , e.g. , @xcite ) .
these studies , principally numerical , indicate that kinetic energy decays as @xmath0 , where @xmath1 is the initial time , i.e. the time when the initial forcing is switched off .
the value of @xmath2 seems to depend on the initial state .
similarly , recent numerical studies of decaying mhd turbulence @xcite also suggest that the total energy decays as power of time
. the phenomena of decaying mhd turbulence is observed in many different situations .
star - forming clouds are astrophysical examples @xcite .
in ref.@xcite we described a new shell model for mhd turbulence , which has all the conservation laws in the inviscid , unforced limit , has the right symmetrires and reduces to the well - known goy shell model for fluid turbulence in absence of any magnetic field . in this paper
we use our shell model to study decaying mhd turbulence .
we begin by solving this set of equations with the external forces and by using the numerical method described there .
we evolve the shell velocities @xmath3 and magnetic fields @xmath4 till we achieve the nonequilibrium , statistical steady state whose properties are described in ref.@xcite .
once we attain this state , say at a time @xmath5 , we use the last set of shell velocities and magnetic fields , i.e. , @xmath6 and @xmath7 , as the initial condition for our shell - model study of decaying turbulence in which we solve our shell - model equations but with all the forcing terms set equal to zero .
we then use the resulting time series for @xmath3 and @xmath4 to determine the total energy @xmath8 and the total crosshelicity @xmath9 where the origin of time is taken to be zero .
we find that @xmath10 , with @xmath11 ; this is illustrated in the log - log plot of fig.1 .
as we have noted above , it is relevant for initial conditions that have the correct statistical properties of homogenous , isotropic mhd turbulence which are well decribed by our shell model .
= 5.5 cm = 5.5 cm in a recent paper @xcite the effects of random initial conditions on the decaying solutions of the unforced navier - stokes equation with hyperviscosity have been discussed : it has been shown that , for an initially gaussian velocity distribution , equal - time two point correlation functions like @xmath12 decay as @xmath13 .
it has also been shown that , for this particular choice of initial conditions , the theory remains finite ( in a field - theoretic sense ) just by renormalisation of the two - point vertex functions . however , it is more relevant to explore the statitical properties of decaying fluid turbulence with ordinary viscosity .
it is also important to find out how the introduction of magnetic fields influences this decay . in this paper
we study decaying mhd with ordinary viscosity ; we work with both the @xmath14mhd equations and a @xmath15-dimensional generalisation of the @xmath16mhd model introduced in ref.@xcite .
we show that both these models give similar results . in particular ,
we show that , in both these models , appropriately defined structure functions decay with a power law in time ; this is modulated by a scaling function whose form depends on the initial state . in section.ii
, we describe the equations we work with and the intial conditions we use . for most of the calculations in this chapter ,
we choose initial correlations which are analytic in the low - wavenumber limit .
we choose ordinary viscosities for both velocity and magnetic fields .
we calculate the _ dynamic exponent _ @xmath17 and show that it is unaffected by the presence of the nonlinearities , i.e. , it remains 2 .
we next show that if there is no initial cross - correlation between velocity and magnetic fields , correlation functions defined at the same spatial point decay with power laws in time and , consequently , two point structure functions also decay likewise . also ,
multipoint correlation functions ( and structure functions ) exhibit similar behaviour , with simple ( and not anomalous ) scaling in time . in other words , the exponents for high - order structure functions are simply related to those of two - point functions .
another interesting feature is that there is no change in the exponents with the introduction of the magnetic field ( i.e. the exponents are the same for fluid and mhd turbulence ) ; only the amplitudes of the correlation or structure functions change . in the section iii
we show that in the presence of suitable cross correlations higher order correlations ( and consequently structure functions ) decay more slowly than when there is no cross correlation , i.e. , they display anomalous scaling . in section
iv we extend and discuss our results in the context of the decay of fully developed mhd turbulence . in section
v we conclude and compare results obtained from both @xmath14mhd and our simplified model .
the dynamical equations which govern the time evolutions of the velocity and magnetic fields in a magnetised fluid are given by the equations of magnetohydrodynamics , namely , @xmath18 where @xmath19 and @xmath20 are the velocity and the magnetic fields , @xmath21 and @xmath22 are fluid and magnetic viscosities , @xmath23 and @xmath24 represent pressure and density , and @xmath25 and @xmath26 are external forces .
we also impose the incompressibility condition @xmath27 and the divergence - free condition on the magnetic field @xmath28 . in the absence of magnetic fields
the mhd equations reduce to usual navier - stokes equation for fluid turbulence .
we now briefly present the ideas that go into the construction of the model , which has already been discussed in details in @xcite .
since this model is an extension of burgers equation for fluid dynamics to mhd , it should have the following properties : * both the equations should be galilean invariant . * the equations should be invariant under @xmath29 , @xmath30 and @xmath31 i.e. the velocity field is odd parity and the magnetic field is even parity . *
the fields are _ irrotational _ vector fields in @xmath15-dimension ( similar to burgers velocity field ) . * the equations in the ideal limit
i.e. in absence of any viscosity and any external forcing should conserve all the scalar conserved quantities of @xmath14mhd namely the total energy and the cross - helicity .
the third conserved quantity i.e. the magnetic helicity is zero for irrotational fields .
the above considerations lead to the following model equations @xmath32 @xmath33 whose relation to the @xmath14mhd equations is the same as the relation of the burgers equation to the navier - stokes equation . here
@xmath19 and @xmath20 are the velocity and magnetic fields with magnitudes @xmath34 and @xmath35 respectively .
these fields are @xmath36 , i.e. , @xmath37 and @xmath38 .
these equations together conserve the total energy @xmath39 and crosshelicity @xmath40 .
we have @xmath41 by homogeneity and isotropy .
we choose the initial correlations to have the simple analytic forms @xmath42 and @xmath43 in @xmath44-space the @xmath14mhd equations are @xmath45 the quantities @xmath46 appear to include the incompressibility , @xmath47 ) is the transverse projection operator and wavevector arguments @xmath48 indicate spatial fourier transforms .
the initial conditions we use for @xmath14mhd are @xmath49 which ensure the divergence - free conditons , namely , @xmath50 we discuss the structure of @xmath51 later . for the time being , we choose @xmath52 .
these simple initial conditions , which are analytic in the @xmath53 limit , are enough to elcidate the issues raised in this chapter .
the formal solutions of the eqs .
( 3 ) and ( 4 ) are given by @xmath54\exp\{-\nu t^{\prime } [ p^2+(k - p)^2]\}\nonumber \\&&[u_m ( { \bf p},0)u_m ( { \bf k - p},0 ) + \exp\{-\mu t^{\prime}[p^2+(k - p)^2]\}\times\nonumber \\ & & b_m({\bf p},0)b_m({\bf k - p},0 ) ] + ... \label{mu1}\end{aligned}\ ] ] @xmath55\exp\{-t^{\prime}[\nu p^2+\mu ( k - p)^2]\}\nonumber \\ & & [ u_m({\bf p},0)b_m({\bf k - p},0)]+ ... \label{mb1}\end{aligned}\ ] ] where ellipsis refer to higher order terms similarly the formal solutions of the @xmath14mhd equations ( [ 3dmhdu ] ) and ( [ 3dmhdb ] ) are given by : @xmath56exp\{-\nu t^{\prime } [ p^2+(k - p)^2]\}\nonumber \\&&[u_j ( { \bf p},0)u_s ( { \bf k - p},0 ) \nonumber \\ & & - \exp\{-\mu t^{\prime}[p^2+(k - p)^2]\}b_j ( { \bf p},0)b_s({\bf k - p},0)]+ ... , \label{3dmu1}\end{aligned}\ ] ] @xmath57\exp\{-t^{\prime}[\nu p^2+\mu ( k - p)^2]\ } \nonumber \\ & & [ u_m({\bf p},0)b_l({\bf k - p},0 ) -u_l({\bf p},0)b_m({\bf k - p},0 ) ] + ... , \label{3dmb1}\end{aligned}\ ] ] where the ellipsis in refer to higher order terms in the series .
we use bare perturbation theory , i.e. , after truncating the series at a particular order , all @xmath19 and @xmath20 are to be replaced by their bare values , i.e. , by @xmath58 and @xmath59 , respectively , where @xmath60 and @xmath61 .
this is similar to ref.@xcite . in the steady state of fluid or mhd turbulence ,
the _ dynamic exponent _
@xmath17 is determined by the form of the correlations of the driving noise and dimensionality of the space .
for example , when driven by noise correlations of the form @xmath62 ( in @xmath14ns ) , one finds @xmath17 ( within a 1-loop dymanic renormalisation group calculation ) to be 2/3@xcite . what is @xmath17 when we study decaying mhd turbulence ?
we calculate this within a 1-loop diagramatic perturbation theory : in the bare theory ( i.e. , when the nonlinear terms are absent ) @xmath63 . if the one - loop corrections to the response functions ( i.e. , to the viscosities ) have infrared divergence at zero external frequency , then @xmath64 ; otherwise @xmath63 .
the diagrams shown in fig.2 contribute at the one - loop level in renormalisation of the fluid viscosity @xmath21 , similar diagrams renormalise @xmath22 .
the effective 1-loop fluid viscosity is given by @xmath65 + { d_{bb}\over\nu k^2}\int d^3 p{p_l p_m { \bf k.(k - p)}\over\ 2\mu { \bf k.p } } [ { 1\over\ 2\mu p^2 } ( 1-{\nu k^2\over\ 2\mu p^2})\nonumber \\ & - & { 1\over\ 2\mu p^2 } ( 1-{\nu k^2\over\ 2\mu p^2})].\end{aligned}\ ] ] we see that , in the long - wavelength limit , 1-loop corrections have no infrared divergence at @xmath66 ; in other words @xmath17 remains 2 . in the case of the @xmath14mhd equations ,
the forms of the integrals remain unchanged , only changes appear in the coefficients of the integrals where , in place of @xmath67 , appropriate projection operators are present .
thus again there is no diagramatic correction to viscosities in the long - wavelenght limit in @xmath14mhd .
hence the exponent @xmath17 is unaffected by the nonlinear terms .
therefore , without any loss of generality , we can put @xmath68 .
we now calculate the effective @xmath69 and @xmath70 for @xmath66 at the one - loop level .
the diagrams shown in fig.3 contribute at the 1-loop level to @xmath69 , these contributions are proportional to @xmath71 which does not have any infrared divergence in _ any dimension_. the one - loop diagram contributing to @xmath70 has the same form .
for @xmath14mhd the forms of the one - loop integrals do not change ; only the coefficients have appropriate factors of projection operators instead of @xmath72 . on purely dimensional grounds for @xmath66 , @xmath73 and @xmath74 similar quantities obtained from @xmath14mhd
also show the same exponents but with different amplitudes.mhd equations and our simplified model as the projection operator @xmath75 that appears in the @xmath14mhd equations is dimensionless and the nonlinearities in both the equations do not renormalise due to the galilean invariance of our simplified model and @xmath14mhd equations ( which in turn keeps the power - counting structure same ) .
amplitudes of course depend upon the specific models .
similarly , energy spectra obtained in the steady - state from different models which give same power - counting in the one - loop integrals , when driven by stochastic forces with correlations having same scaling , have same scaling propertes , see , for e.g. refs.@xcite . ] note that the temporal dependences can be scaled out of the momentum integrals easily as the lower and upper limits of the momentum integrals can be extended to 0 and @xmath76 , respectively .
it is easy to see that all multi - point correlation functions also do not exhibit any infrared divergence in three dimensions : consider , e.g. , 3-point functions .
even though @xmath77 at @xmath78 , it becomes non - zero at @xmath79 as can be easily seen at the one - loop level .
@xmath80 has a tree - level contribution and next a one - loop contribution .
the tree - level part being proportional to @xmath81 is zero because we have chosen a gaussian distribution for the initial conditions .
the one - loop diagrams are shown in fig.([duuu ] ) .
the 1-loop contribution is however nonzero .
it has the form @xmath82 k_{1i}k_{2j}(-k_{1s}-k_{2s } ) d_{uuu},\ ] ] where @xmath83 is the value of the one - loop integral in the limit @xmath84 , i.e. , @xmath85 which is infrared ( ir ) finite .
we can interpret this one - loop result by saying that the nonlinearities in eqs.([mhd1 ] ) and ( [ mhd2 ] ) are absent but there are nonzero initial three - point correlations at @xmath78 , i.e. , the effective linear system behaves as if @xmath86 hence @xmath87 \langle u_i({\bf k_1},t)u_j({\bf k_2},t)u_s({\bf -k_1-k_2},t ) > \nonumber \\ & = & \int d^3k_1 d^3k_2 exp(-k_1 ^2-k_2 ^ 2 -({\bf k_1+k_2})^2)t \exp[i{\bf k_1.(x_1-x_2)+k_2.(x_1-x_3 ) } ] \nonumber \\&&k_{1i}k_{2j}k_{3s}d_{uuu } \sim t^{-9/2}a_{ijs}({{\bf x_1-x_2}\over\sqrt t},{{\bf x_1-x_3}\over\sqrt t}).\end{aligned}\ ] ] the last step can be easily obtained by using the dimensionless variables @xmath88 .
next we consider the four - point functions : we concentrate on @xmath89 ; the diagramatic representations up to the one - loop contributions are given in fig.([duuuu ] ) .
the one - loop integrals in the long - wavelength limit , are @xmath90 these one - loop contributions can be interpreted as in the previous case of three - point functions i.e. as if there were initial four - point correlations in absence of the non - linearities : @xmath91 similar results can be easily obtained for other four - point functions involving velocity and magnetic fields . hence , @xmath92 \exp[i{\bf k_1.(x_1-x_2)+k_2.(x_3-x_4 ) } ] \nonumber \\ & + & d_{uuuu}\int d^3k_1d^3k_2d^3k_3k_{1i}k_{2j}k_{3s}(-k_{1l}-k_{2l}-k_{3l } ) \exp[-\nu\{k_1 ^ 2+k^2+k_3 ^ 2+(k_1 ^ 2+k_2 ^ 2+k_3 ^ 2)\}t ] \nonumber
\\&&\exp[i{\bf k_1.(x_1-x_2)+k_2(x_2-x_3)+k_3(x_3-x_4)}]\nonumber \\ & \sim & d_{uu}^2t^{-5}+d_{uuuu}t^{-13/2}. \label{4u}\end{aligned}\ ] ] in the infinite - reynolds - number limit , i.e. , in the limit @xmath93 , @xmath94 and hence the second term in the last line of eq.([4u ] ) dominates . as in the case for the three - point functions four - point functions ,
are of the form @xmath95 here @xmath96 and @xmath97 are third and fourth rank tensors respectively , and @xmath98 . to study decaying mhd , we extend the definiton of equal - time structure functions as follows : @xmath99|^n>\ ] ] and @xmath100|^n > , \\\ ] ] where @xmath101 and @xmath102 denotes the time that has elapsed after the forcing terms are switched off .
we now look at the general forms of these structure functions .
consider the three - point function first : @xmath103|^3\rangle.\ ] ] on expanding the expressions for , @xmath104 we get the following kind of terms : ( i)@xmath105 and ( ii)@xmath106 .
the first one is given by @xmath107d_{uuu}\nonumber \\ & \sim & t^{-9/2},\end{aligned}\ ] ] whereas the second term has the form @xmath108 \nonumber \\&&\exp[i{\bf k_1.(x_1-x_2)}]d_{uuu } \sim t^{-9/2}\phi(r/\sqrt t),\end{aligned}\ ] ] with @xmath109 .
thus , it is easy to see that @xmath110 has the structure @xmath111 with @xmath112 is a function with dimensionless argument @xmath113 .
we can easily extend our calculation above to see that @xmath114 where @xmath115 and @xmath116 are functions with dimensionless argument @xmath113 .
it is obvious that all these higher - order correlation functions _ do not _ exhibit any _ anomalous scaling _ in time ; their temporal behaviours can be obtained by simple dimensional arguments from the behaviours of the two - point functions .
it is easy to see that @xmath117 i.e. , @xmath118 , at the level of our one - loop approximation .
what happens if we make @xmath119 non - zero ?
let us first establish the form of @xmath51 : recall that @xmath120 has odd parity ( since @xmath121 is a polar vector whereas @xmath20 is an axial vector ) .
hence it is purely imaginary and is an odd function of @xmath44 .
( from now on we will work only in three dimension . )
thus we choose @xmath122 where @xmath123 is the totally antisymmetric tensor .
this is the simplest analytic form with the desired structure .
we consider the cases @xmath124 and 2 .
it is easy to see that the value of @xmath17 does not change if @xmath125 as no new diagram appears .
however , an additional diagram ( fig.4.6 ) contributes to the effective @xmath69 and @xmath70 if @xmath125 .
this diagram yields the following contribution @xmath126 where we have used the identity @xmath127 naive power counting shows that this contribution is finite for @xmath66 , hence @xmath69 and @xmath70 are not renormalised even if @xmath125 , for both @xmath124 and 2 . the same conclusion holds for @xmath14mhd as the one - loop integral remains finite ; only the coefficients change because of factors of @xmath128 .
it is quite interesting to note that for , @xmath129 , there is a diagramatic correction to @xmath130 ; the form of the correction is @xmath131 however this correction vanishes for both @xmath124 and 2 since the integrand is odd in @xmath132 .
hence there is no renormalisation of @xmath130 and two - point structure functions eqs.(4.22 ) and ( 4.23 ) scale as @xmath133 just as they do when @xmath134 .
we now consider the @xmath135-point correlation functions with @xmath136 .
none of the three - point functions have any infrared divergence .
hence they exhibit simple scaling , i.e. , @xmath137 .
let us consider the four - point functions namely , @xmath138 at the 1-loop level , all of them they will have finite parts ( coming from initial auto correlations of velocity and magnetic fields ) .
however , the first three of them will have _ new , infrared - divergent _ , one - loop contributions if @xmath125 .
the analytic structure of the divergent 1-loop integral is illustrated for one of these correlation functions in fig.([fig_4 ] ) whose contribution is @xmath139 isotropy yields @xmath140 , \label{delta}\ ] ] so summing over @xmath141 and @xmath142 separately we obtain @xmath143 for @xmath66 .
hence naive power counting for the integral in eq.([4pt ] ) yields , for @xmath124 , @xmath144 where @xmath145 is a momentum scale arising from the lower limit of the integral .
hence it is linearly infrared ( ir ) divergent for @xmath124 .
for @xmath146 there is no ir divergence as is the case with @xmath134 .
we analyse the case @xmath124 in detail : there is a disconnected diagram also that contributes to @xmath89 , which yields @xmath147 this is finite in the @xmath53 limit , so the contribution from the connected part will dominate in the @xmath148 limit . if we define the effective four - point correlation to be @xmath149 ( with the choice @xmath150 ) , where @xmath151 is just a constant ) then @xmath152 hence one obtains @xmath153 similarly @xmath154 .
the other four - point functions are ( dimensionally ) linearly divergent at the 1-loop level ; however , the 1-loop integral vanishes becauses the integrand is odd in @xmath23 .
hence they will scale as @xmath155 .
it is now obvious that @xmath156 ^ 4>&\sim & t^{-1/2}t^{-13/4}a_u ( { { \bf x_1-x_2 } \over t^{1/2 } } ) \nonumber \\ & \sim & t^{-1/2}<[u_i({\bf x_1},t)-u_i({\bf x_2},t)]^4>|_{d_{ub}=0 } , \\
< [ b_i({\bf x_1},t)-b_i({\bf x_2},t)]^4>&\sim & t^{-1/2}t^{-13/4}a_b ( { { \bf x_1-x_2 } \over t^{1/2 } } ) \nonumber \\ & \sim & t^{-1/2}<[u_i({\bf x_1},t)-u_i({\bf x_2},t)]^4>|_{d_{ub}=0}.\end{aligned}\ ] ] hence we see that these structure functions exhibit anomalous dynamical scaling if @xmath125 .
similarly , it follows that the even - order exponents are @xmath157 where @xmath158 is the _ anomalous _ part of the exponent @xmath159 .
by contrast the odd order exponents @xmath160 do not have an anomalous part . in fig.([formula ] ) we present plots of @xmath161 versus @xmath135 for both even and odd @xmath135 .
in this section we choose initial conditions of the following form : @xmath162 we choose then one - loop effective @xmath69 and @xmath70 will scale as @xmath163.]@xmath164 .
the effective fluid viscosity @xmath165 , and two - point correlations @xmath166 and @xmath166 are given by @xmath167}[{1\over\nu[p^2+(k - p)^2-k^2]}(1- { k^2\over\ p^2+(k - p)^2})\nonumber \\&- & { 1\over\ 2\nu p^2}(1-{k^2\over\ p^2}){\bf k.(k - p ) } + { 1\over\nu k^2}\int { d^3pp^{-s}p_ip_jd_{bb } \over\mu[p^2+k^2-(k - p)^2]}[{1\over\mu[p^2+(k - p)^2-k^2]}\nonumber \\&&(1- { k^2\over\ p^2+(k - p)^2})- { 1\over\ 2\mu p^2}(1-{k^2\over\ p^2}){\bf k.(k - p)},\\ \tilde{d}_{uu}&=&d_{uu}+\delta { d_{uu}^2\over\nu^2 k^{1-s}},\\ \tilde{d}_{bb}&=&d_{bb}+\delta { d_{bb}^2\over\nu^2 k^{1-s}},\\ \tilde{d}_{ub}&=&d_{ub},\end{aligned}\ ] ] where @xmath168 is just a numerical constant .
the self - consistent solution is given by @xmath169 , with the long - wavelength behaviours of @xmath170 and @xmath130 unchanged .
now we can easily calculate the decay of the total velocity and magnetic energy , and cross helicity ; these are : @xmath171 and @xmath172 here @xmath173 are total velocity - field and magnetic - field energies , and total crosshelicity respectively .
while calculating the temporal decay of these quantities it has been implicitly assumed that the lower and the upper limits of the momentum integrals can be extended upto 0 and @xmath76 , respectively ; if this not possible temporal behaviour can not be obtained by scaling ( see next section )
. in fully developed @xmath14mhd turbulence one finds , in the inertial range , @xmath174 of course , a study of higher - order correlation functions reveals that the velocity and magnetic- field distributions are not gaussian @xcite .
if , however , for the purpose of analytical tractability , we assume that the initial distributions are gaussian with the variances as ( for our burges - type model ) @xmath175 such that @xmath176 as observed in fully developed mhd turbulence , we can calculate @xmath17 and effective two - point correlations as we did above in sec ( 4.2.3 ) . for simplicity
we give details only for our simple burgers - type model [ eqs.([mhd1]-[mhd2 ] ) ] .
the effective @xmath21 , @xmath69 and @xmath70 are given by @xmath177 + { d_{bb}\over\nu k^2}\int d^d p{p_l p_m p^{-11/3 - 2 } { \bf k.(k - p)}\over\ 2\mu { \bf k.p } } [ { 1\over\ 2\mu p^2 } ( 1-{\nu k^2\over\ 2\mu p^2})- \nonumber \\ & & { 1\over\ 2\mu p^2 } ( 1-{\nu k^2\over\ 2\mu p^2 } ) ] \\
\tilde{d}_{uu}&=&d_{uu}+\delta { d_{uu}^2\over\nu^2 k^{8/3}}+ \delta { d_{bb}^2\over\mu^2 k^{8/3}}\\ \tilde{d}_{bb}&=&d_{bb}+2\delta{d_{bb}d_{uu}\over\mu^2 k^{8/3}}.\end{aligned}\ ] ] here @xmath178 is a numerical constant .
the self - consistent solutions of these equations are given by @xmath179 and @xmath180 , implying that @xmath181 ; this is the same as in the steady state . with this self - consistent solution
, we see that the 1-loop corrections to @xmath69 and @xmath70 are as singular as the bare quantities .
hence we ignore their renormalisations .
when @xmath130 is non - zero , it will contribute to the 1-loop corrections of @xmath69 and @xmath70 with the same low - wavenumber singularity .
@xmath130 itself does not have any diagramatic correction , as in the case discussed in section iii .
the decay of two - point correlation functions will be of the form @xmath182 it is easy to see that , because of the infrared divergence of the integral , the lower limit can not be extended to 0 , so it will depend upon a lower cut - off ( the inverse of which is the integral scale ) . consequently the @xmath102-dependence can not be scaled out as in the previous case .
this problem is reminiscent of the sweeping divergence that appears in a perturbative drg calculation of fully developed turbulence or mhd turbulence @xcite .
in fact this problem begins when @xmath183 ; then the above integral becomes log - divergent .
this type of initial state can be prepared by randomly stirring the system with a stochastic force with a variance @xmath184 with @xmath185 . in a drg analysis
, sweeping divergences appear when @xmath186 ( but k41 scaling obtains for @xmath187)@xcite .
higher - order correlation functions have more severe divergences , hence their decay also depend strongly upon the integral scale .
we have examined the decay of the total kinetic and magnetic energies , and the crosshelicity in our shell model , these quantities decay with power laws in time . in order to understand this behaviour analytically
, we have employed a one - loop perturbation theory . in the first part of our calculations
, we have chosen random initial conditions with the statistical properties given in eqs.(5),(6 ) and ( 10),(11 ) ( with no crosscorrelations ) .
we have shown that such initial conditions lead to power - law decay of temporal correlations and appropriately defined structure functions . for initial correlations analytic in the long - wavelength limit no higher order structure functions
show any anomalous scaling .
however , on introducing crosscorrelations @xmath188 ( eq.4.43 ) higher - order structure functions exhibit anomalous scaling .
this highlights the importance of crosscorrreations in the statistical properties of mhd turbulence .
we have also extended our calculations to the case of initial correlations with singular variances and discussed the difficulties in the decay of fully developed mhd turbulence .
we have argued , that these difficulties are related to the sweeping effect which arises in the study of the stochsatically driven mhd turbulence ( in the statistically steady state ) .
it is also worth noting that we have obtained similar temporal behaviour from both @xmath14mhd and our simplified burgers - type model
. it would be interesting to see if our results can be checked experimentally and numerically .
the author wishes to thank rahul pandit , sriram ramaswamy and jayanta k bhattacharjee for discussions .
the author acknowledges csir ( india ) for financial support .
99 also at poorna prajna institute of scientific research , bangalore , india .
m. r. smith , r. j. donnelly , n. goldenfeld , and w.f .
vinen , _ phys .
lett._*71*,2583 ( 1993 ) .
g. comte - bellot and s. corrsin , _ j. fluid mech._*25 * , 657 ( 1966 ) .
m. m. low , r. s. klessen , and a. burkert , e - print : astro - ph/9712013 .
m. hossain , p. c. gray , d. h. pontius , jr . , and w. h. matthaeus , _ phys . fluid _ * 7 * , 2886(1995 ) .
s. galtier , h. politano , and a. pouquet , _ phys .
lett _ * 79 * , 2807 ( 1997 ) .
d. biskamp and w. mller , _ phys .
lett . _ * 83 * , 2195 ( 1999 ) .
a. basu , a sain , s. k. dhar and r. pandit _ phys .
lett . _ * 81 * , 2685 ( 1998 ) .
v. periwal , _ phys .
a _ * 231 * , 235 ( 1997 ) .
a. basu , j. k. bhattacharjee , and s. ramaswamy , _ eur .
j b _ * 9 * , 725(1999 ) .
d. forster , d.r.nelson , and m.j .
stephen , _ phys .
a _ * 16 * , 732 ( 1977 ) .
v.yakhot and s.a .
orzag , _ journal of scientific computing _ * 1 * ( 1986 ) .
c .- y . mou and p.b.weichman , _ phys .
rev . e_*52 * , 3738(1995 ) . | we study the effects of random homogenous and isotropic initial conditions on decaying magnetohydrodynamics ( mhd ) .
we show that for an initial distribution of velocity and magnetic field fluctuations , appropriately defined structure functions decay as power law in time .
we also show that for a suitable choice of initial cross - correlations between velocity and magnetic fields even order structure functions acquire anomalous scaling in time where as scaling exponents of the odd order structure functions remain unchanged .
we discuss our results in the context of fully developed mhd turbulence . |
the minimal supersymmetric standard model ( mssm ) is a promising extension of the standard model of particle physics ( sm ) @xcite . at the international linear collider ( ilc )
@xcite , the masses , decay widths , couplings , and spins of the new susy particles can be measured with high precision @xcite .
in particular , the lightest states like pairs of neutralinos , charginos , and sleptons , can be studied in the initial stage of the ilc , with a center - of - mass energy @xmath3 gev , and a luminosity of @xmath4 fb@xmath5 .
the lightest susy state is a pair of radiatively produced neutralinos @xcite @xmath6 the signal is a single high energetic photon , radiated off the incoming beams or off the exchanged selectrons , and missing energy , carried by the neutralinos @xcite .
the main standard model background is photons from radiatively produced neutrinos @xmath1 @xcite . in order to quantify whether an excess of signal photons from radiative neutralino production , @xmath7 , can be observed over the sm background photons , @xmath8
, we define the theoretical significance @xmath9 , and the signal to background ratio @xmath10 @xcite @xmath11 for example , a theoretical significance of @xmath12 implies that the signal can be measured at the statistical 68% confidence level . if the experimental error of the background cross section is 1% , the signal to background ratio must be larger than 1% .
a detection of the signal requires at least @xmath13
for the tree - level calculation of the cross sections @xmath14 for signal and background , we use the formulas for the amplitudes squared as given in ref .
@xcite . to regularize the infrared and collinear divergences , we apply cuts on the photon scattering angle @xmath15 and energy @xmath16
@xcite @xmath17 the upper cut on the photon energy @xmath18 is the kinematical limit of radiative neutralino production .
this cut also reduces much of the on - shell @xmath19 boson contribution to the background from radiative neutrino production @xcite .
note that the ratios @xmath10 and @xmath9 do not depend very sensitively on the choice of the cuts @xmath20 and @xmath21 , since signal and background have very similar distributions in energy @xmath16 and angle @xmath15 .
we present numerical results of the signal and background cross sections with emphasis on their dependence on the beam polarization @xcite , and on the higgsino and gaugino mass parameters @xmath22 and @xmath23 , respectively .
finally we discuss a msugra scenario , where radiative production of neutralinos is the only accessible susy state at @xmath2 gev . in fig .
[ fig : beampol ] , we show the beam polarization dependence of the signal @xmath27 and background cross sections @xmath26 . in the sps 1a scenario
@xcite , the neutralino is mostly bino , such that radiative neutralino production dominantly proceeds via right selectron @xmath28 exchange .
the background , radiative neutrino production , mainly proceeds via @xmath29 boson exchange .
thus positive electron beam polarization @xmath30 and negative positron beam polarization @xmath31 enhance the signal cross section and reduce the background at the same time , such that the significance is greatly enhanced , see fig .
[ fig : beampol ] . for beam polarizations of @xmath32 $ ]
, the signal cross section is @xmath33 $ ] fb , the background is @xmath34 $ ] fb , such that the significance is @xmath35 $ ] , and the signal to background ratio is @xmath36 $ ] . these results should motivate a detailed experimental analysis including monte carlo studies @xcite . in fig .
[ fig : beampol ] , we show contour lines of the signal @xmath27 and background cross section @xmath26 in the @xmath22@xmath23 plane .
the signal is decreasing for increasing @xmath23 , since the neutralino mass @xmath37 and the selectron mass @xmath38 are increasing . for decreasing values of @xmath39 gev ,
the bino component of the neutralino is decreasing , leading to a decreasing signal cross section @xmath27 .
note that the background cross section also depends on @xmath22 and @xmath23 , since the kinematical cuts include the neutralino mass , see eq .
( [ cuts ] ) . in fig .
[ fig : beampol ] , we also indicate the kinematical limits of the production of neutralinos @xmath40 ( dashed line ) and charginos @xmath41 ( dot - dashed line ) . above these lines ,
the radiative production of neutralinos is the only kinematically allowed susy state , which can however be observed with a significance of up to @xmath42 .
consider , as an example , the msugra scenario @xmath43 gev , @xmath44 gev , @xmath45 gev , and @xmath46 . in this scenario
, we have @xmath47 gev , @xmath48 gev , and the particle masses are @xmath49 gev , @xmath50 gev , @xmath51 gev , @xmath52 gev .
the neutralinos , charginos and selectrons are too heavy to be pair produced at @xmath2 gev .
however , neutralinos can still be radiatively produced . in table
[ tab : msusgr ] , we show the cross section and the background from radiative neutrino production for different sets of beam polarizations with @xmath53 and @xmath54 .
polarized beams enhance the signal , in particular the background is strongly reduced by a high degree of electron polarization @xmath55 .
note that without beam polarization , the signal can not be observed .
.cross sections , significance @xmath9 , and signal to background ratio @xmath10 , for different sets of beam polarizations , for @xmath3 gev , @xmath4 fb@xmath5 , and @xmath43 gev , @xmath44 gev , @xmath45 gev , @xmath46 @xcite .
[ cols="^,^,^,^,^,^,^ " , ] a pair of radiatively produced neutralinos @xmath56 is the lightest state of susy particles to be produced at @xmath57 colliders .
the signal is a single high energetic photon and missing energy
. the signal could not be observed at lep due to the large background from radiative neutrino production @xmath58 . at the ilc , however , polarized beams enhance the signal and simultaneously reduce the background .
we have shown that the significance for observing the signal can be as large as @xmath59 , and that the signal to background ratio can be as large as @xmath60 .
these results should motivate detailed experimental studies , to learn as much as possible about supersymmetry through the process of radiatively produced neutralinos .
a. datta , a. datta and s. raychaudhuri , phys .
b * 349 * ( 1995 ) 113 [ arxiv : hep - ph/9411435 ] .
s. ambrosanio , b. mele , g. montagna , o. nicrosini and f. piccinini , nucl .
b * 478 * ( 1996 ) 46 [ arxiv : hep - ph/9601292 ] .
a. datta , a. datta and s. raychaudhuri , eur .
j. c * 1 * ( 1998 ) 375 [ arxiv : hep - ph/9605432 ] .
s. y. choi , j. s. shim , h. s. song , j. song and c. yu , phys .
d * 60 * ( 1999 ) 013007 [ arxiv : hep - ph/9901368 ] .
h. baer and a. belyaev , [ arxiv : hep - ph/0111017 ] .
a. datta and a. datta , phys .
b * 578 * ( 2004 ) 165 [ arxiv : hep - ph/0210218 ] .
h. k. dreiner , o. kittel and u. langenfeld , phys .
d * 74 * ( 2006 ) 115010 [ arxiv : hep - ph/0610020 ]
. h. k. dreiner , o. kittel and u. langenfeld , arxiv : hep - ph/0703009 . g. w. wilson , lc - phsm-2001 - 010 .
a. birkedal , k. matchev and m. perelstein , phys .
d * 70 * ( 2004 ) 077701 [ arxiv : hep - ph/0403004 ] . c. bartels , diploma thesis , _ model - independent wimp searches at the ilc _ , desy 2007
. g. a. moortgat - pick _
et al . _ ,
arxiv : hep - ph/0507011 . b. c. allanach _ et al .
_ , in _ proc . of the aps / dpf / dpb summer study on the future of particle physics ( snowmass 2001 ) _ ed
. n. graf , eur .
j. c * 25 * ( 2002 ) 113 [ econf * c010630 * ( 2001 ) p125 ] [ arxiv : hep - ph/0202233 ] . c. bartels et al .
, in preparation . | we study the impact of beam polarization on radiative neutralino production @xmath0 at the international linear collider .
we show that longitudinal polarized beams significantly enhance the signal and simultaneously reduce the standard model background from radiative neutrino production @xmath1 .
we point out that the radiative production of neutralinos could be the only accessible susy particles , if neutralinos , charginos , sleptons , as well as squarks and gluinos are too heavy to be pair - produced in the first stage of the ilc at @xmath2 gev .
( 10,10 ) ( 45,260)bonn - th-2007 - 04 |
the detection of temperature anisotropies in the cosmic microwave background ( cmb ) provided evidence that large - scale structure formation in the universe was seeded by small density fluctuations generated at early times .
the statistical properties of these seeds are usually modelled with a gaussian random field .
historically the gaussian approximation was introduced for mathematical convenience . in the absence of a solid model for the generation of density fluctuations
the gaussian hypothesis was accepted on the basis of the central limit theorem ( e.g. @xcite and references therein ) .
the advent of inflationary models provided further support for gaussianity .
small - amplitude curvature perturbations generated during a standard inflationary phase ( single field , slow roll ) are very nearly gaussian distributed ( e.g. @xcite and references therein ) .
however , many variants of the inflationary scenario predict appreciable levels of primordial non - gaussianity . in terms of bardeen
s gauge - invariant potential , @xmath7 , most of these models ( but not all , see e.g. * ? ? ?
* ) can be reduced to the form : = + ( ^2-^2 ) , [ deffnl ] where @xmath8 is an auxiliary gaussian random field and @xmath0 quantifies the amount of primordial non - gaussianity .
on subhorizon scales , @xmath9 where @xmath10 denotes the usual peculiar gravitational potential related to density fluctuations via poisson s equation . the parameter @xmath0 thus has the same sign as the skewness of the density probability distribution function .
this local form of non - gaussianity ( note that equation ( [ deffnl ] ) applies in configuration space ) can be obtained from a truncated expansion of the effective inflaton potential @xcite .
the parameter @xmath0 thus encodes information about the inflaton physics .
standard inflation gives @xmath11 @xcite .
however , even in this case , the non - linear evolution of perturbations on superhorizon scales yields an observable @xmath0 of order unity ( which , in reality , should be scale and redshift dependent ; @xcite , see also @xcite ) .
large values of @xmath12 naturally arise in multi - field inflation models ( e.g. @xcite ; for an extensive review see @xcite ) and even in cyclic or ekpyrotic models of the universe with no inflation @xcite .
observational constraints on @xmath0 have been derived studying three - point statistics of temperature fluctuations in the cmb @xcite .
the recent 5-year data from the _ wilkinson microwave anisotropy probe _ ( wmap ) give @xmath13 at the 95 per cent confidence level @xcite .
parallel studies on the same dataset give @xmath14 using minkowski functionals ( komatsu et al .
2008 ) and @xmath15 from wavelet decomposition @xcite .
some recent reanalyses of earlier
3-year wmap data claim substantial evidence for positive @xmath0 : @xmath16 from the bispectrum of temperature fluctuations @xcite and @xmath17 from their one - point distribution function @xcite . on the other hand ,
a study of minkowski functionals on the 3-year data gives @xmath18 @xcite .
higher quality data are needed to improve these constraints .
the upcoming planck satellite should be able to reduce the uncertainty in @xmath0 to @xmath19 @xcite .
alternatively , one might use observational signatures of primordial non - gaussianity imprinted in the large - scale structure ( lss ) of the universe ( e.g. @xcite ) .
ideally , one would like to use high - redshift probes as the non - linear growth of density fluctuations quickly superimposes a strong non - gaussian signal onto the primordial one so that the latter might then be difficult to recover . for instance , the large - scale distribution of neutral hydrogen in the era between hydrogen recombination and reionisation encodes information on @xmath0 @xcite .
this could be probed by detecting the redshifted hyperfine 21-cm transition with very low - frequency radio arrays from space . in principle , an experiment of this kind can limit @xmath0 to @xmath20 ( @xcite , see also @xcite ) . however , it is not clear yet whether such an experiment will ever be possible due to technical complexity and problematic foreground subtraction . at lower redshifts
, @xmath0 can be constrained probing the statistics of rare events , as like as the mass function of galaxy groups and clusters @xcite .
early attempts of using cluster counts to constrain @xmath0 have been rather inconclusive due to low - number statistics ( see e.g. @xcite and references therein ) .
even though cluster - mass estimates are still rather uncertain and massive objects are very rare , the observational perspectives look very promising . a number of galaxy surveys encompassing large fractions of the observable universe are being planned ( e.g. ground - based surveys as des , panstarrs , and lsst , and the satellite missions euclid and adept ) and could potentially lead to solid measurements of @xmath0 ( e.g. * ? ? ?
* ; * ? ? ?
[ cols="^,^,^,^,^,^,^,^ " , ] data in table [ tabbeta ] have an amazing regularity .
apart from a normalisation constant , each column ( row ) shows the same linear trend with @xmath4 ( @xmath0 ) .
this suggests that , within the explored parameter range ( @xmath21 mpc@xmath6 and @xmath22 ) , ( k,)=_0(1-_1 ) ( 1-_2k ) .
[ betafit ] we thus use this equation to fit the original data for the halo bias from gaussian and non - gaussian initial conditions and find @xmath23 at the 68.3 per cent confidence level .
note that we computed the power spectra in finite - sized bins of the wavenumber , so that there is some degree of ambiguity in associating the results with a given value of @xmath4 .
unfortunately the choice plays a role in determining @xmath24 as @xmath25 is a steep function of @xmath4 on the scales of interest . in table
[ tabbeta ] and in equation ( [ betafit1 ] ) , we have used the arithmetic mean of the wavenumbers contributing to a given bin .
if one instead uses the logarithmic center of the bin , @xmath26 is slightly reduced with a best - fitting value of @xmath27 .
the parameters @xmath28 and @xmath29 are unaltered . therefore
, a systematic contribution @xmath30 should be added to the error budget of @xmath26 .
equations ( [ deltab ] ) and ( [ deltabcorrected ] ) assume that the gaussian bias @xmath31 is constant with @xmath4 but this is only approximately true in the simulations .
the fit in equation ( [ betafit1 ] ) , the table [ tabbeta ] and the figures ( [ fig_ngb1 ] ) and ( [ fig_ngb2 ] ) have been obtained by identifying @xmath31 with the actual bias measured in the gaussian simulation at each wavenumber . if , instead , the estimate for @xmath31 introduced in section 5.1 is used , one gets @xmath32 , @xmath33 and @xmath34 , slightly improving the goodness of the fit . the quadratic dependence of @xmath35 on @xmath0 is rather surprising as it can not be straightforwardly derived from the simple models listed above .
it might possibly arise from higher - order terms which have been neglected in the expansion that leads to equation ( [ deltab ] ) .
anyway , it is clearly present in the simulations as it can be seen by looking at the variation of @xmath24 along a given row in table [ tabbeta ] . within the range of @xmath0 of physical interest ,
the effect is rather small : the coefficient @xmath28 only corresponds to a few percent correction .
note that a quadratic term breaks the symmetry in the amplitude of @xmath35 between non - linearity parameters with opposite sign and identical absolute value .
it is hard to directly test this against our simulations as we just have two runs with @xmath36 and both of them correspond to rather small @xmath12 where the uncertainties in @xmath24 are large .
an alternative explanation for a non - vanishing @xmath28 could be that it artificially derives from imposing a linear relation in @xmath37 to data that do not scale linearly for @xmath38 .
indeed , just using datapoints with small values of @xmath31 we derive bigger values of @xmath24 for large @xmath0 ( more or less in line with @xmath39 ) .
therefore , what is robust is that at least one of the scalings with @xmath31 or with @xmath0 is incorrect in equation ( [ deltab ] ) .
we found that a scaling proportional to @xmath40 ( with @xmath41 and @xmath42 two adjustable parameters ) does slightly better ( in terms of reduced @xmath43 ) than @xmath44 , at least for @xmath45 mpc@xmath6 .
however , since the scaling with @xmath37 has a sound theoretical basis @xcite we preferred to quote our results as in equation ( [ betafit ] ) . from the statistical point of view , the parameters ( [ betafit1 ] ) provide an acceptable description of the simulation data to high confidence for all values of @xmath31 .
however , they are particularly accurate for @xmath46 , while @xmath47 ( with the same @xmath26 and @xmath29 ) has to preferred for smaller values of @xmath31 .
the linear correction in @xmath4 should be thought of as the first - order term of a series expansion in the wavenumber .
we attempted to determine the corresponding quadratic term by considering larger values of @xmath4 in the fit ( one bin more , up to @xmath48 mpc@xmath6 ) .
however , values of @xmath35 become small compared with the numerical errors and we found that the quadratic parameter is badly constrained by the data ( @xmath49 ) while the other parameters remain nearly unchanged ( and get larger uncertainties ) . also note that the gaussian bias starts to depart from @xmath31 at @xmath50 mpc@xmath6 and it is not clear whether equation ( [ deltab ] ) should still be expected to hold in this regime . + @xcite derived an expression for @xmath35 which coincides with equation ( [ deltab ] ) but does not include the linear transfer function .
theoretically , this is hard to understand , as non - gaussianity is generated well before matter - radiation equality and one should account for the linear evolution of density perturbations .
anyway , due to the different @xmath4-dependence , their expression for @xmath35 provides a better fit to the simulation data than equation ( [ deltab ] ) when both models are allowed to vary in amplitude with a tunable free parameter . for the model of @xcite and @xmath51 for the model of @xcite ] none of them
, however , provides such an accurate fit to the data as our equations ( [ betafit ] ) and ( [ betafit1 ] ) , which improve the @xmath43 by at least a factor of @xmath52 .
slosar et al . ( 2008 ) have used equation ( [ deltab ] ) to constrain @xmath0 by considering measures of the clustering amplitude of luminous red galaxies ( lrgs ) and quasars from the sloan digital sky survey . combining all datasets
, they found @xmath53 at 95 per cent confidence .
how would this result change based on our simulations ? disentangling the different contributions , the strongest constraints to @xmath0 in slosar et al .
( 2008 ) come from the angular power spectrum of quasars with photometric redshifts in the range @xmath54 and a mean bias of @xmath55 .
weaker limits are also contributed from the power spectrum of spectroscopic lrgs and the angular spectrum of photometric lrgs .
( with a bias of @xmath56 at @xmath57 ) .
figure ( [ fig_ngb2 ] ) and table [ tabbeta ] suggest that at the scales of interest ( @xmath58 mpc@xmath6 ) the model given in equation ( [ deltab ] ) tends to overestimate the scale - dependent bias seen in the simulations . therefore , to match an observed @xmath35 ,
a larger value of @xmath12 is required than predicted by the analytic model . when applied to the data by slosar et al .
( 2008 ) , our correction should thus give somewhat looser limits on @xmath0 .
because of the strong @xmath4-dependence of the function @xmath24 it is impossible to give more precise estimates without fitting the power - spectrum data .
note , however , that a steeper @xmath4-dependence potentially makes determinations of @xmath0 even more competitive with respect to studies of cmb anisotropies .
we use a series of high - resolution n - body simulations to study the mass function and the clustering properties of dark - matter haloes arising from gaussian and non - gaussian initial conditions .
in particular , we follow the formation of structure in a universe characterised by the best - fitting parameters from the third- and fifth - year wmap data releases .
we consider non - gaussianity of the local type and we use eight different values of @xmath0 ( @xmath59 ) enclosing the parameter space currently allowed by studies of the cosmic microwave background .
our main results can be summarised as follows .
\(i ) the mass function of dark - matter haloes varies with @xmath0 .
larger values of the non - linearity parameter correspond to higher abundances of the most massive haloes .
analytical models based on the press - schechter method @xcite are compatible with our simulated results for the ratio of the gaussian and non - gaussian mass functions only if the critical threshold for halo collapse is lowered to @xmath60
. an accurate fit of the gaussian @xmath61 is necessary to derive the non - gaussian mass function from the aforementioned ratio .
\(ii ) we find that , in perfect analogy with the gaussian case @xcite , the halo mass function assumes an approximately universal form .
this means that , when expressed in terms of suitable variables , its dependence on redshift and cosmology is erased to good precision ( nearly 10 per cent ) .
we parameterise the @xmath0-dependence of the universal mass function and provide an accurate fit for its high - mass end . for @xmath62 and for masses
@xmath63 , the best - fitting parameters for the non - gaussian halo mass function in equation ( [ eq_ourfit ] ) are given in equation ( [ linearfit ] ) and table [ tab_mff_small ] .
this fit reproduces the mass function of friend - of - friends haloes with an accuracy of 5 per cent on top of a systematic contribution ( up to 10 per cent ) due to the non perfect universality . for applications requiring higher precision , an additional formula
is provided : for @xmath64 and @xmath65 the fit in equations ( [ eq_ngmfz ] ) , ( [ eq_ourfit ] ) , and ( [ eq_zandfnlfit ] ) has to be preferred to the universal fit . on the other hand , for higher values of @xmath12 and for higher redshifts ,
the universal fit gives a better and more economic ( in terms of parameters ) description of the data . in the gaussian case
, we extend the fit to a larger interval of halo masses ( @xmath66 ) by combining simulations with different box sizes : see equations ( [ eq_ourfit ] ) and ( [ eq_bestgaufit ] ) .
our fitting function provides a precious tool to forecast constraints on @xmath0 from future surveys and to analyze current datasets .
\(iii ) the matter power - spectrum in non - gaussian cosmologies departs from the gaussian one already on very large scales . for values of @xmath0 within the current cmb constraints
these scale - dependent deviations can be as high as two per cent at @xmath67 and increase with wavenumber .
the discrepancy is systematic : models with positive @xmath0 have more large - scale power than the gaussian case and models with negative @xmath0 have less .
this warns against the widespread habit of using the gaussian matter power spectrum to determine non - gaussian bias parameters when high - precision is required .
it also suggests that primordial non - gaussianity modifies the shape and amplitude of the baryonic - oscillation feature in the two - point statistics and the convergence power spectrum in weak - lensing studies .
\(iv ) we present an accurate fitting formula for the linear bias of dark matter haloes arising from gaussian initial conditions extending previous work to larger mass intervals .
this , together with the mass function fit mentioned above , can be used to constrain parameters of halo - occupation models from clustering data .
\(v ) finally , using the halo - matter cross spectrum , we confirm the strong @xmath4-dependence of the halo bias on large scales ( @xmath5 mpc@xmath6 ) which was already detected by @xcite . however , we show that commonly used parameterisations based on the peak - background split overestimate the effect for @xmath68 mpc@xmath6 .
the discrepancy increases with the wavenumber and at @xmath69 mpc@xmath6 @xmath35 in the simulations changes sign with respect to the models . on top of this ,
the analytic model for the scale - dependent part of the bias requires corrections which depend on the non - linearity parameter , the wavenumber and , possibly , also on redshift and clustering strength .
equations ( [ deltabcorrected ] ) and ( [ betafit ] ) with the best - fitting parameters listed in ( [ betafit1 ] ) provide a fitting formula which accurately reproduces the outcome of the simulations for @xmath70 mpc@xmath6 and @xmath71 .
this fit should be employed to constrain @xmath0 from future clustering data at low and high redshift .
all simulations were performed at the swiss national supercomputing center ( cscs ) in manno , switzerland .
ap and oh acknowledge support from the swiss national science foundation .
we thank volker springel for kindly making the lean version of the gadget-2 code available to us .
we acknowledge discussions with robert e. smith and tommaso giannantonio .
while our paper was being submitted , a related work by desjacques , seljak & iliev ( 2008 ) appeared as a preprint .
their results based on a different halo finder are in qualitative agreement with ours with the exception of the non - gaussian bias correction for small halo masses ( @xmath72 ) which they find to exceed the model by slosar et al .
( 2008 ) .
99 afshordi n.,tolley a.j . , arxiv:0806.104 amara , a. , & refregier , a. 2004 , mnras , 351 , 375 bardeen j.m . ,
bond j.r . , kaiser n. , szalay a.s . , 1986 , apj , 304 , 15 bartolo n. , komatsu e. , matarrese s. , riotto a. , 2004 , phr , 402 , 103 bartolo n. , matarrese s. , riotto a. , 2005 , jcap , 10 , 010 bertschinger , e.2001 , apjs , 137 , 1 bond , j. r. , cole , s. , efstathiou , g. , & kaiser , n. 1991 , apj , 379 , 440 buchbinder , e. i. , khoury , j. , & ovrut , b. a. 2008 , physical review letters , 100 , 171302 catelan , p. , lucchin , f. , matarrese , s. , & porciani , c. 1998 , mnras , 297 , 692 matarrese , carbone c. , verde l. , matarrese s. , 2008 , apj , 684 , 1 cohn j.d . , white m. , 2008 , mnras , 385 , 2025 cooray , a. 2006 , physical review letters , 97 , 261301 creminelli p. , senatore l. , 2007 , jcap , 11 , 010 creminelli p. , senatore l. , zaldarriaga m. , 2007 , jcap , 03 , 19 crocce m. , pueblas s. , scoccimarro r. , 2006 , mnras , 373 , 369 curto , a. , martinez - gonzalez , e. , mukherjee , p. , barreiro , r. b. , hansen , f. k. , liguori , m. , & matarrese , s. 2008 , arxiv:0807.0231 dalal n. , dore o. , huterer d.,shirokov a. , 2008 , phys . rev .
d77 , 123514 desjacques , v. , seljak , u. , & iliev , i. t. 2008 , arxiv:0811.2748 evrard , a. e. , et al .
2002 , apj , 573 , 7 falk t. , rangarajan r. , srednicki m. , 1993 , apj , 403 , 1 gangui a. , lucchin f. , matarrese s. , mollerach s. , 1994 , apj , 430 , 447 grinstein , b. , & wise , m. b. 1986 , apj , 310 , 19 grossi m. , dolag k. , branchini e. , matarrese s. , moscardini l. , 2007 , mnras , 382 , 1261 grossi m. , branchini e. , dolag k. , matarrese s. , moscardini l. , 2008 , arxiv:0805.0276 grossi m. , verde l. , carbone c. , dolag k. , branchini e. , iannuzzi f. , matarrese s. , moscardini l. , 2009 , arxiv:0902.2013 heitmann k. , luki z. , habib s. , ricker p.m. , 2006 , apj , 642 , 85 hikage , c. , matsubara , t. , coles , p. , liguori , m. , hansen , f. k. , & matarrese , s. 2008 , mnras , 389 , 1439 huterer , d. , & takada , m. 2005 , astroparticle physics , 23 , 369 jenkins a. , 2001 , mnras , 321 , 372 jeong , e. , & smoot , g. f. 2007 , arxiv:0710.2371 jing , y. p. 1998 , apj , 503 , l9 kang x. , norberg p. , silk j. , 2007 , mnras , 376 , 343 knebe a. , wagner c. , knollmann s. , diekershoff t. , krause f. , 2009 , arxiv:0904.0083 komatsu , e. , & spergel , d. n. 2001 , phrvd , 63 , 063002 komatsu , e. , et al .
2008 , arxiv:0803.0547 koyama , k. , soda , j. , & taruya , a. 1999 , mnras , 310 , 1111 lehners , j .- l . , & steinhardt , p. j. 2008 , phrvd , 77 , 063533 linde a. , mukhanov v. , 1997 , phrvd , 56 , 535 loverde m. , miller a. , shandera s. , verde l. , 2008 , jcap , 04 , 014 lucchin , f. , matarrese , s. , & vittorio , n. 1988 , apjl , 330 , l21 luki z. , heitmann k. , habib s. , bashinsky s. , ricker p.m. , 2007 , apj , 671 , 1160 magliocchetti , m. , & porciani , c. 2003 , mnras , 346 , 186 maldacena j. , 2003 , jhep , 05 , 013 matarrese , s. , lucchin , f. , & bonometto , s. a. 1986 , apj . , 310 , l21 matarrese s. , verde l. , jimenez r. , 2000 , apj , 541 , 10 matarrese s. , verde l. , 2008 , apjl , 677 , 77 mcdonald p. , arxiv:0806.1061 mo , h. j. , & white , s. d. m. 1996 , mnras , 282 , 347 moscardini , l. , matarrese , s. , lucchin , f. , & messina , a. 1991 , mnras , 248 , 424 peebles , p. j. e.1980 , research supported by the national science foundation .
princeton , n.j . , princeton university press , 1980 .
435 p. , pillepich , a. , porciani , c. , & matarrese , s. 2007 , apj , 662 , 1 porciani , c. , catelan , p. , & lacey , c. 1999 , apj , 513 , l99 press w.h . , schechter p. 1974 ,
apj , 187 , 425 pyne t. , carroll s.m . , 1996 , phrvd , 53 , 2920 reed d. , gardner j. , quinn t. , stadel j. , fardal m. , lake g. , governato f. , 2003 , mnras , 346 , 365 robinson , j. , & baker , j. e. 2000 , mnras , 311 , 781 robinson , j. , gawiser , e. , & silk , j. 2000 , apj , 532 , 1 salopek d.s . ,
bond j.r . , 1990 ,
phrvd , 42 , 3936 scoccimarro , r. , sefusatti , e. , & zaldarriaga , m. 2004 , phrvd , 69 , 103513 sefusatti , e. , & komatsu , e. 2007 , phrvd , 76 , 083004 seljak u. , warren m.s . , 2004 , mnras , 355 , 129 sheth r.k .
, mo h.j . , tormen g. , 2001 , mnras , 323 , 1 sheth r.k . , tormen g. 1999 , mnras , 308 , 119 slosar a. , hirata c. , seljak u. , ho s. , padmanabhan n. , 2008 , jcap , 08 , 031 smith r.e . , scoccimarro r. , sheth r.k . 2007 , phrvd , 75 , 3512 smith r.e . , 2008 , arxiv0810.1960s spergel , d. n. , et al . 2007 , apjs , 170 , 377 springel , v. 2005 , mnras , 364 , 1105 stanek , r. , rudd , d. , & evrard , a. e. 2008 , arxiv:0809.2805 taruya a. , koyama k. , matsubara t. , 2008 , arxiv:0808.4085 tinker j.l . , weinberg d.h . , zheng z. , zehavi i. , 2005 , apj , 631 , 41 tinker j.l . , kravtsov a.v . , klypin a. , abazajian k. , warren m.s . , yepes g. , gottlober s. , holz d.e . , 2008 , arxiv:0803.2706 verde , l. , wang , l. , heavens , a. f. , & kamionkowski , m. 2000 , mnras , 313 , 141 warren m.s . , abazajian k. , holz d.e . , teodoro l. , 2006 , apj , 646 , 881 white , m. 2002 , apjs , 143 , 241 willick , j. a. 2000 , apj , 530 , 80 yadav , a. p. s. , & wandelt , b. d. 2008 , physical review letters , 100 , 181301 zeldovich , y. b.1970 , aap , 5 , 84
the initial positions and velocities of the particles in our n - body simulations have been generated using the zeldovich approximation .
this method introduces long - lasting artificial transients in the growth of perturbations which might alter the halo mass function at the high - mass end even at very late epochs @xcite .
it is therefore important to start the simulation at a sufficiently high redshift to ensure that all transients have decayed within the cosmic time at which the simulation output is used for science applications .
alternatively , less stringent requirements on the initial redshift are necessary if one uses second order lagrangian perturbation theory to displace particles at the initial time @xcite . the simpler zeldovich approximation ( which only requires the calculation of the gravitational potential ) is much more widespread . in this case , a few authors have investigated how to compute the optimal starting redshift ( see e.g. @xcite ) as well as quantified the effects of the initial redshift on the halo mass function @xcite and on the internal properties of dark matter haloes @xcite .
these studies show that simulations of the concordance cosmology ( and gaussian initial conditions ) with initial redshifts of @xmath73 ( depending on the simulations specifications ) have converged to the correct solution by @xmath74 ( at least for halo masses @xmath75 ) . even though our @xmath76 is in the right ball park , it is important to test that our results are robust against changing initial redshift .
we thus decided to perform the following simple test : we re - simulated run1.0 and run1.750 of our main series using @xmath76=99 instead of @xmath76=50 and compared the halo mass functions of the two simulations as a function of redshift ( see figure a1 ) . in good agreement with @xcite , for @xmath77 we find that discrepancies are smaller than 10 per cent at @xmath78 and that the ( possible ) effects of zeldovich transients are completely erased by @xmath79 .
our results show that this holds true also for relatively large values of @xmath0 ( see the right panel in figure a1 where @xmath80 ) : even though non - linearities in the initial conditions are slightly enhanced with respect to the gaussian case , at any given redshift the corresponding density field is also more evolved and the effect of the zeldovich transients on the halo mass function are again erased by @xmath81 .
for this reason this paper only uses data from snapshots at @xmath82 which are accurate to better than 5 per cent . | we perform a series of high - resolution n - body simulations of cosmological structure formation starting from gaussian and non - gaussian initial conditions .
we adopt the best - fitting cosmological parameters from the third- and fifth - year data releases of the _ wilkinson microwave anisotropy probe _ and we consider non - gaussianity of the local type parameterised by eight different values of the non - linearity parameter @xmath0 .
building upon previous work based on the gaussian case , we show that , when expressed in terms of suitable variables , the mass function of friends - of - friends haloes is approximately universal ( i.e. independent of redshift , cosmology , and matter transfer function ) to good precision ( nearly 10 per cent ) also in non - gaussian scenarios . we provide fitting formulae for the high - mass end ( @xmath1 ) of the universal mass function in terms of @xmath0 , and we also present a non - universal fit in terms of both @xmath0 and @xmath2 to be used for applications requiring higher accuracy . for gaussian initial conditions ,
we extend our fit to a wider range of halo masses ( @xmath3 ) and we also provide a consistent fit of the linear halo bias . we show that , for realistic values of @xmath0 , the matter power - spectrum in non - gaussian cosmologies departs from the gaussian one by up to two per cent on the scales where the baryonic - oscillation features are imprinted on the two - point statistics . finally , using both the halo power spectrum and the halo - matter cross spectrum , we confirm the strong @xmath4-dependence of the halo bias on large scales ( @xmath5 mpc@xmath6 ) which was already detected in previous studies . however , we find that commonly used parameterisations based on the peak - background split do not provide an accurate description of our simulations which present extra dependencies on the wavenumber , the non - linearity parameter and , possibly , the clustering strength .
we provide an accurate fit of the simulation data that can be used as a benchmark for future determinations of @xmath0 with galaxy surveys .
[ firstpage ] cosmology : theory , dark matter , large - scale structure methods : n - body simulations galaxies : haloes , clusters . |
transporting biological molecules in living cells plays a key role in biochemical interactions , transmembrane signaling , and efficient reactions . in single - particle tracking ( spt ) ,
the motion of proteins or lipids is tracked to determine directly the diffusivity and to understand the biological role of diffusivity .
therefore , it is expected that spt experiments will provide new insight into molecular transport in living cells . in fact
, many spt experiments reveal anomalous dynamics such as subdiffusion , aging , fluctuating diffusivity , and heterogeneous environments in living cells @xcite .
mean square displacement ( msd ) is the most popular observable for characterizing the diffusivity of particles .
there are two different averaging procedures for calculating the msd .
one is the ensemble average , and the other is the time average . the time - averaged msd is defined as @xmath0^{2},\ ] ] where @xmath1 is the position of a particle at time @xmath2 tracked by the spt experiments and @xmath3 is the total measurement time . in stationary stochastic processes , these two averages are equivalent with the aid of the law of large numbers .
this equivalence is one of the properties of ergodicity . while ergodicity is a concept in dynamical systems , an observable in a stochastic system
is called ergodic if the time averages of the observable for different realizations converge uniquely to the ensemble average in equilibrium .
this property ensures the reproducibility of measurements in experiments @xcite : long spt measurements give the same result under the same experimental setup .
however , it was reported in spt experiments in living cells that this reproducibility breaks down @xcite , where the time - averaged msd for a fixed @xmath4 does not converge to a constant but fluctuates randomly across in realizations ( random diffusion coefficient ) .
further , other experiments also reveal that time averages of observables such as occupation time and intensity of fluorescence fail to converge to a constant in some non - equilibrium systems @xcite . while there are several distributional limit theorems related to distributional behaviors of time averages in probability theory @xcite , little
is known about the relationship between the stochastic models used in probability theory and the systems in experiments .
therefore , a theoretical foundation of irreproducibility is an important and challenging problem in statistical physics .
ergodicity gives a mathematical guarantee that time averages are equal to the ensemble average , i.e. , it ensures reproducibility @xcite .
mathematically , infinite ergodic theory generalizes the concept of ergodicity , and states that time - averaged observables remain random even in the long - time limit @xcite .
thus , it is expected that infinite ergodic theory will play a fundamental role in understanding random transport coefficients observed in spt trajectories @xcite .
however , ergodicity in stochastic processes has been studied in a different way .
if there is a highly stuck region in phase space , a particle can not explore the whole phase space due to the trapping in the stuck region .
such a situation is called weak ergodicity breaking @xcite .
when a system shows weak ergodicity breaking , a time - averaged observable does not converge to a constant even when the measurement time goes to infinity @xcite .
however , distributional behavior of time - averaged observables can be observed in homogeneous systems . in simple random walk , the time - averaged occupation time that a random walker resides in positive region
does not converge to a constant but converges in distribution , known as the generalized arcsine law @xcite .
thus , some time - averaged observables in homogeneous environments do not converge to constants but converge in distribution .
such a time - averaged observable is not reproducible but has a distributional reproducibility because the distribution is universal in the sense that it does not depend on initial ensembles . in stochastic models of anomalous diffusion , several distributional limit theorems for random diffusion coefficients
have been studied to elucidate irreproducibility @xcite . however , there are experimental results which can not be explained by such stochastic models @xcite .
the goal of this paper is to fill the gap between experimentally observed irreproducibility and distributional limit theorems in theoretical models . to consider the irreproducibility of the time - averaged msd in diffusion in living cells , we investigate the annealed transit time model ( attm ) @xcite , which has been shown to describe heterogeneous diffusion in living cells @xcite .
the authors of @xcite show anomalous diffusion and aging of the time - averaged msd .
however , the distributional behavior of the time - averaged msd remains an open problem .
moreover , the exact descriptions of the governing equations for the propagator have not been found so far . in this paper
, we describe the equations rigorously and solve them analytically . within this model , we show that the time - averaged msd remains random even in the long measurement times , i.e. , the diffusion coefficients are irreproducible but have distributional reproducibility in the sense that the distribution of the time time - averaged msd is universal .
in living cells , diffusivity strongly depends on space as well as time , that is , it is heterogeneous diffusion .
one of the simplest models describing such a heterogeneous diffusion process is the langevin equation with fluctuating diffusivity @xcite , @xmath5 where @xmath6 is the @xmath7-dimensional position of a particle at time @xmath3 and @xmath8 is a stochastic process .
such a fluctuating diffusivity results from a fluctuating medium driven by fluctuations of friction or temperature @xcite , diffusion in two - layer medium @xcite , or fluctuations of a diffusing particle s shape . in @xcite ,
dichotomous processes are used for @xmath8 to investigate effects of the underlying stochastic process @xmath8 on physical features of the time - averaged msd . to consider heterogeneous diffusion in living cells
, we have to model the stochastic process of @xmath8 . in a previous study ,
the attm was proposed for modeling heterogeneous diffusion in living cells , where the diffusion coefficient is constant for a random sojourn time and the constant depends on the sojourn time @xcite .
when we consider quenched environment with heterogeneous local diffusivities , sojourn times in slow and high diffusive regions will imply long and short times , respectively .
thus , it is physically natural to assume that the sojourn time is inversely coupled to the diffusion coefficient .
moreover , it is important to consider the annealed model like eq .
( [ lefd ] ) because the annealed framework enables us to treat analytical calculations , heterogeneous environments in cells may not provide quenched environments , and the annealed model is considered to be a good approximation for the quenched model . in this paper , we assume that the diffusion coefficient is coupled to the sojourn time , i.e. , @xmath9 ( @xmath10 ) as in @xcite . in non - equilibrium situations ,
the probability density function ( pdf ) @xmath11 of the sojourn time follows a power - law with no finite mean : @xmath12 where @xmath13 is a scale parameter .
a power - law sojourn - time distribution is observed in super - cooled liquids @xcite and can be derived with @xmath14 in the first passage time .
the mean sojourn time diverges for @xmath15 , which means that there is no finite characteristic time in the process . in other words ,
this process is an intrinsic non - equilibrium process .
[ traj_diff ]
here we provide several recurrence time distributions studied in renewal theory @xcite . as shown in fig .
[ traj_diff ] , the underlying diffusion process at time @xmath16 , @xmath17 , is determined by the first sojourn time @xmath18 and not by the forward recurrence time @xmath19 . therefore , one must consider recurrence time distributions to describe the exact equations for the propagator in the equilibrium situation . by the same technique in @xcite ,
the laplace transform of the joint pdf @xmath20 of the sojourn time at time @xmath3 , @xmath21 , and the forward recurrence time at time @xmath3 , @xmath22 , is given by @xmath23 where @xmath24 is the number of changes of states until time @xmath3 , @xmath25 is the time when the @xmath26th change of states occurs , and @xmath27 is the laplace transform of @xmath11 . in equilibrium process ,
the system started at @xmath28 , and we start to observe from @xmath16 @xcite .
thus , we have the double laplace transform of the joint pdf @xmath29 of @xmath30 and @xmath19 in equilibrium process : @xmath31 where @xmath32 is the mean sojourn time .
the inverse laplace transform with respect to @xmath26 and @xmath33 yields @xmath34 where @xmath35 if @xmath36 and @xmath37 otherwise . integrating eq .
( [ joint_pdf_taut_tauf ] ) with respect to @xmath19 yields the pdf @xmath38 of the sojourn times at @xmath16 in equilibrium ( @xmath30 for @xmath39 ) : @xmath40 the mean and the second moments of the initial diffusion coefficient @xmath17 in equilibrium can be calculated as @xmath41 and @xmath42 , and are assumed to be finite in equilibrium processes .
let @xmath43 be the pdf of the position @xmath44 at time @xmath3 and @xmath45 be the pdf of the position @xmath46 conditioned that the state of @xmath8 changes at exactly time @xmath3 .
we assume that @xmath47 .
hence , the pdfs satisfy the following generalized renewal equations : @xmath48 where @xmath49 is the pdf @xmath43 conditioned that the sojourn at @xmath3 , @xmath30 , is given by @xmath50 , @xmath51 is the joint pdf of the displacement @xmath46 and the sojourn time @xmath3 , @xmath52 is the joint pdf of the displacement @xmath46 and the first sojourn time @xmath3 , @xmath53 is the joint pdf of the displacement @xmath46 , the time elapsed @xmath3 from @xmath54 , and the last sojourn time ( the sojourn time at @xmath3 ) @xmath50 . finally ,
the pdf @xmath43 is @xmath55 where @xmath56 is the joint pdf of the displacement @xmath46 and the time elapsed @xmath3 , and the sojourn time @xmath50 , and there is no renewal during @xmath3 . in attm , the joint pdf @xmath51 is given by @xmath57 , where @xmath58 is a gaussian propagator with diffusion coefficient @xmath59 : @xmath60 moreover , the joint pdf @xmath53 is given by @xmath61 , where @xmath62 is a gaussian propagator with diffusion coefficient @xmath63 : @xmath64 and the joint pdf @xmath56 is given by @xmath65 . by the fourier - laplace transform , we have from eqs .
( [ grw_q ] ) and ( [ grw_p ] ) @xmath66 this is the exact representation of the fourier - laplace transform of the propagator , which is a generalization of the random walk framework @xcite .
next , we derive moments of the time - averaged msd . for @xmath67
, we approximate the time - averaged msd as @xmath68 where @xmath24 is the number of changes of states until time @xmath3 , @xmath69 is the @xmath70th sojourn time , @xmath71 , and @xmath72 is the time - averaged diffusion coefficient under the diffusion coefficient @xmath73 : @xmath74 we further assume that @xmath75 : @xmath76 this assumption is considered to be valid in the asymptotic limit for @xmath3 @xcite .
let @xmath77 be the pdf of @xmath78 and @xmath79 be the pdf of @xmath78 where a renewal occurs at exactly time @xmath3 .
we can write the generalized renewal equation for @xmath78 : @xmath80 where @xmath81 is the joint pdf of @xmath78 and the time elapsed @xmath3 , i.e. , @xmath82 , @xmath83 is the joint pdf of @xmath78 and the first renewal at @xmath3 , @xmath84 is the joint pdf of the displacement @xmath85 , the time elapsed @xmath3 , and the last sojourn time given by @xmath50 , i.e. , @xmath86 , and @xmath87 is the joint pdf of the displacement @xmath85 , the time elapsed @xmath3 , and the last sojourn time @xmath50 .
finally , @xmath77 can be obtained : @xmath88 by the double laplace transform , we have @xmath89
in equilibrium processes , the pdfs related to the first recurrence times are given by @xmath90 and @xmath91 . substituting these into eq .
( [ flt_propagator ] ) , we obtain the laplace transform of the msd : @xmath92 .
it follows that the msd grows linearly with time in equilibrium processes : @xmath93 moreover , the pdfs related to the first recurrence times in @xmath78 are given by @xmath94 and @xmath95 .
the laplace transform of @xmath96 , denoted by @xmath97 , is given by @xmath98 .
the inverse laplace transform reads @xmath99 . hence , @xmath100 . to characterize the irreproducibility of the time - averaged msd
, we calculate the relative standard deviation ( rsd ) studied in several diffusion processes @xcite : @xmath101^{2 } \rangle } } { \langle \overline{\delta^{2}(\delta;t ) } \rangle } .\ ] ] we note that the rsd is independent of @xmath4 because the time - averaged msd depends linearly on @xmath4 in the attm [ see eq .
( [ tmsd ] ) ] .
if the time - averaged msd is reproducible , then the rsd approaches zero as the measurement time @xmath3 goes to infinity .
it is important to note the rsd extracts a characteristic time from the system even when the time - averaged msd is reproducible @xcite . in particular , as will be shown below , a crossover time in the rsd is related to a characteristic time of fluctuating diffusivity if the instantaneous diffusivity changes over time . obtaining the laplace transform of @xmath102 and inverting it
, we have the asymptotic behavior of the squared rsd : @xmath103 when @xmath11 is the exponential distribution , the asymptotic behavior of the squared rsd decays as @xmath104 thus , the rsd becomes zero when the measurement time goes to infinity .
in other words , the time - averaged msd is reproducible in the long - time measurements . on the other hand , for measurement times that are small
compared with the characteristic time @xmath105 , the rsd does not decay : @xmath106 thus , there is a transition from constant rsd ( irreproducible ) to reproducible behavior , and the crossover time is related to a characteristic time like the mean sojourn time ( see fig .
[ rsd_exp ] ) .
the crossover time provides useful information on a characteristic time of fluctuating diffusivity , which has not been known so far . ) .
we used exponential distributions with different relaxation times for the sojourn time distribution .
the lines represent the theory , i.e. , eqs .
( [ theory_rsd_asym_exp ] ) and ( [ theory_rsd_flat_exp ] ) , while symbols show the results of numerical simulations . ] here , we assume that the pdf of the sojourn time follows a power - law distribution @xmath11 with exponent @xmath107 . because there is no equilibrium distribution for the forward recurrence time , this stochastic process is an intrinsic non - equilibrium process .
if @xmath108 , we note that the average of @xmath109 with respect to the sojourn time @xmath3 , i.e. , the msd during times when the state does not change , diverges because @xmath110 .
we consider a non - equilibrium situation in which the first renewal occurs at time @xmath16 . in this case , the generalized renewal equation is given by setting @xmath111 and @xmath112 in eqs .
( [ grw_q ] ) and ( [ grw_p ] ) . using the laplace analysis as in the equilibrium case
, we have the asymptotic behavior of @xmath113 for @xmath39 : 0 @xmath114 the inverse laplace transform reads @xmath115 our theory provides the exact form of the msd in the asymptotic limit , which matches perfectly with the results of numerical simulations without fitting the parameters ( see fig . [ msd_noneq ] ) .
the exponent of subdiffusion is the same as that previously obtained ( note that our notations are described by @xmath116 and @xmath117 in their notations ) @xcite .
( @xmath14 and @xmath118 ) .
the lines represent the theory ( [ theory_msd_noneq ] ) , while the symbols show the results of numerical simulations .
note that there are no fitting parameters . ]
next , we derive the ensemble average of the time - averaged msd .
the generalized renewal equation for @xmath78 is given by setting @xmath119 and @xmath120 in eqs .
( [ gre_qd ] ) and ( [ gre_pd ] ) . using the laplace analysis on @xmath77 as in the equilibrium case and using the relation @xmath121
, we have @xmath122 therefore , the ensemble average of the time - averaged msd shows aging : @xmath123 ( @xmath39 ) . figure [ tmsd_aging ] shows that this aging behavior is clearly described by eq .
( [ theory_tmsd_aging ] ) .
this exact form in the asymptotic limit has also been obtained for the first time .
moreover , we obtain the second moment of @xmath78 ( see appendix .
a ) : @xmath124 it follows that the rsd is given by @xmath125 in the asymptotic limit of @xmath126 .
therefore , the rsd does not decay even in the long - time limit for the measurement time .
the theory of the rsd has been confirmed in numerical simulations ( see fig .
[ rsd_crossover ] ) .
this is direct evidence of irreproducibility .
we note that the value of the rsd for @xmath127 is the same as that in a continuos - time random walk @xcite .
furthermore , we can show that all of the higher moments are given by @xmath128 where the coefficient @xmath129 is given in appendix a. because @xmath130 is the time - averaged diffusion coefficient , the distribution of the normalized time - averaged diffusion coefficient , @xmath131 , does not converge to a delta function like ergodic observables but converge to a broad distribution .
moreover , the distribution of time - averaged diffusion coefficients obtained from single trajectories is universal in the sense that it does not depend on the initial conditions nor noise histories
. therefore , time - averaged diffusion coefficient has a distributional reproducibility in the attm when the exponent is less than one . for different @xmath132 ( @xmath14 and @xmath118 ) .
the lines represent the theory ( [ theory_tmsd_aging ] ) , while the symbols show the results of numerical simulations .
note that there are no fitting parameters . ]
( @xmath133 , and 0.75 ) .
in numerical simulations , we use @xmath134 for calculating the rsd .
the dashed lines represent the theory described by eq .
( [ rsd_attm ] ) , while the symbols show the results of numerical simulations . ]
we have described rigorous equations for the propagator , eqs .
( [ grw_q ] ) and ( [ grw_p ] ) , and the time - averaged msd , eqs .
( [ gre_qd ] ) and ( [ gre_pd ] ) , in attm . by solving these equations ,
we have obtained exact solutions for the msd and the moments of the time - averaged msd . in equilibrium processes
, we found a transition from irreproducible to reproducible behavior in the time - averaged msd and extracted the characteristic time using the crossover time .
however , the rsd does not decay at all in non - equilibrium processes .
we have provided theoretical evidence for distributional reproducibility of the time - averaged msd in heterogeneous environments .
distributional behaviors for the time - averaged msd obtained here are closely related to the distributional limit theorem of a non - integrable observation function in infinite ergodic theory @xcite .
this is because the moments obtained here are similar to those in @xcite . in other words ,
the distribution looks the same in shape .
a quenched model , called the quenched radius model ( qrm ) , was also considered in @xcite . by analogy to the relationship between the quenched trap model and the annealed model ( continuous - time random walk ) @xcite , we conjecture that the exponent of the msd as well as the moments of the time - averaged msd in qrm will be the same as those in attm when the dimension is greater than two . however , it should be noted that the msd and the moments of the time - averaged msd differ when the dimension is less than two .
in fact , the subdiffusive exponent in the qrm is given by @xmath135 in the one - dimensional case when the second moment of patch size does not diverge @xcite .
thus , the ergodic properties for one - dimensional qrm will be different from those for attm , which is still an interesting open problem .
t. a. was partially supported by a grant - in - aid for young scientists ( b ) ( 26800204 ) .
10 url # 1#1urlprefix[2][]#2 golding i and cox e c 2006 _ phys . rev .
lett . _ * 96 * 098102 jeon j h , tejedor v , burov s , barkai e , selhuber - unkel c , berg - srensen k , oddershede l and metzler r 2011 _ phys . rev .
lett . _ * 106*(4 ) 048103 weigel a , simon b , tamkun m and krapf d 2011 _ proc .
usa _ * 108 * 6438 tabei s a , burov s , kim h y , kuznetsov a , huynh t , jureller j , philipson l h , dinner a r and scherer n f 2013 _ proc .
usa _ * 110 * 49114916 hfling f and franosch t 2013 _ rep .
phys . _ * 76 * 046602 manzo c , torreno - pina j a , massignan p , lapeyre jr g j , lewenstein m and parajo m f g 2015 _ phys . rev .
x _ * 5 * 011021 barkai e , garini y and metzler r 2012 _ phys . today
_ * 65 * 29 granli a , yeykal c c , robertson r b and greene e c 2006 _ proc .
usa _ * 103 * 1221 wang y m , austin r h and cox e c 2006 _ phys .
lett . _ * 97*(4 ) 048302 brokmann x and _ et al _ 2003 _ phys .
lett . _ * 90 * 120601 stefani f d , hoogenboom j p and barkai e 2009 _ phys . today
_ * 62 * 3439 takeuchi k a and akimoto t 2015 _ arxiv:1509.03082 _ darling d a and kac m 1957 _ trans . am .
soc . _ * 84 * 444458 lamperti j 1958 _ trans .
* 88 * 380387 dynkin e 1961 _ selected translations in mathematical statistics and probability ( american mathematical society , providence ) _ * 1 * 171 birkhoff g d 1931 _ proc .
* 17 * 656660 aaronson j 1997 _ an introduction to infinite ergodic theory _
( providence : american mathematical society ) thaler m and zweimller r 2006 _ probab .
theory relat .
fields _ * 135 * 1552 akimoto t and miyaguchi t 2010 _ phys . rev .
e _ * 82 * 030102(r ) akimoto t 2012 _ phys .
lett . _ * 108*(16 ) 164101 lutz e and renzoni f 2013 _ nat .
phys . _ * 9 * 615619 bouchaud j
p 1992 _ j. phys .
i _ * 2 * 17051713 lubelski a , sokolov i m and klafter j 2008 _ phys . rev .
_ * 100 * 250602 he y , burov s , metzler r and barkai e 2008 _ phys .
lett . _ * 101 * 058101 metzler r , jeon j h , cherstvy a g and barkai e 2014 _ phys
* 16 * 2412824164 feller w 1968 _ an introduction to probability theory and its applications _
vol 1 ( wiley , new york ) miyaguchi t and akimoto t 2011 _ phys .
e _ * 83 * 031926 akimoto t and miyaguchi t 2013 _ phys .
e _ * 87*(6 ) 062134 miyaguchi t and akimoto t 2015 _ phys .
e _ * 91*(1 ) 010102 massignan p , manzo c , torreno - pina j a , garca - parajo m f , lewenstein m and g j lapeyre j 2014 _ phys .
lett . _ * 112 * 150603 uneyama t , miyaguchi t and akimoto t 2015 _ phys .
e _ * 92*(3 ) 032140 akimoto t and yamamoto e 2016 _ phys .
e _ * 93*(6 ) 062109 rozenfeld r , uczka j and talkner p 1998 _ physics letters a _ * 249 * 409414 uczka j , talkner p and hnggi p 2000 _ physica a _ * 278 * 1831 beck c and cohen
e 2003 _ physica a _ * 322 * 267275 uczka j , niemiec m and hnggi p 1995 _ phys .
e _ * 52*(6 ) 58105816 akimoto t and seki k 2015 _ phys .
* 92*(2 ) 022114 miyaguchi t , akimoto t and yamamoto e 2016 _ phys .
e _ * 94*(1 ) 012109 helfferich j , ziebert f , frey s , meyer h , farago j , blumen a and baschnagel j 2014 _ phys .
e _ * 89*(4 ) 042603 cox d r 1962 _ renewal theory _
( london : methuen ) godrche c and luck j m 2001 _ j. stat . phys . _ * 104 * 489524 montroll e w and weiss g h 1965 _ j. math .
* 6 * 167181 scher h and lax m 1973 _ phys .
b _ * 7*(10 ) 44914502 shlesinger m , klafter j and wong y 1982 _ j. stat . phys .
_ * 27 * 499512 shlesinger m f , west b j and klafter j 1987 _ phys .
_ * 58*(11 ) 11001103 akimoto t and miyaguchi t 2014 _ j. stat . phys . _
* 515 deng w and barkai e 2009 _ phys . rev .
e _ * 79 * 011112 akimoto t , yamamoto e , yasuoka k , hirano y and yasui m 2011 _ phys . rev .
lett . _ * 107*(17 ) 178103 miyaguchi t and akimoto t 2011 _ phys .
e _ * 83 * 062101 uneyama t , akimoto t and miyaguchi t 2012 _ j. chem .
phys . _ * 137 * 114903114903 akimoto t , shinkai s and aizawa y 2015 _ j. stat .
phys . _ * 158 * 476493 machta j 1985 _ j. phys .
a _ * 18 * l531 bouchaud j and georges a 1990 _ phys .
* 195 * 127293
the @xmath7th derivative of @xmath136 satisfies the following recursion relation : @xmath137 , \ ] ] where @xmath138 ( @xmath139 ) and @xmath140 . 0 @xmath141 here
, we assume that @xmath142 it follows that @xmath143 \frac{1}{|\gamma(-\alpha)| s^{1 + n\sigma}}. \ ] ] therefore , @xmath144 | local diffusion coefficients in disordered systems such as spin glass systems and living cells are highly heterogeneous and may change over time .
such a time - dependent and spatially heterogeneous environment results in irreproducibility of single - particle - tracking measurements .
irreproducibility of time - averaged observables has been theoretically studied in the context of weak ergodicity breaking in stochastic processes . here ,
we provide rigorous descriptions of equilibrium and non - equilibrium diffusion processes for the annealed transit time model , which is a heterogeneous diffusion model in living cells .
we give analytical solutions for the mean square displacement ( msd ) and the relative standard deviation of the time - averaged msd for equilibrium and non - equilibrium situations .
we find that the time - averaged msd grows linearly with time and that the diffusion coefficients are intrinsically random in non - equilibrium situations .
our findings pave the way for a theoretical understanding of distributional behavior of the diffusion coefficients in disordered systems .
february 2014 _ keywords _ : anomalous diffusion , ergodicity , non - equilibrium processes |
one of the main goal of the study of relativistic heavy ion collisions is to study the deconfined matter , known as quark - gluon plasma(qgp ) , which is expected to form at large densities .
it has been suggested that the transition from hadronic to qgp state can be treated by percolation theory @xcite .
the formulation of percolation problem is concerned with elementary geometrical objects placed on a random d - dimensional lattice .
the objects have a well defined connectivity radius @xmath0 , and two objects can communicate if the distance between them is less than @xmath0 .
several objects can form a cluster of communication . at certain density of the objects a infinite cluster appears which spans the entire system .
this is defined by the dimensionless percolation parameter @xmath1 @xcite .
percolation theory has been applied to several areas ranging from clustering in spin system to the formation of galaxies .
figure 1 shows the transition from disconnected to connected system at high densities . in nuclear collisions
there is indeed , as a function of parton density , a sudden onset of large scale color connection .
there is a critical density at which the elemental objects form one large cluster , loosing their independent existence .
percolation would correspond to the onset of color deconfinement and it may be a prerequisite for any subsequent qgp formation .
the determination of the eos of hot , strongly interacting matter is one of the main challenges of strong interaction physics ( hotqcd collaboration ) @xcite .
recent lqcd calculations for the bulk thermodynamic observables , e.g. pressure , energy density , entropy density and for the sound velocity have been reported @xcite .
a percolation model coupled with hydrodynamics has been utilized to calculate these quantities from the star data at rhic energies in central au+au collisions .
multiparticle production at high energies is currently described in terms of color strings stretched between the projectile and target . hadronizing these strings
produce the observed hadrons .
the strings act as color sources of particles through the creation of @xmath2 pairs from the sea .
the number of strings grows with the energy and with the number of nucleons of participating nuclei .
color strings may be viewed as small discs in the transverse space filled with the color field created by colliding partons .
particles are produced by the schwinger mechanisms @xcite . with growing energy and size of the colliding nuclei the number of strings grow and start to overlap to form clusters @xcite . at a critical density
a macroscopic cluster appears that marks the percolation phase transition .
2d percolation is a non - thermal second order phase transition , but in cspm the schwinger barrier penetration mechanism for particle production and the fluctuations in the associated string tension due to the strong string interactions make it possible to define a temperature . consequently the particle spectrum is `` born '' with a thermal distribution @xcite .
the percolation threshold at which the spanning cluster appear , a `` connected '' system of color sources , identifies the percolation phase transition . with an increasing number of strings
there is a progression from isolated individual strings to clusters and then to a large cluster which suddenly spans the area . in two dimensional percolation theory
the relevant quantity is the dimensionless percolation density parameter given by @xcite @xmath3 where n is the number of strings formed in the collisions and @xmath4 is the transverse area of the single string and @xmath5 is the transverse nuclear overlap area .
the critical cluster which spans @xmath5 , appears for @xmath6 1.2 @xcite . as @xmath1 increases the fraction of @xmath5 covered by this spanning cluster increases .
the percolation theory governs the geometrical pattern of string clustering .
it requires some dynamics to describe the interaction of several overlapping strings .
it is assumed that a cluster behaves as a single string with a higher color field corresponding to the vectorial sum of the color charge of each individual string . knowing the color charge
, one can calculate the multiplicity @xmath7 and the mean transverse momentum @xmath8 of the particles produced by a cluster of strings .
one finds @xcite @xmath9 @xmath10 where @xmath11 and @xmath12 are the mean multiplicity and average transverse momentum squared of particles produced by a single string with a transverse area @xmath13 . in the saturation limit
, all the strings overlap into a single cluster that approximately occupies the whole interaction area , one gets the following universal scaling law @xmath14 this scaling law shows a reasonable agreement with all the experimental data for all projectiles , targets and energies @xcite . in the limit of high density one obtains @xmath15 with @xmath16 being the color suppression factor .
it follows that @xmath17 a similar scaling is found in the color glass condensate approach ( cgc)@xcite .
the saturation scale @xmath18 in cgc corresponds to @xmath19 in cspm .
the net effect due to @xmath20 is the reduction in hadron multiplicity and increase in the average transverse momentum of particles .
the cspm model calculation for hadron multiplicities and momentum spectra was found to be in excellent agreement with experiment @xcite . within the framework of clustering of color sources , the elliptic flow , @xmath21 , and the dependence of the nuclear modification factor on the azimuthal angle show reasonable agreement with the rhic data @xcite
. the critical density of percolation is related to the effective critical temperature and thus percolation may be the way to achieve deconfinement in the heavy ion collisions @xcite .
an additional important check of this interacting string approach was provided by the measurement of long range forward backward multiplicity correlations ( lrc ) by the star group at rhic @xcite .
to obtain the value of @xmath1 from data , a parameterization of p - p events at 200 gev is used to compute the @xmath22 distribution @xcite @xmath23 where a , @xmath24 , and @xmath25 are parameters used to fit the data .
this parameterization also can be used for nucleus - nucleus collisions to take into account the interactions of the strings@xcite @xmath26 where @xmath27 corresponds to the area occupied by the n overlapping strings .
the thermodynamic limit , i.e. letting n and @xmath27 @xmath28 while keeping @xmath1 fixed , is used to evaluate @xmath20 given by eq.(5 ) . for nucleus - nucleus collisions
eq.(8 ) becomes @xmath29 in pp collisions @xmath30 @xmath31 1 due to the low string overlap probability .
the star analysis of charged hadrons had presented the preliminary results for the percolation density parameter , @xmath1 at rhic for several collisions systems as a function of centrality@xcite .
figure 2 shows @xmath1 as function of the number of participant nucleons(@xmath32 ) in au+au collisions at @xmath33 200 and 62.4 gev .
these experimental @xmath1 values are used to get the bulk thermodynamic observables .
the strong longitudinal chromo - electric fields produce schwinger - bialas @xcite like radiation with a thermal spectrum , in analogy with the hawking - unruh radiation @xcite .
both the schwinger - bialas and hawking - unruh derivations lead to the same value of the maximum entropy temperature . above the critical value of @xmath1 , the qgp in cspm consists of massless constituents ( gluons ) .
the percolation density parameter @xmath1 determines the cluster size distribution , the temperature t(@xmath1 ) and the transverse momentum in the collision @xcite .
the connection between @xmath1 and the temperature @xmath34 involves the schwinger mechanism ( sm ) for particle production . in cspm the schwinger distribution for massless particles is expressed in terms of @xmath35 @xmath36 with the average value of string tension @xmath37 .
gaussian fluctuations in the string tension around its mean value transforms sm into the thermal distribution @xcite @xmath38 with @xmath37 = @xmath39 . the temperature is given by @xmath40 the string percolation density parameter @xmath1 which characterizes the percolation clusters also determines the temperature of the system . in this way at @xmath41=1.2 the percolation phase transition at @xmath42 models the thermal deconfinement transition .
in the determination of temperature using eq.(13 ) the value of @xmath20 is obtained using the experimental data @xcite .
we will adopt the point of view that the experimentally determined chemical freeze - out temperature is a good measure of the phase transition temperature , @xmath43 @xcite .
the single string average transverse momentum @xmath44 is calculated at @xmath41 = 1.2 with the universal chemical freeze - out temperature of 167.7 @xmath45 2.6 mev @xcite .
this gives @xmath46 = 207.2 @xmath45 3.3 mev which is close to @xmath47200 mev used previously in the calculation of the percolation transition temperature @xcite . above @xmath41 = 1.2 the size and density of the spanning cluster increases .
we use the measured value of @xmath1 = 2.88 to determine the temperature , before the expansion , @xmath48 = [email protected] mev of the quark gluon plasma in reasonable agreement with @xmath48 = 221@xmath49 from the enhanced direct photon experiment measured by the phenix collaboration@xcite .
among the most important and fundamental problems in finite - temperature qcd are the calculation of the bulk properties of hot qcd matter and characterization of the nature of the qcd phase transition .
the qgp according to cspm is born in local thermal equilibrium because the temperature is determined at the string level .
we use cspm coupled to hydrodynamics to calculate energy density , pressure , entropy density and sound velocity . as mentioned earlier the strings
interact strongly to form clusters and produce the pressure at the early stages of the collisions , which is evident from the presence of elliptical flow in cspm @xcite .
after the initial temperature @xmath50 the cspm perfect fluid may expand according to bjorken boost invariant 1d hydrodynamics @xcite @xmath51 where @xmath52 is the energy density , s the entropy density , @xmath53 the proper time , and @xmath54 the sound velocity . above the critical temperature only massless particles are present in cspm .
the initial energy density @xmath55 above @xmath43 is given by @xcite @xmath56 to evaluate @xmath55 we use the charged pion multiplicity @xmath57 at midrapidity and @xmath27 values from star for 0 - 10% central au - au collisions with @xmath33200 gev @xcite .
the factor 3/2 in eq.(19 ) accounts for the neutral pions .
the dynamics of massless particle production has been studied in qe2 quantum electrodynamics .
qe2 can be scaled from electrodynamics to quantum chromodynamics using the ratio of the coupling constants @xcite .
the production time @xmath58 for a boson ( gluon ) is @xcite @xmath59 in cspm the total transverse energy is proportional to @xmath1 . from the measured value of @xmath1 and @xmath52
it is found that @xmath52 is proportional to @xmath1 for the range @xmath60 , @xmath61 @xmath1 gev/@xmath62 @xcite .
this relationship has been extrapolated to below @xmath63 and above @xmath64 for the energy and entropy density calculations .
figure 3 shows @xmath65 as obtained from cspm along with the lqcd calculations @xcite and the cspm pressure @xmath66 .
for an ideal gas of massless constituents , the energy density and pressure are related by @xmath67 . in lqcd
the basic quantity is the interaction measure @xmath68 , which is also known as the trace anomaly . in
cspm @xmath69 reaches zero for t @xmath70 while in case of lqcd it is strongly interacting even for t @xmath71 @xcite .
the sound velocity requires the evaluation of s and @xmath72 , which can be expressed in terms of @xmath1 and @xmath20 . with @xmath73 = @xmath20
one obtains : @xmath74 then @xmath75 becomes : @xmath76 for @xmath77 , an analytic function of @xmath1 for the equation of state of the qgp for t @xmath78 .
figure 4 shows the comparison of @xmath75 from cspm and lqcd .
the lqcd values were obtained using the eos of 2 + 1 flavor qcd at finite temperature with physical strange quark mass and almost physical light quark masses @xcite . at @xmath79=1
the cspm and lqcd agree with the @xmath75 value of the physical hadron gas with resonance mass truncation m @xmath80 2.5 gev @xcite
. the entropy density @xmath81 is obtained from energy density and speed of sound as shown in fig.5 along with the lqcd results .
cspm is in excellent agreement with the lqcd calculations in the phase transition region for @xmath821.5 .
the percolation analysis of the color sources applied to star data at rhic provides a compelling argument that the qgp is formed in central au - au collisions at @xmath33 200 gev .
it also suggests that the qgp is produced in all soft high energy high multiplicity collisions when the string density exceeds the percolation transition .
the results are also in agreement with lattice qcd in the phase transition region , when the results are plotted with respect to @xmath83 and @xmath84 .
the value of @xmath75=0.14 is in agreement with the physical resonance gas value at the critical temperature . thus clustering and percolation can provide a conceptual basis for the qcd phase diagram which is more general than symmetry breaking @xcite .
this research was supported by the office of nuclear physics within the u.s .
department of energy office of science under grant no .
de - fg02 - 88er40412 .
00 t. celik , f. karsch , h. satz , phys .
b*97 * , 128 ( 1980 ) .
m. b. isichenko , rev .
phys . * 64 * , 961 ( 1992 ) .
a. bazavov et al . , phys .
rev . d*80 * , 014504 ( 2009 ) .
j. schwinger , phys .
rev . * 82 * , 664 ( 1951 ) .
m. a. braun , c. pajares , eu .
j. c*16 * , 349 ( 2000 ) .
m. a. braun , f. del moral , c. pajares , phys .
c * 65 * , 024907 ( 2002 ) .
j. dias de deus , c. pajares , phys .
b*642 * , 455 ( 2006 ) .
a. bialas , phys .
b*466 * , 301 ( 1999 ) .
h. satz , rep .
. phys . * 63 * , 1511 ( 2000 ) .
j. dias de deus , e. g. ferreiro , c. pajares , r. ugoccioni , phys .
b*581 * , 156 ( 2004 ) .
j. schaffner - bielich , d. kharzeev , l. mclerran , r. venugopalan , nucl .
a*705 * , 494 ( 2002 ) .
j. dias de deus , e. g. ferreiro , c. pajares , r. ugoccioni , eur .
j. c*40 * , 229 ( 2005 ) .
p. brogueira , j. dias de deus , j. g. milhano , nucl .
a. * 832 * , 76 ( 2010 ) .
i. bautista , l. cunqueiro , j. dias de deus , c. pajares , j. phys .
g*37 * , 015103 ( 2010 ) .
b. i. abelev et al . , ( star collaboration ) , phys .
* 103 * , 172301 ( 2009 ) .
b. k. srivastava , r. p. scharenberg , t. tarnowsky , ( star collaboration ) , nukleonika * 51 * , s109 ( 2006 ) .
s. w. hawking , commun .
phys . * 43 * , 199 ( 1975 ) .
w. g. unruh , phys .
d*14 * , 870 ( 1976 ) .
m. k. parikh , f. wilczek , phys .
. lett . * 85 * , 5042 ( 2000 ) .
d. kharzeev , k. tuchin , nucl .
a*753 * , 316 ( 2005 ) .
d. kharzeev , e. levin , k. tuchin , phys .
c*75 * , 044903 ( 2007 ) .
p. castorina , d. kharzeev , h. satz , eur .
j. c*52 * , 187 ( 2007 ) .
braun - munzinger , j. stachel , christof wetterich , phys .
b*596 * , 61 ( 2004 ) .
f. becattini , p. castorina , a. milov , h. satz , eur .
j. c*66 * , 377 ( 2010 ) .
a. adare et al . , ( phenix collaboration ) , phys .
104 * , 132301 ( 2010 ) .
j. d. bjorken , phys .
rev . d*27 * , 140 ( 1983 ) .
b. i. abelev et al . , ( star collaboration ) , phys .
c*79 * , 34909 ( 2009 ) . c. y. wong , introduction to high energy heavy ion collisions ( world scientific,1994 ) .
j. schwinger , phys . rev . * 128 * , 2425 ( 1962 ) .
p. castorina , j. cleymans , d. e. miller , h. satz , arxiv : hep - ph/0906.2289v1 .
h. satz , quantum field theory in extreme environments , paris , april 2009 . | possible phase transition of strongly interacting matter from hadron to a quark - gluon plasma ( qgp ) state have in the past received considerable interest .
it has been suggested that this problem might be treated by percolation theory .
the color string percolation model ( cspm ) is used to determine the equation of state ( eos ) of the qgp produced in central au - au collisions at rhic energies .
the bulk thermodynamic quantities- energy density , entropy density and the sound velocity- are obtained in the framework of cspm .
it is shown that the results are in excellent agreement with the recent lattice qcd calculations(lqcd ) .
relativistic heavy - ion collisions , percolation , qgp , eos |
first experimental studies of electron transport through single molecules attached to metal contacts by using a scanning tunnelling microscope ( stm ) or mechanically controlled break junction techniques @xcite , also triggered considerable theoretical activity in this field since the beginning of the new millennium .
the theoretical framework most widely used in this context is a non - equilibrium green s function ( negf ) formalism @xcite , where coherent electron transmission according to landauer s theory is assumed .
the conductance of a molecular junction can then be described in dependence on the incoming electron s energy @xmath1 in terms of the transmission probability @xmath2 , which within negf is defined as @xmath3,\ ] ] where the self energy matrices @xmath4 and @xmath5 contain the coupling of the molecule to the left and right electrodes and @xmath6 and @xmath7 are the retarded and advanced green s functions ( gf ) of the ( extended ) molecule .
the negf formalism has been implemented in a variety of codes , where in combination with density functional theory ( dft ) it allows for a first principles treatment of the conductance of single molecule junctions @xcite@xmath8@xcite .
the usefulness of such calculations , however , still relies on simple models for interpreting them in terms of quantum chemical concepts such as atomic orbitals ( aos ) or molecular orbitals ( mos ) in order to achieve a qualitative understanding of the observed electron transport characteristics in the context of our general knowledge of the electronic properties of molecules or what is regarded as chemical intuition . in principle
, electron transmission can be viewed as a particular manifestation of the more general phenomenon of electronic communication through a molecule , where a green s function describes the propagation of a perturbation and is a measure of the connected paths made available by the bonding pattern of the molecule @xcite .
the zeroth green s function @xmath9 can be used to rewrite the expression for @xmath2 given in eq .
[ negf ] under the assumption that the central molecule is coupled to both metal contacts only via a single ao labeled @xmath10 and @xmath11 on the respective side , because then each matrix @xmath12 contains only one non - zero element , @xmath13 and @xmath14 , respectively and therefore a single term remains from the trace resulting in @xmath15 by evaluating the relevant matrix elements of @xmath9 , chemical understanding of the general properties of molecules can arguably be complemented by studying their transmission properties .
for example , the low - bias conductance through benzene is orders of magnitude lower when it is contacted at positions which are meta ( _ m _ ) with respect to each other when compared with ortho ( _ o _ ) or para ( _ p _ ) @xcite@xmath8@xcite .
this result comes as no surprize for a chemist even without any prior exposure to the theory of molecular conductivity , who knows that the influence of a substituent in a benzene ring is `` felt '' , in both electrophilic and nucleophilic substitution reactions , in o- and p - position to it , while m - positions `` do not communicate '' .
this knowledge might be referred to as chemical intuition but is actually based on rules stemming from resonance theory within the valence bond ( vb ) framework , where the relation of electronic communication to the topology of mos is not self - evident but should be contained in @xmath9 . in their classical 1947 - 1950 series of papers
coulson and co - workers attempted to put the one - electron green s function ( without using the term at the time ) at the heart of chemical theory @xcite@xmath8@xcite .
they demonstrated how starting from a hckel hamiltonian in ao representation basic molecular characteristics such as mo energies , atomic charges , bond orders and response coefficients can be derived directly from the secular determinant without referring to explicit mo vectors , where the relation of this work to electron transport phenomena has been commented on very recently @xcite@xmath8@xcite .
although it was correctly claimed by datta @xcite amongst others that in a single molecule junction the conductance is defined not only by the central molecule but rather by the entire system including the metal contacts , the individual contributions of the components are separable in eq .
therefore , for the purpose of device design the molecular contribution can be optimized independently from the coupling to the metal contacts , a notion which has been recently confirmed in a joint theoretical and experimental work by manrique et al .
@xcite . in this study
it was shown that molecules and even their fragments contribute well defined and transferable factors to electron transport as a crucial observation for the investigation of destructive quantum interference ( dqi ) effects , a phenomenon which has been the topic of a tremendous number of recent articles , where for a rather complete bibliography we refer to ref .
@xcite .
such dqi effects when occurring in the transmission close to the fermi energy @xmath16 can be used for data storage @xcite , inducing thermoelectricity @xcite or the design of logic circuits @xcite .
simple models have been proposed for their analysis , which were derived from tight - binding ( tb ) or topological hckel theory and validated by dft calculations : one of them which we refer to as `` the graphical ao scheme '' in the following has been derived specifically for the prediction of dqi and is based on a graphical analysis of the connectivity matrix of atomic orbitals ( aos ) @xcite@xmath8@xcite , while the other interprets the efficiency of transmission in a broader sense in terms of the signs and amplitudes of molecular orbitals on the atomic sites directly connected to the electrodes @xcite@xmath8@xcite , and the analysis is sometimes limited to a `` frontier orbital approximation '' where only the highest occupied ( homo ) and the lowest unoccupied mo ( lumo ) are taken into account . the aim of this article is to reconcile the predictions from these two conceptually different approaches for an interpretation and analysis of the molecular green s function .
it is expected that by focusing on either an ao or a mo representation of the same quantum - mechanical problem one should obtain the same results .
their reconcilation is akin to that of the vb and mo theories in the earlier days of quantum chemistry by van vleck et al . in 1935
@xcite .
but while vb and mo approaches become variationally equivalent for the ground state only in the limit of full configuration interaction , for electron transport within a single particle framework the representations of the molecular green s function in the ao and mo bases are already strictly equivalent on a single determinant level . from eq .
[ trans ] it can be seen that it is both necessary and sufficient to evaluate the purely molecular quantity @xmath9 for estimating whether the transmission will be finite or zero at any given energy @xmath1 .
the derivation of both `` the graphical ao scheme '' and the mo based scheme mentioned above start from this observation . within a frontier orbital approximation , however ,
only the homo and lumo are taken into accound instead of all mos contained in @xmath9 and this approximation then limits the range of applicability of the mo based scheme to that of the coulson - rushbrooke pairing theorem @xcite as we explain in detail in the next section . if the contributions of all mos and not only the frontier orbitals to @xmath9 in eq . [ trans ] are correctly accounted for on a quantitative level , however , dqi can be analyzed from a mo perspective leading to equivalent results as the graphical ao scheme from refs .
@xcite@xmath8@xcite for all conjugated @xmath0 systems both alternant and non - alternant , with and without hetero - atoms and regardless of which subset of sites the contact atoms belong to , which is the main message of our article .
the paper is organized as follows : in the next section we shortly review the graphical ao scheme and highlight its relation to eq .
[ trans ] .
here we also explain on the basis of the pairing theorem that dqi effects entering @xmath9 can in general only be quantitatively described and understood in terms of the onsite energies of all mos and their respective amplitudes at the contacted atomic sites .
furthermore , we clarify the connection of such a mo centered analysis scheme to larsson s formula , which has been originally proposed for the definition of an effective coupling from the mo contributions to the transfer integral in a marcus theory description of electron hopping @xcite@xmath8@xcite but later on also used for the analysis of coherent electron transport in single molecule junctions @xcite . in the following section we provide computational studies for a variety of test systems in order to substantiate our claim that it is possible to gain understanding of dqi effects in accurate terms for any conjugated @xmath0 system without the limitations of applicability facing the original frontier mo rules .
for all the molecular systems in our article numerical calculations on a dft level exist in the literature and most of them have also been studied experimentally .
since the focus of our work is on topological properties of @xmath9 , for the calculations we present here topological hckel hamiltonians are used in combination with negf . in this numerical chapter
we also present the respective predictions from the graphical ao scheme for all systems as a reference and demonstrate their equivalence to the results obtained from an analysis of mo contributions , where the convergence with respect to the number of mos included plays a prominent role . in the final chapter we provide a summary .
the zeroth green s function @xmath9 in eq . [ trans ] describes the propagation of a tunnelling charge between the atomic sites @xmath10 and @xmath11 mediated by all molecular orbitals ( mos ) which in the weak coupling limit can be formulated as @xcite @xmath17^{-1},\ ] ] where @xmath18 is the molecular hamiltonian , i a unity matrix and @xmath19 an infinitesimal imaginary term introduced in order to avoid divergence of @xmath9 at the eigenvalues of @xmath18 .
since @xmath9 is obtained from the inversion of the hamiltonian matrix @xmath18 , which is defined in an ao representation , one can analyze the properties of @xmath9 from the ratio of one of the minors of @xmath18 and its determinant @xcite@xmath8@xcite where @xmath9 is only equal to zero when the respective minor , as defined by the connection of the leads to two particular atomic sites @xmath10 and @xmath11 on the molecule , is also zero . in this way the graphical ao scheme for the prediction of dqi effects was derived , which has been formulated as the following set of rules : dqi occurs at @xmath20 if it is impossible to connect the two atomic sites @xmath10 and @xmath11 in a molecular topology , i.e. the only two sites with a direct coupling to the leads , by a continuous chain of paths , and at the same time fulfill the conditions ( i ) two sites can be connected by a path if they are nearest neighbors and ( ii ) at all atomic sites in the molecule other than @xmath10 and @xmath11 , there is one incoming and one outgoing path . in other words , for a finite zero - bias conductance all aos of the molecular topology have to be either traversed within a continuous chain of paths from @xmath10 to @xmath11 or be part of a closed loop in the topology , where the latter can be a double line due to the pairing of connected orbitals or a triangle or any larger loop @xcite .
we will demonstrate in the following section how to apply these rules for any given molecular topology in praxis . in a later extension of this scheme
it has been clarified that such defined paths can also cancel each other out in special cases and that therefore a sign has to be attributed to them @xcite .
it has to be noted that these `` paths '' are just mathematical terms coming from forming the minor of @xmath18 and should not be interpreted as the physical path of an electron moving through the molecule .
this graphical ao scheme has the advantage that it allows for the prediction of dqi without any numerical calculations being required simply by a visual assessment of the chemical structure of the central molecule in the junction .
the scheme has been designed for molecules with a conjugated @xmath0 system , because it is only @xmath0 electrons which are taken into account in the topological hckel hamiltonian it was derived from . in praxis this is not really a limitation , since potential functional molecules of interest are usually conjugated systems , where @xmath0-transmission is dominant . in order to allow for a simple analytical treatment , the derivation of the scheme also originally assumed sites with identical onsite energies and couplings to each other @xcite .
this assumption was later lifted in an attempt to generalize the method now also allowing for hetero atoms in the molecular structure but this came at the price of increased mathematical complexity @xcite .
another assumption was that the only energy @xmath1 , where @xmath2 is of interest is the fermi energy because it defines the zero - bias conductance and therefore the rules only apply at @xmath20 .
this latter assumption is rather delicate considering that in the model hamiltonian the graphical rules were derived from the onsite energy of carbon sites were artificially set to @xmath16 .
quite surprisingly , it was found in praxis that this rather crude approximation did not seem to limit the predictive qualities of the model even for cases where hetero atoms such as oxygen were involved in the molecular structures under investigation @xcite as long as the fermi energy defined by the metal leads lies within the homo - lumo gap of the molecule when energy levels are aligned @xcite . in order to gain a mo perspective of @xmath9 instead of an ao one
, @xmath18 has to be looked at in its diagonalized form as @xmath21 , where @xmath22 is the diagonal matrix of mo eigenenergies and @xmath23 is the matrix of the coefficients for the expansion of all mos as a linear combination of all aos in the molecule . inserting this definition of @xmath18 into eq .
[ zero ] gives @xmath24 which is the spectral representation of @xmath9 in a hckel ao basis with @xmath25 the coefficient of the l(r)-th ao in the m - th mo in a sum that runs over all @xmath26 occupied and unoccupied mos , which result from the coupling of the aos defining the basis vectors for @xmath18 .
it should be stressed that eq .
[ spectral ] is exact for any hamiltonian with an orthogonal ao basis and that this spectral representation of @xmath9 served as the starting point for the formulation of the molecular orbital rules for efficient transmission by yoshizawa and co - workers @xcite@xmath8@xcite .
for the special case of alternant hydrocarbons ( ah ) , which are molecules with a conjugated @xmath0 system where carbon atoms can be divided into two subsets , `` starred and unstarred '' , such that the atoms of one subset are bonded only to those from the other , the coulson - rushbrooke pairing theorem @xcite applies which states that ( i ) the @xmath0 electron energy levels are symmetrically distributed about the zero energy level ( which is assumed to be @xmath16 in single molecule junctions ) and ( ii ) that each occupied mo obtained from diagonalizing the corresponding hamiltonian in an orthogonal ao basis with an energy @xmath27 has its mirror image in the unoccupied region with the energy @xmath28 , which regarding its shape differs only in the sign of all ao coefficients of one subset . in the following we focus on molecules with an even number of mos , which we can then group in eq .
[ spectral ] into pairs of the contributions from the mos whose energies are linked by the symmetry relation it defines , i.e ( h , l ) , ( h-1,l+1), ... ,(h-(n/2 - 1),(l+(n/2 - 1 ) ) with h = homo , l = lumo and @xmath29 .
we can then redefine @xmath30 as the sum of these pairs , which in the following we will refer to as coulson - rushbrooke or cr pairs : @xmath31 the pairing theorem now predicts for aos @xmath10 and @xmath11 on carbon atoms of the same subset that @xmath32 because either @xmath33 and simultaneously @xmath34 , or @xmath35 and simultaneously @xmath36 , as all the coefficients in only one subset change their sign when comparing an occupied with its mirrored unoccupied level .
therefore , the terms in every cr pair of eq .
[ coulson ] cancel exactly at @xmath16 for this case and dqi occurs as a result as has also been observed in refs .
@xcite .
if on the other hand the contact aos @xmath10 and @xmath11 belong to carbon atoms from different subsets , then @xmath37 because either @xmath33 and @xmath36 , or @xmath35 and @xmath34 .
for this case , the contributions coming from the two individual parts of each cr pair of mos ( h - k , l+k ) including the homo and the lumo always add up constructively at @xmath16 in eq . [ coulson ] .
although any individual cr pair contribution is therefore nonvanishing , it is important to stress that destructive interference is still possible between cr pairs , as each of them can contribute either a positive or a negative term to @xmath9 . the pairing theorem , however , does not provide the means for an assessment of prediction of such inter - pair interference . the general conclusion from the pairing theorem is therefore that dqi will always occur for the electron transport through junctions containing alternant hydrocarbons when carbon atoms of the same subset are contacted , which is already sufficient to account for the low conductance of a variety of systems such as polyenes with contact atoms of the same parity , meta - contacted benzene or generic cross - conjugated molecules , where these cases can readily be identified from their chemical structure without any deeper analysis of the shapes and signs of their frontier mos . on the other hand , for alternant hydrocarbons contacted on carbons belonging to different subsets ,
i.e. where one contact atom is starred and the other one unstarred or for non - alternant hydrocarbons or for conjugated @xmath0 systems containing hetero atoms , the pairing theorem can neither predict nor rule out dqi . in the literature
these two cases are sometimes distinguished in terms of `` easy zeros '' ( the same subset contacted ) and `` hard zeros '' ( different subsets contacted ) @xcite or linked to the occurence of an odd or even number of zeroes in @xmath2 @xcite .
but for the purpose of our article the important distinction is that for even - membered alternant hydrocarbons contacted at sites of the same subset dqi will always occur , while for all other cases dqi can not be predicted without numerical calculations from a mo perspective .
we note that our discussion above only refers to alternant hydrocarbons with an even number of mos and therefore also an even number of carbon sites .
this is the general case for stable alternant hydrocarbons .
when the total number of mos is odd , which implies the existence of a non - bonding mo at the fermi energy with non - vanishing contributions from only one subset follows from the pairing theorem , which then allows for a conduction peak instead of a dqi induced minimum at @xmath16 when the contacted atoms belong to the subset contributing to this non - bonding mo .
we now turn our attention to the molecular orbital rules derived by yoshizawa and co - workers @xcite@xmath8@xcite , where the starting point was also the spectral representation of @xmath9 given in eq .
[ spectral ] .
these rules are amongst the earliest formulated providing a link between the complex phenomenon of dqi in electron transmission and the standard output of quantum chemical calculations , in this case the sign of the amplitudes of mos . within a frontier orbital approximation
they also become particularly simple to apply because then the entire sum in eq .
[ coulson ] is dominated by only one cr pair , namely the contribution to @xmath30 coming from the homo and the lumo , and then the remaining pairs can all be neglected because their large energetic distance @xmath38 to @xmath16 results in large denominators in the respective terms , thereby making them numerically negligible . from this assumption
, it can be concluded that the transport through a single molecule would be effective , i.e. dqi would be absent , when on the two contact atoms to the two leads i ) the sign of the product of the mo expansion coefficients in the homo ( @xmath39 ) is different from that in the lumo ( @xmath40 ) and ii ) all four involved amplitudes @xmath41 , @xmath42 , @xmath43 and @xmath44 are of significant magnitude .
if these conditions are not fulfilled , then `` inefficient '' transmission due to at least a partial cancellation of the contributions from the homo and lumo was predicted which was not formulated as necessarily the zero transmission which is typical for dqi in a rigid sense .
such a frontier orbital approximation , however , only delivers correct results for the prediction of dqi whre the cr pairing theorem @xcite is applicable . if the atoms contacted by the two electrodes belong to the same subset ( either starred or unstarred ) of carbon atoms in an even - membered ah , the cancellation of the contributions from the homo and the lumo to @xmath30 is a reliable indicator of dqi not necessarily because they are dominant , but because it represents the cancellation of also the contributions within all other cr pairs entering eq .
[ coulson ] .
this is the reason why dqi can be understood in this case in terms of the frontier orbitals alone .
for all other cases all mos in the system need to be considered . if an alternant hydrocarbon is contacted at atomic sites belonging to different subsets , i.e one being starred and one being unstarred according to the cr framework , then although contributions from the homo and lumo can only interfere constructively the tails related to lower lying occupied and higher lying unoccupied mos might still cancel out with those of the frontier orbitals at @xmath16 and cause dqi . for non - alternant hydrocarbons and organic molecules containing hetero atoms , it turns out to be equally insufficient to limit the analysis to just one or even two cr pairs of mo contributions . in the next section
we will provide a range of numerical examples justifying this assertion .
a somewhat simplified form of eq . [ spectral ] has been known for decades as larsson s formula in a different but related context , where it was used for the definition of the transfer integral mediated by a selected set of mos in a marcus theory description of electron hopping @xcite@xmath8@xcite .
more recently , it has been realized @xcite that the same formula can be also employed to define an approximation for @xmath2 in coherent eletron tunnelling as @xmath45 with @xmath46 being an energy dependent effective coupling containing the contributions from all mos of a molecular bridge and defined by @xmath47 here @xmath48 , @xmath49 and @xmath50 are the eigenenergy , and the respective couplings to the left and right contact of the @xmath51th mo and @xmath1 is the kinetic energy of a transferred electron .
it is easy to see by direct comparison that the effective coupling @xmath46 in eq .
[ gamma ] is very much related to the zeroth green s function in eq .
[ spectral ] .
there are only two differences between the two equations .
first , the mo amplitudes @xmath25 have been replaced by couplings @xmath49 and @xmath50 , which describe the overlap of each mo with a respective contact ao on the two leads . since in the negf - tb description
we employ for our numerical studies in the following section the contact ao on the leads is always the same orbital , this difference amounts to just the same constant factor for all the mo terms of the sum in eq .
[ gamma ] and is therefore irrelevant for our study where we just set this value to 1 .
the second difference between the two expressions , namely the dropping of the infinitesimal term @xmath19 just means that @xmath46 diverges at the eigenergies of all mos . in principle
this deficiency can be repaired by introducing an energy dependent normalisation factor as has been derived from the more general theory in ref .
@xcite by sautet et al .
@xcite under the very limiting condition of @xmath52 , which is relevant when the focus is on the analysis of stm measurements . since the qualitative behaviour of @xmath53 does not deviate from that of @xmath2 obtained from negf - tb in all systems investigated in this article , we avoide such a normalisation factor as an unnecessary complication . in the following section
we just truncate @xmath53 as obtained from eq .
[ gamma ] at the poles and scale it with the arbitrary constant of @xmath54 for the purpose of its graphical presentation in the related figures . while the poles of @xmath30 in eq .
[ spectral ] or @xmath46 in eq .
[ gamma ] define the peaks in @xmath2 when these quantities are squared and each of these peaks can be identified with the electron transmission through one individual mo in the absence of degeneracies , the non - resonant transmission for energies between the peaks contains contributions from all mos @xmath51 with the respective couplings @xmath49 and @xmath50 and these contributions can interfere constructively or destructively in dependence on the energy @xmath1 . in the sums of eqns .
[ spectral ] and [ gamma ] the sign of the contribution from any mo is determined by i ) the numerator of its corresponding term in the summation as defined by the product of amplitudes at the contact site or product of couplings to the metal leads and ii ) its denominator which depends on whether the energy @xmath1 for which @xmath2 is evaluated lies below or above the onsite energy of the mo in question . for ahs contacted via carbon atoms belonging to the same subset , dqi at the center of the homo - lumo gap
can be directly concluded from the cr pairing theorem by using eq .
[ coulson ] .
for all other cases , neither the pairing theorem applies nor can reliable predictions on dqi be obtained within a frontier orbital approximation , because the occurrence or absence of dqi for any given value of @xmath1 seems to depend on a fine balance of cumulative contributions with different signs from all mos , where a quantitative description of the decay of their respective tails is crucial as we will demonstrate in the next section .
all the case studies in this article are based on simple models both in ao and mo representations as derived from the topological properties of the molecular hamiltonian @xmath18 in a tight binding approximation with all onsite energies for the carbon aos contributing to the @xmath0 system set to zero , i.e. to the origin of energy assumed to represent @xmath16 , and the couplings between them to the resonance integral @xmath55 from hckel theory which also defines the unit of the energy axis . as an appropriate numerical benchmark for our conclusions we therefore present negf - tb calculations which have been conducted within the atomic simulation environment ( ase ) @xcite with a coupling of @xmath55 between the molecular topology and the semi infinite carbon chains used for the electrodes .
sulfur atoms have been given an onsite energy of 1.11 @xmath55 and c - s bonds a coupling value of 0.69 @xmath55 @xcite in order to account for the effect of the hetero atom .
since the molecules we investigate here all have been chosen due to the recent interest they attracted , we also refer to the relevant literature for each system in order to show that our conclusions harmonize with the results from more realistic negf - dft calculations or experimental conductance measurements . as a first example for our arguments in the last section we consider butadiene contacted at different sites ( fig .
[ fig1 ] ) .
since this molecule with a conjugated @xmath0 system is an even - membered ah , its even- and odd - numbered atoms belong to different starred / unstarred subsets and therefore dqi can be predicted for the ( 1,3- ) connection within a frontier orbital approximation in agreement with the pairing theorem as outlined in section ii b because carbon sites belonging to the same subset are contacted ( fig .
[ fig1]a ) . for all other possible connections , sites belonging to different subsets are contacted , and therefore constructive interference of the contributions from the homo and lumo alone is found according to the pairing theorem . the graphical ao scheme ( fig .
[ fig1]b ) as well as negf - tb calculations for @xmath2 and their estimates as @xmath53 from eq .
[ gamma ] after a diagonalisation of @xmath18 where for both all four mos have been properly accounted for ( fig.[fig1]c ) find dqi not only for the 1,3- but also for the 2,3-connection .
within the graphical ao scheme dqi is predicted if it is not possible to form a continuous line between the two contact sites and have all aos which are not on this line grouped up in pairs or as part of a closed loop .
the single sites which are not crossed or grouped up are marked by green dots for the sake of clarity , which also allows for the simple correspondence that dqi occurs where green dots are unavoidable .
the mo based scheme within a frontier orbital approximation on the other side also correctly predicts dqi for 1,3-positioning of the contacts , but not for the 2,3-connection .
while the amplitudes of both the homo and lumo are low on sites @xmath56 and @xmath57 ( fig . [ fig1]d ) , this justifies a reduced conductance but not a cancelling out to zero which is characteristic for dqi and found for the 2,3-connection for @xmath2 . in fig .
[ fig1]e we plot @xmath53 when only the contributions from the homo and the lumo enter the expression for @xmath46 in eq .
[ gamma ] ( green lines ) .
there it can be seen that indeed zero transmission is found also considering only the two frontier orbitals for the 1,3-connection in agreement with the predictions from the pairing theorem , while the 2,3-case shows a reduced but finite conductance when compared to 1,2- and 1,4-positions of the contacts .
we also plot @xmath53 from the contributions of the homo-1 and lumo+1 alone ( blue lines ) and find that only for the ( 2,3- ) connection they cross those of the frontier orbitals at @xmath16 . since the corresponding sum of terms entering eq .
[ gamma ] for the two pairs have different signs at their crossing point , they cancel each other out and lead to zero transmission .
this is probably the simplest example to contrast a case of dqi , which can be predicted within a frontier orbital approximation with one that is beyond its range of applicability . and @xmath11 and with `` stars '' related to the pairing theorem and involving only the @xmath0 electrons for a ) benzene connected in ortho- ( _ o_- ) , meta- ( _ m_- ) and para ( _ p_- ) positions ,
b ) dinaphthylethene ( dne ) and c ) dithienylethene ( dte ) in their open and closed forms , respectively , where for the latter also the alternant hydrocarbon analogs obtained by removing the s sites are shown . in d ) the chemical structure of azulene is shown and the positions for connecting metal contacts are numbered for making reference to them in the text . ] in figure [ fig2 ] we show the other molecular systems which we investigate as case studies in this article and also the corresponding tb next neighbor connectivity , which provides the basis for all our negf - tb computations as well as the application of the pairing theorem and the graphical ao scheme in the following .
we note that only unsaturated carbon atoms are part of the @xmath0 system and that it is only those which need to be considered in a tb framework . for benzene ( fig .
[ fig2]a ) , it is established knowledge @xcite@xmath8@xcite that the conductance is finite for an _
p_-connected pair of contacts but dqi occurs at @xmath16 for a _
m_-connection , which is also consistent with the chemical understanding of communication through an aromatic ring . in figure
[ fig2]a we illustrate that these findings can be also understood in the context of the pairing theorem ( fig .
[ fig2]b ) because only for the m - connection two `` starred '' carbon sites are contacted in this example of an even - membered ah meaning that each cr pair will provide a contribution of exactly zero in eq . [ coulson ] .
another type of systems where dqi plays an important role are molecular switches @xcite based on `` conducting '' and `` insulating '' isomers that can be transformed into each other in a highly reversible photochemical reaction .
we will consider here one family of such switches , namely diarylethenes @xcite , and in particular the homocyclic dinaphthylethene ( dne ) ( fig .
[ fig2]b ) and the heterocyclic dithienylethene ( dte ) ( fig . [ fig2]c ) . for both systems
the closed isomer is much better conducting than the open isomer in a molecular junction which has been demonstrated experimentally @xcite and confirmed theoretically @xcite , where the formation ( or breaking ) of a single bond distinguishes one from the other in structural terms .
negf - dft calculations can be found e.g. in ref .
@xcite and ref .
@xcite for dne and dte , respectively .
the molecular orbital rules by yoshizawa and co - workers have also been applied for both systems , where although their application in a narrow sense would have suggested constructive interference for the `` closed '' ( conducting ) and `` open '' ( insulating ) form for dne , the differences in conductance between the two forms found with negf - dft has been attributed to the larger orbital amplitudes for the `` closed '' form @xcite . as for the butadiene example we discussed above , this argument explains quantitative differences in the conductance but not the qualitative difference defined by the occurrence or absence of an interference minimum . for dte on the other side ,
the contributions from the homo-1 and lumo+1 had to be added to those from the frontier orbitals in order to reach better agreement with experimental findings @xcite .
as can be seen from their respective tb next neighbor connectivity in figs .
[ fig2]b and c the electrodes are attached to carbon atoms which belong to different subsets of the starring scheme for the open forms of both dne and dte . as a consequence of the mo symmetry properties following from the pairing theorem therefore constructive qi has to be found within all cr pairs of mos defined by eq .
[ coulson ] including the homo and the lumo , and hence dqi can only occur due to cancellation of terms between different cr pairs which can not be assessed by using a frontier orbital approximation , and
this is the reason why we included these systems in the present study .
dte is heteorocyclic , which means that we need to include an onsite energy for the sulfur atoms differing from those of the carbon sites and also a value for the c - s coupling in the negf - tb calculations in the following as specified in the computational details . in order to have a reference point for the application of the graphical ao scheme also for dte
, we also introduce a structure with the s atoms of dte deleted in fig .
[ fig2]c , which transforms it into an even - membered ah , where in the next section we will compare the transmission functions of both dte forms with and without sulfur sites . as the final system for our study we chose azulene ( fig .
[ fig2]d ) , which is a non - alternant hydrocarbon and therefore it is not possible to divide its carbon sites into `` starred '' and `` unstarred '' subsets in relation to the pairing theorem or derive any conclusions regarding destructive or constructive qi within a frontier orbital approximation .
this system is also of interest because it has been wrongly claimed in a joint experimental and theoretical study that for azulene the graphical ao scheme also fails in its predictions at least when the electrodes contact the sites @xmath58 and @xmath57 in fig .
[ fig2]d @xcite .
this claim has later been refuted @xcite , where it was shown that the predictions of the graphical ao scheme were also correct for azulene with @xmath58,@xmath57-contacts when closed loops of aos and not only pairs of them are considered which always has been a central aspect of the scheme @xcite .
negf - dft calculations for azulene containing compounds with different electrode contact sites including the @xmath58,@xmath57- and @xmath59,@xmath60-cases can be e.g. found in ref .
@xcite and ref .
@xcite . ,
@xmath57- and @xmath59,@xmath60-positions .
dqi is predicted if it is not possible to form a continuous line between the two contact sites and have all aos which are not on this line grouped up in pairs or as part of a closed loop .
the single sites which are not crossed or grouped up are marked by green dots for the sake of clarity . ] in fig .
[ fig3 ] , @xmath2 as calculated from negf - tb for all systems introduced in fig .
[ fig2 ] is shown in the left panels for each label and compared with @xmath53 in the right panels , which was obtained from eq .
[ gamma ] with mo onsite energies and amplitudes resulting from a diagonalisation of @xmath18 in the same tb setup , where for both quantities the curves for the systems exhibiting dqi at @xmath16 are shown in red and the others in black . apart from the units , which are in multiples of the conductance quantum @xmath61 for @xmath2 and chosen arbitrarily for @xmath53 the agreement in all cases is excellent , which fully justifies to investigate the absence or occurrence of dqi solely in terms of the contributions entering eq .
[ gamma ] . from both @xmath2 and @xmath53 the m - connection
is correctly identified as the only one with dqi for benzene ( fig .
[ fig3]a ) and zero conductance found only for the open form of dne ( fig .
[ fig3]b ) , a result which needs the inclusion of all mos and not just the frontier orbitals @xcite as we will further argue below . also for dte ( fig .
[ fig3]c ) only the open form exhibits a transmission zero at @xmath16 for the analog alternant hydrocarbons ( right two panels ) , which is shifted to higher energies if the s atoms are included in the calculations ( left two panels ) but still lowers the conductance at the fermi energy in its vicinity quite substantially even then . for azulene ( fig .
[ fig3]d ) there are qi minima across the energy spectrum for the two investigated junctions which differ in their respective contact sites .
but while contacts in the @xmath59,@xmath60-positions ( red lines ) result in zero conductance , the minima for the @xmath58,@xmath57-connected system are not only lying above @xmath16 but are also so narrow that they do not seem to have an impact on @xmath62 .
we note that all these features we summarized here are in good qualitative agreement with the respective negf - dft calculations in the literature we referred to in the last section when introducing the respective molecular structures above . in fig .
[ fig4 ] we demonstrate the application of the graphical ao scheme @xcite for all systems in fig . [ fig2 ] without hetero atoms and its predictions for dqi identify the cases with @xmath63 in the calculations shown in fig .
[ fig3 ] without a single failure , regardless of whether the molecular topology belongs to an alternant or non - alternant hydrocarbon or which subset of carbon atoms in relation to the `` starring '' scheme the electrodes are connected to . in principle
, it would also be possible to account for the presence of hetero atoms within the scheme as it has been done elsewhere @xcite for a treatment of dte containing its sulfur sites but this would come at the price of diminishing the scheme s simplicity and would not provide any important arguments for the present discussion .
this ao scheme considers all orbitals in an ao representation and relies on the structure of the entire hamiltonian thereby strongly reflecting the respective molecular topology .
this is in contrast to any frontier orbital approximation within a mo based scheme where by definition all but two mos are disregarded .
the pairing theorem justifies this omission for the specific case , where each cr pair defined by the respective equal distance of its parts to @xmath16 cancels out individually for symmetry reasons . for the other cases where interference is constructive within each pair dqi can still occur due to cancellation between pairs
this is why all mo contributions are significant in this latter case as we will further explore below . independently of the frontier orbital approximation
as we discuss it in this article , there is the common conception in studies of the conductance of single molecule junctions that @xmath62 is dominated by the mos close to @xmath16 since the tails of the peaks further apart decay rapidly and their contributions can therefore safely be disregarded @xcite .
this assumption is also motivated by the fact that the respective distance of each mo to the fermi energy enters its respective term in eqns .
[ spectral ] and [ gamma ] explicitly in the denominator and its increase can therefore be expected to reduce the terms significance .
there are two reasons why such preconceptions should be questioned regarding their validity in general : i ) while it is true that the denominator of a term in eqns .
[ spectral ] and [ gamma ] increases for high values of @xmath64 , this effect might be outweighed by large couplings or large mo amplitudes at the contact sites ; ii ) the distinction between the occurrence and absence of dqi is often one between an exact value of zero ( at least in the framework of tb where only @xmath0 electrons are considered ) and a rather small number which appears to be bigger than it actually is due to the logarithmic plotting of @xmath2 . in fig .
[ fig5 ] we increase the number of cr pairs of mos included in the sum of eq .
[ gamma ] for the calculation of @xmath53 stepwise for the systems in fig .
[ fig2 ] which exhibit dqi close to the fermi level but where this can not be predicted within a frontier orbital approximation . here we first consider only the homo and the lumo ( label @xmath58 ) , then the homo , the lumo , the homo-1 and the lumo+1 ( label @xmath56 ) and
so on where only the red curve with the highest label includes the contributions from all cr pairs of mos corresponding to the respective molecular topology .
quite contrarily to what might be expected , for the open forms of dne ( fig .
[ fig5]a ) and dte ( fig .
[ fig5]b and c ) the convergence of @xmath53 with the number of included pairs is not smooth but oscillating because the contributions of cr pairs of mos enter with alternating signs . even for the label with only the cr pair with the energies most remote from @xmath16 missing , i.e. the labels @xmath65 , @xmath66 and @xmath59 for dne , dte without and with s , respectively
, the conductance is still far from zero on a logarithmic scale . for the label @xmath57 with the three cr pairs of mos closest to the fermi energy
included it even has a magnitude comparable to that of the conducting closed form of the respective switch , where for dte the convergence behavior seems to be rather unaffected by the presence or absence of the s atoms .
this analysis strongly indicates that in order to capture dqi effects for a particular molecular topology correctly really all mos belonging to its @xmath0 system need to be properly accounted for in order to achieve a reliable theoretical description .
even for the non - alternant hydrocarbon azulene contacted in ( @xmath59,@xmath60- ) positions ( fig .
[ fig5]d ) , where no destructive but also no constructive interference within each pair can be indicated directly from the pairing theorem , the contributions from the frontier orbitals , i.e. the homo and lumo , alone ( label @xmath58 ) do not result in any dqi feature close to @xmath16 .
the inclusion of the second cr pair of mos ( label @xmath56 ) produces this feature but it then again needs the contributions from all cr pairs to position it energetically exactly at @xmath16 . , where @xmath46 is taken as the sum over all five occupied ( blue curve ) or only the three occupied mos closest to @xmath16 ( magenta curve ) in eq .
[ gamma ] for the open form of the ah analog of dte with the s atoms removed . ]
another property that arises from the pairing theorem is that in the case of dqi for even - membered ahs connected at sites belonging to different subsets , the contributions from all occupied and all unoccupied mos to @xmath46 in eq .
[ gamma ] must each cancel out individually at @xmath16 .
this is because in those cases the contribution from each half of a cr pair is equal to the other half in both sign and magnitude at @xmath16 , which means that it is then sufficient to consider either all occupied or all unoccupied mos alone . making use of this knowledge , in fig .
[ fig6 ] we plot @xmath53 from the sum over the occupied states in eqn .
[ gamma ] , where we compare taking all five occupied mos for the open form of the ah analog of dte with the s atoms removed ( blue curve ) with the case where the lowest lying two mos have been excluded from the summation ( magenta curve ) .
as can be seen from the figure , cutting out the lowest lying two mos does not make any difference in the energy regions of the peaks of the three higher lying mos , because the transmission in the vicinity of a peak is always largely dominated by the contribution of the one mo it is related to , but crucially decides whether dqi occurs at the fermi level or not .
this enforces the main message of our article that in order to identify dqi reliably from a mo perspective the contribution of all mos need to be taken into account and not just a selected few of them .
this finding has also high importance for the analysis of negf - dft results , where a cut coupling approach is routinely used to describe dqi in terms of a few mos close to @xmath16 only @xcite .
in summary , we showed that dqi in the electron transport of single molecule junctions can be reliably discussed from a mo perspective if the contributions from all mos are accounted for and not only those from mos close to the fermi level .
this applies in general and not only for even - membered alternant hydrocarbons contacted at carbon sites of the same subset as it is the case for predictions within a frontier orbital approximation . this mo perspective
, however , does not in general allow for the prediction of dqi without prior numerical calculations within a tb framework which puts it into contrast to a recently proposed graphical ao scheme @xcite where such predictions are indeed possible .
on the other hand such a mo based analysis is not limited to the prediction of the zero - bias conductance defined by the transmission at the fermi level and can thus provide essential input for the interpretation of computationally more demanding negf - dft results as well as reconcile findings from single molecule electronics with more traditional concepts from quantum chemistry .
10 s. shaik , h. s. rzepa and r. hoffmann , angew .
ed . * 52 * , 3020 ( 2013 ) . c. joachim , j. k. gimzewski , r. r. schittler , and c. chavy , phys .
letters * 74 * , 2102 ( 1995 ) .
m. a. reed , c. zhou , c. j. muller , t. p. burgin , and j. m. tour , science * 278 * , 252 ( 1997 ) .
j. reichert , r. ochs , d. beckman , h. b. weber , m. mayor , and h. v. lohneysen , phys .
letters * 88 * , 17680 ( 2002 ) .
r. h. m. smit , y. noat , c. untiedt , n. d. lang , m. c. van hemert , and j. m. van ruitenbeek , nature * 419 * , 906 ( 2002 ) .
y. meir and n. s. wingreen , phys .
lett . * 68 * , 2512 ( 1992 ) .
m. brandbyge , j. l. mozos , p. ordejon , j. taylor and k. stokbro , phys .
b * 65 * , 165401 ( 2002 ) .
y. xue , s. datta and m. a. ratner , chem .
phys . * 281 * , 151 ( 2002 ) .
a. r. rocha , v. m. garcia - suarez , s. w. baily , c. j. lambert , j. ferrer and s. sanvito , nature materials * 4 * , 335 ( 2005 ) .
k. s. thygesen and k. w. jacobsen , chem . phys . *
319 * , 111 ( 2005 ) .
y. hrn and j. linderberg , phys .
139 , a1063 ( 1965 ) .
p. sautet and c. joachim , chem .
lett . * 153 * , 511 ( 1988 ) .
c. patoux , c. coudret , j. p. launay , c. joachim and a. gourdon , inorg . chem .
* 36 * , 5037 ( 1997 ) .
m. mayor , h. b. weber , j. reichert , m. elbing , c. von hnisch , d. beckmann , and m. fischer , angew .
* 42 * , 5834 ( 2003 ) . c. a. coulson and h. c. longuet - higgins , proc .
lond . a * 191 * , 39 ( 1947 ) . c. a. coulson and h. c. longuet - higgins , proc .
a * 192 * , 16 ( 1947 ) . c. a. coulson and h. c. longuet - higgins , proc .
lond . a * 193 * , 447 ( 1948 ) . c. a. coulson and h. c. longuet - higgins , proc .
a * 195 * , 188 ( 1948 ) .
b. h. chirgwin and c. a. coulson , proc .
a * 201 * , 196 ( 1950 ) .
b. t. pickup , philos . mag .
b * 69 * , 799 ( 1994 ) .
t. stuyver , s. fias , f. de proft , p. w. fowler and p. geerlings , j. chem . phys . * 142 * , 094103 ( 2015 ) .
t. stuyver , s. fias , f. de proft and p. geerlings , chem .
. lett . * 630 * , 51 ( 2015 ) .
t. stuyver , s. fias , f. de proft and p. geerlings , j. phys .
c * 119 * , 26390 ( 2015 ) .
y. tsuji , r. hoffmann , m. strange and g. c. solomon , proc .
usa * 113 * , e413 ( 2016 ) .
s. datta , _ electronic transport in mesoscopic systems _ , ( cambridge univ . press , cambridge , uk , 1995 ) .
d. z. manrique , c. huang , m. baghernejad , x. zhao , o. a. al - owaedi , h. sadeghi , v. kaliginedi , w. hong , m. gulcur , t. wandlowski , m. r. bryce , c. j. lambert , nat . commun .
* 6 * , 6389 ( 2015 ) .
m. g. reuter and t. hansen , j. chem .
phys . * 141 * , 181103 ( 2014 ) .
r. stadler , m. forshaw and c. joachim , nanotechnology * 14 * , 138 ( 2003 ) .
r. stadler and t. markussen , j. chem . phys .
* 135 * , 154109 ( 2011 ) .
r. stadler , s. ami , m. forshaw and c. joachim , nanotechnology * 15 * , s115 ( 2004 ) . t. markussen , r. stadler and k. s. thygesen , nano lett .
* 10 * , 4260 ( 2010 ) .
t. markussen , r. stadler and k. s. thygesen , phys .
* 13 * , 14311 ( 2011 ) .
r. stadler , nano lett .
* 15 * , 7175 ( 2015 ) .
k. g. l. pedersen , a. borges , p. hedegrd , g. c. solomon and m. strange , j. phys .
c * 119 * , 26919 ( 2015 ) .
k. yoshizawa , t. tada and a. staykov , j. am .
soc . * 130 * , 9406 ( 2008 ) .
y. tsuji , a. staykov and k. yoshizawa , thin solid films * 518 * , 444 ( 2009 ) .
y. tsuji , a. staykov and k. yoshizawa , j. phys .
c * 113 * , 21477 ( 2009 ) .
x. li , a. staykov and k. yoshizawa , j. phys .
c * 114 * , 9997 ( 2010 ) .
y. tsuji , a. staykov and k. yoshizawa , j. am .
133 * , 5955 ( 2011 ) .
k. yoshizawa , acc .
45 * , 1612 ( 2012 ) .
j. h. van vleck and a. sherman , rev .
* 7 * , 167 ( 1935 ) . c. a. coulson and g. s. rushbrooke , math .
cambridge philos .
soc . * 36 * , 193 ( 1940 ) .
i. gutman and o. e. polansky , _ mathematical concepts in organic chemistry _
, p. 57 - 58 ( springer , 1986 ) , isbn : 978 - 3 - 642 - 70984 - 5
. s. larsson , j. am .
soc . * 103 * , 4034 ( 1981 ) . m. a. ratner , j. phys . chem . * 94 * , 4877 ( 1990 ) .
g. kastlunger and r. stadler , phys .
b * 89 * , 115412 ( 2014 ) . p. sautet and m .- l .
bocquet , phys .
b * 53 * , 4910 ( 1996 ) .
r. stadler , phys .
b 80 , 125401 ( 2009 ) .
m. h. garner , g. c. solomon and m. strange , j. phys .
c * 120 * , 9097 ( 2016 ) .
r. stadler , j. phys . : conf .
* 61 * , 1097 ( 2007 ) .
k. g. l. pedersen , m. strange , m. c. leijnse , p. hedegrd , g. c. solomon and j. paaske , phys .
b * 90 * , 125413 ( 2014 ) .
m. strange , j. s. seldenthuis , c. j. o. verzijl , j. m. thijssen and g. c. solomon , j. chem . phys .
* 142 * , 084703 ( 2015 ) .
y. tsuji , r. hoffmann , r. movassagh and s. datta , j. chem .
* 141 * , 224311 ( 2014 ) .
p. sautet and c. joachim , phys . rev .
b * 38 * , 12238 ( 1988 ) .
s. r. bahn and k. w. jacobsen , comput .
sci . eng . * 4 * , 56 ( 2002 ) .
transport.html http://www.hulis.free.fr/ s. j. van der molen and p. liljeroth , j. phys .
: cond . mat .
* 22 * , 133001 ( 2010 ) .
m. irie , chem . rev .
* 100 * , 1685 ( 2000 ) .
k. matsuda and m. irie , j. am .
soc . * 122 * , 7195 ( 2000 ) .
s. fraysse , c. coudret and j.p .
launay , eur .
, 1581 ( 2000 ) d. duli , s. j. van der molen , t. kudernac , h. t. jonkman , j. j. d. de jong , t. n. bowden , j. van esch , b. l. feringa , and b. j. van wees , phys .
* 91 * , 207402 ( 2003 ) . c. van dyck , v. geskin , a. j. kronemeijer , d. m. de leeuw and j. cornil , phys .
chem . phys . * 15 * , 4392 ( 2013 ) .
y. tsuji and r. hoffmann , angew .
int . ed . *
53 * , 4093 ( 2014 ) .
j. xia , b. capozzi , s. wei , m. strange , a. batra , j. r. moreno , r. j. amir , e. amir , g. c. solomon , l. venkataraman and l. m. campos , nano . lett . * 14 * , 2941 ( 2014 ) . f. schwarz , m. koch , g. kastlunger , k. venkatesan , h. berke , r. stadler and e. lrtscher , angew .
, in print ( 2016 ) ; doi : 10.1002/anie.201605559 c. m. gudon , h. valkenier , t. markussen , k. s. thygesen , j. c. hummelen and s. j. van der molen , nat . nanotechnol .
* 7 * , 305 ( 2012 ) . | destructive quantum interference ( dqi ) in single molecule electronics is a purely quantum mechanical effect and entirely defined by inherent properties of the molecule in the junction such as its structure and symmetry .
this definition of dqi by molecular properties alone suggests its relation to other more general concepts in chemistry as well as the possibility of deriving simple models for its understanding and molecular device design .
recently , two such models have gained wide spread attention , where one was a graphical scheme based on visually inspecting the connectivity of carbon sites in conjugated @xmath0 systems in an atomic orbital ( ao ) basis and the other one put the emphasis on the amplitudes and signs of the frontier molecular orbitals ( mos ) .
there have been discussions on the range of applicability for these schemes , but ultimately conclusions from topological molecular hamiltonians should not depend on whether they are drawn from an ao or a mo representation , as long as all the orbitals are taken into account . in this article
we clarify the relation between both models in terms of the zeroth order green s function and compare their predictions for a variety of systems . from this comparison
we conclude that for a correct description of dqi from a mo perspective it is necessary to include the contributions from all mos rather than just those from the frontier orbitals .
the cases where dqi effects can be successfully predicted within a frontier orbital approximation we show to be limited to alternant even - membered hydrocarbons , as a a direct consequence of the coulson - rushbrooke pairing theorem in quantum chemistry .
_ .. you have a program , for god s sake use it , play with it , do a calculation on any small problem related to your problem . let the calculations teach you .
they are so easy ! shall we stop teasing one another about mo and vb ? _
... quantum chemistry has given us two wonderful tools to reason about chemistry , and denying any one of them would impoverish our ability to reason . |
the physical vacuum is a complicated condensed matter @xcite . at the moment we know only the low - energy properties of this substance , i.e. the properties at energies much smaller than the planck energy scale , @xmath1 , where @xmath2 is newton s constant .
we know that at the low - energy edge our vacuum has many different symmetries : @xmath3 and @xmath4 gauge symmetries , lorentz invariance , general coordinate invariance , and dicrete cpt symmetry . with increasing energy more elements of symmetry
are added : @xmath5 symmetry of weak interactions , probably the gut symmetry and even supersymmetry .
when the temperature dicreases such symmetries become spontaneously broken . this is the traditional point of view , which is supported by the observation of the similar symmetry breaking in different ( many - body ) condensed matter systems : superfluids , superconductors , magnets , liquid and ordinary crystals .
however there is another and actually opposite point of view : all symmetries known in the universe spontaneously ( but without any phase transition ) appear at the low - energy corner@xcite . they become more and more pronounced the lower the energy .
the symmetries disappear at higher energies when the planck energy is approached .
this conjecture is also supported by the condensed matter analogy . here
we discuss the topological origin of such enhanced symmetry in the infrared limit .
it is caused by the so - called topologically stable fermi points in the spectrum of the fermionic system .
we show that if such points exist in quantum condensed matter , then at low enough temperature this system exhibits fully or partially the lorentz invariance , general covariance , and gauge invariance
. moreover the collective modes which describe such a system at low @xmath6 all are represented by the chiral relativistic fermions , gauge fields , and gravity .
all are the low - energy phenomena , which are absent at higher energies , where the condensed matter has only a very limited set of the global symmetries . from the second point of view some directions in physics
look artificial . in particular , since the gravity exists only as the infrared phenomenon one should not quantize gravity : only low - energy gravitons can be quantized @xcite .
the same concerns the multidimensional string theories , which also give rise to gravitation in the infrared limit .
the fermi point mechanism does not require a high dimensionality for the space - time : the topologically stable fermi point is just a property of the conventional 3 + 1 dimensional space - time .
the fermi point is a particular case of the topologically stable manifolds of zeroes in the fermionic spectrum .
let us start with the simplest such manifold the fermi surface .
the fermi surface appears in the noninteracting fermi gas , where the energy spectrum of fermions is @xmath7 and @xmath8 is the chemical potential .
the fermi surface bounds the volume in the momentum space where the energy is negative , @xmath9 , and where the particle states are all occupied at @xmath10 . in this isotropic model
the fermi surface is a sphere of radius @xmath11 .
it is important that the fermi surface survives even if interactions between particles are introduced .
such stability of fermi surface comes from the topological property of the feynman quantum mechanical propagator
the one - particle green s function @xmath12 let us write the propagator for a given momentum @xmath13 and for the imaginary frequency , @xmath14 .
the imaginary frequency is introduced to avoid the conventional singularity of the propagator `` on the mass shell '' , i.e. at @xmath15 . for noninteracting particles
the propagator has the form @xmath16 obviously there is still a singularity : on the 2d hypersurface @xmath17 in the 4-dimensional space @xmath18 the propagator is not well defined .
this singularity is stable , i.e. can not be eliminated by small perturbations .
the reason is that the phase @xmath19 of the green s funtion @xmath20 changes by @xmath21 around the path embracing this 2d hypersurface in the 4d - space ( see fig .
[ fermisurfaceasvortex ] ) .
the phase winding number @xmath22 can not change continuously , that is why it is robust towards any perturbation .
thus the singularity of the green s funtion on the 2d - surface in the momentum space is preserved , even when interactions between particles are introduced .
exactly the same topological conservation of the winding number leads to the stability of the quantized vortex in superfluids and superconductors , the only difference being that , in the case of vortices , the phase winding occurs in the real space , instead of the momentum space .
the complex order parameter @xmath23 changes by @xmath24 around the path embracing vortex line in 3d space or vortex sheet in 3 + 1 space .
the connection between the topology in real space and the topology in momentum space is , in fact , even deeper ( see _ e.g. _ ref.@xcite ) . if the order parameter depends on space - time , the propagator in semiclassical aproximation depends both on 4-momentum and on space - time coordinates @xmath25 .
the topology in the 4 + 4 dimensional space describes : ( i ) the momentum space topology in the homogeneous system ; ( ii ) the defects of the order parameter in real space ; and ( iii ) topology of the energy spectrum within the topological defects @xcite . in the more complicated cases ,
when the green s function is the matrix with spin and band indices , the phase of the green s function becomes meaningless . in this case
one should use a general analytic expression for the integer topological invariant which is responsible for the stability of the fermi surface : @xmath26 here the integral is taken over an arbitrary contour @xmath27 in the momentum space @xmath28 , which encloses the fermi hypersurface ; and @xmath29 is the trace over the spin and band indices .
the topological class of systems with fermi surface is rather broad .
in particular it contains the conventional landau fermi - liquid , in which the propagator preserves the pole .
close to the pole the propagator is @xmath30 evidently the residue @xmath31 does not change the topological invariant for the propagator , eq.([invariantforfs ] ) , wihch remains @xmath32 .
this is essential for the landau theory of an interacting fermi liquid ; it confirms the assumption that in fermi liquids the spectrum of quasiparticles at low energy is similar to that of particles in a fermi gas .
it is also important for the consideration of the bosonic collective modes of the landau fermi - liquid .
the interaction between the fermions can not not change the topology of the fermionic spectrum , but it produces the effective field acting on a given particle by the other moving particles .
this effective field can not destroy the fermi surface owing to its topological stability , but it can locally shift the position of the fermi surface
. therefore a collective motion of the particles is seen by an individual quasiparticle as a dynamical mode of the fermi surface .
these bosonic oscillative modes are known as the different harmonics of the zero sound .
an example is shown in the upper part of fig.[collectivemodes ] .
note that the fermi hypersurface exists for any spatial dimension .
in the 2 + 1 dimension the fermi hypersurface is a line the vortex line in the 2 + 1 momentum space as is shown in fig . [
fermisurfaceasvortex](a ) . in the 1 + 1 dimension the green s function looses its pole , but nevertheless the fermi surface is still there @xcite . in 1 + 1
dimension the landau fermi liquid transforms to another states within the same topological class .
example is provided by the luttinger liquid .
close to the fermi surface the green s function for the luttinger liquid can be approximated as ( see @xcite ) @xmath33 where @xmath34 and @xmath35 correspond to fermi velocities of spinons and holons and @xmath36 .
the above equation is not exact but reproduces the topology of the green s function in luttinger fermi liquid . if @xmath37 and @xmath38 , the singularity in the @xmath39 momentum space occurs on the fermi surface , i.e. at @xmath40 .
the topological invariant in eq.([invariantfor fs ] ) remains the same @xmath32 , as for the conventional landau fermi - liquid .
the difference from landau fermi liquid occurs only at real frequency @xmath41 : the quasiparticle pole is absent and one has the branch cut singularities instead of the mass shell , so that the quasiparticles are not well defined .
the population of the particles has no jump on the fermi surface , but has a power - law singularity in the derivative @xcite .
another example of the non - landau fermi liquid is the fermi liquid with exponential behavior of the residue @xcite .
it also has the fermi surface with the same topological invariant , but the singularity at the fermi surface is exponentially weak . in this section
we do not consider the bifurcations which lead to the overall reconstruction of the particle spectrum , which occur , say , during the transition to superconducting state .
we assume that the temperature , though is small compared to the `` planck '' scale @xmath42 , is still above the superfluid transition : @xmath43 . but even in the superconducting region @xmath44 the winding number for the conventional green s function is preserved : the invariant in eq.([invariantfor fs ] ) is again @xmath32 . only instead of pole at @xmath45 in the green s function of the normal state
, one has zero at @xmath45 in the superconducting state : @xmath46 where @xmath47 is the gap in the quasiparticle spectrum .
this demonstrates the stability of the singularity in the green s function , though the fermi surface as the manifold of zeroes in the energy spectrum disappears in the superconductor .
although the systems we have discussed contain fermionic and bosonic quantum fields , this is not the relativistic quantum field theory which we need for the simulation of quantum vacuum : there is no lorentz invariance and the oscillations of the fermi surface do not resemble the gauge field even remotely .
the situation is somewhat better for superfluids and superconductors with fully gapped spectra ; examples , which provide useful analogies with dirac fermions and spontaneously broken symmetry in quantum fields , are conventional superconductors @xcite and superfluid @xmath0he - b @xcite . in @xmath0he -
b the hamiltonian of free bogoliubov quasiparticles is the @xmath48 matrix : @xmath49 where the pauli @xmath50 matrices @xmath51 describe the conventional spin of fermions and @xmath50 matrices @xmath52 describe the bogoliubov isospin in the particle - hole space .
the bogoliubov hamiltonian asymptotically approaches the dirac one in the low @xmath53 limit .
this is however is not the low energy limit , which for the typical @xmath0he - b parameters occurs for @xmath53 close to @xmath54 .
the low @xmath53 and low @xmath55 limits coincide at large negative @xmath56 , which is not the case in @xmath0he - b .
nevertheless the @xmath0he - b also serves as a model system for simulations of many phenomena in particle physics and cosmology including experimental verification @xcite of the kibble mechanism describing formation of cosmic strings in the early universe @xcite . note that even for the fully gapped systems there can exist the momentum - space topological invariants , which characterize the vacuum states .
typically this occurs in 2 + 1 systems , e.g. in the 2d electron system exhbiting quantum hall effect @xcite ; in thin film of @xmath0he - a ( see @xcite and sec.9 of ref.@xcite ) ; and in the 2d superconductors with broken time reversal symmetry @xcite .
the quantum ( lifshitz ) transition between the states with different topological invariants occurs through the intermediate gapless regime @xcite .
now we proceed to the topological class which most fully exhibits the fundamental properties needed for a realization of the relativistic quantum fields , analogous to those in particle physics and gravity .
it is the class of systems , whose representatives are superfluid
@xmath0he - a and the vacuum of relativistic left - handed and right - handed chiral fermions ( another example of this class in condensed matter has been discussed for gapless semiconductors @xcite ) .
this class is characterized by points in the momentum space where the ( quasi)particle energy is zero . in particle physics
the energy spectrum @xmath57 is characteristic of the massless neutrino ( or any other chiral lepton or quark in the standard model ) with @xmath58 being the speed of light .
the energy of a neutrino is zero at point @xmath59 in the 3d momentum space .
the hamiltonian for the neutrino the massless spin-1/2 particle is a @xmath50 matrix @xmath60 which is expressed in terms of the pauli spin matrices @xmath61 .
the sign @xmath62 is for a right - handed particle and @xmath63 for a left - handed one : the spin of the particle is oriented along or opposite to its momentum , respectively .
the bogoliubov matrix for the @xmath0he - a fermions is @xmath64 where @xmath65 and @xmath66 are real vectors , which in equilibrium or ground state are unit orthogonal vectors .
the energy of the fermions @xmath67 is zero at 2 points , at @xmath68 .
let us show that zeroes in the spectrum of the chiral fermions and in the spectrum of @xmath0he - a fermions are topologically stable and described by the same topological invariant .
let us again consider the propagator of the particle @xmath69 on the imaginary frequency axis , @xmath14 .
one can see that this propagator still has a singularity , which is now not on the surface but at the points in the 4d momentum space : @xmath70 and @xmath71 for chiral fermions and @xmath0he - a fermions respectively .
an integer topological invariant which supports the stability of the point zeroes is expressed in terms of the propagator in the following way @xcite : @xmath72 where @xmath73 is the 3-dimensional surface embracing the point node in the 4-momentum space . for the chiral fermions in eq.([neutrino ] ) this invariant is @xmath74 , where the sign is determined by the chirality of the fermion .
the meaning of this topological invariant can be easily visualized .
let us consider the behavior of the particle spin @xmath75 as a function of its momentum @xmath13 in the 3d - space @xmath76 . for right - handed particles @xmath77 ,
while for left - handed ones @xmath78 . in both cases
the spin distribution in the momentum space looks like a hedgehog ( see fig .
[ fermisurfaceasvortex]b ) , whose spines are represented by spins : spines point outward for the right - handed particle and inward for the left - handed one . in the 3d - space
the hedgehog is topologically stable . for the @xmath0he - a fermions in eq.([aphasefermions1 ] )
@xmath79 , where the sign is determined by the position of the node , @xmath80 .
the topological invariant is twice larger because of the double degeneracy of the fermi point over the conventional spin of the @xmath0he atom . for each projection of spin
one has @xmath81 .
note that the bogoliubov spin @xmath82 in @xmath0he - a plays the same role as the conventional spin @xmath83 of chiral fermions in eq.([neutrino ] ) .
on the other hand the conventional spin of the @xmath0he atom is responsible for the degeneracy , but not for chirality , and thus plays the part of the isospin ( see also sec.[degeneracy ] ) .
this means that the origin of the spin responsible for the chirality of the ( quasi)particle is fully determined by the matrix structure of the fermi point . in this sense
there is no principle difference between spin and isospin : changing the matrix structure one can convert isospin to spin , while the topological charge of the fermi point remains invariant . the high - temperature superconductors in cuprates most probably contain zeroes in their quasiparticle energy spectrum .
the arpes experiments @xcite show that these are four lines in the 3d momentum space where the quasiparticle energy is zero or , equivalently , four point zeroes in the 2d cuo@xmath84 planes .
the high - t superconductors thus belong to class of systems with fermi lines : the dimension @xmath85 of the manifold of zeroes is @xmath86 , which is intermediate between a fermi surface with @xmath87 and a fermi point with @xmath88 .
the energy spectrum of quasiparticles near each of the 4 gap nodes can be written as @xmath89 the `` speeds '' of light @xmath90 and @xmath91 are the `` fundamental '' characteristics determined by the microscopic physics of the cuprates .
this means that the system belongs to the same class as 2 + 1 qft with massless relativistic fermions .
as in the other two classes of fermionic systems with the gapless quasiparticles , all low - energy ( infrared ) properties of cuprate superconductors are determined by zeroes .
in particular , the density of the fermionic states is determined by the dimension of the zeroes : @xmath92 many low - temperature properties of these superconductors are obtained from a simple scaling arguments .
for example , an external magnetic field @xmath93 has dimension of @xmath94 and thus of @xmath95 . at finite @xmath93 , the density of states is nonzero even at @xmath96 .
substituting @xmath97 to eq.([dosgeneral ] ) one obtains @xmath98 and the following scaling law for the heat capacity : @xmath99 where @xmath100 is some function with the known asymptotes ( see @xcite ) .
an experimental indication of such scaling with @xmath101 was reported for yba@xmath84cu@xmath102o@xmath103 in ref .
@xcite .
the lines of zeroes generally have no stability : there is no corresponding @xmath104 invariant , which can support the topological stability . the singular line in the momentum space from which the spines ( now the vector @xmath82 ) point outward ( see fig .
[ fermisurfaceasvortex]c ) can be elmininated by the escape of the @xmath82-vector to a third dimension .
this can be accomplished by an operation similar to the folding of an umbrella ( see fig .
[ fermisurfaceasvortex]d ) .
existence of the nodal lines can be prescribed , however , by the symmetry of the ground state .
there are many nontrivial classes of superconductors , whose symmetry supports the existence of nodal lines in symmetric positions in the momentum space @xcite .
the symmetry violating perturbations , such as impurities , an external magnetic field , etc .
, destroy the lines of zeroes @xcite .
one could expect different types of transformations of these lines of zeroes which depend on the perturbation .
impurities , for example , can : ( i ) produce the gap in the fermionic spectrum @xcite ( see fig.[fermisurfaceasvortex]d ) , which corresponds to appearance of mass for the 2 + 1 relativistic fermions ; ( ii ) lead to the finite density of states @xcite , thus transforming the system to class ( 1 ) ; ( iii ) produce zeroes of fractional dimension , which means that the exponent in the density of states @xmath105 is non - integral @xcite and thus corresponds to a fractional @xmath85 of the manifold of zeroes ; and ( iv ) lead to localization @xcite .
an open question is : can the quantum fluctuations do the same , in particular , can they change the effective dimension of the zeroes ?
close to the fermi point @xmath106 in the 4d space one can expand the propagator in terms of the deviations from this fermi point , @xmath107 .
if the fermi point is not degenerate , the general form of the propagator is @xmath108 here we returned back from the imaginary frequency axis to the real energy , so that @xmath109 instead of @xmath14 ; and @xmath110 .
the quasiparticle spectrum @xmath111 is given by the poles of the propagator : @xmath112 thus in the vicinity of the fermi point the massless quasiparticles are described by the lorentzian metric @xmath113 .
it is most important that this is the general form of the energy spectrum in the vicinity of any fermi point , even if the underlying fermi system is not lorentz invariant ; superfluid @xmath0he - a is an example . the fermionic spectrum necessarily becomes lorentz invariant near the fermi point , _
i.e._. if one applies this reasoning to our quantum vacuum , one may conclude that possibly the observed lorentz invariance of the physical laws is not a fundamental but a low - energy property of the vacuum which results from the existence of the topologically stable fermi points .
let us consider the collective modes in such a system .
the effective fields acting on a given particle due to interactions with other moving particles can not destroy the fermi point .
that is why , under the inhomogeneous perturbation of the fermionic vacuum the general form of eqs.([generalpropagator]-[generalenergy ] ) is preserved .
however the perturbations lead to a local shift in the position of the fermi point @xmath106 in momentum space and to a local change of the vierbein @xmath114 ( which in particular includes slopes of the energy spectrum ( see fig .
[ collectivemodes ] ) .
this means that the low - frequency collective modes in such fermi liquids are the propagating collective oscillations of the positions of the fermi point and of the slopes at the fermi point .
the former is felt by the right- or the left - handed quasiparticles as the dynamical gauge ( electromagnetic ) field , because the main effect of the electromagnetic field @xmath115 is just the dynamical change in the position of zero in the energy spectrum : in the simplest case @xmath116 .
the collective modes related to a local change of the vierbein @xmath114 correspond to the dynamical gravitational field .
the quasiparticles feel the inverse tensor @xmath117 as the metric of the effective space in which they move along the geodesic curves @xmath118 therefore , the collective modes related to the slopes play the part of the gravity field ( see fig .
[ collectivemodes ] ) . thus near the fermi point
the quasiparticle is the chiral massless fermion moving in the effective dynamical electromagnetic and gravitational fields .
another important property which results from the above equation is that the fermionic propagator in eq.([generalpropagator ] ) is gauge invariant and even obeys the general covariance near the fermi point .
for example , the local phase transformation of the wave function of the fermion , @xmath119 can be compensated by the shift of the `` electromagnetic '' field @xmath120 .
these attributes of the electromagnetic ( @xmath121 ) and gravitational ( @xmath113 ) fields also arise spontaneously as the low - energy phenomena .
now let us discuss the dynamics of collective bosonic modes , @xmath121 and @xmath113 . since these are the effective fields their motion equations do not necessarily obey gauge inariance and general covariance .
however , in some special cases such symmetries can arise in the low energy corner .
what are the conditions for that ?
the effective lagrangian for the collective modes is obtained by integrating over the vacuum fluctuations of the fermionic field .
this principle was used by sakharov and zeldovich to obtain an effective gravity @xcite and effective electrodynamics @xcite , both arising from fluctuations of the fermionic vacuum .
if the main contribution to the effective action comes from the vacuum fermions whose momenta @xmath13 are concentrated near the fermi point , _ i.e. _ where the fermionic spectrum is linear and thus obeys the `` lorentz invariance '' , the result of the integration is necessarily invariant under gauge transformation , @xmath120 , and has a covariant form .
the obtained effective lagrangian then gives the maxwell equations for @xmath121 and the einstein equations for @xmath117 , so that the propagating bosonic collective modes do represent the gauge bosons and gravitons .
thus two requirements must be fulfilled
( i ) the fermionic system has a fermi point and ( ii ) the main physics is concentrated near this fermi point . in this case
the system acquires at low energy all the properties of the modern quantum field theory : chiral fermions , quantum gauge fields , and gravity .
all these ingredients are actually low - energy ( infra - red ) phenomena .
there is another important symmetry obeyed by massless relativistic weyl fermions , the conformal invariance
the invariance under transformation @xmath122 . in the extreme limit when the vacuum fermions are dominatingly relativistic
, the effective action for gravity must be conformly invariant .
such gravity , the so - called weyl gravity , is a viable rival to einstein gravity in modern cosmology @xcite : the weyl gravity ( i ) can explain the galactic rotation curves without dark matter ; ( ii ) it reproduces the schwarzschild solution at small distances ; ( iii ) it can solve the cosmological constant problem , since the cosmological constant is forbidden if the conformal invariance is strongly obeyed ; etc .
( see @xcite ) .
let us consider what happens in a practical realization of systems with fermi points in condensed matter in @xmath0he - a .
close to the gap nodes , _
i.e. _ at energies @xmath123 , where @xmath47 is the maximal value of the gap in @xmath0he - a which plays the part of the planck energy , the quasiparticles do obey the relativistic equation @xmath124 here @xmath125 is the `` electric charge '' and simultaneously the chirality of the quasiparticles .
let us consider the simplest situation , when the @xmath0he - a is in its vacuum manifold , which is characterized by two unit mutually orthogonal vectors @xmath126 and @xmath127 , and neglect the spin degeneracy of the fermi point . when we consider the low - energy collective modes , these vectors are slowly changing in space - time .
in this situation the effective metric and effective electromagnetic field are given by : @xmath128 where @xmath129 is the superfluid velocity given by @xmath130 .
from above equations it follows that the fields , which act on the `` relativistic '' quasiparticles as electromagnetic and gravitational fields , have very strange behavior . for example , the same texture of the @xmath131-vector is felt by quasiparticles as the effective magnetic field @xmath132 according to eq.([emfieldinaphase ] ) and simultaneously it enters the metric according to eq.([gravfieldinaphase ] ) .
such field certainly can not be described by the maxwell and einstein equations together .
actually the gravitational and electromagnetic variables coincide in @xmath0he - a only when we consider the vacuum manifold : outside of this manifold they split .
@xmath0he - a , as any other fermionic system with fermi point , has enough number of collective modes to provide the analogs for the independent gravitational and electromagnetic fields . but some of these modes are massive in @xmath0he - a .
for example the gravitational waves correspond to the modes , which are different from the oscillations of the @xmath131-vector . as distinct from the photons ( orbital waves propagating oscillations of the @xmath131-vector ) the gravitons are massive @xcite .
all these troubles occur because in @xmath0he - a the main contribution to the effective action for bosonic fields come from the vacuum fermions at the `` planck '' energy scale , @xmath133 .
these fermions are far from the fermi points and their spectrum is nonlinear .
that is why in general the effective action is not symmetric .
there are , however , exclusions , for example , the action for the @xmath131-field contains the term with the logarithmically divergent factor @xmath134 @xcite .
it comes from the zero charge effect , experienced by the vacuum of the massless fermions for whom the @xmath131-field acts as electromagnetic field . due to its logarithmic divergence
this term is dominanting at low frequency @xmath135 : the lower the frequency the larger is the contribution of the vacuum fermions from the vicinity of the fermi point and thus the more symmetric is the lagrangian for the @xmath136-field . as a result , in the very low - energy limit , when the non - logarithmic contributions can be neglected , the effective lagrangian for the @xmath121 becomes gauge invariant and even obeys the general covariance : @xmath137 where @xmath138 is the strength of the effective electromagnetic field @xmath121 from eq.([emfieldinaphase ] ) and @xmath113 is the effective gravitational field from eq.([gravfieldinaphase ] ) . in this regime
the @xmath121 field does obey the maxwell equations in a curved space . on the other hand the `` einstein '' action for @xmath117 is highly contaminated by many noncovariant terms , which come from the integration over the `` nonrelativistic '' high energy degrees of freedom at `` planck '' scale . in this sense
the @xmath0he - a , with its given physical parameters , is not a perfect model for quantum vacuum . to remove the polluting noncovariant terms
, the integration must be spontaneously cut - off at energies much below the `` planck '' scale , @xmath139 , for example , due to strong quasiparticle relaxation .
the main reason why @xmath0he - a is not a good substance , is that the fermi points of the left particles , i.e. at @xmath140 , and the fermi points of the right particles , i.e. at @xmath141 , are far from each orher in equilibrium .
the `` perfect '' condensed matter would be such where all the fermi points are at the origin , at @xmath142 , as it happens in the standard model .
however , inspite of the absence of general covariance even in the low - energy corner , many different properties of the physical vacuum with a fermi point , whose direct observation are still far from realization , can be simulated in @xmath0he - a .
one of them is the chiral anomaly .
the chiral anomaly is the phenomenon which allows the nucleation of the fermionic charge from the vacuum @xcite .
such nucleation results from the spectral flow of the fermionic charge through the fermi point to high energy .
since the flux in the momentum space is conserved , it can be equally calculated in the infrared or in the ultraviolet limits . in @xmath0he -
a it is much easier to use the infrared regime , where the fermions obey all the `` relativistic '' symmetries . as a result one obtains the same anomaly equation , which has been derived by adler and by bell and jackiw for the relativistic systems .
the rate of production of quasiparticle number @xmath143 from the vacuum in applied electric and magnetic fields is @xmath144 here @xmath145 and @xmath146 is the density of the right and left quasiparticles ; @xmath147 and @xmath148 are their charges ; and @xmath149 is the dual field strength . note that the above equation is fully `` relativistic '' .
this equation has been verified in @xmath0he - a experiments @xcite , where the `` magnetic '' @xmath132 and `` electric '' @xmath150 fields have been simulated by the space and time dependent @xmath136-texture . in particle physics
the only evidence of axial anomaly is related to the decay of the neutral pion @xmath151 , although the anomaly is much used in different cosmological scenaria explaining an excess of matter over antimatter in the universe ( see review @xcite ) . in @xmath0he -
a the fermi point ( say , at the north pole ) is doubly degenerate owing to the ordinary spin @xmath83 of the @xmath0he atom .
this means that in equilibtium the two zeroes , each with the topological invariant @xmath152 , are at the same point in momentum space .
let us find out what can be the consequences of the fermi point degeneracy .
it is clear that the collective motion can split the fermi points : positions of the two points can oscillate separately .
moreover , since the propagator is now the @xmath48 matrix there can be the cross terms .
if we neglect the degrees of freedom related to the vierbein then the collective variables of the system with the doubly degenerate fermi point enter the fermionic propagator as @xmath153 the new effective field @xmath154 acts on the chiral quasiparticles as @xmath5 gauge field .
thus in this effective field theory the ordinary spin of the @xmath0he atoms plays the part of the weak isospin @xcite .
the `` weak '' field @xmath154 is also dynamical and in the leading logarithmic order obeys the maxwell ( actually yang - mills ) equations .
this implies that the higher symmetry groups of our vacuum can in principle arise as a consequence of the fermi point degeneracy .
for example , the 4-fold degeneracy of the fermi point ( which implies that there are , say , 4 left - handed fermionic species , or 2 left - handed + 2 right - handed ) can produce the @xmath155 gauge group . in particle physics
the collective modes related to the shift of the 4-momentum are discussed in terms of the `` generalized covariant derivative '' @xcite . in this theory the gauge fields , the higgs fields , and yukawa interactions , all
are realized as shifts of positions of the degenerate fermi point , with degeneracy corresponding to different quarks and leptons . in the eq.([nonabelian fieldpropagator ] ) we did not take into account that dynamically the vierbein can also oscillate differently for each of the two elementary fermi points . as a result
the number of the collective modes could increase even more .
this is an interesting problem which must be investigated in detail .
if the degenerate fermi point mechanism has really some connection to the dynamical origin of the non - abelian gauge fields , we must connect the degeneracy of the fermi point ( number of the fermionic species ) with the symmetry group of the gauge fields .
naive approach leads to extremely high symmetry group .
that is why there should be some factors which can restrict the number of the gauge and other bosons .
for example there can be some special discrete symmetry between the fermions of the degenerate point , which restricts the number of massless bosonic collective modes .
another source of the reduction of the number of the effective field has been found by chadha and nielsen @xcite .
they considered the massless electrodynamics with different metric ( vierbein ) for the left - handed and right - handed fermions . in this model
the lorentz invariance is violated .
they found that the two metrics converge to a single one as the energy is lowered .
thus in the low - energy corner the lorentz invariance becomes better and better , and at the same time the number of independent bosonic modes decreases .
since the connections between qft in the standard model and in @xmath0he - a has been discussed in the earlier publications @xcite , we concentrate here on some problems related to gravitational analogy .
as we have seen from eq.([gravfieldinaphase ] ) the gravitational field can be simulated in @xmath0he - a by the motion of the liquid with the superfluid velocity @xmath156 and by the @xmath131-texture @xcite . in this section
we consider the situation when the @xmath131-vector is uniform and thus does not produce the effective gravitational field , so that the gravitational effects come only from the motion of the superfluid vacuu .
the propagation of fermions in the moving liquid obeys the same equation as propagation of relativistic particles in the gravitational field .
the same happens for propagating sound waves in normal fluids @xcite and phonons in superfluid @xmath157he . in the simplest case of the radial motion of superfluid @xmath157he
, the effective metric is expressed in terms of the radial superfluid velocity @xmath158 as @xmath159 where @xmath58 now is a speed of sound in @xmath157he ( phonon velocity ) . for @xmath0he - a fermions
the spherically symmetric metric occurs if @xmath131-field is radial .
then it follows from eq.([gravfieldinaphase ] ) @xmath160 kinetic energy of superflow plays the part of the gravitational potential : @xmath161 .
if one chooses the velocity field corresponding to the potential of the point body of mass @xmath162 @xmath163 one obtains the panlev - gullstrand form of schwarzschild geometry ( see e.g. ref.@xcite ) . here
@xmath164 denotes the position of the event horizon , where the velocity reaches the `` speed of light '' @xmath58 . if the fluid moves towards the origin , the low - energy quasipartilcles are trapped behind the horizon , since their speed @xmath58 with respect to the fluid is less than the velocity of the fluid .
such sonic black hole was first suggested by unruh for ordinary liquid @xcite . however
since all the known normal liquids are classical , the most interesting quantum effects related to the horizon can not be simulated in such flow .
also the geometry is such that it can not be realized : in such radial flow inward the liquid is accumulated at the origin , so that this sonic black hole can not be stationary . in the other scenario a horizon appears in moving solitons , if the velocity of the soliton exceeds the local `` speed of light '' @xcite .
this scenario has the same drawback : in finite system the motion of the soliton can not be supported for a long time . in a draining bathtub geometry suggested in ref.@xcite the fluid motion
can be made constant in time .
however the friction of the liquid , which moves through the drain , is the main source of dissipation .
the superfluidity of the liquid does not help much in this situation .
horizon does not appear since the `` superluminal '' motion with respect to the boundaries of the drain tube is not allowed : the superflow becomes unstable and superfluidity collapses ( see @xcite ) .
let us suggest a scenario , in which this collapse is avoided .
the superfluid motion becomes quasi - stationary and exhibits the event horizon ; the life time of the flow state is determined by hawking radiation .
the stationary black hole can be realized in the following geometry , which is the development of the bathtub geometry of ref.@xcite ( see fig . [ 2dbh](a ) ) .
the superfluid @xmath0he - a film is moving towards the center of the disk , where it escapes to the third dimension due to the orifice ( hole ) .
if the thickness of the film is constant , the flow velocity increases towards the center as @xmath165 and at @xmath166 reaches the speed of light ( now @xmath167 denotes the radial coordinate in the cylindrical system ) . outside the orifice
the motion of the liquid is two dimensional and the effective metric for the low - energy bogoliubov quasiparticles is @xmath168 here we took into account that the @xmath169 vector in the film is fixed along the normal to the film , so that the `` speed of light '' for quasiparticles propagating along the film @xmath170cm / sec is much smaller than the fermi velocity @xmath171 which corresponds to the `` speed of light '' for quasiparticles propagating along the normal to the film . note that @xmath58 is much smaller than the speed of sound in @xmath0he - a , that is why the motion of fluid has no effect on the density of the liquid . the important element of the construction in fig . [ 2dbh ]
is that the moving superfluid @xmath0he - a film is placed on the top of the superfluid @xmath157he film .
this is made to avoid the interaction of the @xmath0he - a film with the solid substrate .
the superfluid @xmath157he film effectively screens the interaction and thus prevents the collapse of the superfluid flow of @xmath0he - a with `` superluminal '' velocity : since the interactions with walls is removed the local observer moving with the superfluid velocity can not detect the relative motion with respect to the wall .
the motion of the superfluid @xmath0he - a with respect to superfluid @xmath157he film is not dangerous : the superfluid @xmath157he is not excited even if @xmath0he - a moves with its superluminal velocity : @xmath58 for @xmath0he - a is much smaller than the landau velocity for radiation of quasiparticles in superfluid @xmath157he , which is about @xmath172 m / sec . in this consideration
we neglected the radiation of surface waves , assuming that the thickness of @xmath157he film is small enough .
finally one can close the superflow by introducing the toroidal geometry in fig . [
2dbh](b ) , so that the superfluid condensate can circulate . in this case in addition to the black hole horizon
the white hole horizon appears on the path where the superfluid @xmath0he - a flows out from the orifice .
since the extrinsic mechanism of the friction of @xmath0he - a film the scattering of quasiparticles on the roughness of substrate is abandoned , we can consider now intrinsic mechanisms of dissipation .
the most interesting one is the hawking radiation related to existence of a horizon .
let us consider the simplest case of the 2d motion along the film in the bathtub geometry of fig .
[ 2dbh](a ) .
this can be easily generalized to the motion in the torus geometry .
there are two important reference frames : ( i ) the frame of the observer , who is locally comoving with the superfluid vacuum . in this frame
the local superfluid velocity is zero , @xmath173 , so that the energy spectrum of the bogoliubov fermions in the place of the observer is ( here we assume a pure 2d motion along the film ) @xmath174 in this geometry , in which the superflow velocity is confined in the plane of the film , the speed @xmath58 coincides with the landau critical velocity of the superfluid vacuum , @xmath175 . the vacuum as determined by the comoving observer is shown on fig . [
vacua](a ) : fermions occupy the negative energy levels in the dirac sea ( the states with the minus sign in eq.([energycomovingframe ] ) . it is the counterpart of the minkowski vacuum , which is however determined only locally : the comoving frame can not be determined globally .
moreover for the comowing observer , whose velocity changes with time , the whole velocity field @xmath176 of the superflow is time dependent . this does not allow to determine the energy correctly .
\(ii ) the energy can be well defined in the laboratory frame ( the rest frame ) . in this frame
the system is stationary , though is not static : the effective metric does not depend on time , so that the energy is conserved , but this metric contains the mixed component @xmath177
. the energy in the rest frame is obtained from the local energy in the comoving frame by the doppler shift : @xmath178 in case of the radial superflow @xmath179 one has @xmath180 figs .
[ vacua](b - c ) show how the `` minkowski '' vacuum of the comoving frame is seen by the rest observer ( note that the velocity is negative , @xmath181 ) . in the absense of horizon , or outside the horizon the local vacuum does not change : the states which are occupied ( empty ) in the minkowski vacuum remain occupied ( empty ) in the rest frame vacuum ( see fig .
[ vacua](b ) ) . in the presense of horizon behind which the velocity of superflow exceeds the landau critical velocity
the situation changes : behind the horizon the vacuum in the rest frame differs from that in the comoving frame .
let us for simplicity consider the states with zero transverse momentum @xmath182 on the branch @xmath183 in the rest frame .
if the system is in the minkowski vacuum state ( i.e. in the ground state as viewed by comoving observer ) , quasiparticles on this branch has reversed distribution in the rest frame : the negative energy states are empty , while the positive energy states are occupied ( see fig .
[ vacua](c ) ) . for this branch
the particle distribution corresponds to the negative temperature @xmath184 behind horizon .
since the energy in the rest frame is a good quantum number , the fermions can tunnel across the horizon from the occupied levels to the empty ones with the same energy .
thus if the system is initially in the minkowski vacuum in the comoving frame , the tunneling disturbs this vacuum state : pairs of excitations are created : the quasiparticle , say , is created outside the horizon while its partner the quasihole is created inside the horizon .
this simulates the hawking radiation from the black hole . to estimate the tunneling rate in the semiclassical approximation ,
let us consider the classical trajectories @xmath185 of particles , say , with positive energy , @xmath186 , for the simplest case when the transverse momentum @xmath187 is zero fig .
[ tunnelingtraj ] .
the branch @xmath188 describes the incoming particles with @xmath189 which propagate through the horizon to the orifice ( or to the singularity at @xmath190 , if the orifice is infinitely small ) without any singularity at the horizon .
the classical trajectories of these particles are @xmath191 the energies of these particles viewed by the comoving observer are also positive : @xmath192 .
another branch @xmath183 in fig .
[ tunnelingtraj ] contains two disconnected pieces describing the particle propagating from the horizon in two opposite directions : @xmath193 the eq.([tunnelingtrajectory2 ] ) describes the outgoing particles the particles propagating from the horizon to the exterior .
the energy of these particles is positive in both frames , comoving and rest .
the eq.([tunnelingtrajectory1 ] ) describes the propagation of particles from the horizon to the orifice ( or to the singularity ) .
though for the rest frame observer the energy of these particles is positive , these particles , which live within the horizon , belong to the minkowski vacuum in the comoving frame . the classical trajectory in eqs.([tunnelingtrajectory2],[tunnelingtrajectory1 ] ) is thus disrupted at the horizon .
there is however a quantum mechanical transition between the two pieces of the branch : the quantum tunneling .
the tunneling amplitude can be found in semiclassical approximation by shifting the contour of integration to the complex plane : @xmath194 this means that the wave function of any particle in the minkowskii vacuum inside the horizon contains an exponentially small part describing the propagation from the horizon to infinity .
this corresponds to the radiation from the minkowski vacuum in the presence of the event horizon .
the exponential dependence of the probability on the quasiparticle energy @xmath195 suggests that this radiation looks as thermal .
the corresponding temperature , the hawking temperature , is @xmath196 the radiation leads to the quantum friction : the linear momentum of the flow decreases with time .
this occurs continuously until the superfluid minkowski vacuum between the horizons is completely exhausted and the superfluid state is violated .
this leads to the phase slip event , after which the number @xmath197 of circulation quanta of superfluid velocity trapped by the torus is reduced .
this process will repeatedly continue until the two horizons merge . the above construction in fig .
[ 2dbh](b ) allows us ( at least in principle ) to obtain the event horizon in the quasi - stationary regime , when the main source of nonstationarity is the dissipation coming from the hawking radiation . as for the practical realization , there are , of course , many technical problems to be solved . on the other hand ,
if the black hole analog can in principle exist in condensed matter as the quasi - stationary object , its prototype the real black hole can also exist , at least in principle ( though it is not so easy to find the scenario of how this object can be obtained from the gravitational collapse of matter @xcite ) .
it appears that in some range of the energies of hawking - radiated particles , close to the `` planck '' scale , where the fermionic spectrum becomes `` nonrelativistic '' , the hawking radiation does not exist any more .
the particles which are radiated in the fully relativistic case , will be now andreev scattered back to the black hole .
thus both partners ( particle and hole ) of the hawking radiation will remain within the horizon , so that the particle creation in high gravity field will disturb the quantum vacuum inside the horizon without any radiation outside . in principle
the pair creation inside the horizon can be more important for the dissipation in the superfluid 3he film under discussion than the hawking radiation .
the body moving in the vacuum with linear acceleration @xmath198 is believed to radiate the thermal spectrum with the unruh temperature @xmath199 @xcite .
the comoving observer sees the vacuum as a thermal bath with @xmath200 , so that the matter of the body gets heated to @xmath201 ( see references in @xcite ) .
linear motion at constant proper acceleration ( hyperbolic motion ) leads to velocity arbitrarily close to the speed of light . on the other hand
uniform circular motion features constant centripetal acceleration while being free of the above mentioned pathology ( see the latest references in @xcite ) .
the latter motion is stationary in the rotating frame , which is thus a convenient frame for study of the radiation and thermalization effects for uniformly rotating body .
zeldovich @xcite was the first who predicted that the rotating body ( say , dielectric cylinder ) amplifies those electromagnetic modes which satisfy the condition @xmath202 here @xmath135 is the frequency of the mode , @xmath203 is its azimuthal quantum number , and @xmath204 is the angular velocity of the rotating cylinder .
this amplification of the incoming radiation is referred to as superradiance @xcite .
the other aspect of this phenomenon is that due to quantum effects , the cylinder rotating in quantum vacuum spontaneously emits the electromagnetic modes satisfying eq.([zeldovichcondition ] ) @xcite . the same occurs for any rotating body , including the rotating black hole @xcite , if the above condition is satisfied .
distinct from the linearly accelerated body , the radiation by a rotating body does not look thermal .
also , the rotating observer does not see the minkowski vacuum as a thermal bath .
this means that the matter of the body , though excited by interaction with the quantum fluctuations of the minkowski vacuum , does not necessarily acquire an intrinsic temperature depending only on the angular velocity of rotation .
moreover the vacuum of the rotating frame is not well defined because of the ergoregion , which exists at the distance @xmath205 from the axis of rotation .
the problems related to the response of the quantum system in its ground state to rotation@xcite , such as radiation by the object rotating in vacuum @xcite and the vacuum instability caused by the existense of ergoregion @xcite , etc .
, can be simulated in superfluids , where the superfluid ground state plays the part of the quantum vacuum .
the quantum friction due to spontaneous emission of phonons in superfluid @xmath157he and bogoliubov fermions in superfluid @xmath0he - b has been discussed in @xcite . here we extend an analysis to the radiation of quasiparticles in @xmath0he - a .
let us consider a cylinder of radius @xmath206 rotating with angular velocity @xmath204 in the ( infinite ) superfluid liquid .
in this situation there are again two important reference frames .
one of them is the laboratory frame .
the energy of quasiparticles is not well determined in the laboratory frame , because , if the rotating body is not the perfect axysymmetric cylinder , it produces the time dependent perturbations of the liquid .
however at distances far enough from the surface of the cylinder the influence of the rotating cylinder on the liquid can be neglected . in this approximation
the superfluid vacuum is in rest with respect to the laboratory frame .
this means that the superfluid velocity @xmath173 in the laboratory frame .
the quasiparticle energy in this frame is that as in the comoving frame of previous section .
it is @xmath207 if we consider superfluid @xmath157he , or the eq.([energycomovingframe ] ) , if we consider a pure 2d motion of @xmath0he - a with @xmath208 .
such energy spectrum corresponds to the effective minkowski metric of flat space @xmath209 here @xmath210 for isotropic @xmath157he and @xmath211 for anisotropic @xmath0he - a . in the 3d case of @xmath0he
- a there can be another effective metric far from the body .
it can be caused by the normal boundary condition on the @xmath212-vector , which prefers the radial orientation of @xmath212-vector . in the case of the radial distribution , @xmath213 , the effective metric for the quasiparticles moving outside the cylinder follows from eq.([gravfieldinaphase ] )
: @xmath214 such an effective space is conical : the space outside the cylinder is flat , but the proper length of the circumference of radius @xmath167 around the cylinder is not equal to @xmath215 . in the relativistic theories such
conical metric can arise outside the local cosmic strings , where there is an angle deficit . in our case
the length of the circumference is @xmath216 , which is much larger than @xmath215 : this effective space exhibits a `` negative angle deficit '' ( for details see ref.@xcite ) .
since the system far outside the rotating cylinder is not disturbed by the rotation of the body , the system will remain to be in or close to the minkowski vacuum ( or vacuum in the conical space in the case of radial @xmath217-vector ) as viewed in the laboratory frame .
however this does not hold in the region ajacent to the cylinder , where the superfluid velocity field is disturbed and time - dependent .
also the energy in the laboratory frame is not well determined because of the time - dependent perturbations .
the energy is well determined in the frame corotating with the cylinder . in the corotating frame
the cylinder is at rest , and thus the perturbations caused by rotation are stationary .
the metric in the coorotating frame is simplest far outside the rotating body , where the superfluid velocity in the corotating frame is @xmath218 . substituting this @xmath129 into eq.([gravfieldinaphase ] ) one obtains the interval @xmath219 , which determines the propagation of quasiparticles in the corotating frame : @xmath220 here again @xmath210 for phonons in isotropic @xmath157he and @xmath211 for fermionic quasiparticles in anisotropic @xmath0he - a with @xmath208 .
these metrics correspond to the interval in the rotating frame discussed in relativistic theories .
it is convinient to write the quasiparticle spectrum in this frame in two different approximations . in classical description one
has @xmath221 . in the other description , an azimuthal motion of quasiparticles
is quantized in terms of the angular momentum @xmath222 , while the radial is still treated in the quasiclassical approximation . in this case
the energy spectrum of phonons in the corotating frame is @xmath223 and the energy spectrum of the bogoliubov fermions @xmath224 the radius @xmath205 , where @xmath225 in the eq.([intervalrotation ] ) , marks the position of the ergoplane . in the ergoregion , i.e. at @xmath226 , the energy of quasiparticle in eq.([rotbogolonspectrum ] ) can be negative for any @xmath227 .
this means that in the ergoregion the minkowski vacuum is not the vacuum for the corotating observer .
situation is similar to the case discussed in previous section .
however the relevant superfluid velocity is azimuthal now instead of radial . as a result
there is an ergoplane instead of a horizon . from the point of view of the corotating observer
, the minkowski vacuum is unstable towards the filling of the negative energy states in the ergoregion , which means the radiation of the quasiparticles from the rotating cylinder .
however we need a real process , which leads to such radiation .
this radiation can be caused only by the interaction between the superfluid minkowski vacuum and the rotating object .
let us consider the slow rotations @xmath228 .
in this case the linear velocity of the cylinder at the surface of the cylinder , @xmath229 , is much smaller than the landau critical velocity for nucleation of quasiparticles , @xmath230 .
thus quasiparticles can not be nucleated near the surface of cylinder .
the ergoregion , where @xmath231 and quasiparticles can be nucleated , is far from the cylinder , @xmath232 .
the interaction with the cylinder , which produces the matrix element for the radiation , is very small . in this situation
the most effective mechanism of the quasiparticle radiation is the tunneling of quasiparticles from the liquid ajacent to the surface of the rotating body , which plays the part of the rotating detector , to the ergoregion .
the simplest rotating detector is the fermionic system , which is rigidly connected to the rotating body . in superfluids
the rotating body can be effectively substituted by the rigidly rotating cluster of quantized vortices ( see fig . [
ergoregion](a ) ) .
such clusters , that experience the solid - body rotation , are experimentally investigated in both phases of superfluid @xmath0he ( see e.g. @xcite ) .
the vortex cluster rotating in the infinite superfluid liquid represents the vacuum in the corotating frame .
however this vacuum state is not extended to the exterior of the cluster , where one has the analogue of minkowski vacuum .
that is why such state is not equilibrium , but it is quasistationary metastable state which can live for a macroscopically long time . at @xmath10
the dominating mechanism of the relaxation of angular velocity @xmath204 of the cluster is the process of the radiation of quasiparticles .
[ ergoregion](b ) shows the distribution of superfluid velocity @xmath233 in the laboratory frame . within the vortex cluster ,
i.e. at @xmath234 , the superfluid velocity , being averaged over the vortices , follows the velocity of the solid body rotation of the cluster : i.e. @xmath235 in the laboratory frame and thus @xmath236 in the frame corotating with the cluster . outside the cluster the superfluid velocity decays as @xmath237 , where @xmath238 is number of vortices in the cluster and @xmath239 is superfluid circulation around individual vortex . within the cluster one has @xmath236 in the corotating frame
, so one can expect that the spectrum of quasiparticles , which live within the cluster , is the same as the energy spectrum in the frame comoving with the superfluid velocity , @xmath240 .
in other words this spectrum has no @xmath241 shift of the energy levels , as distinct from quasiparticle spectrum far from the body , eqs.([quasiclassicalphononspectrum],[rotbogolonspectrum ] ) , which is measured in the same corotating frame .
this is not exactly true , since the equation @xmath236 does not imply that @xmath173 locally : it essentially depends on the position in the vortex lattice . the main feature is nevertheless preserved , when one considers the spectrum of quasiparticles that live in the vortex core .
their energy spectrum in the corotating state does not depend on rotation velocity @xmath204 : @xmath242 , where @xmath243 is the generalized angular momentum ; @xmath244 is the so called minigap , which depends on the core structure .
this branch crosses zero energy level as a function of @xmath243 , if one considers this quantum number as continuous , and thus represents the fermion zero mode ( see left part of fig . [
ergoregion](c ) ) .
the energy spectrum of such chiral fermions , that live in the vortex core , has been first calculated by caroli , de gennes and matricon for the abrikosov vortex in @xmath245-wave superconductors @xcite ; on the relation between the number of fermion zero modes and the winding number of the vortex see @xcite . for us it is important that for vortices in @xmath0he the quantum number @xmath243 is integer and thus one has the states in the detector with zero energy in the corotating frame . the radiation of bogoliubov quasiparticles can be considered as the process in which the particle from the zero energy state in the detector , @xmath246 , tunnels to the scattering state at the ergoplane , where also its energy is @xmath247 ( fig .
[ ergoregion](c ) ) . in the quasiclassical approximation
the tunneling probability is @xmath248 , where at @xmath249 and @xmath228 : @xmath250 thus all the particles with @xmath251 are radiated , but the radiation probability is smaller for higher @xmath203 : @xmath252 here @xmath253 is energy ( frequency ) of the radiated quasiparticles in the laboratory frame .
if @xmath58 is substituted by the speed of light , eq.([fermiontunnelingprobability ] ) is proportional to the superradiant amplification of the electromagnetic waves by rotating dielectric cylinder derived by zeldovich @xcite . since each radiated fermion carries the angular momentum @xmath203 , the vortex cluster rotating in superfluid vacuum at @xmath10 is loosing its angular momentum and thus experiences the quantum rotational friction .
the radation also leads to excitation of the detector matter . in principle
the radiation can occur without excitation of the detector vacuum , via direct interaction of the particles in the minkowski vacuum with the rotating body ( cluster ) , but the contribution of this process the quantum friction is smaller @xcite .
the rotational friction experienced by the body rotating in superfluid vacuum at @xmath10 , is caused by the spontaneous quantum emission of the quasiparticles from the rotating object to the `` minkowski '' vacuum in the ergoregion .
the emission is not thermal and depends on the details of the interaction of the radiation with the rotating body . in the quasiclassical approximation
it is mainly determined by the tunneling exponent , which can be approximately characterized by the effective temperature @xmath254 .
the vacuum friction of the rotating body can be observed only if the effective temperature exceeds the temperature of the bulk superfluid , @xmath255 .
for the body rotating with @xmath256rad / s , @xmath6 must be below @xmath257k .
however , high rotation velocity can be obtained for clusters .
the cluster containg two vortices rotate around their center of mass with @xmath258 , where @xmath206 is the radius of the circular orbit .
if the radius @xmath206 is of order of superfluid coherence length , the effective temperature can reach @xmath259k .
the process discussed in this section occurs only if there is an ergoplane in the rotating frame .
if the superfluid is contained in a finite external cylinder of radius @xmath260 , this process occurs only at high enough rotation velocity , @xmath261 , when the ergoplane is within the superfluid . on the instability of the ergoregion in quantum vacuum towards emission
see also in ref.@xcite .
if @xmath262 and ergoregion is not present , then the interaction between the coaxial cylinders via the vacuum fluctuations becomes the main mechanism for dissipation .
this causes the dynamic casimir forces between the walls moving laterally ( see review @xcite ) . as in @xcite
the nonideality of the cylinders , i.e. violation of the rotational symmetry by the body , is the necessary condition for quantum friction .
in the above examples of the nontrivial space , the effective gravitational field acts as a fixed external field .
the dynamics of this field has not been discussed here . in most cases
the effective gravity field do not obey the einstein equations .
this is the main drawback of superfluid @xmath0he - a : since the fermi points in @xmath0he - a are too far apart from each other , the dynamical equations for the gravitational field are clumsy .
nevertheless the above textures allow us to simulate many phenomena related to quantum vacuum in the presence of the strong gravity field .
this is because many properties of the quantum vacuum in curved space , which are determined by the geometry , do not depend on the dynamical origin of the geometry .
for example , it is well known that the hawking radiation is a purely kinematic effect and occurs in any geometry , if it exhbits an event horizon @xcite .
that is why the @xmath0he - a quantum vacuum is a right object for simulation of many aspects of physics of vacuum in a curved space .
for example what is the effect of the breakdown of the lorentz invariance at higher energy on hawking radiation .
the entropy of the black hole can be also investigated using the above model , since the microstates within the horizon are well determined and ( at least in principle ) are completely known in the whole energy range including the `` transplanckian '' region .
all these are different aspects of the problem of stability of the vacuum in strong gravitational and other fields .
the superfluids provide many examples of the instability of the superfluid vacuum , and thus allow us to investigate different mechanisms of relaxation of the physical vacuum the ether . | we discuss topological properties of the ground state of spatially homogeneous ensemble of fermions .
there are several classes of topologically different fermionic vacua ; in each case the momentum space topology of the vacuum determines the low - energy ( infrared ) properties of the fermionic energy spectrum . among them
there is class of the gapless systems which is characterized by the fermi - hypersurface , which is the topologically stable singularity .
this class contains the conventional landau fermi - liquid and also the non - landau luttinger fermi - liquid .
another important class of gapless systems is characterized by the topologically stable point nodes ( fermi points ) .
superfluid @xmath0he - a and electroweak vacuum belong to this universality class .
the fermionic quasiparticles ( particles ) in this class are chiral : close to the fermi points they are left - handed or right - handed massless relativistic particles . since the spectrum becomes relativistic at low energy , the symmetry of the system is enhanced in the low - energy edge .
the low - energy dynamics acquires local invariance , lorentz invariance and general covariance , which become better and better when the energy decreases .
interaction of the fermions near the fermi point leads to collective bosonic modes , which look like effective gauge and gravitational fields .
since the vacuum of superfluid @xmath0he - a and electroweak vacuum are topologically similar , we can use @xmath0he - a for simulation of many phenomena in high energy physics , including axial anomaly .
@xmath0he - a textures induce a nontrivial effective metrics of the space , where the free quasiparticles move along geodesics . with @xmath0he
- a one can simulate event horizons , hawking radiation , rotating vacuum , conical space , etc .
[ @twocolumnfalse ] |
ever since _ uhuru _ @xcite , x - ray astronomy missions have had a tradition of publishing catalogs of detected x - ray sources , and these catalogs have provided the fundamental datasets used by numerous studies aimed at characterizing the properties of the x - ray sky . while source catalogs are the primary data products from x - ray sky surveys ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , the _ einstein _
ipc catalog @xcite demonstrated the utility of catalogs of _ serendipitous _ sources identified in the fields of _ pointed - observation _ x - ray missions .
more recent serendipitous source catalogs ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) have further expanded the list of sources with x - ray data available for further analysis by the astronomical community .
source catalogs typically include a uniform reduction of the mission data .
this provides a significant advantage for the general scientific community because it removes the need for end - users , who may be unfamiliar with the complexities of the particular mission and its instruments , to perform detailed reductions for each observation and detected source . when compared to all previous and current x - ray missions , the _ chandra _ x - ray observatory ( e.g. , * ? ? ?
* ; * ? ? ?
* ) breaks the resolution barrier with a sub - arcsecond on - axis point spread function ( psf )
. launched in 1999 , _ chandra _ continues to provide a unique high spatial resolution view of the x - ray sky in the energy range from @xmath4 to @xmath5 , over a @xmath6@xmath7 square arcminute field of view .
the combination of excellent spatial resolution , a reasonable field of view , and low instrumental background translate into a high detectable - source density , with low confusion and good astrometry .
_ chandra _ includes two instruments that record images of the x - ray sky .
the advanced ccd imaging spectrometer ( acis ; * ? ? ? * ; * ? ? ? * ) instrument incorporates ten @xmath8 pixel ccd detectors ( any six of which can be active at one time ) with an effective pixel size of @xmath9 on the sky , an energy resolution of order @xmath10 at the al - k edge ( @xmath11 ) , and a typical time resolution of @xmath12 . the high resolution camera ( hrc ; * ? ? ?
* ) instrument consists of a pair of large format micro - channel plate detectors with a pixel size @xmath13 on the sky and a time resolution of @xmath14 , but with minimal energy resolution .
the wealth of information that can be extracted from identified serendipitous sources included in _
observations is a powerful and valuable resource for astronomy .
the aim of the ( csc ) is to disseminate this wealth of information by characterizing the x - ray sky as seen by _
chandra_. while numerous other catalogs of x - ray sources detected by _
chandra _ may be found in the literature ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , the region of the sky or set of observations that comprise these catalogs is restricted , and they are typically aimed at maximizing specific scientific goals . in contrast , the csc is intended to be an all - inclusive , uniformly processed dataset that can be utilized to address a wide range of scientific questions .
the csc is intended ultimately to comprise a definitive catalog of x - ray sources detected by _
chandra _ , and is being made available to the astronomical community in a series of increments with increasing capability over the next several years .
the first release of the csc was published in 2009 march .
this release includes information about 135,914 source detections , corresponding to 94,676 distinct x - ray sources on the sky , extracted from a _ subset _ of public imaging observations obtained using the acis instrument during the first eight years of the _ chandra _ mission .
the distribution of release 1 sources on the sky is presented in figure 1 .
we expect that the csc will be a highly valuable tool for many diverse scientific investigations .
however , the catalog is constructed from pointed observations obtained using the _ chandra _ x - ray observatory , and is neither all - sky nor uniform in depth .
the first release of the catalog includes only point and compact sources , with observed extents @xmath0 . because of the difficulties inherent in detecting highly extended sources and point and compact sources that lie close to them , and quantifying in a consistent and robust way the properties of such sources , we have chosen to exclude entire fields ( or in some cases , individual acis ccds ) containing such sources from the first release of the csc , as described in 3.1 .
therefore , the catalog does not include sources near some of the most famous _ chandra _ targets , and there may be selection effects that restrict the source content of the catalog and which therefore may limit scientific studies that require unbiased source samples .
the minimum flux significance threshold for a source to be included in the first release of the csc is set conservatively , and corresponds typically to @xmath15 detected source photons ( on - axis ) in the broad energy band integrated over the total exposure time .
this conservative threshold was chosen to maintain the spurious source rate at an acceptable level over the wide variety of _ chandra _ observations that are included in this release of the catalog .
we expect to relax this criterion in future releases based on experience gained constructing the current release .
a number of other _ chandra _ catalogs do include sources with fewer net counts than the csc .
such fainter thresholds are attainable typically either because of specific attributes of the observations included in those catalogs , or because of the assumptions made when constructing the catalog . as an example of the former category , the xbootes survey catalog @xcite includes sources that are roughly a factor of two fainter than the csc flux significance threshold .
that survey is constructed from short ( @xmath16 ) observations obtained in an area with low line - of - sight absorption .
this results in a negligible background level that substantially simplifies source detection and enables identification of sources with very few counts .
catalogs derived from observations with the range of exposures comparable to those that comprise the csc ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) also include fainter sources
. however , in these cases the additional source fractions are in general not large , typically adding @xmath17 more sources below the csc threshold , as described in detail in 3.7.1 . for other _
catalogs , visual review and validation at the source level is a planned part of the processing thread ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . in some cases ( e.g. , * ? ?
* ) , visual review may be used to adjust processing parameters for individual sources .
such manual steps are time - consuming , but enable lower significance levels to be achieved while maintaining an acceptable spurious source rate .
in contrast , the csc catalog construction process requires that the processing pipelines run on a wide range of observations with a minimum of manual intervention .
the scope of the csc is simply too large to require manual handling at the source level .
we do not manually inspect individual source detections , nor do we adjust source detection or processing parameters based on manual evaluation . instead , the csc uses a largely automated quality assurance approach , as described in 3.14 . the sky coverage of the first catalog release ( fig .
2 ) totals @xmath18 square degrees , with coverage of @xmath19 square degrees brighter than a @xmath20@xmath21 flux limit of @xmath22 , decreasing to @xmath23 square degrees brighter than @xmath24 , and @xmath25 square degrees brighter than @xmath26 .
these numbers will continue to grow as the _ chandra _ mission continues , with a 15 year prediction of the eventual sky coverage of the csc of order 500 square degrees , or a little over 1% of the sky . in this paper
we describe in detail the content and construction of release 1 of the csc . however , where appropriate we also discuss in addition the steps required to process hrc instrument data used to construct release 1.1 of the catalog , since the differences in the algorithms are small .
release 1.1 of the catalog is scheduled for spring 2010 .
this paper is organized into 5 sections , including the introduction . in 2 , we present a description of the catalog .
this includes the catalog design goals , an outline of the general characteristics of _ chandra _ data that are relevant to the catalog design , the organization of the data within the catalog , approaches to data access , and an outline of the data content of the catalog .
section 3 , which comprises the bulk of the paper , describes in detail the methods used to extract the various source properties that are included in the catalog , with particular detail provided when the algorithms are new or have been adapted for use with _
chandra _ data .
a brief description of the principal statistical properties of the catalog sources is presented in 4 ; this topic is treated comprehensively by f. a. primini et al .
( 2010 , in preparation ) .
conclusions are presented in
finally , appendix a contains details of the algorithm used to match source detection from multiple overlapping observations , as well as the mathematical derivation of the multivariate optimal weighting formalism used for combining source position and positional uncertainty estimates from multiple observations .
the csc is intended to be a general purpose virtual science facility , and provides simple access to a carefully selected set of generally useful quantities for individual sources or sets of sources matching user - specified search criteria .
the catalog is designed to satisfy the needs of a broad - based group of scientists , including those who may be less familiar with astronomical data analysis in the x - ray regime , while at the same time providing more advanced data products suitable for use by astronomers familiar with _
chandra _ data .
the primary design goals for the csc are to ( 1 ) allow simple and quick access to the best estimates of the x - ray source properties for detected sources , with good scientific fidelity , and directly support scientific analysis using the individual source data ; ( 2 ) facilitate analysis of a wide range of statistical properties for classes of x - ray sources ; ( 3 ) provide efficient access to calibrated observational data and ancillary data products for individual x - ray sources , so that users can perform detailed further analysis using existing tools such as those included in the _ chandra _ interactive analysis of observations ( ciao ; @xcite ) portable data analysis package ; and ( 4 ) include all real x - ray sources detected down to a predefined threshold level in all of the public _
datasets used to populate the catalog , while maintaining the number of spurious sources at an acceptable level . to achieve these goals , for each detected x - ray source the catalog records the source position and a detailed set of source properties , including commonly used quantities such as multi - band aperture fluxes , cross - band hardness ratios , spectra , temporal variability information , and source extent estimates .
in addition to these traditional elements , the catalog includes file - based data products that can be manipulated interactively by the user .
the primary data products are photon event lists ( e.g. , * ? ? ?
* ) , which record measures of the location , time of arrival , and energy of each detected photon event in a tabular format .
additional data products derived from the photon event list include images , light curves , and spectra for each source individually from each observation in which a source is detected .
the catalog release process is carefully controlled , and a detailed characterization of the statistical properties of the catalog to a well defined , high level of reliability accompanies each release .
key properties evaluated as part of the statistical characterization include limiting sensitivity , completeness , false source rates , astrometric and photometric accuracy , and variability information .
both acis and hrc cameras operate in a photon counting mode , and register individual x - ray photon events . for each photon event
, the two dimensional position of the event on the detector is recorded , together with the time of arrival and a measure of the energy of the event . in most operating modes ,
lists of detected events are recorded over the duration of an observation , typically between @xmath27 and @xmath28 , and are then telemetered to the ground for subsequent processing . to minimize the effect of bad detector pixels , and to avoid possible burn - in degradation of the camera by bright x - ray sources , the pointing direction of the telescope
is normally constantly dithered in a lissajous pattern , with a typical scale length of about @xmath29 on the sky and a period of order @xmath27 , while taking data .
the motion of the telescope is recorded via an `` aspect camera '' @xcite that tracks the motion of a set of ( usually 5 ) guide stars as a function of time during the observation . the coordinate transformation needed to remove the motion from the event ( photon ) positions
is computed from the aspect camera data and applied during data processing .
breaking down the 4-dimensional x - ray data hypercube into spatial , spectral , and temporal axes provides a natural focus on the properties that may be of interest to the general user , but also identifies some of the complexities inherent in _ chandra _ data that must be addressed by catalog construction and data analysis algorithms .
spatially , the _ chandra _ psf varies significantly with off - axis and azimuthal angle ( with the former variation dominating ) , as well as with incident photon energy ( fig . 3 ) .
close to the optical axis of the telescope , the psf is approximately symmetric with a 50% enclosed energy fraction radius of order @xmath30 over a wide range of energies , but at @xmath31 off - axis the psf is strongly energy - dependent , asymmetric , and significantly extended , with a 50% enclosed energy fraction radius of order @xmath32 at @xmath33 . for the widely used acis detector ,
the instrumental spectral energy resolution is of order @xmath34@xmath35 , and depends on incident photon energy and location on the detector . because the energy resolution is significantly lower than the typical energy width of the features and absorption edges that define the effective area of the telescope optics ( and therefore the quantum efficiency of the telescope plus detector system ) , a full matrix formulation that considers the redistribution of source x - ray flux into the set of instrumental pulse height analyzer bins must be used when performing spectral analyses .
this is in contrast to the more familiar scenario from many other wavebands , where the instrumental resolution is often much higher than the spectral variation of quantum efficiency , enabling the commonly used implicit assumption that the flux redistribution matrix is diagonal ( and is therefore not considered explicitly ) .
we note in passing that _ chandra _ is equipped with a pair of transmission gratings that can be inserted into the optical path , and is therefore capable of performing high spectral resolution ( slitless spectroscopy ) observations .
however , such observations are not included in the current release of the csc .
time domain analyses must consider the impact of spacecraft dither within an observation .
strong false variability signatures at the dither frequency can arise because of variations of the quantum efficiency over the detector , or because the source or background region dithers off the detector edge or across a gap between adjacent acis ccds .
corrections for these effects , as well as for cosmic x - ray background flares that can be highly variable over periods of a few kiloseconds , must be applied when computing light - curves .
the extremely low photon event rates common for many faint x - ray sources typically require time domain statistics to be evaluated using event arrival - time formulations instead of rate - based approaches .
an additional level of complexity occurs because many astronomical sources of interest that will be included in the catalog are extremely faint .
rigorous application of poisson counting statistics is required when deriving source properties and associated errors , separating x - ray analyses from many other wavebands where gaussian statistics are typically assumed .
the tabulated properties included in the csc are organized conceptually into two separate tables , the _ source observations table _ and the _ master sources table_.
distinguishing between source detections ( as identified within a single observation ) and x - ray sources physically present on the sky is necessary because many sources are detected in multiple observations and at different off - axis angles ( and therefore have different psf extents ) .
each record included in the source observations table tabulates properties derived from a source detection in a single observation .
these entries also include pointers to the associated file - based data products that are included in the catalog , which are all observation - specific in the first catalog release .
each record in the source observations table is further split internally into a set of source - specific data and a set of observation - specific , but source - independent , data .
the latter are recorded once to avoid duplication .
a description of the data columns recorded in the source observations table for each source detection is provided in table 1 . -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein -0.75truein because of the dependence of the psf extent with off - axis angle , multiple distinct sources detected on - axis in one observation may be detected as a single source if located far off - axis in a different observation ( fig . 4 ) .
during catalog processing , source detections from all observations that overlap the same region of the sky are spatially matched to identify distinct x - ray sources .
estimates of the tabulated properties for each distinct x - ray source are derived by combining the data extracted from all source detections and observations that can be uniquely associated , according to the algorithms described in 3 .
the best estimates of the source properties for each distinct x - ray source are recorded in the master sources table .
a description of the data columns recorded in the master sources table for each source is provided in table 2 .
each distinct x - ray source is thus conceptually represented in the catalog by a single entry in the master sources table , and one or more associated entries in the source observations table ( one for each observation in which the source was detected ) .
all of the tabulated properties included in both the master sources table and the source observations table can be queried by the user .
bi - directional links between the entries in the two tables are managed transparently by the database , so that the user can access all observation data for a single source seamlessly .
if a source detection included in the source observations table can be related unambiguously to a single x - ray source in the master sources table , then the corresponding table entries will be associated by `` unique '' linkages .
source detections included in the source observations table that can not be related uniquely to a single x - ray source in the master source table will have their entries associated by `` ambiguous '' linkages ( fig .
the data from ambiguous source detections are not used when computing the best estimates of the source properties included in the master sources table . in the case of acis observations , source detections for which the estimated photon pile - up fraction @xcite exceeds @xmath36
will not be used if source detections in other acis observations do not exceed this threshold . using the linkages between the entries in the two tables
, the user will nevertheless be able to identify all of the x - ray sources in the catalog that could be associated with a specific detection in a single observation , and vice - versa .
these linkages may be important , for example , when identifying candidate targets for follow - up studies based on a data signature that is only visible in the observation data for a confused source . the primary user tool for querying the csc is the cscview web - browser interface @xcite , which can be accessed from the public catalog web - site .
the user can directly query any of the tabulated properties included in either the master sources table or the source observations table , display the contents of an arbitrary set of properties for matching sources , and retrieve any of the associated file - based data products for further analysis .
cscview provides a form - based data - mining interface , but also allows users to enter queries written using the astronomical data query language ( adql ; @xcite ) standard directly .
query results can be viewed directly on the screen , or saved to a data file in multiple formats , including tab - delimited ascii ( which can be read directly by several commonly used astronomical applications ) and international virtual observatory alliance ( ivoa ) standard formats such as votable @xcite .
automated access to query the catalog from data analysis applications and scripts running on the user s home platform was identified as being needed for several science use cases .
vo standard interfaces , including simple cone search @xcite and simple image access @xcite , provide limited query and data access capabilities , while more sophisticated interactions are possible through a direct url connection .
support for vo workflows using applicable standards will be added in the future as these standards stabilize .
an interface that integrates catalog access with a visual sky browser provides a simple mechanism for visualizing the regions of the sky included in the catalog , and may also be particularly beneficial for education and public outreach purposes .
since _ chandra _ is an ongoing mission , the csc includes a mechanism to permit newly released observations to be added to the catalog and be made visible to end users , while at the same time providing stable , well - defined , and statistically well - characterized released catalog versions to the community .
this is achieved by maintaining a revision history for each database table record , together with flags that establish whether catalog quality assurance and catalog inclusion criteria are met , and using distinct views of the catalog databases that utilize these metadata .
`` catalog release views '' provide access to each released version of the catalog , with the latest released version being the default .
catalog releases will be infrequent ( no more than of order 1 per year ) because of the controls built in to the release process , and because of the requirement that each release be accompanied by a detailed statistical characterization of the included source properties .
once data are included in a catalog release view , then they are frozen in that view , even if the source properties are revised or the source is deleted in a later catalog release .
a source may be deleted if the detection is subsequently determined to be an artifact of the data or processing , but the most likely reason that a source is deleted from a later catalog release is that additional observations included in the later release resolve the former detection into multiple distinct sources .
ccccccc ultra - soft & _ u _ & 0.20.5 & 0.4 & 7.362.24 & 68.723.0 & + soft & _ s _ & 0.51.2 & 0.92 & 216155 & 411274 & + medium & _ m _ & 1.22.0 & 1.56 & 438401 & 539493 & + hard & _ h _ & 2.07.0 & 3.8 & 15901580 & 16801670 & + broad & _ b _ &
0.57.0 & 2.3 & 22402140 & 26302440 & + wide & _ w _ & 0.110 & 1.5 & & & 605 + `` database views '' provide access to the catalog database , including any new content that may not be present in an existing catalog release .
because on - going processing is continually modifying the catalog database , tabulated data and file - based data products in a database view may be superseded at any time , and the statistical properties of the data are not guaranteed .
we anticipate that users who require a stable , well - characterized dataset will choose primarily to access the catalog through the latest catalog release view . on the other hand , users who are interested in searching the latest data to identify sources with specific signatures for further study will likely use the latest database view .
the first release of the csc includes detected sources whose flux estimates are at least 3 times their estimated @xmath1 uncertainties , which typically corresponds to about 10 net ( source ) counts on - axis and roughly 2030 net counts off - axis , in at least one energy band . in this release , multiple observations of the same field are _ not _ combined prior to source detection , so the flux significance criterion applies to each observation separately .
for each source detected in an observation , the catalog includes approximately 120 tabulated properties .
most values have associated lower and upper confidence limits , and many are recorded in multiple energy bands .
the total number of columns included in the source observations table ( including all values and associated confidence limits for all energy bands ) is 599 .
roughly 60 master properties are tabulated for each distinct x - ray source on the sky , generated by combining measurements from multiple observations that include the source .
combining all values and associated confidence limits for all energy bands yields a total of 287 columns included in the master sources table .
the tabulated source properties fall mostly into the following broad categories : source name , source positions and position errors , estimates of the raw ( measured ) extents of the source and the local point spread function , and the deconvolved source extents , aperture photometry fluxes and confidence intervals measured or inferred in several ways , spectral hardness ratios , power - law and thermal black - body spectral fits for bright ( @xmath37 net counts ) sources , and several source variability measures ( gregory - loredo , kolmogorov - smirnov , and kuiper tests ) . also included in the csc
are a number of file - based data products in formats suitable for further analysis in ciao .
these products , described in table 3 , include both full - field data products for each observation , and products specific to each detected observation - specific source region .
the full - field data products include a `` white - light '' full - field photon event list , and multi - band exposure maps , background images , exposure - corrected and background - subtracted images , and limiting sensitivity maps .
source - specific data products include a white - light photon event list , the source and background region definitions , a weighted ancillary response file ( the time - averaged product of the combined telescope / instrument effective area and the detector quantum efficiency ) , multi - band exposure maps , images , model ray - trace psf images , and optimally binned light - curves .
observations obtained using the acis instrument additionally include low - resolution ( @xmath38@xmath39 , depending on incident photon energy and location on the array ) source and background spectra and a weighted detector redistribution matrix file ( the probability matrix that maps photon energy to detector pulse height ) .
the energy bands used to derive many csc properties are defined in table 4 .
the energy bands are chosen to optimize the detectability of x - ray sources while simultaneously maximizing the discrimination between different spectral shapes on x - ray color - color diagrams .
the effective area of the telescope ( including both the _ chandra _ high resolution mirror assembly [ hrma ] and the detectors ) is shown in figure 6 as a function of energy , together with the average acis quiescent backgrounds derived from blank sky observations @xcite . the effective area is measured at the locations of the nominal `` acis - s '' aimpoint on the acis s3 ccd , and the nominal `` acis - i '' aimpoint on the acis i3 ccd . where possible , the energy bands are chosen to avoid large changes of effective area within the central region of the band , since such variations degrade the accuracy of the monochromatic effective energy approximation described below .
for example , the m - edge of the iridium coating on the hrma has significant structure in the @xmath40@xmath41 energy range that provides a natural breakpoint between the acis medium and hard energy bands .
note however , that large effective area variations are unavoidable within the acis broad and soft energy bands and the hrc wide energy band . weighting the effective area by the source spectral shape and integrating over the bandpass provides an indication of the relative detectability of a source in the energy band . selecting energy band boundaries so that source detectability is roughly the same in different energy bands more uniformly distributes poisson errors across the bands , and so enhances detectability in the various bands .
several different source types were simulated when selecting energy bands .
these included absorbed non - thermal ( power - law ) models with photon index values ranging from 1 to 4 , absorbed black - body models with temperature varying from @xmath42 to @xmath43 , and absorbed , hot , optically - thin thermal plasma models @xcite with @xmath44@xmath45 .
in all cases , the hydrogen absorbing column was varied over the range @xmath46@xmath47 .
detected x - ray spectra were simulated using pimms @xcite , and then folded through the bandpasses to construct synthetic x - ray color - color diagrams ( see fig . 7 for example color - color diagrams based on the final band parameters ) .
energy bands chosen to fill the color - color diagrams maximally provide the best discrimination between different spectral shapes . for detailed x - ray spectral - line modeling ,
the @xcite models have been superseded by more recent x - ray plasma models ( e.g. , * ? ? ?
* ; * ? ? ?
. however , since the radiated power of the newer models as a function of temperature is not significantly different from the 1993 versions of the @xcite models used here , the latter are entirely adequate for the purpose of evaluating coverage of the x - ray color - color diagrams and the task is greatly simplified because of their availability in pimms .
@xcite compared broad band x - ray photometry with accurate acis spectral fits and found that model - independent fluxes could be derived from the photometry measurements to an accuracy of about 50% or better for a broad range of plausible spectra .
they used similar but not identical energy bands to those adopted for the csc , but did not use the method of deriving fluxes from individual photon energies employed herein . combining all of these considerations @xcite yields the following selection of energy bands for the csc .
the acis soft ( _ s _ ) energy band spans the energy range 0.5@xmath48 .
the lower bound is a compromise that is set by several considerations .
acis calibration uncertainties increase rapidly below @xmath49 , so this establishes a fairly hard lower limit to avoid degrading source measurements in the energy band . as shown in figure 8 , below about @xmath50 the background count rate begins to increase rapidly , while the integrated effective area rises very slowly resulting in few additional source counts .
while pushing the band edge to higher energy will result in a lower background , the integrated effective area drops rapidly if the lower bound is raised above @xmath51 , reducing the number of source counts collected in the band .
we choose to set the lower bound equal to @xmath49 since doing so enhances the detectability of super - soft sources , while not noticeably impacting measurements of other sources . the upper cutoff for the soft energy band
is set equal to @xmath52 , which balances the preference for uniform integrated effective areas amongst the energy bands with the desire to maximize the area of x - ray color - color plot parameter space spanned by the simulations .
the lower bound of the acis medium ( _ m _ ) energy band matches the upper bound of the soft energy band .
we locate the upper band cutoff at @xmath43 since this value tends to maximize the coverage of the x - ray color - color diagram .
this value also moves the iridium m - edge out of the sensitive medium band , and instead placing it immediately above the lower boundary of the acis hard ( _ h _ ) energy band .
the high energy boundary of the latter band is set to @xmath21 .
this cutoff provides a good compromise between maximizing integrated effective area and minimizing total background counts ( fig .
8) . above @xmath21 ,
the background rate increases rapidly at the acis - s , while below this energy the integrated effective area decreases rapidly at the acis - i aimpoint . placing the hard energy band cutoff at @xmath21 also has the advantage that the @xmath53 line is included in the band , allowing intense fe line sources to be detected without compromising the measurement quality for typical catalog sources .
the acis broad ( _ b _ ) band covers the same energy range as the combined soft , medium , and hard bands , and therefore spans the energy range 0.5@xmath21 .
simulations indicate that an additional energy band extending below @xmath54 is beneficial for discriminating super - soft x - ray sources in color - color plots .
the acis front - illuminated ccds have minimal quantum efficiency below @xmath55 , while the response of the back - illuminated ccds extends down to @xmath56 .
hydrocarbon contamination is present on both the hrma optics @xcite and the acis optical blocking filter @xcite .
the latter reduces the effective area at low energies , and enhances the depth of the carbon k - edge .
an acis ultra - soft ( _ u _ ) band covering @xmath57@xmath54 is added to provide better discrimination of super - soft sources .
source detection is _ not _ performed in this energy band , because of the typical lower overall signal - to - noise ratio ( snr ) and the resulting enhanced false - source rate . finally ,
since the hrc ( particularly hrc - i ) has minimal spectral resolution , a single wide ( _ w _ ) band that includes essentially the entire pulse height spectrum ( specifically , pi values @xmath58 ) , roughly equivalent to 0.1 - 10 kev , is used for hrc observations .
while bands in these general energy ranges give the best balance of count rate and spectral discrimination , our simulations indicate that the exact choice of band boundary energies is not critical at the 10% level . in principle ,
the variations of hrma effective area , detector quantum efficiency , and ( for acis ) focal plane contamination , with energy imply that energy - dependent data products such as exposure maps or psfs should be constructed by integrating the source spectrum over the energy band
. this approach would be both extremely time - consuming , and require knowledge of the source spectrum that is typically not available _ a priori_. in practice , a monochromatic effective energy is chosen for each energy band to be used to construct energy dependent data products @xcite .
the monochromatic effective energy for each band is determined using the relation @xmath59 where @xmath60 is the energy , @xmath61 is the effective area of the hrma , @xmath62 is the detector quantum efficiency , @xmath63 is the reduction in transmission due to focal plane contamination , @xmath64 is a power - law spectral weighting function of the form @xmath65 , and the integral is performed over the energy band .
the monochromatic effective energies for each energy band were calculated for sources located at the acis - i and acis - s aimpoints , and also for the nominal aimpoint on the hrc - i detector . since the csc is constructed from observations acquired throughout the _ chandra _ mission , acis focal plane contamination models with both zero contamination ( appropriate for observations obtained early in the mission ) and the contamination level current as of late 2009 were employed .
power - law spectral weighting functions with @xmath66 varying from 0.0 to 2.0 were used .
setting @xmath67 gives a spectral weighting function that approximates an absorbed @xmath68 power - law spectrum , and the limits for @xmath66 were chosen to span the typical range of values determined from fits to a canonical subset of _ chandra _ datasets .
the remaining parameters in equation ( [ eqn : eeff ] ) are extracted from the _ chandra _ calibration database ( caldb ; * ? ? ?
* ; * ? ? ?
the monochromatic effective energies for acis were chosen to be the approximate arithmetic means of the @xmath69 values derived for the acis - i and acis - s aimpoints , with zero and late 2009 focal plane contamination .
for acis energy bands other than the the broad band , the monochromatic effective energies computed for a single value of @xmath66 all agree within @xmath70 .
the dependence on @xmath66 is similarly small , except for the hard energy band , where varying @xmath66 from 0.0 to 2.0 changes the monochromatic effective energy from @xmath71 to @xmath72 . for the acis broad energy band ,
the agreement between the different models for a single value of @xmath66 is @xmath73 .
however , for this band the dependence on @xmath66 is more significant , varying from @xmath74 for @xmath75 to @xmath76 for @xmath77 .
the monochromatic effective energies used to construct the csc are reported in table 4 .
although the use of a single monochromatic effective energy for each energy band simplifies data analysis by removing the dependence on the source spectrum , some error will be introduced for sources that have either _ extremely _ soft or _ extremely _ hard spectra compared to the canonical @xmath78 power - law spectral weighting function .
knowledge of the expected magnitude of the error that may be introduced is helpful when evaluating catalog properties . for both the acis medium and hard energy bands ,
neither extremely soft nor extremely hard source spectra induce variations in exposure map levels that are greater than @xmath36 , so photometric errors due to source spectral shape should not exceed this value . in the acis soft energy band , very soft
spectra may produce deviations of order 520% , with the largest excursions expected for the front - illuminated ccds .
these differences increase to @xmath7935% for the acis ultra - soft energy band , with the largest values once again associated with the front - illuminated ccds . for all of the acis
narrow energy bands , the errors induced by extremely hard spectra are much smaller than those caused by extremely soft spectra
. the presence of the iridium edge and the large energy ranges included in the acis broad and hrc wide energy bands may produce significantly larger variations for extreme spectral shapes .
very soft spectra can alter exposure map values by @xmath8090% in the acis broad energy band , although there is little impact in the hrc wide energy band .
conversely , extremely hard spectra may induce changes up to @xmath81 in the hrc wide energy band , and @xmath8230% in the acis broad energy band . as described in 4.4 , model - based statistical characterization of csc source fluxes ( f. a. primini et al .
2010 , in preparation ) produces results that are generally consistent with these expectation , with the exception that flux errors in the acis broad energy band appear to be @xmath36 for most sources . when computing fluxes for point sources , an aperture correction is applied to compensate for the fraction of the psf that is not included in the aperture .
since the extent of the _ chandra _ psf varies with energy , using a monochromatic effective energy can introduce a flux error because the energy dependence of the psf fraction is not considered .
this error can be bounded by a _
post facto _ comparison of psf fractions for catalog source detections in the 5 acis energy bands . the majority of variations _ between _ energy bands fall in the range 48% , with 90% of source detections showing @xmath83 differences .
these values represent an upper bound on the error introduced _ within _ an energy band by the use of a monochromatic effective energy .
as described previously , x - ray photon event data are recorded in the form of a photon event list .
the pixel position on the detector where a photon was detected is recorded in the `` chip '' pixel coordinate system .
event positions are remapped to celestial coordinates through a series of transforms , as described by @xcite .
the first step in this process remaps chip coordinates to a uniform _ real - valued _ virtual `` detector '' pixel space by applying corrections for the measured detector geometry , and instrumental and telescope optical system distortions recorded in the caldb .
subsequent application of the time - dependent aspect solution removes the spacecraft dither motion , and maps the event positions to a uniform virtual `` sky '' pixel plane .
the latter has the same pixel scale as the original instrumental pixels , but is oriented with north up ( @xmath84 direction ) and is centered at the celestial coordinates of the tangent plane position for the observation . as an aid to users , the location of each event in each coordinate system is recorded in the calibrated photon event list .
a simple unrotated world coordinate system transform maps sky positions to icrs right ascension and declination by applying the plate scale calibration to the difference between the position of the source and a fiducial point , which is typically the optical axis of the telescope .
the celestial coordinates of the fiducial point are determined from the aspect solution .
sky images are constructed from the calibrated photon event lists by binning photon positions in sky coordinates into a regular , rectangular image pixel grid .
a consequence of constructing images by binning in sky coordinates is that _ chandra _ images are always oriented with north up .
the choice of image blocking factor determines the number of sky pixels that are binned into a single image pixel .
full field image products associated with acis observations are constructed by binning the area covered by the inner @xmath85 sky pixels at single pixel resolution , then binning the inner @xmath86 sky pixels at block 2 , and finally binning the entire @xmath87 sky pixel field at block 4 .
the corresponding blocking factors for hrc - i observations are 2 , 5 , and 12 .
using a constant blocked image size of @xmath85 pixels reduces overall data volume , while preserving resolution in the outer areas of the field of view where the psf size is significantly larger than a single pixel .
in this section , we describe in detail the methods used to derive the x - ray source properties that are included in the csc , with particular detail provided in cases where the algorithms are new or have been newly adapted for use with _
chandra _ data . the principal steps necessary to generate the catalog consist of processing the data for each observation s full field - of - view , detecting x - ray sources included within that field of view , and then extracting the spatial , photometric , spectral , and temporal properties of each detected source .
figure 9 is a depiction of the high - level flow used to perform these steps . in the figure
, each block references the section of the text that describes in detail the methods used .
the physical properties associated with each source detection are recorded as a separate row in the catalog source observations table .
once the source detections from each observation have been evaluated , they are correlated with source detections from all other spatially overlapping observations to identify distinct x - ray sources on the sky . the steps required to perform the source cross - matching , and then combine the data from multiple observations of a single source to evaluate the source s properties , follow a similar flow to the one presented in figure 10 . many of the elements that comprise the second flow are built on the foundations developed for the related steps from the first flow . for convenience and continuity of notation ,
the former are described in the same text sections as the latter .
the properties for each distinct x - ray source are included as a separate row in the catalog master sources table .
data processing for release 1 of the csc was performed using versions 3.03.0.7 of the _ chandra _ x - ray center data system ( cxcds ; * ? ? ?
* ; * ? ? ?
* ) catalog processing system ( `` cat '' ) , with calibration data extracted from caldb version 3.5.0 .
the observation recalibration steps included in cat3.0 correspond _ approximately _ with those included ciao 4.0 .
in several cases , programs developed for cat3.0 to evaluate source properties have been repackaged with new interfaces for interactive use in subsequent ciao releases ( see table 5 ) . while the csc ultimately aims to be a comprehensive catalog of x - ray sources detected by _
chandra _ , all of the functionality required to achieve that goal are not included in the release 1 processing system .
a set of pre - filters is used to limit the data content to the set of observations that the catalog processing system is capable of handling . for release 1
, only public acis `` timed - exposure '' readout mode imaging observations obtained using either the `` faint , '' `` very faint , '' or `` faint with bias '' datamodes are included .
acis observations that are obtained using ccd subarrays with @xmath88 rows are also excluded , because there are too few rows to ensure that source - free regions can be identified reliably when constructing the high spatial frequency background map .
hrc - i imaging mode observations are not included in release 1 of the catalog , but are included in incremental release 1.1 .
hrc - s observations are excluded because of the presence of background features associated with the edges of `` t''-shaped energy - suppression filter regions that form part of the uv / ion - shield .
observations of solar system objects are not included in the csc .
all observations included in the csc must have been processed using the standard data processing pipelines included in version 7.6.7 , or later , of the cxcds .
this version of the data system was used to perform the most recent bulk reprocessing of _ chandra _ data , and includes revisions to the pipelines that compute the aspect solution that is used to correct for the spacecraft dither motion and register the source events on the sky .
observations must have successfully passed the `` validation and verification '' ( quality assurance ) checks that are performed upon completion of standard data processing .
the largest scale lengths used to detect sources to be included in the csc have angular extents @xmath89 .
sources with apparent sizes greater than this are either not detected , or may be detected incorrectly as multiple close sources .
prior to catalog construction , all observations are inspected visually for the presence of extended sources that may be detected incorrectly , and such observations are excluded from catalog processing . for acis observations , if the presence of any spatially extended emission is restricted to a single ccd only , then the data from that ccd are dropped , and sources detected on any remaining ccds are typically included in the catalog .
the latter rule allows many sources surrounding bright , extended cores of galaxies to be included in the catalog , rather than having the entire observation rejected outright .
while the visual inspection and rejection process is inherently subjective in nature , an attempt was made to calibrate the method by constructing a `` training set '' of several hundred observations that were processed through a test version of the catalog pipelines .
the training set observations included a wide variety of point , compact , and extended sources , with differing exposures and snr , which were classified as accept / reject based on the actual results of running the pipeline source detection and source property extraction steps . these observations and classifications
were then used to train the personnel who performed the visual inspection process .
although all observations included in the csc have been processed through the cxcds standard data processing pipelines , we nevertheless re - run the instrument - specific calibration steps as the first step in catalog construction .
one reason for reapplying the instrumental calibrations is that they are subject to continuous improvements , and may have been revised since the last time the observations were processed or reprocessed .
a second reason is to ensure that a single set of calibrations are applied to all datasets , so that the resulting catalog will be calibrated as homogeneously as possible . for acis ,
the principal instrument - specific calibrations that are re - applied are the ( time - dependent ) gain calibration and the correction for ccd charge transfer inefficiency ( cti ) .
the former calibration maps the measured pulse height for each detected x - ray event into a measurement of the energy of the corresponding incident x - ray photon .
cti correction attempts to account for charge lost to traps in the ccd substrate when the charge is being read out .
this effect is considerably larger than anticipated prior to launch because of damage to the acis ccds caused by the spacecraft s radiation environment .
additionally , observation - specific bad pixels and hot pixels are flagged for removal , as are `` streak '' events on ccd s4 ( acis-8 ) .
the latter apparently result from a flaw in the serial readout electronics @xcite .
pixel afterglow events , which arise because of energy deposited into the ccd substrate by cosmic ray charged particles , are removed using the acis_run_hotpix tool that is also included in ciao .
although this program can miss some real faint afterglows , such events are very unlikely to exceed the flux significance threshold required for inclusion in the catalog .
the default 0.5 pixel event position randomization in chip coordinates is used when the calibrations are reapplied .
lcl & 4.1 & calculate source aperture photometry properties + dmellipse & 4.1 & calculate ellipse including specified encircled fraction + eff2evt & 4.1 & calculate energy flux from event energies + lim_sens & 4.1 & create a limiting sensitivity map + mkpsfmap & 4.1 & look up psf size for each pixel in an image + acis_streak_map & 4.1.2 & create a high spatial frequency background map + dither_region & 4.1.2 & calculate region on detector covered by a sky region + evalpos & 4.1.2 & get image values at specified world coordinates + glvary & 4.1.2 & search for variability using gregory - loredo algorithm + pileup_map & 4.1.2 & create image that gives indication of pileup + modelflux & 4.1.2 & calculate spectral model energy flux + srcextent & 4.1.2 & compute source extent + create_bkg_map & 4.2 & create a background map from event data + dmimgpm & 4.2 & create a low spatial frequency background map + the main instrument specific calibrations for hrc data relate to the `` degapping '' correction that is applied to the raw x - ray event positions to compensate for distortions introduced by the hrc detector readout hardware .
several additional calibrations compensate for effects introduced by amplifier range switching and ringing in the hrc electronics , and a number of validity tests are performed to flag x - ray event positions that can not be properly corrected due to amplifier saturation and other effects . since data are recorded continuously during an observation , a `` mission time line '' is constructed during standard data processing that records the values of key spacecraft and instrument parameters as a function of time .
these parameters are compared with a set of criteria that define acceptable values , and `` good time intervals '' ( gtis ) that include scientifically valid data are computed for the observation .
the gti filter from standard data processing is reapplied without change as part of the recalibration process .
background event screening performed as part of catalog data recalibration is somewhat more aggressive than that performed as part of standard data processing , typically reducing the non - x - ray background . for a @xmath90 observation ,
the median catalog background rate is roughly 80% of the nominal field background rates @xcite , although there is considerable scatter .
f. a. primini et al .
( 2010 , in preparation ) include a detailed statistical analysis of the improvements to the non - x - ray background afforded by this screening .
the reduction of the background event rate is achieved by removing time intervals containing strong background flares .
these time intervals are identified separately for each chip .
first , the background regions of the image are identified by constructing a histogram of the event data , determining the mean and standard deviations of the histogram values , and rejecting all pixels that have values more than 3 standard deviations above the mean .
an optimally - binned light curve of the background pixels is then created using the gregory - loredo algorithm ( see 3.12.1 ) .
time bins for which the count rate exceeds @xmath91 the minimum light curve value are identified .
the corresponding intervals are considered to be background flares , and the gtis are revised to exclude those periods .
we emphasize that the objective of this procedure is to remove only the most intense background flares , which occur relatively infrequently .
time intervals that include moderately enhanced background rates are not rejected by this process , since their contributions increase the overall snr .
the aggregate loss of good exposure time exceeds 25% for less than 1.5% of the observations included in the catalog ; the loss is greater than 10% for 3% of the observations , and greater than 5% for 5% of the observations
. for each observation included in the csc , the recalibrated photon event list is archived , together with several additional full - field data products .
these include multi - resolution exposure maps computed at the monochromatic effective energies of each energy band and the associated ancillary data products ( aspect histogram , bad pixel map , and field of view region definition ) , used to construct them ( see table 3 ) .
for the first release of the csc , background maps are used for automated source detection .
they are created directly from each individual observation with the necessary accuracy .
the general observation background is assumed to vary smoothly with position , and is modeled using a single low spatial frequency component .
although this assumption is in general satisfied across the fields of view included in this catalog release , there may be localized regions where the background intensity has a strong spatial dependence , and therefore where the detectability of sources may be reduced .
several different approaches were considered for constructing the low spatial frequency background component , including spatial transforms , low pass filters , and data smoothing .
however , the most effective and physically meaningful technique is a modified form of a poisson mean .
this method , described below , estimates the local background from the peak of the poisson count distribution included in a defined sampling area .
the dimensions of the sampling area act effectively as a spatial low pass filter that determines the minimum angular size that contributes to the background .
high spatial frequency linear features , commonly referred to as `` readout streaks , '' result when bright x - ray sources are observed with acis .
these streaks arise from source photons that are detected during the ccd readout frame transfer interval ( @xmath92 per row ) following each exposure ( @xmath12 per exposure for a typical observation ) .
all pixels along a given readout column are effectively exposed to all points on the sky that lie along that column during the frame transfer interval , so that columns including bright x - ray sources have enhanced count rates along their length . unless accounted for by the source detection step , the increased counts in the bright readout streak are detected as multiple sources .
although readout streaks are comprised of mis - located source photons , we choose to model them as a background component .
background maps computed for acis observations include contributions from both components , while hrc background maps include only the low spatial frequency component .
the reader should note that background maps are _ not _ used when deriving source properties such as aperture photometry . instead
, a local background value determined in an annular aperture surrounding the source is used , as described in 3.4.1 .
significant spatial variations of the observed x - ray flux on the scale of the background aperture will increase the background local variance , thus reducing the significance of the source detection , perhaps below the threshold required for inclusion of the source in the catalog .
this effect is seen in some galaxy cores , where the unresolved emission contributes x - ray flux to the annular background apertures surrounding each source .
the algorithm described here is a refinement of method used by @xcite to address the impact of readout streaks on source detection .
the streak map is computed at single pixel resolution independently for each acis ccd and energy band .
the first step is to identify the bright - source - free regions on the detector . for ease of computation the orientation of the @xmath93-axis
is defined to be along the chip rows ( perpendicular to the readout direction ) and the @xmath94-axis is defined to be along the direction of the readout columns . to identify the source - free regions , the photon event totals ,
@xmath95 summed along the @xmath93-axis are constructed , and the median ( @xmath96 ) , mode ( @xmath97 ) , and standard deviation ( @xmath98 ) of the distribution of the @xmath95 values are computed .
these values provide a basic characterization of the background . from an examination of many data histograms , the maximum value of @xmath95 which can still considered background dominated is given by @xmath99,\ ] ] where @xmath100 is set to 1 .
rows for which @xmath101 include a substantial bright source contribution .
all rows with @xmath102 ( excluding off - chip and dither regions ) are considered to comprise the source - free regions and are used to calculate the streak map .
the average number of events per pixel is calculated separately for each readout column ( @xmath94-axis direction ) from all of the rows in the source - free regions .
these values are replicated across each ccd row to create an image that includes the sum of the readout streak contribution and the mean one - dimensional low spatial frequency background component .
the latter must be accounted for when combining the high spatial frequency readout streak map with the two - dimensional low spatial frequency background map . for the algorithm to obtain a good measure of the background , of order 100
bright - source - free rows are required .
this condition is satisfied for most observations
. observations with too few source - free rows poorly sample the background .
this can lead to erroneously low intensities for bright readout streaks in the resulting background map , which may enhance the false source rate along these streaks .
faint sources that fall in the source - free rows will be considered to be part of the background , which can lead to similar results .
nevertheless , the algorithm is remarkably effective , even in crowded regions such as the orion complex and the galactic center fields . an example broad - band acis streak map , created for ( m81 ) ,
is shown in figure 11 , _
left_. for each observation , a low spatial frequency background map is constructed separately for each energy band and image blocking factor ( see 3.4 ) . @xcite
provide an initial discussion of this algorithm and general background map creation . as described above , for acis observations the high spatial frequency background map includes a component that represents the one - dimensional average of the low frequency background over the rows used to create the streak map .
this component , as well as the high spatial frequency background , are removed by subtracting the streak map from the original image from which it was created . for each image blocking factor , the difference image
is constructed by subtracting the appropriately regridded streak map from the corresponding blocked original image . for each pixel in the resulting difference image , a centered sampling region with dimensions @xmath103 pixels is defined .
spatial scales smaller than @xmath104 pixels are attenuated .
the sampling regions are truncated at the edges of the images , and so some higher frequency information may propagate into the background map .
however this effect has not been found to have any significant impact on the utility of the resulting map .
a histogram of the count distribution is constructed from the pixels included in the sampling region associated with each image pixel .
the first histogram bin will typically span the count range from @xmath105 to @xmath106 for acis observations , since the readout streak map has been subtracted and there will be some negative pixels .
the low spatial frequency background at this image pixel location is computed using a modified form of a poisson mean @xmath107,\ ] ] where @xmath108 is the number of counts in histogram bin @xmath109 , @xmath110 is the bin with the maximum number of histogram counts , and @xmath111 and @xmath112 are the lower and higher bins immediately adjacent to @xmath110 .
the low spatial frequency background map is formed by computing @xmath113 for each pixel location in the image . for acis observations , @xmath114 pixels , corresponding to a spatial scale of order @xmath115 for images blocked at single pixel resolution .
figure 11 , _
right _ displays the acis broad - band low spatial frequency map for observation 00735 ( m81 ) that corresponds to the streak map shown in the left hand panel of the figure .
the first step in creating the total background map is to correct the readout streak map ( for acis observations only ) for the effects of reduced exposure near the edges of the observation that arise due to the spacecraft dither , by dividing by the appropriate band - specific normalized exposure map .
similarly , the low spatial frequency background map is corrected by dividing by the smoothed , band - specific normalized exposure map .
the smoothing that is applied to the normalized exposure map in the latter case matches the smoothing applied when constructing the low spatial frequency background map .
finally , the two background components for each energy band are summed to produce the total exposure - normalized background map that is required for source detection . figure 12 displays the central region of the broad - band acis image of m81 ( observation 00735 ) , with source detections overlayed .
the source detections shown in the left - hand panel are those that result if the background is modeled internally by wavdetect ( see 3.4 , below ) ; the panel on the right shows the source detections resulting from using the total background from fig .
11 . using the background map
has eliminated the false sources detected on the readout streak .
the total background maps for each energy band are also archived and accessible through the catalog .
these maps differ from those used for source detection in that they have been multiplied by the normalized band - specific exposure map , and are therefore recorded in units of counts . for the convenience of the user , we also store multi - resolution photon - flux images for the full field of each observation , created by filtering the photon event list by energy band , binning to the appropriate image resolution , subtracting the total background map appropriate to the energy band , and dividing by the corresponding exposure map .
candidate sources for inclusion in the csc are identified using the ciao wavdetect wavelet - based source detection algorithm @xcite .
wavdetect has been used successfully with _
data by a number of authors ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and its capabilities and limitations are well known ( e.g. , * ? ? ?
early in the catalog processing pipeline development cycle , several different methods for detecting sources were evaluated .
in addition to wavdetect , these included the ciao implementations of the sliding cell @xcite and voronoi tessellation and percolation @xcite algorithms , and a version of the sextractor package @xcite modified locally to use poisson errors in the low count regime .
the voronoi tessellation and percolation algorithm was quickly discarded because of the significant computational requirements and complexities for automated use .
a series of simulations was used to compare the performance of the remaining methods with respect to source detection efficiency for isolated point sources , the efficiency with which close , equally - bright pairs of point sources with @xmath116 and @xmath117 separations are resolved , and false source detection rate ( a. dobrzycki , private communication ; * ? ? ?
the first two properties were evaluated for point sources containing 10 , 30 , 100 , and 2000 counts , with off - axis angles 0@xmath118 with @xmath115 spacing , and nominal background rates for exposure times of 3 , 10 , 30 , and @xmath3 .
the false source rate was evaluated as a function of off - axis angle for the same exposure times .
all three detection algorithms performed reliably for bright , isolated sources located close to the optical axis . compared to the remaining methods , wavdetect had better source detection efficiency for faint sources located several arcminutes off - axis , and was able to resolve close pairs of sources more reliably than the sliding cell technique .
the locally modified version of sextractor provided inconsistent results , in some cases detecting large numbers of spurious sources .
these simulations were performed early in the catalog processing pipeline development cycle , as an aid in selecting the source detection algorithm to be used for catalog construction .
they did not make use of the background maps described in the previous section .
the actual performance of the source detection process used to construct the csc is established from more detailed and robust simulations , as described in 4 and references therein .
based on the results of the simulations , wavdetect was selected as the source detection method of choice for the csc .
the wavdetect algorithm does not require a uniform psf over the field of view , and is effective in detecting compact sources in moderately crowded fields with variable exposure and poisson background statistics . to detect candidate sources in a two - dimensional image @xmath119
, wavdetect repeatedly constructs the two - dimensional correlation integral @xmath120 for a set of marr ( `` mexican hat '' ) wavelet functions , @xmath121 , with scale sizes that are appropriate to the source dimensions to be detected .
the elliptical form of the marr wavelet may be written in the dimensionless form @xmath122 where @xmath123 and the parameters @xmath124 define the semi - major and semi - minor radii and rotation angle of the mexican hat . a localized clump of counts in the image @xmath119 will produce a local maximum of @xmath63 if the scale sizes defined by @xmath125 are approximately the same as , or larger than , the dimension of the clump . to determine whether a local maximum of @xmath63 is due to the presence of a _ source _ , the detection significance , @xmath126 , in each image pixel @xmath127
is determined from @xmath128 where @xmath129 is the number of background counts within the limited spatial extent of @xmath121 , and @xmath130 is the probability of @xmath63 given the background @xmath131 . if @xmath132 , where @xmath133 is a defined limiting significance level , then pixel @xmath127 is identified as a source pixel .
the limiting significance level used to generate the csc is set to @xmath134 .
this formally corresponds to @xmath135 false source due to random fluctuations per @xmath136 pixel image , although due to the heuristics of the algorithm , the actual number of false sources may be lower .
the situation is further complicated in our case because the final candidate source list output from the csc source detection pipeline is a combination of several wavdetect runs in different energy bands ( see below ) .
we note that reliable quantitative estimates of the false source rates and detection efficiency can only be provided through simulations , as discussed in
4 . as described in 2.5 , we impose an additional restriction on the flux significance of a source . to ensure that the flux significance requirement is the defining criterion for a source to be included in the catalog
, we have verified that the flux significances of sources that pass our wavdetect threshold extend well below that required to satisfy the flux significance rule ( see figure 13 ) .
we estimate that roughly @xmath137 of all the sources detected by wavdetect fall below this threshold .
source detection is performed recursively by applying wavdetect to multiply - blocked sky images constructed as described in 2.5.3 .
the use of a constant blocked image size maintains algorithm efficiency while not compromising detection efficiency in the outer areas of the field of view where the psf size is significantly larger than a single pixel .
applied to the csc , wavelets with scales @xmath138 , @xmath139 , @xmath140 , @xmath141 , and @xmath142 ( blocked ) pixels are computed for each image blocking factor and each energy band except for the acis ultra - soft band .
this combination of wavelet scales and image blocking factors provides good sensitivity for detection of sources with observed angular extents @xmath0 .
some point sources with extreme off - axis angles , @xmath143 , may not be detected because the size of the local psf exceeds the largest wavelet scale / blocking factor combination .
f. a. primini et al .
( 2010 , in preparation ) calibrate this effect statistically .
source detection is not performed in the acis ultra - soft energy band .
this band is impacted heavily both by increased background and by decreased effective area because of acis focal plane contamination ( the ratio of integrated background to effective area is 12 orders of magnitude larger for the _ u _ band when compared to the other acis energy bands ) . under these circumstances
we are limited by the accuracy of the background map determination ; small errors in the background map result in an unacceptable fraction of spurious source detections .
the wavdetect algorithm incorporates steps to compare nearby correlation maxima identified at multiple wavelet scales to ensure that each source is counted only once .
after duplicates are eliminated , a source cell that includes the pixels containing the majority of the source flux is constructed .
although a source cell may have an arbitrary shape , for simplicity an elliptical representation of the source region is used throughout the csc .
the lengths of the semi - axes of this source region ellipse are set equal to the @xmath144 orthogonal deviations of the distribution of the counts in the source cell .
source region ellipses for candidate sources detected within a single observation from images with different blocking factors or in different energy bands are combined outside of wavdetect to produce a single merged source list .
this step rejects any detections that have rms radii smaller than the 50% enclosed counts fraction radius of the local psf , calculated at the monochromatic effective energy of the band in which the source is detected .
such detections are likely artifacts arising from cosmic ray impacts .
candidate source detections whose centroids are closer than the local psf radius , or that are closer than @xmath145 of the mean detected source ellipse radii , are deemed to be duplicates .
if any duplicates are identified , then the detection from the image with the smallest blocking factor is kept , and if the image blocking factors are equal , then the detection with the highest significance is used .
this approach ensures that data from the highest spatial resolution blocked image will be used to detect point and compact sources .
however , knotty emission that is located on top of extended structures will tend to be identified as distinct compact sources , while the extended emission is not recorded .
numerous source - specific catalog properties are evaluated within defined apertures .
we define the `` psf 90% ecf ( enclosed counts fraction ) aperture '' for each source to be the ellipse that encloses 90% of the total counts in a model psf centered on the source position . because the size of the psf is energy dependent ,
the dimensions of the psf 90% ecf aperture vary with energy band .
we define the `` source region aperture '' for each source to be equal to the corresponding @xmath144 source region ellipse included in the merged source list , scaled by a factor of @xmath146 . like the psf 90% ecf aperture
, the source region aperture is also centered on the source position , but the dimensions of the aperture are _ independent _ of energy band .
evaluation of model psfs with off - axis angles @xmath147 demonstrates that the dimensions of the source region aperture correspond _
approximately _ to the dimensions of the psf 90% ecf ellipses for the acis broad energy band .
this is confirmed _ a posteriori _ by examining the distribution of psf aperture fractions in source and background ( see below ) region apertures of all individual catalog sources with acis broad band flux significance @xmath148 .
figure 14 demonstrates that the source region apertures typically include @xmath14995% of the psf , while the background region apertures contain @xmath15010% .
we emphasize that while these fractions are typical , the actual psf fractions , determined by integrating the model psf over the source and background region apertures and excluding regions from contaminating sources , are used for the actual determination of source fluxes ( see 3.7 ) .
comparison of the source fluxes within the psf 90% ecf aperture and the source region aperture provide a crude indication whether a source is extended .
if the flux in the source region aperture is significantly greater than the flux in the psf 90% ecf aperture , then the source region determined by wavdetect is considerably larger than the local psf , and the source is likely extended .
both the psf 90% ecf aperture and the source region aperture are surrounded by corresponding background region annular apertures . in both cases ,
the inner edge of the annulus is set equal to the outer edge of the corresponding source aperture , while the radius of the outer edge of the annulus is set equal to @xmath151 the inner radius of the source region aperture .
although the background region apertures defined in this manner include @xmath15210% of the x - rays from the source , this contamination is accounted for explicitly when computing aperture photometry fluxes .
overlapping sources could contaminate any measurements obtained through the source and background apertures . to avoid this ,
both types of apertures are modified to exclude areas that are included in any overlapping source region apertures , or that fall off the detector .
areas surrounding acis readout streaks are also excluded from the modified background apertures . _
aperture - specific catalog quantities are derived from the event data in the appropriate modified aperture . _
the fractions of the local model psf counts that are included in the modified apertures are recorded in the catalog for each source , and are used to apply aperture corrections when computing fluxes , under the assumption that the source is well modeled by the psf .
the modified source region and background region aperture definitions are recorded as fits files using the spatial region file convention @xcite .
ciao @xcite can be used to apply these regions as spatial filters to extract the photon event data for the source ( or background ) from the archived photon event list . to simplify access to file - based data products ( see table 3 ) for individual sources
, we also separately store the source region photon event list , per - band exposure maps , and per - band source region images .
these products include data from the rectangular region of the sky that is oriented north
south / east west and that bounds the background region .
each source record in the csc master sources table is constructed by combining source detections included in the source observations table from one or more observations .
a necessary first step in this process requires matching the source detections from all of the observations that include the same region of the sky .
cross - matching algorithms ( e.g. , * ? ? ?
* ; * ? ? ?
* ) are often focussed on efficiently matching large catalogs , and typically use criteria on the position difference distribution , or cross - correlation approach techniques , for identifying matches . in many cases ,
these approaches assume ( often implicitly ) that the source psf is at least approximately spatially uniform across the field of view , and comparable between the datasets being matched . however , when matching source detections across multiple _ chandra _ observations , the strong dependence of the psf size with off - axis angle must be considered explicitly , since source detections that are well off - axis in one observation are often resolved into multiple sources close to the optical axis in other observations . under these circumstances
the source positions determined by wavdetect are _ not _ comparable , and can not be used for source matching . instead , the source matching approach used for the csc
is based on the overlaps between the psf 90% ecf apertures of the source detections from the individual observations .
although empirical in nature , this algorithm works well for matching compact source detections between _ chandra _ observations .
the detailed algorithm is described in appendix a. the method identifies the overlap fractions between the psf 90% ecf apertures of overlapping source detections from the observations , and separates them into three different categories .
the first category is the simplest , where the source detections from the various observations have apertures that all mutually overlap ( fig .
15 , _ left _ ) .
this is the most common situation , and corresponds to the case where the source detections all uniquely match a single source on the sky .
roughly 90% of the @xmath153 sources in the master sources table that are linked to more than one source detection in the source observations table fall into this category .
each of the matching entries in the latter table will be associated with the corresponding master sources table entry with a `` unique '' linkage , as described in 2.3 .
in the second category , the aperture associated with a source detection in one observation overlaps the apertures associated with multiple distinct source detections from other observations .
this circumstance typically arises because source detections from a single observation are always assumed to be distinct ; this assumption can fail very far off - axis ( @xmath154 ) , where the psf size exceeds the maximum wavdetect wavelet scale / image blocking factor combination .
this category is illustrated in figure 15 , _
center _ , and arises most often because a source detection in one observation is resolved into multiple sources by one or more of the overlapping observations .
the unresolved source detection in the source observations table will be connected to all master sources table entries associated with the matching resolved source detections via `` ambiguous '' linkages , and the detection will be flagged as confused .
the x - ray photon events associated with the unresolved detection can not be distributed across the matching resolved sources . in release 1 of the csc
, source properties derived from the detection will not be used to compute the source properties included in the master sources table .
upper limits for photometric quantities could in principle be extracted from the unresolved source detection , and these would be quite valuable for variability studies . this capability will be included in a future release of the csc . in a few cases , a set of aperture overlaps can not be resolved automatically using the current algorithm .
this third category typically occurs when there are multiple overlapping , confused source detections . in this case
, the source detections are flagged for review by a human , who is then responsible for resolving the matches .
only 415 ( out of 94,676 ) master sources include source detections that required manual review , and the majority of these were readily resolved as confused `` pairs of pairs . '' the latter case , which is illustrated in figure 15 , _ right _ commonly occurs because a pair of source detections included in a single observation both overlap a pair of source detections in another observation .
each of the source detections in the first observation overlaps both source detections in the second observation , making the detections confused , and vice - versa .
if a manual review is required to complete a match for a specific source , then a flag is set in the catalog to indicate this fact .
the actual processing required to perform the matches is complex . whenever new overlapping source detections are identified during catalog processing , the set of matches is recomputed using all of the observations processed so far .
the algorithm then queries the prior state of the catalog , and determines the set of updates that are necessary to migrate from that state to the newly determined state .
this procedure works regardless of the order in which observations processed , and also permits an already - processed observation to be reprocessed should an error have occurred .
each distinct x - ray source included in the master sources table is assigned a name that is derived from the source s location on the sky .
catalog sources are designated `` cxo j@xmath155 , '' where @xmath156 and @xmath157 are the icrs right ascension and declination , respectively , of the source position , _ truncated _ to the indicated precision .
this format complies with the international astronomical union ( iau ) recommendations for nomenclature .
the `` cxo '' prefix is registered with commission 5 of the iau for exclusive use in source designations issued by the _
chandra _ x - ray center .
the name assigned to a source is determined from the combined source position once the detections of the source are merged according to 3.4.2 .
if the source has never been included in a released version of the csc , then the source name may be revised if a source detection in a subsequently processed observation modifies the combined source position .
therefore , the name assigned to a source that is visible in a database view _ may _ change as new observations are processed , if the source has never been included in a catalog release .
once a source is included in a released version of the csc , then the name of that source is frozen .
the name will not be changed in either future catalog release views ( i.e. , subsequent catalog releases ) or database views , even if additional observations refine the source position . therefore minor discrepancies can arise between the latter and the source designation . however
, if an observation included in a later catalog release resolves an apparently single source included in an earlier version of the catalog into multiple distinct sources , then the previous source designation is retired and new names are assigned to the resolved sources . within a single observation ,
the detected source positions are those assigned by the wavdetect algorithm .
their accuracy can be estimated by evaluating the mean ( signed ) coordinate differences between the wavdetect positions and the positions of matching sources in the seventh data release ( dr7 ) of the sloan digital sky survey ( sdss ; * ? ? ?
the latter are extracted from the csc / sdss cross - match catalog ] .
as shown in figure 16 , the mean coordinate differences demonstrate good position agreement between the csc and sdss for sources with @xmath158 ( where we have restricted the comparison to include only individual observations of csc sources with at least 500 net counts to minimize statistical errors )
. for larger off - axis angles , figure 16 suggests that there may be a systematic offset between the wavdetect source positions and the sdss source positions .
the measured mean position difference is @xmath159 for @xmath160 , but appears to increase with off - axis angle .
the exact cause of this offset is uncertain .
some authors ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) have reported a similar effect , which they attribute to centroiding errors introduced by of the asymmetric nature of the _ chandra _ psf at large off - axis angles .
high quality psf simulations generated using the saotrace ( formerly saosac ) ray - trace code @xcite confirm that the asymmetry can displace the measured centroid from the requested location of the simulated psf on a uniform pixel grid .
we designate the vector orientation from the measured centroid position to the requested location of the psf as @xmath161 .
as shown in figure 16 , the expected centroid displacement along @xmath161 computed from the simulations is not a good measure of the actual mean coordinate difference , which is a factor of order 2 times smaller than predicted by the models .
one possible reason for the disagreement between the model and actual measurements is that the _ chandra _ plate scale calibration was derived from observations of ngc 2516 and lmc x-1 using source centroid measurements that were _ not _ corrected for the asymmetry of the psf @xcite . as described in 2.5.3 , the celestial coordinates of a source observed by _
chandra _ are computed by applying a series of transforms to the measured position of the source on the detector .
the final step in this process applies the measured plate scale calibration to the difference between the position of the source on the virtual sky pixel plane and a known fiducial point , which is typically the telescope optical axis .
if all of the star - star baselines used to calibrate the plate scale were oriented parallel to @xmath161 , then the lack of correction for the psf asymmetry would to first order compensate for the linear component of the systematic position offset when measuring real sources whose locations are fixed in world coordinates rather than virtual pixel plane coordinates .
since not all star - star baselines were so aligned , some residual systematic position offset may be expected , but at an undetermined level that is less than predicted from the psf simulations .
systematic position offsets can also arise from uncertainties in the detector geometry . as an example , consider imaging observations that use the nominal acis - i aimpoint .
sources with @xmath158 will be located on the same ccd array as the aimpoint . as @xmath162 increases , an increasing fractions of sources will instead be positioned on acis - s array ccds , until for @xmath163 all sources
will be located on the acis - s array .
uncertainties in the relative positions and tilts of the acis - i and acis - s arrays would therefore introduce systematic position offsets for sources with large off - axis angles , while not impacting sources that fall on the same ccd array as the aimpoint .
this signature is consistent with the absence of mean position differences measured for @xmath158 . because of the small magnitude of the mean position differences measured for @xmath160 , we have chosen not to apply an uncertain correction to source positions in release 1 of the catalog .
we plan to investigate in detail the cause of the systematic position offsets at large off - axis angles , and adjust source positions in future catalog releases if appropriate .
in addition to reporting measured source positions , wavdetect also reports positional errors associated with each source detection .
the reported errors are based on a statistical moments analysis , and do not consider instrumental effects such as pixelization , aspect - induced blur , or asymmetrical psf structure that may contribute to the total positional uncertainties .
simulations that compare the reference positions of artificially generated sources with their positions determined by wavdetect indicate that the positional uncertainties computed by wavdetect are underestimated for sources with large off - axis angles .
the simulation results , which quantify the dependence of positional uncertainties of simulated sources on off - axis angle , were found to be consistent with the more extensive simulations uses to construct the _
chandra _ multiwavelength project ( champ ) x - ray point source catalog @xcite . in the first release of the csc
, source position error ellipses are substituted by error circles computed using the champ positional uncertainty relations @xmath164 in these equations , @xmath165 is the positional uncertainty in arcseconds , @xmath162 is the off - axis angle in arcminutes , and @xmath166 is the source net counts reported by wavdetect . these relations were derived to characterize the positional uncertainties at the 95% confidence level of x - ray point sources in the champ x - ray point source catalog , which includes @xmath167 x - ray sources detected in 149 _ chandra _ observations .
the values of @xmath168 computed using equations ( [ eqn : pu ] ) are not equal at the boundary where @xmath169 ( roughly 138 net counts ) .
however , this error is negligible for @xmath170 .
although hrc observations are not included in the first release of the csc , we have used a series of simulations to derive an improved positional uncertainty relation that is appropriate for sources detected in hrc - i observations .
the simulations include @xmath171 point sources spanning @xmath172 and @xmath173 .
the best fit surface for the 95% position uncertainty quantile is @xmath174 the simulations do not sample the region with @xmath143 and @xmath175 , and so we impose an upper bound of @xmath176 on this relation , to cap the uncertainty in this regime . in release 1.1 of the csc , positional uncertainties for sources detected in acis observations
are computed using equation ( [ eqn : pu ] ) while positional uncertainties for sources detected in hrc - i observations are computed using equation ( [ eqn : hpu ] ) .
the positional uncertainties from equations ( [ eqn : pu ] ) and ( [ eqn : hpu ] ) provide a good measure of the statistical uncertainty of the location of the source in the frame of the observation , but do not consider potential sources of error that are external to the observation .
these include the error in the mean aspect solution for the observation , the astrometric errors in the axaf ( _ chandra _ ) guide and acquisition star catalog @xcite , and the calibration of the geometry of the spacecraft and focal plane . as described in 4.3 ,
@xcite has recently used the csc / sdss cross - match catalog to calibrate the combined external error by analyzing the statistical distribution of the measured separations of csc point source detections from individual observations with their counterparts in sdss dr7 @xcite .
the resulting external astrometric error is @xmath177 ( @xmath1 ) , or @xmath178 ( 95% confidence ) .
the latter must be added in quadrature to the position uncertainties from equations ( [ eqn : pu ] ) and ( [ eqn : hpu ] ) to compute the absolute position error for csc sources . in release 1 of the catalog , the positional error reported in the catalog tables
is taken directly from equation ( [ eqn : pu ] ) , so the quadrature addition of the external astrometric error component _ must be performed by the user .
_ release 1.1 of the csc will _ include _ this error component directly in the tabulated values .
a _ post facto _ histogram of the 95% confidence positional uncertainties , including the external astrometric error , for source detections included in release 1 of the csc is presented in figure 17 .
the figure demonstrates that statistical errors due to lack of net source counts does dominate the positional uncertainty for all but the brightest sources .
however , for bright sources detected at small off - axis angles , the external astrometric error limits the accuracy of the derived source positions .
any error introduced by using uncorrected source centroid positions from wavdetect is negligible for the overwhelming majority of source detections
. improved estimates of the position and positional uncertainty of each x - ray source are determined from the statistically independent source detections included in the set of individual observations using a multivariate optimal weighting formalism .
this technique is effective in cases where simple averaging fails , for example when the area defining the source position varies significantly from observation to observation .
we express the uncertainties of the estimates in the form of error ellipses centered upon the estimated source positions .
an equivalent approach has been used for weapons targeting @xcite . to our knowledge
the usage here is the first documented application to astrophysical data . in the multivariate optimal weighting formalism ,
given a set of estimates , @xmath179 , of the mean of some two - dimensional quantity , and the @xmath180 covariance matrices , @xmath181 , associated with these estimates , an improved estimate , @xmath93 , of the mean , and the associated covariance matrix , @xmath182 , are ( e.g. , * ? ? ? * ) @xmath183^{-1}.\end{aligned}\ ] ] for the application described here , we take @xmath179 to be the @xmath184th estimate of the source position , projected onto a common tangent plane ( which is constructed at the mean position of the ellipse centers ) .
the corresponding covariance matrix is @xmath185 where @xmath186 and @xmath187 are the lengths of the semi - minor and semi - major axes , respectively , of the @xmath184th error ellipse projected onto the common tangent plane , and @xmath188 is the angle that the major axis of the @xmath184th error ellipse makes with respect to the tangent plane @xmath189 axis .
the derivation of equation ( [ eqn : cov ] ) is presented in appendix b. once the covariance matrices corresponding to the error ellipses for each individual source observation , equation ( [ eqn : cov ] ) , are computed , the error ellipses are combined using equation ( [ eqn : mow ] ) .
this yields the optimally weighted source position and position error ellipse for the combined set of observations , on the common tangent plane .
mapping these back to the celestial sphere provides the combined source position and error ellipse estimates .
the observed spatial extent of a source is estimated using a rotated elliptical gaussian parameterization of the form @xmath190,\end{aligned}\ ] ] where @xmath191 where @xmath192 is the cartesian center location of the gaussian , and the parameters @xmath193 are the @xmath1 radii along the major and minor ellipse axes , and the position angle of the major axis of the ellipse , respectively .
the parameters of @xmath64 are determined using a wavelet - based approach that is similar to that used for source detection .
the methods differ in their details and assumptions , however . for source detection ,
the choice of wavelet scales used by wavdetect to detect sources is determined _ a priori_. because of the strong variation of psf size with off - axis angle , multiple wavelet scales and input image blocking factors are required to search for sources at all off - axis angles , as described in 3.4 .
stepping between the discrete wavelet scales and image blocking factors as a function of off - axis angle introduces small but systematic biases in the derived dimensions of the source region ellipses ( and therefore the source region apertures ; see figure 18 ) .
photometric , spectral , and temporal properties determined from the x - ray events included in the aperture are not impacted by these effects , since the aperture dimensions are sufficiently large that they typically enclose @xmath194 of the psf counts , and a correction factor is applied for the fraction of the psf that falls outside of the aperture . however ,
these biases render the source region aperture dimensions unsuitable for use as a measure of the source extent .
the wavelet - based approach used for estimating the source extent determines the optimal wavelet scale size directly from the data under the assumption that a source exists at approximately the location determined by wavdetect .
the raw source extent estimates derived using this approach vary smoothly with off - axis angle ( fig .
the two - dimensional correlation integral @xmath195 where the region of interest is @xmath196 and @xmath197 , and @xmath121 is again specified by equation ( [ eqn : mexhat ] ) , is computed first .
we choose a coordinate system in which the peak of @xmath64 is centered at the origin .
the quantity @xmath198 has a maximum value at the origin when @xmath199 and @xmath200 @xcite .
the source parameters are determined by maximizing @xmath201 . in practice ,
equation ( [ eqn : corrint2 ] ) is evaluated as a discrete sum over the pixels of the image . although integration of equation ( [ eqn : mexhat ] ) does not yield a simple closed - form solution , and numerical integration
is computationally expensive , for the purpose of optimizing @xmath202 , a rectangular approximation for the integral over each pixel is sufficient . in this approximation ,
@xmath203 where @xmath204 is evaluated at the center of each pixel and the pixel area is @xmath205 .
a small sub - image of the source is extracted centered on the source position determined by wavdetect .
the accuracy of this source position is refined by searching the center of the sub - image for the coordinates @xmath206 that maximize @xmath207 . a new sub - image
is then extracted using the improved source position .
finally , the size and orientation of the elliptical gaussian source parameterization are derived by maximizing @xmath208 .
choosing good initial values for @xmath209 helps to ensure that this optimization step converges reliably .
we set the initial values @xmath210 $ ] , where @xmath211 are the ellipse semi - axes derived by wavdetect , and the value of @xmath212 is obtained by examining @xmath213 on a grid of @xmath110 values spanning the half - width of the source image ; @xmath212 is usually the smallest @xmath110 that corresponds to either a local maximum or an inflection point .
this choice is motivated by the observation that when a single source is present , the location of a local maximum provides a good estimate of the source size ( see figure 19 ) .
similarly , when the source of interest is blended with other nearby sources , an inflection point where @xmath214 often occurs near the `` edge '' of the central source .
when the first occurrence of @xmath215 occurs at a local minimum , @xmath212 is the smallest value of @xmath110 on the pixel grid .
the spatial extent of the local psf is determined for comparison with the observed source extent , and as an aid to assessing the intrinsic extent of the source .
since the size of the _ chandra _ psf is a strong function of off - axis angle , a ray - trace model is constructed at the measured off - axis and azimuthal angles @xmath216 separately for each detected source .
although the shape of the psf is energy - dependent , within each energy band the ray - trace model is computed only at the monochromatic effective energy ( see 2.5.2 ) of the band .
this approximation results in an error that is dependent on the actual spectrum of the source , but that does not exceed @xmath36 for typical power - law or black - body source spectra . the ray - trace model is computed using version 1.0.0 of the saotrace simulation code @xcite with the latest hrma optical coefficients calibration model configuration . ] derived from calibration observations .
a ray density of @xmath217 is used for the ray - trace .
the rays are then projected onto the detector focal plane , a gaussian blur is applied to account for the degradation due to de - dithering the telescope motion ( @xmath218 for acis , @xmath219 for hrc ) , and the image is resampled onto the pixel plane of the detector .
each psf image is recorded in the csc as a file - based data product ( see table 3 ) , which can be retrieved by the catalog user and compared directly to the corresponding source region image .
the ray density used in the ray - trace models typically produces a total of @xmath25@xmath220 counts ( with a full range of @xmath221@xmath222 counts ) in the resulting psf model in the acis broad , hard , and soft energy bands , where the combined hrma / instrument effective area at the monochromatic effective energies of the bands is @xmath223@xmath224 .
the number of counts in the acis medium energy band psf models is about 60% higher , because of the higher effective area at that band s monochromatic effective energy . for the acis ultra - soft energy band
the total psf counts may be only a few hundred because of the poor quantum efficiency in the band .
although the psf models computed here are sufficient for their intended purposes of providing basic estimates of source extent and point source aperture corrections for aperture photometry , they are not suitable for analyses such as image deconvolution that require detailed psf models .
once the image of the local model psf is constructed , the rotated elliptical gaussian parameterization , @xmath225 , of the spatial extent is computed in the same way that the observed source extent , @xmath226 , is computed from the source image . using
the rotated elliptical gaussian parameterizations derived above , the observed source extent , @xmath227 , can be treated as the convolution of the local psf , @xmath225 with the intrinsic source extent , @xmath228 , where we parameterize the latter similarly to equation ( [ eqn : egaus ] ) . in general , @xmath229 , since the psf - convolved ellipse need not have the same orientation as the intrinsic source ellipse . in principle
, one can determine the parameters , @xmath230 , of the intrinsic source ellipse by solving a nonlinear system of equations involving the psf parameters , @xmath231 , and the observed source parameters , @xmath232 .
however , because these equations are based on simple assumptions regarding the source and psf profiles , and because the input parameters are often uncertain , such an elaborate calculation seems unjustified .
a much simpler and more robust approach makes use of the identity @xmath233 which applies to the convolution of two elliptical gaussians having arbitrary relative sizes and position angles . using this identity , one can define a _ root - sum - square _ intrinsic source size , @xmath234^{1/2 } , \end{aligned}\ ] ] that depends only on the sizes of the relevant ellipses and is independent of their orientations .
this expression is analogous to the well - known result for convolution of one - dimensional gaussians and for convolution of circular gaussians in two dimensions .
the factors of @xmath235 ensure that the statistic value gives the radius of the source image when applied to circular source images . using equation ( [ eqn : arss ] )
, one can derive an analytic expression for the uncertainty in @xmath236 in terms of the measurement errors associated with @xmath237 and @xmath238 .
because @xmath237 and @xmath238 are non - negative , evaluating the right - hand side of equation ( [ eqn : arss ] ) using the corresponding mean values should give a reasonable estimate of the mean value of @xmath236 . a taylor series expansion of the right - hand side of equation ( [ eqn : arss ] ) evaluated at the mean parameter values
therefore yields the uncertainty @xmath239^{1/2 } , \end{array}\ ] ] where @xmath240 represents the variance in @xmath93 , and where @xmath241 measurements of the mean intrinsic source extent derived from multiple independent observations , @xmath242 , are combined using the multivariate optimal weighting formalism , equations ( [ eqn : mow ] ) .
the minimum variance estimator of the intrinsic source size is the variance - weighted mean , @xmath243\sum_i{\rm var}[a_{{\rm rss},i}]^{-1}a_{{\rm rss},i},\ ] ] where @xmath244 = ( \delta a_{{\rm rss},i})^2 $ ] .
the variance in @xmath245 is @xmath246 = \left[\sum_i{\rm var}[a_{{\rm rss},i}]^{-1}\right]^{-1}.\ ] ] net source counts , count rates , and photon and energy fluxes for point sources are computed from counts and exposure data accumulated in independent source and background apertures , @xmath247 and @xmath248 .
typically , these apertures are simple elliptical regions and surrounding elliptical annuli , but arbitrary areas from either aperture may be excluded to avoid contamination from nearby sources , or missing data due to detector edges .
the net aperture areas , @xmath249 and @xmath250 , and the fractions , @xmath66 and @xmath251 , of source counts expected in both apertures are determined from @xmath252 and @xmath253 we use the psfs described in 3.6.1 to estimate @xmath66 and @xmath251 .
although the finite number of psf counts leads to some uncertainty in the estimate for @xmath251 , the effect of this uncertainty on the derived net counts , count rates , and fluxes is small ( typically @xmath254 ) .
if a uniform background over the scale of @xmath247 and @xmath248 is assumed , then the net source counts with aperture corrections applied , @xmath64 , can be determined by solving the simultaneous set of linear equations @xmath255 where @xmath63 and @xmath131 are the total counts in @xmath247 and @xmath248 , respectively , @xmath111 represents the background in @xmath247 , and @xmath256 .
the solution is @xmath257 in general , selecting a background aperture so that @xmath258 is difficult , since the inner radii of such annuli could range from @xmath259 to @xmath260 , depending on @xmath162 and energy .
such large background apertures would be subject to errors due to intrinsic background variations , diffuse source emission , and background contributions from multiple detector chips .
we choose rather to use smaller apertures , containing @xmath15210% of the source flux ( see fig .
14 ) , whose effects can be modeled more accurately . to apply a consistent statistical approach in determining confidence bounds for all photometric quantities ( see below ) , we assume that the generic photometric quantity @xmath64 ( whether counts , count rate , photon flux , or energy flux ) can be converted to counts by multiplying by appropriate generic conversion factors @xmath261 and @xmath262 defined below , averaged over @xmath247 , @xmath248 , and generalize equations ( [ eqn : getc ] ) to include these terms , @xmath263 for example , if @xmath64 represents a count rate , @xmath264 and @xmath265 , where @xmath266 and @xmath267 represent average exposure times in @xmath247 and @xmath248 , respectively .
the corresponding definitions for photon flux are @xmath268 and @xmath269 , where @xmath270 and @xmath271 are the average exposure map values ( in @xmath272 ) , computed at the monochromatic effective energy of the band , in @xmath247 and @xmath248 , respectively . for energy flux ,
@xmath273 and @xmath274 , where @xmath275 and @xmath276 represent average fluxes ( in @xmath277 ) in @xmath247 and @xmath248 , respectively .
the values @xmath275 and @xmath276 are determined by applying quantum efficiency and effective area corrections to individual event energies in @xmath247 and @xmath248 , and computing the averages of the resulting quantities .
finally , we relax the assumption of uniform background over @xmath247 and @xmath248 by defining @xmath278 for source rate , @xmath279 for photon flux , and @xmath280 for energy flux . with these definitions , the general solution for @xmath64
may be written as @xmath281 to determine confidence bounds for @xmath64 , the background marginalized posterior probability density is computed first , with the assumption that @xmath63 and @xmath131 are poisson - distributed random variables whose means are @xmath282 and @xmath283 , respectively . the posterior probability density for @xmath64
may then be written @xmath284 to determine @xmath285 , we use bayes theorem to write the joint posterior probability density for @xmath286 , taking advantage of the fact that @xmath247 and @xmath248 are independent : @xmath287 the likelihoods are simple poisson probabilities , @xmath288 and we use generalized @xmath289-priors for @xmath290 and @xmath291 : @xmath292 where the parameters @xmath293 , @xmath294 , @xmath295 , and @xmath296 define the function shapes , and @xmath297 is determined through normalization of @xmath286 . once @xmath286 is known , @xmath285 may be found from simple substitution of variables , @xmath298 details of the derivation may be found in v. kashyap & f. a. primini ( 2010 , in preparation ) , but here we merely cite the final results under the additional assumption of non - informative priors @xmath299 and @xmath300 : @xmath301 because of the computationally intensive nature of this expression , @xmath302 is approximated with an equivalent gaussian distribution when @xmath303 counts .
the validity of this approximation is verified through simulations .
examples of @xmath304 for three csc sources are shown in figure 20 . by using the different definitions for @xmath261 and @xmath262 as described above , probability densities for net counts , rates , and photon and energy fluxes
can then be computed .
confidence bounds are determined by numerically integrating @xmath304 in alternating steps above and below its mode until the desired confidence level is achieved .
the values of @xmath64 at these points then determine the confidence bounds .
if the value of @xmath305 is reached before summation is complete , or if the mode itself is 0 , integration continues for @xmath64 above the mode , and the resulting bound is considered an upper limit .
we note that in our approach , photon flux and energy flux are determined somewhat differently .
while average exposure map values are used when computing photon flux , in the case of energy flux the values @xmath275 and @xmath276 are determined by applying quantum efficiency and effective area corrections to individual event energies .
for sources with few counts in either the source or background region , a single photon detected at an uncharacteristically low or high energy ( where the _ chandra _ effective area is small ) can make a dominant contribution to the estimated energy flux . in such cases ,
the true uncertainty will be significantly larger than our estimated errors .
a _ post facto _ comparison of energy flux estimates computed in this manner with energy fluxes calculated using an assumed canonical power - law spectral model ( see 3.10 , below ) , indicates that fewer than 1% of acis broad energy band fluxes are affected by this problem .
a more detailed analysis of the statistical accuracy of the energy flux determinations is provided by f. a. primini et al .
( 2010 , in preparation ) .
significances for all aperture photometry quantities are determined directly from the probability densities @xmath304 .
our goal is to provide a simple statistic that is robust to calculate , easily interpretable by non - expert users , and consistent with the classical snr definition for high count sources . to this end , we compute the @xmath306 of @xmath304 , since the latter has a well - defined width even for low - significance sources in the catalog , as shown in figure 20 . if @xmath305 is reached before the half - maximum point below the mode is found , the @xmath307 is computed from values above the mode and @xmath306 is set equal to @xmath308 .
the @xmath309 is then used to compute the `` equivalent @xmath310 '' for a gaussian probability density , @xmath311 the flux significance value that is reported in the catalog for a source is defined to be @xmath312 , where @xmath64 is determined from equation ( [ eqn : sgeneral ] ) .
this value must be at least 3.0 in at least one energy band for an observation of a source to be included in the first release of the catalog .
the flux significance threshold that we use imposes a conservative limit on sources included in the csc , which we deem necessary to reduce the number of spurious sources at low count levels to an acceptable value . comparing our results to those of other large _ chandra _ surveys whose
source lists are derived from wavdetect , but whose detection procedures differ , is useful . in figure 21 , we compare the distribution of net counts for csc sources detected in the acis broad ( @xmath20@xmath21 ) energy band with distributions of similar quantities for four other _ chandra _
catalogs derived from a range of acis exposures comparable to those in the csc : aegis - x ( * ? ? ?
* @xmath20@xmath21 ) , the galactic center catalog ( * ? ? ?
* @xmath20@xmath313 ) , c - cosmos ( * ? ? ?
* @xmath20@xmath21 ) , and champ ( * ? ? ?
* @xmath20@xmath313 ) .
we note that while these other catalogs do include sources with fewer net counts than the csc , the additions are in general not large , comprising @xmath314 , @xmath315 , @xmath315 , and @xmath316 , for aegis - x , the galactic center catalog , c - cosmos , and champ , respectively .
we attribute the larger percentage in champ to the restricted fields - of - view and the careful manual screening of source detections used when constructing that catalog .
the csc appears to fare worse in comparison to the xbootes survey ( * ? ? ?
* @xmath20@xmath21 ) , most of whose sources have fewer than 10 net counts . however , the xbootes survey is composed of many @xmath16 non - overlapping observations for which the very low acis background allows a lower count threshold .
in contrast , the csc is constructed from observations comprising a wide range of exposures , @xmath81 of which are greater than @xmath16 and @xmath36 of which are greater than @xmath317 . finally , as mentioned in 3.4 , @xmath137 of all sources detected by wavdetect in the acis broad energy band fall below the flux significance threshold .
however , we expect that a substantial fraction of these sources are spurious
. ideally , one should compute aperture photometry quantities for combined observations by computing a joint probability density , using @xmath304 from one observation as the prior for the next .
however , this approach is difficult to implement and computationally expensive , especially when probability densities from individual observations do not overlap significantly .
we have therefore chosen simply to combine aperture data from various observations and compute a single @xmath304 from those data .
we compute @xmath318 to cast these in the same form as equations ( [ eqn : cgeneral ] ) , we define @xmath319 , where the @xmath238 are determined from the solutions to equations ( [ eqn : cgeneral ] ) for individual observations .
one can then write @xmath320 which are identical in form to equations ( [ eqn : cgeneral ] ) . the combined aperture photometry quantities and bounds can then be determined as described earlier . for the purposes of the catalog , `` limiting sensitivity ''
is defined to be the flux of a point source that meets but does not exceed the flux significance threshold for inclusion in the catalog . limiting sensitivity is a function of source position , background , and the algorithm used to calculate flux and flux significance . at any point within the field of view of an observation , the limiting sensitivity can be used as a simple x - ray flux limit for individual sources detected at other wavelengths . for the catalog ,
a full - field sensitivity map is provided for each observation and energy band as a file - based data product ( see table 3 ) .
these data are also required to calculate sky coverage histograms ( solid angle surveyed as a function of limiting flux ) , which are themselves needed to calculate luminosity functions and source surface brightness versus number density relationship . as described above
, the flux significance of a catalog source is defined to be the ratio of the source flux to the equivalent @xmath310 determined from the width of the flux s posterior probability density .
there is no equivalent quantity for sensitivity , and for simplicity and ease of computation , we use a technique similar to that developed by @xcite for the _ einstein _ observatory medium sensitivity survey , namely , we approximate significance using the aperture photometry relations of 3.7 , under the assumption of gaussian statistics , and use the model background maps , randomized to provide statistics appropriate to the observation in question , to determine aperture counts .
recall from equation ( [ eqn : sgeneral ] ) that the flux may be written as @xmath321 since @xmath63 and @xmath131 are independent random variables , the variance on @xmath64 may be written @xmath322 assuming gaussian statistics .
the significance , @xmath323 may then be written as @xmath324 the limiting sensitivity is found by determining the minimum number of counts @xmath325 in the source aperture that yields the flux significance threshold @xmath326 in equation ( [ eqn : snr ] ) , @xmath327 whose solution is @xmath328 and the limiting sensitivity for that aperture is then given by equation ( [ eqn : flux ] ) , @xmath329 where we have approximated @xmath330 since the limiting sensitivity maps are computed from background maps with no real sources , information about real source apertures is unavailable .
rather , for each element in the map , circular source and annular background apertures appropriate to the 90% ecf source aperture at that location are constructed , and used to determine @xmath131 , @xmath331 , and @xmath261 for use in equation ( [ eqn : ls ] ) .
the assumption of gaussian statistics , and the subsequent simplification in the algorithm , is made of necessity , since limiting sensitivity must be computed not for each source but for each pixel in each of 5 energy band images .
we have , however , verified the performance of the algorithm by comparing detected source fluxes with values of limiting sensitivity at the source locations , for thousands of catalog sources in all energy bands ( f. a. primini et al .
2010 , in preparation ) , and find good agreement . for observations of sources with at least 150 net counts in the energy band @xmath20@xmath332 obtained using the acis detector
, we further characterize the intrinsic source properties by attempting to fit the observed counts spectrum with both an absorbed black - body spectral model and an absorbed power - law spectral model .
these two models represent basic spectral shapes of thermal and non - thermal x - ray emission .
the standard forward fitting method used in x - ray spectral analysis computes the predicted counts produced by the spectral model with the observed counts in the detector channel space , and iteratively refines the model parameters to improve the quality of the fit .
instrumental response functions @xcite define the mappings between physical ( source ) space and detector space .
@xcite describe two of these calibration files , the detector redistribution matrix file ( rmf ) and the ancillary response file ( arf ) .
the former specifies the energy dispersion relation @xmath333 that defines the probability that a photon of actual energy @xmath60 , location @xmath334 , and arrival time @xmath335 will be observed with an apparent energy @xmath336 and location @xmath337 , while the instrumental effective area @xmath338 is recorded in the latter .
the final dispersion relation is the photon spatial dispersion @xmath339 transfer function due to the instrumental point spread function . with these definitions , the model @xmath340 that describes the expected distribution of counts arriving at the detector is then @xmath341 where @xmath342 is the physical model that defines the physical energy spectrum , spatial morphology , and temporal variability of the source .
we follow standard practice by ignoring the dependency on photon arrival time , and instead consider only the total number of photons that arrived during the observation in the forward fitting process .
the source position and shape are taken as known , and we assume that the source photons are collected from the detector area containing an entire source region of interest .
the latter assumption is valid provided that sources are spatially separated on scales of order the size of the psf or larger . in crowded fields , or for sources that have a complex diffuse structure , the contribution from the other sources are important . with the assumptions listed above , equation ( [ eqn : spmod ] )
reduces to @xmath343 where the source emitted spectrum @xmath344 depends on the source physics .
the forward fitting procedure solves for the best fit parameters for @xmath344 , assuming a pre - defined fit statistic .
since spectral fitting is only performed for sources with a minimum of 150 net counts , a @xmath345 fit statistic is used , but note that this assumes a gaussian distribution for the source counts . for _ all _
sources observed using the acis detector ( i.e. , not just those with at least 150 net counts in the broad energy band ) , the catalog processing pipelines extract the observed energy spectra of the photons included in the source and background regions of each detected source and store these in a standard format ( pha file ; * ? ? ? * ) .
an appropriate associated arf and rmf are computed by weighting the instrumental responses based on the history of how the source and background regions move over the surface of the detector due to the spacecraft dither motion .
the extracted spectra , and associated arf and rmf are stored as file - based data products ( see table 3 ) and can be retrieved by the user for further analysis such as low - count spectral fitting or spectral stacking . to fit the background subtracted data , each pha spectrum is grouped to a minimum of 16 counts per channel bin , and the source model parameters are varied to minimize the @xmath345 statistic ( assuming data variance , @xmath346 ) .
two models are applied to the data in order to evaluate source properties : ( 1 ) an absorbed blackbody model @xmath347 ; and ( 2 ) an absorbed power law model @xmath348 . in these models
, @xmath349 is the equivalent hydrogen column density , @xmath350 is the photo - electric cross - section based on @xcite and metal abundances from @xcite , @xmath61 is the model normalization at @xmath351 , @xmath352 is the blackbody temperature , and @xmath353 is the power - law photon index . forward fitting is performed using the _ sherpa _ fitting engine @xcite .
_ sherpa _ finds the best fit model parameters and calculates two - sided confidence intervals for each significant parameter .
the model flux for the best fit parameters over the energy range @xmath20@xmath354 is also calculated .
the @xmath355 confidence limits for each parameter are calculated using the `` projection '' method in _
sherpa_. this method finds the two - sided confidence bounds independently for each parameter .
the algorithm assumes that the current model has been fitted , and that all of the parameters are at the values corresponding to a best fit which is at the minimum of the fit statistic ( @xmath356 ) . for each parameter of interest ,
the search for the lower or upper bound starts at the best fit value , which is then varied along the parameter axis . at each new value ,
the parameter of interest is frozen and a new best fit model is determined by minimizing @xmath345 over the remaining thawed parameters .
the new @xmath345 statistic , @xmath357 , is determined and the difference between the new and the minimum statistics , @xmath358 , is calculated . a change in @xmath359 equal to 1 corresponds to a @xmath360 confidence bound @xcite , so the parameter of interest is varied until this value of @xmath359 is obtained .
we note here that energy - dependent aperture corrections are not applied when performing the spectral model fits .
since the _ chandra _ psf is somewhat more extended at higher energies , the lack of correction has the effect of slightly softening the calculated spectral slope .
the correction in @xmath353 is approximately 0.030.05 for power - law spectra with a wide range of spectral indices . for sources included in release 1 of the csc for which spectral fits have been performed ,
the error in @xmath353 introduced by not applying energy - dependent aperture corrections is roughly six times smaller than the median computed @xmath1 confidence limits .
about 2.5% of sources with spectral fits have computed confidence limits @xmath361 , and these cases appropriate caution should be exercised when using the spectral fit properties . for most source properties , values recorded in the master sources table
are computed by combining the relevant data from the set of observations in which the source is detected . however , for simplicity , the spectral model fit parameter values recorded in the master sources table are taken directly from the single observation of the source that has the highest significance , equation ( [ eqn : snr ] ) . in this case , data from multiple observations are not combined to compute the master source table spectral fit properties .
spectral model fits are not performed for sources with @xmath362 net counts . however , for all sources we estimate energy fluxes using canonical absorbed power - law and black - body spectral models . for a canonical source model @xmath344 whose integral over the energy band
is @xmath363 , an corresponding band count rate , @xmath364 in @xmath365 , can be computed from the effective area calibration , @xmath366 , and the rmf , @xmath367 .
the count rate is @xmath368 , where the integral is performed over the energy band . for hrc observations ,
a diagonal rmf is assumed .
the actual flux of a source can be estimated from @xmath363 by scaling the latter by the ratio of the measured and modeled source aperture count rates in the energy band .
since the only free parameter in this case is the normalization of the model , the calculation can be performed for sources with too few counts for a reliable spectral fit .
the canonical power - law spectral model has a fixed photon index @xmath68 , which falls in the range of values ( @xmath3692.5 ) that are typical of agn spectra @xcite .
the value chosen matches the photon index used to convert count rates to energy fluxes in the second xmm - newton serendipitous source catalog ( 2xmm ; * ? ? ?
* ) , to simplify comparison of csc and 2xmm source fluxes .
since we anticipate that the majority of sources with spectra that are best fit by a power - law model are agn , we fix the total neutral hydrogen absorbing column @xmath370 equal to the galactic column , @xmath371 , under the assumption that this represents a lower limit to the true column density .
the canonical black - body spectral model has a fixed temperature @xmath372 and total neutral hydrogen column density @xmath373 .
the latter value matches the median column density identified by @xcite , and also corresponds to the typical column density found within @xmath374 of the sun @xcite . as discussed by @xcite ,
the choice of black - body temperature is a compromise between the possible ranges of values for different classes of thermal x - ray emitters .
sources for which a thermal model best represents the data will likely lie in our galaxy , and so in this case setting @xmath375 would overestimate the total neutral hydrogen absorbing column . similar to spectral model fits , master source spectral model energy fluxes are taken directly from the single observation of the source that has the highest significance . while the spectral model fits described in 3.9
provide detailed information about a source s spectral properties , only about 10% of the source observations included in the csc have sufficient net counts to perform the fitting process . as an aid to characterizing the spectral properties of the remaining catalog sources ,
hardness ratios are computed between the hard , medium , and soft energy bands for all sources observed with the acis detector . the spectral hardness ratio for the pair of energy bands @xmath109 and @xmath189 is defined as @xmath376 where @xmath377 and @xmath378 are the photon fluxes measured in the energy bands @xmath109 and @xmath189 respectively ( @xmath109 is always the higher - energy band of the pair ) , and @xmath379 is the photon flux in the acis broad energy band , @xmath380 . a catalog source may be readily detected in one or more energy bands , but remain undetected or include very few total counts in other bands .
since hardness ratios are cross - band measures , a technique that applies rigorous statistical methods in the poisson regime is required to compute these values and their associated confidence limits robustly .
the hardness ratios included in the csc are computed using a bayesian approach developed by @xcite , which should be consulted for a detailed description of the algorithm . to ensure that the poisson errors are propagated correctly ,
the conversion between counts and photon flux for each energy band is modeled as a linear process , with a scale factor that is determined from the effective area of the telescope / instrument combination computed at the monochromatic effective energy of the band .
this implies that the photon fluxes in equation ( [ eqn : hr ] ) may not match exactly the aperture photometry fluxes derived in 3.7 . specifically , we model the observed total and background counts , @xmath381 and @xmath382 , in the hard , medium , and soft acis energy bands as @xmath383 , \\
b_x & \!\sim\ ! & { \rm poisson}[r e_x\xi_x],\end{aligned}\ ] ] where @xmath109 represents the energy band ( one of _ h _ , _ m _ , or _ s _ ) ; @xmath384 and @xmath385 are the expected source and background counts intensities , respectively ; @xmath386 are the conversion factors that scale counts to photon fluxes ; and @xmath331 is the ratio of the background aperture area to the source aperture area . with these definitions , the spectral hardness ratio for the pair of energy bands @xmath109 and
@xmath189 is determined by computing the expectation value @xmath387 ccl 0 & @xmath388 & definitely not variable + 1 & @xmath389 and @xmath390 and @xmath391 & not considered variable + 2 & @xmath392 and @xmath390 and @xmath391 & probably not variable + 3 & @xmath393 and ( @xmath394 or @xmath395 ) & may be variable + 4 & @xmath396 and ( @xmath394 or @xmath395 ) & likely to be variable + 5 & @xmath392 and ( @xmath394 or @xmath395 ) & considered variable + 6 & @xmath397 and @xmath398 & definitely variable + 7 & @xmath399 & definitely variable + 8 & @xmath400 & definitely variable + 9 & @xmath401 & definitely variable + 10 & @xmath402 & definitely variable + following the lead of @xcite , the joint posterior probability distribution can be written as @xmath403 where we have made use of the fact that the @xmath384 are independent .
marginalizing over nuisance variables yields the posterior distribution for the hardness ratios ( equivalent to equation of * ? ? ?
* ) : @xmath404,\end{aligned}\ ] ] where @xmath405 and @xmath406 .
the spectral hardness ratios that are included in the csc are determined separately for each observation in which a source is detected , and also from the ensemble of all observations of the source .
the former quantities are recorded in the source observations table , while the latter are recorded in the master sources table .
the prior probability distributions used to derive the bayesian posterior probabilities are computed differently in these two cases . for a single observation , non - informative conjugate @xmath289-prior distributions @xcite
are used for the source and background intensities .
these distributions ensure that the posterior probabilities conjugate to the expected poisson distributions of counts with no other prior information . when multiple observations are combined , the ensemble hardness ratios are computed by stepping through all of the observations of a source in order of increasing net broad - band source counts .
the posterior probability distribution computed from each observation is used as the prior probability distribution for the subsequent step .
if the propagated prior probability distribution is not consistent with the observed counts in any step , then a conjugate @xmath289-prior is used instead , and a catalog flag is set to indicate that the source spectrum is variable .
the csc includes estimates of the probability that the flux from a source is temporally variable both within a single observation and between two or more observations in which the source was detected .
these estimates are distinguished not only by the fact that they measure variability on different time scales , but also because their definitions differ fundamentally . within an observation
, we measure the probability that the source flux is not consistent with a constant level during a ( largely ) continuous observation , which is equivalent to estimating the probability that the source is variable , and is a positive statement with respect to variability .
the inter - observation variability estimates measure the probability that the average flux levels during the different observations are consistent with a uniform source intensity .
this provides only a lower limit to the probability of the source being variable , since we have no information about the source s behavior during the gaps between the observations , which are often widely separated . the intent of the various variability measures included in the csc is to provide users a means to easily select _ potentially _ variable sources .
the individual source light curves or event lists should be assessed to reveal the true nature of the source s temporal characteristics .
moderately intense background flares that are not rejected as part of the enhanced background event screening ( see 3.2 ) may cause sources to be incorrectly identified as variable .
this possibility can be evaluated by comparing the structure of the source and background light curves .
the probability that a source is variable is estimated separately in each energy band using the gregory - loredo and kolmogorov - smirnov ( k - s ) algorithms , and kuiper s variation on the latter .
a brief description of each of the three algorithms is provided below .
gregory - loredo probabilities are used to construct intra - observation variability indices that provide a shorthand measure of variability .
all three algorithms directly use the photon event arrival times to compute the variability probabilities , and apply corrections for variations of the geometric areas of the source and background region apertures due to the spacecraft dither - induced motion during the observation .
the latter corrections are necessary since a source region that is moving across the edge of the detector or over a bad detector region might otherwise be erroneously classified as variable .
optimal resolution light curves are generated as by - products of the gregory - loredo test , and their power spectra are evaluated for the presence of the fundamental spacecraft pitch and yaw dither frequencies or associated beat frequencies .
if there is a peak in the power spectrum at one of these frequencies that is at least @xmath151 the rms value , then a catalog warning flag is set for the source observation to indicate that the intra - observation variability properties are unreliable .
the k - s test @xcite is a familiar and well established robust test for comparing two distributions that are a function of a single variable . in the simplest case
we compare the cumulative sum of photon events , as a function of time , against a linearly increasing function that represents a constant flux .
this null - hypothesis function is modified as necessary to account for data gaps and variations in effective area . for an observation with @xmath407 events ,
let @xmath408 be the cumulative sum of detected events as a function of time @xmath335 , and @xmath409 the cumulative function that represents a constant flux .
the k - s statistic @xmath410 is defined as @xmath411 the k - s derived probability that the two distributions @xmath408 and @xmath409 do not belong to the same population , and therefore that the source is variable , is given by @xmath412 where @xmath413 equation ( [ eqn : pvar ] ) is strictly valid only in the asymptotic limit as @xmath414 . in practice @xmath415
is `` large enough , '' especially if conservative significance levels @xmath416 are required ( e.g. , * ? ? ?
@xcite proposed a variation on the k - s test that involves replacing the expression for @xmath410 by the difference between the largest positive and negative deviations , @xmath417 - \inf_t[s_n(t ) - p(t)].\ ] ] folding this expression into equation ( [ eqn : pvar ] ) yields the kuiper derived probability that the source is variable .
while the k - s test is primarily sensitive to differences between the median values of the cumulative distribution functions , the kuiper test statistic is as sensitive to differences in the tails of the distributions . in many cases ,
this makes the kuiper test a more robust variation of the traditional k - s test for evaluating the probability that a source is variable .
the gregory - loredo test @xcite is based on a bayesian approach to detecting variability .
the method works very well on photon event data and is capable of dealing with data gaps .
we have incorporated the capability to include temporal variations in effective area . although the algorithm was developed for detecting periodic signals , it is a perfectly suitable method for detecting random variability by forcing the period to equal the length of the observation .
briefly , the gregory - loredo algorithm bins the @xmath407 observed photon events into a series of histograms containing @xmath418 bins , where @xmath418 runs from 2 to @xmath419 . if the observed distribution of events across the @xmath418 histogram bins is @xmath420 , then the probability that this distribution came about by chance can be determined from the ratio of the multiplicity of the distribution , @xmath421 , to the total number , @xmath422 , of possible distributions . the inverse of this ratio is a measure of the significance of the distribution .
following @xcite , we calculate an odds ratio @xmath423 for @xmath418 bins versus a flat light curve as @xmath424 where we have rewritten the multiplicity of the distribution , @xmath425 , as @xmath426 data gaps are accounted for through the binning factor , @xmath427 , which is ( appendix b of * ? ? ?
* ) @xmath428 where @xmath429 @xmath430 is the amount of good exposure time in bin @xmath431 , and @xmath432 is the total good exposure time for the observation .
the odds are summed over all values of @xmath433 to determine the odds that the source is time - variable .
@xmath434 is chosen for each case in such a way that the odds ratios corresponding to higher values of @xmath418 contribute negligibly to the total .
the probability , @xmath435 , of a particular binning , @xmath418 , is simply @xmath436 summing over bins @xmath437 corresponding to a non - constant source flux yields the gregory - loredo variability probability @xmath438 where @xmath439 , and we have made use of the fact that @xmath440 .
ccccc 0 & & @xmath441 & & @xmath442 + 3 & @xmath443 & @xmath444 & @xmath445 & @xmath446 + 4 & @xmath447 & @xmath446 & @xmath448 & @xmath449 + 5 & @xmath450 & @xmath451 & @xmath452 & @xmath453 + 6 & @xmath454 & @xmath455 & @xmath456 & @xmath457 + 7 & @xmath458 & @xmath459 & @xmath460 & @xmath461 + 8 & @xmath462 & & @xmath463 & + the gregory - loredo algorithm bins the events into a series of light curves of varying resolution , corresponding to the number of bins , @xmath418 , in the range 2 to @xmath434 . using the definitions above , the bins that comprise the normalized light curve , @xmath464 , associated with a specific value of @xmath418 are @xmath465 and the corresponding standard deviations derived from the posterior distribution are @xmath466 as described by @xcite , an optimal resolution , light curve , @xmath467 ,
can be obtained by combining the individual light curves , @xmath464 , weighted by the probabilities , @xmath435 : @xmath468 the optimal resolution light curve computed from the events included in the source region aperture for each source is recorded as a file - based data product ( see table 3 ) that is accessible through the catalog . as well as the light curve , @xmath467 , this data product includes the uncertainty , @xmath310 , and upper and lower confidence intervals , @xmath469 , and @xmath470 , respectively . to allow the users to verify the significance of features that may be present in the light curve
, the file also includes the corresponding quantities derived from the events extracted from the background region aperture , using the same binning .
careful judgement should be applied when assessing the reliability of source variability indicators using the source and background light curves . since the background region aperture may contain up to @xmath36 of the source flux ( see 3.4.1 ) , the background and source light curves may appear similar for very bright sources .
the psf wings of unrelated but nearby strongly variable sources may contaminate both the source and background region apertures of the source being investigated . an observation may have experienced background variations intense enough to be noticeable when compared to the target source s count rate , but not strong enough to have been removed by background screening during observation recalibration . in the first case
the source is truly variable , but this is not necessarily so in the latter two examples . a helpful ,
though not definitive , test is to scale the amplitude of the flux variations in the source and background region apertures by their respective areas ( recorded in the fits keyword aperture ) .
if the variations of the two scaled amplitudes are similar , then there is a good chance that a background problem is responsible .
if the source region scaled amplitude is considerably larger than the background region scaled amplitude and the source is strong , then one is likely to have a truly variable source .
the gregory - loredo test appears to provide a more uniform and reliable measure of variability than either the k - s or kuiper tests , although the gregory - loredo algorithm is more `` conservative '' than the other tests . in cases where the k - s and/or kuiper tests detect variability , but the gregory - loredo test does not , close inspection of the light curve often , but
not always , demonstrated that the level of variability does not exceed the @xmath144 bounds on the light curve . in cases where there are considerable data gaps , gregory - loredo is not always be able to detect variability on time scales comparable to those gaps . to provide the user with a short - hand measure of variability that allows selection of sources on different degrees of confidence , the csc includes a set of integer `` variability index '' values in the range @xmath471 $ ] .
these indices are based on a combination of the gregory - loredo probability , @xmath472 , the logarithm of the odds ratio , @xmath473 , and a secondary criterion that addresses the overall deviation of the light curve from the mean value .
the latter criterion is based on the parameters @xmath474 and @xmath475 , which are the fractions of the light curve that falls within @xmath144 and @xmath476 of the average rate , respectively .
table 6 defines the mapping of the test parameters to variability index values .
inter - observation variability is based on comparison of source region aperture photon fluxes , and their confidence intervals , from multiple observations in which the source is detected .
the catalog provides a probability that the data are not consistent with a constant - flux source , as well as an inter - observation variability index that is similar to the index defined for intra - observation variability .
these measures of variability are assessed for each spectral energy band independently , and consequently no cross - instrument comparison is performed . in the first release of the csc , observations that cover the same region of the sky , but in which the source is not detected , are _ not _ considered when computing inter - observation variability .
these observations should enter into the variability assessment as flux upper limits , since they could conceivably be inconsistent with a constant source flux .
a future release of the catalog will address this limitation .
as mentioned above , the inter - observation variability probability must be interpreted differently from the intra - observation variability probability . whereas the light curve can be used to declare a source to be variable or non - variable within the time range of a single observation
, one can never conclude that a source does not vary between multiple observations . if inter - observation variability is detected ,
then the source is definitely variable ; however the converse is not true .
ll conf_code & 0 : source is not confused + & 1 : multiple source in source region + & 2 : source region overlaps another source region + & 4 : source region overlaps another background region + & 8 : background region overlaps another source region + & 16 : background region overlaps another background region + edge_code & 0 : source does not dither off detector boundary + & 1 : source position dithers off detector boundary + & 2 : source region dithers off detector boundary + & 4 : background dithers off detector boundary + multi_chip_code & 0 : source does not dither between detector chips + & 1 : source position dithers across 2 chips + & 2 : source region dithers across 2 chips + & 4 : background region dithers across 2 chips + & 8 : source position dithers across @xmath477 chips + & 16 : source region dithers across @xmath477 chips + & 32 : background region dithers across @xmath477 chips + var_code & 0 : intra - observation source variability not detected in any band + & 1 : intra - observation variability detected in the acis ultrasoft ( _ u _ ) energy band + & 2 : intra - observation variability detected in the acis soft ( _ s _ ) energy band + & 4 : intra - observation variability detected in the acis medium ( _ m _ ) energy band + & 8 : intra - observation variability detected in the acis hard ( _ h _ ) energy band + & 16 : intra - observation variability detected in the acis broad ( _ b _ ) energy band + & 32 : intra - observation variability detected in the hrc wide ( _ w _ ) energy band + the inter - observation variability probability is simply based on the reduced @xmath345 of the distribution of the source region aperture photon fluxes of the individual observations and their confidence intervals . for a source detected in @xmath100 separate observations ,
we first use the source region aperture photon flux , @xmath478 , and the associated lower and upper @xmath1 confidence limits , @xmath479 and @xmath480 , respectively , to compute an initial estimate of the variance - weighted mean source region aperture photon flux @xmath481 where we take @xmath482 . using this estimate of the mean flux
, we define the `` effective @xmath310 '' for the @xmath184th observation of the source as @xmath483 a refined estimate of the variance - weighted mean flux is then given by @xmath484 and the reduced @xmath345 is @xmath485 the inter - observation variability index is assigned on the basis of the reduced @xmath345 , according to table 7 .
note that the values 1 , 2 , 9 , and 10 are not used .
each entry in both the master sources table and the source observations table includes several source - specific flags and codes that identify specific circumstances that may be of relevance to the catalog user .
some flags and codes are used to encode source properties that are commonly searched for by users , as an aid to simplify catalog queries . however , in most cases flags and codes are intended to warn the user of conditions that may degrade the quality of measured source properties , or that may limit the usefulness of the source detection for some investigations . the codes and flags included in the source observations table are defined in table 1 .
flags are boolean quantities that describe `` yes / no '' or `` true / false '' properties , whereas codes are multi - bit data values that encode several levels of information .
translations of the bit - encodings can be found in table 8 .
the extent and variability codes require additional explanation .
the former encodes a conservative estimate of whether the intrinsic extent of a source , @xmath236 [ equation ( [ eqn : arss ] ) ] , is inconsistent with the extent of the local psf in each energy band .
specifically , a source is considered extended in an energy band if @xmath486 in that energy band , where @xmath487 is the uncertainty in @xmath236 , given by equation ( [ eqn : errarss ] ) .
the variability code bit corresponding to a specific energy band is set if the intra - observation variability index ( table 6 ) @xmath488 .
a zero code therefore implies that the source is either definitely not variable , not considered variable , or probably not variable , depending on the value of the variability index .
similarly , a non - zero code implies that the source either may be variable , is likely to be variable , is considered variable , or is definitely variable .
the remaining codes and flags all warn of conditions that may affect derived source properties to some extent .
the streak source flag , if set , indicates that the source detection is located on an acis readout streak .
if the readout streak is associated with a bright source , then there is a significant probability that the source properties may be compromised .
this is particularly true of aperture photometry values .
if the streak is especially intense , then the source detection may not be real . if the saturated source flag is set , then the source is definitely real , but is so bright that photon pile - up has eroded the core of the source image so that a single source has a `` cratered '' appearance ( fig .
all source properties are compromised .
the master sources table includes only flags , and these are defined in table 2 . in most cases , master sources table flags summarize the corresponding source observations table flags and codes for all of the source detections that have `` unique '' linkages to the master source .
the confusion flag is an exception to this , in that it is set for a master source if the confusion codes for any of the associated source detections indicate that multiple sources are present in the source region or that the source region overlaps another source region , _ or _ if there are any source detections that have `` ambiguous '' linkages to the current master source .
the master extent and variability flags are set if the corresponding codes for _ any _ uniquely matched source detections indicate that the source is extended or variable ( as appropriate ) in any energy band .
the remaining master source flags are set only if the corresponding source observations table flags are set for _ all _ uniquely associated source detections , indicating that the corresponding warning criteria are violated in all observations of the source .
ccc acis-012367 & 9 & 0.0 + acis-235678 & 10 & 0.02 + acis-012367 & 29 & 0.0 + acis-235678 & 30 & 0.12 + acis-235678 & 51 & 0.21 + acis-012367 & 68 & 0.22 + acis-235678 & 118 & 1.2 + acis-012367 & 125 & 1.28 + the scientific integrity of the csc is guaranteed through a set of quality assurance steps that are performed as part of the catalog construction process @xcite .
many of these analyses are executed automatically at the completion of each stage of catalog pipeline processing , so that any issues can be identified and corrected before they can affect downstream processing .
these mechanisms detect pipeline processing errors , and identify potential data quality issues by comparing key diagnostic output products with predefined standards .
each standard that is violated will either trigger a human review to determine how to proceed , or will initiate one or more automated actions .
the latter typically result in termination of the processing thread for a subset of the input data .
the vast majority of violations that occur because of data quality issues address the reality of detected sources , and are typically resolved without human intervention . following the source detection step , detected source regions that are either significantly smaller than the dimensions of the local psf or significantly larger than the maximum expected source size , or which exceed a maximum ellipticity threshold ,
are deemed to be artifacts , and the processing thread for the source region is terminated immediately to avoid evaluating source properties unnecessarily .
sources that have too few counts , or that have a detection significance that is too low to pass the catalog snr threshold are similarly discarded .
the cores of sources observed with acis that are sufficiently bright can be eroded by photon pile - up .
the source detection algorithm incorrectly detects bright spots on the ring surrounding the dark center of the image as distinct sources .
saturated sources are identified using a sliding matched filter algorithm .
the source detections are manually adjusted so that a single source centered on the crater is included in the source observations table for the source , and the source properties are flagged as having been manually modified , so that the user can exclude such sources if they so wish .
the tabulated source position errors are unreliable for sources whose regions have been manually modified .
a more detailed comparison of the source dimensions with the local psf is performed after the source properties are computed , to identify sources that are statistically smaller than the psf in all energy bands .
the fraction of the local psf that is included within the modified source region aperture ( as defined in
3.4.1 ) must be sufficiently large that the source location and aperture photometry can be computed from the fraction of the aperture that is not contaminated by overlapping sources .
finally , the snr of the source is evaluated and compared with the minimum required for inclusion in the catalog .
human review is primarily required to address the occurrence of unexpected pipeline warnings or errors .
although an automated process performs the laborious task of scanning the log files associated with each processing pipeline to identify problems , the wide diversity of possible error conditions require human intelligence to assess the reason for the failure and determine how to proceed .
the typical response is to terminate the current processing thread , perform any needed repairs , and initiate reprocessing of the thread .
other conditions that trigger a human review are precautionary in nature , and include cases where the local spatial density of detected sources is too high , or the total number of sources detected in the field of view exceeds a predefined threshold .
although these conditions most likely arise because of field crowding , they could indicate an error in the source detection process .
errors that generate a large number of sources would require substantial cleanup if processing was allowed to continue incorrectly .
a detailed characterization of the statistical properties of the csc is beyond the scope of this paper , but is the subject of a comprehensive discussion by f. a. primini et al .
( 2010 , in preparation ) . here
, we merely present a summary of the principal statistical properties of the catalog for reference
. statistical characterization of catalog source properties is accomplished primarily by using simulated and empty field ( blank - sky ) observations , together with datasets consisting of empty fields that have simulated sources with known properties added .
these datasets are processed by the catalog pipelines in the same manner as real observations . to estimate false source rates , a series of blank - sky simulations with exposure times of @xmath15 , @xmath489 , @xmath490 , and @xmath491
were constructed for typical acis imaging ccd configurations . for each simulation , a template background event list for each active ccd
was used to define the overall spatial variation of the background , and the total number of background events was determined from the nominal field background rates @xcite and the simulated exposure time .
for all ccds except chip s4 ( acis-8 ) the template background event lists recorded in the instrumental calibration database were used .
chip s4 is significantly affected by a variable pattern of linear streaks that appear to be caused by a flaw in the serial readout which randomly deposits significant amounts of charge along pixel rows as they are read out @xcite .
because of this issue , no adequate template is available for chip s4 , and so one was constructed by combining several csc event lists that do not include bright sources on that ccd .
each simulated blank - sky event list was then processed through the csc pipeline source detection steps .
the false source rates derived from these simulations are reported in table 9 . from these data
we derive a simple linear relation for the number of false sources per field as a function of livetime , namely @xmath492 where @xmath493 is the false source rate , and @xmath494 is the exposure livetime in units of ks . using this relation , we estimate that @xmath495 sources ( @xmath496 ) included in the catalog are spurious . as can be seen from the table , the false source rate is appreciable only for exposures longer than @xmath497 .
there is some evidence for a clustering of false source detections near chip edges and at the boundaries between the back- and front - illuminated ccds .
this should not be surprising since the low spatial frequency background is poorly constrained or changing rapidly in these locations . to investigate these effects further
, the false source rates near the chip edges and interfaces were examined separately for the longest simulated exposures .
figure 23 demonstrates that false source rates are enhanced in these regions for the @xmath498 simulation .
source detection efficiency is characterized using point source simulations .
a spatially random distribution of point sources is added to the blank sky simulations described above using the marx simulator @xcite to generate the incident x - ray photons .
separate simulations were generated for non - thermal sources with a power - law spectral distribution @xmath499 , and for thermal black - body sources with temperature @xmath500 , spectra .
a neutral hydrogen absorbing column @xmath501 was assumed for all sources .
source fluxes were drawn from a power - law @xmath502 distribution with index 1.5 .
the overall normalization was adjusted to yield a few hundred detectable sources per simulation , a compromise aimed at reducing source confusion while limiting the total number of simulations required to obtain good statistics .
the effects of photon pile - up @xcite and observation - specific bad pixels were included by post - processing each simulation with marxpileup and acis_process_events , respectively .
the source events from the marx simulations were then merged with the appropriate simulated blank - sky event lists , keeping only marx - simulated source events that fell on active ccds for the observation . as with the blank - sky simulations ,
simulated event lists were then processed through the csc pipeline source detection and source properties extraction steps , and the resulting sources that would have been included in the catalog were tabulated . finally ,
these sources were cross - referenced with the input source lists to allow a source - by - source comparison of input and derived properties . source detection efficiency is determined by comparing the measured @xmath502 and input @xmath502 distributions .
the ratio of these two distributions represents the fraction of input sources of a given incident flux that are actually detected .
results of the comparison for the acis broad energy band detections from the shortest and longest acis-012367 power - law spectral distribution simulation sets are shown in figure 24 .
the standard csc processing pipeline further combines acis source detections from the broad , soft , medium , and hard energy bands to construct the final detected source list .
this step was not performed as part of these simulations . however , since the simulated source spectra are homogenous and well detected in the broad energy band , the difference is not significant in this case . as mentioned in 3.5.1 ,
the absolute astrometric accuracy of release 1 of the csc was evaluated _
post facto _ by cross - matching catalog sources with their counterparts from the sdss dr7 @xcite . like the csc
, the sdss dr7 is referenced to the international celestial reference system ( icrs ; * ? ? ?
* ) , and has statistical positional uncertainties @xmath503 ( mas ) rms per coordinate for bright stars , with systematic errors @xmath504 @xcite . only csc - sdss source pairs with more than 90% match probability , evaluated according to the bayesian probabilistic formalism described by @xcite , were evaluated , resulting in 6,310 source pairs associated with 9,476 sources detected in individual observations .
full details of the analysis and results are presented by @xcite and f. a. primini et al .
( 2010 , in preparation ) . here
we summarize the main result . for each matching csc - sdss source pair ,
the separation , @xmath505 , and the total @xmath1 positon error are computed , summing in quadrature the csc and sdss errors ( and remembering that csc position errors are reported as 95% uncertainties ) .
we then examine the value of reduced @xmath506 for bins in @xmath507 covering the range @xmath508@xmath116 .
the value of the reduced @xmath345 is reasonably close to 1 , except for @xmath509 ( indicating that the errors are underestimated in that range ) . adding a systematic astrometric error component of @xmath177 to @xmath507 yields reduced @xmath345 near 1 for all values of @xmath510 .
we therefore adopt that value as the systematic astrometric error present in release 1 of the csc .
the distribution of the normalized separations for the csc - sdss source match pairs is shown in the left panel of figure 25 , together with the theoretical rayleigh distribution for the same number of sources .
the overall shape of the curve agrees with the rayleigh distribution , although there is a slight deficit at high values of normalized separation , suggesting that the overall error may be overestimated for sources at large off - axis angles . in the right panel of figure 25 , we present the average csc - sdss separation as a function of off - axis angle , with the systematic astrometric error included .
the average csc @xmath1 positional error ranges from @xmath219 on - axis to @xmath511 at @xmath512 off - axis . to assess the accuracy of csc source fluxes , the measured source region aperture photon fluxes are compared with the input photon fluxes of the simulated point sources .
figure 26 presents the comparison of acis broad energy band photon fluxes for simulated sources with a power - law spectrum .
inspection of the figure reveals good agreement for sources with off - axis angles within @xmath118 of the aimpoint . for sources beyond @xmath118
, photon fluxes appear to be systematically overestimated by a factor of @xmath513 for sources fainter than @xmath514 .
the systematic error in the faint flux bins is more prominent in the acis soft energy band , in which the measured fluxes appear under - estimated for all simulated input flux levels .
further investigation of this effect will be reported by f. a. primini et al .
( 2010 , in preparation ) .
preliminary analysis suggests that the effect results from the use of a monochromatic exposure map ( computed at the effective energy of the band ) when determining source fluxes .
models based on this assumption reproduce the general features of the apparent systematic errors , and for the assumed model power - law spectrum the error is @xmath36 in the broad , medium , and hard energy bands , @xmath515@xmath516 in the soft energy band , and @xmath517 in the ultra - soft energy band .
the _ chandra _ source catalog is a general purpose virtual x - ray astrophysics facility that provides access to a carefully crafted set of scientifically useful quantities for individual x - ray sources observed by the _
chandra _ x - ray observatory .
the first release of the catalog was published to the astronomical community in march 2009 , and includes source properties for 94,676 point and compact x - ray sources detected in a subset of public acis imaging observations from roughly the first eight years of the _ chandra _ mission .
this release of the catalog includes sources with observed spatial extents are @xmath0 , and whose flux estimates are at least 3 times their estimated @xmath1 uncertainties .
observations that include substantially extended sources are not included in the first release of the catalog . for each x - ray source ,
the catalog tabulates about 60 distinct measured and derived source properties , generally with associated lower and upper confidence limits , in several energy bands .
these properties are generally derived from all of the observations in which a source is detected . however , in the first catalog release , multiple observations are not _ combined _ prior to source detection , so the depth of the catalog is limited by the duration of the longest single exposure of a field .
the catalog further tabulates roughly 120 observation - specific properties for each observation of a source , again with associated lower and upper confidence limits , and in several energy bands .
tabulated source properties include source position , spatial extent , multi - band aperture fluxes computed in several different ways , x - ray hardness ratios and spectral model fits , and intra- and inter - observation variability measures . in addition to these `` traditional '' catalog elements , for each source detection the catalog includes an extensive set of fits format file - based data products that can be manipulated interactively by the user , including source images , event lists , light curves , and spectra from each observation in which a source is detected . looking towards the future ,
release 1.1 of the catalog , scheduled for spring 2010 , will include data from public hrc - i imaging observations and newly public acis imaging observations , but will otherwise retain the same limitations as release 1 . in release 2 , we plan to co - add multiple observations of the same field that use the same or similar instrument configurations , and that have similar spacecraft pointings ( within @xmath89 ) prior to source detection , to achieve fainter limiting sensitivities in many fields .
we anticipate that new algorithms will allow this release to have a significantly fainter source detection threshold than release 1
. this release should also provide limited improvements in the area of extended source handling ( for example allowing for the inclusion of exposures containing moderately extended emission from galaxy cores up to @xmath518 spatial scale ) , as well as numerous algorithm enhancements that will refine field and source property calculations .
the authors would like to thank the _ chandra _ source catalog project review visiting panel , who endorsed the catalog and proposed several key recommendations that have guided the development of the catalog .
we would like to acknowledge the support and guidance of the _ chandra _ x - ray center director , harvey tananbaum , and manager , roger brissenden .
former _ chandra _ source catalog project team members who have contributed significantly to the definition and development of the catalog include martin elvis , stphane paltani , adam dobrzycki , johnathan slavin , dan harris , peter freeman , and michael wise .
we would like to thank taeyoung park for his support enhancing and testing the behr algorithm for use in the catalog processing pipelines .
lisa paton and the cxc systems group provided extensive installation and operational support for the catalog production beowulf cluster .
the authors would also like to thank the anonymous referee , who performed a very careful and comprehensive review of the first version of the manuscript , and whose suggestions materially improved the content and quality of the paper . the development and operational construction of the catalog made extensive use of the ciao , chips , and sherpa software packages developed by the _
chandra _ x - ray center , and the saoimage ds9 imager developed by the smithsonian astrophysical observatory .
support for development of the _ chandra _ source catalog is provided by the national aeronautics and space administration through the _ chandra _ x - ray center , which is operated by the smithsonian astrophysical observatory for and on behalf of the national aeronautics and space administration under contract nas 8 - 03060 .
the procedure , referenced in 3.4.2 , for matching source detections from multiple observations that overlap the same region of the sky , is described here .
these steps must be executed for each observation , but the outcome does not depend on the order in which the observations are processed .
the algorithm defines the _ overlap ellipse _ of a source detection to be the psf 90% ecf aperture in the energy band that has the highest number of aperture source counts .
the following two assumptions are made : ( a ) if a source detection in one observation is resolved into multiple source detections in a second observation , then the overlap ellipse corresponding to the former detection will overlap all of the overlap ellipses corresponding to the latter detections , and ( b ) multiple source detections in a single observation correspond to distinct sources on the sky , even if the overlap ellipses intersect spatially . for the set @xmath519 of source detections identified in the current observation ,
the following 9 steps are performed .
( 1 ) identify the sets @xmath520 of _ candidate matching source detections _ in observations @xmath521 that overlap the current observation .
candidate matching source detections are those source detections that have radial separations on the sky from any member of the set @xmath519 that are smaller than some predefined radius , @xmath331 . for convenience , we designate the union @xmath522 as the set @xmath523 .
( 2 ) compute the overlap ellipses , defined above , for each member of the sets @xmath519 and @xmath523 .
two source detections @xmath110 and @xmath111 in observations @xmath524 and @xmath525 , respectively , are deemed to _ overlap _ _ if and only if _ @xmath526/a[a ] > 0.15 & { \it or\/ } & a[a \cap b]/a[b ] > 0.15 , \end{aligned}\ ] ] where @xmath527 $ ] is the area of the overlap ellipse of source detection @xmath110 , and @xmath528 $ ] is the area of the spatial intersection of the overlap ellipses of @xmath110 and @xmath111 .
this has the effect of dismissing very small overlaps .
in addition , _ if _ a source detection @xmath110 in observation @xmath524 overlaps multiple source detections @xmath529 ( @xmath530 ) in observation @xmath525 , _ and _ @xmath531/a[a ] > 0.67 & { \it and\/ } & a[a \cap b_j]/a[a \cap b_i ] < 0.33 \end{aligned}\ ] ] for all @xmath532 , _ then _ only the overlap between source detections @xmath110 and @xmath238 is recognized .
the remaining source detections @xmath533 included in observation @xmath525 are deemed _ not _ to overlap source detection @xmath110 .
this has the effect of recognizing only a single dominant overlap and ignoring additional smaller overlaps from the same observation .
( 3 ) compute the subset @xmath534 of source detections that do not overlap any member of the set @xmath523 of candidate matching source detections .
_ @xmath535 comprises the set of source detections that must be added to the master sources table as newly identified master sources .
_ ( 4 ) compute the set @xmath536 of members of @xmath537 that comprise _ partial unambiguous matches_. a source detection @xmath110 included in observation @xmath524 is a partial unambiguous match to a source detection @xmath111 included in a different observation @xmath525 _ if and only if _ ( a ) @xmath110 overlaps @xmath111 , ( b ) @xmath110 does not overlap any other source detection included in observation @xmath525 , and ( c ) @xmath111 does not overlap any other source detection included in observation @xmath524 .
sources included in @xmath536 are uniquely matched between _
pairs _ of overlapping observations , but are not necessarily uniquely matched between _ all _ overlapping observations .
( 5 ) compute the subset @xmath538 of _ unambiguous matches_. an unambiguous match between source detections @xmath539 included in observations @xmath540 occurs when all pairs @xmath541 , [ a , c ] , [ a , \ldots ] , [ b , c ] , [ b , \ldots ] , \ldots)$ ] of source detections individually form partial unambiguous matches . _ @xmath542 comprises the set of source detections that are uniquely matched to existing master sources in the master sources table .
_ figure 15 , _ left _ is an example of an unambiguous match .
( 6 ) compute the set @xmath543 of members of @xmath537 that comprise _ confused matches_. a confused match results when a source detection @xmath110 included in observation @xmath524 overlaps multiple source detections @xmath544 that either ( a)(i ) are included in a single observation @xmath525 _ and
_ ( ii ) overlap no other source detection in observation @xmath524 than @xmath110 , _ or _ ( b ) consist of partial unambiguous matches such that there is at least one observation that is common amongst the partial unambiguous matches for all of the sources @xmath544 .
_ @xmath543 comprises the set of confused source detections that must be flagged as confused , and linked ambiguously to the corresponding master sources in the master sources table .
_ an example of case ( a ) above is shown in figure 15 , _
center_. in the case ( b ) above , note that source detection @xmath110 is not required to overlap all of the individual source detections that comprise each of the partial unambiguous matches ( e.g. , fig .
( 7 ) once the set @xmath543 of confused matches is determined , steps ( 2)(5 ) should be re - applied to the set @xmath545 to identify additional members of the sets @xmath535 and @xmath542 that were previously missed because they were overlapped by one or more confused source detections .
( 8) after all members of the sets @xmath535 and @xmath542 have been identified , re - examine members of the set @xmath543 to verify that the overlaps of source detections that caused each member to become assigned to set @xmath543 are overlaps with members of @xmath535 and @xmath542 .
_ if _ any source detection which is not a member of @xmath535 or @xmath542 overlaps a member of @xmath543 , _ and _ that overlap was relied on to assign the member to @xmath543 , _ then _ remove the member from set @xmath543 .
the removed member will revert to an uncategorized source detection .
( 9 ) at this point , the set @xmath546 consists of source detections that can not be merged or linked to master sources using the above rules .
these source detections typically ( but not exclusively ) overlap at least two other sources that were observed in different observations and that do not overlap each other .
we designate members of this set @xmath547 as _ human - review matches_. manual review is required to disambiguate the source matches .
figure 15 , _ right _ is an example of a human - review match .
as described in 3.5.2 , a multivariate optimal weighting formalism is used to improve the estimates of the position and positional uncertainty of each x - ray source by combining the statistically independent source detections included in the set of individual observations .
the source position error uncertainties are expressed in the form of error ellipses centered upon the estimated source positions .
details of the derivations can be found in @xcite .
the improved estimates of the source position , @xmath93 , and associated covariance matrix , @xmath182 , are @xmath548^{-1},\end{aligned}\ ] ] where @xmath179 represents the @xmath184th estimate of the mean of the two - dimensional source position , and @xmath181 denotes the @xmath180 covariance matrix , equation ( [ eqn : appcov ] ) below , associated with this estimate .
before the covariance matrix @xmath182 can be computed , the individual error ellipses must be mapped from the celestial sphere onto a common tangent plane .
the @xmath184th estimate of the source position is specified as a confidence - ellipse centered upon the celestial coordinates @xmath549 , with the major axis of the ellipse making an angle @xmath550 ( @xmath551 ) with respect to the local line of declination at the center of the ellipse .
the celestial coordinates @xmath552 correspond to a unit vector @xmath553 on the celestial sphere , where @xmath554 are orthonormal basis vectors oriented such that @xmath555 points to the origin of right ascension on the celestial equator , @xmath556 points to the north celestial pole , and @xmath557 completes the right - hand cartesian system .
the common tangent plane is constructed on the celestial sphere at the position @xmath558 , which is taken to be the arithmetic mean of the ellipse centers @xmath559 : @xmath560 the tangent plane coordinates ( @xmath561 , @xmath562 ) corresponding to @xmath549 are @xmath563 where @xmath564 and @xmath565 are orthonormal basis vectors parallel to the local lines of right ascension and declination at @xmath558 , _
i.e. _ , @xmath566 where @xmath567 are the celestial coordinates that correspond to @xmath558
. similarly , the unit vectors on the celestial sphere corresponding to the end - point positions of the semi - minor and semi - major axes of the ellipse are given by @xmath568 where @xmath569 and @xmath570 are the arc - lengths of the semi - minor and semi - major axes , respectively , and @xmath571 and @xmath572 are unit vectors that point along the directions of increasing right ascension and declination , respectively , at the position @xmath559 .
the lengths of the semi - minor and semi - major axes on the tangent plane are given by @xmath573 respectively , where we have denoted the tangent plane coordinates of @xmath574 and @xmath575 as ( @xmath576 , @xmath577 ) and ( @xmath578 , @xmath579 ) , respectively . the angle that the semi - major axis makes with respect to the local line of declination is @xmath580 the three parameters that specify the geometry of each projected error ellipse are the lengths of the semi - major and semi - minor axes , and the position angle @xmath581 that the major axis of the ellipse makes with respect to the tangent plane @xmath189 axis .
the semi - major and semi - minor axis lengths correspond to the @xmath1 confidence intervals along these axes . in a basis
whose origin is at the center of the ellipse , and whose @xmath189 axis is along the major axis of the ellipse , the covariance matrix is @xmath582 where @xmath583 and @xmath584 are the @xmath585 confidence values along the minor axis and major axis of the ellipse , respectively ( @xmath586 ) .
the form of the covariance matrix in the unrotated system is @xmath587 where @xmath188 is the angle that the major axis of the ellipse makes with respect to the tangent plane @xmath189 axis . at this point ,
equation ( [ eqn : appcov ] ) can be used to compute the covariance matrices from the lengths of the semi - minor and semi - major axes of the source position error ellipses in the tangent plane , equations ( [ eqn : apptpl ] ) .
the error ellipses for the individual source observations are then combined using equation ( [ eqn : appmow ] ) to compute the optimally weighted source position and position error ellipses _ on the tangent plane _ for the combined set of observations .
the mapping of the optimally weighted error ellipse from the tangent plane to the celestial sphere can be performed using the inverse relations of equations ( [ eqn : apptpc ] ) , ( [ eqn : apptpam ] ) , ( [ eqn : apptpl ] ) , and ( [ eqn : apptpt ] ) .
abazajian , k. n. , et al .
2009 , , 182 , 543 aldcroft , t. l. , karovska , m. , cresitello - dittmar , m. l. , cameron , r. a. , & markevitch , m. l. 2000 , , 4012 , 650 alexander , d. m. , et al .
2003 , , 126 , 539 anders , e. , & grevesse , n. 1989 , geochimica et cosmochimica acta , 53 , 197 arias , e. f. , charlot , p. , feissel , m. , & lestrade , j .- f .
1995 , , 303 , 604 arnaud , k. a. , george , i. m. , & tennant , a. f. 2009 , the ogip spectral file format , ogip memo ogip/92 - 007 , http://heasarc.gsfc.nasa.gov/docs/heasarc/ofwg/docs/ summary / ogip_92_007_summary.html avni , y. 1976 , , 210 , 642 balucinska - church , m. , & mccammon , d. 1992 , , 400 , 699 bautz , m. w. , et al .
1998 , , 3444 , 210 bertin , e. , & arnouts , s. 1996 , , 117 , 393 brandt , w. n. , et al .
2001 , , 122 , 2810 brassington , n. j. , et al .
2008 , , 179 , 142 broos , p. s. , et al .
2007 , , 169 , 353 budavri , t. , & szalay , a. s. 2008 , , 679 , 301 calderwood , t. , dobrzycki , a. , jessop , h. , & harris , d. e. 2001 , in asp conf .
238 , astronomical data analysis software and systems x , ed .
f. r. harnden , jr . , f. a. primini , & h. e. payne ( san francisco , ca : asp ) , 443 x - ray center .
2009 , the _ chandra _ proposers observatory guide , version 11.0 ( cambridge , ma : cxc ) , http://cxc.cfa.harvard.edu/proposer/pog/ conroy , m. a. , garcia , m. r. , mandel , e. g. , roll , j. , & worrall , d. m. 1992 , in asp conf . ser .
325 , astronomical data analysis software and systems i , ed .
d. m. worrall , c. biemesderfer , & j. barnes ( san francisco , ca : asp ) , 17 damiani , f. , maggio , a. , micela , g. , & sciortino , s. 1997 , , 483 , 350 davis , j. e. 2001a , , 548 , 1010 davis , j. e. 2001b , , 562 , 575 davis , j. e. 2007a , pile - up fractions and count rates , http://cxc.cfa.harvard.edu/csc/memos/files/davis_pileup.pdf davis , j. e. 2007b , combining error ellipses , http://cxc.cfa.harvard.edu/csc/memos/files/davis_ellipse.pdf devereux , d. , abel , d. j. , power , r. a. , & lamb , p. r. 2005 , in asp conf . ser . 347 , astronomical data analysis software and systems xiv , ed .
p. shopbell , m. britton , & r. ebert ( san francisco , ca : asp ) , 346 dickey , j. m. , & lockman , f. j. 1990 , , 28 , 215 doe , s. , et al .
2007 , in asp conf .
376 , astronomical data analysis software and systems xvi , ed .
r. a. shaw , f. hill , & d. j. bell ( san francisco , ca : asp ) , 543 ebeling , h. , & wiedenmann , g. 1993 , , 47 , 704 elvis , m. , plummer , d. , schachter , j. , & fabbiano , g. 1992 , , 80 , 257 elvis , m. , et al . 2009 , , 184 , 158 evans , i. n. , evans , j. d. , glotfelty , k. j. , hall , d. m. , plummer , d. a. , & zografou , p. 2008 , in asp conf . ser .
394 , astronomical data analysis software and systems xvii , ed . r. w. argyle , p. s. bunclark , & j. r. lewis ( san francisco , ca : asp ) , 414 evans , i. n. , et al . 2006 , , 6270 , 62701u evans , j. d. , et al .
2006 , , 6270 , 62700n forman , w. , jones , c. , cominsky , l. , julien , p. , murray , s. , peters , g. , tananbaum , h. , & giacconi , r. 1978 , , 38 , 357 freeman , p. , doe , s. , & siemiginowska , a. 2001 , , 4477 , 76 freeman , p. e. , kashyap , v. , rosner , r. , & lamb , d. q. 2002 , , 138 , 185 fruscione , a. , et al .
2006 , , 6270 , 62701v garmire , g. p. , bautz , m. w. , ford , p. g. , nousek , j. a. , & ricker , g. r. , jr .
2003 , , 4851 , 28 george , i. m. , arnaud , k. a. , pence , b. , & corcoran , m. f. 2007 , the calibration requirements for spectral analysis , ogip calibration memo cal / gen/92 - 002 , http://heasarc.gsfc.nasa.gov/docs/heasarc/caldb/docs/memos/
cal_gen_92_002/ george , i. m. , & corcoran , m. f. 2005 , the heasarc calibration database , ogip calibration memo cal / gen/91 - 001 , http://heasarc.gsfc.nasa.gov/docs/heasarc/caldb/docs/memos/
cal_gen_91_001/ giacconi , r. , kellogg , e. , gorenstein , p. , gursky , h. , & tananbaum , h. 1971 , , 165 , l27 giacconi , r. , murray , s. , gursky , h. , kellogg , e. , schreier , e. , & tananbaum , h. 1972 , , 178 , 281 giacconi , r. , et al .
2002 , , 139 , 369 gioia , i. m. , et al . 1990 , , 72 , 567 graessle , d. e. , evans , i. n. , glotfelty , k. j. , he , x. ( helen ) , evans , j. d. , rots , a. h. , fabbiano , g. , & brissenden , r. j. , 6270 , 62701x gray , j. , szalay , a. , budavri , t. , lupton , r. , nieto - santisteban , m. , & thakar , a. 2006 , cross - matching multiple spatial observations and dealing with missing data , microsoft research technical report msr tr 2006 - 175 , arxiv.org/cs/0701172 gregory , p. c. , & loredo , t. j. 1992 , , 398 , 146 grimm , h .- j . , mcdowell , j. , fabbiano , g. , & elvis , m. 2009 , , 690 , 142 hain , r. , et al . 2004 , in asp conf .
314 , astronomical data analysis software and systems xiii , ed .
f. ochsenbein , m. allen , & d. egret ( san francisco , ca : asp ) , 420 harnden , f. r. , jr . ,
fabricant , d. g. , harris , d. e. , & schwartz , j. 1984 , sao report no .
393 harris , d. e. 1990 , the einstein observatory catalog of ipc x - ray sources ( cambridge , ma : smithsonian astrophys .
houck , j. c. 2000 , removing streaks from acis - s4 , cxc memorandum , http://cxc.cfa.harvard.edu/ciao/download/doc/destreak.ps ishibashi , w. , & courvoisier , t. j .- l .
2010 , , 512 , 58 jerius , d. h. 2005 , effect of contamination of the hrma on the optics @xmath588 , http://cxc.cfa.harvard.edu/cdo/cuc/cuc_file05/jan05/ jeriuscuc-20050125.pdf jerius , d. h. , freeman , m. , gaetz , t. , hughes , j. p. , & podgorski , w. 1995 , in asp conf .
77 , astronomical data analysis software and systems iv , ed .
r. a. shaw , h. e. payne , & j. j. e. hayes ( san francisco , ca : asp ) , 357 jerius , d. h. , et al . 2004 , , 5165 , 402 kenter , a. , et al .
2005 , , 161 , 9 kim , m. , et al .
2007 , , 169 , 401 kuiper , n. h. 1962 , proc .
koninklijke nederlandse akademie van wetenschappen , series a , 63 , 38 laird , e. s. , et al .
2009 , , 180 , 102 lehmer , b. d. , et al .
2005 , , 161 , 21 liszt , h. s. 1983 , , 275 , 163 luo , b. , et al .
2008 , , 179 , 19 maccacaro , t. , et al . 1982 , , 253 , 504 markevitch , m. 2001a , acis background , http://cxc.cfa.harvard.edu/contrib/maxim/bg/index.html markevitch , m. 2001b , plate scale and relative chip ( plate ) positions , http://cxc.cfa.harvard.edu/cal/hrma/focal_length/ geom_public.html marshall , h. l. , tennant , a. , grant , c. e. , hitchcock , a. p. , odell , s. l. , & plucinsky , p. p. 2004
, , 5165 , 497 massey , f. j. , jr .
1951 , j. amer .
statist . ass .
, 46 , 68 mccollough , m. l. 2007 , _ chandra _ source catalog energy bands , http://cxc.cfa.harvard.edu/csc/memos/files/ mccollough_csc_ebands.pdf mccollough , m. l. , & rots , a. 2005 , in asp conf . ser . 347 , astronomical data analysis software and systems xiv , ed .
p. shopbell , m. britton , & r. ebert ( san francisco , ca : asp ) , 478 mccollough , m. l. , & rots , a. 2008 , in asp conf .
394 , astronomical data analysis software and systems xvii , ed .
r. w. argyle , p. s. bunclark , & j. r. lewis ( san francisco , ca : asp ) , 559 mccollough , m. l. 2010 , default spectral fit parameters for the chandra source catalog , http://cxc.cfa.harvard.edu/csc/memos/files/ mccollough_def_spec_parm_csc.pdf mcdowell , j. 2001 , coordinate systems for analysis of on - orbit _ chandra _ data paper i : imaging , http://cxc.cfa.harvard.edu/contrib/jcm/ncoords.ps mewe , r. , kaastra , j. s. , & liedahl , d. a. 1995 , legacy , 6 , 16 mukai , k. 2009 , portable , interactive multi - mission simulator , http://heasarc.gsfc.nasa.gov/docs/software/tools/pimms.html muno , m. p. , et al . 2009 , , 181 , 110 murray , s. s. , et al .
2000 , , 4012 , 467 orechovesky , j. r. , jr .
1996 , single source error ellipse combination , master s thesis ( monterey , ca : naval postgrad .
school ) ortiz , i. , et al .
2008 , ivoa astronomical data query language ( int .
alliance ) , http://www.ivoa.net/documents/latest/adql.html ochsenbein , f. , & williams , r. 2009 , votable format definition ( int .
alliance ) , http://www.ivoa.net/documents/latest/vot.html park , t. , kashyap , v. l. , siemiginowska , a. , van dyk , d. a. , zezas , a. , heinke , c. , & wargelin , b. j. 2006 , , 652 , 610 press , w. h. , flannery , b. p. , teukolsky , s. a. , & vetterlin , w. t. 1986 , numerical recipes the art of scientific computing ( 1st ed . ; cambridge : cambridge univ . press ) raymond , j. c. , & smith , b. w. 1977 , , 35 , 419 romano , p. , campana , s. , mignano , r. p. , moretti , a. , mottino , m. , panzera , m. r. & tagliaferri , g. 2008 , , 488 , 1221 rots , a. , & mcdowell , j. 2008 , fits region binary table design asc - fits - region-1.2 , http://fits.gsfc.nasa.gov/registry/region.html rots , a. 2009 , determining the astrometric error in csc source positions , http://cxc.cfa.harvard.edu/csc/memos/files/ rots_cscastrometricerror.pdf saxton , r. d. 2003 , a statistical evaluation of the epix flux calibration , xmm - soc - cal - tn-0023 , http://xmm.vilspa.esa.es/docs/documents/ cal-tn-0023-2-1.ps.gz schmidt , d. , & green , p. 2003 , in asp conf .
ser . 295 , astronomical data analysis software and systems xii , ed .
h. e. payne , r. i. jedrzejewski , & r. n. hook ( san francisco , ca : asp ) , 81 smith , r. k. , brickhouse , n. s. , liedahl , d. a. , & raymond , j. c. 2001 , , 556 , l91 tody , d. , & plante , r. 2009 , simple image access specification ( int .
alliance ) , http://www.ivoa.net/documents/latest/sia.html ueda , y. , ishisaki , y. , takahashi , t. , makishima , k. , & ohashi , t. 2005 , , 161 , 185 valtchanov , i. , pierre , m. , & gastaud , r. 2001 , , 370 , 689 van dyk , d. a. , connors , a. , kashyap , v. l. , & siemiginowska , a. 2001 , , 548 , 224 voges , w. 1990 , adv .
space res . , 13 , 391 voges , w. , et al . 1999 , , 349 , 389 watson , m. g. , et al .
2008 , , 493 , 339 weisskopf , m. c. , brinkman , b. , canizares , c. , garmire , g. , murray , s. , & van speybroeck , l. p. 2002
, , 114 , 1 weisskopf , m. c. , tananbaum , h. d. , van speybroeck , l. p , & odell , s. l. 2000 , , 4012 , 2 white , n. e. , giommi , p. , & angelini , l. 1994 , , 185 , 4111 williams , r. , hanisch , r. , szalay , a. , & plante , r. 2008 , simple cone search ( int .
alliance ) , http://www.ivoa.net/documents/latest/conesearch.html wise , m. w. , davis , j. e. , huenemoerder , d. p. , houck , j. c. , & dewey , d. 2003 , marx 4.0 technical manual , revision 3.0 ( cambridge , ma : cxc ) , http://space.mit.edu/cxc/marx/ zezas , a. , fabbiano , g. , baldi , a. , schweizer , f. , king , a. r. , ponman , t. j. , & rots , a. h. 2006 , , 166 , 211 zografou , p. , van stone , d. , harbo , p. , & tibbetts , m. 2008 , in asp conf .
394 , astronomical data analysis software and systems xvii , ed .
r. w. argyle , p. s. bunclark , & j. r. lewis ( san francisco , ca : asp ) , 365 | the _ chandra _ source catalog ( csc ) is a general purpose virtual x - ray astrophysics facility that provides access to a carefully selected set of generally useful quantities for individual x - ray sources , and is designed to satisfy the needs of a broad - based group of scientists , including those who may be less familiar with astronomical data analysis in the x - ray regime .
the first release of the csc includes information about 94,676 distinct x - ray sources detected in a subset of public acis imaging observations from roughly the first eight years of the _ chandra _ mission .
this release of the catalog includes point and compact sources with observed spatial extents @xmath0 .
the catalog ( 1 ) provides access to the best estimates of the x - ray source properties for detected sources , with good scientific fidelity , and directly supports scientific analysis using the individual source data ; ( 2 ) facilitates analysis of a wide range of statistical properties for classes of x - ray sources ; and ( 3 ) provides efficient access to calibrated observational data and ancillary data products for individual x - ray sources , so that users can perform detailed further analysis using existing tools .
the catalog includes real x - ray sources detected with flux estimates that are at least 3 times their estimated @xmath1 uncertainties in at least one energy band , while maintaining the number of spurious sources at a level of @xmath2 false source per field for a @xmath3 observation . for each detected source
, the csc provides commonly tabulated quantities , including source position , extent , multi - band fluxes , hardness ratios , and variability statistics , derived from the observations in which the source is detected .
in addition to these traditional catalog elements , for each x - ray source the csc includes an extensive set of file - based data products that can be manipulated interactively , including source images , event lists , light curves , and spectra from each observation in which a source is detected . |
adaptation is fundamental to all organisms for their survival and evolutionary success in a changing environment .
in the course of evolution , living systems have gradually attained and developed more active and efficient strategies for adaptation , which generally accompany more effective use of environmental information .
understanding how the use of information is linked to the efficiency of adaptation is crucial to clarify the fundamental limits and universal properties of biological adaptations@xcite .
the most primitive strategy for adaptation is to randomly generate genetic and phenotypic heterogeneity in a population @xcite .
provided that a sufficiently large heterogeneity is constantly generated in the population , a fraction of organisms can , by chance , have the types adaptive to the upcoming environmental state and circumvent extinction of the population at the cost of others with non - adaptive types@xcite .
such a strategy is known as bet - hedging or phenotypic diversification and works even if the organisms are completely blind to the environment , without any _ a priori _ knowledge of its dynamics .
the bet - hedging is a passive and _ a posteriori _ adaptation in the sense that the adaptation is achieved extrinsically by and after the impact of environmental selection@xcite .
the evolutionary advantage of the bet - hedging strategy is demonstrated by the persistence of bacteria , pathogens , and cancer cells to antibiotic or anticancer drug treatments@xcite .
the gain of fitness by bet - hedging can be optimized if the population evolves to generate an appropriate pattern of heterogeneity by learning the environmental statistics@xcite .
nevertheless , the gain of fitness by bet - hedging is fundamentally limited because of the passive and _ a posteriori _ nature of the strategy , in which the individual organisms have no access to the microscopic information of which environmental states will actually be realized . with any access to such information
, the loss can be avoided further by decision making : directly sensing the current environmental state , predicting the upcoming state , and switching into the phenotypic state that is adaptive to that state @xcite .
the strategy of adaptation via sensing is active and _ a priori _ in a sense that adaptation is intrinsically achieved by the predictive actions of the organisms@xcite . in biologically relevant situations ,
both passive and active aspects of adaptation are intermingled because perfect sensing and prediction of environment are impossible with the limited capacities of biological systems . at an analogical level
, the problem of the fundamental law and the limits of adaptation and evolution shares several aspects with physics , especially with thermodynamics , which drove the long - lasting attempts to establish the thermodynamics of biological adaptation and evolution @xcite . among other areas ,
the fundamental limit of fitness in a changing environment and the value of environmental information have been a major focus in evolutionary biology @xcite .
haccou and iwasa may be the first who linked , albeit implicitly , environmental information with the gain of fitness in a stochastic environment@xcite .
bergstrom and lachmann pursued the fitness value of information by directly incorporating mutual information@xcite .
others also pointed out some quantitative relations between fitness and information measures such as relative entropy and jeffreys divergence@xcite .
more recently , rivoire and liebler conducted a comprehensive analysis by employing an analogy between bet - hedging of organisms and horse race gambling @xcite , the link of which to information theory was revealed in the seminal work by kelly in 1956 @xcite .
however , all previous works either imposed certain restrictions on their models to derive the information - theoretic measures of fitness value @xcite or had to introduce phenomenological measures for the value of information to accommodate more general situations@xcite , because they lacked an appropriate method to handle the mixture of the passive and active aspects in adaptation .
we recently resolved this problem @xcite by combining a path integral formulation of population dynamics @xcite , a retrospective characterization of the selected population @xcite , and a variational structure in population dynamics @xcite .
the results we obtained generalized the limits of fitness gain by sensing and revealed that the gain satisfies fluctuation relations ( frs ) that fundamentally constrain not only its average but also its fluctuation .
these relations , alongside a previous work in the line of neutral theory @xcite , imply that fitness in the fluctuation environment shares , at least mathematically , similar structures to those of stochastic and information thermodynamics@xcite . in our frs , the fluctuation of fitness of a given population is evaluated by the difference from the fitness that achieves the maximum average fitness over all possible phenotypic histories of organisms .
conceptually , this means that we postulate a darwinian demon , an imaginary organism , that can exhibit any type of behavior without imposing any constraint not only on biological capacity but also on the causality of dynamics .
the frs characterize the loss of fitness of a realistic organism from such an idealized organism .
thus , understanding the properties of the darwinian demon and the deviation from it by a realistic organism are central to a deeper understanding of the behavior of populations in a changing environment .
however , the implicit definition of the demon as the maximizer of the average fitness hampers the explicit characterization of the demon and obscures the formal link to stochastic thermodynamics , in which a variational characterization is not common@xcite . more practically , without an explicit characterization , we are unable to simulate possible behaviors of the demon even numerically . in this paper , we resolve these problems by deriving frs of fitness without using the variational approach .
to this end , we first formulate and generalize the problem of adaptation in a changing environment so that individual organisms can change not only their strategy of switching phenotypic states but also the strategy of allocating metabolic resources to each phenotypic state ( sec .
[ sec : modeling ] ) . combined with the path integral formulation of population dynamics ,
this generalization enables us to obtain a decomposition of fitness with a combination of time - forward ( chronological ) and time - backward ( retrospective ) path probabilities ( sec .
[ sec : pathformdecomp ] ) .
the decomposition naturally spells out an explicit representation of the upper bound of the average fitness , which was implicitly defined in our previous work@xcite . for the bet - hedging problem without a sensing environment ( sec .
[ sec : bethedging ] ) , the decomposition directly leads to frs of the fitness loss , which has a very similar form as the frs of entropy production in stochastic thermodynamics@xcite .
after numerically verifying the derived frs ( sec .
[ ssec : verfrbf ] ) , we investigate the biological meanings and achievability of the frs ( sec . [
sec : bmfrbh ] ) .
the average fr is related to the evolutionary stable state ( ess ) , under which the strategy with maximal average fitness can not be invaded by any other strategies .
the detailed and integral frs generalize the ess by giving the probability that a suboptimal strategy outperforms the optimal one within a finite time interval owing to rare fluctuation of the environment@xcite . by using a dualistic relation between phenotypic and environmental dynamics
, the detailed fr is shown to be represented as the ratio of the path probability of the actual environment and that of the conjugate environment under which a given strategy of the organisms becomes optimal .
the duality also clarifies that the average loss of fitness is directly related to the imperfectness of the adaptive behavior of the organisms , originating both from physical constraints and from the suboptimality of the behaviors . the introduction of a sensing signal extends the frs to accommodate the mutual information between the environment and the signal as the gain of fitness by sensing , in the same manner as mutual information bounds the negative gain of entropy production in information thermodynamics@xcite ( sec .
[ sec : frs ] ) .
although the extended frs cover very general situations , the frs are not tight and therefore the mutual information overestimate the value of fitness by sensing . by explicitly assuming a causal relation between the environment and the signal , the frs
are further modified to involve the directed information as a tighter bound of fitness gain ( sec .
[ sec : cfrs ] ) .
this modification clarifies how the loss of fitness from the upper bound is related to the causality , the inaccessibility to perfect information of the environment , and the imperfect implementation of information processing .
finally , three quantities are introduced to account for the fitness loss of inappropriate sensing and the imperfectness of metabolic allocation and phenotypic switching strategies in general situations ( sec .
[ sec : mmis ] ) . the summary and future directions are described in discussion ( sec .
[ sec : sum ] ) .
let @xmath0 , @xmath1 , and @xmath2 be the phenotype of a living organism , the state of the environment , and the state of the sensing signal at time @xmath3 , respectively ( fig .
1 ( a ) ) . for simplicity ,
possible phenotypic , environmental , and signal states are assume to be discrete as in references @xcite .
the paths ( histories ) of the states up to time @xmath3 are defined as @xmath4\ } \in { \mathfrak{s}_{{\mathpzc{x}}}}{:=}{\mathfrak{s}_{x}}^{\times ( t+1)}$ ] , @xmath5\}\in { \mathfrak{s}_{{\mathpzc{y}}}}{:=}{\mathfrak{s}_{y}}^{\times ( t+1)}$ ] , and @xmath6\}\in { \mathfrak{s}_{{\mathpzc{z}}}}{:=}{\mathfrak{s}_{z}}^{\times ( t+1)}$ ] , respectively .
time is also treated as discrete in this work . the phenotype of an organism , in general , switches stochastically over time , depending on its past phenotypic state and the sensed signal ( fig .
1 ( a ) and ( b ) ) .
the switching dynamics is modeled , for example , by a markov transition probability @xmath7 , which satisfies @xmath8 for all @xmath9 and @xmath10 .
although we mainly focus on the markov switching , our result can be extended for causal switching @xmath11 in which the next phenotypic state @xmath9 depends on both past phenotypic and signal histories @xmath12 and @xmath13 .
next , we model the strategy of metabolic resource allocation of the organisms .
each organism is assumed to duplicate asexually to produce @xmath14 daughter organisms on average . for each state of environment @xmath15 , the organisms have a maximum replication rate , @xmath16 , that can be achieved only when the organism allocates its all metabolic resources for adapting only to that environmental state .
because the environment changes , however , the organism usually distributes its resources for different environmental conditions@xcite .
to represent this situation , we introduce a conditional probability @xmath17 that quantitatively represents the fraction of resources allocated to the environmental state @xmath15 in a phenotype @xmath18 .
an instantaneous replication rate @xmath19 of the phenotype @xmath18 under the environmental state @xmath15 is then assumed to be represented as @xmath20 ( fig . 1 ( c ) ) .
note that such a decomposition of @xmath21 can date back to haccou and iwasa@xcite , at least .
this relation between the resource allocation strategy and the replication rate may appear to be restrictive because of the linear relation between the allocation strategy @xmath17 and the actual replication rate @xmath21 .
nevertheless , this decomposition of @xmath19 is general enough because , for a given @xmath19 , we can find a pair of @xmath16 and @xmath17 as long as the possible phenotypic states are fewer than those of the environment ( refer to appendix [ ap : decomp ] )
. such a situation is biologically plausible because the environment is usually more complex than the phenotype of an organism .
moreover , our setting includes the special situation that has been intensively investigated in previous works@xcite .
for example , when the numbers of phenotypic and environmental states are equal as @xmath22,@xmath23 corresponds to kelly s horse race gambling @xcite , in which each phenotypic state allocates all metabolic resources to a certain environmental state and , as a result , can survive and grow only when the phenotypic state matches the realized environmental state .
we finally model the sensing process .
we consider the case that organisms can obtain a sensing signal @xmath24 , which correlates with the environment @xmath25 .
the sensed signal @xmath24 is assumed to be common to all the organisms in the population ( fig .
1 ( a ) ) .
a biologically relevant situation is that @xmath24 is a vector of concentrations of extracellular signaling molecules that can not be consumed as metabolites but correlate with the available metabolites .
another situation is that @xmath24 is a subset of @xmath25 to which the organisms have sensors . in either situation ,
the sensing noise should be negligibly small because all the organisms receive the same sensing signal @xmath24 .
even though sensing of a common signal can not cover all biologically realistic situations such as individual sensing with noisy receptors@xcite , the common signal has been investigated in various works @xcite . in this work
, we mainly focus on the common sensing problem and touch on the individual sensing in discussion .
schematic diagrams for population dynamics of organisms with multiple types and sensing in a changing environment .
( a ) a tree representation of a growing population in a changing environment .
colors indicate different phenotypic states , environmental states , and signal over time . by tracking the phenotypic history from @xmath26 to @xmath3 in a time - forward manner , we can effectively obtain a sample @xmath12 with probability @xmath27 $ ] . by tracking the phenotypic history retrospectively from @xmath3 to @xmath26 in a time - backward manner
, we can obtain a sample @xmath12 with probability @xmath28 $ ] .
note that we assume that the size of the population is sufficiently large when we introduce the population dynamics of the organisms eq .
( [ eq : popdyn ] ) .
( b ) phenotypic switching from @xmath9 to @xmath29 in response to environmental signal @xmath10 .
the switching probability is represented by @xmath7 .
( c ) replication of organisms with phenotypic state @xmath9 under environmental state @xmath30 .
each organism generates @xmath31 descendants on average . ] by combining the phenotype switching , the metabolic allocation , and the sensing strategies , we can explicitly derive the dynamics of the population of the organisms . because both environmental and sensing histories are external factors of the organisms , the population dynamics of the organisms is described for a given pair of environmental and signaling histories , @xmath32 and @xmath33 .
let @xmath34 be the number of organisms whose phenotypic state is @xmath9 at time @xmath3 under the realization of environmental and signal histories @xmath32 and @xmath33 .
the population size of the organisms is assumed to be sufficiently large so that @xmath35 can be well approximated as a continuous variable . because the large population size enable us to effectively ignore the demographic fluctuation due to finite number of the organisms , we can obtain the update rule@xcite of @xmath35 as @xmath36 if we need to work on a population , the size of which is not sufficiently large , modeling of the population e.g. , by a branching process is required .
the statistical properties of the environmental and signal histories are generally characterized by a joint path probability @xmath37 $ ] .
by using @xmath35 in the previous section , we define the fitness of the population and derive its path integral formulation . the cumulative fitness @xmath38 $ ] of the population at @xmath3 under environmental and signal histories @xmath32 and @xmath33 is defined by the exponential expansion of the total population size as follows : @xmath39 { : = } \ln\frac{\sum_{x_{t}}{\mathcal{n}}_{t}^{{\mathpzc{y}},{\mathpzc{z}}}(x_{t})}{\sum_{x_{0}}{\mathcal{n}}_{0}^{{\mathpzc{y}},{\mathpzc{z}}}(x_{0 } ) } = \ln\frac{{\mathcal{n}}_{t}^{{\mathpzc{y}},{\mathpzc{z}}}}{{\mathcal{n}}_{0}^{{\mathpzc{y}},{\mathpzc{z}}}}.\end{aligned}\ ] ] we here define @xmath40 .
the ensemble average of the cumulative fitness for different realizations of the environmental and signal histories is represented as @xmath41 \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}.\end{aligned}\ ] ] as derived in @xcite , the cumulative fitness at time @xmath3 can be represented with a path - wise ( path integral ) formulation .
let us first define the time - forward path probability of the phenotype switching as @xmath42 & { : = } \left[\prod_{\tau=0}^{t-1}{\mathbb{t}}_{f}(x_{\tau+1}|x_{\tau},z_{\tau+1})\right]p(x_{0}),\end{aligned}\ ] ] where @xmath43 .
we here use kramer s causal conditioning @xmath44 rather than the usual conditioning @xmath45 in order to indicate that the path probability @xmath27 $ ] is causally generated by the markov transition matrix @xmath7 , which depends only on the past phenotypic state @xmath9 and the signal @xmath10@xcite .
the single bar @xmath45 is also used for the normal conditioning of a path probability that does not necessarily satisfy the causal relation between conditioning and conditioned histories .
we also define the path - wise ( historical ) fitness of a phenotypic history @xmath12 under an environmental history @xmath32 as @xmath46{:=}\sum_{\tau=0}^{t-1}{k}(x_{\tau+1 } , y_{\tau+1}).\end{aligned}\ ] ] where @xmath47 $ ] is defined over all @xmath48 . with these path - wise quantities@xcite
, we obtain the population size of the organisms at time @xmath3 , the past phenotypic history of which is @xmath12 as @xmath49 = e^{{k}[{\mathpzc{x}}_{t},{\mathpzc{y}}_{t}]}{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathcal{n}}_{0}^{{\mathpzc{y}},{\mathpzc{z}}}.\ ] ] because @xmath50 $ ] , the cumulative fitness with sensing @xmath51 $ ] is explicitly described as @xmath39 & = \ln\frac{{\mathcal{n}}_{t}^{{\mathpzc{y}},{\mathpzc{z}}}}{{\mathcal{n}}_{0}^{{\mathpzc{y}},{\mathpzc{z}}}}=\ln{\left < e^{{k}[{\mathpzc{x}}_{t},{\mathpzc{y}}_{t } ] } \right>}_{{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]}.\label{eq : cmff}\end{aligned}\ ] ] from this representation , the fitness can be described variationally @xcite as @xmath39 & = \max_{{\mathbb{p}}[{\mathpzc{x}}_{t}]}\left[{\left < { k}[{\mathpzc{x}}_{t},{\mathpzc{y}}_{t } ] \right>}_{{\mathbb{p}}[{\mathpzc{x}}_{t}]}-{{\mathcal{d}}\big[{\mathbb{p}}[{\mathpzc{x}}_{t}]\big|\big|{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]\big ] } \right],\end{aligned}\ ] ] where @xmath52{:=}\sum_{{\mathpzc{x}}_{t}}{\mathbb{p}}[{\mathpzc{x}}_{t}]\ln { \mathbb{p}}[{\mathpzc{x}}_{t}]/{\mathbb{p}}'[{\mathpzc{x}}_{t}]$ ] is the kullback - leibler divergence ( relative entropy)@xcite between two path measures @xmath53 and @xmath54 .
it should be noted that both kl divergence and kramer s causal conditioning use the double bar @xmath44 but their meanings are different . the maximization is achieved with the time - backward retrospective path probability defined by @xmath55 & { : = } e^{{k}[{\mathpzc{x}}_{t},{\mathpzc{y}}_{t}]-{\psi}^{{s}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t } ] { = \frac{{\mathcal{n}}_{t}^{{\mathpzc{y}},{\mathpzc{z}}}[{\mathpzc{x}}_{t}]}{{\mathcal{n}}_{t}^{{\mathpzc{y}},{\mathpzc{z } } } } } .
\label{eq : pbc}\end{aligned}\ ] ] if the phenotypic switching does not depend on the sensing signal as @xmath27={\mathbb{p}}_{f}[{\mathpzc{x}}_{t}]$ ] , which corresponds to the bet - hedging by random phenotypic switching , @xmath56is reduced to @xmath57 & { : = } e^{{k}[{\mathpzc{x}}_{t},{\mathpzc{y}}_{t}]-{\psi}^{b}[{\mathpzc{y}}_{t}]}{\mathbb{p}}_{f}[{\mathpzc{x}}_{t } ] = \frac{{\mathcal{n}}_{t}^{{\mathpzc{y}}}[{\mathpzc{x}}_{t}]}{{\mathcal{n}}_{t}^{{\mathpzc{y}}}},\label{eq : pbb}\end{aligned}\ ] ] where the superscript @xmath58 denotes bet - hedging and @xmath59 { : = } \ln{\left < e^{{k}[{\mathpzc{x}}_{t},{\mathpzc{y}}_{t } ] } \right>}_{{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}]}$ ] . because @xmath60 $ ] is the number of organisms with phenotypic history @xmath12 in the population at time @xmath3 , the second equality in eq .
( [ eq : pbc ] ) indicates that @xmath61 $ ] is the fraction of the organisms that has phenotpyic history @xmath12 .
this property of @xmath61 $ ] leads to an interpretation of @xmath61 $ ] as the probability of observing a certain phenotypic history @xmath12 under a realization of the environmental and signal histories @xmath32 and @xmath33 when we sample an organism in the population at time @xmath3 and track its phenotypic history in a time - backward manner , retrospectively@xcite . because the organisms grow more if their phenotypic histories are more adaptive than others for the given environmental realization @xmath32 , the chance to observe a certain phenotypic history @xmath12 after selection under @xmath32 is biased to @xmath61 $ ] from the probability @xmath27 $ ] with which the same phenotypic history is intrinsically generated . because the selected phenotypic histories strongly depend on the actual realization of the environmental history , @xmath61 $ ] is conditional on @xmath32 .
it should be noted that @xmath62 $ ] is not necessarily causal , because of which we use the normal conditioning @xmath45@xcite . in order to understand the relation between fitness and information obtained by sensing , we decompose the cumulative fitnesses into biologically relevant components . to obtain the decompositions ,
we first define a constant @xmath63 and a probability distribution @xmath64 by using @xmath65 as @xmath66 and @xmath67 . from these definitions ,
@xmath65 can be described as @xmath68 . by defining @xmath69{:=}\prod_{\tau=0}^{t-1}q_{0}(y_{\tau+1})$ ] ,
@xmath70{:=}\prod_{\tau=0}^{t-1}{\mathbb{t}}_{{k}}(y_{\tau+1}|x_{\tau+1})$ ] , and @xmath71 , we obtain the following decomposition of @xmath72 : @xmath46 = \phi_{0 } + \ln \frac{{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]}{{\mathbb{q}}_{0}[{\mathpzc{y}}_{t } ] } \label{eq : decombk},\end{aligned}\ ] ] where we use eq .
( [ eq : ek ] ) . by defining @xmath73{:=}\sum_{\tau=0}^{t-1}{k}_{max}(y_{\tau+1})=\phi_{0}-\ln { \mathbb{q}}_{0}[{\mathpzc{y}}_{t}]$ ] ,
@xmath47 $ ] can also be described as @xmath46 = { k}_{\max}[{\mathpzc{y}}_{t } ] + \ln { \mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}].\end{aligned}\ ] ] with eqs .
( [ eq : pbb ] ) and ( [ eq : decombk ] ) , for the bet - hedging problem , we obtain a decomposition of the fitness @xmath74 = \phi_{0 } - \ln { \mathbb{q}}_{0}[{\mathpzc{y}}_{t } ] -\ln \frac{{\mathbb{p}}^{b}_{b}[{\mathpzc{x}}_{t}| { \mathpzc{y}}_{t}]}{{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}]}.\end{aligned}\ ] ] for a given environmental statistics @xmath75 $ ] , the fitness is represented by the ratio of the time - forward and time - backward path probabilities as @xmath74 = { \psi}_{0}[{\mathpzc{y}}_{t } ] -\ln \frac{{\mathbb{p}}^{b}_{b}[{\mathpzc{x}}_{t } , { \mathpzc{y}}_{t}]}{{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}]},\label{eq : fdb_1}\end{aligned}\ ] ] where @xmath76{:=}{\mathbb{p}}^{b}_{b}[{\mathpzc{x}}_{t}|{\mathpzc{y}}_{t}]{\mathbb{q}}[{\mathpzc{y}}_{t}]$ ] is the time - backward joint probability of the phenotypic and environmental histories @xmath12 and @xmath32 .
here we also define @xmath77{:=}\phi_{0 } + \ln \frac{{\mathbb{q}}[{\mathpzc{y}}_{t}]}{{\mathbb{q}}_{0}[{\mathpzc{y}}_{t } ] } = { k}_{\max}[{\mathpzc{y}}_{t}]+\ln { \mathbb{q}}[{\mathpzc{y}}_{t}].\label{eq : phi0}\end{aligned}\ ] ] if the organisms can perfectly foresee that the environmental state at time @xmath78 becomes @xmath79 and if they can choose the phenotype that allocates all metabolic resource to the environmental state @xmath79 , the maximum replication rate at time @xmath80 reaches @xmath81 .
therefore , @xmath73 $ ] is interpreted as the maximum replication over an environmental path @xmath32 that can be achieved only when the organisms perfectly foresee what kind of environmental history will be realized in advance .
in contrast , @xmath82 $ ] is the entropic loss of fitness due to the lack of knowledge of which environmental history will be realized@xcite .
therefore , @xmath83 $ ] is the maximum replication when the organisms can not know which environmental state will be realized but know the statistics of the future environmental state .
the relevance of this interpretation and the biological meaning of some quantities such as @xmath84 and @xmath69 $ ] are explicitly shown by using the frs derived in the following sections . for the case with the sensing signal
, we can similarly obtain a decomposition of @xmath51 $ ] as @xmath39 = { \psi}_{0}[{\mathpzc{y}}_{t } ] -\ln \frac{{\mathbb{p}}_{b}^{{s}}[{\mathpzc{x}}_{t},{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}{{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t}|{\mathpzc{y}}_{t } ] } \label{eq : fitnessdecombcommon}\end{aligned}\ ] ] where @xmath85 { : = } { \mathbb{p}}_{b}^{{s}}[{\mathpzc{x}}_{t}|{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]$ ] is the time - backward joint path probability among @xmath12 , @xmath32 , and @xmath33 .
it should be noted that @xmath70{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t}|{\mathpzc{y}}_{t}]$ ] is not a joint path probability because of the circular noncausal dependency among @xmath12 , @xmath32 , and @xmath33 .
finally , by using the decomposition in eq .
( [ eq : decombk ] ) , the time - backward conditional path probabilities , eq . ( [ eq : pbb ] ) and eq .
( [ eq : pbc ] ) , are reduced to @xmath57 & = \frac{{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}]}{{\mathbb{p}}_{{k},f}[{\mathpzc{y}}_{t}]},\label{eq : pbb2}\\ { \mathbb{p}}^{{s}}_{b}[{\mathpzc{x}}_{t}|{\mathpzc{y}}_{t},{\mathpzc{z}}_{t } ] & = \frac{{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]}{{\mathbb{p}}_{{k},f}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]},\label{eq : pbc2}\end{aligned}\ ] ] where the normalization factors are @xmath86 & { : = } \sum_{{\mathpzc{x}}_{t}}{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}],\\ { \mathbb{p}}_{{k},f}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t } ] & { : = } \sum_{{\mathpzc{x}}_{t}}{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}].\end{aligned}\ ] ] because @xmath87 $ ] is the probability to intrinsically generate the repertoire of phenotypic histories in the population and @xmath70 $ ] is the probability that an organism with a phenotypic history @xmath12 allocates resources to each history of environment @xmath32 , the normalization factor @xmath88 $ ] can be interpreted as the marginal resource allocation to the environmental history @xmath32 at the population level .
@xmath89 $ ] can similarly be interpreted as the population - level resource allocation to @xmath32 when signal history @xmath33 is received . by using frs in the next section
, we clarify that @xmath87 $ ] and @xmath89 $ ] also have meaning as the conjugate environment under which the given strategy @xmath90 becomes optimal .
by rearranging the decomposition of @xmath59 $ ] in eq .
( [ eq : fdb_1 ] ) , we can immediately obtain a detailed causal fr for fitness difference @xmath83-{\psi}^{b}[{\mathpzc{y}}_{t}]$ ] as @xmath91-{\psi}^{b}[{\mathpzc{y}}_{t } ] ) } = \frac{{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}]}{{\mathbb{p}}_{b}^{b}[{\mathpzc{x}}_{t},{\mathpzc{y}}_{t}]}= \frac{{\mathbb{p}}_{{k},f}[{\mathpzc{y}}_{t}]}{{\mathbb{q}}[{\mathpzc{y}}_{t } ] } , \label{eq : dfrb}\end{aligned}\ ] ] where we use eq .
( [ eq : pbb2 ] ) to obtain the last equality .
the first equality means that the fitness difference is the log ratio of time - forward and time - backward path probabilities @xmath70{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}]$ ] and @xmath92 $ ] .
@xmath70{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}]$ ] is the time - forward probability of observing an organism that takes phenotypic history @xmath12 and then allocates @xmath70 $ ] of metabolic resources to @xmath32 _ a priori _ to selection by conducting time - forward tracking of the histories .
@xmath92 $ ] is the time - backward probability of observing the realization of environmental history @xmath32 and the time - backward phenotypic history @xmath12 _ a posteriori _ to selection by conducting time - backward tracking of the histories .
the second equality also indicates that the fitness difference is the log ratio of the percentage of resource allocated to environmental history @xmath32 at the population level and the probability of observing environmental history @xmath32 . by averaging eq .
( [ eq : dfrb ] ) with respect to @xmath92 $ ] or @xmath75 $ ] , we can derive an integral fr as @xmath93-{\psi}^{b}[{\mathpzc{y}}_{t } ] ) } \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t}]}= 1\label{eq : ifrb}.\end{aligned}\ ] ] if we average eq .
( [ eq : dfrb ] ) after taking the logarithm of both sides , we obtain an average fr as @xmath94 where @xmath95 \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t}]}$ ] , @xmath96 \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t}]}$ ] , and @xmath97\big|\big|{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}]\big ] } \\ & = { { \mathcal{d}}\big[{\mathbb{q}}[{\mathpzc{y}}_{t}]\big|\big|{\mathbb{p}}_{{k},f}[{\mathpzc{y}}_{t}]\big ] } \label{eq : dloss}\end{aligned}\ ] ] because of the non - negativity of the relative entropy @xmath98 , we can easily see that @xmath99 is an upper bound of the average fitness @xmath100 of a bet - hedging strategy : @xmath101 where we use the fact that @xmath83 $ ] is dependent neither on the phenotype switching strategy @xmath102 nor on the metabolic allocation strategy @xmath103 .
these frs are basically the same as those we derived in our previous work by using a variational approach .
it should be also noted that the frs derived by mustonen and lssig@xcite are different from ours because their relations are based on a model describing the dynamics of an ensemble of populations whereas ours is one describing the dynamics of a population . from eq .
( [ eq : phi0 ] ) , the upper bound of the average fitness , @xmath105 admits two different representations : @xmath106=\phi_{0 } + { { \mathcal{d}}\big[{\mathbb{q}}\big|\big|{\mathbb{q}}_{0}\big]},\end{aligned}\ ] ] where @xmath107 { : = } - { \left < \ln { \mathbb{q}}[{\mathpzc{y}}_{t } ] \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t}]}$ ] is the entropy of @xmath75 $ ] and @xmath108 is the average fitness under environmental statistics @xmath75 $ ] that is attained only when organisms have perfect knowledge of the future environment .
therefore , the first equality means that the randomness of the environment quantified by the entropy @xmath107 $ ] works as the inevitable loss of fitness due to the lack of knowledge of which environmental history will be realized in the future .
if the environment fluctuates more unpredictably , we have a higher @xmath107 $ ] and a lower upper bound of the average fitness .
note that these properties of @xmath105 have been pointed out previously and repeatedly @xcite .
schematic representations of environmental dynamics , phenotypic switching , and sensing used for numerical verification of frs in figs . 3 and 5 .
( a ) characteristics of the environment .
the environment has three states : @xmath109 . @xmath110 and @xmath111 are nutrient rich states that have different sources of nutrients .
@xmath112 is a nutrient - poor state under which replication of the organisms is severely restricted .
the maximum growth under each environment is @xmath113 , @xmath114 , and @xmath115 .
the transition rates @xmath116 between the states are shown on the arrows . the environmental states usually fluctuate between @xmath110 and @xmath111 but occasionally flip to @xmath112 with a @xmath117 chance .
( b ) characteristics of phenotypic states without sensing .
organisms have two phenotypic states : @xmath118 .
the transition rates @xmath119 between the states are shown on the arrows .
@xmath120 is the phenotypic state allocating more resources to the environmental state @xmath110 ( @xmath121 ) than @xmath122 ( @xmath123 ) whereas @xmath122 allocates more to @xmath122 ( @xmath121 ) than @xmath120 ( @xmath123 ) .
both states allocate @xmath124 of the resources to the starving environmental state @xmath112 .
( c ) characteristic of sensing signal . the sensing signal has two states as @xmath125 .
when the environmental state is @xmath110 ( @xmath111 ) , the organisms obtain @xmath126 ( @xmath127 ) as the sensing signal with a @xmath128 accuracy .
if the environment is in @xmath112 , the sensing signal produces @xmath126 or @xmath127 with equal probability .
( d ) characteristic of phenotypic switching with sensing .
when the organism obtains the sensing signal @xmath126 ( @xmath127 ) , it switches its phenotypic state into @xmath120 ( @xmath122 ) @xmath129 of the time . ]
the meaning of @xmath84 and @xmath69 $ ] in the second equality becomes explicit by considering the minimization of @xmath130 with respect to @xmath75 $ ] as follows : @xmath131 \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t } ] } \ge \min_{{\mathbb{q}}}\max_{\{{\mathbb{t}}_{f},{\mathbb{t}}_{{k}}\}}{\left < { \psi}^{b}[{\mathpzc{y}}_{t } ] \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t } ] } , \end{aligned}\ ] ] where we use @xmath132 \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t}]}=\phi_{0 } + \min_{{\mathbb{q}}}{\mathcal{d}}[{\mathbb{q}}||{\mathbb{q}}_{0 } ] = \phi_{0}. \end{aligned}\ ] ] this relation indicates that @xmath84 is the minimum of the maximum average fitness and that @xmath69 $ ] is the worst environment for the organisms under which the maximum average fitness is minimized .
this min - max characterization of @xmath84 and @xmath69 $ ] has been clarified in the context of game theory with a matrix formulation@xcite .
equations ( [ eq : dfrb][eq : afrb ] ) indicate that , under quite general situations , the fitness difference @xmath83-{\psi}^{b}[{\mathpzc{y}}_{t}]$ ] satisfies the frs @xcite as the entropy production does in stochastic thermodynamics@xcite .
to demonstrate the relations , we consider an organism with two phenotypic states growing in a markovian environment with three states as depicted in fig .
2 ( a ) and ( b ) .
the three environmental states , @xmath110 , @xmath111 , and @xmath112 , describe nutrient a rich , nutrient b rich , and nutrient - poor conditions , respectively ( fig 2(a ) ) .
the two phenotypic states , @xmath120 and @xmath122 , employ strategies to allocate @xmath133 of the metabolic resources to @xmath110 and @xmath111 , respectively .
both states allocate @xmath134 resources to the rest of two states ( see appendix [ ap : num ] for more details ) .
this setting abstractly and simply represents the fact that organisms generally have much less phenotypic and sensing states than possible environmental states because of the limited physical complexity of the organisms@xcite .
figures 3 ( a ) and ( b ) show the population dynamics of the organisms with the phenotypic states @xmath120 and @xmath122 under two different realizations of the environmental history alongside @xmath59 $ ] and @xmath83 $ ] . depending on the actual realization of the environment
, the relative relations among @xmath135 , @xmath136 , @xmath137}={\mathcal{n}}_{t}^{{\mathpzc{y}}}$ ] , and @xmath138}$ ] change over time stochastically . in fig 3 ( a ) ,
@xmath138}$ ] is mostly greater than @xmath139}$ ] , which reflects the average relation @xmath140 . on the contrary , @xmath141 frequently becomes greater than @xmath138}$ ] in fig 3 ( b ) .
figures 3 ( c ) and ( d ) show that the environmental fluctuation induces a large fluctuation in both @xmath59 $ ] and @xmath83 $ ] . as shown in fig 3 ( e ) , although most environmental fluctuations result in positive fitness differences , that is , @xmath83-{\psi}^{b}[{\mathpzc{y}}_{t}]>0 $ ] , rare environmental fluctuations lead to a negative fitness difference in a finite time interval , meaning that the fitness of the suboptimal strategy @xmath59 $ ] outperforms the average upper bound @xmath83 $ ] .
this is analogue to the reversed heat flow in a small thermal system@xcite .
such rare events are balanced to satisfy the integral fr in eq .
( [ eq : ifrb ] ) as verified numerically in fig 3 ( f ) . numerical simulation of the population dynamics without sensing defined in fig .
( a , b ) two sample histories of environment @xmath32 and the dynamics of @xmath135 , @xmath136 , @xmath142}$ ] , and @xmath138}$ ] under @xmath32 obtained by solving eq .
( [ eq : popdyn ] ) .
we set @xmath143 .
the color bars on the graphs represent the state of the environment at each time point .
the correspondence between the colors and environmental states is as shown in fig . 2 ( a ) .
cyan and yellow lines represent @xmath135 and @xmath136 , respectively .
the dashed blue line with filled grey style is @xmath141 .
the red line is @xmath144 $ ] .
( c ) stochastic behaviors of population fitness @xmath59 $ ] for @xmath145 independent samples of the environmental histories .
each colored thin line represents @xmath137}={\mathcal{n}}_{t}^{{\mathpzc{y}}}/{\mathcal{n}}_{0}^{{\mathpzc{y}}}$ ] for each realization of environmental history .
the thick black line is @xmath146 \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t}]}}$ ] .
( d ) stochastic behaviors of @xmath83 $ ] for the same @xmath145 samples of environmental histories as those in ( c ) .
each colored thin line represents @xmath138}$ ] for each realization of the same environmental history as in ( c ) .
( e ) stochastic behaviors of @xmath147-{\psi}^{b}[{\mathpzc{y}}_{t}])}$ ] for the same @xmath145 samples of environmental histories as those in ( c ) .
each colored thin line represents @xmath147-{\psi}^{b}[{\mathpzc{y}}_{t}])}$ ] for each realization of the same environmental history as in ( c ) .
( f ) @xmath148-{\psi}^{b}[{\mathpzc{y}}_{t } ] ) } \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t}]}$ ] calculated empirically by the numerical simulations of sample paths of @xmath147-{\psi}^{b}[{\mathpzc{y}}_{t}])}$ ] .
the thick line is the average of @xmath149 independent samples of environmental histories . to illustrate the fluctuation ,
the @xmath149 histories are dissected into 100 groups of histories , each contains @xmath150 histories .
each thin line is obtained as the average of @xmath150 histories in each group . ]
in the original detailed fr over paths@xcite , the entropy production is the log ratio of the path probability of a system s trajectory and its time reversal .
the average entropy production attains its minimum @xmath151 only when the time reversibility of the system holds in the sense that the probabilities of observing the time - forward and the time - reversed trajectories are equal .
thus , the frs are related to the extent of the time reversibility of the system . in contrast , the detailed frs for the fitness difference ( eq .
( [ eq : dfrb ] ) ) is the log ratio between the probability @xmath75 $ ] of observing the environmental history and the percentage @xmath88 $ ] of the marginal resource allocation to @xmath32 , or that between the time - backward path probability @xmath92 $ ] and time - forward path probability @xmath70{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}]$ ] . by investigating the frs
, we clarify a dualistic structure , a conjugacy of these quantities , and the time - reversal condition for the equality attained in eq .
( [ eq : afrbbound ] ) .
the average fr in eq . ( [ eq : afrb ] ) implies that the maximization of @xmath100 with respect to the strategies is dual to the minimization of the relative entropy @xmath152 as follows : @xmath153 because @xmath83 $ ] is independent of @xmath154 and @xmath155 .
this duality indicates that maximizing the average fitness by choosing the best strategy is equivalent to the organisms to implicitly learn and prepare for the environmental statistics @xmath75 $ ] so that the marginal resource allocation @xmath156 $ ] to @xmath32 becomes close to @xmath75 $ ] because @xmath157||{\mathbb{p}}_{{k } , f}[{\mathpzc{y}}_{t}]]$ ] .
the upper bound of the average fitness in eq .
( [ eq : afrbbound ] ) is achieved as @xmath59={\psi}_{0}[{\mathpzc{y}}_{t}]$ ] if and only if @xmath158 satisfies @xmath75={\mathbb{p}}_{{k } , f}^{\dagger}[{\mathpzc{y}}_{t}]$ ] where @xmath159 { : = } \sum_{{\mathpzc{x}}_{t}}{\mathbb{t}}_{{k}}^{\dagger}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{t}}_{f}^{\dagger}[{\mathpzc{x}}_{t}]$ ] , meaning that the environmental statistics and the marginal resource allocation match perfectly . for
a given environment , the strategy that achieves the bound may not always exist .
in contrast , for a given pair of strategies @xmath161 , there always exists the environment @xmath162 $ ] under which the pair achieves the bound @xmath163 \right>}_{{\mathbb{q}}^{\dagger}[{\mathpzc{y}}_{t } ] } \ge \max_{\{{\mathbb{t}}'_{f},{\mathbb{t}}'_{{k}}\}}{\left < { { \psi}^{b}}'[{\mathpzc{y}}_{t } ] \right>}_{{\mathbb{q}}^{\dagger}[{\mathpzc{y}}_{t } ] } , \label{eq : bhbound}\end{aligned}\ ] ] and @xmath164 \right>}_{{\mathbb{q}}^{\dagger}[{\mathpzc{y}}_{t}]},\end{aligned}\ ] ] is satisfied . because of the duality shown in eq .
( [ eq : optb ] ) , @xmath162 $ ] is explicitly obtained as @xmath162={\mathbb{p}}_{f,{k}}[{\mathpzc{y}}_{t}]$ ] .
therefore , @xmath156 $ ] can also be regarded as the conjugate environment to the strategy @xmath161 or @xmath165 under which they are optimal .
therefore , the fitness difference in eq .
( [ eq : dfrb ] ) is the log ratio of the actual environmental statistics @xmath75 $ ] and that of the conjugate environment @xmath162={\mathbb{p}}_{{k } , f}[{\mathpzc{y}}_{t}]$ ] of the given strategy .
@xmath59 $ ] is bounded by @xmath83 $ ] on average , and therefore , the optimal strategy that attains @xmath83 $ ] can not be invaded by any other strategy if we consider an infinitely large population and the asymptotic dynamics .
thus , the optimal strategy is a version of ess in a fluctuating environment . within a finite time interval ,
however , @xmath59 $ ] becomes greater than @xmath83 $ ] under certain realizations of environment @xmath32 that satisfy @xmath75<{\mathbb{q}}^{\dagger}[{\mathpzc{y}}_{t}]$ ] .
in stochastic thermodynamics , such realizations correspond to the temporal reversed heat flow in a small thermal system@xcite . in biology , they temporally violate ess because a suboptimal strategy @xmath166 under @xmath167 can outperform the optimal strategy or near optimal strategies with the aid of the environmental fluctuation within a finite time interval .
the integral fr in eq .
( [ eq : ifrb ] ) tells us that the violation of ess can always occur with a small but finite probability in a finite time interval .
if the population size of the optimal strategy is finite , such violation can leads to extinction of the optimal population with a finite probability@xcite .
moreover , the detailed fr in eq .
( [ eq : dfrb ] ) implies that a greater violation can occur under a realization of environmental history @xmath32 if @xmath75 $ ] is small but @xmath162 $ ] is large .
this fact can be intuitively understood as follows : the greater violation is induced by the environmental history @xmath32 that rarely occurs in the actual environmental statistics @xmath75 $ ] but is adaptive and advantageous for the given strategy .
a crucial fact is that this intuitive understanding is supported by a quantitative relation as in eq .
( [ eq : dfrb ] ) .
furthermore , the detailed fr suggests that greater violation can occur for specific suboptimal strategies than for others if @xmath75 $ ] contains many rare environmental histories .
in contrast , if @xmath75 $ ] is perfectly random as @xmath75={\mathrm{const.}}$ ] , the extent of the violation is limited and all the suboptimal strategies have only an even chance of violation .
this implies that structured environmental fluctuation promotes impactful violation by some specific suboptimal strategies .
if @xmath59={\psi}_{0}[{\mathpzc{y}}_{t}]$ ] holds , moreover , the time - backward probability @xmath168 $ ] and the time - forward path probability @xmath169{\mathbb{p}}_{f}^{\dagger}[{\mathpzc{x}}_{t}]$ ] become equal : @xmath170={\mathbb{p}}_{b}^{b,\dagger}[{\mathpzc{x}}_{t}|{\mathpzc{y}}_{t}]{\mathbb{q}}[{\mathpzc{y}}_{t}]={\mathbb{p}}_{{k}}^{\dagger}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}^{\dagger}[{\mathpzc{x}}_{t}]\end{aligned}\ ] ] where we use the first equality in eq .
( [ eq : dloss ] ) .
marginalization of this equality with respect to @xmath32 indicates that the marginalized time - backward path probability @xmath171 $ ] satisfies the consistency condition @xmath172={\mathbb{p}}_{f}^{\dagger}[{\mathpzc{x}}_{t}]$ ] as shown previously@xcite .
thus , the optimal strategies to achieve the bound have time reversibility in the sense that the ensembles of the time - forward and time - backward phenotypic histories are indistinguishable without knowing the actual environmental history that the population experienced . while the definition of time reversibility is different , this result is closely related to the fact that the average entropy production attains @xmath151 when the time reversibility is satisfied .
biologically , this result is quite important because we can evaluate the optimality of an organism in a changing environment by just observing its phenotypic dynamics without directly measuring the environment that the organisms experience .
it should be noted , however , that @xmath172={\mathbb{p}}_{f}^{\dagger}[{\mathpzc{x}}_{t}]$ ] is a necessary but not sufficient condition for @xmath59={\psi}_{0}[{\mathpzc{y}}_{t}]$ ] .
the duality eq .
( [ eq : optb ] ) indicates that the achievability of the bound in eq .
( [ eq : afrbbound ] ) depends on the actual property of the environmental statistics and biological constraints on the selectable strategies . if the environment is a time - homogeneous markov process as @xmath75=\prod_{\tau=0}^{t-1}{\mathbb{t}}_{e}(y_{\tau+1}|y_{\tau})q(y_{0})$ ] and
if the number of possible phenotypic states is the same as that of the environmental states , that is , @xmath173 , the bound can be achieved by a pair of strategies : @xmath174 where @xmath175 and @xmath176 is the kronecker delta .
this pair is equivalent to the optimal bet - hedging strategy in kelly s horse race gambling because @xmath177 means that organisms can survive only when their phenotypic state matches the current environmental state ; they die out otherwise . to be specific , we call @xmath23 kelly s strategy of metabolic allocation . under biologically realistic constraints
, however , kelly s strategy can not be the optimal strategy because the possible phenotypic states are usually much fewer than those of the environment as in fig 2 and 3 .
allocating all resources to a specific environment easily leads to extinction if the phenotypic states can not cover all the possible environmental states . with a limited capacity in possible phenotypic states
, @xmath178 can be attained only if the environmental fluctuation has a hidden structure the dimensionality of which is sufficiently low .
this is manifested by the fact that the conjugate environment @xmath179 is the environment with a hidden dynamics @xmath180 $ ] that generates the actual environmental history as @xmath162=\sum_{{\mathpzc{x}}'_{t}}{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}|{\mathpzc{x}}'_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}'_{t}]$ ] .
because such a low dimensional structure may not always exist , however , @xmath181 is generally not zero but finite and positive under a biological constraint that the possible phenotypic states are fewer than environmental ones .
therefore , @xmath105 is generally attained only by a darwinian demon that can not perfectly foresee the future environment but has sufficient capacity in its phenotypic properties to perfectly learn and prepare for the environmental fluctuation @xmath75 $ ] .
even when the bound is not achieved , @xmath182 has explicit meaning as the component in the environmental fluctuation that can not be learned or represented by the dynamics of the cell s strategy .
for example , when @xmath102 is memoryless and @xmath75 $ ] is a stationary markov process with the stationary probability @xmath183 , that is , @xmath75=\prod_{\tau=0}^{t-1}{\mathbb{t}}_{e}^{f}(y_{\tau+1}|y_{\tau})q(y_{0})$ ] , then @xmath184 } \right]\\ & = { \mathcal{i}}^{x_{t } ; x_{t+1}},\end{aligned}\ ] ] where @xmath185}$ ] is the mutual information that measures the correlation of environmental states between consecutive time points , which can not be learned or mimicd by the memoryless phenotypic switching .
next , we consider the case in which the organisms can exploit the information obtained from the sensing signal
. the fitness decomposition in eq . ( [ eq : fitnessdecombcommon ] ) can be rearranged as @xmath39 & = { \psi}_{0}[{\mathpzc{y}}_{t } ] + i[{\mathpzc{y}}_{t } ; { \mathpzc{z}}_{t } ] -\ln \frac{{\mathbb{p}}_{b}^{{s}}[{\mathpzc{x}}_{t},{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}{{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t } ] } \\ & = { \psi}_{0}[{\mathpzc{y}}_{t } ] + i[{\mathpzc{y}}_{t } ; { \mathpzc{z}}_{t } ] -\ln \frac{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}{{\mathbb{p}}_{{k},f}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t}]},\end{aligned}\ ] ] where @xmath186{:=}\ln { \mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]/{\mathbb{q}}[{\mathpzc{y}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t}]$ ] is the bare mutual information between @xmath32 and @xmath33 and we use eq .
( [ eq : pbc2 ] ) to derive the last equality . from this
, we can similarly obtain detailed , integral , and average frs with information as follows : @xmath91 + i[{\mathpzc{y}}_{t } ; { \mathpzc{z}}_{t}]-{\psi}^{{s}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t } ] ) } = \frac{{\mathbb{p}}_{{k},f}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t}]}{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]},\label{eq : dfc1}\end{aligned}\ ] ] @xmath93 + i[{\mathpzc{y}}_{t } ; { \mathpzc{z}}_{t}]-{\psi}^{{s}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t } ] ) } \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t } ] } = 1,\end{aligned}\ ] ] and @xmath187 where @xmath188 \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}$ ] is the path - wise mutual information between @xmath32 and @xmath33 , and @xmath189\big|\big|{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t}]\big ] } \notag \\ = & { { \mathcal{d}}\big[{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]\big|\big|{\mathbb{p}}_{{k},f}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t}]\big ] } \label{eq : dsloss}.\end{aligned}\ ] ] the way information terms involved in eqs .
( [ eq : dfc1])-([eq : afrc1 ] ) is the same as the way those appearing in the sagawa - ueda relations , where the maxwell demon and feedback regulation are involved@xcite . because of the non - negativity of @xmath190 , @xmath191 is upper bounded by @xmath192 : @xmath193,\end{aligned}\ ] ] where @xmath194 { : = } - { \left < \ln { \mathbb{q}}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t } ] \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}$ ] is the conditional entropy of @xmath37 $ ] .
if the history of signal @xmath33 has perfect information on the history of @xmath32 , the upper bound reaches @xmath195 . as in the bet - hediging situation , maximization of the average fitness @xmath191 with sensing
is also dual to the minimization of the relative entropy @xmath190 : @xmath196 where we use @xmath197 to denote the optimal @xmath102 and @xmath103 with sensing .
the duality indicates that @xmath191 achieves the bound @xmath192 only when @xmath198={\mathbb{q}}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]$ ] holds .
as in the bet - heding , if @xmath198={\mathbb{q}}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]$ ] holds , the time backward path probability @xmath199 $ ] equals the time - forward path probability : @xmath200 = { \mathbb{p}}_{{k}}^{*}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}^{*}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t}].\label{eq : cd_common1}\end{aligned}\ ] ] marginalization of this equation leads to the consistency condition @xmath201={\mathbb{p}}_{b}^{{s},*}[{\mathpzc{x}}_{t}|{\mathpzc{z}}_{t } ] ( { : = } \sum_{{\mathpzc{y}}_{t}}{\mathbb{p}}_{b}^{{s},*}[{\mathpzc{x}}_{t}|,{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}])$ ] derived in reference@xcite .
moreover , @xmath202{\mathbb{q}}[{\mathpzc{z}}_{t}]$ ] is the conjugate environment and signal of a given pair of strategies @xmath90 under which it achieves the bound . from the detailed fr in eq .
( [ eq : dfc1 ] ) , we also see that the fitness of a given strategy can exceed the bound by chance as @xmath51>{\psi}_{0}[{\mathpzc{y}}_{t}]+i[{\mathpzc{y}}_{t } ; { \mathpzc{z}}_{t}]$ ] when the realized pair of environmental and signal histories @xmath203 appears more frequently in the conjugate environment than in the actual environment as @xmath202{\mathbb{q}}[{\mathpzc{z}}_{t } ] > { \mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]$ ] .
the necessary and sufficient condition @xmath198={\mathbb{q}}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]$ ] for achieving the bound means that the optimal metabolic allocation and phenotype switching strategy together implement the bayesian computation of the posterior distribution @xmath204 $ ] of @xmath32 given the history of the sensed signal @xmath33 . under the constraint that @xmath202
$ ] satisfies a causal relation as @xmath202=\sum_{{\mathpzc{x}}_{t}}{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]$ ] , @xmath198={\mathbb{q}}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]$ ] can not be realized in general because @xmath204 $ ] does not necessarily satisfy the causality relation between @xmath32 and @xmath33 .
if , for example , the metabolic allocation strategy is of kelly s type as @xmath205 , the phenotypic switching strategy must satisfy @xmath206={\mathbb{q}}[{\mathpzc{x}}_{t}|{\mathpzc{z}}_{t}]$ ] to achieve the bound where @xmath207{:=}\left.{\mathbb{q}}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]\right|_{{\mathpzc{y}}_{t}={\mathpzc{x}}_{t}}$ ] . by definition ,
@xmath27 $ ] should satisfy the causal relation that @xmath208 depends only on the past and/or current states of @xmath24 .
however , @xmath204 $ ] may not necessarily be causal because the conditioning by histories biases the past states of the conditioned history .
for example , consider the case that the signal @xmath33 is causally generated from the environment @xmath32 and there is no feedback from @xmath33 to @xmath32 as in fig.2(a ) and ( c ) . if we observe @xmath209 , we may infer that the environmental state changes from @xmath110 to @xmath111 at time @xmath3 .
however , if we further observe @xmath210 , then we may change our prediction such that @xmath211 and @xmath212 was simply an error of the signal .
this intuitive observation illustrates that the inferred state of @xmath25 can be affected by the future observation of the signal , e.g. , @xmath213 and @xmath214 . for exactly the same reason , in
@xmath204 $ ] , the past state in the environmental history @xmath32 is affected by the future observations of the signal in @xmath33 .
because of the dependency on future state of the signal , @xmath204 $ ] can not be represented causally in general .
an exception is that the environmental history is causally and memorylessly generated from the signal as @xmath215=\prod_{\tau=0}^{t}{\mathbb{t}}_{e}(y_{\tau}|z_{\tau})$ ] , which is not realistic because we usually expect the signal to be generated from environment and not _
vice versa_. thus , @xmath216 is not generally zero , and @xmath190 contains not only the loss due to suboptimality of strategies , but also the loss from the causal constraints in exploiting the information of @xmath33 for phenotype switching .
causal dependency of environmental and signal histories .
( a ) a causal relation between environment and signal , in which the environment causally drives the signal .
black solid arrows indicate that @xmath217 is affected by its one - step past state @xmath30 and that of the signal @xmath218 .
blue dashed arrows represent that @xmath10 is affected by the current environment @xmath217 and the one - step past signal @xmath218 .
( b ) the causal decomposition of @xmath37 $ ] with the reversed causal relation between the environment and the signal , in which the signal causally drives the environment .
black solid arrows indicate that @xmath10 is affected by its one - step past environment @xmath30 and that of the signal @xmath218 .
blue dashed arrows represent that @xmath217 is affected by the current and one - step past signal @xmath10 and @xmath218 and the one - step past environment @xmath30 .
( c ) the causal relation between the environment and the signal described in figs .
2 ( a ) and ( c ) , and used in the numerical simulation ( figs . 3 and 5 ) . ]
by taking account of the causality relation between @xmath32 and @xmath33 in more detail , we can obtain tighter frs , which illustrate the problem of the causality more explicitly .
let us additionally assume that @xmath37 $ ] has a causal and markov relation as depicted in fig .
the relation is represented as @xmath219 & = \left[\prod_{\tau=0}^{t-1}{\mathbb{t}}_{e}^{f}(z_{\tau+1}|y_{\tau+1},z_{\tau}){\mathbb{t}}_{e}^{f}(y_{\tau+1}|y_{\tau},z_{\tau})\right]q(y_{0},z_{0}),\notag \\ & { : = } { \mathbb{q}}_{f}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t}]{\mathbb{q}}_{f}[{\mathpzc{y}}_{t}||{\mathpzc{z}}_{t-1}],\label{eq : qcausal}\end{aligned}\ ] ] where @xmath220 { : = } \prod_{\tau=0}^{t-1}{\mathbb{t}}_{e}^{f}(z_{\tau+1}|y_{\tau+1},z_{\tau})q(z_{0}|y_{0})$ ] and @xmath221 { : = } \prod_{\tau=0}^{t-1}{\mathbb{t}}_{e}^{f}(y_{\tau+1}|y_{\tau},z_{\tau})q(y_{0})$ ] . here , the signal @xmath24 and environment @xmath25 are markovean and dependent on the past states of the environment and the signal . by applying bayes theorem ,
@xmath37 $ ] admits another causal decomposition ( fig .
4 ( b ) ) with reversed causality between @xmath32 and @xmath33 : @xmath219 = & \left[\prod_{t=0}^{t-1}{\mathbb{t}}_{e}^{b}(y_{t+1}|z_{t+1 } , y_{t } , z_{t})\right]\left[\prod_{t=0}^{t-1}{\mathbb{t}}_{e}^{b}(z_{t+1}|y_{t},z_{t})\right],\notag \\ & \times q(y_{0},z_{0})\notag \\ { : = } & { \mathbb{q}}_{b}[{\mathpzc{y}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1 } ] \label{eq : qdecombb},\end{aligned}\ ] ] where @xmath222 @xmath223{:=}\prod_{t=0}^{t-1}{\mathbb{t}}_{e}^{b}(y_{t+1}|z_{t+1 } , y_{t } , z_{t } ) q(y_{0}|z_{0})$ ] , and @xmath224{:=}\prod_{t=0}^{t-1}{\mathbb{t}}_{e}^{b}(z_{t+1}|y_{t},z_{t})q(z_{0})$ ] .
because of the meaning of bayes theorem , @xmath225 is interpreted as the posterior distribution of @xmath217 inferred from observations , @xmath10 , @xmath30 , and @xmath218 .
thus , @xmath223 $ ] is equivalent to the posterior path probability of the environment inferred by the sequential bayesian inference of @xmath217 given the observation of the signal @xmath218 and the past state of the environment @xmath30 .
note that @xmath225 is slightly different from the usual sequential bayesean inference to estimate the hidden state @xmath25 only from observation @xmath33 in which we can not use the past environmental state @xmath30 for inference .
moreover , @xmath223 $ ] is also different from @xmath226 $ ] because the posterior path probability @xmath226 $ ] is computed after observing the whole history of @xmath33 rather than by conducting bayesian inference sequentially as in @xmath223 $ ]
. equation ( [ eq : qdecombb ] ) yields a tighter decomposition of the fitness : @xmath39 & = { \psi}_{0}[{\mathpzc{y}}_{t } ] + i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t } ] \notag \\ & -\ln \frac{{\mathbb{p}}_{b}^{{s}}[{\mathpzc{x}}_{t } , { \mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}{{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1}]}.\end{aligned}\ ] ] where @xmath227 { : = } \ln { \mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]/{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1}]{\mathbb{q}}[{\mathpzc{y}}_{t}]$ ] is the bare directed information from @xmath33 to @xmath32@xcite .
this decomposition leads to the following detailed fr : @xmath228+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]-{\psi}^{{s}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t } ] ) } \notag \\ & \qquad \qquad = \frac{{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1}]}{{\mathbb{p}}_{b}^{{s}}[{\mathpzc{x}}_{t } , { \mathpzc{y}}_{t},{\mathpzc{z}}_{t}]},\notag \\ & \qquad \qquad = \frac{{\mathbb{p}}_{{k},f}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1}]}{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}.\label{eq : dfrcsens}\end{aligned}\ ] ] because @xmath70{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1}]$ ] forms a joint path probability among @xmath12 , @xmath32 , and @xmath33 , which satisfies @xmath229{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1}]=1 $ ] , we obtain the integral and average frs as @xmath93+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]-{\psi}^{{s}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t } ] ) } \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t } ] } = 1 , \label{eq : ifrcsens}\end{aligned}\ ] ] and @xmath230 where @xmath231\big|\big|{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1}]\big]},\notag \\ & = { { \mathcal{d}}\big[{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]\big|\big|{\mathbb{p}}_{{k},f}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1}]\big]}\end{aligned}\ ] ] and @xmath232\big|\big|{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1}]{\mathbb{q}}[{\mathpzc{y}}_{t}]\big]}$ ] .
@xmath233 is the directed information that quantifies the amount of information @xmath33 has for inferring @xmath32@xcite .
if @xmath223={\mathbb{q}}[{\mathpzc{y}}_{t}]$ ] , @xmath234 as well as @xmath227=0 $ ] hold , which means that @xmath33 is useless for inferring @xmath32 causally .
@xmath235 in eq .
( [ eq : afrcsens ] ) is a tighter bound for the average fitness with sensing because @xmath236}{{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1}]{\mathbb{q}}[{\mathpzc{y}}_{t}]}+\ln\frac{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}{{\mathbb{q}}_{b}[{\mathpzc{y}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t } ] } \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}\notag \\ & = { \mathcal{i}}^{{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t } } + { \mathcal{i}}^{{\mathpzc{y}}_{t } \to { \mathpzc{z}}_{t } } \ge { \mathcal{i}}^{{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}}\end{aligned}\ ] ] holds , where @xmath237\big|\big|{\mathbb{q}}_{b}[{\mathpzc{y}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t}]\big]}$ ] is the directed information from @xmath32 to @xmath33 .
moreover , @xmath238 is the residual information in @xmath239 that has no fitness value under the causal constraints . to demonstrate the validity of frs in eqs .
( [ eq : dfrcsens])-([eq : afrcsens ] ) , we conduct a modified numerical simulation of the population dynamics to incorporate sensing ( see also appendix [ ap : num ] )
. the stochastic laws of environmental dynamics @xmath240
( fig.2 ( a ) ) and the metabolic allocation strategy @xmath241 ( fig.2 ( b ) ) are the same as those of the simulations in fig .
3 . the sensing signal is assumed to have two states , @xmath126 and @xmath127 ( fig .
2 ( c ) ) .
if the environment is either in @xmath110 or in @xmath111 , the signal gives the corresponding @xmath126 and @xmath127 with @xmath128 accuracy .
if the environment is in @xmath112 , the signal produces either @xmath126 or @xmath127 with eqaul probability ( fig .
2 ( c ) ) . because the signal is memorylessly generated from the environment at every time point , the environment and signal together form a markov relation as depicted in fig .
by employing the sensed signal , the organisms switch to the phenotypic state @xmath120 with a @xmath129 chance when the sensed signal is @xmath126 and to @xmath122 when the signal is @xmath127 as shown in fig .
2 ( d ) . figures 5 ( a ) and ( b ) show two trajectories of the population size under the same realizations of environment as in figs .
3 ( a ) and ( b ) , respectively . in fig . 5 ( a )
, the total population size @xmath242}$ ] ( dashed line ) is less than @xmath138+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]}$ ] ( red line ) whereas @xmath243 frequently becomes greater than @xmath138+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]}$ ] in fig . 5 ( b ) .
even with sensing , both @xmath51 $ ] and @xmath83+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]$ ] fluctuate substantially as shown in figs . 5 ( c ) and ( d ) depending on the realization of the environment and signal . due to this fluctuation ,
@xmath51 $ ] sometimes become greater than @xmath138+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]}$ ] as in fig . 5 ( e ) to satisfy the integral fr ( [ eq : ifrcsens ] ) as demonstrated in fig . 5 ( f ) .
numerical simulation of the population dynamics with sensing defined in fig .
( a , b ) two independent sample histories of environment @xmath32 and signal @xmath33 .
the dynamic of @xmath35 , @xmath244}$ ] , and @xmath245+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]}$ ] under the two realizations of the environmental and signal histories obtained by solving eq .
( [ eq : popdyn ] ) .
the color bars on the graphs represent the states of the environment and the signal at each time point .
the correspondence of the colors and environmental and signal states is as shown in fig .
2 ( a ) and ( c ) . cyan and yellow lines represent @xmath246 and @xmath247 , respectively .
the dashed blue line with filled grey style is @xmath243 .
the red line is @xmath248+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]}$ ] .
( c ) stochastic behaviors of population fitness @xmath38 $ ] for @xmath145 independent samples of environmental and signal histories .
each colored thin line represents @xmath249}={\mathcal{n}}_{t}^{{\mathpzc{y}},{\mathpzc{z}}}/{\mathcal{n}}_{0}^{{\mathpzc{y}},{\mathpzc{z}}}$ ] for each realization of the environmental and signal histories .
the thick black line is @xmath250 \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}}$ ] .
( d ) stochastic behaviors of @xmath83+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]$ ] for the same @xmath145 samples of environmental and signal histories as those in ( c ) . each colored thin line represents @xmath138+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]}$ ] for each realization of the same environmental and signal histories as in ( c ) .
( e ) stochastic behaviors of @xmath147+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]-{\psi}^{s}[{\mathpzc{y}}_{t}])}$ ] for the same @xmath145 samples of environmental and signal histories as those in ( c ) .
each colored thin line represents @xmath147+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]-{\psi}^{s}[{\mathpzc{y}}_{t}])}$ ] for each realization of the same environmental and signal histories as in ( c ) .
( f ) @xmath148+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]-{\psi}^{s}[{\mathpzc{y}}_{t } ] ) } \right>}_{{\mathbb{q}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]}$ ] calculated empirically by the numerical simulations of sample paths of @xmath147+i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t}]-{\psi}^{s}[{\mathpzc{y}}_{t}])}$ ] .
the meaning of the thick and thin lines are the same as in fig .
3 . ] the bound in eq .
( [ eq : afrcsens ] ) can be achieved when @xmath251={\mathbb{q}}_{b}[{\mathpzc{y}}_{t}||{\mathpzc{z}}_{t}]$ ] holds , which also means that @xmath252{:=}{\mathbb{q}}^{*b}[{\mathpzc{y}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}_{b}[{\mathpzc{z}}_{t}||{\mathpzc{y}}_{t-1}]$ ] is the conjugate environment for a given @xmath161 where @xmath253={\mathbb{p}}_{{k } , f}[{\mathpzc{y}}_{t}|{\mathpzc{z}}_{t}]$ ] . because @xmath223 $ ] is the path obtained by a type of sequential bayesian inference @xmath254 , the bound is attained if phenotype switching with sensing and metabolic allocation , as a whole , implement the same bayesian computation as @xmath225 .
however , to estimate @xmath217 , the bayesian inference represented by @xmath254 uses not only @xmath10 and @xmath218 but also the error - less information on the past environmental state @xmath30 . if organisms do not have any way to obtain perfect information on the past environment by sensing as they do not for biologically realistic situations , the metabolic allocation strategy should be of kelly s type as @xmath255 to achieve the bound without additional constraints on the environment . with kelly s strategy
, the organisms can effectively obtain information on the past environment because the past phenotypic state @xmath208 of the survived organisms becomes the same as the past environmental state @xmath25 under this situation . the optimal phenotype switching with kelly s strategy for metabolic allocation
should be @xmath256 to attain the bound .
another situation in which the bound can be achieved is when the environmental state @xmath217 depends not on @xmath30 but only on @xmath218 as @xmath257 .
we then have @xmath258 where @xmath225 is reduced to be independent of @xmath30 as @xmath259 .
thus , the bound can be attained if the phenotypic switching and metabolic allocation strategies satisfy @xmath260 note that the phenotypic switching need not be dependent on the past phenotypic state in this situation because @xmath213 and @xmath24 contain all the relevant information on the future state of @xmath261 . for both cases
, the achievability of the bound is directly linked to the accessibility of the organisms to the past information that directly drives the current environmental state .
in addition , even when the bound is not achieved , @xmath262 explicitly represents the loss of fitness associated with the inaccessibility . for general metabolic allocation strategies
other than kelly s type , the phenotypic states of the survived organisms contain only imperfect information on the past environmental history .
in addition , the environment and signal may not always form the circular dependency @xmath263 .
in such situations , the loss is most generally represented by @xmath190 , which quantifies the difference between the time - backward path probability @xmath264 $ ] and the time - forward path probability @xmath70{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]{\mathbb{q}}[{\mathpzc{z}}_{t}]$ ] .
the loss from different sources can be represented by using the following dissection of the fitness : @xmath39 & = { \psi}_{0}[{\mathpzc{y}}_{t } ] + i[{\mathpzc{y}}_{t } , { \mathpzc{z}}_{t } ] -i_{b}^{{s}}[{\mathpzc{y}}_{t } , { \mathpzc{z}}_{t}|{\mathpzc{x}}_{t } ] \\ & -\ln \frac{{\mathbb{p}}_{b}^{{s}}[{\mathpzc{y}}_{t}|{\mathpzc{x}}_{t}]}{{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]}- \ln\frac{{\mathbb{p}}_{b}^{{s}}[{\mathpzc{x}}_{t}|{\mathpzc{z}}_{t}]}{{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]},\label{eq : decomretc}\end{aligned}\ ] ] where @xmath265 { : = } \ln { \mathbb{p}}_{b}^{{s}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}|{\mathpzc{x}}_{t}]/{\mathbb{p}}_{b}^{{s}}[{\mathpzc{y}}_{t}|{\mathpzc{x}}_{t}]{\mathbb{p}}_{b}^{{s}}[{\mathpzc{z}}_{t}|{\mathpzc{x}}_{t}]$ ] is the bare conditional backward mutual information . although this representation does not immediately admit detailed and integral frs
, we can obtain the following decomposition of @xmath190 into three terms : @xmath266 where @xmath267 \right>}_{{\mathbb{p}}_{b}[{\mathpzc{x}}_{t},{\mathpzc{y}}_{t},{\mathpzc{z}}_{t}]},\\ { \mathcal{d}}_{{\mathrm{loss}}}^{{s},{k}}{:=}&{\left < { { \mathcal{d}}\big[{\mathbb{p}}_{b}^{{s}}[{\mathpzc{y}}_{t}|{\mathpzc{x}}_{t}]\big|\big|{\mathbb{p}}_{{k}}[{\mathpzc{y}}_{t}||{\mathpzc{x}}_{t}]\big ] } \right>}_{{\mathbb{p}}_{b}^{{s}}[{\mathpzc{x}}_{t } ] } , \\ { \mathcal{d}}_{{\mathrm{loss}}}^{{s},f}{:=}&{\left < { { \mathcal{d}}\big[{\mathbb{p}}_{b}^{{s}}[{\mathpzc{x}}_{t}|{\mathpzc{z}}_{t}]\big|\big|{\mathbb{p}}_{f}[{\mathpzc{x}}_{t}||{\mathpzc{z}}_{t}]\big ] } \right>}_{{\mathbb{q}}[{\mathpzc{z}}_{t}]}.\end{aligned}\ ] ] @xmath268 is the residual information on the environment that the signal still has even if we know the retrospective history of the phenotype .
because the retrospective correlation between the phenotype and environment is induced by the selection , this residual information represents the information that is not used in the selection .
if the signal contains certain information on the environmental fluctuation that is nothing to do with the replication of the organisms , for example , the information on the existence of unmetabolizable artificial molecules , such information can not be exploited in the phenotypic switching to choose a better phenotype for the current environmental state .
thus , @xmath268 measures the amount of such useless information for fitness that the sensing signal conveys .
@xmath269 accounts for the imperfectness of the metabolic allocation strategy , and @xmath270 quantifies the imperfectness of phenotype switching due to the causal constraint and suboptimality . for example , in the case of the causally optimal strategy under kelly s metabolic allocation in eq .
( [ eq : coptkelly ] ) with the causal structure in @xmath37 $ ] ( eq .
( [ eq : qcausal ] ) ) , these quantities are reduced to @xmath271 the first equation , @xmath272 , holds because organisms with the metabolic allocation strategy of kelly s type can survive only when their phenotypic history is identical to the actual environmental history , and the retrospective phenotypic history @xmath12 contains perfect information on @xmath32 .
thus , the residual @xmath268 becomes @xmath151 .
similarly , @xmath273 because only organisms with their phenotypic history identical to their environmental hisotry can survive , and @xmath274 = \delta_{{\mathpzc{x}}_{t},{\mathpzc{y}}_{t}}$ ] holds for kelly s strategy .
in contrast , @xmath270 can not be zero because the phenotypic switching strategy is causally constrained .
the minimum loss due to this constraint is the directed information @xmath238 that measures the causally useless information in @xmath239 . in a general situation ,
these three quantities are mutually related .
for example , the amount of useless information for fitness @xmath268 depends on the choice of strategies .
if phenotype switching does not use any information obtained from @xmath24 , then @xmath268 becomes @xmath275 .
they are appropriate information - theoretic quantities that account for the irrelevance of sensing and the imperfectness of metabolic allocation and phenotypic switching for exploiting the sensed information .
in this paper , we clarified the stochastic and information thermodynamic structures in population dynamics with and without sensing , to derive the bound of fitness and the fitness gain by sensing , in the form of frs .
the detailed and integral frs obtained substantially generalized the previous results on the average fitness value of information by showing that not only average but also the fluctuation of fitness is generically constrained as is the entropy production .
such constraints manifest the possibility and the condition that the fitness of organisms with suboptimal strategies can be greater than that of the optimal one by chance due to environmental fluctuations , just as there is a finite probability of observing reversal of heat flow in a small thermal system .
such rare events can be regarded as a violation of ess because they may induce the takeover of the optimal organism by a suboptimal one in a finite population .
nevertheless , the violation is somehow ruled to follow the integral and detailed frs .
moreover , the directed information is derived to be the tighter measure of the fitness value of information , in which the causal structure in the problem is explicitly accounted for .
the condition for achieving this bound is shown to be related to the ability of the organisms to conduct or implement a type of bayesian sequential inference from the sensed information . finally , we derived three quantities that can measure the irrelevance of sensing and the imperfectness of metabolic allocation and phenotypic switching for exploiting the sensed information .
all these results and generalizations are derived by employing the path - wise and the time - backward representation of population dynamics . among others , pivotal
is the duality between the maximization of the average fitness and the minimization of the difference between the time - forward and time - backward path probabilities ( eqs .
( [ eq : optb ] ) and ( [ eq : optw ] ) ) .
the minimum loss of the average fitness due to causality is clearly described in this dual problem as the deviation between the causal time - forward path probability and the non - causal time - backward path probability ( eq .
( [ eq : dsloss ] ) ) .
we believe that the path - wise and retrospective formulation of the population dynamics also play indispensable roles in addressing other biologically relevant problems . in this paper as well as in most of the previous works ,
the processes of attaining better strategies by mutation or learning are rarely considered directly and explicitly@xcite , except in a reference@xcite .
as we have demonstrated , the average optimal strategy may not dominate a population all the time in a fluctuating environment because of a rare event : the takeover by suboptimal strategies .
thus , the organisms and their corresponding strategies are expected to change non - stationarily over time .
in such a situation , an advantage may be gained by organisms that acquire an ability to adaptively change and learn the strategies as our brain and immune system do@xcite .
however , the fitness to be increased by adaptation is a population - level quantity whereas adaptation is conducted at the individual - level .
the duality might be used to resolve this problem . as demonstrated in kelly s strategy in the most extreme fashion
, the survived organisms can retrospectively obtain certain information on the environment they experienced .
as shown recently@xcite , such information can be used to adjust time - forward strategies to switch phenotype and to allocate resource so as to reduce the discrepancy between time - forward and backward phenotypic histories rather than directly increasing the population fitness . by employing our path - wise formulation
, we may obtain more general results on the adaptive learning of strategies and its evolutionary advantages .
another problem that should be addressed is the type of sensing . in this paper ,
the sensing signal is treated as an extrinsic factor by assuming that all organisms have the same sensing signal .
such an assumption is valid only when the sensing noise of the organisms is negligibly small .
a more general and biologically realistic situation is the individual sensing in which each organism receives different sensing signals owing to the sensing noise intrinsic to the organism@xcite .
the fitness value of individual sensing has been rarely addressed , with the exception of a pioneering work@xcite , which shows that the fitness value of individual sensing can be greater than the mutual information under restricted situations@xcite .
a generalization of this result may be achieved by using our pat - hwise and retrospective formulation .
as we have clarified , thermodynamics and adaptation share the same fundamental mathematical structure . nevertheless , most attempts to bridge thermodynamics and adaptation or evolution , including ours , are just formal in the sense that the similarity shown is at the level of mathematics @xcite .
however , an actual thermodynamics underlies the processes of replication and sensing of organisms@xcite .
the thermodynamics must constrain the rate of replication and the efficiency of sensing , and the latter was investigated intensively in the context of stochastic and information thermodynamics , recently@xcite . our relation between fitness and information
must be consistent with the constraints imposed by the thermodynamics of these processes .
integration of the relation between entropy and information with that between fitness and information would be an indispensable step toward establishing the real thermodynamics of adaptation and evolution .
we acknowledge yuichi wakamoto , takashi nozoe , and yoichiro takahashi for useful discussions . this research is supported partially by an jst presto program , jsps kakenhi 16h06155 and 16k17763 , the platform for dynamic approaches to living system from mext and amed , japan , and the 2016 inamori research grants program , japan .
let us consider a given replication rate @xmath19 .
@xmath21 determines a @xmath276 matrix , and we define its row vectors by @xmath277 . when the number of the environmental states is equal to or greater than the phenotypic ones as @xmath278 , @xmath279 can form a hyperplane , and we can fine a vector @xmath280 that is orthogonal to the plane .
let @xmath281 be the normalization of @xmath282 as @xmath283 . by definition
, @xmath281 satisfies @xmath284 for all @xmath285 where @xmath286 is a constant .
this means that we can define a conditional probability @xmath17 as @xmath287 thus , we have the decomposition of @xmath21 in eq .
( [ eq : ek ] ) and its variant used in eq .
( [ eq : decombk ] ) as @xmath288 where @xmath289 .
thus , when the environmental states are more complex than the phenotypic ones as @xmath278 , the decomposition of the replication rate , eq .
( [ eq : ek ] ) , is general enough even if we do not explicitly assume the relation between metabolic allocation strategy and replication rate as in eq .
( [ eq : ek ] ) .
if @xmath290 , the decomposition does not necessarily exist .
such a situation may occur when an organism has redundant phenotypic states , a fraction of which can be effectively realized by linear combinations of the others . while the redundant phenotypic states can appear in the process of mutation ,
@xmath290 is not biologically realistic because the environment is generally more complex than phenotype .
finally , we note that the decomposition , eq .
( [ eq : decompk ] ) , is not unique except when @xmath22 and @xmath279 are linearly independent each others .
nevertheless , our frs hold irrespective of the choice of the decomposition .
achievability of the maximum average fitness , eq . ( [ eq : afrbbound ] ) , is affected by the choice because @xmath84 depends on the decomposition and because @xmath291 must be changed so that it satisfies eq .
( [ eq : decompk ] ) for a given @xmath19 . while we apparently have a freedom to choose the metabolic allocation strategy , @xmath17 , for maximization of @xmath100 in eq .
( [ eq : afrbbound ] ) , eq .
( [ eq : afrbbound ] ) is virtually reduced to the problem of the maximization of the average fitness by changing only phenotypic switching strategy @xmath154 as @xmath292 because @xmath293 is independent of the choice of the decomposition of @xmath21 .
this is the problem addressed in our previous work@xcite .
tight upper bound of @xmath294 can be obtained by minimization of @xmath105 as @xmath295 we again stress that our integral and detailed frs generally hold for any given @xmath102 and @xmath103 .
for the numerical simulations shown in figs 3 and 5 , we consider the case that the environmental dynamics is markovean as @xmath116 , and that the signal is memorylessly generated from the environment as @xmath296 .
realizations of the environmental history @xmath32 and the sensing signal @xmath33 are obtained by conducting a finite - state markov transition by following @xmath116 and @xmath297 . for a given pair of realizations @xmath203 ,
the fitnesses @xmath59 $ ] ( without sensing ) and @xmath298 $ ] ( with sensing ) are calculated recursively as @xmath299&={\psi}^{b}[{\mathpzc{y}}_{t } ] + \ln \frac{{\mathcal{n}}^{{\mathpzc{y}}}_{t+1}}{{\mathcal{n}}^{{\mathpzc{y}}}_{t}},\\ { \psi}^{{s}}[{\mathpzc{y}}_{t+1},{\mathpzc{z}}_{t+1}]&={\psi}^{{s}}[{\mathpzc{y}}_{t},{\mathpzc{z}}_{t } ] + \ln \frac{{\mathcal{n}}^{{\mathpzc{y}},{\mathpzc{z}}}_{t+1}}{{\mathcal{n}}^{{\mathpzc{y}},{\mathpzc{z}}}_{t}},\end{aligned}\ ] ] where @xmath300 and @xmath301 are obtained by solving eq .
( [ eq : popdyn ] ) recursively for the given @xmath203 .
similarly , @xmath83 $ ] is recursively computed as @xmath302={\psi}_{0}[{\mathpzc{y}}_{t } ] + \left(\phi_{0 } + \ln \frac{{\mathbb{t}}_{e}^{f}(y_{t+1}|y_{t})}{q_{0}(y_{t+1 } ) } \right).\end{aligned}\ ] ] the point - wise directed information @xmath227 $ ] in eq .
( [ eq : dfrcsens ] ) and eq .
( [ eq : ifrcsens ] ) is recursively obtained as @xmath303 = i[{\mathpzc{z}}_{t } \to { \mathpzc{y}}_{t } ] + \ln \frac{{\mathbb{t}}_{e}^{f}(z_{t+1}|y_{t+1})}{{\mathbb{t}}_{e}^{b}(z_{t+1}|y_{t})},\end{aligned}\ ] ] where @xmath304 .
note that , when a feedback relation exists between @xmath30 and @xmath218 , the computation of @xmath83 $ ] and @xmath227 $ ] becomes more demanding because @xmath75 $ ] has to be obtained by marginalizing @xmath33 in @xmath37 $ ] .
83ifxundefined [ 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 @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ , some theoretical explorations ( , ) @noop * * , ( ) @noop * * , ( ) @noop , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop , ( ) while the difference between phenotypic and environmental states is only @xmath305 , this does not affect our numerical verifications because our theory is proved for more general situations
. however , if we consider more environmental states , we need more samples to verify the fluctuation relations , numerically . @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) _ _ , @noop ph.d . thesis ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | adaptation in a fluctuating environment is a process of fueling environmental information to gain fitness .
living systems have gradually developed strategies for adaptation from random and passive diversification of the phenotype to more proactive decision making , in which environmental information is sensed and exploited more actively and effectively .
understanding the fundamental relation between fitness and information is therefore crucial to clarify the limits and universal properties of adaptation . in this work ,
we elucidate the underlying stochastic and information - thermodynamic structure in this process , by deriving causal fluctuation relations ( frs ) of fitness and information .
combined with a duality between phenotypic and environmental dynamics , the frs reveal the limit of fitness gain , the relation of time reversibility with the achievability of the limit , and the possibility and condition for gaining excess fitness due to environmental fluctuation .
the loss of fitness due to causal constraints and the limited capacity of real organisms is shown to be the difference between time - forward and time - backward path probabilities of phenotypic and environmental dynamics .
furthermore , the frs generalize the concept of evolutionary stable state ( ess ) for fluctuating environment by giving the probability that the optimal strategy on average can be invaded by a suboptimal one owing to rare environmental fluctuation .
these results clarify the information thermodynamic structures in adaptation and evolution . |
measurements of the gas temperature in clusters of galaxies is a potentially powerful tool for testing cosmological models since the gas temperature provides a direct measurement of the cluster mass .
the cluster mass function is directly traceable to the underlying cosmology and can be used to determine the normalization and slope of the initial mass fluctuation spectrum in the early universe ( peebles et al .
1989 , henry & arnaud 1991 , lilje 1992 , bartlett & silk 1993 , white et al .
1993 , colafrancesco & vittorio 1994 , hattori et al .
1995 , eke et al .
1995 , eke et al .
1996 , pen 1996 , viana & liddle 1996 , oukbir et al .
1997 , oukbir & blanchard 1997 , bahcall et al .
the usefulness of this method is limited by the uncertainty in the parameters used to fit the theory of cluster formation to the observed cluster temperature distribution function ( colafrancesco et al .
1997 ) . observations of very hot clusters can alleviate this problem to a certain extent .
the cluster temperature function is predicted to have an exponential cutoff beyond a critical temperature which depends on the cosmological model , so the differences between the theoretical predictions increase exponentially beyond the critical temperature .
here we report on optical and x - ray observations of a newly discovered cluster , 1e0657 - 56 .
the measured temperature @xmath4 makes it a strong candidate for the hottest known cluster .
the interpretation of this temperature is complicated by evidence that the cluster is in the process of merging with a subcluster ( see e.g. schindler & muller , 1993 , roettiger et al 1997 ) . since the subcluster has a mass on the order of one - tenth the mass of the primary cluster , we argue that the high temperature is a true indication of a large primary mass .
if this interpretation is correct , then the existence of such a cluster at the observed redshift @xmath0 is very difficult to reconcile with current theoretical models for a critical density ( @xmath5 ) universe .
this letter is organized in the following manner .
the optical and x - ray observations are presented and analyzed in @xmath6 , and the implications of the exceptionally high temperature of the hot gas in 1e0657 - 56 are considered in @xmath7 .
the cluster 1e0657 - 56 , was detected as an extended source on a scale of @xmath8 by the imaging proportional counter ( ipc ) onboard the einstein observatory .
the source remained unidentified until a ccd image of this source was obtained with the las campanas 2.5 meter telescope as part of a program to search for possible failed clusters of galaxies , i.e. , clusters with hot gas but no galaxies ( tucker , tananbaum & remillard 1995 ) .
the ccd image revealed the presence of a rich cluster of galaxies and a luminous arc about 12 arcsec long .
follow - up imaging and spectroscopic observations with the eso ntt telescope in december 1993 confirmed the presence of the arc and yielded a cluster redshift of @xmath9 .
the corresponding luminosity distance is @xmath10 gpc for @xmath11 , where @xmath12 is the hubble constant in units of 50 km s@xmath2 mpc@xmath2 .
we use our ntt spectroscopic data to calculate the cluster velocity dispersion .
the method of danese et al ( 1980 ) is used to include the effects of measurement errors . from the 24 measured redshifts we only include galaxies with velocities between 80,000 km s@xmath2 , and 100,000 km s@xmath2 , and within @xmath13 arcmin of the brightest cd galaxy in the cluster .
the resulting velocity dispersion is @xmath14 km s@xmath2 , based on 13 galaxies .
when we vary @xmath15 between 1 arcmin and 4 arcmin , we find @xmath16 varies from 830 to 1515 km s@xmath2 , respectively , and the corresponding number of galaxies varies from 7 to 15 . at the lower limit , the number of galaxies is too small to yield useful errors ; at the upper limit , the errors are comparable to those with @xmath17 arcmin .
the rosat hri observed 1e0657 - 56 for a total live time of 57.7 ksec between 8 february 1995 and 27 june 1995 .
a contour image of the background subtracted hri image is shown in figure 1 .
this figure shows that the x - ray emission is extended and double - peaked , indicating either a merger event or two spatially distinct groups .
the x - ray flux of the secondary peak , which is also an extended source , comprises only 4% of the total flux .
the surface brightness profile of the primary cluster was obtained by extracting the net counts in concentric annuli about the primary x - ray peak after excluding the western quadrant ( which contains the secondary peak ) and all point sources .
we then fit the observed x - ray surface brightness profile to the standard hydrostatic - isothermal @xmath18 model , @xmath19 with the addition of a flat background component .
the best fit is obtained with a core radius of @xmath20 = 62@xmath21 ( corresponding to @xmath22 kpc ) , @xmath23 , and a central electron number density @xmath24 @xmath25 .
integrating the best fit @xmath18 model gives a gas mass within @xmath26 mpc of @xmath27 . the net flux and luminosity of the primary cluster
is then determined by multiplying the flux from the 3 quadrants centered on the primary peak by 4/3 to account for the flux underlying the secondary peak .
using the gas temperature derived from the asca data gives @xmath28 erg @xmath29 s@xmath2 and @xmath30 erg s@xmath2 .
standard screening criteria were applied to the sis and gis data to exclude data taken during earth occultations , periods of high particle flux , and the portions of asca orbit that pass through the south atlantic anomaly .
hot and flickering pixels were removed with sisclean and gisclean .
the resulting screened exposure times are : sis0 ( 20.7 ksec ) ; sis1 ( 20.0 ksec ) ; gis2 ( 21.7 ksec ) ; and gis3 ( 21.7 ksec ) .
source spectra were extracted using a rectangular region of @xmath31 by @xmath32 for the sis0 data , a rectangular region of @xmath31 by @xmath33 for the sis1 data , and circular regions of radii @xmath34 and @xmath35 for the gis2 and gis3 data , respectively .
background spectra were taken from available blank sky observations .
all four background subtracted spectra are then simultaneously fit to a single - temperature raymond thermal plasma model . the best fit model ( @xmath36 for 778 spectral bins ) and residuals are shown in figure 2a .
the best fit model has a temperature of @xmath37 ( 90% confidence level ) and an fe abundance of @xmath38 solar ( @xmath39 ) .
a red - shifted 6.9 kev line , corresponding to the fexxvi @xmath40 line is also detected , but only at the 2.5 @xmath41 level .
figure 2b shows a comparison of the sis0 data from 1e0657 - 56 with the best fit single - temperature raymond model for a2163 ( @xmath42 kev ; markevitch et al 1996 ) .
1e0657 - 56 is clearly hotter than a2163 .
the best fit temperature is also greater than that of ms1054 - 53 , reported to have a temperature of @xmath43 kev ( donahue et al 1997 ) . at @xmath44 kev ,
1e0657 - 56 is to our knowledge the hottest known cluster of galaxies .
for a temperature of 17 kev and the velocity dispersion given above , we find @xmath45 , which agrees well with the value of @xmath18 determined from the fit to the hrihri surface brightness profile .
an absorbed power - law model ( photon index = 1.5@xmath46 0.05 ) also provides an acceptable fit to the asca data , with a reduced chi - squared of 0.932 .
however the rosat hri image shows that 1e0657 - 56 is extended in the 0.1 to 2.4 kev band , so a pure power law from an active galactic nucleus ( agn ) can not explain the entire 0.5 to 10 kev asca spectrum .
the fwhm of the asca / sis point response function is comparable to the extent of the cluster , so we can not determine if the hard emission outside of the pspc bandpass is extended . however , the centroids of the asca image in both the 1 - 4 kev and 4 - 10 kev bands are coincident with the centroid of the rosat image , supporting the assumption that the rosat and asca emission have the same origin , namely the hot cluster gas . assuming the gas is isothermal gives a total gravitating mass within @xmath47 mpc of @xmath48 , based on the equation of hydrostatic equilibrium , the best fit @xmath18 model , and the asca determined gas temperature .
this gives a gas mass fraction of @xmath49 , consistent with other rich clusters ( david , jones , & forman 1995 ; white & fabian 1995 ) .
the thermal energy content of the cluster gas within a radius of @xmath26 mpc is approximately @xmath50 ergs .
it is difficult to account for this energy content by any process other than gravitational collapse . assuming a thermal energy input of 10@xmath51 ergs per supernova , @xmath52 supernovae would be required to heat the gas .
assuming an equal mixture of type ia supernovae ( fe yield of @xmath53 ) and type ii supernovae ( fe yield of @xmath54 ) , the resulting iron abundance would be more than ten times the solar abundance and thirty times greater than implied by the asca data .
a quasar can generate @xmath55 ergs s@xmath2 of energy for @xmath56 years ; ten such quasars would be required to explain the thermal energy content of 1e0657 - 56 , but then the space density of quasars in the cluster would have to be @xmath57100 times greater than the cosmic average . the cluster is observed to be a radio source ; griffith and wright ( 1993 ) detected a 1.7mjy radio source at 4850 mhz .
inverse compton scattering of the microwave background off the relativistic electrons in the radio source could produce an extended , hard x - ray source .
the intensity of this source in the 0.1 - 10 kev band is of the order of numerical simulations of merging clusters ( e.g. roettiger , et al .
1997 , evrard et al .
1996 , schindler & muller , 1993 ) show that temperatures of order 20 kev can be achieved in mergers , a conclusion confirmed by the asca observations of abell 754 ( henriksen & markevitch 1996 ) .
however , the simulations and the observations both indicate that for mergers with high cluster / subcluster mass ratio , the high temperature regions are usually localized , and do not affect significantly the average cluster temperature . in a754 , the average temperature is @xmath60 kev , even though a temperature in the range @xmath61 kev is found in one region ( henriksen & markevitch 1996 ) .
this is in agreement with the numerical simulations of roettiger et al .
( 1997 ) for a cluster / subcluster mass ratio of 8:1 .
their calculations show little change of the emission weighted temperature , except for a peak at the time of the merger which is confined to the central 300 kpc for roughly a few hundred million years . for 1e0657 - 56 , we can estimate the mass ratio from the hri observation , which indicates that 4% of the flux comes from the secondary peak which occupies 10% of the volume , @xmath62 . since the luminosity @xmath63 , the mass ratio is @xmath57 16:1 , so the effect of the merger on our emission weighted temperature should not be large for this particular cluster .
one of the strongest correlations between the x - ray properties of clusters of galaxies is that between the gas temperature and x - ray luminosity ( david et al .
1993 ; mushotzky & scharf 1997 ) .
figure 3 shows that , within the errors , the temperature of 1e0657 - 56 is consistent with this relation , providing further support that the high temperature reflects a global property of the cluster , and is not due to a small scale temperature enhancement . in summary , the numerical simulations do illustrate the importance of having spatial as well as spectral information before using cluster temperatures as an indicator of virial masses , but in the present case we conclude that the unusually high temperature of 1e0657 is due to an unusually deep potential well and hence a large virial mass . every cluster that is discovered with a temperature greater than about 15 kev places an additional straw on the `` camel s back '' of cosmological theories for the formation of clusters of galaxies that require @xmath64 = 1 . the standard normalization for primordial density fluctuations
is derived from the number density of 5 - 8 kev clusters ( henry & arnaud 1991 , eke et al 1996 , pen 1996 , viana & liddle 1996 )
. differences between cosmological models are strongly amplified at temperatures of 17 kev .
for example , the expected number of clusters with a temperature greater than 17.4 kev , in a volume bounded by a redshift @xmath65 is @xmath66 for @xmath67 and @xmath68 for @xmath69 for a cdm spectrum normalized to fit the cluster number density at 5.5 kev ( u. pen , private communication ) .
the discovery of just one high temperature cluster thus effectively rules out models with a cold dark matter ( cdm ) type of power spectrum of fluctuations and @xmath70 .
other authors ( donahue et al .
1997 , luppino & gioia 1995 , ostriker 1993 , and references cited therein ) have noted that standard cdm models with @xmath70 can not explain the cobe measurements of cmbr temperature anisotropies ( smoot et al 1992 ) , as well as the abundances of clusters and galaxies . in the future
we can anticipate that the discovery and study of very hot gas in high redshift clusters of galaxies with asca and future observatories such as axaf and xmm will provide severe tests for cosmological models for the formation of large - scale structure in the universe .
we are grateful for the assistance of k. arnaud , p. gorenstein , u. pen , and r. mushotzky .
this research was supported by nasa contract nas8 - 39073 and the smithsonian institution .
we benefited from the use of the nasa extragalactic data base to locate the radio source identified with 1e0657 - 58 . | we identify the extended einstein ipc x - ray source , 1e0657 - 56 , with a previously unknown cluster of galaxies at a redshift of @xmath0 .
optical ccd images show the presence of a gravitational arc in this cluster and galaxy spectra yield a cluster velocity dispersion of @xmath1 km s@xmath2 .
x - ray data obtained with the rosat hri and asca indicate that 1e0657 - 56 is a highly luminous cluster in which a merger of subclusters may be occurring .
the temperature of the hot gas in 1e0657 - 56 is @xmath3 , which makes it an unusually hot cluster , with important cosmological implications . |
transition - edge sensor ( tes ) detectors are showing great promise as fast bolometer arrays in astronomical applications from the sub - millimeter through the x - rays ( wollman , _
et al._2000 ) . in the near - ir / optical / uv
range these devices offer good broad - band quantum efficiency ( qe ) , high time resolution and modest energy resolution and saturation count - rate ( cabrera _ et al._1998 , romani _
et al._1999 ) . along with competing cryogenic technologies ( e.g. superconducting tunnel junction devices , stjs ; perryman , foden & peacock 1993 , perryman _
et al._1999 ) these sensor arrays offer the potential for important new capabilities , particularly for the study of rapidly varying compact object systems .
we are developing sensitive tes spectrophotometer systems for such applications and report here on test observations of the crab pulsar with a prototype fiber - coupled array . the basic principles and present performance of thin - film tungsten ( w ) tes devices have been recently summarized in miller _
et al._(2000 ) . in brief
, the systems routinely achieve energy resolutions of @xmath1ev at @xmath2ev , photon arrival time resolution of @xmath3ns and single pixel count rates of @xmath4khz without serious pile - up problems . _ en route _ to a high sensitivity , general purpose camera suitable for faint object high speed spectrophotometry , we have scheduled a number of astronomical demonstrations .
observations at the mcdonald observatory 2.7 m harlan j. smith telescope were made in february 2000 with a system that incorporated a number of substantial improvements over the single pixel , fiber - coupled system used for the first astronomical observations described in romani , _
et al._1999 .
the detector system used in these observations employs a @xmath5 array of w tes pixels on a si substrate with a @xmath6 m pitch , fabricated by our group using the stanford nanofabrication facility . for ease of fabrication @xmath7 m wide
al leads , the ` voltage - bias rails ' , were connected in the pixel plane which means that the fill factor decreases towards the sides of the array as the active tungsten area is reduced to accommodate these leads ( figure 1 ) .
the rails , superconducting during operation , are covered by an extension of the active w thin film of the tes pixel , providing electrical connectivity .
rails and pixels are separated by @xmath7 m gaps . because of the underlying al , photons absorbed above these rails in fact couple a larger fraction of the the deposited energy to the w @xmath8 system than those absorbed directly on the adjacent tes .
these ` rail hits ' thus increase the array fill fraction , but produce a ` satellite peak ' in the energy psf , which complicates the spectral analysis ( 3 ) .
the data system used for these tests had six read channels so in practice a @xmath9 sub - array along the midline was used .
these uniform @xmath10m@xmath11 m pixels gave an effective fill factor of 96% .
the geometrical peculiarities of the arrays used in these measurements are not intrinsic to the tes system .
new arrays have masks to eliminate the rail events .
buried wiring and focusing collimators have been designed that can further maintain uniform pixel size and high effective fill factors over substantially larger tes arrays .
configuration are not wired and serve only to mask the substrate . ] in operation , the si substrate is cooled well below the @xmath12mk w transition temperature , while the tes sensor is biased in a circuit with a fixed voltage . the resulting current allows the w electron system to self - heat into the middle of the superconducting - normal transition .
the tes sensors are thus operating in a regime of strong negative electrothermal feedback ( irwin 1995 ) .
a major advance of the present apparatus is the first implementation of a nist - developed digital feedback system which simultaneously linearizes the squid ammeters and functions as the data acquisition system .
the digitized feedback signal monitors the tes device current .
the data stream is processed through an fpga with adjustable peak shaping and baseline - restore algorithms .
in addition to accurate gps - stamped arrival times and pulse height measurements for each photon , the system provides a number of quality control indicators including adjustable pile - up flags .
four of the pixels were read with independent digital channels of this new design ; these provide the photon set analyzed in this paper .
this system is directly scalable to large numbers of read channels and can be modified to allow multiplex reading of several pixels per read channel ( chervenak _ et al._1999 ) .
there are two serious challenges in coupling these tes detectors to a typical ground - based astronomical telescope .
one is to filter the incident beam so that the large flux of thermal photons ( at @xmath13 m ) from the warm optics does not saturate the tes sensor system , which has a modest maximum count rate .
the second challenge is to maximally couple the beam of a large telescope to our present small ( effectively 46@xmath14 m ) detector array . in the system
described here we used a focal reducing train of fiber and imaging optics ( figure 2 ) .
given the relatively large plate scale ( @xmath15m / arcsec at the f/17.7 cassegrain focus of the 2.7 m telescope ) and best expected image of 1 - 1.5@xmath16 fwhm we adopted a 400@xmath14 m ( 1.7 arcsec ) entrance aperture and fed the light through a 2:1 reducing taper .
this taper , along with the @xmath17 m run of 200@xmath14m - core fiber to the refrigerator bulkhead , were fabricated from low oh , low impurity ( polymicro fsu ) si / si step - index fibers , providing both good ir and uv transmission . at the bulkhead of the cryostat , a portable @xmath18he/@xmath19he dilution refrigerator , an st - connector vacuum feedthrough passed the light to a @xmath20 m length of high oh ` wet ' 200@xmath14 m core fiber .
the bulk of this fiber was spooled in thermal contact with the 1k stage of the cryostat . in this way
the strong oh absorption bands of the wet fiber ( humbach 1996 ) acted to provide a cold filter with effective blocking beyond 1.7@xmath14 m and several strong shorter wavelength absorption bands ( e.g. at 0.95@xmath14 m ) .
these absorptions correspond to wavelengths of strong atmospheric oh emission .
the system thus provided transmission in the astronomically interesting j and h bands as well as reasonable throughput to 400 nm where scattering in the fibers began to compromise transmission . to illuminate the sensor array at the cold stage
, we collimated the fiber beam with a 1.8 mm 1/4 pitch gradient index ( grin ) lens and then focused with a 1.0 mm diameter spherical lens to provide the best coupling to the central @xmath21-pixel section of the tes array .
despite inevitable focal ratio degradation in the fiber train , the final @xmath22 focus with modest chromatic aberration ensured that we had reasonable coupling into our central 160@xmath14m@xmath23 of active w tes .
we measured the absolute quantum efficiency of the system from the cryostat bulkhead to final recorded photons , including the @xmath24 absorptivity of the grey tungsten surface , using a calibrated 850 nm light source , calibrated attenuators and an absolute power meter .
these measurements determined a total 850 nm qe of 0.097 .
this was confirmed by comparison with ccd integrations , using cameras of known qe .
the ` rail ' hit photons contribute an additional 0.01 to the system qe .
inevitably , the fiber train to the telescope incurs additional losses .
unfortunately , these were exacerbated by a defect in the st couplers of the taper section . as a consequence ,
the fiber section imposed a factor of @xmath25 decrease in the peak throughput , relative to the back end of the telescope .
the energy scale for each pixel was set by interrupting the observations every few hours to take short @xmath26s exposures of a calibration lamp passed through a grating monochrometer and fiber fed to the tes array .
typically we integrated with the 1@xmath27 order set at @xmath280.5ev , 1.0ev and 2.0ev .
observations of the order peaks provided a quick calibration of the system energy scale and non - linearity .
in addition , to monitor any slow drifts in the energy gain as might occur due to baseline temperature drifts in the dilution refrigerator , a series of calibrated ` heat pulses ' with a range of energies were introduced to the w @xmath8 system on a regular cadence throughout the data exposures .
these pulses produce photon - like signals of known arrival time and amplitude interleaved through the data stream .
these heat pulses were excised and monitored for any sign of gain drift . in practice ,
operation was sufficiently stable that no detectable drifts were generally seen in the heat pulse monitor .
the monochrometer calibration spectra were used to form the response matrices @xmath29 across detector energy bins @xmath30 to the incoming photon energies @xmath31 for each pixel @xmath32 . % . for comparison
the count spectrum observed from a 2.06ev setting of a grating monochrometer is shown .
the low energy events ( dashed line ) include a rising continuum due to substrate ( si ) hits , as well as a thermal ir h - band peak from the warm optics and lamp . ] to complete the spectrophotometric calibration , we observed the bruzual - persson - gunn - stryker spectrophotometric standard bd+30 2344 at low airmass during each observing night of the run .
although convenient for its well calibrated absolute ir - optical - uv spectral energy distributions , at @xmath33 the star had to be observed with substantial de - focus to avoid detector non - linearities at high count rates .
we therefore also observed the @xmath34 blue kpno flux standard pg 0939 + 262 ( massey _ et al._1988 ) to allow an improved estimate of the system uv response and a check on the point - source total efficiency , including aperture losses .
the fiber filter has strong absorption bands whose width is comparable to or smaller than our energy resolution . also for steep red spectra ,
the satellite ` rail hit ' peak at 1.26 times the main peak energy contributes significantly to the count rate .
this made it impossible to extract the flux scale by directly dividing the predicted standard star counts by the observed channel counts . instead the calibration star spectrum was convolved with the response function , the fiber absorption curve measured at high resolution , and the w absorption spectrum to produce model calibrator count spectra . comparing in the data space ,
an effective area function varying slowly with energy was fit to model the additional focus and absorption losses compared to stellar fluxes at the top of the atmosphere . combining these factors , we derive model efficiency functions @xmath35 for each active pixel @xmath32 .
after summing these we have a total ( detector+telescope+atmosphere ) efficiency for the system ; this peaks at @xmath36% . in figure 3
we correct this efficiency for the estimated atmospheric extinction , telescope inefficiencies and aperture losses to produce the effective response of the system ( fiber optics + detector ) at the back end of the telescope . for comparison ,
the energy ranges corresponding to johnson broad - band colors are indicated .
variation between the calibration sources indicate that the systematic errors in the flux scale are @xmath37 for the range 1 - 2.5ev . in the blue ,
the calibration spectra have lower s / n and larger contribution from the rail hit events , so the systematic errors in the flux scale rise , probably reaching 30% above 3.5ev .
comparison with archival crab data below support the idea that our flux scale is systematically high by approximately this amount in the u band .
imprecision in the estimates of the telescope and aperture losses also produce uncertainties exceeding 10% in the absolute flux scale .
we observed a number of compact object sources , both accretion powered binaries and isolated spin - powered pulsars .
the relatively bright ( @xmath38 ) crab pulsar was used to tune and characterize the system performance .
our gps - stamped crab photon arrival times were reduced to the solar system barycenter employing the widely used tempo pulsar timing package ( from pulsar.princeton.edu/tempo ) .
we used the monthly crab pulsar radio ephemeris published by jodrell bank ( lyne , pritchard & roberts 1999 ) to phase the photon arrival times .
the light curves from our four nights of observation are phased within @xmath39s ; the absolute phase agrees with the predicted infinite frequency pulse arrival time ( referenced to the peak of the main pulse ) to within @xmath40s .
this absolute precision is rather better than the nominal @xmath41s accuracy of the published monthly ephemeris and is completely adequate for our purposes , allowing a detailed comparison of our phase bin spectra with other ( e.g. space - based ) data sets .
we focus here on a relatively long ( 3500s ) coherent data set taken on february 6 , 2000 ( mjd 51581 ) for which the system parameters were stable and the observing efficiency was high .
this data set provides much of our total s / n for the crab pulsar and represents a unique time- and energy - resolved measurement of this well - studied object .
a sample optical ( 1.5ev-3.5ev ) light curve is shown for reference in figure 4 .
-bin light curve from 3500s of tes crab data .
phase 0 is referenced to the main peak of the radio ephemeris .
the upper histogram shows the same data at 100 bins ( zero suppressed ) , magnified to show the baseline variation .
our selected background region is shown ( label bg ) . ]
there is appreciable structure in our light curve baseline even at pulse minimum .
previous studies also infer magnetospheric emission throughout the pulse period ( e.g. sanwal _ et al._1998 ) .
recently golden , shearer & beskin ( 2000 ) have argued from mama imaging data that the phase interval 0.75 - 0.825 represents the dc pulse minimum with @xmath42% of the integrated pulse emission .
our data show significant structure in this region , but given the substantial background in our our 1.7@xmath16 aperture , we select the phase interval @xmath43 for our ` background ' spectrum , recognizing that there will be a small @xmath44% magnetospheric contamination ; this will be dominated by poisson fluctuations in nearly all of our extracted spectra . ; line histogram ) and observed ( filled dots ) count spectra are shown for comparison ( right scale ) . ] in figure 5 we show the phase - averaged spectrum after background subtraction , convolution through the response matrix and application of the flux calibration solution of 3 . to maximize the independence of the spectral bins , we first compute the expected count spectrum of the best - fit simple absorbed power - law model and then derive the crab fluxes by multiplying the model flux by the ratio of the observed to predicted counts for each channel .
this ` pre - whitening ' avoids spreading fluctuations in the observed count bins between final spectral bins and ensures that sensitivity to errors in the response matrices are of second order .
nonetheless , the derived spectrum represents only one model ( the smoothest ) that fits the count data .
the error bars shown are poisson only , derived from the independent counts in the energy bins . for comparison , we plot optical / ir phase - averaged broad - band photometry tabulated in eikenberry _
et al._(1997 ) .
we also show the phase - averaged spectrum fit by sollerman _
et al._(2000 ) to optical / uv data as the solid histogram this is an @xmath45 powerlaw subject to e(b - v)=0.52 , r=3.1 interstellar reddening displayed as a predicted count histogram .
this may be compared with the observed spectrum ( solid dots ) .
notice that , in agreement with the ir photometry of eikenberry _
et al._(1997 ) , this powerlaw under - predicts the j / h flux .
it is evident that our derived spectrum shows a somewhat steeper ( redder ) optical spectrum than the sollerman _
et al._power - law fit and also lies below the b & u colors of percival _
et al._(1993 ) .
this supports the suspicion of a systematic over - estimate of our sensitivity function above 2.5ev ( possibly due to incomplete modeling of the rail - hit component of the psf and/or pile - up ) .
the notch in the spectrum at @xmath46ev is also likely an artefact caused by the rapid variation in the system sensitivity at the oh absorption band .
interestingly , we find that our @xmath47ev spectrum is however very well modeled by an extincted @xmath48 power law , which is actually the same spectrum found originally by oke ( 1969 ) .
the true spectral index is evidently quite sensitive to the reddening parameters and energy range chosen and , indeed , the optical / uv spectrum of sollerman _ et al._shows appreciable curvature .
however , given the mismatch with the broad - band photometry in the blue , slight flux calibration systematics are the likely culprit .
we can , of course , study spectral variations in the crab by assuming that the phase - averaged spectrum is smooth and using our observed total pulsar spectrum as the flux calibrator .
this differential measurement allows a much improved removal of small scale variations in the sensitivity function and provides robust detection of any spectral phase variations .
figure 6 shows a selection of phase bin spectra , referenced to the total ( phase - averaged ) spectrum .
there are evidently subtle but continuous variations in the broad - band spectral index and a distinct change in the behavior in the near - ir band ( see , for example the break at @xmath49ev in the leading edge of the main pulse , spectrum 1 ) .
the spectra show little fine structure except at the strong fiber absorption bands .
smooth continuum variations dominate , as expected for a broad - band emission process such as optical synchrotron .
errors are computed by propagating poisson fluctuations in both the background subtracted spectra and the reference ( phase - averaged ) spectrum . ) for several phase bins from the 13 contiguous bins indicated in the panel to the right . ]
we can measure these smooth variations in smaller @xmath50 bins by fitting .
figure 7 shows a double power - law fit with a fixed break at 1.3ev . to obtain these spectral indices , we correct by a mean ( phase - averaged ) powerlaw . in view of the calibration uncertainties above , we have chosen to reference the spectra using a phase - averaged index of @xmath51 .
the spectral index fit errors are derived from the poisson spectral bin errors above . the data show a rapid sweep in the spectral index , especially in the main peak . an optical light curve and peak phase marks are provided for reference . at the leading edge of the main peak
the ir index does not reverse , but the spectrum becomes increasingly blue .
this indicates strong break or depression of the ir at the leading edge of the main pulse .
this is the origin of the ir variation in the leading pulse half - width noted by eikenberry , _
et al._(1997 ) . while the phase bins and error flags are large in the bridge between the pulses , the optical and ir indices also differ strongly in this region , with a steep red ir spectrum passing to a much harder component in the blue .
this increase of the bridge flux at high energies is in accord with the strengthening of the bridge emission in the x - ray and @xmath52-ray bands .
phase - dependent spectral indices were first noticed in the crab in the x - ray band ( pravdo & serlemitsos 1981 ) .
the reference _ rossi x - ray timing explorer ( rxte ) _ data above and more extensive bepposax opservations reported by massaro _
et al._(2001 ) show rapid variation in the hard x - ray power - law spectrum , especially near the first peak .
the phase coincidence of the optical / ir variations imply a common origin , but there is evidently no simple relation between the indices in the two band . at higher @xmath52-ray energies
limited count rates do not allow the fine phase bins used in the x - ray analysis , but a correlated spectral variation is seen across the crab pulse profile ( fierro 1995 ) . in the optical ,
many authors starting with muncaster & cocke ( 1972 ) have used used time - resolved color photometry to suggest that the bridge region is bluer than the pulses .
this is seen most clearly in the largely unpublished ubvr observations reported in sanwal _ et
al._(1998 ) .
recently , golden _
et al._(2000 ) have reported a _ reddening _ in the bridge region , based on ubv photometry , but this does not accord with other recent measurements .
note that with the exception of the sanwal _
al._(1998 ) data , limited count statistics forced all observations to use very coarse phase bins , typically summing over the entire pulse or interpulse .
the @xmath53 sweeps we see are located largely within these components ; this is a likely explanation for the non - detection of spectral variation in several studies ( e.g. carraminana , cadez & zwitter 2000 , perryman _
et al._1999 ) .
thus the optical @xmath53 variations through the pulses are seen here clearly for the first time .
the simultaneous ir measurements of this data set also provide a new window on the pulsar physics , given the rapid variation in the spectral break below @xmath54ev .
simultaneous optical / ir spectra were first reported in romani _
et al._(1999 ) .
recent simultaneous dual band ( v & h ) light curves have also been reported by moon _
et al._(2001 ) , but the low efficiency and limited s / n of these test data likely preclude measurement of the subtle spectral variations . although some interesting indications of spectral variations can clearly be seen in the broad - band ( non - simultaneous ) ir photometry of eikenberry _
et al._(1997 ) , the data reported above are the first to show these fine - resolution spectral variations extending into the infrared . while the variations are highly significant , longer integrations with better s
/ n are certainly desirable .
nonetheless , these data already provide some constraints on the radiation processes , which we describe briefly in 7 .
at our roughly kilohertz count rates , we detected @xmath2825 photons in each main pulse , giving us substantial sensitivity to photon correlations . in the radio
most pulsars exhibit a gaussian or log - normal distribution of pulse intensities .
the crab , however , also produces occasional radio pulses with up to @xmath55 the mean pulse energy .
these giant pulses are short duration bursts of flux , appearing randomly in narrow windows of pulsar phase .
et al._(1995 ) have searched for and placed limits on any correlation between giant radio pulses and @xmath52-ray emission , while patt _
et al._(1999 ) have found that no giant pulses are present in the x - ray band .
et al._(1993 ) have also found no non - poisson pulse - to - pulse variations in the optical flux of the individual pulse components .
lacking simultaneous radio data , we examined our data for any short - time scale photon correlations .
we searched for non - poisson photon distributions by predicting the expected counts in a range of lag bins following each observed photon . because each detector has a small @xmath39s dead time associated with the recovery from a photon pulse , we searched the _ cross_-correlations between our pixels , allowing us to extend the analysis to microsecond times . to test the null poisson hypothesis , we need an accurate estimate of the expected count rate in the face of the varying pulsed emission and background in our aperture .
this was obtained by forming a 100-bin light curve for each second of data and fitting the counts to a background and pulsed flux amplitude . for each photon arriving during this second
, we used the two fit amplitudes to _ predict _ the photon arrival rate in the other three detectors on a range of timescales .
the model flux in each lag bin was determined by an integration through the appropriate phases of a high - resolution ( @xmath56bin figure 4 ) light curve from all the data in a selected energy range , scaled to the instantaneous detected flux .
the observed counts can be compared to the poisson expectation for photons of a given energy range on various timescales and over various windows of pulsar phase .
no strong energy dependency was noted , so we show the statistics of the full low background 1 - 4ev energy range .
the grey scale image shows the ratio in independent lag bins , spaced as even octaves ( vertically ) , vs. pulse phase ( horizontally ) .
light shows high correlation and dark shows sub - poisson counts .
bins with @xmath57 counts are set to 1 . ]
figure 8 shows as a gray intensity scale the observed count rate in units of the poisson predicted rate in 100 bins of pulsar phase .
the vertical axis of the grey scale shows correlation bins covering time lags @xmath58 , where @xmath59s .
@xmath60 runs from 1 to 11 ( @xmath61s to @xmath62s ) in the gray scale plot shown .
the lag bins are thus completely statistically independent . a significant correlation ( or anti - correlation ) on short time scales would be expected to show as a coherent bright ( or dark ) region at the appropriate phase , fading to grey in longer timescales .
the data show fluctuations in the individual amplitudes when the counts are low .
there is a @xmath63% enhancement on timescales @xmath64s leading the first peak by @xmath54ms ; this is of @xmath65 significance .
there is also a @xmath4% enhancement at @xmath66s leading the second peak , but this is only @xmath67 in the most significant lag bin .
a few percent decrement at lag @xmath68 appears correlated with noise in the 33@xmath14s bins in the fine template .
all other significant features are smaller than 1% .
we conclude that our photon arrival times are consistent with poisson and that there is no evidence for strong correlations on short timescales , such as giant pulses , at optical energies .
ultimately , complete exploitation of astronomical fluxes requires time - resolved spectrophotopolarimetry .
the strong , rapidly varying polarization of the crab pulsar makes it an attractive test target .
phase resolved polarimetry has been presented in the optical ( smith _ et al._1988 ) and the near uv ( smith _ et al._1996 ) , which give a basis for comparison .
the double position - angle sweep reported in these papers has been shown to be a good quantitative match to a relativistic outer magnetosphere version of the rotating vector model by romani and yadigaroglu ( 1995 ) . in this model emission from a range of altitudes forms caustics in photon arrival phase that define the pulses .
further , a combination of two distinct emitting regions can contribute to the interpulse phases .
if the radiating populations ( and hence spectral indices and breaks ) differ between these regions we would expect energy dependence of the polarization fraction . as a test
, we observed the crab pulsar on february 7 , 2000 ( mjd 51582 ) with a simple polaroid filter placed before the focal plane .
we integrated at two full cycles of position angle with the following exposures : @xmath69 ( 1370s,630s ) , @xmath70 ( 520s,690s ) , @xmath71 ( 740s,620s ) , @xmath72 ( 960s,680s ) with angles referenced to the equatorial coordinate system .
of course , varying background and varying aperture losses between the exposures mean that the light curves must be renormalized before forming @xmath73 and @xmath74 . to subtract the substantially polarized nebular emission in our aperture , we employed our usual background phase window ( 0.76 - 0.87 ) , following the prescription in ( smith _ et al._1996 ) .
while residual pulsar emission at these phases may be highly polarized ( smith _ et al._1988 ) , the amplitude is small enough that its contamination of the peaks and bridge region should be smaller than our statistical errors .
an examination of the smith _
et al._(1988 ) data ( kindly supplied in digital form by f.g .
smith ) showed a polarization minimum during the bright first pulse in the region @xmath750.005 - 0.025 , consistent with 0 , within the error bars .
this provides a convenient exposure monitor based on the pulsed emission itself .
these data were taken in a broad optical band with the ` rgo people s photometer ' ; we synthesized this band in our data , selecting the appropriate energy bins and using the polarization minimum flux as a monitor of the total crab exposure at each polarization position angle . in figure 9 , we show the resulting polarization fraction @xmath76 corrected for the wavelength dependent polarization efficiency of the filter ( measured using monochrometer light passed through crossed filters ) .
we also corrected for the bias due to non - normal statistics at small polarization , with the observed polarization @xmath77 corrected to @xmath78^{1/2}$ ] .
in the upper panel we show the polarization position angle @xmath79 .
the error flags shown are the propagated poisson errors . differences between the two sets of position angle data were within the statistical errors , so these were summed in the polarization amplitude and sweep shown .
limited statistics restricted us to 18 phase bins through the pulse , so fine structure is likely not fully resolved .
the overall polarization behavior follows that measured in previous optical / uv experiments quite closely .
we show the optical data divided into three energy bands .
the rapid decrease in polarization and strong position angle sweep of the first pulse appear quite energy independent .
there are small differences in the polarization and position angle of the bridge and second pulse , but these are of marginal significance .
given that the phase variations in the spectral index and break are strongest below 1.5ev ( 4 ) we might expect the strongest spectral effects in the near - ir .
indeed , our near ir data do appear to show significantly different polarization properties .
unfortunately , the increased ir background and difficulty in calibrating the rapidly varying ir polarization response of the filter preclude a quantitative comparison with the optical data in the present experiment .
serious phase - resolved spectropolarimetry will require imaging solutions for local background subtraction ; as part of our next - generation imaging system ( see below ) , we plan to install wollaston prisms for imaging spectrophotopolarimetry over reduced fields .
the phase variations through the crab profile described above clearly provide new tests of the physics of the optical / ir radiation .
the polarized ir through hard x - ray non - thermal emission from the crab is generally held to be synchrotron radiation from a population of @xmath80 propagating and cooling in the outer magnetosphere ( e.g. lyne and smith 1990 and references therein ; romani 1996 ; wang , ruderman & halpern 1998 ) .
the broad - band phase - averaged sed is smooth , with @xmath53 varying from @xmath81 ( hard x - rays ) through @xmath82 ( soft x - rays ) to @xmath83 at optical frequencies .
there is evidently a peak in the near - ir and then the spectrum drops off dramatically at lower energies ( middleditch , pennypacker & burns 1983 ) , although this important result definitely requires confirmation . recently , crusius - waetzel _
et al._(2001 ) have provided a synchrotron interpretation that provides a plausible explanation of the soft - x - ray through ir phase averaged sed in terms of a single @xmath84 population ( presumably produced through an @xmath80 pair cascade in the outer magnetosphere ) with @xmath85 . in the x - ray band , the emission is assumed to be classical relativistic synchrotron radiation with @xmath86 ; this gives the observed @xmath87 spectrum observed for the powerlaw x - ray emission of the crab and other young pulsars .
however , crusius - waetzel _
et al._additionally argue that the @xmath80 population should have a component with very low pitch angle @xmath88 , for which the synchrotron emission formulae give @xmath89 , when averaged over the radiating beam .
for the crab , they infer @xmath90 and @xmath91 , and connect the @xmath92 spectrum with that seen in the optical .
finally , they suggest that below @xmath93ev the particles dominating the emission spectrum have even lower @xmath52 , so that the observed pulse width @xmath94 samples only a subset of the @xmath95 radiation cone and argue that this leads to a @xmath96 spectrum . this index is taken to be , at least in part , an explanation of the steep rise in the mid - ir .
if we take the x - ray spectral index to be a _ local _ measurement of the electron powerlaw @xmath97 then this picture predicts the correlated optical spectrum in the low pitch angle limit . in figure 10 , we show the tes data compared with this prediction , based on the 0.1 - 4.0kev data of massaro _
et al._(2001 ) . clearly
this phase bin application of the low pitch - angle picture is not adequate : while the index and trend for the second pulse are reasonably well predicted , the spectra of the bridge region and , especially , the first pulse show no match to the extrapolations of soft x - ray data .
when one considers that the spectral index in all phase bins steepen into the hard x - rays , then one must conclude that a simple application of powerlaw synchrotron emission , even with a phase - varying @xmath97 , is not adequate .
one factor likely important in a more complete model is the inclusion of the local beaming of the radiation cones along the emission surface .
given the sharp pulses and the observation at a single line - of - sight , the radiation must differ from the emission - cone averaged synchrotron expressions above .
models including these effects are sensitive to the detailed geometry . as an example , consider the strong ir break at the leading edge of the main pulse . in the outer magnetosphere caustic picture of the pulse formation , the leading edge of the main pulse arises from emission propagating along the radiating surface for the longest distance . hence , one would expect this to be the region first showing self - absorption suppression .
while the overall ir suppression of the main peak reported by penny ( 1982 ) suggestive of uniform self - absorption is not seen in our data , the observed leading edge suppression and steepening may be a more restricted instance of self - absorption .
polarization measurements extended into the ir will be a good discriminant between absorption effects and emission breaks , as a restricted sampling of the synchrotron emission cone leads to polarization signatures varying significantly in accord with the spectral breaks .
we are developing a focal plane camera based on tes detector arrays , with the potential for time resolved imaging spectral polarimetry . as the data of golden _
et al._(2000 ) and perryman _
et al._(1999 ) show , the absolute photometry and background subtraction of such true imaging measurements can provide much improved isolation of faint pulsar phenomena .
with the clean monoenergetic energy psf of a ` masked ' tes array , improved system qe and improved energy resolution , observations of the crab ( and other fainter pulsars ) should allow a detailed examination of the phase - resolved phenomena discovered in the exploratory measurements reported here .
this work was supported in part by grants from nasa ( nag5 - 3775 and nag 5 - 3263 ) , from the doe ( de - fg03 - 90er40569 ) , from the stanford otl fund and from the research corporation .
this work also made use of the stanford nanofabrication users network funded by the national science foundation under award number ecs-9731294 .
we thank kent irwin and r. welty of nist for continued collaboration on tes and squid technology , david doss of mcdonald observatory for assistance with the observational set - up and colin bischoff for assistance with the data analysis .
cabrera , b. , _
_ 1998 , _ appl .
_ , * 73 * , 735 .
carraminana , a. , cadez , a. zwitter , t. 2000 , _ apj _ , * 542 * , 947 .
chervenak , j.a . , _
_ 1999 , _ appl .
_ , * 74 * , 4043 .
crusius - wtzel , a.r , kunzl , t. & lesch , h. 2001 , _ apj _ , * 546 * , 401 eikenberry , s.s . , _
et al._1997 , _ apj _ , * 476 * , 281 .
fierro , j.m .
1995 , phd thesis , stanford university .
golden , a. , shearer , a. & beskin , g.m .
2000 , _ apj _ , * 535 * , 373 .
golden , a. , _ et al._2000 , _ aa _ , * 363 * , 617 .
humbach , o. , _
et al._1996 , _ journal of non - crystaline solids _ , * 203 * , 19 .
irwin , k.d .
1995 , _ appl .
_ , * 66 * , 1998 .
lundgren , s.c .
_ et al._1995 * apj * , _ 453 _ , 433 .
lyne , a.g . , pritchard , r.s . and roberts , m.e .
1999 , ( http://www.jb.man.ac.uk/@xmath28pulsar/crab.html ) lyne , a.g .
& smith , f.g .
1998 , _ pulsar astronomy _
( cambridge : cambridge ) .
massaro , e. , cusumano , g. , litterio , m. & mineo , t. 2001 , _ aa _ , * * , massey , p. , strobel , k. , barnes , j.v & anderson , e. 1988 , _ apj _ , * 328 * , 315 .
middleditch , j. , pennypacker , c. & burns , m.s .
1983 , _ apj _ , * 273 * , 261 .
miller , a.j . , _
et al._2000 , _ nimpa _ , * 444 * , 445 .
moon , d .- s . ,
pirger , b.e . & eikenberry , s. 2001 , pasp , * 113 * , 646 muncaster , g.w . & cocke , w.j .
1972 , _ apj _ , 178 , l13 oke , j.b .
1969 , _ apj _ , * 156 * , l49 patt , b.l . , ulmer , m.p .
, zhang , w. , cordes , j.m . &
arzoumanian , a. 1999 , _ apj _ , * 522 * , 440 .
penny , a.j .
1982 , _ mnras _ , * 198 * , 773 percival , j.w .
_ et al._1993 , _ apj _ , * 407 * , 276 perryman , m.a.c . ,
foden , c.l . , & peacock , a. 1993 , _ nucl .
_ , a325 , 319 .
perryman , m.a.c .
, favata , f. , peacock , a. , rando , n. & taylor , b.g . 1999 , _ aa _ , 346 , l30 .
pravdo , s.h . & serlemitsos , p.j .
1981 , _ apj _ , * 246 * , 484 .
pravdo , s.h .
angellini , l. & harding , a.k .
1997 , _ apj _ , * 491 * , 808 .
romani , r.w .
1996 , _ apj _ , * 470 * , 469 .
romani , r.w . , _
et al._1999 , _ apj _ , * 521 * , l153 .
romani , r.w .
& yadigaroglu , i .- a .
1995 , _ apj _ , * 438 * , 314 sanwal , d. , robinson , e.l . & stiening , r.f .
1998 , _ baas _ , * 30 * , 1420 .
smith , f.g . ,
jones , d.h.p . ,
dick , j.s.b .
& pike , c.d . 1988 , _ mnras _ , * 233 * , 305 .
smith , f.g . , _
et al._1996 , _ mnras _ , * 282 * , 1354 sollerman , j. _ et al._2000 , _ apj _ , * 537 * , 861 .
wang , f.y .- h . , ruderman , m. & halpern , j.p .
1998 , _ apj _ , 498 , 373 wollman , d.a .
_ et al._2000 , _ nimpa _ , * 444 * , 145 . | we are developing time- and energy - resolved near - ir / optical / uv photon detectors based on sharp superconducting - normal transition edges in thin films .
we report observations of the crab pulsar made during prototype testing at the mcdonald 2.7 m telescope with a fiber - coupled transition - edge sensor ( tes ) system .
these data show substantial ( @xmath0 ) , rapid variations in the spectral index through the pulse profile , with a strong phase - varying ir break across our energy band .
these variations correlate with x - ray spectral variations , but no single synchrotron population can account for the full spectral energy distribution ( sed ) .
we also describe test spectrophotopolarimetry observations probing the energy dependence of the polarization sweep ; this may provide a new key to understanding the radiating particle population . |
nuclear - polarized @xmath0he has proven to be useful in a number of different areas of research . in electron scattering , polarized @xmath0he provides a means for studying spin - dependent interactions involving neutrons .
this is because , to first approximation , a @xmath0he nucleus is comprised of two protons whose spins are paired , and a single neutron that accounts for most of the nuclear spin @xcite .
an important early example of the use of polarized @xmath0he in electron scattering came during an experiment to measure the internal spin structure of the neutron at the stanford linear accelerator center ( slac ) , e142 @xcite . polarized
@xmath0he has also been used to measure the electric form factor of the neutron @xmath2 , including a recent experiment at jefferson laboratory ( jlab ) in newport news , va @xcite .
important applications of polarized @xmath0he have also included its use as a neutron polarizer @xcite , and , together with polarized @xmath3xe , as a source of signal for magnetic resonance imaging @xcite .
there are two predominant techniques by which high levels of nuclear polarization are produced in @xmath0he . in one technique , often known as metastability exchange optical pumping ( meop ) , metastable states of @xmath0he
are optically pumped directly , and subsequently transfer their polarization to other ground - state @xmath0he nuclei during metastability - exchange collisions @xcite . in a second technique , known as spin - exchange optical pumping ( seop ) , a vapor of alkali - metal atoms is optically pumped and subsequently transfers its polarization to @xmath0he nuclei via a hyperfine interaction during spin - exchange collisions @xcite . an important difference between the two techniques is that meop is performed at pressures that are quite low , around a few torr , whereas seop is often done at pressures as high as roughly ten atmospheres .
when a high - density is required , a target based on meop inevitably involves a compressor of some sort .
in contrast , high - density targets based on seop typically utilize a sealed glass cell with no moving parts and hence have an advantage from the perspective of simplicity . in considering relative merits ,
however , one also needs to consider the speed with which the gas is polarized . here
targets based on meop have done quite well , with a recent target at the mainz microtron reporting a polarization rate of 2 bar - liters per hour@xcite . in short , both techniques for polarizing @xmath0he have been quite successful and have complimentary advantages . for electron - scattering experiments
in which @xmath0he is polarized using seop , the targets typically utilize a sealed glass cell with two distinct chambers : a pumping chamber " in which the gas is polarized , and a target chamber " through which the electron beam passes ( see fig . [
fig : genstyle ] ) .
this design ensures that ionization due to the electron beam does not adversely affect the optical pumping process , as well as providing in the pumping chamber a geometry that lends itself well to illumination with lasers .
the two chambers are connected by a transfer tube " , and gas that is polarized in the pumping chamber migrates into the target chamber largely by diffusion .
several conditions need to be met in targets with designs such as that shown in fig . 1 in order to maintain high polarization .
first , the rate at which the @xmath0he is polarized must be relatively fast compared to the rate at which the @xmath0he is depolarized . while this is true in all work involving seop , it is of particular significance when considering depolarization of the @xmath0he due to an electron beam .
there are also issues having to do with polarization dynamics that are unique to the two - chamber design . as has been pointed out by chupp _
et al._@xcite , it is important that diffusion between the two chambers is rapid compared to other @xmath0he - related time constants in the system . as we will show in detail ,
a matter of special importance is maximizing the ratio of the diffusion rate to the relaxation rate that is specific to the target chamber .
put differently , it is critical that the polarized gas in the target chamber is replenished much faster than it is depleted through depolarization mechanisms such as , for example , ionization from the electron beam .
a failure to replenish the polarized gas quickly enough will result in a lower polarization in the target chamber than is the case in the pumping chamber , a condition we refer to herein as a _
polarization gradient_. until relatively recently , polarization gradients in two - chambered cells have been at most a few percent relative .
advances in seop , however , have made it possible to run polarized @xmath0he targets at increasingly high luminosities . during slac e142 , where luminosities were in the range of @xmath4 ,
the polarization gradients were on the order of 1% . during recent experiments at jlab , however , with luminosities in the range of @xmath5 , polarization gradients were as high as nearly 9% ( relative ) . for future experiments with luminosities in the range of @xmath6 , polarization gradients well in excess of 10% are likely if no changes are made to the basic target - cell design .
polarization gradients can also be difficult to quantify accurately , an issue that can lead to uncertainties in polarimetry .
the fact that polarized @xmath0he targets are being run at increasingly high luminosities is due largely to advances in seop .
one example is the use of hybrid mixtures of alkali metals ( typically potassium and rubidium ) instead of a single alkali metal ( typically rubidium)@xcite .
this technique , often known as alkali - hybrid seop , greatly improves the efficiency with which the @xmath0he is polarized .
another important advance has followed from the availability of commercial spectrally - narrowed high - power diode - laser arrays .
these new lasers result in significantly higher optical pumping rates for a given amount of light .
collectively , these advances have made it possible to significantly increase the rate at which the @xmath0he is polarized .
it has thus become possible to achieve higher polarizations even when using higher beam currents and thicker targets .
higher currents , however , make it necessary to replenish the polarized gas in the target chamber more quickly . if the advances in seop are to be fully exploited , it is essential that the designs of polarized @xmath0he targets evolve .
we report here on the successful implementation of a new design for polarized @xmath0he target cells based on seop .
the design incorporates the ability to circulate the gas between the pumping chamber and the target chamber using convection instead of diffusion , an idea discussed by wojtsekhowski in ref .
@xcite in anticipation of both advances in seop as well as the need for higher luminosities
. the convection is achieved by maintaining a temperature differential between different parts of the target cell , and does not involve pumps or other moving parts .
we have shown that the velocity of the gas moving through the target chamber can be varied between 580@xmath7 in a simple controllable manner .
the advent of a means to circulate a polarized noble gas in a sealed vessel without the use of pumps has great potential for high - luminosity polarized @xmath0he targets .
the simplicity of the approach has advantages from the perspective of reliability , and fast transport of gas between the two chambers makes it possible to greatly increase the electron beam current without causing a polarization gradient .
convection - based target cells also open the possibility of physically separating the pumping chamber and the target chamber by much larger distances than was previously possible , something that offers several practical advantages .
before considering the formalism for target cells with two chambers , we begin by considering the simpler example of a single - chambered cell , where the equation describing the time evolution of the polarization is given by : @xmath8 where @xmath9 is the @xmath0he polarization , @xmath10 is the polarization of the alkali vapor , @xmath11 is the rate at which the @xmath0he is polarized due to spin exchange , and @xmath12 is the spin - relaxation rate of the @xmath0he due to all other processes .
the solution to eqn.[eq : singlecell ] is given by @xmath13 where @xmath14 is the @xmath0he polarization at @xmath15 .
it has been shown by babcock _
_ that one of the components of @xmath12 is a relaxation rate that empirically appears to be proportional to @xmath11 @xcite .
one can accordingly write that @xmath16 , where @xmath17 is a proportionality constant .
we can thus write the saturation polarization associated with eqn .
[ eqn : psingle ] as @xmath18 where we note that the denominator of eqn .
[ eqn : psingle_w_x ] is also the rate that characterizes the buildup of polarization in [ eqn : psingle ] , rewritten in terms of @xmath19 and @xmath17 .
the existence of a relaxation mechanism proportional to @xmath11 is unfortunate , as it implies that , even with a 100% polarized alkali vapor , in the limit that @xmath20 , @xmath21 . while not well understood ,
the existence of this additional relaxation is now well established , and is important for understanding the overall polarization achieved .
it is straightforward to project the performance of a target at an arbitrary range of beam currents if we know how it performed at a particular beam current @xmath22 ( true even if @xmath23 ) .
let us assume that we have data describing polarization as a function of time while polarizing a target from an initial polarization of zero .
a plot of data of this sort is something we refer to herein as a
spin - up curve " , and can be fit to yield the spin - up rate " that appears in the denominator of eqn .
[ eqn : psingle_w_x ] ( as well as the exponent in eqn .
[ eqn : psingle ] , @xmath24 ) . in keeping with the notation introduced above with respect to eqn .
[ eqn : psingle_w_x ] , we define a quantity @xmath25 that characterizes the spin - up rate at a particular current @xmath26 : @xmath27 let us further define @xmath28 as the equilibrium polarization associated with that spin - up . using eqns .
[ eqn : psingle_w_x ] and [ eq : gamma_su_def ] , we find @xmath29 the quantities @xmath28 and @xmath25 can be determined by fitting data from a spin up , and the quantity @xmath30 , where the three terms are spin - relaxation rates due to wall collisions ( at room temperature ) , dipole interactions due to @xmath0he-@xmath0he collisions , and ionization of the electron beam respectively .
the sum of the first two terms is essentially the target s room - temperature spin - relaxation rate in the absence of beam , a quantity that is quite easy to measure .
we note , however , that @xmath31 will be slightly different under operating conditions than it is at room temperature both because of heating the pumping chamber and the related differences in densities .
this correction is easily calculated using the calculation of newbury _
et al . _ from ref .
the contribution @xmath32 is also easily computed @xcite
. we note that calculations of @xmath32 have shown good agreement with experiment at a level of roughly 10% or better @xcite .
we can thus use eqn .
[ eq : model ] to project the performance of a particular target at an arbitrary beam current .
it is instructive to investigate how different targets that have been used during past experiments would fare at significantly increased current , for example , at @xmath33 .
here we ignore the effect of polarization gradients .
the first time liter - type quantities of polarized @xmath0he were used in a target was the aforementioned experiment at slac , e142 @xcite .
in the presence of @xmath34 of electron beam , the @xmath0he polarization averaged about 33% , with values for @xmath35 of about 15 - 20 hours . if this same target were instead exposed to @xmath33 of electron beam , eqn .
[ eq : model ] suggests that the @xmath0he polarization would drop to just over 10% . in contrast , during the more recent experiment at jlab that measured @xmath2 @xcite , the cell - averaged polarizations were around 50% with @xmath36 of beam and @xmath35 was in the range of @xmath37 . here
[ eq : model ] suggests that at @xmath33 the resulting @xmath0he polarization would be around 3738% .
the improved projected performance is due both to the shorter values for @xmath35 , as well as the fact that the target cells were significantly larger , making them more resistant to depolarization from the electron beam .
we note , however , that even though the cell - averaged polarization would be fairly reasonable , the polarization that one would have in the target chamber would be much lower because of polarization gradients .
for a full description of a double - chambered cell , the polarization build up must be described by the coupled differential equations first described in ref .
@xcite : @xmath38 where @xmath39 ( @xmath40 ) is the @xmath0he polarization in the pumping ( target ) chamber , @xmath11 is the spin - exchange rate in the pumping chamber , and @xmath41 and @xmath42 are the @xmath0he spin - relaxation rates due to all other processes in the pumping and target chambers respectively .
the transfer rate @xmath43 ( @xmath44 ) is the probability per unit time per nucleus that a nucleus will exit the target ( pumping ) chamber and enter the pumping ( target ) chamber .
we note that we do not include the transfer tube as a separate volume in this analysis .
the transfer rates are related by @xmath45 where @xmath46(@xmath47 ) is the fraction of atoms in the pumping ( target ) chamber , and @xmath48 . for the dynamic studies reported in ref .
@xcite , the authors were able to neglect terms involving @xmath11 and @xmath12 relative to terms involving @xmath44 and @xmath43 .
for the discussion here , however , we must retain these terms , requiring an analysis essentially identical to that considered by jones _
et al . _
@xcite and later by kominis @xcite .
we refer the reader to those two references for details .
we find that the solutions to eqns .
[ eqn : ppcdot1 ] and [ eqn : ptcdot1 ] are given by @xmath49 and @xmath50 where @xmath51 and @xmath52 are the initial polarizations in the pumping and target chambers respectively , @xmath53 and @xmath54 we have chosen to write eqn .
[ eq : pcsatpol ] in the form shown to emphasize that @xmath55 where @xmath56 is the spin - exchange rate averaged throughout the double - chambered cell ( @xmath57 , since the spin - exchange rate is @xmath11 in the pumping chamber and is zero in the target chamber ) , and @xmath58 is the spin - relaxation rate averaged throughout the cell .
[ eq : limit ] has the same form as the saturation polarization of eqn .
[ eqn : psingle ] , as we would expect in the limit of infinitely fast transfer .
the coefficients @xmath59 and @xmath60 are given by @xmath61 and @xmath62 where @xmath63 , @xmath64 , @xmath65 and @xmath66 .
we note that in the fast - transfer limit , the quantities @xmath59 and @xmath60 are given by @xmath67 and @xmath68 the constants @xmath69 and @xmath70 represent slow and fast rates respectively that govern the time evolution of the polarization .
it is useful to write @xmath69 in the form @xmath71 where the quantity @xmath72 is generally small and goes to zero in the limit of infinitely fast transfer between the two chambers .
we can see that the limiting form of @xmath69 has the same form as the time constant that appears in eqn .
[ eqn : psingle ] ( @xmath73 ) .
it can be shown that @xmath74\ ] ] where @xmath75 and @xmath76 for most of the situations we would normally consider , the quantity @xmath77 is fairly small .
this is due to two things .
first , the spin - exchange rate @xmath11 is typically slow compared to the sum of the two transfer rates @xmath44 and @xmath43 , and second , both @xmath41 and @xmath42 must be relatively small compared to @xmath11 , or the polarization of the target would not be high .
it is thus reasonable to expand @xmath72 in a taylor series in @xmath77 : @xmath78
lastly , we consider @xmath70 which can be written as @xmath79 in the fast - transfer limit , @xmath80 ; under these conditions , eqns .
[ eqn : ppct ] and [ eqn : ptct ] reduce to the form of eqn .
[ eqn : psingle ] .
data illustrating the time evolution of polarization ( what we referred to earlier as a spin - up ) are shown in fig .
[ fig : brady ] for both the pumping and target chambers of a double - chambered cell we refer to herein as brady " .
the polarization was measured every three minutes using the nmr technique of adiabatic fast passage ( afp ) @xcite .
we note that under normal operating conditions , nmr measurements would only be made once every few hours , in part because each measurement results in a small loss ( @xmath81 ) of polarization .
the frequent measurements shown in fig .
[ fig : brady ] strongly limit the saturation polarization because of accumulating losses .
also shown in fig .
[ fig : brady ] , but obscured beneath the many data points , is a fit to the data using double - exponential functions of the form given in eqns .
[ eqn : ppct ] and [ eqn : ptct ] .
the fit clearly describes the data quite well .
finally , we note that in the context of the types of cells that have been used in electron - scattering experiments ( see fig .
[ fig : genstyle ] ) , the mechanism behind the transfer rates @xmath43 and @xmath44 is overwelmingly diffusion .
l + -0.2truein some of the parameters discussed earlier can be readily determined by studying spin - up curves of the sort shown in fig .
[ fig : brady ] . to extract values for the transfer rates @xmath44 and @xmath43
, it is particularly valuable to examine the spin - up curves for the initial time period during which the polarization is growing . for small values of the time @xmath82 , it is readily apparent from fig .
[ fig : brady ] that the nature of the time evolution in the two chambers is quite different . under the assumption that the time @xmath83 ( in this case , @xmath84hrs )
, we can expand eqns .
[ eqn : ppct ] and [ eqn : ptct ] in a taylor series . to second order ,
for the case of @xmath85 , this expansion simplifies to : @xmath86 and @xmath87 in fig .
[ fig : bradyearly ] , we show only the first 24 minutes of the data shown in fig .
[ fig : brady ] .
it can be seen that the initial shape of the spin - up curve appears to be linear in the pumping chamber and quadratic in the target chamber , in agreement with expectations from eqns .
[ eq : taylorpc ] and [ eq : taylortc ] . to empirically determine @xmath43 , we note first that the slope of the nearly - linear polarization buildup in the pumping chamber is equal to the product @xmath88 . with a fit to this slope , along with a fit to the coefficient characterizing the quadratic polarization buildup in the target chamber
, we can extract a value for the transfer rate @xmath89hrs@xmath90 .
as will be discussed in the next subsection , a value for @xmath43 can also be computed from first principles given the dimensions of the cell .
the comparison of empirically determined and calculated values for @xmath43 provides insight into our understanding of the diffusion processes taking place in our cells .
-0.15truein -0.2truein using gas kinetic theory , the dimensions of the target cell , the density of @xmath0he and other gases when the cell was filled , the operating temperatures of the pumping and target chambers and the assumption that the temperature gradient along the transfer tube is linear , it possible to compute @xmath44 and @xmath43 from first principles .
we begin similarly to the discussion in ref .
@xcite with an equation describing the net polarization flux @xmath91 through the transfer tube : @xmath92 where @xmath93 , @xmath94 and @xmath95 are the @xmath0he number density , self - diffusion constant and polarization respectively , all shown as a function of position @xmath96 along the transfer tube .
as discussed by romalis @xcite and later by zheng @xcite , using data on the self - diffusion constant of @xmath97he by kestin _
et al . _
@xcite , the self diffusion constant for @xmath0he is well approximated by the expression @xmath98 where @xmath99 , @xmath100k , @xmath101 and @xmath102 ( 1 amg = @xmath103@xmath104 ) .
the assumption of a linear temperature gradient along the transfer tube between the pumping and target chambers , as was assumed in refs .
@xcite , allows us to express @xmath105 and hence @xmath93 explicitly . with this assumption , and substituting eqn .
[ eq : diff_constant ] into eqn .
[ eq : diff_diff_eq ] , we can solve for @xmath91 to find @xmath106 @xmath91 is the total rate of polarization transfer per unit area , whereas we want the rate per atom .
multiplying by the transfer tube cross sectional area @xmath107 , dividing by the number of particles in each chamber , and finally dividing by @xmath108 , we have @xmath109 of specific interest here is the value for @xmath43 implied by eqn .
[ eqn : dtc ] for the target cell brady . for brady , @xmath110@xmath111 , @xmath112 cm , @xmath113cm@xmath0 , and @xmath115 . using temperatures that correspond to the tests illustrated in figs .
[ fig : brady ] and [ fig : bradyearly ] , we find @xmath116hrs@xmath90 , a value that agrees fortuitously with the value found by fitting the polarization buildup curves .
an issue of considerable practical importance for polarized @xmath0he targets is the polarization gradient between the pumping and target chambers .
dividing both sides of eqn .
[ eq : polgrad ] by @xmath117 , we find @xmath118 it is also useful to define the quantity @xmath119 , the amount by which the polarization in the target chamber is lower than that of the pumping chamber . here
we express this difference as a fraction of the pumping chamber polarization : @xmath120 we can see that @xmath119 approaches 0 for cells in which @xmath42 is quite slow and @xmath43 is quite fast .
the fact that it is @xmath42 and not the cell - averaged relaxation rate @xmath121 that appears in this expression is quite important .
when a target is subjected to high electron - beam currents , the overall cell - averaged relaxation rate @xmath121 may not be strongly affected even though the local relaxation rate in the target chamber @xmath42 is . as stated earlier
, what is important is that the polarized gas in the target chamber is replenished at a rate that is much faster than the rate at which the gas is depolarized in the target chamber .
looking at specific examples , we find that the targets used during e142 at slac had polarization gradients of just over 1% with no beam current , and would hypothetically have had polarization gradients of around 13% at @xmath33 .
for the targets used to measure @xmath2 at jlab , the gradients were worse , a little over 6% with no beam current , and we project they would be about 21% at @xmath33 .
the larger gradients in the case of the @xmath2 cells were mostly due to faster intrinsic cell - relaxation rates , but also to having longer transfer tubes , something that is difficult to avoid with a cell that has a larger overall size .
shortly we will return to the question of what can be done to better minimize polarization gradients when using a single transfer tube .
regardless of the magnitude of the polarization gradient , it is also difficult to quantify accurately because of uncertainty in the quantity @xmath42 .
when characterizing a target cell , the quantity that is most straightforward to measure is the cell - averaged room - temperature spin - relaxation rate @xmath122 , which is typically due to three primary contributions : @xmath123 where @xmath124 is the spin relaxation rate due to wall collisions , @xmath125 is spin relaxation due to dipolar interactions during @xmath126 collisions and @xmath127 is the spin relaxation due to the electron beam , which can be taken to be zero if the cell s spin - relaxation rate is measured in the absence of an electron beam . here
we ignore relaxation due to magnetic field inhomogeneities which with some care can be made quite small
. the wall relaxation rate will be the sum of the wall relaxation rates in the target and pumping chambers respectively , weighted by the fraction of @xmath0he atoms that are in each chamber : @xmath128 for the purposes of this discussion , it is convenient to introduce a parameter r , representing the ratio of @xmath129 to @xmath130 , so that @xmath131 from eqns .
[ eq : contributions1][eq : contributions3 ] , taking @xmath132 , we find an expression for @xmath129 : @xmath133 unfortunately , we have no direct measurement of @xmath134 , and wall - relaxation rates are notoriously variable .
one plausible assumption is that @xmath135 , in which case @xmath136 , and eqn .
[ eq : contributions4 ] simplifies to @xmath137 .
this is , in fact , the assumption that has been made in polarized @xmath0he experiments ( those using the basic design shown in fig .
[ fig : genstyle ] ) prior the aforementioned @xmath2 experiment .
several authors have shown , however , that for uniform wall relaxation per unit area , overall wall relaxation is proportional to the surface - to - volume ( s / v ) ratio of the vessel containing the gas @xcite .
this dependence was recently exploited by anger _
et al . _ who successfully constructed storage cells for polarized @xmath3xe with unusually long , perhaps unprecedented , relaxation times @xcite .
if wall relaxation were uniform throughout a target , we would expect @xmath134 to be the ratio of the s / v ratios of the pumping and target chambers respectively , a quantity we will refer to here as @xmath138 .
a conservative approach , then , might be to take @xmath139 , with an error on @xmath119 that includes the full range of @xmath140 .
the uncertainties in @xmath42 , and consequently @xmath119 , can in some situations translate into a systematic uncertainty in polarimetry .
one of the best techniques for determining the absolute polarization of @xmath0he is the method of measuring shifts in the electron paramagnetic resonance frequencies of the alkali - metal atoms due to the effective magnetic field caused by the polarized @xmath0he @xcite .
this can only be performed in the pumping chamber ( where significant alkali - metal vapor is present ) , despite the fact that the quantity of interest in an electron scattering experiment is the polarization in the target chamber .
thus , it is often the case that nmr measurements are made directly on the target chamber , but are calibrated against frequency shift measurements in the pumping chamber .
when this is done , the calibration requires a knowledge of @xmath119 . for the case of the targets used in the @xmath2 experiment discussed earlier ,
the uncertainty in @xmath119 , following the prescription described earlier , translated into a 12% ( relative ) uncertainty in polarimetery . while not catastrophic , such systematic uncertainties would be nice to avoid .
two points emerge regarding polarization gradients : 1 ) they limit target polarization at high beam currents , and 2 ) they are somewhat uncertain in their size , which in some circumstances , results in systematic uncertainties in polarimetry . if we want to make polarization gradients smaller while retaining the basic cell design illustrated in fig .
[ fig : genstyle ] , there are two things that can be considered : increasing the cross sectional area of the transfer tube and decreasing the length of the transfer tube .
we note that with the large - volume cells that are more resistant to beam current , design constraints make it difficult to significantly shorten the transfer tube , and they are currently at most 50% longer than the shortest lengths used in early two - chambered targets such as those employed during e142 .
we will thus focus here on the cross sectional area . to better understand the limitations of single transfer - tube configurations for future targets , we consider a hypothetical target design for an approved experiment at jlab ( e12 - 06 - 016 ) that will measure @xmath2 up to @xmath141 @xcite .
the experiment will run at @xmath142 with a target length of 60 cm instead of 40 cm , a luminosity equivalent to running @xmath143 into the target of fig .
[ fig : genstyle ] . here
we will assume a single transfer tube is employed instead of the convection - based design that is actually specified in the proposal .
the single best way to make a target less susceptible to beam current is to make the cell larger and use more lasers , thus making relaxation from the beam a smaller perturbation to the target as a whole . the proposal for e12 - 06 - 016
calls for a target containing roughly 7 stp liters of gas , in contrast to the roughly 3 stp liters contained in the targets illustrated by fig .
[ fig : genstyle ] . using fig .
[ fig : genstyle ] as a starting point , we will assume a transfer tube diameter equal to the diameter of the target chamber ( about as large as is practically realizable ) .
we further assume the transfer tube has the same length as before , and that the pumping chamber is larger ( around 12.5 cm ) so as to satisfy the criterion of having the target contain 7 stp liters of gas .
such a cell would have a polarization gradient of around 12% in @xmath142 of beam .
if the cell - averaged polarization were around 70% with no beam ( more on this in section iv ) , and 62% in beam , the polarization in the target chamber would be about 55% .
in addition to the reduction of the in - beam polarization , the gradients would also have the potential of introducing uncertainties in polarimetry at the level of 24% .
in short , this hypothetical design falls well short of optimized performance . and
while we have assumed here the same transfer - tube length as is shown in fig .
1 , as will be discussed more in section iv , there are compelling reasons to increase the distance between the pumping and target chambers .
this would naturally require a longer transfer tube , something that would further aggravate the problem of polarization gradients .
we describe next a variant of the target cell geometry depicted in figure 1 .
there are still two chambers , a pumping chamber and a target chamber , but the two chambers are connected by two transfer tubes instead of one . with this design ,
it is possible to induce convection , thus causing rapid transfer of gas between the two chambers .
furthermore , all that is required to induce convection is to maintain a temperature differential between the vertical segments of the two transfer tubes . by controlling the temperature differential
, the speed of the convection can be adjusted . with rapid mixing of gas between the two chambers ,
the aforementioned polarization gradients can be made negligible , even if the distance between the pumping and target chambers is substantially increased . to demonstrate the feasibility of convection - driven polarized @xmath0he target cells , we have constructed a prototype with the geometry and dimensions illustrated in fig .
[ fig : convectionstyle ] .
the cell was constructed entirely out of aluminosilicate glass ( ge 180 ) , and was sealed after being filled with 7.2 amg of @xmath0he and a 0.11 amg of n@xmath144 .
the pumping chamber also contained several tens of milligrams of a hybrid mixture of potassium and rubidium .
the @xmath0he gas in the new prototype was polarized in the same manner as in our other target cells .
the pumping chamber was surrounded by a forced - hot - air oven , constructed largely out of ceramic and glass . while polarizing the @xmath0he , the oven was maintained at temperatures that were typically between @xmath145c , resulting in a vapor pressure of alkali - metal atoms corresponding to a number density on the order of @xmath146 .
the rubidium atoms were optically pumped using laser light from high - power diode - laser arrays with a wavelength of 795 nm , and as described in refs .
@xcite and @xcite , quickly shared their polarization with the potassium atoms that were also present .
subsequent spin - exchange collisions with the @xmath0he atoms resulted in the buildup of substantial nuclear polarization .
the temperature differential used to induce convection was maintained by using a second forced - hot - air convection heater " , installed on one of the transfer tubes as illustrated on fig .
[ fig : convectionstyle ] ( we note that , in reality , the convection heater also covered much of the horizontal portion of the transfer tube , but that this portion contributed negligibly to the speed of the convection ) . with a portion of one of the transfer tubes at an elevated temperature , the gas contained therein had a lower density than the corresponding gas in the other transfer tube , and thus experienced a small buoyant force which induced convection . by controlling the temperature of the convection heater , the gas flow could be controlled in a stable and reproducible fashion .
the flow of the gas was monitored using an nmr tagging technique . a slug " of gas within a small section of one of the transfer tubes
was depolarized by subjecting it to a pulse of rf tuned to the larmor frequency of the @xmath0he nuclei .
the rf was delivered using a small coil ( labeled in fig .
[ fig : convectionstyle ] as the zapper coil " ) wrapped directly around one of the transfer tubes .
nmr signals were then detected at each of four locations along the target chamber as indicated on fig .
[ fig : convectionstyle ] using small pickup coils " . the movement of the slug of gas
could then be tracked by monitoring nmr signals from each of the four pickup coils .
signals from the pickup coils were obtained once every two seconds using the nmr technique of adiabatic fast passage ( afp)@xcite .
representative examples of such signals are plotted in fig .
[ fig : zapper ] as a function of time .
time zero in this plot corresponds to the moment when a slug of gas was tagged .
-.3truein it is readily apparent from fig .
[ fig : zapper ] that a transient dip occurs in each of the signals from the four pickup coils .
this dip corresponds to the passage of the depolarized slug of gas , and it can be seen that the transient occurs at successively later times for each of coils 14 . given the known positions of the regularly spaced pickup coils , the difference in time between the transients associated with each of the four pickup coils provides a measure of the speed with which the tagged slug of gas was moving .
the measurement illustrated in fig .
[ fig : zapper ] corresponds to a target - chamber gas velocity of @xmath147 . also apparent from fig .
[ fig : zapper ] is the fact that each successive dip becomes wider and more shallow .
this is due in part to the fact that the gas flow within the cell is characterized by the classic parabolic hagen - pouiselle velocity distribution .
that is , the velocity as a function of the distance @xmath148 from the middle of the tube has the functional form @xmath149 .
hence , the slug of gas , which is initially fairly localized to the region around the zapper coil " , becomes increasingly spread out as it moves through the target chamber .
additional spreading also occurs because of diffusion .
we note also that the nmr signal decreases as a function of time .
this is due largely to polarization losses that occur with each afp measurement .
while the loss from each individual measurement is quite small ( on the order of 1% ) , the accumulation of many such losses is quite substantial .
we note that in the normal operation of a polarized @xmath0he target measurements are typically made not every two seconds , but rather every two to four hours .
hagen - pouiselle flow occurs whenever a pressure differential between two ends of a pipe causes the laminar flow of a viscous gas or fluid @xcite . in equilibrium
, the driving force from the pressure differential @xmath150 must be equal to the retarding force @xmath151 from the viscosity : @xmath152 for the case of a pipe that is circular in cross section , eqn .
[ eqn : f ] must be satisfied for each annular ring of fluid of thickness @xmath153 , a condition which leads to the equation @xmath154 where @xmath155 is the pressure differential , @xmath156 is the viscosity of the fluid and @xmath157 is the length of the pipe . imposing the boundary condition that the velocity of flow must go to zero at the perimeter of the pipe , the solution to this differential equation is @xmath158 where @xmath134 is the radius of the pipe .
it is the velocity distribution given by eqn .
[ eqn : vr ] that is often referred to as hagen - pouiselle flow . in the case of our convection cell
the driving force is due to the small buoyancy force that results from maintaining a vertical portion of one transfer tube at a higher temperature than the corresponding section of the other transfer tube : @xmath159 where @xmath160 is the volume of the vertical portion of the transfer tube that is being heated , @xmath161 is the difference between the average densities of the heated and unheated portions of the transfer tubes , and @xmath162 is the acceleration due to gravity
. we can express @xmath161 as follows : @xmath163 where @xmath164 is the density of the gas in those portions of the cell that are not heated , @xmath165 is the temperature of those portions of the cell that are not heated , and @xmath166 is the temperature of the portion of the transfer tube that is being heated ( both temperatures are in kelvin ) .
for the case being discussed here , the pressure differential that appears in the left - hand side of eqn .
[ eqn : f2 ] is thus given by @xmath167 where @xmath168 is the cross sectional area of the transfer tube , and @xmath169 is the length of the portion of the transfer tube that is heated .
if our convection cell could be treated as a long straight tube , we could simply substitute eqns .
[ eq : del_rho ] and [ eq : del_p ] into eqn .
[ eqn : vr ] to obtain an expression for the velocity .
our convection cells are more complex , however which complicates the expression that appears on the right - hand side of eqn .
[ eqn : f2 ] .
there are multiple sections of tubing , each with its own radius , as well as bends etc .
the velocity of the gas will be different in each section , and even the viscosity will be different depending on the temperature .
luckily , however , as will be shown in the appendix , the continuity equation ensures that the velocity in each section is related in a simple linear fashion to the velocity in the other sections , so it is still possible to solve eqn .
[ eqn : f2 ] exactly .
in essence , the quantity @xmath170 that appears on the right - hand side of eqn .
[ eqn : f2 ] must be replaced by a single quantity @xmath171 , which still has the dimensions of viscosity times length , but incorporates the full complexity of the cell .
the solution can accordingly be written in the form @xmath172 where @xmath148 is the radial coordinate in the target chamber , and @xmath134 is the radius of the target chamber .
the temperature dependence of @xmath173 is largely dominated by the factor @xmath174 .
the quantity @xmath171 , however , is also dependent on temperature , although for the range of values of @xmath166 that we consider , the temperature dependence of @xmath171 on @xmath166 is relatively weak .
it is important to understand the relationship between the observed velocity of the gas , @xmath175 , as indicated by data of the sort shown in fig .
[ fig : zapper ] , and the velocity distribution given by eqn .
[ eqn : vel_convec ] .
the natural way to compute @xmath175 is by taking the physical separation of adjacent pickup coils , and dividing by the separation in time between the minima of the corresponding transients .
if we consider the limit in which the distance between the zapper coil and the pickup coils is long compared to the length of the zapper coil itself , it is straightforward to show that , to a good approximation , the above method of computing @xmath175 corresponds to the maximum value of the velocity given by eqn . [
eqn : vel_convec ] , @xmath176 , which results from setting @xmath177 . to a good approximation
, we can express @xmath176 in the simplified form @xmath178 where all quantities not dependent on @xmath166 have been absorbed into the constant @xmath179 , and the temperature dependence of @xmath171 is accounted for by the factor @xmath180 , where @xmath181 . for the conditions we have considered , @xmath182 is on the order of @xmath183 c. in fig .
[ fig : convectiondata ] we have plotted @xmath175 as a function of @xmath166 for a range of temperatures . for each point , the velocity was computed using data such as those shown in fig .
[ fig : zapper ] .
we also show in fig .
[ fig : convectiondata ] with a solid black line a fit of the data to a function of the form of eqn .
[ eq : temp_fit ] .
the quality of the fit is clearly quite good , and yields the values @xmath184k@xmath185cm / min , @xmath186 , and @xmath187 c. for comparison , using our best knowledge of the cell geometry and densities , with @xmath188c , we compute @xmath189k@xmath185cm / min and @xmath190 c ( see appendix a ) . given the uncertainties of some of the quantities with which we are working , particularly in describing the retarding forces associated with our relatively complicated cell geometry , this agreement is quite reasonable .
more importantly , the agreement is more than sufficient to suggest that we have an acceptable quantitative understanding of the parameters influencing the convective flow from a practical perspective .
-0.3truein ultimately , the value of a convection - driven target cell is measured by the degree to which polarization gradients can be avoided between the pumping chamber and the target chamber .
it is critical that , as gas is depolarized by an electron beam , freshly polarized gas is delivered from the pumping chamber sufficiently quickly . ideally , one would like the ratio of the polarizations of the two chambers to be as close to unity as possible . fortunately ,
when convection is used to transfer gas between chambers , the polarization gradient is suppressed because the transfer rates are quite high . for the range of velocities measured in fig .
[ fig : convectiondata ] , we find values for @xmath43 in the range of 4.9 81 hrs@xmath90 .
even with relatively fast relaxation in the target chamber ( for example , consider @xmath191hrs@xmath90 ) , the polarization gradient ( eqn .
[ eq : delta ] ) will be very small in the presence of such fast transfer rates ( @xmath192 , or @xmath193 ) . in fig .
[ fig : brady ] , the @xmath0he polarization in both chambers of a traditional target cell is plotted as a function of time .
the ratio of target chamber to pumping chamber polarization for this plot begins at zero , and gradually climbs to a value significantly less than unity , reflecting a substantial polarization gradient . in fig .
[ fig : convectionratio ] , we plot this ratio for a convection - driven cell for three different operating conditions . in all cases ,
the temperature of the oven was held at @xmath194 .
the three curves correspond to different temperatures @xmath166 of the heated transfer tube . for the data shown with the open squares , the transfer - tube set temperature was @xmath195 , the same as the other ( unheated ) transfer tube .
this case corresponds to no driven convection , and is associated with a polarization gradient of 8% . for the data shown with the filled triangles and the filled circles , the set temperatures were @xmath196 and @xmath197 respectively .
these two conditions corresponded to target - chamber gas velocities of approximately 19.9 cm / min and 48.5 cm / min . in both of these cases , the ratio of the polarizations of the target chamber and the pumping chamber quickly reached a value approaching unity .
it is notable that there is very little difference between these last two curves despite substantially different gas velocities . in short , as soon as convection rather than diffusion is responsible for the gas transfer between the two chambers , polarization is relatively uniform throughout the cell .
with advances in seop , it is now possible to run polarized @xmath0he targets at significantly higher luminosities .
using a single transfer tube , however , even currently approved experiments would suffer polarization gradients of 12% or worse , and subsequent experiments at even higher luminosities would be even more limited .
the work described here demonstrates that with only minor changes to current designs , convection - based cells can be built in which polarization gradients are extremely small . in many ways ,
however , the most compelling reason to adopt convection - based cells is the enormous flexibility that is gained in being able to choose the distance between the pumping and target chambers .
experience has shown that the probability of a cell rupturing goes up markedly after 46 weeks exposure to something on the order of @xmath198 of beam .
most commonly , the part of the cell that breaks is the pumping chamber .
stress in a sphere goes up linearly with the radius .
this makes the relatively large pumping chamber particularly vulnerable to radiation damage . with convection - based cells the distance between the pumping chamber and the target chamber
can be greatly increased with minimal implications for performance . among other things ,
it becomes practical to incorporate radiation shielding for the pumping chamber .
there are at least two additional reasons that flexibility in the configuration of target cells is highly desirable .
as discussed earlier , the primary way to make targets more tolerant of beam is to make them larger .
the more gas that is being polarized per unit time , the less relevant it is that some small portion of that gas is being depolarized by the beam .
it is not practical , however , to make the pumping chamber arbitrarily large . as the stress in the glass walls increases , the cell becomes more prone to rupturing , even if the wall thickness is increased .
furthermore , the intensity of the optical pumping radiation goes like total power over the square of the radius while the volume goes like the cube of the radius .
hence , since the required laser power scales with volume , the required intensity of light scales linearly with radius .
there are already indications that we are near a threshold at which laser intensity damages the cells from the perspective of spin relaxation .
if multiple pumping chambers were used instead of one , it would become possible to limit both laser intensity and the size of each individual pumping chamber .
if diffusion were the only mechanism for moving polarized gas inside the target , the implementation of multiple pumping chambers would be quite challenging .
we close by citing a concrete example that underscores the desirability of adopting convection - based target cells .
we show in fig .
[ fig : bradyspinup ] results from a bench test of the cell brady " ( discussed earlier ) in which the primary goal was to achieve the highest polarization possible .
brady , like the the targets used for the @xmath2 measurements @xcite , contained an alkali - hybrid mixture of k and rb .
unlike during the @xmath2 experiment , however , the results shown in fig . [ fig : bradyspinup ] were obtained using diode - laser arrays that were spectrally - narrowed such that their line widths were about ten times narrower than their broadband counterparts .
the polarization in this test saturated just over 70% .
brady was one of two target cells used to collect data on single - spin asymmetries in semi - inclusive deep inelastic scattering , and was similar in design , although slightly smaller , than the configuration shown in fig .
[ fig : genstyle ] . during the experiment ,
the in - beam polarization of the @xmath0he averaged around 55% @xcite , despite the fact that the cell - averaged polarization , averaged over the entire experiment , was over 60% .
if the targets used for this experiment had had convection - based gas mixing ( which was just being explored at the time ) , significant gains in performance would have resulted .
-0.2truein in conclusion , with the advances in target performance brought about by a combination of alkali - hybrid spin - exchange optical pumping and spectrally - narrowed diode - laser arrays , the point has been reached when it is desirable to explore target - cell geometries that go beyond that shown in fig . 1 .
we suggest that the convection - based gas mixing demonstrated in this work can play an important role in implementing high - luminosity polarized @xmath0he targets in the future . not only will the use of convection reduce polarization gradients , it will also make possible greater flexibility in the placement of the pumping chamber with respect to the target chamber . among other things ,
this flexibility will enable the use of radiation shielding , and even the use of multiple pumping chambers .
convection - based polarized @xmath0he target cells open new possibilities that collectively address several of the important challenges facing the next generation of high - luminosity polarized @xmath0he targets .
this work was supported by the u.s .
department of energy under contract numbers de - fg02 - 01er41168 and de - ac05 - 060r23177 .
jefferson science associates , llc , operates jefferson lab for the u.s .
doe under u.s .
doe contract no .
de - ac05 - 060r23177 .
we would also like to thank michael souza of princeton university for his patience and expert glass blowing .
here we formulate estimates for the parameters that appear in eqn .
[ eq : temp_fit ] .
since our temperatures are in the classical regime ( @xmath199 3 k ) , the viscosity @xmath200 of @xmath0he can be calculated from the viscosity @xmath201 of @xmath97he using @xcite : @xmath202 we parameterize the @xmath97he viscosity in the range of @xmath203 c following kestin et al .
@xcite , @xmath204 where @xmath205 has units of @xmath206 c and @xmath207 at @xmath208 , @xmath209pa@xmath185s .
the flow in a pipe is laminar if the reynold s number is below 2300 @xcite .
the reynolds number is defined as @xmath210 where @xmath211 is the density of the fluid .
a pipe @xmath212 cm wide that is filled with 8 amagats @xmath0he will have laminar flow at @xmath213 c ( @xmath214 ) provided @xmath215 cm / min .
the flow in the convection cell arises from a forced density difference between the two transfer tubes
one tube is maintained at room temperature while the other is heated ( see fig . [
fig : convectionstyle ] ) .
we modeled the convection cell as five contiguous pipes as is illustrated in fig .
[ fig : cylinders ] .
[ eqn : f2 ] becomes @xmath216 where @xmath169 is the vertical length of transfer tube that is held at an elevated temperature . in this model
we approximate the pumping chamber as a cylinder with transfer tubes entering axially , and identify five distinct regions in the cell as is indicated in fig .
[ fig : cylinders ] .
each region is identified as a pipe of length @xmath217 , radius @xmath218 , cross sectional area @xmath219 and temperature @xmath220 .
we further assume that @xmath221 , @xmath222 and @xmath223 .
finally , we identify @xmath224 and @xmath225 .
we note that both the density and viscosity of the gas are temperature dependent .
( [ eqn : pouisellelike1 ] ) assumes that there are no minor losses " in the system , where a minor loss represents a pressure drop due to a sudden change in flow from a pipe fitting or a pipe expansion or contraction ; the term minor loss " refers to a loss that is small relative to the overall length of pipe under consideration @xcite . in our case , due to the relatively short length of the cell , the minor losses are actually significant . the actual cell , as illustrated in fig . [
fig : convectionstyle ] , consists of four elbow bends , two tees , and four expansions / contractions .
the retarding forces these losses exert on the gas can be approximated by considering instead an equivalent length of straight pipe .
the expansions / contractions have a relatively small loss , which is equivalent in magnitude to the loss that would be incurred passing through a pipe of length @xcite @xmath234 where r is the tube radius , @xmath235 is the friction factor ( @xmath236 for laminar flow ) and @xmath237 is the fluid - independant resistance coefficient . for sudden expansions / contractions ,
@xmath238 gas flowing through the transfer tube / target chamber junction has @xmath239 , @xmath240 . at @xmath241cm / min ( @xmath242 ) , this gives a negligible @xmath243 cm
gas flowing between the pumping chamber and the transfer tube will have an even smaller @xmath244 .
the losses in the bends , however , are much greater .
we model the loss coefficient in the bends using the 3-k method of darby @xcite , @xmath245 where @xmath246 is the diameter of the pipe ( in inches ) and @xmath247 are geometry - dependent loss coefficients .
we approximate our glass bends as flanged , welded bends with @xmath248 ( here , @xmath249 is the radius of the bend ) ; such bends have @xmath250 . for laminar flow ,
the 3-k method gives @xmath251\ \ .\ ] ] we treat the transfer tube / target chamber tee junctions as elbows ( effectively ignoring the dead - end branch of the tee ) .
the system therefore has five bends in temperature - region 1 ( which have a total equivalent length of approximately @xmath252 ) and one bend in temperature - region 2 ( which has an equivalent length of approximately @xmath253 ) . finally , using eqn .
[ eq : del_rho ] for @xmath254 we find @xmath255 where @xmath256\quad \\
l_2 ' & = & l_2 + \frac{2r_2}{64}\left[k_1 + re_2k_\mathrm{i}\left(1+\frac{k_d}{d_2^{0.3}}\right)\right]\ \ .\end{aligned}\ ] ] note that the reynold s number is dependent on the velocity of the gas .
table [ tab : bestguesses ] lists values for r and @xmath157 .
measurements of @xmath134 ( which is the inner diameter of the tube ) require a knowledge of the thickness of the glass tube .
we measured the thickness of the glass by observing interference patterns using a scannable single - frequency laser . using this information , eqn .
[ eqn : pouisellelike ] predicts velocities that agree within 20% of the measured value ( see fig .
[ fig : convectiondata ] ) . where @xmath259 \ \ , \\ \beta_2 & = & 0.8681 l_2'\frac{r_1 ^ 2}{r_2 ^ 2}\left[\frac{b + 2c(t_1 - 273)}{t_1 } + c\right],\\ \mathrm{and } \\
\beta_3 & = & 0.8681 l_2'\frac{r_1 ^ 2}{r_2 ^ 2}\left[\frac{c}{t_1}\right ] \ \ .\end{aligned}\ ] ] in the above equations , all temperatures are in kelvin . to describe the velocity in region 4 ( the target chamber ) instead of in region 1
, we use eqn .
a9 and replace @xmath260 with @xmath261 .
this is the quantity that appears in eqn .
[ eq : temp_fit ] that characterizes the magnitude of the gas velocities shown in fig .
[ fig : convectiondata ] . evaluating the above equations in terms of our best guesses for cell dimensions and temperatures , with @xmath262c gives @xmath189k@xmath185cm / min , @xmath263 , @xmath264 , and @xmath265 .
it is clear from these values that it is quite reasonable to neglect the terms involving @xmath266 and @xmath267 , which results in eqn .
[ eq : temp_fit ] from section iii - b . | the dynamics of the movement of gas is discussed for two - chambered polarized @xmath0he target cells of the sort that have been used successfully for many electron scattering experiments .
a detailed analysis is presented showing that diffusion is a limiting factor in target performance , particularly as these targets are run at increasingly high luminosities .
measurements are presented on a new prototype polarized @xmath0he target cell in which the movement of gas is due largely to convection instead of diffusion .
nmr tagging techniques have been used to visualize the gas flow , showing velocities along a cylindrically - shaped target of between @xmath1 .
the new target design addresses one of the principle obstacles to running polarized @xmath0he targets at substantially higher luminosities while simultaneously providing new flexibility in target geometry . |
during the last decade , gamma - ray burst classification has emerged as a promising tool to understand these events . in the pre - batse data ,
there were indications of a bimodality in the temporal durations of non - repeating bursts @xcite , as well as a modest correlation of duration with spectral hardness @xcite .
these results were confirmed by data from the burst and transient source experiment ( batse ) which show that short bursts inhabit a distinct region of the duration / spectral hardness parameter space @xcite .
more recently , theoretical work has made use of the batse archive to attempt to determine the processes which distinguish short and long bursts .
there is some evidence that short grbs arise from internal shock processes @xcite , and that they are differentiated from long bursts by the ejecta shell geometry ( eg .
@xcite ) . alternatively
, these two classes may arise from different accretion disk states around progenitor black holes @xcite .
indications that very short bursts exhibit an unusually uniform spectral hardness distribution may hint at yet a third class of grb @xcite . despite what has been gleaned from the gamma - ray emission ,
the study of short bursts has not been benefited by the more recent advances brought about by optical afterglow ( eg .
@xcite , @xcite ) and burst @xcite observations , all of which have identified only long - duration bursts .
one of the main reasons for this bias stem from difficulty in obtaining well - localized final positions for these events . to date , the main source of well - localized burst notices for late optical study has been the bepposax satellite @xcite , but the 1 s integration time of the grbm prevents efficient detection of short bursts .
in addition , prompt wide - field searches such as rotse - i and lotis @xcite have been hampered by the practical difficulty of searching their ~250 sq .
degree fields . at a sensitivity of 15th magnitude
, there are of order @xmath1 stars in these fields , and approximately a few million photometric measurements in a typical rotse - i trigger response .
burst finding necessitates well - controlled photometry across an entire field , as well as a detailed knowledge of the hardware behavior ( eg .
bad pixels , position resolution ) .
although this limitation was overcome for long bursts by using localizations from the inter - planetary network ( ipn , @xcite ) to reduce the search area @xcite , no such localizations were available for the short bursts on which rotse - i triggered . to overcome this limitation , we have designed an analysis specifically to confidently analyze many observations of a large field .
this paper presents the results of a search for optical counterparts to five bursts in the rotse - i trigger data sample using these techniques .
an additional burst , studied with our previous method , is also presented .
the rotse - i array used in these observations consists of four telephoto lenses comounted in a @xmath2 geometry on a lightweight platform which allows rapid slews and excellent pointing accuracy relative to the @xmath3 total field - of - view ( 14.4 pixel scale , see @xcite for more details ) . from march , 1998 until the deorbiting of the compton gamma - ray observatory ( cgro ) in june 2000
, 57 grb triggers were received via the grb coordinates network ( gcn , @xcite , @xcite ) derived from batse data @xcite . to date
, we have successfully analyzed data for a subset of seven long duration bursts possessing small localization errors , yielding one detected prompt optical counterpart @xcite and six non - detections @xcite .
this paper discusses a search in a set of six additional triggers taken during the full period of batse activity while rotse - i was automated .
five of these triggers were selected for this analysis because ( 1 ) they fall into the short duration class of bursts , ( 2 ) their most probable known position is within @xmath4 of the rotse - i pointing , or ( 3 ) we have a thorough , photometric - quality set of images from most of the cameras .
the sixth burst , grb 000201 , is included because the analysis of the known ipn diamond localizing this burst was straightforward , and other optical follow - up has been performed @xcite .
preliminary results for grb 980527 and grb 000201 have been previously presented in @xcite , respectively .
this paper supersedes and improves upon those results .
some important burst parameters from the batse observations , as well as the rotse - i spatial and temporal coverages are itemized in table 1 .
the spatial coverage is defined as the probability that the burst location was imaged in our search , and it is based on the statistical and systematic uncertainties in the best batse or ipn localizations .
the observations scheduled in response to a trigger vary somewhat depending on trigger type and delay from the time of the burst .
the three trigger types involved in this analysis are ` original ' which arrives around 7 seconds after the burst start , ` final ' which arrives approximately 1 minute later , and ` maxbc ' which arrives about 5 to 10 minutes after the burst . for all of the trigger responses taken with rotse - i ,
5 s exposures are taken during the first minute of the burst .
prior to december 1998 , the exposures were lengthened to 25 s and then 125 s as the delay progressed from 1 minute to approximately 10 minutes .
after january 1999 , these longer exposure lengths were reduced to 20 s and 80 s , respectively .
there are occasional gaps in temporal coverage for the less accurate ` original ' and ` maxbc ' triggers because we hedged our bets and scheduled tiling sequences during the burst response .
grb 980527 was one of the first well - localized bursts observed by rotse - i .
this is a very intense , spectrally hard burst lasting approximately 0.1 s. the first images were taken within 12 s of the burst start and consist of two groups of five exposures of 5 s and 25 s length with a tiling gap in between .
after this , five 25 s exposures were taken centered on the subsequent maxbc localization . unfortunately , the maxbc localization is more distant from the most likely position for this burst , reducing our coverage at later times .
in addition , camera ` a ' was not functioning which reduces our coverage to 86% ( stat.+sys . ) .
the only trigger localizations provided for both grb 990323 and grb 990808 were of the maxbc variety , and so our imaging begins more than 10 minutes after these bursts . both grb 991028 and grb 991228 possess ` original ' and ` final ' localizations .
grb 991028 is a spectrally hard short burst and these on - line positions are a good match to that determined off - line by the batse team .
hardware behavior suffered in two ways for the grb 991228 trigger response : camera ` c ' did not take useable data , and camera ` d ' exhibited a mild charge - transfer problem . for grb 000201 ,
we only triggered on the ` final ' localization since the ` original ' position was deemed unobservable by the automated scheduler .
the best localization for this burst consists of an ipn diamond derived from batse , ulysses and near data , and is well - covered by the rotse - i images . observing conditions for all of these bursts
were good .
the reduction of the data proceeds as in @xcite .
raw images are dark subtracted and flat - fielded by using darks and sky - flats generated on the night of the trigger .
clustering of the corrected image is performed with sextractor @xcite utilizing a background mesh segmented by 32 pixel increments .
raw magnitudes are based on @xmath5 pixel apertures .
astrometric calibration and determination of the overall zero - point for the image are performed by comparison to the hipparchos catalog @xcite . during this step , a bad pixel template , derived from the raw darks for each camera ,
is used to flag objects containing such pixels within their apertures .
a systematic error is assigned to the source based on the observed fluctuations of the constituent bad pixels .
individual calibrated source lists are matched up to one another to form preliminary lightcurves .
any observation of an object is tagged as bad in which the position lies over one half pixel from the source s mean location .
final photometric calibration across the image is then performed by determining the median magnitude for a set of template sources for each 100 pixel @xmath6 100 pixel image subregion , and using these to derive a map of offsets and offset variances in each subregion .
all objects are then corrected by a bilinear interpolation of this relative photometry map , and a systematic error is assigned based on the offset variance .
the analysis of the burst fields proceeded on two paths : a small field search applied to grb 000201 , and a wide field search applied to the other five . in both , candidate objects are rejected if they do not occur in at least one consecutive pair of observations .
in addition , an object s individual observations are ignored if they are saturated or at an image edge . from here , the two analyses differ .
because the ipn localization for grb 000201 dramatically reduced the background for a counterpart search , we employed a loose lightcurve selection . during the observations in which a candidate is detected
, there must be at least one pair of good measurements which exhibits a variation in excess of @xmath7 , where @xmath8 is the sum in quadrature of the statistical and systematic errors of the two observations involved .
because it was the best localization early , we searched within the 1 @xmath8 ( stat . ) limits of the later batse locburst position .
we have also checked the location of a purported optical counterpart detection @xcite .
no counterparts were identified .
an initial ipn localization became available on feb .
9 , and we searched within a surrounding box having sides with @xmath9 and @xmath10 .
the final ipn diamond which was obtained later comprises the inner 20% of this region .
we found no counterparts in this region .
this ipn - based analysis gives insufficient control of backgrounds in wide - field searches . to address this
, we first attempted an image differencing approach on grb 980527 @xcite .
this method s effectiveness was reduced , however , because blending is not a critical problem in rotse images away from the galactic plane , and the typical psf is very undersampled . in addition
, it removes the diagnostic information present in the original image with which we can either correct occasional mild photometric variations , or flag problem observations . as a result
, we perform our wide - field search by extending our ipn - based analysis with a more restrictive lightcurve variation selection , as well as a more stringent requirement on local observation quality .
individual observations are rejected if : a bad pixel lies within the aperture , there is a large offset from the mean position ( ie.@xmath11 pixel ) , photometry in the local image subregion has a large ( @xmath12 mag ) standard deviation , or there are @xmath13 template stars per local subregion .
lightcurve cuts were chosen based on a comparison of simulated bursts with power - law lightcurves vs. typical backgrounds observed in rotse trigger data .
we require at least one variation passing a logical and of @xmath11 mag . and @xmath14 ( stat.+sys . ) .
in addition , we require that the overall lightcurve ( using good observations only ) differs from constant by a @xmath15 per degree - of - freedom , where the largest variation is excluded from the calculation .
this last criterion also implies that the source is detected in 3 good , not necessarily consecutive , observations .
no candidates were observed for grb 990323 , grb 990808 or grb 991028 after these selections .
a few objects passed the selection for grb 980527 , but most of these correspond to known sources in the usno a-2.0 catalog @xcite .
examination of the images for the remaining candidates revealed them to be incompletely removed bad pixels . for grb 991228 ,
the source of the candidates was solely due to a charge - transfer problem giving rise to spurious transients in the trailing charge .
the limits for these bursts are shown in figure 1 , and are itemized in table 2 for up to three discrete epochs .
all three short bursts have very prompt trigger responses , and two , grb 980527 and grb 991028 , provide a probable coverage of the allowed error region .
the earliest limits from these two bursts are 13.1 and 13.8 mag at 9.9 s and 12.2 s , respectively .
later limits range from magnitude 14 to 15 .
our non - detections for these bursts , which have very different durations , suggest that short bursts do not usually give rise to optical emission brighter than 13th or 14th magnitude in the first few minutes after the burst .
unfortunately , we do not have data for the first 10 seconds of the burst which may be a critical phase for short duration events . among the three long duration bursts analyzed
, two have a high coverage probability , and one ( grb 000201 ) has images starting soon after the burst .
the best limits for this sample are 14.0 mag at 95.9 s and 15.8 mag late .
the non - detection of a counterpart in these cases reinforces our previous conclusion that early optical emission from grbs is not typically brighter than 14th magnitude , and it is fainter than 16th magnitude around 1 hour after the burst . in an effort to compare this data to the only burst in which prompt
optical emission has been observed , we have replotted the limits in figure 2 after adjusting for their peak flux in 64 ms binning of the 50 - 300 kev batse data , as compared to grb 990123 .
an optical burst from grb 980527 , grb 981028 and grb 000201 would have been observable if optical flux and gamma - ray flux were highly correlated .
more prompt optical data are clearly needed .
this analysis , which is the first attempt in this direction , signals a hopeful future for more sensitive work .
the methods described here are directly applicable to the rotse - iii imaging which will be the mainstay of this effort in the future .
it should now be possible to perform ipn - based searches to much deeper magnitudes than with rotse - i . more importantly for early optical study
, we are now implementing an improved pipeline to find an optical transient in a fully automated and rapid way based on hete-2 triggers .
we thank the batse team for their grb data .
we also thank the near and ulysses teams for their grb 000201 data .
rotse is supported by nasa under @xmath16 grant nag5 - 5101 , the nsf under grants ast-9703282 and ast-9970818 , the research corporation , the university of michigan , and the planetary society .
work performed at lanl is supported by the doe under contract w-7405-eng-36 .
work performed at llnl is supported by the doe under contract w-7405-eng-48 .
kevin hurley is grateful for ulysses support under jpl contract 958056 , and for near support under nasa grant nag5 - 3500 and nag5 - 9503 .
hurley , k. 1992 . in aip conf .
265 , gamma - ray bursts , ed .
w. paciesas and g. fishman .
kouvelioutou , c. , et al .
1993 , apj , 413 , l101 .
dezalay , j. , et al .
1992 , in `` gamma - ray bursts '' , p304 - 309 .
nakar , e. and piran , t. 2001 . subm .
to mon . not .
r. astron .
soc . , astro - ph/0103192 .
kobayashi , s. 2000 , apj , 545 , 807 .
van putten , m. , and ostriker , e. 2001 , acc . by apjl , astro - ph/0010440 cline , d. , matthey , c. , and otwinowski , s. 1999 .
apj , 527 , 827 .
metzger et al .
kulkarni et al .
nature , 393:35 - 39 .
akerlof , c. , et al .
1999 , nature , 398 , 400 1997 . in proc .
3114 , _ euv , x - ray , and gamma - ray instrumentation for astronomy viii _ , 186 + .
park , h. , et al .
1999 , aaps , 138 , 577 .
hurley , k. , et al .
1999 , apjs , 122 , 497 .
akerlof , c. , et al .
2000 , apj , 532 , l25 . 1995 .
_ astrophysics and space science , _ * 231*:235238 .
1998 . http://gcn.gsfc.nasa.gov / gcn/. kehoe , r. et al .
2000 , in the stsci may symp . , the largest explosions since the big bang : supernovae and gamma ray bursts , ed .
m. livio , n. panagia , & k. sahu ( baltimore : stsci ) paciesas , w. , et al . 1999 ,
apjs , 122 , 465 .
boer , m. , atteia , j. , and klotz , a. , 2000 , gcn circ . 535 kehoe , r. wren , j. , and lee , b. , 2000 , gcn circ . 536 bertin , e. , & arnouts , s. 1996 , a&as , 117 , 393 hog , e. , et al , 1998 , a&a , 335 , l65 .
usno catalog , http://www.nofs.navy.mil/. .characteristics of six bursts responded to by rotse - i . corresponding information for grb990123 is given in @xcite ) .
the columns specify : grb date , batse trigger number , duration in seconds where endpoint fluxes are 50% of peak , duration where endpoint fluxes are 10% of peak , peak flux ( @xmath17 , in @xmath18 ) , fluence ( @xmath19 ) , coverage of the grb probability ( % ) , and start time ( @xmath20 , in sec . ) for first image recorded .
the peak flux is taken using 64 ms binning , except for grb 990323 and grb 990808 for which only 1024
ms binning was available .
coverages for epochs with pointings different than the first epoch are indicated in parentheses .
[ tab : sixgrbs ] [ cols="^,^,^,^,^,^,^,^",options="header " , ] | gamma - ray bursts of short duration may harbor vital clues to the range of phenomena producing bursts .
however , recent progress from the observation of optical counterparts has not benefitted the study of short bursts .
we have searched for early optical emission from six gamma - ray bursts using the rotse - i telephoto array .
three of these events were of short duration , including grb 980527 which is among the brightest short bursts yet observed .
the data consist of unfiltered ccd optical images taken in response to batse triggers delivered via the gcn .
for the first time , we have analyzed the entire @xmath0 field covered for five of these bursts .
in addition , we discuss a search for the optical counterpart to grb 000201 , a well - localized long burst .
single image sensitivities range from 13th to 14th magnitude around 10 s after the initial burst detection , and 14 - 15.8 one hour later .
no new optical counterparts were discovered in this analysis suggesting short burst optical and gamma - ray fluxes are uncorrelated . |
Subsets and Splits
Data Mining Abstracts
This query searches for and retrieves 2000 documents containing the phrase 'data mining' in the abstract, providing limited but specific document samples.